““ “H “““““““

“I“ “ I “I“ I

IIII.

II “““I'M “I““I'

IIIIIIIII“IIII“IIIIILIIIIIIII-IIIIIIIIIII.IIIIIIIIIIIIIII I.

“I.

I

I

IIIII,IIIII

r3:v“f7°;..,

IIII _,‘:i"-.

in“. II“I I “IIIIIII: I

l I

I I'II I

II“II““I V“ “““I“I' I

IIILI‘II“I

I‘I'III

I

l '“I“'“"““

.

II

.

“II“

IIIIII‘II

'I“III|IIIIIIIIIIIII'I IIII'II“II IIIIII I

IIIIII'II

“III“IIIIIII“IIIIIIIIIIIIIIII ”I”III“I’IIIIIIII‘“IIIIII“L’I“.‘II_ “:2“._.II‘I““L““I: "NflI

IIIIIIIIH“IIIIIIIIIII “III“IIIIIIIIIIIIIIHIIIIIIIIII“III“ITIIIII“II:~I“I“|I'IIIIIIIII_IIIII

;'

IIIIIIIIII IIIIIII

IIJ'II ‘

IIIIIIIII I

  

'.

I“

I'IIIIIIH“.I “III'I‘IIIIIIII'IIIII“IIIII.“““I'IIIII

III I

.III‘IINIII‘IIIxIIIIIIIIIi,

““III'I“IIII«:;&‘

'I

I

“' I

I " “' “I“II’I'III

"'“IIIIIIII '3“" “II’II‘II'III

II ”III “II.“ “I“.I ’I "I .-.;.

I “IIIIIIII

“III' I

III

     

IIIIIIII ”I “““I“I“““I “““III.

IIIIII

“II'. IIIII-I":

IIIIIII IIIIIII3’

  

_.

7

III II

“III lIII'II'IT‘I“'III ‘t'EI'IIII II .““.;‘ 1"“..“II?”

II...

  

III“ “II“IIII“,“““II‘II'IIII"I-‘“I‘IIII'III“III’7:le"I III I

“III III“ “II“I‘I'VIII»? ‘

'I'IIIII'I“““II|‘I““I'““““,“,.”I III

  

’~' ‘

  

III

(III

I

'

“

I

"
I

I“““III““I“I“““I““II“IIILI'“IIIIIIIIHIIwt ..IIIIIIIIIQ “I“I'qugfII“““’““““IIII-II“IU“III

I

I

III

I“

II'II“

III

"I

Ifr“

‘

II““'

I

I

“I“I“““

IIIIIIIII‘II III

IIIIIIII‘IIII

.IIIIII»

I5“

I31

III I

III“

I
“I I I -
:II

I“
III“ #“I .

III

“I

,

"I “III“. ’

III;I LIIéII

I “I“;I,

-:-

'I

I

‘vI" “.

-'
'-

'

L

“L

“ :

3,:"..“

uffr “NH“ 3
'

IT.

.‘I‘II ”IN
'“iII' I“~"

I.“

I

.

I.

I.,..
“.
III“

I

.

'

‘I

I=~
.

.

I

’

. "1“

“g ;JPI‘

“figs“

I,

if

‘1'

LINUIH:,‘I-

IvIIIvIII

: "I

I

-

“

I

t

.2“: u
“iv

.-I“'

31;”!

,

'.

'

,

QIIIIII

.‘.

“III““““““IIIII“?

"I

~I:IIII'I‘IIIIIIIII’III ..:I

.

.

‘

I

Ian III‘III

.IIII'IHIS” “If"

WWI-1.35:2“.

..-

‘

-'

-

I

1'“ ”“1II“IIII""“‘“III".

III"

-’—.L"I..-

.

d“IIIIIII“IIILII

IIHIIIII‘II

LII. I ,

I.

J

“I“,

“ '“I““

“11.1.“qu I I W I

“I“.

 

:'I

I

I“

"J“II
WI

*IiII‘II’IIII"II”II'“III I“IIIII
v.41 f‘IIt'II“

7-.“I 34‘‘¢“.‘I1H“II'I“'““'“ J! “I.“

“II.

‘I I --

“ I I I I I“ I I I I I I I I I II “ “ “ . I I “ I I I

NIBIIII' “IIII

Inmr..7“‘.LI“j~LI""I';

THESES

IImm;MIMI“Lm1lemum“(mayIflll

‘

 

I fl\\

OVERDUE FINES:

25¢ per day per item

"f {{f““}.\ .“
x . “3:; --,2'

(

1

x

"om,"

-

mamas umav mums:
Place in book return to remove
charge from circulation records

 

 

 

  

A SYSTEMS ANALYSIS OF RURAL DEVELOPMENT ORGANIZATION

THE COMILLA STRATEGY:

By

Blake W. H. Smith

A DISSERTATION

Submitted to

Michigan State University

in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Administration and Higher Education

1979

ABSTRACT

THE COMILLA STRATEGY:

A SYSTEMS ANALYSIS OF RURAL DEVELOPMENT ORGANIZATION

By

Blake W. H. Smith

The Pakistan Academy for Rural DeveloPment in Comilla, East

Pakistan

was a center of international attention in the nineteen

sixties.

Experts in agricultural and rural development flocked there

from around the world to find out how it was that the programs and

institutions sponsored by the academy had succeeded, in one of the

poorest regions on earth, in raising agricultural output, cutting

birth rates, educating women, and generating employment in the villages.

While the innovative tradition of the rural development academy at

Camilla has been carried on under the government of independent

Bangladesh, the golden era of the Comilla experiment was in the nine-

teen sixties.

The author was privileged to spend a year at the Academy

for Rural Deve10pment in Comilla in 1965-66, when he served there as an

Africa-Asia Public Service Fellow.

Expanding upon infbrmation garnered in Comilla in 1965-66 about

the economics of the rural credit cooPeratives there, this study

examines the rural development organizations in Comilla, and principally

the c00peratives, from the standpoint of modern organization theory.

Through a modification of an analytic paradigm proposed by Fremont Kast

and James Rosenzweig, the cooperatives are regarded as a composite

of several subsystems, each of which is examined in a separate chapter.

A theoretical justification for this analytic approach is developed in

the third chapter.

The utility of these system-analytic techniQues is demonstrated

by Showing how they lead to a fuller understanding of the rural

development organizations in Comilla than can be obtained from the

best alternative schema:

the model of effective organization for

rural development proposed by Norman Uphoff and Milton Esman in their

summary of a study of rural local government in Asia.

Comparisons are

also drawn with the paradigm of the Institution Building school, and

with the model of rural organization elaborated by the National

Front for the Liberation of South Vietnam.

©

Copyright by

Blake Wales Hendee Smith

1979

Dedicated to

William Wirt Lockwood

1906 - 1978

Professor at the

WOodrow Wilson School of

Public and International Affairs

Princeton University

He was determined that I should write something of value.

TABLE OF CONTENTS

LIST OF TABLES .

LIST OF FIGURES

INTRODUCTION .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

CHAPTER 1:

THE COMILLA EXPERIMENT .

.

.

.

.

Founding of the Comilla Academy .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

The Academy Director

.

.

.

.

.

.

.

.

The Rural Setting of the Academy

.

.

.

Initiatives in Rural Development Organization

CHAPTER 2:

THE NEED FOR MODELS

.

.

.

.

.

.

.

The World Bank and the Rural Poor .

.

The Deficiency of Present Theory

.

.

.

.

The Cornell Rural Development Study .

.

.

.

.

Critique of Uphoff and Esman's Approach .

.

.

.

CHAPTER 3:

A SYSTEMS APPROACH TO ORGANIZATION

THEORY

Overview

.

.

.

.

.

Basic Bertalanffy .

.

.

.

.

.

.

.

.

.

.

.

.

Boulding's Typology of Systems

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

BOUlding's Typology Revised .

. .‘. .

.

.

Static Stmctures D

O

I

O

C

O

O

O

O

O

I

0

Simple Machines .

.

.

.

.

.

.

.

.

.

.

.

.

Machines with Input

organic syStemS a

o

a

a

a

a

a

o

o

a

o

e

o

Systw ContrOl

.

.

.

C

C

.

O

.

O

.

.

C

.

ii

Page

vii

10

16

21

30

8
1
‘

33

86

86

88

91

96

98

99

108

118

135

CHAPTER 3:

(Continued)

A Complete Typology of Systems

.

Feedback and Operational Control

.

.

Elementary Subsystems .

.

.

.

.

.

.

CHAPTER 4:

THE STRUCTURAL SUBSYSTEM .

.

Introduction

.

.

.

.

.

.

.

.

.

.

Relevant Research .

.

.

.

.

.

.

.

.

.

The Structural Subsystem at the Village Tier

The Structural Subsystem at the Thana Tier

A Systems Interpretation

.

.

.

.

.

CHAPTER 5:

THE TECHNICAL SUBSYSTEM

.

Introduction

.

.

.

Relevant Research .

.

.

.

.

.

.

.

.

.

.

.

.

The Technical Subsystem at the Village Tier

The Technical Subsystem at the Thana Tier .

A Systems Interpretation

.

.

.

.

.

CHAPTER 6:

LINKAGE TO THE ENVIRONMENT .

Introduction

.

.

.

.

.

.

Relevant Research .

.

.

.

.

.

.

.

.

.

.

.

.

.

Linkages with the Environment at the Village Tier

Linkages with the Environment at the Thana Tier .

CHAPTER 7:

THE PSYCHOSOCIAL SUBSYSTEM .

Introduction

.

.

.

Relevant Research .

"

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

The Psychosocial Subsystem at the Village Tier

The Psychosocial Subsystem at the Thana Tier

137

141

1u3

167

168

172

185

188

202

235

235

239

2&6

254

276

347

BA?

349

355

378

394

395

400

#16

431

CHAPTER 3:

(Continued)

Page

.

.

.

A Complete Typology of Systems

.

Feedback and Operational Control

Elementary Subsystems .

.

.

.

.

.

CHAPTER N:

THE STRUCTURAL SUBSYSTEM .

Introduction

.

.

.

.

.

Relevant Research .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

The Structural Subsystem at the Village Tier

The Structural Subsystem at the Thana Tier

A Systems Interpretation

.

.

.

.

.

.

CHAPTER 5:

THE TECHNICAL SUBSYSTEM

Introduction

.

.

.

Relevant ResearCh .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

The Technical Subsystem at the Village Tier .

.

The Technical Subsystem at the Thana Tier .

.

A Systems Interpretation

.

.

.

.

.

CHAPTER 6:

LINKAGE TO THE ENVIRONMENT .

Introduction

.

.

.

Relevant Research .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Linkages with the Environment at the Village Tier

Linkages with the Environment at the Thana Tier .

CHAPTER 7:

THE PSYCHOSOCIAL SUBSYSTEM .

.

.

.

.

.

.

Intmuct 1 on

I

I

I

I

I

I

I

I

I

I

I

I

I

Q

I

I

I

I

R 61 emt R e S earCh I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

O

f

The Psychosocial Subsystem at the Village Tier

The Psychosocial Subsystem at the Thana Tier

.

.

.

137

141

1143

167

168

172

185

188

202

235

235

239

286

251,

276

347

347

349

355

378

394

395

too

1:16

1,31

.

.

.

.

.

.

CHAPTER 8:

THE MANAGEMENT SUBSYSTEMS

Introduction

.

.

.

Relevant Research .

.

.

.

.

.

.

.

.

.

.

Management at the Village Tier

Management at the Thana Tier

CHAPTER 9:

CONCLUSIONS

.

.

.

.

.

.

.

.

.

.

.

.

.

.

APPENDIX A:

MATHEMATICAL DERIVATIONS

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

APPENDIX B:

COOPERATIVE ASSOCIATION(QUESTIONNAIRE .

.

.

APPENDIX C:

COOPERATIVE CREDIT SIMULATION .

.

.

.

.

.

.

.

.

Page

440

440

446

452

460

475

482

491

498

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

BIBLIOW I

I

I

I

I

I

I

I

I

I

I.

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

510

Table

LIST OF TABLES

2-1

Rural Development Channels in East Pakistan

.

.

.

2—2

Rural Development Channels in Comilla Kotwali Thana

'Page

45

46

Forced Solutions for Three Input Signals to the Rein-

forcement Mbdel of Dyadic Interaction

.

.

.

.

115

4-1

T-tests for Minimum in Percent of Supervisors in the

1966 Projects

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

210

4-2

Regression Estimates of Structural Subsystem Para-

meters for the 1966 Projects .

.

.

.

.

.

.

.

.

.

.

.

217

4-3

Structural Subsystem Regressions for the 1966 Pro-

jects in the Ayub Khan Era .

.

.

.

.

.

.

.

.

.

4—4

A Conservative Interpretation of the Structural Sub-

system Parameters for the 1966 Projects

.

.

.

221

223

5-1

Growth of Cooperatives, Loan Transactions, and

Inspectors in the Field Supervision Section of

the KTCCA (1962-66) and in the.Agricultural C00per-

atives Federation (1967-70)

.

.

.

.

.

.

.

.

.

265

Distribution of Agricultural Cooperatives Federation

Loans by Purpose (Percent of Annual Loan Issue

Per Purpose) .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

267

Growth in Membership, Cooperatives, and Loan Issue

in

K012188121. Thana

o

e

o

o

o

a

I

a

o

o

e

a

a

o

a

0

Capital Formation in the Agricultural COOperatives

in

KOtMi Than I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

Regression Estimates for the Capital Generation

Ehuation in Kotwali Thana and in the 1966 Pro-

jects..............o.....

Regression Estimates for the Capital Generation

Equation in Seven Thanas .

.

.

.

.

.

.

.

.

.

.

Comparison of Actual and Estimated Operating Costs

an Average Thana Cooperative Association .

.

.

269

273

280

282

300

Table

Page

viii

5-8

Comparison of Actual and Estimated Values for the

Fidelity Hm I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

Year 10 Values of Village Capital, Loans Outstanding,

and Annual Profits for Different Rates of Growth

in the Number of Village Cooperatives

.

.

.

.

.

5-10

Year 10 State Variable Values for Different Rates of

Deposit in the Fixed Deposit Account .

.

.

.

.

.

5-11

Breakaway values for Different Settings of PRLATE

”d BEI‘A I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

Managers and Model Farmers Trained Each Year for the

Agricultural Cooperatives of Comilla Kotwali

mm 0

e

o

I

a

0

0

o

I

I

O

o

0

O

o

a

o

o

o

o

o

8-1

Irrigated Acreage for Cooperatives With Good,

8-2

8-3

8-4

Average and Bad Management .

.

.

.

.

.

Background Characteristics of Inspectors .

.

.

Inspectors' Ranking of Their Activities

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Comparison of Managers' and Inspectors' Activities

inKotwaliThana......-.-.......

303

326

329

336

427

457

462

464

466

Figure

LIST OF FIGURES

2-1

The Territorial HierarChy of Administration

2-2

Levels of Administrative Interest

.

.

.

.

.

.

.

3-1

A Simplified System Typology .

.

.

.

.

.

.

.

.

.

.

.

3-2

Phase Plane for a Closed System Dyadic Interaction

MOdel

.

.

.

.

.

.

.

.

.

.

.

.

.

3-3

A Step Function With Amplitude B

.

.

.

.

.

.

.

.

.

.

.

.

.

.

3-4

Phase Plane for an Open System Dyadic Interaction

Model, WithU (t)=e=8a. .

.

.

3-5

A Ramp Function With Slope B .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Page

36

42

96

105

110

111

113

3-6

Phase Plane for an Open System Model of Dyadic Inter-

action With a Delayed Step Input .

.

.

.

.

.

.

114

3-7

Phase Plane for an Open System Model of Dyadic Inter-

action With a Ramp Input .

3-8

A Complete Systems Typology

.

.

.

.

.

.

.

.

.

.

.

.

116

138

138

140

144

147

153

155

156

157

159

160

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

3-9

Comparison With the Boulding Typology

3-10

Comparison With Mathematical Models

.

3-11

Kast and.Rosenzweig's Paradigm of Organizational

Subsystems .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

3-12 A Second View of the Organization as Comprised of

Subsystems .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

3-13 The Organization as a Composite of Levels of Control

3-14 The Organization as a Composite of Subsystems

3-15

Exploded View of the Organizational Paradigm .

.

.

3-16

Subsystem Components of Organizational Systems .

3-17

Chapters in the Context of the Systems Typology

3-18

Systemic DeveIOpment .

.

.

.

.

.

.

.

.

.

.

.

.

.

vii

Figure

4-1

A Three Level Bureaucracy

4-2

A Two-Tiered Organization

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

4-3

Causal Loop Diagram Showing the Interactions Among

Structural Variables in the Hummon Model .

.

.

.

4-4

Simplified Causal Loop Diagram of the Hummon Mbdel .

4-5

Organization Chart of the Kotwali Thana Central

Cooperative Association (KTCCA) as of 1961 .

.

.

4-6

Organization Chart of the Agricultural Cooperatives

Federation (ACF) as of 1969

.

.

.

4-7

The Plan and Boulding Effects

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

4-8

A Curvilinear Version of the Blau and Boulding

Effects

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

4-9

Size Versus Percent of Supervisors in the Agricul-

tural Cooperatives Federation

.

.

.

.

.

.

.

.

.

4-10

Scattergram of Percent of Supervisors Versus Size in

th91966PrOjeCtSace-00000000...

4-11

Evolution of Percent of Supervisors Versus Size in

Thana Cooperative Associations .

.

.

.

.

.

.

.

.

170

171

180

181

189

191

203

204

206

208

211

4-12

Causal Loop Diagram for the Structural Subsystem of

the 1966 Projects

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

220

4-13

Causal Loop Diagram of Structural Subsystem Para-

meters for the Period 1966 - 1968

.

.

.

.

.

.

.

222

4-14 Dynamic Model Predictions Compared With Static

Findings for the 1966 Projects .

.

.

.

.

.

.

.

.

4-15

Size Versus Differentiation in the Thana Coopera-

tiveASSOCiationScacao-0000.0...

5-1

Block Diagram of the Loan Repayment Delay

5-2

Eulerian Integral of the Function f (t)

.

.

.

.

.

.

.

5-3

Causal LooP Diagram of the Financial Equations .

5-4

The Capital Generation Function, XSAV

.

.

.

.

.

.

.

.

.

.

.

.

5-5

The Loan Issue Equ-ations O

O

O

I

I

0

o

0

I

o

o

e

e

O

a

5-6

Short Run Profit Projection for the Simulated Thana

Cooperative ASSOCiation

a

o

a

o

a

a

a

I

o

a

a

a

225

228

287

290

292

293

296

305

Figure

Page

xi

5-7

Projections of Loans Outstanding (TVLOAN), and

Bad Debts (BDSM) and Village Capital (TSAV)

From the Extended Run

.

.

.

.

.

.

.

.

.

.

.

.

.

5-8

Profit Projections From the Extended Run .

.

.

.

.

.

.

.

5-9

Phase Plane of Village Loans Outstanding Plus Bad

Debts (TOTLOAN) Versus Village Capital (TSAV)

From the Ext endfl Run

I

I

I

I

I

I

I

I

I

I

I

I

I

I

5-10

Phase Portrait of TVLOAN Versus TSAV From the Extended

Run

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

IVI

I

5-11

A More Complete Causal Loop Diagram for the Coopera-

tive Credit System .

5-12

Comparison of Trajectories for the Extended Run and

the Basel-ine Run I

I

I

I

I

I

.I

I

I

I

I

I

I

I

I

I

I

5-13

Phase Plane of TSAV Versus TOTLOAN From the Closed

sySteIn Rm I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

5-14

Projections of Loans Outstanding, Bad Debts and

Village Capital From the Closed System Run .

.

.

.

Trajectories for the No-Growth Run, the No-Capital

Loans Run, and the Baseline Run

.

.

.

.

.

.

.

.

.

The XCOOPS motion I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

5-15

5-16

5-17

307

309

311

312

314

316

318

319

321

323

Phase Portrait of System Response to Three Rates of

Growth in Village Cooperatives .

.

.

.

.

.

.

.

.

.

324

5-18

Phase Portrait of System Response to Different Values

Of BEI‘A

o

I

o

o

o

e

o

o

o

o

I

o

o

o

o

a

o

o

0

O

0

5-19

Phase Portrait of System Response to Different Values

of VLIR, the Village Loan Interest Rate

.

.

.

.

.

5-20

Phase Portrait of Village Capital Versus Loans Out-

standing for the Kotwali Thana Cooperatives

Exploded View of the Organizational Paradigm .

The Institution-Building Universe

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

The Managerial Task:

Strategic, Coordinative, and

Operating Subsystems .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

Examples of Activities in a Business Organization

Included in Major Framework Headings .

.

.

.

.

.

.

328

333

338

396

401

449

INTRODUCTION

Having grown up among the children of scientists in the posh,

protected think tank of Princeton, New Jersey, I assimilated the

antiquated assumption, which is more appropriate to the eighteenth

than to the twentieth century, that the world is rational.

In Castro's

revolutionary Cuba, where I traveled with some Swarthmore college

friends a few months before the Bay of Pigs invasion, the orderly social

universe to which I was accustomed was thrown up into the bright

Carribean sky, with liberated sugar cane workers and student radicals

laughing as the cracked shards of my universe fell about them.

Ever

since that trip to Cuba I have been seeking to reconstruct the old

verities in a new world of poverty, political passion, and civil war.

This book is an effort to bring order to the interpretation of two

tours of inspection of that tumultuous other world.

The first tour was

to the country that is now called Bangladesh and was then called

East Pakistan.

As an Africa-Asia Public Service Fellow, I was

stationed at the ACademy for Rural Development in Camilla, East Pakistan,

in 1965-66.

Perhaps there are places on this planet where the people

have less clothing, housing and food than they did in East Pakistan in

the nineteen sixties, but nowhere else could the wretched earthlings be

so proximate to one another across such a vast plain.

Yet even there,

amidst those tiny huts and farms, the rural people had hope of abundanCe.

Their inspiration was in the feats of road building, canal dredging and

crop improvement which they discovered they could accomplish with one

1

2

another in the rural development organizations sponsored by the Academy

for'Rural Development at Camilla.

It is those organizations -- mainly

the structure of rural c00peratives -- which are at the center of

attention in this book.

The second inspection tour was four times longer, lasting from 1967

to 1971.

Its impact upon me was greater -- almost devastating -- yet

its influence upon this book is more like that of a leitmotif than a

central theme.

The place where I lived on that second tour was also a

part of rural Asia, and in my eyes there were commonalities in the methods

employed there to organize and mobilize the rural masses.

Yet I have

already discovered, in defending this manuscript when it was submitted

as a doctoral thesis at Michigan State University, that American experts

on rural development resent analogies between their'pet projects for

uplifting the rural poor and the organizations for revolutionary change

in Vietnam.

The intellectual window through which I have examined the Comilla

cooperatives is that of organization theory.

Through a theoretical

interpretation, I have sought to generalize the explanation of the rural

develOpment organizations founded in Comilla.

The significance of those

organizations for rural development projects elsewhere, and for organiza-

tions in general, has thereby been demonstrated more clearly than would

have been the case if a purely anecdotal exposition had been employed.

As if the mixture of Bengal, Vietnam, and organization theory were

not complex enough, the third chapter of this book delves into other

topics not generally thought to be germane to rural development.

The

second law of thermodynamics -- the so-called entropy law -- is discussed

in the third chapter, as are arcane issues such as the difference between

static and dynamic systems.

Although the flights of theorizing in the

3

third chapter'do carry the discussion far from the agricultural coopera-

tives at Comilla, I contend that the third chapter makes the case for a

new version of organization theory which is adaptable to the interpreta-

tion of rural development organizations, and to the interpretation of the

corporate industrial organizations which are the more typical objects of

discussion in contemporary essays in organization theory.

In the five chapters fellowing the theoretical chapter I have

applied the concepts elaborated in the third chapter to the explanation

of the Comilla cooperatives.

These chapters treat the cooperatives'

structure (Chapter 4), their functioning (Chapter 5), their external

linkages (Chapter 6), their ideological cohesion (Chapter 7) and their

management (Chapter 8).

In two of these chapters, those concerning

structure and function, I have formalized the interpretation via

mathematical models.

The second of these models, that concerning the

credit Operations of the cooperatives, was translated into a computer

program which tracked the long term solution of the simultation model

under a range of assumptions about factors such as rates of interest and

loan repayment.

I have laid out this roadmap so that the reader will be less confused

when he takes unconventional intellectual routes through the subject

matter of this book.

If the reader is professionally concerned with

agricultural modernization in the developing countries he will have

heard of the Comilla experiments in rural uplift, but he may be unpre-

pared to contemplate analogies between the National Liberation Front in

Vietnam and the agricultural cooperatives in the villages of,Comilla.

Or

if the reader is interested in the theory of formal organizations, he

may be taken aback to find that theory applied here to organizations

populated by Asian peasants rather than American executives.

And no

A

matter what the reader's professional training, I expect he will recog-

nize that this is not a lean book.

One of the members of my thesis

advisory committee at Michigan State University, an educational psycho-

logist named John Vinsonhaler, remarked that the manuscript was:more

appropriate as a life's work than as a thesis.

Dr. Vinsonhaler's

assessment may have been correct.

I regard this book as a summary of

the inner meaning of the bizarre combination of experiences and education

which I have been privileged to enjoy in the last decade and a half.

This is my mid-life memoir.

Those who assisted in the completion of this work are mostly

associated with Michigan State University in East Lansing, with the

government of Bangladesh, or with the Academy for Rural Deve10pment in

Comilla.

One man, who is quoted frequently in this volume, has been

associated at various times with the faculty of Michigan State University

and with the administration of the rural academy in Comilla.

I am

grateful to Akhter’Hameed Khan for his inspiration, not only as an

architect of rural development organizations but also as a monumental

exemplar of ceaseless curiosity.

The professor who served during the

nineteen sixties as the head of Michigan State University's Pakistan

Project, Richard Niehoff, guided me in factual research about the

Academy for Rural Development at Comilla, especially in its relationship

to Michigan State.

I am grateful to Professor Vandel C. Johnson,

chairman of the Department of Administration and Higher Education at

MSU, for his liberality in allowing me to pursue such an unconventional

topic for thesis research.

And to John E. Hunter, Professor'of

Psychology at MSU and a member of my doctoral committee, I must say, with

a wince, that I appreciate his straightforward criticisms.

On the pages of

an earlier draft of this manuscript, Professor Hunter’penned remarks such

5

as "Unnecessary, uninformative, and irritating degression," or "More

junk, out of place."

Professor Hunter also had a hand in improving my

statistical interpretations.

For their assistance in completing the questionnaire from which

the information for the analysis of the structure of the thana coopera-

tive associations was drawn, I am gmateful to A. M. Abdul Mannan Bhuiya,

formerly Director of the Integrated Rural Development Program of the

Government of Bangladesh, and to

M. Baset, Project Director of the

Integrated Rural Development Program at Comilla.

Within the Academy for

Rural Development at Comilla, Zaker Hussein, Aziz Khan, Md. Solaiman,

Badaruddin Ahmed, and Ali Akbar were especially helpful in providing

data for this research project.

For their labor in looking through

organizational records of a decade ago so as to fill out my Cooperative

Association Questionnaire, I am endebted to Rafinul Islam (Gaibandha

Thana Central Cooperative Association), the project officer of the

Gouripur Thana Central Cooperative Association, Mostafizur Rahman (Laksam

T.C.C.A.), Md. Mosharat Hossain (Chandina T.C.C.A.), Md. Ali Akbar Sarkar

(Sarail T.C.C.A.), Md. Abdur Rahim (Hajiganj T.C.C.A.), Md. Zahedul Islam

(Brahman Baria T.C.C.A.), H. M. Abdul Mabud (Chandpur T.C.C.A.),

Swapan Kumar Guha (Chouddagram T.C.C.A.), Ahmmed Ullah Khan (Barura

T.C.C.A.), Abdul Motalib (Burichang T.C.C.A.), B. P. Barua (Bebidwar

T.C.C.A.), Md. Manirul Haque Khandaker (Muradnagar T.C.C.A.),

Md. Altaf Hossain (Daudkandi T.C.C.A.), M. A. Wadud (Homna T.C.C.A.),

Md. Bahar-Ul-Islam (Kachua T.C.C.A. ), Md. Delwar Hussain (Matlabganj

T.C.C.A. ), Kafil Uddin Ahmed (Faridganj T.C.C.A. ), Niaz Mohammed Bhuiya

(BanchaMpur T.C.C.A.), Md. Zawadul Huque (Nabinagar T.C.C.A.), and to

the project officer for the Nasirnagar T.C.C.A.

These were busy offICials

in the Integrated Rural Development Program in Bangladesh, and their

6

willingness to contribute to my research on the structure of thana

cooperative associations is deeply appreciated.

Finally, I should like to thank my wife and children for their

patience.

On many a night I returned from the Michigan State University

computer center at two or three in the morning.

Wakening, and trying

her best to sound helpful, my wife Margo would ask, "Did the computer

program run?"

I would grunt "no" and pray that she would have faith in

my incompetence a little longer.

My young son, Anson, gave up waiting

for the thesis to be done; as a birthday present, he gave me a small

stapled book of blank white pages, titled in red crayon on the cover,

"A Book of Thesis."

East Lansing, Michigan

October, 1979

CHAPTER 1

The Comilla Experiment

...Small farmers and the landless have no weapon

against the system but organization.1

In the Chilalo valley of Ethiopia the Swedish International Develop-

ment agency and the Ethiopian government collaborated in the late

nineteen sixties to create a model of intensive agricultural develop-

ment, the Chilalo Agricultural Development Unit, or CADU.

In a review

of the accomplishments and failings of CADU, John M. Cohen explained

some of the intellectual origins of the project.

The Swedish and

Ethiopian planners who designed CADU drew heavily on ideas about rural

development organization that had originated in a pilot project which

began five years earlier, in a place called Comilla, in East Pakistan.

"In the history of international development," Cohen remarked, "the

Comilla project...must be ranked as one of the most successful and

influential agricultural development programs in the third world."2

Founding 9f thg Comilla Academy

The town of Comilla, located midway between the major cities of

Dacca and Chittagong, is a medium-sized commercial and governmental

center for a large hinterland of rice-farming villages.

In the wet

7

8

season the fields surrounding the town appear to stretch in some places

for miles, undulating in a green blanket of rice plants or sheeted with

floodwaters.

The winter dry season reveals those same fields as

divided into miniscule diked paddies.

Mudwalled thatch houses are

scattered here and there among groves of cocoanut palms and bamboo.

A

westerner who stands on a roadbank to survey this idyllic scene may

think himself virtually alone, but will soon be surrounded by banks upon

banks of staring, brown-skinned villagers who would seem to have

materialized out of the earth.

In this setting, which except for its extreme density of population

appeared to be similar to the impoverished, tradition-bound rice growing

areas of many other regions of Asia, the Pakistan Government and the Ford

Foundation collaborated to found the Pakistan Academy for VilLage

Development in 1959.

A sister training institution was founded at

Peshawar in what was then West Pakistan.

The Comilla academy was later

renamed the Pakistan Academy for Rural Development, and is now called

the Bangladesh Academy for Rural Development, or BARD for short.

Planning for the 1959 Opening of the Comilla academy had been

underway since 1956, when a four man team from Michigan State University

visited Pakistan, at the request of the prime minister, to draw up a

plan for the creation of two rural development training institutions.

Their idea was that public officials responsible for the implementation

of rural development programs would come to the academies to study

public administration and to learn about village life.

The faculty of

the academies were to be Pakistanis trained in western social science;

the academy faculty would do field research in the villages in the

vicinity of the academies, and would edify the government trainees by

imparting to them the findings of their surveys.

This original

9

conception was pursued fairly strictly in the rural development academy

at Peshawar, but in the Comilla academy it was to be substantially

amplified.

In June of 1957 the Pakistan government accepted the plan for the

establishment of the training academies.

The Ford Foundation designated

Michigan State University as a back up agency which would assist the

academies in the initial recruitment and training of personnel, staff

development, the purchase of library materials, and other such support

services.

Michigan State was also to provide a rotating staff of

resident project advisors.

For the technical assistance which Michigan

State University provided to the two academies between 1956 and 1969,

the Ford Foundation paid M.S.U. $1,935,082.

To launch the Comilla

academy, the foundation contributed a $210,000 grant for architects' fees

and imported construction materials for the dormitory, cafeteria (the

only one in East Pakistan at the time), library, classroom buildings,

administration building, and faculty residences at the academy Campus at

Kotbari, seven miles from the town of Comilla.

A later Ford Foundation

grant of $810,000 to the Government of East Pakistan supported the

foreign exchange costs of some of the academy-sponsored projects for

agricultural modernization.

Operational costs of the Comilla academy

were born by the Pakistan government, and totalled somewhat more than five

million rupees, or about a million dollars, from 1960 to 1968.

The

Pakistan government invested a similar amount for the construction of

the academy facilities.3

Compared to the funds that might have been

expended over a comparable period by a new college in the United States,

this level of support was modest.

But by the standards of public

institutions in East Pakistan, the funding provided by the Ford

Foundation and the Pakistan government was extraordinary.

The Academy Director

10

The most important decision that was made in the period in the late

fifties when the Comilla academy was being planned was the selection of

the academy director.

Akhter Hameed Khan, the principal of Comilla

Victoria College, was chosen for this critical post.

He brought to

the job an exceptional knowledge of the administrative policies of

Pakistan, gleaned from his participation in the Indian Civil Service

in the late thirties, and from his leadership of the East Pakistan

V-AID program in the mid fifties.

He also had a deep compulsion toward

innovation, and a compassion for the small farmers of the Comilla area.

Akhter Hameed Khan was born in 1914 in Agra, India.

He was the

eldest son of a police official of the government of British India.

His parents were Pathans:

the proud, tall Muslim people of India's

northwest frontier.

In comparison to most Muslims of that era, his

parents were exceptionally well read and cultivated, and so they

encouraged.Akhter Hameed Khan's rampant intellectual curiosity.

A

younger brother recalls that when Akhter Hameed Khan was taking food at

the dinner table with his right hand, a book would often be in his left

hand.

He explained to his younger brother that books were as essential

to him as food.

He continued his non-conformist, syncretic search for truth in

Meerut College, where he studied history, philosophy, and English

literature, and at Agra University, from which he received his M.A.

in English literature in 1934.

He was fond of posing provocative

questions such as "Why should I believe in the old dogmas of. my

religion if they are against logic?", and read widely in Arabic,

Persian, and Sanskrit in his query into religious fundamentals.

11

In 1934 he competed successfully in the examinations for the

Indian Civil Service, the most prestigous service in British India.

The ICS sent him for two years study at Maudline College, Cambridge.

Upon his return from England he showed himself to be as engrossed in

religious studies as ever, and thereby disappointed his father, who

had hoped that Akhter Hameed would assume the imperial bearing of an

ICS officer.

He also disappointed his British superiors in the ICS, who

took exception to the simplicity of his dress and life style.

A senior

ICS officer once admonished Akhter Hameed Khan that "the other officers

suffer by contrast."

His first posting in the ICS was to Comilla, where he was assistant

Magistrate in 1938.

He set about learning the Bengali language, which

he later spoke fluently, but with a decided Urdu accent.

His education

in the rural problems of East Bengal began in the late thirties and

was specially intensive when he served as the sub-divisional officer

for Netrokana in Mymensingh, another locality in what was to become

East Pakistan and then Bangladesh.

As the result of the worldwide great

depression,rural indebtedness had increased markedly in East Bengal,

and the inability of farmers to repay their debts pulled the under-

pinnings from the cooperative movement which had flourished in the region

in the period of the First World War.

As sub-divisional officer,

Akhter Hameed Khan was responsible for organizing Debt Settlement

Boards for the scaling down of debts.

He also served as the appellate

court for debt settlement cases.

In later years, Akhter Hameed Khan

observed that his understanding of the problems of rural credit in

East Pakistan derived from his personal experience with hundreds of

debt settlement cases.

Akhter Hameed Khan resigned from the Indian Civil Service in 1944.

12

His disenchantment with the ICS derived from his perception of the

inadequacy of the service, and of his participation in it, in the face

of the poverty, injustice, and oppression of wartime India.

The

populace under his administration had been decimated by the great Bengal

Famine of 1943, a catastrophe which must have been harrowing for

anyone to witness.

His restlessness in government service had been

augmented

by his marriage in 1939 to Hamida Mashraqui, the daughter of

the founder and head of a semi-military political organization known as

Khaksar.

Carrying out his resolution to lead the life of a poor man,

Akhter Hameed Khan moved his family to Mamoola, a village near.Aligarh.

He rented a small house and acquired a goat and a buffalo cow.

Finding

that he couldn't make a living as a milkman, he apprenticed himself to

a blacksmith to learn the art of lock making.

When he had gained

proficiency as a locksmith, he organized a cooperative workshop for

lock manufacture.

When the Indian colonial era came to an end in 1947, Akhter Hameed

Khan returned to public service in the role of an educator.

He served

as headmaster of a secondary school in Delhi, as a lecturer in Islamic

history and English literature at Islamia College in Karachi, and in

1950 was appointed principal of Comilla Victoria College.

His youthful

interest in religious studies persisted; as principal, he sometimes

requested the faculty to leave him unmolested in his office for weeks

at a time so that he might pursue his readings of Islamic and Buddhist

literature.

During his years in the Comilla principalship, Akhter Hameed Khan

served for several terms as president of the East Pakistan Non-Government

Teachers Association.

In his presidential addresses he called for

13

reform of the structure and curriculum of East Pakistan education so as

to make it more germane to the needs of a rural society.

His speeches

often burst the bounds of educational issues, and pointed to the need

for broader social change.

In 1952, for instance, he observed:

Look, for example, how the village is neglected.

The village is a place without opportunities:

there you have under-employment, sweat labour,

uncertain produce, squalor, disease and boredom.

Naturally as many as can want to leave the

village.

But unfortunately the village is the

foundation of our national life.

It is dangerous

to weaken the foundation.“

An opportunity to test out his rural reconstructionist ideals would

seem to have been presented to Akhter Hameed Khan in 1954, when he took

leave of his college duties for a year to serve as the first director

for East Pakistan of the Village Agricultural and Industrial DevelOpment

(V-AID) program.

His recollections about the experience, however, are

generally negative.

Understanding Akhter Hameed Khan's disagreements

with the V-AID program is important, because many of the innovations

which he introduced a decade or so later in the Comilla academy's rural

development projects were justified, in his mind, as responses to the

deficiencies of V-AID.

The Village Agricultural and Industrial Development program was a

classic example of the community develoPment fad which in the nineteen

fifties swept through the less developed countries on the wings of

American money and persuasion.

The international prototype of a

community development program of national scope was launched in India

in 1952.

The Ford Foundation had supported a plan to set up 15

community development projects covering 100 Indian villages each; the

projects were to be built on the model of Albert Mayer's Etawah pilot

14

project in the state of Uttar Pradesh.

When the Indian government got

wind of the possibility of massive funding for community development

through United States foreign aid, the scheme was inflated to include

55 projects each covering 300 villages.5 And so it went, in India,

in Pakistan, and in many other countries.

With the fertilization of

American foreign aid, and with the blessings of several United Nations

agencies, the community deve10pment movement grew phenomenally.

By

1960 over sixty countries in Asia, Africa, and Latin America had

initiated major programs for community development, constructed for the

most part according to the Indian example.

The demise of the international community development movement was

even speedier.

By 1965 most countries, including India, had cut back

or terminated their community development programs.

The life cycle of

this international movement, then, was about 13 years.

In Pakistan

the V-AID program was abandoned in 1961, nine years after it was

started.

The course of the precipitous decline of the community develop-

ment movement was in large part the reverse of the cause for its

meteorotic ascent:

the United States government had lost interest, and

was putting its foreign aid money elsewhere.

Disillusionment had set’in among national leaders in the developing

countries as well.

They had been led to expect that community deve10pment

would yield a plenitude of happy consequences ranging from increased

agricultural productivity to the political stabilization of rural areas.

They observed, after a decade of utopian promises, that few of these

benefits actually accrued.

Within the national government, moreover,

the community development programs occasioned unending jurisdictional

and professional disputes, since the more established ministries, such

as agriculture, generally viewed the community development ministry as

15

a nest of technically incompetent intruders.

In the absence of

American funding or rural results, most national leaders were glad to

be rid of the community development headache.

Although Akhter Hameed Khan has sharpened his hindsight with the

aid of the many critiques of community development that have been

published in recent years, it appears that he did perceive many of the

inadequacies of the community deve10pment approach in the mid nineteen

fifties, when the international community development movement was in

the first stages of proliferation.

As V-AID director, he was disturbed

to find that the American backers of the program (field representatives

of the International Cooperation Administration, a predecessor of the

U.S. agency for International Development) believed that all of the

methods of community development were known, and that the only task was

to implement them.

Akhter Hameed Khan requested, but was not granted,

a small pilot project rural area in which to experiment with different

approaches to community deve10pment.

In the district of Khulna he

did try out one experiment for which he was reprimanded by the community

development authorities:

he authorized the expenditure of 450,000 rupees

for the construction of a flood control embankment.

If the embankment

had not been built, he contended, the Khulna villagers would not have

bought fertilizer for their fields or installed the sanitation facili-

ties and fishponds which the V-AID program promoted.

Akhter Hameed

Khan also decried the distrust with which the community development

experts viewed the traditional civil administration.

His American

advisor, a sociologist named John Green, told Akhter Hameed Khan that

community development work must be conducted independently of the

government establishment.

Lastly, Akhter Hameed Khan recognized that

the economic content of the community development program was

16

insufficient.

In a speech which he gave to a conference at the

Massachusetts Institute of Technology in 1964, he remarked, "So as I

look back on these last ten years my own advice would be that nobody

should be allowed to talk about community development unless he has

passed a tough course in public administration and economic planning."7

Because he found himself disagreeing with so many of the precepts

of the Village Agricultural and Industrial Development program, Akhter

Hameed Khan left the job of provincial V-AID director in 1956, and

returned to his principalship at Comilla Victoria College.

After his selection in 1958 as the first director of the Comilla

Academy for Village Development, Akhter Hameed Khan and the newly

recruited faculty of the academy (all of whom had M.A.s, and some of

whom had Ph.D.s) were flown to Michigan State University for nine

months of training, brainstorming, and planning for the job ahead.

MSU

faculty led discussions on community deve10pment, social change, public

administration, and other aspects of the intellectual melange of

American scholarship on problems of rural development in the post-

colonial countries.

Among the ideas which had the most lasting impact

on the personnel of the Comilla academy were the concepts of the land

grant college, an institution linking rural community involvement with

academic investigation, and the idea of action research for problem

solving.

The thought of applying social science action research to

issues in public administration seemed particularly intriguing.

The Rural Setting 2f the Academy

After their'arrival in Comilla in May of 1959, the faculty of the

newly founded academy set about studying the nearby villages.

The

17

setting of the academy, they discovered, was bleak.

The rich rice land

around Comilla town was desperately overcrowded, with a population

density of over 1,500 per square mile in the rural areas.

In propor-

tion to cultivated area, the population density was over 2,000

persons per square mile.8

The average farm per rural household was

only 1.7 acres in area.

These tiny holdings were further divided and

scattered here and there in paddies, ponds, and houselots.

In the Comilla area, as elsewhere in East Pakistan, the difficulty

of making a living on such a crowded landscape was compounded by the

deterioration of earthworks which had traditionally protected the

farm land.

Flood embankments had been breached, and drainage channels

had become choked with silt.

During the colonial era, the maintenance

of these earthworks had been mainly the responsibility of the zamindars,

a peculiar class of hereditary tax collectors cum landlords.

But the

traditional arrangement for taxation and public works had been

disrupted by the depression, the second world war, and independence,

leaving the earthworks unrepaired for decades.

Economic conditions had grown so harsh in some of the most flood-

prone villages near Comilla that farmers had had to sell their bullocks

and were attempting to heave wooden plows through the soil with the

force of their emaciated limbs.

Wives had sold their jewelry, one of

the most important liquid assets for a peasant family.

Some families

were contemplating migration to Burma as the only way to stay alive.

Social deterioration had proceeded apace with family economic

decline.

Beggars and thieves were numerous in the Comilla area in the

ninteen fifties.

Many farmers were unwilling to make investments for

the improvement of their yields for fear their crops would be cut in

the night and stolen.

The folk method of preserving order and settling

18

disputes via a council of elders known as sgmgj had become nearly

impotent.

Peter Bertocci, who conducted an intensive anthropological

study of Comilla villages, reports that "...throughout my entire

period of contact with these villages I heard continual lamenting

of the fact that the sardari system of conflict resolution had 'broken

down,‘ that it had 'gone bad' and that the £11m had been destroyed."9

The population was generally illiterate; 73 percent of all 15 year

olds -- and 86 percent of all 15 year old females -- could not read in

1960.

Poor health debilitated many villagers, the common afflictions

being cholera, typhoid, tuberculosis, dysentery, and worm infections.

Economic status in the Comilla villages was determined mainly on

the basis of the size of landholdings.

At the bottom of the heap were

landless laborers.

Their numbers at the time the academy began have

unfortunately not been recorded, but in 1967 Bertocci estimated them

as about 10 percent of a sample of 102 Comilla thanacultivators.10

The landless laborers lived hand to mouth, selling their services at

planting or harvest time.

They were so poor that they could not

accumulate the wherewithal to rent land to cultivate.

They were

unemployed about half of each year.

Their destitution was magnified

by the coincidence of their dry-season unemployment, when there was

no work to be done for hire in the fields, with the period of highest

rice prices.

As might be expected, many of these landless laborers

deserted the villages in search of more stable employment in the towns.

The land poor were at the next level of economic stratification in

the Comilla villages.

In Bertocci's classification, these farmers

owned less than the median landholding of 1.15 acres.

They had only

the most precarious toehold on the agricultural economy, and could not

raise enough rice on their own plots to feed an average family of

19

five.

To sustain themselves, these land poor cultivators would have to

sell their labor to other farmers.

The land poor farmers constituted

34 percent of Bertocci's sample.

The next group in the economic scale, accounting for 47 percent

of Bertocci's 1967 sample, was variously denoted as the middle farmers

or the small farmers.

Bertocci bracketted the landholdings of this

group as between 1.15 acres (his sample median) and 4 acres.

Akhter

Hameed Khan has described the small farmers as owning between 1.5 and

5 acres.11

When raising two cr0ps of rice per year without an

affliction of flood, drought, or pests, the small farmer could supply

his family with rice and market a small surplus.

With this little

margin of cash, the small farmer could purchase agricultural supplies

such as fertilizer, and meet a few of his family's consumption needs.

Since he had so little reserve to fall back on, the small farmer usually

sold his crop at harvest time, when the price of rice might be half

the value it attained in the winter dry season.

To meet his need for

cash prior to the fall harvest of the gmgg rice crop, the small farmer

frequently sought loans from rural moneylenders, and should things go

badly for him in any season, the small farmer might have to turn over

part of his tiny landholdings to the moneylender (in usufructory

mortgage) in order to obtain a large enough loan to tide him over.

The

pressures of natural disaster, exploitation by rice merchants, high

interest rates, and the division of landholdings among sons were con-

stantly pressing the small farmer downward toward the category of land-

poor cultivators, or even into the nether pool of landless laborers.

At the top of the lowly economic pyramid in the Bengali village

was the class of surplus farmers or big farmers.

Nine percent of the

cultivators in the two Comilla villages in Bertocci's study were in

20

this category.

This small group of farmers were the lucky few with

enough land to feed their families, meet their needs for agricultural

supplies and consumption goods, and still have a margin of surplus to

use as equity capital for moneylending and rice trading.

In the

Comilla villages of the nineteen fifties and early sixties these

farmers had a vested interest in the status quq.

They could invest in

 

loans to land poor and small farmers, or in purchasing and stocking

rice for resale at higher prices, and thereby make substantial profits

with very small risk.

If the small farmer or the land poor farmer

could not repay the principal of his loan along with the 80-100 percent

per year interest, the surplus farmer often retained the use of the

mortgaged land.

He might lease the mortgaged property pending repay-

ment of the loan, and thereby accumulate still more wealth at little

risk.

When urged by village level extension workers to ad0pt better

agricultural practices, these surplus farmers would gladly go along with

the use of fertilizers or pesticides if these novel commodities were

offered for free: after the subsidy for the demonstration ended, the

surplus farmers would typically revert to the traditional practices

which brought them such a relatively comfortable way of life.

East

Pakistani programs for rural uplift in the nineteen fifties, most

notably V-AID, had been captured by this village elite; as a result

the programs had neither boosted agricultural productivity nor arrested

the downward economic spiral in which most villagers were caught.

Much of the preceding analysis of class structure in a Comilla

village is drawn from a speech given by Akhter Hameed Khan to a

conference of personnel of the United States agency for International

Development, held in Comilla in June of 1963.12

In this speech to the

Americans, he described a vision of rural society that was close to

21

that of Mao Tse Tung.13 Many years later, after the collapse of the

conservative East Pakistan regime and.after his own retreat from the

province because of the ethnic hatreds that boiled up during the war of

independence, Akhter Hameed Khan admitted that he had been reading

Mao's works in secret for many years.

In his speeches to Pakistan

government officials and to Comilla villagers during the nineteen

sixties, Akhter Hameed Khan was less explicit about his vision of rural

class conflict; for those domestic audiences he stressed the importance

of vigorous administration or of community solidarity.

Time has born

out the accuracy of Akhter Hameed Khan's radical inner vision, the

outline of which is taken for granted in most informed discussions of

agrarian problems in contemporary Bangladesh.

Initiatives lg Rural Development Organization

In August of 1959, three months after they had regrouped in

Comilla after their training in Michigan, in England, and elsewhere

abroad, the faculty of the Comilla academy faced their first class of

trainees.

These were all, as Akhter Hameed Khan recollects, "experienced

government officers who had been connected with village life possibly

for fifteen or twenty years....

It was interesting for me to watch how

nervous the instructors were in facing these officers and how sure, on

the other hand, the officers were that the instructors did not know

anything about village life which was relevant to them."11+ The trainees

were determinedly sceptical; they said that the odds against significant

rural change were too great, that the instructors' foreign ideas had no

chance of being implanted among the quarrelsome, downtrodden villagers

of East Pakistan.

22

The upshot of this contretemps was that Akhter Hameed Khan

obtained the backing of the academy faculty for the establishment of a

working demonstration of rural development.

He asked the East Pakistan

government to allocate the rural region adjoining the academy,

Comilla Kotwali Thana, as a laboratory and demonstration area to be

attached to the academy.

Obtaining approval for this unusual request

required some lobbying, and Akhter Hameed Khan did not hesitate to make

use of his connections with high government officials who had been his

colleagues in the Indian Civil Service in the nineteen thirties.

One

old friend, the fist Pakistan Chief Secretary, told him, "You are a fool,

but a good foOl, so I will let you go ahead."

When the request for a laboratory area was granted, the academy

assumed the responsibility, which was (and still is) quite rare for an

academic and training institution, of making many of the most crucial

decisions as to how an administrative area was virtually to be governed.

The academy did not assume full executive control over the thana,

since representatives of the civil administration, police, and nation

building departments still reported to their respective superiors above

the thana level.

The academy obtained, nonetheless, a very considerable

latitude for rearranging administrative relationships and for launching

pilot projects that were both rural and urban in nature.

(The staff of

the West Pakistan Academy, in Peshawar, did not take on comparable

responsibility for the experimental development of a specific area.

Not until Akhter Hameed Khan associated himself with the Peshawar

academy in 1974 did it undertake to combine theory and practice in the

demonstration development project in Daudzai thana.)

The size of a thana, according to Akhter Hameed Khan, was deter-

mined originally by the capacities of a horse; the borders of the thana

23

were drawn in British imperial days by estimation of the distance that

a man on horseback could travel from a police station and return in

one day.

In area, the thanas averaged 100 square miles.

Comilla

Kotwali Thana contained 107 square miles, a territory which the academy

staff thought might be manageable for the purposes of deve10pment

administration, just as it had proven manageable for police admini-

stration.

Explaining the selection of the thana as the experimental project

area, Akhter Hameed Khan said, "The community deve10pment people had a

kind of allergy to the existing administrative structure.

We said

don't create a new unit, but take the old law and order unit, and make

it the basis of the reform.

For over a hundred and fifty years the

police station had been the symbol of the presence of the government as

a maintainer of law and order.

We said we would put the development

center next to the police station, to show the new concern of the

government."15

A still more important reason for seeking rural development models

which were implementable on a thana scale was that the thana was the

lowest level to which the civil administration -- the structure

descended from the steel frame of government that had been manned by

the Indian Civil Service in imperial days -- posted its officers.

The

thana-level administrative officer was the circle officer, and he was

the man Akhter Hameed Khan had his eye on as the coordinator of local

development projects.

"Nobody except a captain can really coordinate

from the public administration point of view," he commented.' "Again,

if I was talking the language of community development I could speak

about human relations and tact and all that.

But, in public admini-

stration, it really does not work.

There must be a captain, and the

captain is the coordinator.

He is the chief."16

2h

A second fundamental decision about the organization of rural

development in Comilla was also related to the territorial hierarchy of

administration, but was slower to crystallize.

In the first years of

trial and error experimentation, the academy staff organized farmer's

cooperatives in more than 20 villages of Comilla Kotwali Thana.

The

aim of the cooperatives was to mobilize the small farmers, and to

provide them with a means of group

protection against rice merchants

and moneylenders.

It was realized during 1960 and 1961 that these

independent village level cooperatives could not muster the resources,

the technical expertise, or the bargaining strength to make break-

throughs in agricultural modernization on the one hand, or to deal with

exploitative rice merchants and moneylenders on the other hand.

Accord-

ingly, the Kotwali Thana Central Cooperative Association (KTCCA) was

organized and registered as a cooperative in 1962.

With the establishment of the KTCCA as a thana level umbrella organ-

ization, the two-tier administrative structure which was to become the

hallmark of the Comilla cooperatives had been initiated.

As will become

apparent in later chapters of this book, the elements at each tier and

the relationships between the tiers were successively transformed

through the nineteen sixties.

But the basic pattern of village level

primary cooperatives and a thana level cooperative association remained

fixed.

The two-tiered structure is the aspect of the Comilla coopera-

tives which has been most widely imitated, as in the Integrated Rural

Development Program in Bangladesh, and in the new cooperatives which

Akhter Hameed Khan helped to bring into being in Daudzai thana of

Peshawar, in Pakistan.

A third organizational decision was also closely related to the

25

pattern of imperial administration via a territorial hierarchy.

This

was to create a Thana Training and Development Center in which local

government, administrative services, and projects of non-formal education

could be coordinated.

Akhter Hameed Khan described the rationale for

the creation of the center as follows:

The Academy's first concern was to improve the

quality of rural administration.

In this field

the Empire had left a poor legacy.

The Imperial

symbol was the Police Station.

We erected a new

symbol - the Thana Training and Development

Center.

Offices of the so-called nation build-

ing departments - Agriculture, Animal Husbandry,

Fishery, Health, Education, etc. - were housed

together in a spacious new building.

For inten-

sive planning and coordination a local government

council was created at the thana level and also

located at the Center.

Chairmen of the next

lower tier, the Union Councils, and the Depart-

mental officers were the constituent members.

The assumptions were that the people should be

mobilized through their elected leaders; that the

officers should coordinate departmental activities

with each other as well as with the councillors.

The Thana Center was to be the focus not only of

planning and coordination but also of training.

Special meeting halls and classrooms were added

and officers were encouraged to teach.

The Thana

Center assumed that a good job requires a good

tool.17

_At about the time that the Thana Training and DevelOpment Center

really started functioning, the academy staff itself removed to a new

and far more spacious campus at Kotbari, five miles away from the

operational center at Abhoy Ashram.

The visual contrast between the

two centers was striking:

many of the one story buildings at the old

Ghandian retreat of Abhoy.Ashram were walled in the village manner

with woven bamboo, whereas the multi-storied concrete structures at

Kotbari had been designed, under Ford Foundation contract, by the Greek

city planner Constantine Doxiadis.

The social worlds of the two

26

campuses were also distinct.

The academy at Kotbari was the nexus of

contact with government officers from Dacca and from other districts,

who often flocked to the academy for in-service training in rural

development.

To the Abhoy Ashram campus came village cooperative

managers dressed in shirt and wrap around luggi, bearded Muslim Imégs,

and shy rural women who might seldom have ventured from their houses

before.

Although the Abhoy Ashram center'appeared the more indigenous of

the two institutions, the ideas behind the pilot projects which were

headquartered there were of cosmopolitan origin.

The village coopera-

tives, for example, were inspired by those which Friedrick Wilhelm

Raiffeisen had organized in Germany in the mid nineteenth century.

Raiffeisen had emphasized the need for a trade union type of joint

action by the small German farmers, who had been grossly exploited

by moneylenders and merchants.

To an extent which he was reluctant to admit while he served the

conservative government of East Pakistan, Akhter Hameed Khan was also

inspired by the example of communist China.

In a speech which he gave

to an International Seminar on Rural Development held at Michigan State

University in 1975, he compared the Comilla peasant cooperatives to the

Chinese mutual aid teams, and the Thana Training and Development Center

to the commune.

He averred that he and others at the academy had

admired the extraordinary expertise of the Chinese communists in organ-

izational and educational concepts and methods.

"The Chinese leaders

were veteran experts of rural organization.

Their idealism and sincerity

had survived fiery ordeals.

Their long experience gave them superb

insights."18

In the same speech on his lessons from China, Akhter Hameed Khan

27

drew attention to one of the deep-seated contradictions in the develop-

ment strategy which was devised at Comilla.

The thrust of the Camilla

movement derived from the hunger of the small farmers for protection

against natural disaster and economic exploitation.

The Comilla

cooperatives were organized as instruments for the small farmers to

use against the elites which had always dominated and squeezed them.

Yet those very elites were in control of the Muslim League (the govern-

ment political party), and an aristocratic psychology characterized

those in control of the central government.

"Since 1947 our rulers

were aristocrats, civil servants and generals who felt more at home

with landlords and industrialists than with peasants and laborers.

Our parties in power were permeated with privilege.

Government by

bribery was as common as government by intimidation.

There was little

integrity and much corruption, few examples of discipline and much

demagoguery.

There was beggary and wasteful pomp instead of self-

reliant austerity."19

If the government was in the hands of men such as those, was.it

possible to forge a reformist phalanx in which the circle officers

were the captains and the small farmers were the men?

Would not the

officers of the civil administration, acting as agents of the ruling

classes, assure the emasculation of any genuinely popular movement?

The analysis is appealingly simple and partially true.

The argument

was sufficiently false, however, as to allow for the great strides

taken in Comilla toward popular solidarity, rural welfare, and

agricultural modernization.

Notes for Chapter 1

The Comilla Experiment

1 Carl H. Gotsch, "Relationships between Technology, Prices,

and Income Distribution in Pakistan's Agriculture:

Some Lessons from

the Green Revolution," In Robert D. Stevens, Hamza Alavi, and Peter J.

Bertocci, eds., Rural Development in Bangladesh and Pakistan, The

University of Hawaii, Honolulu, 1976, p. zéél

2 John M. Cohen, "The Chilalo Agricultural Development Unit as

a Program Intermediary for Foreign Assistance in Ethiopia," AID Spring

 

Review of Small Farmer Credit, Vol. VIII, February 1973, p. 9.

 

3 Arthur F. Raper, et al., Rural Development in Action:

The

Comprehensive Experiment—atComilla, East Pakistan,Cornell University

Press, Ithica, 1970.

Includes a more detailed discussion of the

financing of the academy, and of its relationship to M.S.U.

4 Akhter Hameed Khan, Presidential Address to the East Bengal

 

Teachers Conference, Comilla, 20 November 1952 (Eimeographed).

5 Sudhir Sen, g Richer Harvest:

New Horizons for Develgping

 

Countries, Orbis Books, Maryknoll, New York, 1974, p. 113.

6 Holdcroft, Lane E., "The Rise and Fall of Community Development,"

Department of Agricultural Economics, Michigan State University, 1976,

P- 3-

? Akhter Hameed Khan, "The Essence of the Comilla Program,"

speech delivered at the MIT Conference on Productivity and Innovation

in Agriculture in the Underdeveloped Countries, July, 1964 (mimeo-

graphed).

Akhter Hameed Khan was amused at the anti-communist cold

war ideology which impelled the American community development experts.

He recalls that his American advisor was wont to drive him through the

villages in a jeep: the advisor would wave his arm at the scene and

say, "We have only ten more years before all this explodes."

8 Peter J. Bertocci, Elusive Villages, Social Structure and

memunipy Organization lg Rural East Pakistan, Ph.D. thesis,,Michigan

State University, 1970, p. 150.

9 Ibid., p. 150.

10 Ibid., p. 71.

28

_

29

11 Akhter Hameed Khan, "An Analysis of the East Bengal Village,"

a speech to the U.S.A.I.D. seminar held at Comilla in June, 1963.

The

speech is summarized in Comilla-USAID Conference Report, (mimeographed).

12 Ibid.

13 See, for example, "An Analysis of the Various Classes of the

Chinese Peasantry and their Attitudes Toward Revolution," in Stuart R.

Schram, The Political Thought 9; Mac Tse Tung, Frederick A. Praeger,

New York, 1963, pp. 172-177.

Mao divides the pro-revolutionary Chinese

peasantry into eight economically-defined classes:

the big landlords,

small landlords, peasant landholders, semi-landholders, share croppers,

poor peasants, farm laborers and rural artesans, and elements declasses.

My own rough translation of Mao Tse Tung's classification into the

scheme presented by Akhter Hameed Khan would equate the big landlords

with zamindars, the small landlords with big farmers, and the peasant

landholders

of China with the small farmers of East Pakistan.

For the

rest, Mao's discrimination is much finer than Khan's.

1“ Akhter Hameed Khan, "The Essence of the Comilla Program,"

22 cit, p. 2.

15 Author's lecture notes from Agricultural Economics 865, "Rural

Development Administration," Michigan State University, Winter term,

1975.

16 Akhter Hameed Khan, "Principles of Rural DeveIOpment," speech

sponsored by the Asian Studies Center, Michigan State University,

June 11, 1964 (mimeographed), p. 6.

17 Akhter Hameed Khan, Reflections 23.232 Comilla Rural Develgpment

Projects, Overseas Liaison Committee, American Council on Education,

Washington, 1974, p. 11.

18 Akhter Hameed Khan, "Learning from China, A Pakistani Experience,"

speech delivered at the International Seminar on Rural Development,

Michigan State University, 1975. p. 2 (mimeographed).

19 Ibid., pp. 2, 3.

CHAPTER 2

The Need for Models

Many of us have seen the farmers of these lands.

We have watched while they strained to guide

clumsy wooden plows across meagre fields.

We

have seen them, clad in cotton drawers or breech

clouts, thigh deep in paddy field mud, pain-

stakingly tending young rice plants.

Our arms

and bodies have ached in sympathy as we observed

the laborious efforts to break heavy clods with

primitive hand tools.

Perhaps some of us have

wished we could somehow penetrate the minds that

direct these bent, straining, fragile bodies

through endless steaming days of toil.

But un-

fortunately our efforts at understanding place

us somewhere between guess and conjecture as we

return to our desks - our papers and endless

reports to satisfy those who understand less than

we.

The World Bank and the Rural Poor

The major issue which this book addresses was posed, hardly for

the first time but perhaps more prominently than before, by Robert

McNamara in a speech to the Board of Governors of the World Bank in

Nairobi in 1973.

McNamara said that the most difficult problem in

fashioning a strategy for the world—wide development of small holder

agriculture is the design of an appropriate organizational structure.

He asserted the need, which he called the most critical of all, for

"new forms of rural institutions and organizations that will give as

30

31

much attention to promoting the inherent potential and productivity of

the poor as is generally given to protecting the power of the

privileged....

What is required is the organization of local farm

groups which will service millions of farmers at low cost, and the

creation of intermediate institutions through which governments and

commercial institutions can provide the necessary technical assistance

and financial resources for them."2

A year and a half after McNamara's pathbreaking Nairobi address,

the World Bank published a sector policy paper on rural development.3

The policy paper made mention of several examples of rural deve10pment

projects, such as the Gezeira land settlement scheme in the Sudan, the

Puebla project in Mexico, and the Comilla projects in Bangladesh.

Yet

anyone who reads the rural development policy paper in search of answers

to the organizational problems posed by the bank president will be

drawn to the conclusion that the World Bank staffers who prepared the

paper have only the vaguest clues as to how an acceptable rural develop-

ment organization might be put together.

The Rural Development Sector Policy Paper states that there is a

growing consensus on the need for national planning for rural deve10p-

ment, coordination at the central government level, decentralization at

the regional and local level, and participation by the rural poor.

Beyond these platitudes, the authors of the policy paper offer little

concrete advice.

They note that the minimum package program has

worked in Ethiopia, that the Republic of China has done well with its

Joint Commission on Rural Reconstruction, and that area development

schemes have been successful in places such as Comilla.

Among the few

conclusions which they draw from these examples is a negative assess-

ment of the potential dangers of area development schemes, which they

32

say can unduly concentrate resources and "distort priorities in the

allocation of resources among sectors."4 Plaintive commentary of that

sort achieves little.

The intellectual road to which Robert McNamara

pointed in Nairobi still lies Open and untravelled.

The Deficiency 9: Present Theory

 

The road ahead is not, however, well mapped or even clearly per—

ceivable.

Only a few yards in front of us the verge is blurred by

our essential ignorance of peasant society.

And the road to rural

development appears to divide, with one lane twisting through the

thicket of land reform, another through the jungle of green revolution,

and still another over the rough hills of cooperation.

In the distance

the route is lost in a shifting fog from which flash as disturbing

beacons predictions of doubling world population, oil depletion and

fertilizer shortage, Malthusian starvation, and.aauaclysmic retribution.

We might traverse the road to rural development more expeditiously

if we could rely on some social or economic theory to guide our

stumbling pace, much as the Chinese found guidance, or at least solace,

in the thoughts of Mao Tse Tung.

Unfortunately, the stock of theory on

rural development, and especially on rural development organization, is

meagre.

Perceptive commentary on the subject is extremely rare.

The

intellectual situation is so bad that the late Frederick H. Harbison

of Princeton University remarked that, "...for massive rural development,

the skills are uncertain, the organizational structures are not

developed, the leadership is unidentified, and the techniques for

motivating change are largely unexplored.

Both with respect to theory

and practice, organization for rural transformation is probably the

33

most underdeveloped area in the entire field of economic growth."5

If the subject of organization for rural development is virtually

unexplored territory, the larger topic of rural deve10pment theory is

only slightly better known.

With the exception of a few economists,

theoretically inclined academicians seldom write on the subject.

There is a small literature of ideas that have been transplanted upon

the wasteland from well-watered academic disciplines, but there are no

cases of thriving thought.

As recently as the early nineteen seventies, when the International

Council for Educational Development conducted their World Bank -

sponsored research project on non-formal education and.rumal development,

they found the field as intellectually arid as ever.

"Although much

attention has been given by scholars to evolving general theories of

national development (mainly the economic elements)," they commented,

"surprisingly little has been given to describing and explaining the

nature and process of rural development.

We were obliged, therefore,

to improvise."6

The Cornell Rural Development Study

One of the few enlightened discussions of the genesis of rural

development is contained in Norman T. Uphoff and Milton J. Esman's

summary of the findings of a wide-ranging Cornell University study of

18 cases of rural local government in Asia.

In that summary, Uphoff

and Esman explain their empirical finding that "local organization is

a necessary if not sufficient condition for accelerated rural develop-

ment, especially development which emphasizes improvement in the

productivity and welfare of the majority of rural peOple.

Those cases

34

in which there was more organization reaching down to the local level,

accountable to the local people, and involved with rural development

functions...have accomplished rural development objectives more success-

fully with respect to the available resource base than have those with

less rural organization."7

The studies supervised by the Cornell Rural Development Committee,

which was supported for the research project by the U.S. Agency for

International Development, examined cases of rural local government

from Japan to Egypt, including Bangladesh.

Upon review of the case

study reports, and of the discussions about them which had taken place

among members of the rural development committee, Uphoff and Esman

extracted seven general principles for successful organization for

rural development.

These seven principles comprise the closest analog

of which I am aware to an operational theory of rural development

organization.

To demonstrate the utility of theory in the design of rural

development organizations, I have quoted Uphoff and Esman's seven

principles below, and shown how each applies to the Comilla experience.

Background information which will set the scene for later chapters is

presented in the course of this exposition of the seven principles.

These principles will be amplified, and linked to the larger literature

of organization theory, in subsequent chapters.

Because those chapters

on organizational structure, technology, and so forth contain extensive

reviews of the literature, and in two cases explanations of dynamic

system models as well, the material on the Comilla experience in each

chapter is necessarily compressed to an exclusive focus on the coopera-

tives.

To give the reader a more comprehensive understanding of the

experiment in rural development organization that began in Comilla,

35

the illustrations in this chapter are drawn from two organizations

which flourished there in the nineteen sixties:

the Basic Democracies

system of local self-government, and the Comilla cooperatives.

Many

more rural deve10pment programs were launched in Comilla Kotwali

Thana -- as, for instance, the programs for family planning, tubewell

irrigation, and milk processing -- but I have chosen to concentrate in

this chapter on the two organizational innovations which best exemplify

the seven design principles for rural development which Uphoff and

Esman extracted from their field research.

1.

"Local institutions should have more than one level 2:

organization, probably a two—tier pattern, in which the lower tier

performs functions at the neighborhood or small group level, while the

other undertakes more complex business activities that require relatively

large scale operations."

(Uphoff and Esman, 1974, p. 99).

The multi-tiered approach to administrative organization has been

characteristic of the governments of the Indian sub-continent for

centuries past.

In the sixteenth century the Moghul Emperor Akbar

created the prototype of a territorial hierarchy of administration by

inserting above the existing network of sarkars (which themselves

resemble

the districts of India, Pakistan, and Bangladesh) another

level of provinces.

Each province, or §gp§h, was ruled by a senior

imperial official called a subadar, who was responsible for security.

The subadar's counterpart in the province was the gépgp, who collected

revenue.

The British rulers of Indian perpetuated the Moghul two-tiered

administrative system of provinces and districts.

The two administrative

36

layers which the British added to the system -- the division between

the province and the district, and the subdivisions below the district --

never attained the same bureaucratic status or popular legitimacy as the

provinces and districts which were tracable to Emperor.Akbar.

The

units of the territorial hierarchy of administration which was

fashioned by the British, and retained by the government of Pakistan,

is summarized in Figure 2-1.

Administrative

Number of Units in

Unit

East Pakistan

Principal Officer

Province

Division

District

Subdivision

Thana

Union

1

4

17

54

Governor

Commissioner

Deputy Commissioner

Sub-divisional Officer

413

Civil Officer

4,053

Union Council Chairman

Village

64,000

No legal head

Figure 2-1.

The Territorial Hierarchy of Administration

In the British province of Bengal generally, and in its Muslim

eastern hinterlands especially, the careful balance which Akbar had

struck between revenue collection and other aspects of administration

was upset.

The British rulers of Bengal acquiesced in a pattern of

indirect revenue collection which had been established early in the

eighteenth century by the Moghul diwan of Bengal, Murshid Quli Khan.9

In 1793 Lord Cornwallis, head of the British East India company, enacted

what came to be called the Permanent Settlement.

This arrangement

formalized for the next century and a half the system of collecting land

revenues in Bengal through a class of tax agents known as zamindars.

37

The zamindars assumed some proprietary right to the lands in their tax

estates.

A consequence of this delegation of governmental tax collect-

ing responsibility was that the formal institutions of the British

administrative system, especially at the crucial district collectorate

level, were truncated in Bengal.10

The province of Bengal was

characterized, according to the British rulers themselves, by admini-

strative starvation.

Elliot Tepper has described the administration of

the Muslim districts of British Bengal as "a strong superstructure with

weak underpinnings."10

After Pakistan obtained its independence of Britain, and of India,

the abrogation of the 1793 permanent settlement was widely supported

in East Pakistan.

The pepularity of the land reform measures of

1950-51 was enhanced by the fact that most of the zamindars had been

Hindu, whereas East Pakistan was newly launched as a wing of a Muslim

nation.

The demolition of the structure through which land revenue had

been collected for a century and a half left an administrative void,

nonetheless, since the hierarchies of agents established by each

zamindar had been among the few approximations of a link between the

Bengali villagers and the exalted nexus of governmental power.

In some

areas of the province it seems that the zamindars had at one time per-

formed, albiet at the cost of exploitation, positive governmental

functions such as the coordination of maintenance work on embankments

and drainage canals.

Rehman Sobhan remarks that these functions were

progressively abandoned as the zamindars left their country estates for

a more comfortable city life, leaving the rural infrastructure of

embankments and canals in a state of general decay.11

A signal attempt at filling in the administrative void below the

district level was the multi-tiered system of Basic Democracies,

38

promulgated by Field Marshall Ayub Khan, the Pakistan Chief of State in

1959.

Basic Democracies were Ayub Khan's greatest and fondest inno-

vation.

He conceived of the layered system of councils, directly

elected at the bottom, or union, level and increasingly appointed and

bureaucratically controlled at higher levels, as a mechanism for the

political socialization of the Pakistani populace, for their gradual

education to the values and burdens of democracy and nationhood.

A

period of tutelage by the civil servants who dominated the higher

level councils was anticipated, after which Ayub Khan averred that the

Basic Democracies could become more democratic and less basic.12

From

Ayub Khan's perspective, one of the most valuable functions of the

80,000 basic democrats who were elected as members of union councils

was the legitimation of his own rule: in 1962 and 1965, the basic

democrats formed themselves into an electoral college, in accordance

with the constitution which Ayub Khan had promulgated, and on each

occasion selected him as president of Pakistan.

The form and philosophy of Basic Democracies were more of a

novelty in West than in East Pakistan.

So far as the rural populace

of the eastern wing was concerned, Basic Democracies bore a resemblance

to systems of local government through which they had, in a limited

way, been resolving their affairs for three quarters of a century.

In

the late nineteenth century, which Elliot Tepper describes as "the

earliest, and most paternalistic, period of British experimentation"13

with local government in Bengal, the villages had been grouped into

unions.

The British imperial governors of Bengal sought to devolve

upon committees and boards at the union level some of the governmental

functions which the regular civil service and the zamindars were so

ill-prepared to handle.

39

The first imperial initiative in this direction was a flop.

The

Chaukidari Act of 1871 sought to have union committees tax the villagers

to pay for a staff of chaukidars, or village policemen.

The plan was

abandoned following several violent protests from villagers who per—

ceived the chaukidars and the union taxes as a new form of petty

tyranny.

The Local Self-Government Act of 1885 was more favorably

received, since it gave local councils, in addition to the power of

taxation and responsibility for the maintenance of rural order, a modest

capability for carrying out more positive functions such as the build-

ing of roads.

The three tiered system of councils created by the Local

Self-Government Act prefigured Ayub Khan's Basic Democracies.

Struc-

turally, the major difference was that the intervening council in the

1885 arrangement was at the subdivisional level (the next level below

the district in the British administrative hierarchy, and the lowest

level to which members of the elite Indian Civil Service were posted),

whereas Ayub Khan placed it one step lower down, at the thana level.

Discussions about the powers and composition of the Union Committees,

Sub-divisional Boards, and District Boards which proceeded the 1885

enactment also foreshadowed many of the controversies that were to be

engendered by Basic Democracies.

William Gladstone, leader of the English liberal party in the late

nineteenth century, expressed his opinion that "It is our weakness and

our calamity that we have not been able to give to India the blessings

of free institutions."1u When he returned to the prime ministership

in 1880, Gladstone sent out to India as Governor General his disciple

Lord George Frederick Ripon.

In his history of India, Percival Spear

asserts that from Ripon's four year term of office, from 1880 to 1884,

"dates the beginning of India-wide urban and rural self-government,

40

which still follows in the main the original pattern."15 Ripon advocated

the establishment of a network of union and subdivisional councils, at

least three fourths of the members of which were to be elected.

The

union councils in Ripon's plan were to have strong financial powers.

In defense of his proposal Lord Ripon wrote:

The task of administration is yearly becoming more

onerous....

The cry is everyWhere for increased

establishments.

The universal complaint in all

departments is that of overwork.

Under these cir-

cumstances it becomes imperatively necessary to

look around for some means of relief: and the

Governor-General in Council has no hesitation in

stating his conviction, that the only reasonable

plan open to the government is to induce the people

themselves to undertake, as far as may be, the

management of their own affairs: and to develop,

or create if need be, a capacity for self-help in

respect of all matters that have not, for imperial

reasons, to be retained in the hands of represent-

atives of the Government.1

Before it was officially enacted in 1885, after he had left office,

Ripon's scheme was revised to suit the disposition of the imperial

power elite.

The strongest of the boards established by the Local

Self-Government Act was at the district level, and it was chaired by

the district officer, a member of the Indian Civil Service.

The

district officer was given the power to appoint members of the district

board and of the lower boards.

Financial and programatic powers of the

union committees were severely restricted.

The same issues of bureaucracy versus democracy,of appointment

versus election, and of the relative power of the union, district, and

intermediate councils were to emerge in critiques of Ayub Khan's

system of Basic Democracies.

The system as originally proposed by

Ayub Khan resembled the conservative, bureaucratically—dominated

structure of the 1885 Local Self-Government Act.

Many Bengalis

41

perceived the Basic Democracies system in the form in which it was

first promulgated as a regression, since under the Local Self-Government

Act of 1919 the democratic character and financial powers of the union

committees (then renamed Union Boards) had been strengthened.

Akhter Hameed Khan played something of the role of the liberal

Lord Ripon in urging, throughout the Ayub era and thereafter, the

democratization and vitalization of the lowest levels of the Basic

Democracies structure, the Union Councils and the Thana Councils.

The elaboration of rural administration over the centuries from

Akbar to Ayub Khan has exhibited two themes consistently:

the adherence

to a territorial hierarchy of administrative power, and the descent of

innovation to lower and lower levels of that hierarchy.

For Akbar,

the issue was the relationship between the ggp§h_and the gggkgpg or,

in modern analogy, between the province and the district.

For Lord

Ripon, who was ahead of the times, the issue was the relationship

between the subdivision and the union: his successors, who actually

enacted and implemented the Local Self-Government Act of 1885, put

the emphasis on the relationship between district and subdivision.

Ayub Khan's own thinking did not advance much beyond what the British

imperialists had enacted in 1885.

Akhter Hameed Khan emphasized the

lowest dyad in the official hierarchy, the relationship between thana

and union.

Schematically, this evolutionary descent is summarized in

Figure 2‘2 0

Akbar

Province 

42

   

Division  British Imperial

Administration

Local Self-Government§--——District \

Act Of 1885

\\

\\\

\ \

  

‘N‘Nj7AyutiKhan

///

Lord Ripon

--—-—+-5$ubdivision

// //

:;

\

‘/,/ //’

\\ \\

\

\

\

Thana”

-

i’

 

-

Akhter Hameed Khan

// //

/ /

Figure 2-2.

Levels of Administrative Interest

Even the thana-union dyad, however, does not match the recommenda-

tions of Uphoff and Esman for the proper scale of rural development

organizations.

The primary organizational unit for rural development,

- they suggest, should be the neighborhood, including from 30 to 100

households.

The unions of East Pakistan, each of which included, on

the average, between 12 and 13 thousand people, exceeded this ideal

size by a very large margin.

To attain the mobilization of the rural

populace, organization on a more intimate scale was required.

The

Comilla village level primary cooperatives, which had an average of

34 members during the nineteen sixties, were of a more appropriate

size.

Since each cooperative member typically represented one house-

hold, the primary village cooperatives actually stood near the lower

end of Uphoff and Esman's recommended scale of between 30 and 100

households for the primary organizational unit.

Peter Bertocci, who has examined the relationship of rural society

43

to cooperative organization in Comilla more closely than anyone else,

has asserted:

The Comilla cooperative project has from its

inception taken the "primary village community"

as its grassroots organizational focus.

Indeed,

it seems fair to say that most development

activities in Comilla Kotwali Thana have in

effect bypassed the union level of political

and administrative organization, with the

single important exception of the Rural Public

Works Program.

Most cooperative societies in

Comilla Thana would appear to be coterminous

with the area's many "local villages."

And

where in the larger-sized villages cooperative

societies are formed, it is common for such

communities to have more than one cooperative,

each of these established in parts of the vil-

lage where neighborhood loyalties, as against

those felt toward the larger collectivity, are

strong.

In a word, Comilla cooperatives tend

to overlap with the project area's "natural"

social groupings.1

2.

"Local communities should be linked to higher level decision

centers by multiple channels, both to achieve the benefits of speciali-

zation in communication and to enjoy alternative avenues of influence."

(Uphoff and Esman, 1974, p. 100).

The best evidence for the satisfaction in Comilla Kotwali Thana of

this second of Uphoff and Esman's criteria for rural development

organization is provided in the appendices to their own report.

In

those appendices they describe the major methodological innovation of

the Cornell Rural Development Committee's study as being the development

of quantitative measures for the comparison of configurations of rural

local organizations in different places.

One of these quantitative

measures dealt with the relative importance of different organizational

channels in each of the 18 Asian cases studied.

A standard matrix for

44

each case was devised, with five categories of channels constituting

the columns, and ten rural development functions forming the rows.

The matrix for East Pakistan during the nineteen sixties, which was

among the case studies in the Rural Development Committee's sample,

is shown in Table 2-1.18

The 10 points alloted to each rural development function in the

"Total" column are standard for each of the case study matrices.

Entries

in each row show how these 10 points were judged by the Cornell Rural

Development Committee to have been divided among the five organizational

channels of communication.

For example, nearly all planning and goal

setting in East Pakistan was performed by the state administration,

which was given 9 out of the 10 points for this function.

The sums for

each column (44 for state administration) can be read as the percentage

of rural development functions performed by each channel.

The local

government column in the matrix represents the system of Basic

Democracies, which operated throughout Pakistan in this era.

The

column for associational organizations represents the otiose state-

sponsored Union Multipurpose Cooperative system, plus miscellaneous

farmers' organizations.

political organizations, it might be noted,

‘were just about impotent at the local level in East Pakistan.

It is

important to recognize that the entries in the matrix indicate the

proportional distribution of rural development functions in the place

concerned (East Pakistan in this instance), and do not indicate the

level or amount of activity that took place.

Because Comilla Kotwali Thana was such an exception in the area

of rural development organization in East Pakistan in the nineteen

sixties, the Cornell Rural DeveloPment Committee prepared a special

matrix for the thana.

(See Table 2-2.)

45

Table 2-1.

Rural Development Channels in East Pakistan

Local

Organizations

Local

Assoc.

State

Private

Pol.

GOVt e

Org ' S 0

Admin 0

S eCtOI'

Org ' é;

Tetal

Plannipgpand

 

Goal Setting

 

Resource

Mobilization

 

Provision.gf_

Services

Water

Fertilizer

Credit

Extension

Marketing

Igtegration 9:

 

Services

Control Over

 

Administration

 

Claim Makipg

 

1

O

1

1

0

0

0

0

3

1

1

0

1

8

3

2

1

2

O

1

0

1

9

3

O

6

19

22

6

3

2

8

0

6

5

2

O

5

7

0

1O

0

3

4

0

0

0

0

0

0

0

0

O

1

2

10

10

50

10

10

10

10

10

10

10

10

7

11

44

35

3

100

18

46

Table 2-2.

Rural Development Channels in Comilla Kotwali Thana

Local

Orggnizations

Local

Assoc.

State

Private

Pol.

Govt.

Org's.

Admin.

Sector

Org's.

Total

Planning‘ggg

Gog;_Setting

 

2

Resource

Mobilization

 

1

'

Provision 9:

Services

Water

Fertilizer

Credit

Extension

Marketing

Integration 2;

 

Services

Control Ova;

 

Administration

 

Claim Making

4

2

1

0

1

0

3

1

1

3

2

21

4

3

a

8

2

5

2

3

5

3

9

4

2

2

1

0

2

5

2

0

4

16

0

4

L:

0

8

O

1

3

O

O

0

0

O

o

0

0

O

1

1

10

10

50

10

10

10

10

10

10

10

10

12

36

26

24

2

100

47

The major difference in the interpretation of the Kotwali Thana

matrix is that the associational organizations column includes the

Kotwali Thana Central Cooperative Association and its legions of local

member cooperatives, an institutional innovation that was not present

in most of the remainder of East Pakistan during the nineteen sixties.

All in all, the channels of local organizations (local governments

plus associational organizations) performed 48 percent of rural develop-

ment functions in Comilla Kotwali Thana, as compared with 18 percent in

East Pakistan as a whole.

The proportionate influence of the state

administration was cut down sharply in Kotwali Thana.

With the con-

tinuing exception of political organizations, the distribution of

functions among organizational channels was more even in Kotwali Thana

than in East Pakistan, showing that the thana did indeed meet Uphoff

and Esman's second criterion for successful rural development organiza-

tion.

3.

"Though specialization may be indicated for some activities,

local institutions should ordinarily be vested with multiple functippg

in order to achieve the benefits of scale and to facilitate the local

integration of services....

The organizational design for successful

rural development is likely to be a limited number of local institutions

which are multi-level and multifunctional and which communicate through

multiple channels."

(Uphoff and Esman, 1974, pp. 100, 101).

In Uphoff and Esman's conceptual framework, the important functions

of rural development organizations are those which make up the rows of

the matrices displayed in Tables 2-1 and 2-2, i.e. planning and goal

setting, resource mobilization, the provision of services, the

48

integration of services, control over administration, and claim making.

A complete explanation of how these six types of functions were per-

formed in Comilla Kotwali Thana by the basic democracies and by the

agricultural cooperatives would be a voluminous undertaking.

Rather

than catalog all of the.functions of these organizations in accordance

with Uphoff and Esman's typology, I would call the reader's attention

once more to the entries in the first two columns of Table 2-2.

Read

vertically, the distribution in both columns are fairly even, indicating

that just about the whole gamut of rural development functions was

performed by local government organizations and local associational

organizations in Comilla Kotwali Thana during the nineteen sixties.

The major examples of these organizations in the thana at the time were

the basic democracies and the cooperatives.

The rural cooperatives at the village and thana level were

especially distinctive for the diversity of their activities.

Among

their more salient activities were the disbursal and collection of

loans; the installation of irrigation tubewells and the management of

water distribution: projects in literacy, family planning, and women's

uplift: agricultural demonstration and extension: processing and market-

ing of milk, rice, and other crops: and massive, unending training of

all levels of personnel.

These wide-ranging activities were fused at

the level of the primary cooperative, which served as a single conduit

into the village, but were divided out into specialized sections of the

thana central cooperative association.

Managerial talent, which was

perhaps one of the scarcest attributes of the Comilla villagers, was

thereby husbanded by the primary cooperatives.

The thana cooperative

association, which employed professionals who were not of the village

mold, took full advantage of the efficiencies ensuing from organizational

49

specialization.

Like a modern corporation set up on functional

lines, the thana cooPerative association had organizational units for

the creamery, for the cold storage plant, for tubewell sinking, for

tractor driving and repair, and for other such functions.

A further issue that deserves comment is why so many activities

should have been undertaken by the cooperative association, and pressed

upon the villagers by means of the primary cooperatives.

At the thana

level also there were limits to management capabilities, and it might

be argued that the cooperatives would have done better to concentrate

on a few key projects -- such as tubewells, extension and credit --

rather than dissipating financial resources and executive energies in

so many directions.

The germ of a rejoinder to such a criticism is

to be found in Arthur Raper's book about the comprehensive rural develop-

ment experiment at Comilla.

In his final chapter of reflections Raper

observes that "the quality of intersupportiveness permeates the Comilla

program as one activity leads to or strengthens another, and so each

enhances the other."19

Lane Holdcroft, an officer of the Agency for International Develop—

ment who took a sabbatical at M.S.U., has said that mutually reinforcing

components or activities are characteristic of successful rural develop-

ment programs around the world.

He compared this characteristic comple-

mentarity to the interaction of factors of production in an economic

production function.

When such complementarities are present, "the

application of two or more improved inputs, e.g. seed, fertilizer, and

water, result in more total output than the sum of the increases which

can be attributed to each individual input."2O

Mathematically, such

interaction effects can be expressed as the multiplication of the factors

of production one by another, causing the production function to be

non-linear.

50

One of the complementary input relationships which Akhter Hameed

Khan has come to believe in ever more firmly is that between land

improvement and technological innovations in agriculture.

For the

peasants, he asserts, land protection and land improvement is the

primary issue, and for them all else must follow after.

In a speech

which he gave in 1975 about lessons from China and their influence on

the Comilla rural development projects, he said:

During our investigations we had discovered that

the losses caused by flood and drought were

immense and frequent.

The villagers themselves

attributed their destitution primarily to these

losses.

Their first demands were always for

roads, drains, embankments, canals and wells.

Our agronomists, on the other hand, emphasized

the use of "miracle" seeds and chemical fertili-

zers.

Through intimate knowledge of our ecology,

we comprehended why the Chinese were mobilizing

millions of men, year after year, for land im-

provement, a priority which we had neglected.

Learning from them we organized a rural works

program which simultaneously built the infra-

structure, provided gainful employment to other-

wise idle agricultural laborers, and prepared

the ground for a stable and progressive agricul-

21

ture (including the miracle seeds and fertilizers).

This expost facto interpretation of the intellectual origins of

 

the Rural Public Works program turns the tables of history quite

violently around.

While it seems to be true that the Chinese example

influenced the thinking of Akhter Hameed Khan as long ago as the

early nineteen sixties, when the Comilla rural development projects

were just getting started, American advisors have customarily been

given much of the credit for initiating the Rural works Program.

In October of 1960 Dr. Richard V. Gilbert, the chief of the

Harvard Advisory Group to the Planning Commission of Pakistan, took a

51

boat ride with Akhter Hameed Khan through flood-stricken regions of

Comilla Kotwali Thana.

Because of breaches in poorly maintained levees

and the ensiltation of drainage canals, flooding in a 25 square mile

area south of Comilla town had for four years damaged the spring rice

crop, driving the villagers there to utter destitution.

Akhter Hameed

Khan explained the need for repairing the embankments and canals.

Gilbert suggested that a leaf might be taken from Franklin D. Roosevelt's

book, in that a dry season public works program could be organized

similar to the Works Progress Administration of the depression years in

the United States.

As Gilbert put the case somewhat later:

"Your

villages are in a state of depression.

There is a lack of employment,

and there is a lack of local capital works.

Why is it not possible, in

the slack work season, to put your idle people to work building roads,

drainage canals, and an irrigation system - the very things the

villagers need so badly?"22

Gilbert took the idea of a rural works program for East Pakistan

back to the headquarters of the planning commission, in Karachi.

His

proposal was to finance the works program with American surplus food

commodities, donated under United States Public Law 480, which could be

given to village laborers directly as wages, or which could be sold

for rupees to be paid to the workers.

To Gilbert's consternation, the

proposal elicited a barrage of opposition.

It was asserted, for

example, that a massive works program would either overburden the

civil service, or, if its implementation were entrusted to union

councils, would lead to rampant corruption in the rural areas.

Hearing

such criticism, Gilbert fulminated, "A works program, in the context

of the labor surplus which, in the East wing, as in India, is probably

of the order of 25 percent, is so obvious a necessity that I find

52

myself getting furious at the array of argument, pseudoanalytical or

practical, that is thrown up."23

Gilbert soon swung the Planning

Commission, the cabinet, and the president to his side, however, and

encouraged negotiations with the United States government for an

expanded program of P.L. 480 assistance.

In August of 1961 an agree-

ment was completed for $621 million of U.S. surplus agricultural

commodities, to be used, in part, for a works program in the rural areas.

In September of 1961 the Planning Commission formally inquired

whether'a pilot rural works program might be undertaken in Comilla

Kotwali Thana.

Akhter Hameed Khan replied affirmatively, and brought

the idea before the Kotwali Thana council in October.

Proposals for

projects were solicited from the union councils, and were speedily

consolidated into a rural works program for the thana.

As it happened,

delays in the implementation of the U.S.-Pakistan P.L. 480 agreement

made it impossible to finance the pilot works plan from the intended

source.24 Through the intervention of Richard Patten, an East Pakistan

advisor associated with the Harvard Advisory Group, financing was

secured from the minor irrigation project fund of the provincial

Agriculture Department.

In Kotwali Thana, 44 project committees, each

chaired by a union council member resident in the vicinity of the

particular project, carried out the pilot works program during the dry

season in the early months of 1962.

The results of even such a hastily

assembled pilot scheme were

impressive.

Eight million cubic feet of earth were moved by hand

labor.

In keeping with a long-standing concern of the Comilla villagers

for flood control and drainage, 34 miles of drainage canals were

cleared and 14 miles of embankments were constructed.

The three-month

spurt of land-improvement activity in the pilot works program mobilized

53

46,000 man days of labor, at a total cost of 202,000 rupees, or’about

42,000 dollars.

The Pakistan government and the American aid mission responded to

these results enthusiastically, some have said too enthusiastically,

and replicated the rural works program throughout East Pakistan in

1962—63, and throughout the country the following year.

Annual works

programs implemented through the Basic Democracies structure became

a characteristic feature of rural life in Pakistan for the remainder of

the Ayub Khan era.

In modified form, the rural works program was

perpetuated by Ayub Khan's successors, and by the government of

Bangladesh.

Appraisals of the rural works program in East Pakistan during the

nineteen sixties are quite varied.

Edward S. Mason entered one of

the strongest encomiums, calling the program "probably the most success-

ful attempt to use effectively the services of underemployed agricultural

workers that has been undertaken in any less developed country."25

An astringent judgement has been rendered by Walter P. Falcon, who

contends that the program was ineffective in producing rural employ-

ment except at the margin, since only 3.5 percent of the estimated

rural unemployed were given jobs.26

A criticism of the works program which was at first stridently

expressed -- that the rural councils were technically incapable of

completing such a vast enterprise —- by and large subsided after 1963.

On the basis of his field research in eight dispersed thanas in East

Pakistan, Thomas concluded that works program rural roads were con-

structed in a logical pattern, coinciding with economic and administra-

tive needs, and that 95 percent of them were maintained.

About 10

percent of the works program bridges, in Thomas's estimation, were

54

inadequate for their intended carrying capacity.

Undeniably a

tremendous quantity of earth was moved by hand labor, mostly for the

purpose of building embankments to elevate dirt roads above the annual

flood level.

In 1963 East Pakistan had but 3,411 miles of useable

roads.

Over a hundred thousand miles of roads were built or repaired

in the province between 1963 and 1967.27

A commoner's assessment of the impact of the new roads has been

recorded by Mahamoodur Rahman,of the faculty of the Academy for Rural

Development in Comilla.

In 1966 he interviewed a cross section of East

Pakistan citizens to ascertain their views of the works program.

A group

of?

rickshaw pullers told him that rickshaws had increased in number

in recent years, and "were replacing the traditional time consuming

and laborious palanquins which were used until recently for carrying

women from one place to another.

Outsiders can now easily reach the

remote interior areas of the villages....

Some rickshaw pullers said

that their incomes had also risen due to increased use of their rick-

shaws by the people.

They added that the building of roads had brought

about great changes in the villages."28

One of the criticisms of the Rural Works Program which has stood

up better than many others is that it was at bottom an American give-

away program, and did not require the commitment of indigenous resources.

The works program funds which flowed to the union councils, thana

councils, and district councils of the Basic Democracies structure

were American, and not Pakistani, in origin.

The works program did not;

require any reallocation of indigenous tax revenues away from the

departments and personnel of the established administration.

Nor did

it evoke a financial commitment from the rural people who were its

beneficiaries.

Tax assessments and collections by the union councils

55

did not increase during the years of the works program.

Despite

administrative directives placing the responsibility for maintenance

of works projects upon the union councils and their local resources,

maintenance was typically arranged through new works program alloca-

tion.

In a conversation with the author in 1977. Akhter Hameed Khan

observed that in East Pakistan only the people were mobilized for the

reconstruction of the land, whereas in China land improvement schemes

carried out by the commoners had mobilized both people and resources.

Even if it is granted that the union and thana councils mobilized

no more than people and dirt for the Rural Public Works Program, that

still stands as a considerable endeavor.

I find it hard to understand;

therefore, why the functional matrix in Table 2-1 shows zero for the

mobilization of resources by local governments in East Pakistan during

the nineteen sixties.

Even in the Cornell committee's matrix for Comilla

Kotwali Thana, shown as Table 2-2 above, local governments are credited

with only one out of 10 points for resource mobilization.

In the

province of East Pakistan and in the special case of Comilla Kotwali

Thana, local governments should have been given more recognition for

their accomplishments.

4.

"Politics, the competition among groups for influence and

resources, must be accepted as inevitable and legitimate in rural local

organization....

It is not required that all benefit equally -— nowhere

is this achieved.

What must be avoided is the politics of exclusion or

monopoly, where practices of patronage out certain persons or groups

off from the benefits of government."

(Uphoff and Esman, 1974, p. 101).

During the nineteen sixties, which were the golden years of

56

worldwide admiration for the Comilla projects, little notice was given

to the political competition underway in the union councils and village

cooperatives.

This inattention was to a large degree a reflection of

the anti-political attitude of the Ayub Khan regime, in which it was

considered proper to focus on overcoming the administrative problems

and resolving the policy issues involved in economic development, but

not to discuss the interplay of social classes and political factions.

In an occasional address, such as the 1963 speech to the U.S.

agency for International Development conference referred to in the last

chapter, Akhter Hameed Khan explained his vision of class conflict in

the villages of East Pakistan.

But his vision was seldom followed up

in that era by empirical studies of the patterns of influence of

farmers owning differing amounts of land.

This is not to say that data

on the landholdings of the leaders of the village cooperatives and

union councils was not collected: but the information, once reported,

was usually passed over without incisive commentary.

An academy-

sponsored evaluation of the rural public works program in East Pakistan

during 1962-63, for example, found that five sixths of the rural

works project committee chairmen and a little less than three quarters

of the project committee secretaries owned five acres of land, or more.

The academy faculty who wrote the evaluation report noted that "the

amount of land owned ranges up to more than 35 acres, indicating a

high correlation between land ownership and rural leadership."29 And

that was all they said on the matter.

Subsequent reviews, both by Bengalis and by Western scholars, have

been less evasive and much more critical.

Urban residents and

intellectuals in East Pakistan were never fond of the system of Basic

Democracies, since they correctly divined that Ayub Khan had engineered-

57

the system so as to reduce their share of national political power.

It is not surprising, therefore, to find that a study of Basic

Democracies originating in the Bureau of Economic Research of the

University of Dacca presented a scathing characterization of the union

councillors as venal representatives of a corrupt rural oligarchy.

Yet the study, supervised and written up by Rehman Sobhan, cannot be

easily dismissed.

The empirical evidence about the rural works program

which he amassed from interviews with 1,819 persons in 114 randomly

chosen unions in East Pakistan in 1964-65 is compelling.

Sobhan found,

for example, that 64 percent of his respondents thought the union council

members to be corrupt, and 63 percent thought that the union council

chairmen misused funds from the public works program.30 The slant of

his discussion is intriguing; he devotes much of one chapter to a

description of Akhter Hameed Khan's 1963 address to the U.S.A.I.D. con-

ference in Comilla, and contends that the Basic Democracies, far from

liberating the small farmers, increased the dependence of the small

farmers on the surplus farmers who dominated the union councils.

In

his chapter on "The Politics of the Works Programme" Sobhan remarked:

In the perhaps genuine belief that they were

building up the infrastructure of the villages,

the government poured money not into the hands

of the villagers but into the hands of the 'sur-

plus' farmers in their role as union council

members.

This served a mutually beneficial

purpose.

It bound the B.D.'s* more closely to

the government both in general and more specifi-

cally in the machinery of local government.

At

the same time it greatly strengthened the hands

of this class at the village level.

It made

them into direct dispensers of financial patron-

age....

Now not only were they the moneylenders

and land leasers of the village but also the

dispensers of jobs and fringe benefits.

Here it

* Basic Democrats, another term for Union Council members.

58

is important to note that their control over

the landless or land poor labourer in particular

must have increased even more than over the

generality of villagers.

It was this class of

villager who was most dependent on the 'surplus'

farmer for a bit of land to farm or for imple-

ments or for casual jobs during the harvest.

Now his landlord and creditor also become the

arbiter of whether he would work and hence eat

in the winter months.31

Whether Sobhan's sweeping characterizations should be accepted

;5 toto is open to question.

He has the biases of an armchair socialist

who relished a theoretical affinity with the rural proletariat, and who

deeply resented the Ayub Khan regime.

He tended to play up the dero-

gatory findings of his field research, and neglected to remark upon

those findings which were complimentary to the Basic Democracies.

He

found, for example, that 72 percent of his interviewees described union

council members who were reelected in 1964 as honest and sincere, yet

that information appears only in a table and not in the text of his

book.

His conclusions contradict, moreover, the assertions of many

enthusiasts for Basic Democracies and the rural works program.

John Thomas, in a monograph which he wrote for the Development Advisory

Service of Harvard University, commented that the works program "has

helped to create a new, modernizing leadership for the rural areas,

and it has provided the framework for participation in the governmental

process.

"32

Between the analyses of rural politics proposed by Sobhan and

Thomas, my own reluctant inclination is to place greater credence on

Sobhan's, since events subsequent to 1968, when both authors,published

their research, have tended to bear out the characterization of the

union councillors during the Ayub era, and especially of the union

council chairmen, as members and agents of the rural economic elite.

59

The implications of the events of late 1968 and early 1969, when a

prolonged series of riots and demonstrations culminated in Ayub Khan's

resignation, have been summarized by Akhter Hameed Khan in cutting

terms:

The Ayub model of rural local government was a

failure:

it did not mobilize the rural people,

and it alienated the urban people.

The result

was a revolt of the urban people who with the

help of the army overthrew Ayub Khan.

There

was not a ripple of protest in the rural areas,

because the union councils had been captured

anyway by the landlord class, who in any event

were in league with the urban economic elite.33

Viewing the Basic Democracies experiment in retrospect, Akhter

Hameed Khan noted in 1975 that local councils are no longer’advocated

with such enthusiasm by rural development planners, since it has become

apparent that they promote neither mobilization nor stability in rural

areas.34

Rehman Sobhan concluded his acidulous book with an affirmation of

hope in village cooperatives of the Comilla variety as the potential

gmardians of the interests of the small farmers of East Pakistan.

He

conceded that the cooperatives neglected those at the bottom of the

economic scale, the landless laborers, but he contended that with some

reorientation, perhaps including land reform, cooperatives of the

Comilla stripe might circumvent the oligarchical pitfalls of the

union councils.

This line of thinking accords very well with what

Akhter Hameed Khan has argued for many years:

that the village

cooperatives should serve as a shield, or even as a trade union, for the

small farmers to protect them against exploitation by the large farmer'

moneylenders and by the traders.

And it is quite possibly true, as

Akhter Hameed Khan has further argued, that the village c00peratiwes

60

could serve as the training ground for a new corps of rural leaders

who would eventually unseat the oligarchs on the union councils: such

seemed to be the trend, at least, in the Union Council elections in

Comilla Kotwali Thana in 1964.

Unfortunately for those utopians who foresaw the salvation of

rural society as emerging from village cooperatives, the same critical

attack which Rehman Sobhan made against the union councils has been

made against the Comilla village cooperatives by Harry W. Blairy a

political scientist from Bucknell University.

The Cornell Rural

Development Committee commissioned Blair to write one of their 18 case

study reports on the topic of rural local government in Bangladesh.

He visited Bangladesh in the summer of 1973 to do field research, and

read widely on the many institutional approaches to rural development

which have been attempted in the country.

Since Bangladesh was at the

time of his visit still reeling from the war for independence from

Pakistan, most of Blair's material concerns the more stable decade of

the nineteen sixties, and especially the agricultural cooperatives

which then took shape in Comilla Kotwali Thana.

Blair's thesis is that many of the primary cooperatives

in the

thana were corrupted from about 1967 onwards when surplus farmers with

landholdings of five acres or more muscled their way into the leader-

ship of those cooperatives, and appropriated to themselves a major

share of the cooperative loans.

Blair concludes that after a promising

start in the direction of assisting small farmers, the Comilla coopera-

tives went the way of almost every other experiment in institution-

building for rural development in South Asia:

toward local control by

elite interests.35

I shall have more to say about the details of Blair's critique in

61

later chapters.

Suffice it to say for now that he has surfaced the

question of whether the politics of the Comilla cooperatives remained

open to and protective of the small farmers who comprised the largest

bloc of the rural population of Comilla Kotwali Thana.

He has

thereby cast doubt on whether the fourth of Uphoff and Esman's criteria

for successful rural development organization was fulfilled by the

Comilla cooperatives.

5.

"Sanctions to control the acts of leaders of local organizations

should be applied from above and from below to get best performance."

(Uphoff and Esman, 1974, p. 102).

The most significant sanction in both systems, Basic Democracies

and the Comilla type cooperatives, was constituent pressure, which was

encouraged by requirements for open meetings, publication of financial

records, and periodic elections.

Thomas reports, for instance, that

each project undertaken in the rural works program was expected to be

discussed beforehand in public meetings, that works program allocations

to the union councils were supposed to be made public knowledge, and

that signboards were to be put up at the site of each project to

identify the scope of work, the amount to be spent, and the project

committee chairman.

He contends that these requirements, with the

exception of the signboards, were generally fulfilled.

His field

research in a sample of eight thanas showed that 100 percent of his

village respondents could identify at least one rural works project in

their area, and 87.5 percent could identify four or more.36 'He asked

a random sample of 227 villagers to indicate whether they thought that

1) demonstrated ability to serve the community, 2) age. 3) money, or

62

4) important family connections was the most important prerequisite

for election to a union council.

Demonstrated ability to serve the

community garnered 45 percent of the responses.

Money was the runner

up, with 33 percent.37

Although publicity, constituent pressure for good performance,

and the possibility of electoral defeat did control the union councils

to a degree, these were about the only sanctions that were applied.

There was an official requirement that rural works project committees

maintain an accounts book.muster roll, measurements register'and

daily attendance register, and that these records be audited by the

Circle Officer at thana headquarters.

An evaluation survey of the

operation of the rural works program in 62 unions of 19 thanas, con-

ducted by the Comilla academy in 1964, revealed that the circle officers

had found that these registers were improperly maintained in nine tenths

of the cases.

For the most part the circle officers threw up their

hands at these irregularities, as did the academy evaluation team.

The

team excused the lapse in record keeping on the grounds that "(1) village

people, who are completely ignorant about keeping registers, have been

required for the first time to keep six different registers properly:

and (2) no necessary correlation between improper maintenance of

registers and misappropriation of funds can be established."38

A major cause of the laxity of administration in the rural works

program was the breakneck pace of its expansion.

After the completion

of the pilot works program in Comilla Kotwali Thana in 1961-62, the

Academy for Rural Development recommended that a works program be

attempted the following year in the 54 thanas adjacent to sub~divisional

headquarters, so that administrative supervision and assistance with

technical engineering problems could be assured.

The academy was also

63

concerned that union councilors and thana officers receive sufficient

training in the recommended procedures for project planning, the

formation of project committees, the maintenance of records, and so

forth.

All of the union councilors and thana officers who had been

involved in the pilot works program in Comilla were carefully trained

at the academy.

Project plans for the pilot project had been checked

out for technical feasibility and integrated into a thana-wide scheme.

As the result of such precautions, few irregularities of works con-

struction or of record keeping had been detected in the pilot program.

All in all, the deliberate expansion schedule for the rural works

program which was proposed by the academy projected that the entire

province of East Pakistan would be included in the program after eight

years.

The decision taken by the East Pakistan government, instead, was

to spread the rural works program across the whole province in one

year.39

Training of the participants, under such circumstances, was

perfunctory.

A crash three day orientation was arranged for 295

circle officers, while the circle officers from the 54 thanas included

in the academy's earlier planning were treated to a two week course at

Comilla.

Sobhan comments sarcastically:

The union council members, for their education,

had to be content with invitations to Dacca and

Rajshahi to the chairmen and three members from

each union to hear a public address by the

President and the Director of the academy.

We

are not told how these inspiring addresses were

translated into the know-how necessary to carry

out the programme in the 4053 unions of East

Pakistan.“0

The replication of the Comilla cooperative experiment, by contrast,

proceeded at a pace which was much more restrained.

Comilla-type

64

cooperatives were organized in three new thanas outside of Comilla

district in 1963-64, and seven more thanas within Comilla district

established cooperatives in 1965-66.

The balance of 13 thanas in

Comilla district organized the two-tiered cooperatives three years

later, in 1968-69.

Only at the very end of the decade of the sixties,

when the government of East Pakistan was itself on the verge of

extinction, was a plan promulgated for spreading Comilla-type coopera-

.tives to all the thanas of the province under the auspices of’a scheme

called the Integrated Rural Development Program, which was subsequently

picked up by the government of Bangladesh.

The most important reason for the gingerly pace of the expansion

of the Comilla cooperatives during the nineteen sixties was an alter-

nating hesitancy on the part of Akhter Hameed Khan and the provincial

government.

On several occasions in the early sixties the academy was

urged to duplicate its cooperatives widely, but Akhter Hameed Khan

resisted because he was not confident that a good model for replication

had yet been developed.

In 1965, after encouraging reports had come

in from the first three additional thanas, and after the organizational

pattern in Comilla Kotwali Thana appeared to have clarified, he began

to press for expansion.

Three months after the next seven thanas in

Comilla district were taken up, war broke out between Pakistan and

India, throwing all planning into disarray.

When the country settled

down again in 1966, the governor of East Pakistan, Abdul Monem Khan,

began to heed conservatives in the Muslim League who were hostile to

the Academy for Rural Development and its organizational endeavors

among the poor.

The requirements of recruiting, training, and supervising the

project directors for the new Thana Central COCperative Associations

65

(TCCA's) were also a cause

for caution in the expansion program.

For

the cooperatives, with which he identified far more closely and person—

ally than with Basic Democracies or the works program, Akhter Hameed

Khan was determined to avoid the fiasco of lassitude and corruption

which resulted in India when the community development program was

spread at maximum speed throughout the country in the nineteen fifties.

Young and relatively idealistic officers from the East Pakistan Civil

Service were chosen for the TCCA project director posts.

Each was

trained for six months in Comilla, in classroom sessions at the academy

and through observation of the many field activities of the coopera-

tives.

Once the TCCA's were established, their key staff assembled

monthly in Comilla to discuss problems and progress.

Beyond the

competence of the TCCA project directors, their probity was a matter of

real concern.

Because of the powers that were concentrated in the Thana

Central Cooperative Associations, the Comilla-type c00peratives were

very vulnerable to corruption at the upper echelon.

The two officers

who bore the major responsibility for building up the Kotwali Thana

Central Cooperative Association, Zakir Hussain and Shamsul Huq, were

tested and incorruptable veterans, Zakir Hussain from the police service

and Shamsul Huq from the provincial cooperative department.

But men

of their ilk were exceedingly rare.

It would be all too easy, Akhter

Hameed Khan knew, for a project director to succumb to the blandish-

ments of politically or economically powerful men in their thana.

To

guard against project directors being influenced unduly by relatives

or old acquaintances, the directors were never sent to their home

thana.

Although the control of the project directors themselves was a

precarious business, the control mechanisms and sanctions employed by

66

TCCA's over the primary village cooperatives were explicit.

Accurate

and reasonably up-to-date accounts for each village cooperative

society were maintained by staffs of village accountants paid by the

central association.

These accounts were annually, though typically

rather tardily, audited by officers of the cooperative department of

the provincial government.

Village cooperatives could be denied loans

if they were judged to be too far in arrears on loan repayment, or

incompetent in their management.

The TCCA could initiate legal proceed-

ings to seize the land of individual members of the village cooperatives

who were too deeply in default.

Because of tight controls such as

these, the village cooperatives in all the experimental thanas established

a reputation for being less corrupt than the union councils.

The way of the cooperatives, as Akhter Hameed Khan stressed many

times, was a hard way requiring group discipline and personal sacrifice

on the part of the villagers, and the highest standards of integrity on

the part of the project officers.

The attitude of the government of

Pakistan toward the union councils, by contrast, was strongly colored

by Ayub Khan's desire to secure a base of political support among the

council members, the basic democrats who comprised the electoral college

for the presidency and for the legislatures.

Many a Pakistani citizen

has contended that Ayub Khan used the American P.L. 480 funds which

financed the rural works program to buy his way to reelection in

1965.

In actuality, the Pearsonian correlation between the distribution

of rural works expenditures and of votes for Ayub Khan was statistically

insignificant, but as Harry W. Blair points out, the important fact

was that the regime believed in the political efficacy of rural works,

and acted accordingly.1+1

In such an atmosphere, discipline of corrupt

union council members was hardly an urgent concern.

67

6.

Decentpglization of operating decisions within a system of

centrally determined policies provides the best way of combining

authority with the most relevant information and most informed judge-

ment."

(Uphoff and Esman, 1974, p. 102).

Both Basic Democracies and the Comilla-type cooperatives score well

on this criterion.

Integral to both systems was a real devolution of

responsibilities to lower, more local levels of government and society.

The reforms which gave rise to Basic Democracies and to the new cooper-

atives were all the more striking in the Pakistani context because the

government which formed a backdrop for these experiments was centralized

and elitist, and its connections with the rural populace were poorly

developed.

Although Union Boards had been in existence in one form or another

for three quarters of a century before the launching of Basic Democra-

cies, their fiscal and administrative powers had been slim.

The

imperial administration made it clear to the Union Boards that their

major responsibility was the maintenance of a local force of

Chaukidars, or village policemen.

As a consequence, 60 percent of the

taxes gathered by the Union Boards were spent for the salaries of these

rural police.u2

The union councils brought into existence under Basic

Democracies were financially stronger, being empowered to levy 23 kinds

of taxes.

Their officially sanctioned role was expanded, to include

civil administrative functions, judicial functions, and the stimulation

of economic development, in addition to the traditional function of

maintaining the rural police.“3

During the interval between the promulgation of the Basic Democra-

cies order in 1959 and the launching of the province-wide rural works

68

program in 1962-63, illusory expectations about the revamped local

government institutions were widespread.

The departments of the East

Pakistan provincial government generated a flurry of directives for

collaboration with the union councils.

A goodly portion of this

paperwork ended up in the hands of the union councils; in 1960-61 one

union in Comilla thana, Kalirbazar, was the recipient of 206 letters

from various officials and agencies.

Much of this correspondence

suggested that the union councils prepare plans for the amelioration

of whatever condition the government department was charged with

improving.

With naive enthusiasm, many of the union councils forwarded

ambitious schemes to the departmental officials, who, in the time

honored manner of South Asian functionaries, filed them away and took

no action.

A discrepancy between organizational form and financial power

was the major reason why the wheels of the provincial administration

and of the Basic Democracies at first spun so rapidly and so infre-

quently engaged each other.

The union councils on the one hand and the

provincial departments and authorities on the other (as for agriculture,

animal husbandry, water and power development, etc.) were sharply

different kinds of organizations, and they operated at different levels

of the territorial hierarchy.

The thana was the lowest administrative

level to which the departments posted their regular officers, though

the agriculture department employed extension assistants at the union

level as well.

The first response of the provincial government to this,discrepancy

was to distribute works program funds so as to reinforce the financial

position of the union councils and of the district councils, giving

them more bargaining power.

Akhter Hameed Khan protested to

69

Musa Ahmed, the East Pakistan Secretary for Basic Democracies and

Local Government, that the crucial intervening layer of the thana was

being neglected.

The share of works program funds expended by the

thana councils thereafter rose steadily, from a low of 6 percent in

1962-63 to 58 percent in 1966-67.

The union councils share of public

works expenditures held steady throughout most of this same period at

around 20 percent.

District councils and municipal and town committees

were the losers._

The response proposed at Comilla for the organizational discre-

pancy between Basic Democracies and administrative departments was the

Thana Training and Development Center, which was described briefly in

the first chapter.

When the shift of financial and organizational

power to the thana level had been effected in Comilla Kotwali Thana,

and in other thanas which copied the model organization there, the East

Pakistan countryside was treated, quite unbeknownst to its inhabitants,

to what the architect of many of these organizational innovations

considered to be the South Asian administrative equivalent of a

Communist Chinese commune.

It was in the commune and not the produc-

tion team, at the thana level and not the village level, that Akhter

Hameed Khan believed that an organizational engine of rural progress

could properly be constructed.

The predominence of the thana level organizations, both in the

rural works program and in the Comilla-type cooperatives, is an

instance of what Uphoff and Esman refer to as controlled decentraliza-

tion.

They assert that such a pattern is more effective in the Asian

rural context than is complete decentralization.

To restrain villagers

from abusing their access to governmental resources, Uphoff and Esman

contend that "...effective decentralization is usually best constrained

70

by firm policy guidelines and centrally established standards enforced

by regular inspections and other forms of audit and control, while

leaving operations in the hands of local personnel.“+5 They were

referring, it seems. to relationships between central and local

governments, but the same argument can be made within the microcosm of

a thana.

The greatest accomplishments and the most prominent failings of

Basic Democracies revolved around the issue of decentralization.

The

devolution of decision making power and the distribution of relevant

resources to local bodies was in many cases dramatically efficacious.

It is universally agreed, for example, that the earthworks and even

the small concrete bridges put up by rural works project committees

of the union councils costalfraction.of what would have been charged by

private contractors working for the Department of Communication and

Buildings.

In many instances the villagers were demonstrably wiser than

were the salaried civil engineers in agencies such as the Water and

Power Development Authority, which once in Comilla Kotwali Thana built

a drainage canal that brought in more water than it carried.away.

Villagers knew the lay of the local landscape more intimately, and

were seldom caught making such glaring errors.1+6

Despite the parochial-

ism of so many of the decision makers in the works program, the overall

pattern of roads which emerged was sensible, in that it connected

small markets to larger markets, and government headquarters to one

another.

Transportation costs fell, commerce increased, and the rural

sector of East Pakistan was finally involved in the market economy.

Thomas estimated a benefit-cost ratio for the rural works program of

4.0, which is far above what can be claimed for highway or canal

building schemes in developed countries.

71

One of the most widely noted innovations of the Comilla coopera—

tives, to take another example of decentralization, was the design of

a new organizational system for agricultural extension.

The community

development program in India had promoted the model of a village level

worker, or gramasavak, who would go to the village and act as a catalyst

for change, including the improvement of agricultural practices: this

was the model that was imitated in the ill-fated Pakistani program for

Village Agricultural and Industrial Development, or VeAID.

The

American extensionists, who exerted a strong influence on the Agricul-

tural Department, promoted the similar idea of the county

extension

agent.

Neither of these designs for an extension officer who was a

local appendage of a central bureaucracy proved successful in rural

conditions such as those of East Pakistan.

Villagers distrusted the

sanctimonious outsiders, and the extension workers quickly learned

that the way to personal success was to pay more attention to the

office politics of bureaucratic advancement than to the needs of their

clients.

In Comilla, each village cooperative society chose one of their

members, on a rotating basis, to be the village model farmer.

At the

demonstration farm which Japanese agricultural experts had set up at

Abhoy Ashram, the headquarters of the Kotwali Thana Central Cooperative

Association, these model farmers were instructed in better ways of

planting, irrigating, and fertilizing their rice crop, and in subjects

such as pisiculture and the growing of vegetables.

The model farmers

returned to their villages, adopted some of these innovations on their

own farms, and in turn instructed their fellow farmers in the local

cooperative society.

This decentralized method of agricultural

education worked admirably, and was one of the elements of the Comilla

72

experiment which was most widely imitated, as in the Chilalo Agricul-

tural Development Unit in Ethiopia.

The reason that decentralization was a factor in.the failings of

the works program and of Basic Democracies is paradoxical.

Unsuper-

vised decentralization of decision making to the union councils had

several undesirable consequences.

Among them was the tendancy on the

part of the local councils to favor road projects, which were technically

simple (since what had to be done to build a road, for the most part,

was to pile dirt in a long heap) and which favored the large farmers

whose land values rose near the roads and who were involved to a greater

extent in the market economy.

Flood control, drainage and irrigation

projects were often more difficult to execute without technical assist-

ance and without coordination of planning among several union councils,

or even among thana councils: although such projects were frequently

felt by smaller farmers to be more important than the building of roads,

they tended to be neglected by the union councils.

Yet it was the very centralization of power at the thana head-

quarters which led to the discrediting of the works program and of

Basic Democracies in the eyes of the Bengali people.

After the election

of 1965, particularly, the conservative, elitist political coalition

with which President Ayub Khan and Governor Abdul Monem Khan allied

themselves attempted to bring into its fold the circle officers who

chaired the thana councils, and with them the chairmen of the union

councils, who all served concurrently as members of the thana councils.

Once power had been concentrated at the thana level, the regime dis-

covered there a ready made instrument for the corruption and manipula-

tion of rural politics.

The consequence was that the thana councils

and the thana training and development centers are remembered only by

73

the most visionary of commentators as bearing any resemblance to the

Chinese communes.

7.

"Distribution 2: §§sets and income poses a serious political

issue whenever raised, but our studies indicate the importance of this

issue and the need to address it in shaping rural development strategies

and institutions.

More equitable distribution appears to be a

necessary if not sufficient condition for extensive rural development,

though enclave development can occur in its absence."

(Uphoff and

Esman, 1974, p. 103).

The overwhelming impoverishment of the rural populace of what was

East Pakistan, and is now Bangladesh, comes to the fore when the

distribution of assets and income is considered.

In rural East Pakistan

almost everyone was poor.

Once the zamindars had been swept from office

in 1951, there remained no class of rural landlords who were rich by

the standards of, say, the Latin American latifundistas.

According to

the 1960 Census of Agriculture, which sampled all farmers living in 10

percent of the revenue villages in East Pakistan, only 0.4 percent of

landholdings in the province were over 25 acres, and less than five

percent were over ten acres.

The average landholding in 1960 was 3.5

acres.

Many observers have remarked that institutional experiments

such as Basic Democracies and the Comilla cooperatives were substantially

assisted by the resultant egaliterianism of poverty.

On what Thomas

Jefferson would have considered a microscopic scale, it might appear

that East Pakistan was a yeoman's democracy.

As I have sought to explain, the distinctions among farmers owning

less than 10 acres of land were nonetheless strong enough as to set

74

off the surplus farmers (owning, perhaps, 7.5 to 10 acres of land in

1960, and maybe as little as 5 acres by 1970, when productivity per

acre had risen substantially due to the Green Revolution) into a

category of villagers who were domineering, self-serving, conservative,

and somewhat exploitative.

It was upon the rock of these larger

surplus farmers that Rehman Sobhan claimed the Basic Democracies

foundered: Harry W. Blair said that the Comilla cooperatives did

likewise.

If there had been no cooperatives in Comilla Kotwali Thana, how-

ever, it seems likely that the smaller farmers there would have fared

much more poorly during the nineteen sixties.

The agricultural

technology of the Green Revolution, including especially improved seed

varieties and dry season water supplies, was rapidly disseminated

through the cooperatives in the second half of the decade.

In an

analysis of the agricultural transformation that took place in the thana

from 1966 to 1970, Faidley and Esmay remarked that landless and near

landless cooperative members (those owning no land or less than one

acre of land) had engaged in the growing of high yielding varieties of

irrigated dry season rice in about the same proportion as had coopera-

tive members owning larger acreages.

The landless and near landless

farmers who were not members of cooperatives, by contrast, had found

it difficult to obtain access to irrigation water, and had lagged

dramatically in their participation in the Green Revolution.

Among

all the farmers in the thana, cooperative members owning between one

and three acres led the pack in the percentage of land that was winter

cropped.1+7

It was precisely this latter group of small farmers whom

the cooperative organizers had aimed to assist from the beginning.

In as much as the small farmers, and even some of the landless and

75

near landless farmers, of Comilla Kotwali Thana did benefit from the

cooperatives, these organizations were bucking the tide of change in

rural Asia.

From the perspective of the nineteen seventies, it has

become commonplace for observers of the Green Revolution to complain

that the new high yielding seeds and the new methods of cultivation

were adopted mainly by the larger farmers.

The Green Revolution in

Asia, it has been charged, augmented the income of the rich farmers.

and in some places actually diminished the abysmal standard of living

of the rural poor.

Rich farmers in most places had better'access to

agricultural credit, and to the expensive inputs (pumps, fertilizer,

insecticides, and sometimes tractors) that were needed to grow the

improved seeds.

Tenant farmers were sometimes displaced by landowners

who suddenly found it profitable to cultivate their own fields, and

agricultural wages were depressed in some places due to the substitu-

tion of tractors for hand labor.)+8

Cooperatives of the Comilla variety gave the small farmers more

equitable access to credit and to inputs of all types, as well as to

processing and marketing services.

Through the organizational vehicle,

farmers in Comilla who owned as little as an acre of land were able to

participate fully in the Green Revolution.

In a precient article written just before the collapse of East

Pakistan, Princeton Lyman pointed out that it was no mean achievement

for the Comilla cooperatives to have helped the small farmers to make

real gains in their literacy, production,and standard of living, but

that the cooperatives had nonetheless neglected the large bleck of

landless and near landless residents of the thana.

Forty-one percent

of the farmers in Comilla Kotwali Thana owned less than 0.8 acres in

1969, whereas only fourteen percent of the cooperative members were in

76

49

the same category.

Faidley and Esmay, while they demonstrate

that those landless and near landless farmers who did join the

cooperatives were able to benefit from their membership (especially

by obtaining access to irrigation water), admit that it was nonethe-

less difficult for the most impoverished villagers to meet the

standards of cooperative membership, as by making regular savings

deposits.

Cooperative loans, moreover, were issued to individuals on

the basis of a guarantee of land which could be appropriated in case of

a default; the landless, obviously, could not meet this requirement

for credit worthiness. and so could not benefit from the most important

service which the cooperatives provided.

Due to the inexorable pressure of a growing population against

a fixed land base, there seems to have been a steady increase over the

past decade or two in not only the number but the proportion of rural

people who cannot claim title to even the tiniest plot of land.

Exact

statistics on the absolutely landless are hard to come by, because many

of the standard rural surveys, such as the 1960 East Pakistan Census

of Agriculture, covered only the landed farmers.

In 1963, when Akhter

Hameed Khan spoke to the USAID convention in Comilla, he estimated the

landless laborers and near landless farmers as comprising from 15 to

25 percent of the rural population, and increasing.

The majority

element in the rural population was at that time the small farmers,

the rural group on which the Comilla academy based its village c00pera-

tives.

By 1969, when Abdul Muyeed collected the data referred to by

Princeton Lyman and by Faidley and Esmay, the landless and near

landless sector of the rural population in Kotwali Thana had apparently

swollen to 41 percent.

As the army of miserable landless and near landless paupers has

7?

grown from year to year, the spectre of a radicalized rural proletarrat

has flashed into many minds.

A revolution by the utterly indigent

may have been forestalled by the debility resulting from malnutrition.

The egaliterianism of rural poverty has also served to defuse the

revolutionary impulse: although there are many surplus farmers there

is no prominent clique of large landlords which might serve as a

radicalizing focus for political hostility.

One clear consequence of the proletarianization of the rural

populace is that the Comilla model of cooperative organization has

come under heavy fire as inappropriate to the requirements of Bangladesh

in the nineteen seventies.

Criticism of this sort first came to a

head under the regime of Sheik Mujibur Rahman, who governed Bangladesh

as prime minister and then as president from 1971 to 1975.

In an

absolutely contradictory manner, the political foundation of Mujibur

Rahman's Awami League, and later of his single state party named

Baksal (Bangladesh Krishak Awami League), rested on the vested interests

of landlords and traders which had always controlled the Bengali

villages, yet the ideology which the Mujib regime promoted was that of

socialism.

Armchair socialists who believed in the efficacy of

administrative edicts as a means of restructuring the social order

were especially influential.

Upon the advice of some members of his government who had recently

visited Eastern Europe, Mujibur Rahman declared, at a mass rally in

Dacca on March 26, 1975. that the 65,000 villages of Bangladesh should

be organized within five years into compulsory multipurpose coopera-

tives in which the landless would have an important say.

Very shortly

thereafter, Mujibur Rahman's proposal was forced as a discussion topic

upon an assembly of internationally prominent agriculturalists who

78

had gathered at the Bangladesh Academy for Rural Development: they had

come to Comilla to discuss "The Impact of High Yielding Varieties on

Income Distribution."

Mahbub Alam, who had assumed Akhter Hameed Khan's

old job at the Comilla academy, stressed to the conference participants

that the local institutions developed during the nineteen sixties,

including the Comilla cooperatives, had been dominated by the large

landowners.

The landless and near landless villagers, who were

burgeoning in numbers, had been ignored.

A new cooperative system

was called for, said Mahbub Alam and other governmental representatives,

and would the gentlemen who were so conveniently gathered at the academy

kindly assist in designing it?

The ensuing hubub has been described

by Robert Stevens:

At the conference in Comilla there ensued a great

deal of speculation about what the President

planned or intended about the internal organiza-

tional form of the new village cooperatives....

With respect to speculation about the forms of

the cooperatives, a large official committee

including many appropriate ministry officials has

been charged with making recommendations.

How-

ever, in view of the fact that the committee had

not been consulted about and was surprised by the

March 26 announcement, many guess others are ad-

vising the president on these matters....

Some

participants...insisted that the President's

intention was clear - a pooling of all land and

all other resources in each village.

Others

seemed to think the scope of internal c00pera-

tive organization remained much wider so as to

include owner operatorship of certain amounts of

land within the new cooperative structure.5O

Given the frantic surrealism of its inception, it is just as well

that nothing further came of Mujib's proposal for compulsory,coopera-

tives.

After his March speech he did not push the idea again.

A few

months later he was assasinated.

Critique pf Uphoff and Esman'g Approach

79

The foregoing analysis of East Pakistani experiments in rural

development organization has revealed many points of importance.

Uphoff

and Esman's seven criteria touched upon critical issues which show up

the comparative strengths and weaknesses of the Basic Democracies

system and of the Comilla-type cooperatives.

Yet there is something

missing in the description of rural development organizations which

emerges from the consideration of these seven criteria.

Like Uphoff

and Esman's book itself, these are summary measures of’a macroscopic

nature, generalizations derived from the study of a wide range of cases

of rural development organizations in many countries.

And.although

these generalizations from case study evidence comprise the closest

existing approximation to a theory of rural development organization,

they are not linked to other theoretical approaches to the social

sciences.

The seven criteria and the other generalizations in Uphoff

and Esman's book seem to come at us out of thin air, as if they were

reified out of the cross-talk above the Cornell Rural Development

Committee's conference table.

To take these generalizations to the microscopic scale and really

explain the workings of a specific rural development organization

would be difficult.

The difficulty arises from the lack of connection

between Uphoff and Esman's observations and more established ways of

looking at organizations.

The disjuncture is hardly surprising.

Organization theory, after all, has been built up in order to model,

and to manage, entities like General Motors.

The corporate executives

who are usually at the center of concern in organization theory are at

the opposite end of almost every sociological spectrum from the peasant

80

farmers who were the concern of the Cornell Rural Development Committee.

Organization theorists, it can safely be said, have not gone out of

their way to think about peasants.

In most accounts of the history of organization theory, a French

businessman named Henri Fayol is credited as being the father of the

administrative management school.

In a book which he wrote in 1916,

Fayol asserted that administration is divisible into five primary

functions:

planning, organization, command, coordination, and control.

He also laid down fourteen principles of sound management, among which

were chestnuts such as:

an employee should receive orders from one

superior only: and the organization should provide an orderly place

for every individual.

Fayol's pronouncements have been criticized by

subsequent writers as obiter dicta derived from a narrow, mechanistic,

 

and highly centralized view of business organizations.

They have had

a profound influence on management thought nonetheless.

One good reason why Henri Fayol extracted his principles of manage-

ment mainly from his own executive experience, rather than from the

management literature, was that there was no management literature

when he wrote his book.

Norman Uphoff and Milton Esman, I would

suggest, were in a comparable situation when they derived their set

of six primary rural development functions, and their list of seven

criteria for successful rural deve10pment organizations.

What I have sought to do in this book is to move the theory of

rural development organization beyond the Fayol phase.

At the begin-

ning of each of the analytic chapters which follow the next I have

referred to one or several of Uphoff and Esman's seven principles,

and I have sought to set them within the context of modern organization

theory.

The relevant experience from the Comilla cooperatives is then

81

reviewed in detail, and in some instances interpreted by means of a

mathematical model.

I confess that my own intellectual bent is more

toward analysis than prescription, so the pattern that emerges from

this mosaic is not an eight fold path to rural righteousness to

replace the Cornell dogma.

Some theoretical connections have instead

been demonstrated, so that someone who is either evaluating or con-

structing a rural development organization could proceed with confidence.

My efforts at applying organization theory to rural development

organization have by no means been a one way process.

In attempting

to extend organization theory into such alien territory, I have

discovered many inadequacies in the state of organization theory

itself.

The result has been a thoroughgoing revision of organization

theory, as is described in the next chapter.

Notes for Chapter 2

The Need for Models

1 Lawrence Hewes, Rura; Development, World Frontiers, The Iowa

State University Press, Ames, Iowa. 1974, p. xviii.

Reprinted with

permission.

2

'

September g3, 1223, International Bank for Reconstruction and Develop-

ment, Washington. 1973’ PO 179 180

3 Rural Developpent Sector Policy Paper, World Bank, February. 1975-

“, Ibide ’ p. “'70

5 Frederick H. Harbison, Human Resources fig the Wealth pf Nations.’

Oxford University Press, New York, 1973. P. 41.

Reprinted with permission.

6 Philip H. Coombs and Manzur Ahmed, Attackipg Rural Poverpy,

How Non Formal Education Can Help, The John Hopkins UniVersity Press.

Baltimore, 1974, p. 14.

7 Norman T. Uphoff and Milton J. Esman, Local Organizations for

Rural Development:

Analysis 9f the Asian Experience, Rural Development

Committee, Cornell University, Ithica, New York, 1974, p. xi.

8 Ibid.. pp. 99-103.

9 Philip B. Calkins, "Stability and Change in Landholding and

Revenue Systems in Bengal," in Robert D. Stevens, Hamza Alair, and

Peter J. Bertocci, editors. Rural Development ighBangladesh.§§Q

ng;stan, The University Press of Hawaii, anolUIu, 1976.

10 Elliot L. Tepper, "The Administration of Rural Reform:

Structural

Constraints and Political Dilemmas," in Stevens, gp_9;t, p. 38.

11 Rehman Sobhan, Basic Democracies, Works Programme ggd Rural

Development ig_East Pakistan, Oxford University Press, Dacca, 1968,

p. 66.

The embankment which protects the town of Comilla from innundation

by the river Gumti, which runs by the town, was given over in 1845 by the

government to the Raja of Hill Tippera, zamindar of the area, who assumed

responsibility for its maintenance.

See W. W. Hunter, A Statistical

Account 9; Bengal, Trubuer and Company, London, 1876 (reprinted by

Raj Bandhu Industrial Company, New Delhi, in 1973), p. 364.

That the

Gumti river maintained into the mid-twentieth century its reputation

82

83

as "Comilla's sorrow" does not speak well for the Raja's conscientious-

ness in repairwork.

12 Lawrence Ziring, _'I_‘_h_e Aypb Khan pg, Politics 19. Pakistan 1958-

1962, Syracuse University Press, Syracuse, New York, 1971, p. 16.

13 Tepper, 92 93, p. 47.

14 Quoted in Percival Spear, India, A MOdern History, The Univer-

sity of Michigan Press, Ann Arbor, 1961, p. 305.

15 Ibid., p. 305.

16 Quoted in Tepper. gp_gi£, pp. 38. 39.

17 Peter J. Bertocci, "Social Organization and Aglimlltural Develop-

ment in Bangladesh," in Stevens, pp pip, p. 166.

An East-West Center Book.

Copyright 1976 by the University of Hawaii.

Reprinted with permission.

18 Tables 1 and 2 are from Uphoff and Esman, 9p pip, p. 122.

19 Arthur F. Paper, 23 al., Rural Development ;g_Action, the Com-

prehensive Experiment g3 Comilla, East Pakistan, Cornell University

Press, Ithica, New York, 1970, p. 273.

20 Lane F. Holdcroft, "Rural Development Today:

A Practical Per-

spective," Department of Agricultural Economies, Michigan State Univer-

sity, 1976 (mimeographed), p. 7.

21 Akhter Hameed Khan, "Learning from China - A Pakistan Experience,"

paper presented at the International Seminar on Rural Development,

Michigan State University, 1975, p. 6.

22 Quoted in A. T. R. Rahman, pp §_l_., gn_ Evaluation 2; _tpe Rural

Public Works Programme, East Pakistan 1262-63, Pakistan Academy for

Rural Development, Comilla, 1963, p. x.

23 Quoted in John w. Thomas, Rural Pubpi_c Works app East Pakistan's

Development, Development Advisory Service. Harvard University,

Cambridge, Mass., 1968, p. 20.

The quotation is from a letter written

to Gustav F. Papenek on December 29. 1960.

2“ Richard v. Gilbert, "The Works Program in East Pakistan,"

International Labor Review, V. 9, NO. 3, March, 1964.

25 Thomas, 9p pip, p. 3.

26 Walter P. Falcon, "A Perspective on Rural DeveloPment and the

Green Revolution in Asia," 1975 (mimeographed).

27 Thomas, 9p gig, p. 48.

28 Mahamoodur Rahman, "Works Programme and Peoples' Opinion," in

The Works Programme ip Comilla:

A Case Study, Pakistan Academy for

Rural DeveIOpment, Comilla, 1966, p. 164.

8t

29 A. T. R. Rahman, et al., An Evaluation of the Rural Public

 

Works Programme, East Pakistan 1962fi_3, PakistanAcademy for Rural

Development, Comilla, EastPakistan, 1963, p. 46.

30 Sobhan, pp gig, p. 121, 123.

31 Ibid., pp. 242, 243.

32 Thomas, pp 3;}, p. 39.

33 Akhter Hameed Khan, Agricultural Economics 865, Rural Develop-

ment Administration, Winter Term, 1975, Michigan State University.

From author's class notes.

3“ Ibid.

35 Harry w. Blair, The Elgsgveneps p; pguity:

Institutional

Approaches 39 Rural Development i3 Bangladesh, Rural Development

Committee, Cornell University, Ithica, New York, 197M.

36 Thomas, pp gig, p. 33.

The number of respondents was in this

case 112.

37 Ibid., p. 40.

38 A. T. Rafiqur Rahman, Ag EValuation‘Qg the Rural Works Prpgramme,

Eagfi Pakistan, 1963:63, Pakistan Academy for Rural Development, Comilla,

19

. p. 73

39 In retrospect, Akhter Hameed Khan says that he does not regret,

and in fact was persuaded to support, the government decision to spread

the works program from one thana to #13 thanas in a single year.

He

recalled, in a 1977 conversation with the author, how Musa Ahmed, the

East Pakistan Secretary for Basic Democracies and Local Government,

and Richard Patten, East Pakistan advisor with the Harvard Advisory

Group, explained to him that if the program were not expanded immediately

to the whole province much of the year's allocation of 10 million

rupees for public works from P.L. 480 sales would be lost.

He accepted

their advice, and helped to mount a crash effort for propogating, the

works program.

no Sobhan, pp 913, p. 110.

“1 Blair, ppp_i_t_, p. 97.

#2 K. M. Tipu Sultan, "Workings of Local Councils in British Era,"

in The Works Programme in Comilla: A Case Study, Pakistan Academy for

Rural Development, Comilla, 1966, p. 21.

43 A. K. M. Mohsen, The Comilla Rural Administration Experiment,

Histogy and Annual Report1962e_3, Pakistan Academy for Rural Develop-

ment,Comilla,1963, p. 11.

an Ibid., p. 23.

“5 Uphoff and Esman, 22 _c_1_jc_, p. 76.

85

“6 Zakir Hussain, the project director of the Kotwali Thana Central

Cooperative Association, told me in 1966 that government engineers

working on the Sonaichari irrigation project in Comilla Kotwali Thana

had predicted that water would flow one way in a new diversion channel,

when in fact it flowed the Opposite way.

The villagers in the area

had told the civil engineers what would happen, but the engineers would

not listen.

“7 LaVern Faidley and Merle L. Esmay, "Introduction and Use of

Improved Rice Varieties:

Who Benefits?" in Stevens, pp git.

In case

the idea of landless farmers engaged in farming seems odd, it should be

noted that there was some land available for rent.

The major source of

income for the landless. however, was the sale of their services as

agricultural laborers.

“8 One of the best-documented arguments of this sort has been put

together by Keith Griffin, of Oxford University, in 2g; Political

Economy ggiggrarian Change, §g_Essay‘gg the Green Revolution, Maxmillan,

London, 197“.

49 Princeton Lyman, "Issues in Rural Development in East Pakistan,"

Journal 9: Comparative Administration. May, 1971.

A. Z. Obaidullah Khan,

a civil servant who has held many important rural development posts in

East Pakistan and in Bangladesh, plagarized Lyman's article extensively.

and without attribution, in a monograph titled "Rural Deve10pment in

Bangladesh:

Problems and Prospects."

50 Robert D. Stevens, "Whither Bangladesh in Agriculture -

Cooperatives or Communes," mimeographed report of a conference held at

BARD, Comilla on April 9-11, 1975.

CHAPTER 3

A Systems Approach to Organization Theory

Modern organization theory and general systems

theory are closely related, with organization

theory a special element of general systems

theory.

Systems theory and organization theory

are both concerned with the investigation and

performance of the organization as an inte-

grated whole.1

Overview

In the United States and Britain, at least, modern organization

theory is closely linked to systems theory.

Organization theorists

draw many of their interpretative images, such as "input-output?

"equilibrium," and "feedback," from one branch or another of systems

theory.

Some of the more enthusiastic writers, such as the trio of

Johnson, Kast and Rosenzweig who penned the passage above, aspire to

witness the ultimate mingling, in some Olympian grove of the intellect,

of organization theory and general systems theory.

Sceptics such as

David Berlinski, on the other hand, would contend that whosoever

mingles with general systems theory mingles with a fog.

The more

cautious advocates of a systems view of organizations, such as

Cleland and King, note that management practitioners have resisted

systems concepts because they are based on analogical reasoning rather

86

87

than empirical research; the proponents of the middle way contend,

nonetheless, that the application of some system concepts to organiza-

tional issues can be productive not only of aesthetically gratifying

insights, but even of useful management techniques.

A balanced and sensible view of the two schools of thought should

proceed from a recognition of their dominant failings.

The general

systems theorists are dreamers of the most grandiose, and oft times

of the fuzziest, sort.

Their aspiration is to create a science of

theoretical universals, a set of laws about laws which would be appli-

cable equally to the understanding of living organisms, mechanical and

electronic apparatus, and human organizations.2

The general systems

theorists disdain the lower ground on which most citizens and researchers

tread.

They would build arches, instead, from mountaintop to mountain-

top.

The problem with the organization theorists, in a sense, is that

they, too, have been attracted to heights.

The summit toward which

the organization theorists clamber is that of the office of the president

or of the corporate boardroom.

Organization theorists have flitted

happily toward the source of corporate power like moths to a light.

Ecological irresponsibility, corrupted connections with politics, price

fixing -- none of these disagreeable attributes of the modern corporate

behemoth are mentioned by the organization theorists.

They are the

faithful servants of bureaucrats, organizers, planners, and managers,

criticizing only enough to draw attention to their wares, flashy

baubles such as matrix management, computerized information systems,

and program budgeting.

Basic Bertalanffy

88

"Shall I compare thee to a summer's day?"

The question was

Shakespeare's.

Some of the academic cranks who have indulged in

theorizing about systems in general have shown an analogical inventive-

ness nearly equal to the bards: they have not, unhappily, shown a

comparable concern for intellectual discipline such as is inherent in

the construction of a sonnet.

Consider, for example, a passage of

typically unrestrained comparisons posed (in the disguise of a defini-

tion) by James G. Miller:

An artifact is an inclusion in some system, made

by animals or man.

Spider webs, bird nests.

beaver dams. houses, books, machines, music,

paintings, and language are artifacts.3

The man who coined the term "General Systems Theory," and who is

widely regarded as the founding father of the general systems movement,

Ludwig von Bertalanffy, felt compelled almost at the outset to defend

his school of thought against the charge that general systems theory

was nought but a package of analogies.

In an article published in

1955 in Main Currents in Modern Thought, von Bertalanffy wrote:

But general system theory is not a search for

vague and superficial analogies.

Analogies as

such are of little value since besides similari-

ties between phenomena, dissimilarities can

always be found as well.

The isomorphism under

discussion is more than mere analogy.

It is a

consequence of the fact that, in certain

respects, corresponding abstractions and con-

ceptual models can be applied to different

phenomena.

Only in view of these aspects will

system laws apply.

What, then, are the conceptual models and system laws that have

been discovered by the general system theorists?

The substance of

89

general system theory is a will 0' the wisp.

The chief preoccupations

of Ludwig von Bertalanffy are exploring the intellectual antecedants

of general system theory, and defending it against sceptics.

He

devotes pages to the description of developments which are closely

connected with system theory, such as cybernetics, information theory,

game theory, decision theory, graph theory, and factor analysis (his

list).

Seldom, though, does von Bertalanffy stand forth to state the

ideas around which his expository web is woven.

In my opinion David Berlinski comes down too hard on von Bertalanffy

and his school when he avers that the general system theorists have

done no more than stumble on the modeling capabilities of simultaneous

differential equations.5 The evidence for Berlinski's contention is

certainly there.

In a key chapter on "Some System Concepts"

von Bertalanffy nods to the thought that "a system can be defined

mathematically in various ways," but proceeds to present as his major

example of the representation of a system a set of equations of the

form:

dQl

_a-E'

f1 (Q19 QZDOOOQn)

dQ

2

3.: f2 (Q1! Q29'0'Qn)'

(3'1)

dQn

"a"?= fn(Q1’Q2""Qn)'

Solutions to linear versions of equations such as these;

von Bertalanffy takes pains to demonstrate, can illustrate system

characteristics such as stability, periodic fluctuations and growth.6

Such an observation is not trivial.

It is the basis, in fact, of a

9O

highly productive model-building movement whose practitioners range

from mathematical purists such as Lotka and Volterra, who developed

mathematical models of prey-predator relationships, to devotees of

computer simulation such as Jay W. Forrester, who has written a book

apiece about his respective simulations.of industry, cities, and the

world.

In the interest of historical accuracy, nonetheless, it must be

said that Ludwig von Bertalanffy did not discover the modeling potential

of simultaneous linear differential equations.

Forrester's ambitious

books on system dynamics, for example, derive from a well-established

body of theory in electrical engineering, where such equations have been

the stuff of shop talk for half a century.

And it might also be noted

that laws of the differential calculus are, after all, laws of mathe-

matics and not necessarily of systems in general.

As David Berlinski

puts it, in his critique of von Bertalanffy, "the passage from model to

theory and back to model again is made with oleaginous ease."7

Yet these mathematioo-theoretical fumblings are not the reason

that Ludwig von Bertalanffy's name is celebrated by system theorists.

He is remembered for a non-mathematical idea which he derived from his

professional work as a biologist.

The idea is that living organisms

are open systems which import energy from their environment so as to

sustain the internal organization required for productive work, and so

as to carry on the tasks of the organism.

While such an assertion

might seem to the layman to be obviously true, its significance stands

out clearly when the open system idea is contrasted with scientific

views which prevailed until about the middle of the twentieth century.

The dominant perspective outside of physics, which had moved on, was

akin to that of Newtonian mechanics, in which the solar system, for

example, could be modeled as a closed system containing all the

91

variables required for explaining the movement of the planets about

the sun.

Within such closed systems, it was believed, the tendancy was

toward ever greater disorganization, to conclude in the celestial case with

the heat death of the universe, when the order represented by the

regular orbits of the planets would have been broken down into random

movement.

By pointing to the openness and orderliness of the organism

as a whole, as against the internal and degenerating interactions of its

separate parts, von Bertalanffy pointed the way to a new conception of

biology and, he insisted, of the social sciences.

In 1954 four men sat down around a luncheon table at the Center

for Advanced Study in the Behavioral Sciences in Palo Alto, California.

Out of that luncheon conversation between a biologist, a biomathemati-

cian, a physiologist, and an economist came the idea of founding a

society for the furtherance of general system theory.

The four

originators of what came to be called the Society for General Systems

Research were Ludwig von Bertalanffy, Anatol Rapoport, Ralph<3enard,

and Kenneth Boulding.

The society set itself the task of promoting

the unity of science through the elaboration of general system concepts

and through the improvement of communication among specialists.

Unfor-

tunately the first of these methods for attaining a unified world view

has been crowded out by the second.

The society's ponderous yearbook,

General Systems, has been neatly described by David Berlinski as con-

taining a great variety of articles which, while interesting, seldom

have anything to do with general system theory.

Boulding's Typology 9f Systems

Kenneth Boulding, the only social scientist among the four origina-

tors of the Society for General Systems Research, made but a fleeting

92

sally into the field as a serious author.

His major contribution

was published in 1956 as an article in Management Science.8

In that

essay, "General Systems Theory, the Skeleton of Science," Boulding

delineated a nine-layered hierarchy of systems.

It has since become

‘Qg rigeur for authors of books on organizational systems to summarize

the Boulding hierarchy as an example of the explanatory power of

general system theory.

The hierarchy is never criticized or expanded

upon, and it's real value as a typology is very seldom tested by a

comparison of categories with cases.

Quoted below is one of the clear-

est of such summaries, contained in Kast and Rosenzweig's excellent

text, Organization and Management:

A Systems Approach.

Like most such

authors, Kast and Rosenzweig hold up the Boulding hierarchy as if it

were an archeological artifact to be gasped at, discuss it in one

following paragraph, and then reverently return it to its crystalline

case.

1.

The first level is that of static structure.

It might be called the level of frameworks;

for example, the anatomy of the universe.

2.

The next level is that of the simple dynamic

system with predetermined, necessary motions.

This might be called the level of clockworks.

3.

The control mechanism or cybernetic system,

which might be nicknamed the level of the

thermostat.

The system is self-regulating

in maintaining equilibrium.

4.

The fourth level is that of the "open system,"

or self-maintaining structure.

This is the

level at which life begins to differentiate

from not-life:

it might be called the level

of the cell.

5.

The next level might be called the genetic-

societal level; it is typified by the plant,

and it dominates the empirical world of the

botanist.

93

6.

The animal system level is characterized by

increased mobility, teleological behavior,

and self-awareness.

7.

The next level is the human level, that is,

of the individual human being considered as

a system with selfeawareness and the ability

to utilize language and symbolism.

8.

The social system or systems of human organ-

ization constitute the next level, with the

consideration of the content and meaning of

messages, the nature and dimensions of value

systems, the transcription of images into

historical record, the subtle symbolizations

of art, music, and poetry, and the complex

gamut of human emotion.

9.

Transcendental systems complete the classifi-

cation of levels.

These are the ultimates

and absolutes and the inescapable unknowables.

and they also exhibit systematic structure and

relationship.9

Several things need to be said about this typology.

In the first place, it is a linear.

The set of categories is

appropriately displayed as a straight line on the printed page, in the

same manner as the numbers 1 through 9 appear above.

The typology has

one structural dimension.

Yet the most famous of all typologies,

Mendeleyev's chemical table of the elements, is two dimensional (if

the rare earths are excluded).

What reason is there to believe that

a typology of systems should be one dimensional?

There is no inherent

reason, of course, to suspect that such a typology mgst have the form

of a two dimensional matrix like the table of the elements, or, for

that matter, that it should have the form of a cube or of some n-

dimensional figure.

But raising the issue of dimensionality does open

up possibilities for reconstructing the typology.

The second point which deserves to be made is that Boulding asserted

that every level of his typology subsumed those below it.

Interpretation

94

of this remark leads one to recognize that the one dimensional hierh

archy is printed upside down in the preceding quotation, but that

defect is easily dismissed as a gesture of convenience in accordance

with our European reading habits, inasmuch as our eyes are trained to

trace a course from the top to the bottom of a page.

Another interesting aspect of this typology is that it shows a

sharp break between the third and fourth levels.

The first, second,

and third levels, which represent closed systems of increasing complex-

ity, are all mechanistic.

From the fourth level upwards the references

all involve living things.

For Boulding, as for von Bertalanffy, the

' break between life and non-life was identical to the distinction between

closed and open systems.

Another break in the typology is evident

between the seventh and eighth levels, in that the focus shifts from

individual organisms to social systems.

As Boulding himself admitted,

the ninth level of the typology was added on as a final turret to the

typology, which he included at the risk of being "accused at this point

of having built Babel to the clouds."10

Study of these discontinuities in Boulding's typology is in-

structive.

I am led to the suspicion that he pasted together two

hierarchies, one for mechanistic systems and one for living systems,

and to the result attached social systems and transcendental systems

for the sake of decoration.

The most serious flaw in the Boulding schema is that it confuses

the classification of theoretical models with the classification of

actual objects.

In discussing clockwork systems at the second level of

his typology, for instance, Boulding cites examples of levers, pulleys,

steam engines and dynamos in one sentence, and in the following sentence

asserts that the theories of physics, chemistry and economics are also,

95

for the most part, constructed as clockwork systems.

Perpetuation of

this confusion between the theoretical and the actual can lead to major

fallacies.

Since clockwork systems are the second of Boulding's

categories, preceeding Open systems which he places at the fourth

level, one might be led to conclude from his presentation that steam

engines -- which he places at the second level -- are closed systems.

Yet in some respects such a classification is obviously erroneous.

Steam engines must be provided with inputs of coal to tman.and of

water to boil, so they can also be regarded as Open systems.

The most obvious sorts of errors that can result from the confusion

between models and reality derive from the first level of the Boulding

typology.

Paintings, whether on cave walls or on canvases hung in a

museum, are indisputably static.

A painting of a woman's face looks

much the same today as it did hundreds of years ago when it was created

by Rubens.

The painting is a static model.

Yet no sensible observer

would be led to conclude from the stillness of the painting that the

Flemish women of the seventeenth century were incapable of motion.

The static model may represent one aspect of reality, but it does not

tell the whole story.

A.virtue of Boulding's article is that he admits that the hierf

archy he has sketched is largely heuristic, in that "it gives us some

idea of the present gaps in both theoretical and empirical knowledge.

Adequate theoretical models extend up to about the feurth level, and

not much beyond."11

The theoretical models of self maintaining

structures at the fourth level to which Boulding refers are those of

open system theory, which was, at the time that Boulding wrote his

article, pretty much limited to the writings of Ludwig von Bertalanffy.

"Beyond the fourth level," Boulding wrote, "it may be doubted whether

96

we have even the rudiments of theoretical systems....Up to now, what-

ever the future may hold, only God can make a tree."12

If theoretical reference points are indeed so scarce at the higher

levels of Boulding's typology, then perhaps those higher levels should

be discarded.

What we should be seeking, after all, is something more

than an arbitrarily designed, cut and paste typology.

Every category

of a typology of systems should be theoretically justifiable and

logically distinct.

And the relationship between the categories,

ideally, should itself be orderly, much as the transition from one

category to the next in the chemical table of the elements is governed

by explicit rules.

Boulding's Typology Revised

As a first approximation of a replacement for Boulding's hierarchy,

I would propose a simple four level schema.

1.

Static structures

2.

Simple machines

3.

Machines with input

4.

Organic systems

Figure 3-1:

A simplified system typology

Systems above the dotted line in Figure 3-1 are closed;

those

below are open.

Systems at the first level are static, those at the

second, third and fourth levels are dynamic.

To avert the sort of confusion about the classification of models

versus actual objects which permeates the original Boulding typology,

97

let me assert at the outset that the first three levels of static

structure, simple machines, and machines with input shown in Figure

3—1 represent categories of models.

These are classifications of

theoretical constructs, not of items from the real world.

The fourth

level of the simplified system typology, organic systems, occupies an

ambiguous position between theory and reality.

This level of classifi-

cation is intended to correspond to the fourth level of Kenneth Boulding's

typology, which he stated was the level at which life begins to be

differentiated from non-life, and at which systems are first capable of

self-maintenance and reproduction.

Boulding contended, in the passage

quoted above, that adequate theoretical models extend up to about this

level and not much beyond.

In the absence of adequate theoretical

models of self maintaining, reproductive systems it is difficult to

define a classification for theoretical models of real world living

systems.

However, the concept of organic systems should not be thought

of as limited to biological examples.

As I shall attempt to demonstrate,

human organizations belong in this classification as well.

The third level in Figure 3-1, machines with input, represents a

theoretical construct which is omitted entirely from Boulding's

typology.

In

company with many writers in the fields of system

theory and organization theory, Kenneth Boulding seemed to presume

that open system models can be applied only to living organisms.

Yet

the classic examples of open system models were developed to explain

electrical circuits such as those that are found in a radio.

A radio

has inputs of electricity and radio waves, internal transformational

processes, and an output of sound.

A radio is an open system, and it

is most appropriately modeled as such.

Yet a radio is assuredly not

alive-

98

The name which I have applied to the first of these four levels

of systems, static structures, is used by Boulding, and the content of

that category I see as identical to Boulding's level of frameworks.

The names of the second and third levels in Figure 3-1, simple machines

and machines with input, are taken from John Hunter, a mathematical

psychologist at Michigan State University whose views of systems are

strongly influenced by the writings of W. Ross Ashby.

The term "organic

systems," which I have applied to the fourth level of the typology, is

a throwback to nineteenth century writers such as Auguste Comte, who

described human organizations through a biological, organismic analogy.

MW

Static structures, or frameworks, form the intellectual home of

Copernicus.

This level of system includes, in Boulding's words, "the

geography and anatomy of the universe -- the patterns of electrons

around a nucleus, the pattern of atoms in a molecular fermula, the

arrangement of atoms in a crystal, the anatomy of the gene, the cell,

the plant, the animal, the mapping of the earth, the solar system, the

astronomical universe.

The accurate description of these frameworks is

the beginning of organized theoretical knowledge in any field, for

without accuracy in this description of static relationships no accurate

functional or dynamic theory is possible."13

A little thought should make it plain that for any one of

Boulding's own examples the static description is only a partial

representation of reality.

The orbits of electrons around a nucleus

may be immovably displayed on a printed page, yet we know that the atom

in actuality is a blur of motion.

The point of this observation is well

99

made by Boulding himself in the passage above.

Theoretical constructs

of static structures preceed the more comprehensive models of dynamic

systems.

Copernicus preceeded Newton.

Time is not an issue in the fixed universe of static structures.

Relationships among variables can be expressed in algebraic equations,

such as the economists' production functions, or the regression equa-

tions which are so fascinating to empirical social scientists.

For the

linear case, the general form of a system of such algebraic equations

can be written in matrix notation as:

AX = K,

(3'2)

in which A is a matrix of coefficients, X is a vector of variables, and

K is a vector of numbers.

Simple Machines

Sir Isaac Newton and the eighteenth century deists would have

felt intellectually at home among system constructs at Boulding's

second level of clockworks, containing dynamic systems with predeter-

, mined, necessary motions.

According to Boulding:

"Simple machines such

as the lever and pulley, even quite complicated machines like steam

engines and dynamos fall mostly under this category."14 I have

appropriated the term "simple machines" to describe this category,

because it matches Boulding's meaning just as well as "clockworks," and

because it carries a definite (and identical) meaning for cybernetic

theorists. 15

Just because simple machines are at such a lowly level in the

systems hierarchy, or just because they are called "simple," models of

100

this sort should not be regarded with contempt.

As Boulding himself

points out, most of the theoretical constructs in physics, chemistry,

and economics fit into this category of the typology.

Moreover, the

comparison between statics and dynamics -- which corresponds to the

comparison between static structures and simple machines -- has been

an extremely important one in the history of social thought.

Auguste Comte, the French philosopher who is considered the fbunder

of sociology, divided his new science of social relationships into

static sociology and dynamic sociology, a categorization which he

based on the established division of biology into anatomy and physio-

logy.

A century after Comte published his treatises in the 1850's,

sociologists were still absorbed in the static-dynamic comparison,

posing the question as a debate on the relative importance of

structure versus function.

Following the usage of economics, Boulding contends that simple

dynamic systems may be analyzed via the method of comparative

statiCs, in which "we compare two equilibrium positions of the

system under different values for the basic parameters.

These

equilibrium positions are usually expressed as the solution of a set

of simultaneous equations.

The method of comparative statics is to

compare the solutions when the parameters of the equations are

101

changed."16

But this is really a slight of hand.

If there is any

coherence to the behavior of the simple machine over time, then some

underlying time function must describe the changes in the system

parameters.

A more comprehensive way to describe a simple machine

mathematically is to display the relationships among system variables

as a set of differential or difference equations.

A set of such

equations would be of the form of 3.1, which was Ludwig von Bertalanffy's

mathematical example of a system.

The solutions to such equations are

written as time functions for each system variable.

This is, as

Boulding says, the method of true dynamics.

If the functions referred to in 3.1 are all linear, then the mathe-

matical representation of a simple machine operating in continuous

time can be written as:

it: AX.

(3-3)

in which K signifies a vector of time derivatives of the system

Xm dX2

dX3

EEE

variables (

dt 9 dt I dt '0 no dt f; A is a. matrix Of CoefflClen'tsg and

X is a vector of system variables.

For discrete time, the same system

can be represented as

AX = AX,

(3A)

in which 15X is a vector of discrete changes in system variables

(Axl, AXZ, AX3,...AXn).

In a dynamic system the function describing the rates of change

or the changes in the system variables need not be linear.

I have

used the illustration of linear functions mainly because of the brevity

of notation provided by matrix algebra.

The vector-matrix notation

102

also leads to a symbolically simple comparison between static struc-

tures, simple machines, and machines with input.

A more general, set theoretic definition of’a simple machine has

been formulated by John Hunter.

His definition, which I would take to

be the overall description of systems at the second level of the

Figure 3-1 typology, does not imply that the functional relationships in

a simple machine must be restricted to the mathematically tractable

linear case.

According to Hunter:

M is a simple machine with states S governed by

T if and only if:'

1.

S is the set of all possible descriptions

of M;

2.

T is a function whose domain is S and whose

values are in S, i.e. for each s in S, T(s)

is an element in s.1

For an illustration of a social science theory built according to

the principles of the simple machine, or closed dynamic system, I turn

again to John Hunter, who has developed many mathematical models of

this type.

He refers to them as cybernetic models.

The model of

dyadic interaction is about the simplest conceivable two variable

dynamic mathematical model, and Hunter frequently employs it as an

example in his lectures and mimeographed class notes.

Suppose two people interact in a closed room.

The only stimuli

they receive emanate from each other.

Suppose that neither person masks

his feelings, so the messages that each person sends about the other

are directly proportional to his feelings about that other person:

if

he hates the other person he will send a hateful message, and if he

loves him he will send a loving message.

According to the reinforce-

ment theory of attitude change, positive messages will move the

103

attitude of the receiver in a positive direction.

A positive message,

in this case, will result from a positive feeling on the part of the

sender.

If x is the attitude of the first person toward the second, and

y is the attitude of the second person toward the first, a mathematical

model of the mutual changes in these attitudes can be stated as:

it:

(my

o
<
‘

I
I

0x,

(3.5)

In this excruciatingly elementary model, the coefficient (3 represents

the relationship of changes in each person's attitude to the messages

he receives from the other person, which are in turn related to the

other person's attitude; all of these relationships are compressed

into the single coefficient <1, which is taken to be the same for both

parties.

In terms of the notation of the matrix equation 3.3, in which a

linear continuous time simple machine was symbolized as X = AX, the

vector X for the dyadic interaction model consists of the elements

(x, y)T, the vector X of time derivatives is (x, y)T, and the differen-

tial matrix A is

O

a

c

0

.

[3:113 2]

[:1

The time functions which are the solutions to this system of

differential equations are:

x .+ y

x

- y

_

x: 02 oeat+ o2 oeat

x0+y0

at

yo-xo -at .

(3-7)

The simplicity of the matrix notation is even more apparent for

this set of solutions, which can be summarized as

x0.

(3.8)

X0 in the above equation is a vector of the initial values of the two

variables, x0 and yo, which in the dyadic interaction model indicate

the attitudes which the two persons held toward one another when they

entered the experimental chamber.

The expression eAt is a wonderous

entity, representing a potentially huge amount of information in a few

symbols.

Deriving the exact content of e

can be a devilishly

At

difficult process in complicated models with many variables and coeffi-

cients.

For the dyadic interaction model,

r-

eat+ e- at

at

-at

e

-

   

2

At

e

:-

2

at.

-<1t

eclt + e-cxt

(3-9)

e

- e

2

 

   
.1 

2

Just what do equations such as 3.7 or 3.8 imply about the behavior

of this mathematical model of the attitudes of person number one and

person number two?

That can be a tricky question to answer for even

such an elementary mathematical model.

For any given set of initial

attitudes, x.o and yo, it is a simple matter to plot the ensuing

trajectories of the attitudes of the two parties.

The form of those

trajectories varies considerably, though, with changes in the values

 

 

  

106

of the initial attitudes.

A way out of this dilemma is provided by

the device of the phase plane, in which the values of the system

variables x and y are plotted against each other, rather than separately

against time.

For’a cybernetic or closed system mathematical model

the phase plane summarizes all the trajectories that can result from

all the different sets of initial conditions.

A phase plane for the

dyadic interaction model of attitude change is shown in Figure 3—2.

(In order to plot a phase plane for the model, a value had to be

assigned to the parameter :1.

Following Hunter's example, I chose the

value <1 = ln 2, or the natural logarithm of 2.

This value simplifies

the calculations which are involved in the plotting of the phase plane.)

For the sake of visual clarity, only a few sample curves have been

drawn in Figure 3-2.

It should be understood, though, that the complete

phase plane for the dyadic interaction model would consist of an

infinite family of curves (hyperbolas, in this case) nested one inside

the other.

One of the most striking implications of the dyadic interaction

phase plane is that the two participants in the psychological experiment

end up loving or hating each other with infinite intensity.

If the

initial attitudes (x0, yo) of the participants are similar to those at

point A, they will eventually zoom off past point B toward Nirvana.

From a slightly different combination of initial attitudes such as is

to be found at point C, the miserable experimental subjects will plunge

past point D toward the depths of limitless hatred.

Although it is

often the case that intimate two-person relationships such as marriage

tend toward emotional extremes, the evidence for infinite sentiments

is scanty outside of the realm of poetry.

Empirical psychological studies lend more support to another

107

implication of the dyadic interaction model, which is that the attitudes

of the two parties will tend toward equality with one another.

Graphi-

cally, this is indicated by the asymptotic approach of all the phase

plane trajectories toward the line y'= x.

The tendancy toward

reciprocity of attitude among interacting individuals is a well-

established psychological finding.

From the perspective of system theory, an important aspect of most

closed system mathematical models such as X = AX is that the direction—

ality of time is immaterial to their functioning.

The models would

run just as well if time were reversed.

They are like a movie film,

which can be run through the projector backwards as easily as forwards.

To predict the sequence of action in the motion picture, no matter

which way the film runs through the projector, all that must be known

is the frame at which the movie starts.

For systems which are simple

machines, similarly, the state variables of the system, the transforma-

tional relationship, and the starting position are all that need be

known to project the behavior of the system either forwards or backwards

in time.

The arrowheads indicating the orientation of the trajectories

on the phase plane in Figure 322 are in this respect misleading.

So

long as time goes forwards, the arrowheads point the direction in

Which the system will proceed along each trajectory.

Yet there is

nothing inherent in the mathematical model represented by the phase

plane that indicates time must go forwards.18

Sir Isaac Newton and Pierce Simon de Laplace were the most

brilliant and persistently influential mathematical designers of simple

machines.

In the great clockwork systems of equations which they

constructed, the directionality of time was irrelevant.

"In Laplace's

world," David Layser has remarked, "there is nothing that corresponds

108

to the passage of time....

The past and future coexist on equal terms,

like the two rays into which an arbitrarily chosen point divides a

straight line."19

Machines With Input

In the simplified systems typology presented above in Figure 3-1,

machines with input were presented as the least sophisticated type of

open system.

For a definition of this category of systems, I rely

once again on John Hunter.

M is a machine with input E governed by T if and

only if:

1.

S is the set of all possible descriptions of

M:

2.

E is the set of all possible environments of

M:

3.

T is a function which maps S x E into S, i.e.

for each s in s and for each e in E, T (s, e)

is an element in 5.20

As with simple machines, there are many ways to model systems of

this sort mathematically.

For the sake of simplicity and clarity, I

will continue to resort to the linear, continuous time case for

examples.

A mathematical model of such a system can be put in the

general form:

x (t) = AX (t) + GU (t) .

(3.10)

The meaning of equation 3.10 is that the vector of time derivatives of

the system state variables at time t, x (t), is equal to a coefficient

matrix A times the value of the state variables at time t, x (t), plus

a coefficient matrix C times an input vector in the form of a function

109

of time U (t).

To simplify notation, this general matrix equation for

a linear machine with input may be written as

X = AX + GU,

(3.11)

which is the form that will be used hereafter in this book.

It

should be understood, though, that if the system represented by 3.11

is to have a deterministic solution, as will all the examples discussed

in this chapter and in Chapters 4 and 5, then the input vector U has

prescribed values for all t from t = 0 to the end of the time interval

over which the mathematical model is being tested.

To illustrate the application of equation 3.11, let us imagine

that the two participants in the dyadic interaction experiment are

subjected to various stimuli, or inputs, which influence their

attitudes toward one another.

The simplest such input would be a

constant signal, let us say pleasant background music.

The music

constitutes a stream of positive messages to which the two subjects

will respond.

As before, the subjects will also respond to messages

from each other.

If the stream of incoming messages is at a constant

value

8, an Open system model of dyadic interaction could be expressed

as:

X: ay+ Bu(t-a)

.

Y

a y-+ B‘u (t - a)

(3.12)

Since the input is identical for the two subjects, the input

coefficient matrix C becomes a (2 x 1) vector,

3 =

,

(3.12)

110

which premultiplies the input vector U (t) = u (t - a).

In this

expression, u (t - a) is the standard notation for a step input in

which the steady input signal commences at time a, as is shown

graphically below.

U (1:)

B

A

 

u (t)

 

_
-
-
.
.
b

a

t

Figure 3-3.

A step function with amplitude B .

To simplify the presentation as much as possible, let us assume

that a,= 0, which implies that the pleasant background music is

switched on at the instant when the open system dyadic interaction

experiment begins.

The general form for a solution of the matrix equation X.= AX + GU,

which is commonly referred to in control system literature as the

state equation for the system, can be written as:

X = e

X +

o

t

f

o

eA<t ' T ) GU ( :) dr .

(3.13)

in which T is a dummy variable of integration, and in which the initial

values of the system (state) variables are assumed to have been

measured at time t = 0.

This is a fOrmidable mathematical expression.

The second term on the right hand side, which is a convolution integral,

is particularly imposing.

The first term on the right hand side of 3.13 is recognizable as

the solution to the linear continuous time model of a simple machine.

'1 .01! iii,

[illilfl‘

1.1.. [’lllftcl-‘Il

:

J

 

 

  
 

hr] . ’untliflennrllhnf

3-1"“;

’1:

. «aloll.:H”u .. I
Flt,

.I’-::’

  

 

  

 

 

 

      

 

 

A NIH“...

:1»):

11:!

0.0:

III I!!!” all

-.

.‘L

tumulutflnlliiall

OI." '.l|‘.l‘.1" Tali
1lx.ol.ll .rll 111.1m't61:
{tv'T

:1 III“

. I1 11].! .1.. .

.0..fl31.1

.I.

.1- it.

.JI 1’, v

Irl.

r‘lw ll:r

:I?‘ 11‘

     

112

i = AX.

In the context of a machine with input, this first term of the

total solution is called the free response of the system.

It represents

the trajectory which the system would follow if there were no input,

which is to say if the elements of the input vector U were always zero.

A look at the convolution integral comprising the second term on

the right hand side of 3.13 reveals that the matrix A, which is the

differential coefficient matrix for the closed system model X.= AX,

plays a role in that term as well.

This second term of the total

solution is called the forced response of the system: it indicates how

the system would behave if the values of all the system variables were

zero at time t = 0, so the system would be responding only to input

signals.

Even the forced response of the system, it appears, is

strongly influenced by the closed system characteristics of the system,

as summarized in the coefficient matrix.A.21

The linkage between closed

and open systems is strong indeed.

For the open system dyadic interaction model presented above, the

general solution given in 3.13 can be simplified to the expression:22

-

At

x- e xo+ a a,

_5_

(3-14)

e at - 1

where C =

at

.

(3-15)

6

- 1

For this two variable, constant input model, the easiest way to

grasp the implications Of the solution set given by 3.14 is to plot

representative trajectories once again on a phase plane.

This is what

has been done in Figure 3-4.

Again, the value Of<1 has been taken to be

ln 2.

Additionally, it was assumed that

B = 8 a.

The influence of the closed system upon the open system is apparent

113

from a comparison of Figures 3—2 and 3-4.

For this special case in

which the input to the open system dyadic interaction model is a

constant, and influences both parties in the psychological experiment

equally, the solution of the open system model is a shifted version of

the solution to the closed system model.22

The intersection of the

axes of the hyperbolas on which the phase plane trajectory is traced

has been shifted from the origin to the point (- {éw - %}0.

Since it

was assumed that

B = 8a , the point of intersection of the axes is

at (-8, -8).

What happens if the input signal to the dyadic interaction system

has a form other than that of a step beginning at t = 0?

There is no

end to the possible fOrms of alternative input signals, and I will

illustrate only two cases:

a delayed step input and a ramp.

The

delayed step corresponds to the case in which the parameter a in

Figure 3-3 has a value greater than zero.

An undelayed ramp input is

shown in Figure 3-4.

U (is)

 

r (t

slope of B

h

7

t

Figure 3-5.

A ramp function with slope 8 .

From the general solution of a linear Open system model given by

3.13, it is evident that no matter what the form Of the input signal.

the first term of the overall solution will be eAtxo.

When x0 = 0

(indicating that the trajectory under consideration starts at the

origin of the phase space) this first term disappears and we are left

114

1?

40

30

an

[0

are

6,95:

4:90

 

49 '30

so -

'

.0

so

an

40 '71.

4 /

9"“:

I

,é'l o/o

/'

’.

19‘-

.

//, [I’Lt'i

:47

x/ ’/f

l

//

//

/ 3‘"

a”

_

’9

 

J

(,e'c

Anew

—.-— MICWRY ””77! W

_-—- 77441::7”” ”(I 77/007 /NPW

#603! 306

MA!!! M! m: M am new” M0054 0!

0740/6 /~rtucr/orv WIfi/ .4 012.4740 371.0

man/7:

115

with the forced solution of the model.

Forced solutions for the step,

delayed step and ramp input signals are tabulated below.

Derivations of

these solutions are given in Appendix A.

Table 3-1.

Forced solutions for three input signals to the

reinforcement model of dyadic interaction.

Input Signal

Model

Forced Solution

 

Step

X: ay+8u(t)
r: ax+Bu(t)

_§_ e::-1
GILL -1

_j,

DelayedStep

X: ay+Bu(t-a)

___B_ 62E::a)-1

Y: cx+8u(t-a)

c.3—

a

-_1_

Ramp

.
.
2‘: ay+8r(t)

Y: 0x+8r(t)

8 eat" at-i

Te - (It-1

1.

 

‘7

 

Phase planes showing the response of the dyadic interaction system

to the delayed step and ramp inputs are shown in Figures 3-6 and 3-7.

The inherent logic of the delayed step phase plane, Figure 3-6, is

more readily apprehended than is that of the ramp phase plane, Figure

3—7.

Trajectories in Figure 3-6 behave as if they were in a closed

system dyadic interaction model, such as was shown in Figure 3-2,

between t = 0 and t = a.

In this example, a.= 3.

Thereafter, the

trajectories behave as do those of Figure 3-3, the open system model

with the undelayed step input.

For example, a trajectory which starts

at (-1, -1). which is point A in Figure 3-6, reaches (-8, -8) at t = 3

and stays at that equilibrium point, designated B in the figure,

forever after.

If it had not been for the application of the step

input at t = 3, the trajectory would have continued down the line y'= x

toward negative infinity.

Another such special trajectory is CDB, which

at t = 3 turns a corner and heads in a straight line toward the equili-

brium point at B = (—8, -8).

It should be noted that the three

trajectories AB, CDB and EFG in Figure 3—6 are all special cases, since

116

 

  

-'- mJICfORY W/ffl M9907

--- WJICT‘OKY W/f/iour INpur

13/60“ 3" 7

FHA—fl PUNE ”I AIV OP!” 5”:72/11 IMMA
0" 0740/6 /N7€£A€7’/OIV WI)?! .4 MM lit/PU":

117

they begin or end on straight line paths.

A more typical trajectory

is HIJ, whose hyperboloid curvature is only slightly distorted (or,

more exactly, shifted) by the application of the input signal after

t = 3.

Trajectories resulting from a ramp input, shown in Figure 3-7,

are no longer recognizable as hyperbolas, or as any other'members of

the classic set of conic section curves.

The geometry of the closed

system dyadic interaction model is obliterated, at least in the

vicinity of the origin, by the ramp input.

Only one line, y'= x,

remains unaltered.

And even on that line the model can behave in

unfamiliar ways.

A trajectory commencing at A, where (X0, Y0) =

(-10, -10), goes down the line y'= x to point B and then reverses

itself, so by the time t:= 6 it has gone up to c at (39.1. 39.1).

There is no point on the phase plane at which the system remains at

rest, as it did at the origin in the closed system phase plane, and at

(-8, -8) in the phase plane for a step input.

In contrast to simple machines, which are in most cases completely

time reversible, machines with input can run backwards only over a

restricted range:

the range of time in which the input signal is

explicitly defined.

In accordance with standard conventions, the step

and ramp functions in Figures 3-3 and 3-5 are defined only for positive

values of time.

A system with such functions as inputs cannot run

backwards beyond t = 0, since the input signals are undefined for

negative time.

The phase plane itself has a different meaning for closed system

and for open system models.

For a closed system model puch as the one

represented by the phase plane in Figure 3-2, the starting point for

time (t = 0) along any trajectory is arbitrary.

No matter where the

118

system starts along any trajectory in a closed system phase plane, the

system will follow that same trajectory forever.

Except in special

cases, this is not true for a phase plane of a machine with input.

For example, in the phase plane for the dyadic interaction model with a

delayed step input, Figure 3-6, if the system starts at point A (-1, -1)

it will proceed to point B (-8, -8), and stay there.

However, if the

system starts at point B, it will march on down the line y'= x toward

infinitely negative values for both variables.

To take another example

from the same phase plane, the system will not pursue the straight

trajectory

DB in Figure 3-6 if it starts at any point other than C

(-32, 30) along the trajectory segment CD.

From these examples it is

apparent that for a machine with input a phase plane trajectory has a

definite beginning, and indicates the course the system will fellow

from that defined starting point.

In a closed system phase plane, the

location along any trajectory of the point where t = O is fixed rather

than arbitrary.23

Organic Systems

At the fourth level of the simplified systems hierarchy presented

in Figure 3-1 are organic systems.

The label "organic" for this

category of systems is drawn from the nineteenth century French

sociologist Auguste Comte, who described human societies as organisms.

Like machines with input, organic systems are Open systems:

that is,

they receive and respond to inputs from their environment.

That is not

to say, however, that all open systems have the characteristics of

organic systems.

An amazing number of prominent social theorists have used the

119

term "open systems" as a label for a category of systems which corres-

ponds more or less to what I have called organic systems.

Their

terminology is misleading, to say the least.

Walter F. Buckley, for

example, has written:

The transition from mechanical systems to adap-

tive, information processing systems is closely

related to the transition from the relatively

closed to the Open type of system.

That a sys-

tem is Open means, not simply that it engages

in interchanges with its environment, but that

this interchange is ap essential factor under-

lying the system's viability, its reproductive

ability or continuity, and its ability to

change.24

The originator Of this pervasive and egregiously improper termi-

nology is the grandfather of general systems theory himself,

Ludwig von Bertalanffy.

Since his professional training was as a

biologist, it was natural for von Bertalanffy to have the image of

living organisms in mind when he wrote of Open systems.

The major

advances in the mathematical theory of open systems, however, have been

made by electrical engineers.

The electronic circuits of resistors and

capacitors which are their mental referants for open systems can be

adequately modeled as machines with input.

(The mathematical examples

given by von Bertalanffy, curiously enough, are all models of closed

systems of the form X = AX.)

Unfortunately, the social theorists

who have sought to apply open system ideas to topics such as human

organization have lacked the mathematical training to be responsive to

the contributions of the electrical engineers, and so has been

propagated the misconception that open systems are to be found only in

biology and in the social sciences.

Fremont Kast and James Rosenzweig,

for instance, wrote in this ill-informed tradition when they asserted

that:

120

Bertalanffy made another major contribution in

setting forth a distinction between closed sys-

tems and open systems.

Physical and mechanical

systems can be considered as closed in relation-

ship to their environment.

Thus, the first three

levels in Boulding's hierarchy are closed sys-

tems.

On the other hand, biological and social

systems are not closed but are in constant inter-

action with their environment.

This view of

biological and social phenomena as open systems

has profound importance for the social sciences

and organization theory.

Traditional theory

assumed the organization to be a closed system,

whereas the modern approach considers it an open

system in interaction with its environment.25

Each of the categories in Boulding's hierarchy, and in the simplified

typology presented in Figure 3-1, is more comprehensive than the one

before it, and in a sense includes the category before it plus something

else.

The static structure of a still photograph may be one frame in a

movie film, a closed dynamic system.

And as we have seen in the

preceding discussion of the dyadic interaction model, the solution of

a mathematical model of an open dynamic system, or machine with input,

may contain and be strongly influenced by the solution to a closed

system model of the same phenomenon.

In very simple terms. a simple

machine is like a static structure plus motion.

A machine with input

is like a simple machine plus input.

What, then, is the additional

something which distinguishes an organic system from a machine with

input?

The distinguishing additional feature is, I would contend, the

capacity of an organic system to maintain and reproduce itself.

So an

organic system is like a machine with input plus maintenance gpd

pgproduction.

This assertion is easily misunderstood, and deserves

amplification.

121

First of all, adherents of what has been called the reductionist

view of science might object that self maintenance and reproduction can

be modelled as processes behaving like simple machines or'machines with

input.

I would disagree with that contention, but since it springs

from a common line of argument, it is probably best to fend off this

objection at an early stage of the exposition of organic systems.

In September of 1972, an international group of eighteen geneti-

cists, zoologists and philosophers gathered at the Rockefeller

Foundation Conference Center at Villa Serbelloni in Italy to deliberate

on Problems of Reduction in Biology.

The papers they delivered there

were later reproduced as a volume titled Studies ;p_pp§ Philosophy 9:

Biology, Reduction gpg Related Problems.26

The issue which the

conferees discussed was whether the laws of biology were entirely

reducible to the laws of chemistry, or, in a still more basic reduction,

to the principles Of physics.

That issue is similar to the challenge

posed above to the special identity of organic systems.

For if biology

is in fact no more than applied chemistry and physics, then its legiti-

macy as a distinct realm of inquiry is open to question.

And if all

that is true of organic systems can be studied through the modelling

techniques of simple machines or machines with input, then there would

be no need to present organic systems as a distinctive category in our

typology of systems.

The conference participant who presented the clearest resolution

Of this problem Of reduction was Peter Medawar Of the Clinical Research

Centre in Harrow, Middlesex, England.

Without any thanks to.

Auguste Comte, who had advanced much the same preposition a century

earlier, Medawar asserted that the sciences could be ranked in the

fbllowing hierarchy:

122

1.

Physics

2.

Chemistry

3.

4.

Biology

Ecology/SeeiolOSY-

All of the laws of physics apply in chemistry, in biology, and in

ecology and sociology, just as all the laws of chemistry apply in the

sciences at levels 3 and 4, and so forth.

Yet at each level of this

hierarchy, Medawar asserts, new principles emerge which gg,pgp apply at

the lower levels.

For an analogic illustration of this proposition, Medawar'presented

a hierarchy of four geometries.

1.

Topology

2.

3.

Projective geometry

Affine geometry

4.

Euclidian geometry.

As the most general, or fundamental, Of all geometries topology includes

properties such as "inside" and "outside" which still apply in the more

specialized geometries at levels 2, 3 and 4.

The property of linearity

emerges in projective geometry, while in affine geometry it becomes

meaningful to refer to a rectangle or an ellipse.

The full ponoply

of familiar geometric figures are definable in Euclidian geometry.27

The transition from the second to the third level of Medawar's

hierarchy of the sciences, the transition from chemistry to biology,

is more or less equivalent to the transition from the study of non-

living to the study of living things.

(True enough, one can study the

chemistry of organisms, but in doing so one would only be demonstrating

the inclusion in each level of the hierarchy of the sciences below

it.)

The participants in the Serbelloni conference on Reduction in

Biology did not reach a consensus on which are the special qualities

123

that distinguish life from non-life, or biology from chemistry.

Speakers noted that it used to be thought that the capability of

reproduction was unique to living organisms, but that this dividing

line between the living and the non-living had been cast aside when it

was discovered that some otherwise inert molecules can reproduce them-

selves.

Giuseppe Montalenti of the Instito di Genetica in Rome

asserted that individuality is the distinguishing characteristic of

life.

Natural selection and life itself begin, MOntalenti said, when

a self-reproducing molecule mutates to an individual different from

others Of the same sort.

This brings us to the second way in which misunderstandings can

arise about the assertion that organic systems are distinguished by

the capacity for self maintenance and reproduction.

It is quite possible

that Kenneth.Boulding was wrong when he claimed, in his article on

general systems theory published in 1956, that self-maintenance and

self-reproduction are the principal and critical features which set

life apart from not-life.

His assertion was in tune with biological

thought in the nineteen fifties, but it may be out of date.

Or

else, indeed, biologists may never settle upon a definitive set of

criteria to discriminate the living from the non-living.

That is my

impression after reviewing the material from the Serbelloni conference,

plus a few additional works such as those of Jacques Monod

29 and of

s. E. Luria.3°

Instead of perpetually readjusting the definition of

organic systems to accord with the most recent fads in biological

thought, it seems more sensible to stick with one set of standards.

Boulding's standards seem as good as any.

If the point at issue were to distinguish actual organic systems

from actual inorganic systems, then the BOUlding standard might be less

124

easily defended.

But that is not the issue.

The problem in defining

organic systems is to set them apart from other sorts of theoretical

models, and especially from machines with input.

Some kinds of actual

systems can be described by models having the form of machines with

input, but machines with input are not actual systems.

Living things,

likewise, can be interpreted by organic system models. yet the actual

and the theoretical realms remain apart.

The form which an organic system model would assume is not clearly

definable, since as Kenneth BOUlding observed adequate theoretical

models peter out at the level of complexity of living things.

Some

characteristics which such a theory might possess can be inferred by

analogy with the other sorts of models which we have described pre-

viously.

In the model of a linear simple machine given by X = AX, for

example, the structure of the systemic relations is given by the

coefficient matrix A.

The properties of self maintenance and self

reproduction, I would suggest, have a lot to do with the stability of

that coefficient matrix.

For the reinforcement model of dyadic interaction, for instance, we

might ask for how long a period of time we could expect the matrix

A =E 3 to govern the alteration of the attitudes of the experimental

subjects.

Returning to the assumptions underlying the construction of

the model, we recall that the parameter (r was taken to symbolize two

factors concurrently:

1) the relationship of the attitude of one of the

subjects to the messages which he sends to the other subject, and b) the

relationship between the messages a subject receives and the,changes

in his attitudes toward the sender.

To cap off the model-building

simplification, the parameter c: was taken to be the same for both

experimental subjects.

There is reason to question the stability of

125

this all-embracing parameter.

If the subjects' attitudes traced a

trajectory toward unmitigated hatred, for example, they might reflect

upon their increasingly unpleasant experience and spontaneously alter

the psychological principles governing their relationship.

In other’kinds of simple machines. likewise, internal fOrces of

disorganization will inevitably come into play.

Gears will start to

wobble on their bearings, pistons will wear unevenly from abrasion, and

in multitudinous ways the machine would start to break down.

As the

mechanism wore out, its internal Operating parameters would change-

The coefficient matrix A in the model i = AX would reflect the relation-

ship of the system variables to one another with ever-diminishing

accuracy.

Exactly the same line of argument can be applied to a machine

with input modelled by the matrix equation X = AXr+ GU.

There also

the stability of the matrix A is open to question.

Even resistors and

capacitors deteriorate, so the real response of an electrical network

will drift away from model predictions.

It is important to keep clearly in mind here the distinction between

a real world entity or process about which we are conceptualizing, and

a model of that reality.

It is the models, the intellectual constructs,

which have been categorized as static structures, simple machines, and

machines with input.

Those models are never inclusive of all aspects

of reality.

The models may hold over a certain range, but in the end

they can be expected to fail.

Predictability will give way to uncer-

tainty.

In a short section of his treatise on the geometrical interpretation

of differential equations. Solomon Lefschetz discusses the concept of

structural stability.

A structurally stable system, he says, is one

126

for which the portrait of trajectories in phase space is invariant

under small changes in the system coefficients.31

To the best of my

knowledge, that is about as far as the mathematical discussion has

been carried.

For all his faults of overstatement and imprecision, Ludwig von

Bertalanffy must be given credit for perceiving the importance of order-

ing influences in the biological and social sciences.

He paraphrases

approvingly, for instance, a comment made by Warren Weaver, one of the

founders of mathematical information theory.

Classical physics, Weaver said, was highly suc-

cessful in developing the theory of unorganized

complexity.

Thus, for example, the behavior of

a gas is the result of the unorganized and indi-

vidually untraceable movements of innumerable

molecules: as a whole it is governed by the laws

of thermodynamiCs.

The theory of unorganized

complexity is ultimately rooted in the laws of

thermodynamics.

In contrast, the fundamental

problem of today is that of organized complexity.

Concepts like those of organization, wholeness,

directiveness, teleolOSY: and differentiation

are alien to conventional physics.

However, they

pOp up everywhere in the biological, behavioral

and social sciences, and are, in fact, indispen-

sible for dealing with living organisms or

social groups.

Thus a basic problem posed to

modern science is a general theory of organiza-

tion.32

Ludwig von Bertalanffy did not get very far on his own in the for-

mulation of a general theory of organization.

But at least he pointed

to the need for one.

His call for such a theory has been heeded, it

seems, more by the social than by the natural scientists.

Indeed, among

members of the profession to which he himself belonged, the biologists,

Bertalanffy is regarded these days as something of an irresponsible

mystic.

He was characterized as such by some of the participants in

the Serbelloni conference on Reductionism in Biology.

127

At the root of the theory of unorganized complexity is the second

law of thermodynamics, the entropy law of physics whose implications

for biology and for the social sciences have puzzled so many philo-

sophers.

And the implications of the entropy law for organic systems

pose the third major source of misunderstanding about our simplified

typology of systems.

According to the second law of thermodynamics, which was discovered

by the French engineer Sadi Carnot in 1824 (and later refined and

restated by Rudolf Clausius, Lord Kelvin, and Ludwig Boltzmann), the

entropy of closed systems tends toward a maximum.

In the statistical

form of the law which was fOrmulated by Boltzmann, the assertion is

that particles in a closed system tend toward their most likely distri-

bution, which is one of randomness.

Paul Chambodal has remarked that "No other quantity has given rise

to so much argument as has entropy.

Likewise, of all physical quanti-

ties, entropy is the one most productive of paradoxes."33

The concept

of entropy is exceedingly slippery.

Explanations of the second law of

thermodynamics seem sensible enough at first.

Yet upon reflection

these explanations dissolve into ambiguities.

Following, for instance,

is a concise explanation of the entropy law by a staff writer for

Science magazine, Nicolas Wade.

Note that soon after he explains the

law Wade calls its logic into question.

The law is a broad, almost philOSOphical, concept

which has had many fOrmulations in its 110 year

history.

Central to all of them is the notion of

irreversibility, that certain processes go in one .

direction only and can never be repeated except

at far greater cost on the whole.

A given lump

of coal, for example, can be burned only once.

There is of course the same amount of energy in

the heat, smoke, and ashes as there was in the

lump of coal (that is stipulated by the first law

128

of thermodynamics governing the conservation of

matter-energy), but the energy bound up in the

combustion products is so dissipated that it is

unavailable for use, unlike the "free" energy in

the coal, and the process cannot be reversed.

Entropy is a measure of this bound or dissipated

energy.

The entropy law says that the entropy of

a closed system always increases, the change be-

ing from free energy to bound, not the other way

around.

Entropy is also a measure of disorderli-

ness (dissipated energy represents a more chaotic

situation than that before the lump of coal was

burned).

So the entropy law is also saying that

the natural state of things is to pass always

from order to disorder.

Whence the notion of

entropy as time's arrow. . . .

For a deep law of physics, the entropy law's dis-

tinction between free and bound, available and

unavailable energy may sound strangely anthro-

pomorphic.

And indeed it is anthropomorphic.

A

pure intellect would not comprehend the distinc-

tion:

it would just see energy shifting about.

The difference is important only to living organ-

isms, because they exist on the slope between low

entropy and high.

They absorb low entropy by

feeding, directly or indirectly, on sunlight, and

they give out high entropy in the form of waste

and heat.

Our subjective experience informs us that lumps of coal do not

regenerate themselves from ashes, smoke. and heat, but why is such a

recombination so very unlikely as never to have been observed? The

process of combustion, after all, can be interpreted as a reversible

chemical reaction.

Why does not our subjective experience include

instances in which heat energy flowed into the combustion chamber so

that water, carbon dioxide, and other residues could recombine to form

coal?

This is the conundrum of entropy.

Ludwig Boltzmann's answer to

the riddle was that a lump of coal is a highly ordered and therefore

highly improbably state of matter, and that the tendency of events is

toward a state of maximum probability, or randomness.

If that

assertion smacks of tautology or teleology, so be it.

Such is the

status of entropy theory.

129

Ludwig Boltzmann's critics pointed out that he had postulated a

law of irreversibility on the basis of fully time-reversible physical

laws.

In all the laws of mechanics which govern the motion of the gas

molecules which Boltzmann was discussing, the sign of the expression

for time, t, can be changed from positive to negative without affecting

the validity of the mechanical laws.

In response to these strenuous

objections to his entropy theory, Ludwig Boltzmann commited suicide.

The expression S = k log N, which was his formulation of the statisti-

cal value of entropy, S, is inscribed on his grave in the Central

Cemetery in Vienna.

A century later, Boltzmann is regarded as one of

the great physicists of the nineteenth century, and as one of the

founders of the field of statistical mechanics.

The inexorability of increasing entropy is a profoundly discourag-

ing idea.

The human mind seems to relish order in the universe.

The

thought that the cosmos is slowly, inevitably coming apart at the seams

is disheartening.

I recall that my roommate during my freshman year in

college, Frans van der Bogert, had read about the ascendancy of entropy

and the disintegration of order in Norbert Weiner's book Cybernetics

éfli Society.

During the required weekly physical education class, at

Swarthmore, we freshmen were told to run across a field, around a tree

at the far end, and back again.

Frans asked everybody in the P.E.

class to run around the tree in the same direction, so as to preserve

for a little while longer the sum of order in the universe.

He said it

was a gesture of existential defiance.

Herbert Spencer, the highly influential social philosopher. was

taken aback when he was informed of the implications of the second law

of thermodynamics.

Spencer formulated his basic ideas aboui; the

130

inevitability of progress in the early nineteenth century, when the

first law of thermodynamics dealing with the conservation of energy was

known.

In Spencer's sociological interpretation of physical ideas, the

first law together with processes of biological evolution led inevit-

ably to the establishment of perfection, order and universal happiness.

During the period in which Spencer was elaborating his ideas in his

major opus, First Principleg, Rudolf Clausius and Lord Kelvin announced

their*discoveries of the second law of thermodynamics, the entropy

law.

When it was pointed out to Spencer that the second law implies

that the ultimate state of the universe will be one of omnipresent

disorder and death, he replied, in a letter, "Regarding, as I have

done, equilibrium as the ultimate and highest state of seeiety, I have

assumed it to be not only the ultimate but also the highest state of

the universe.

And your assertion that when equilibrium was reached

life must cease staggered me....

I still feel unsettled about the

matter."35

Spencer resolved the contradiction, in the end, by brushing

the second law under the rug.

Like the British and American industrial

expansionists who lionized him for his theoretical justification of

laissez faigg, Spencer was unprepared to observe any malignancy in the

social organism.

The existence of biological organisms and of human organizations

would seem to contradict the implications of the second law of thermo-

dynamics.

Both organisms and organizations are highly complex entities,

and they often evolve towards states of higher order, or less random-

ness.

How can the evidence for increasing order in the biological and

organizational spheres be reconciled with the predictions of the

entropy law?

The paradox of order in the midst of entropy is one of the deepest

131

enigmas of modern science.

Astrophysicists ask themselves why stars

group themselves into galaxies.

Biologists inquire how protoplasm,

the stuff of life, could coalesce out of the primordial soup in the

early oceans of the earth.

I do not pretend to offer a full resolution

of this paradox.

I can only point to the ways in which some outstanding

contemporary scientists have addressed the problem, and draw some rough

inferences from their thought.

(Following up on Sir Arthur Eddington's remark that the increase

of entropy shows which way time's arrow is pointing, a Harvard astro-

physicist named David Layzer has made a distinction between two arrows

of time:

the historical arrow and the thermodynamic arrow.36 Historical

processes in the evolution of the universe, such as the formation of

galaxies, have transformed simple states into more complex states, and

have thereby generated order.

The thermodynamic arrow is the one which

Ludwig Boltzmann described. Layzer's illustration of the thermodynamic

arrow is less than galactic in scale.

He remarks that as a lump of

sugar dropped into a cup of tea dissolves, the macroscopic infbrmation,

or order, represented by the aggregation of the sugar molecules

gradually disappears; entropy increases.

These two arrows would seem to point in opposite directions; the

historical arrow toward increasing order, and the thermodynamic arrow

toward increasing randomness.

How can these two arrows of time coexist

in the same universe?

Through a complex argument centered on what he

terms the "strong cosmological principle," David Layzer argues that

as the universe has evolved from a hypothetical state of local thermo-

dynamic equilibrium both macroscopic information (the historical arrow)

and entropy (the thermodynamic arrow) could be generated simultaneously-

According to Layzer's estimate, the universe was in a state of local

132

thermodynamic equilibrium during the first fraction of a microsecond

after the big bang, the cosmological singularity.

In 1977 the Nobel Prize in Chemistry was won by a Belgian scientist,

Illya Prigogine, in recognition of his path—breaking contributions to

the theory of chemical thermodynamics.

Prigogine has been misinter-

preted by some reporters as having disproved the second law of thermo-

dynamics.37 Anyone who has read the lecture which Prigogine gave upon

the acceptance of the Nobel Prize would realize that that assertion is

erroneous.

Prigogine actually professed an exceptional reverence fer

the second law.

He compared it with Einstein's theory of relativity,

since both theories set absolute limits on physical experiments.

The

relativity theory asserts that it is impossible to propagate a signal

that travels faster than light, and the second law proves that a

perpetual motion machine cannot be constructed.

The misconception that Prigogine has thrown over the traces of the

entropy law arises from his contention that open systems which are not

in a state of thermodynamic equilibrium can generate order.

The fUnda-

mental assertion of the second law, by contrast, is that as closed

systems move toward a condition of thermodynamic equilibrium they

generate entr0py, or randomness.

Most of Prigogine's examples of such

non-equilibrium thermodynamics involve chemical reactions, but he is

cognizant of the possible application of his ideas to the explanation

of order in biological systems, and even in human organizations.38

The prevalence of open system non-equilibrium thermodynamics, Prigogine

suggests, is so widespread that nature seems more prone to developing

order than disorder.

Another Nobel Prize winner, the French biochemist,Jvaues Monod,

describes how biological systems evade the degenerative implications of?

133

the entropy law in transferring large amounts of genetic information at

the expenditure of miniscule amounts of energy.

Energy, after all, is

what thermodynamics is all about.

If energy expenditure can be con—

served, then thermodynamic dissipation can be minimized.39

While there are suggestive commonalities in the theories of

Layzer, Prigogine, and Monod (both Layzer and Prigogine, for instance,

explain how order arises in the context of non-equilibrium thermodynamic

processes), it is beyond my capability to integrate their ideas.

What

I would suggest, though, is that the root of a general theory of

organization -- such as Warren Weaver and Ludwig von Bertalanffy called

for several decades ago -- will be feund in a new and more general formu-

lation of the second law of thermodynamics.

Illya Prigogine has

advanced farther toward such a refbrmulation of the entropy law than

has anyone else.

But it appears that his ideas have yet to be tested

and refined by application to the ultimate intricacies of the macro-

molecular chemistry of genetics and biological regulation.

Once a

general thermodynamic theory of organization is discovered and

verified

in the biological case, I would expect that the theory will

have utility in the interpretation of human organizations as well.

One of the subtler implications of the theories advanced by

Prigogine and by Layzer is that they both result in the specification

of the direction of time.

In mathematical models of simple machines,

time is simply a geometric parameter.

In a simple machine, time can

run backwards just as well as forwards (-t can be substituted for t),

and like a reversed movie the model will still yield internally

consistent results.

We saw that in a machine with input some limita-

tions are imposed on the reversal of time.

In the non-equilibrium

thermodynamic theories of Layzer and Prigogine, the directionality of

134

time's unfolding through history becomes an essential property of their

models.

David Layzer's article on "The Arrow of Time" concludes with the

observation that "the intuitive perception of the world as unfblding in

time captures one of the most deep-seated properties of the universe."

According to Jacques Monod, the directionality of time has a

special significance for biological organisms.

Drawing upon his studies

of the molecular theory of the genetic code, Monod posits three pro-

perties which he says are essential and distinctive characteristics of

all living things.

The unfolding of the organism through time is

integral to all three of these closely interlinked properties.

- Teleonomy:

the property of an object's being endowed with a pur-

pose or project.

In their structure and in their*performances living

things act projectively, even if their behaVior is directed toward no more

than the most basic of biological projects:

to mature and multiply.

- Autonogous morphogenisis:

like crystals, living things take on a

form which is internally determined, rather than being shaped by outside

forces-

- Invariant reproduction:

the egg of a frog grows up to become

another frog, and not a bull.

Unlike the other two authors cited, Layzer and Prigogine,

Jacques Monod does not seek to explain the direction of time, thermo-

dynamics, and the creation of order and disorder within one over-

arChing theory.

I suspect that these three concepts -- time, thermo-

dynamics, and order -- will come to be united in a general theory of

organization, when such a theory is fOrmulated.

In the absence of a general theory of organization, the distinc-

tion in our typology between machines with input and organic systems

135

will have to remain imprecise.

It is conceivable that from the general

theory of organization could be derived attributes of organic systems

which would be more fundamental than those proposed by Kenneth Boulding

in 1956.

For example, it might be possible, in light of the general

theory, to stipulate that organic system models are characterized by

qualities such as the irreversibility of time.

I cannot defend that

proposition, since it is no more than a non-scientist's wild guess at

the intellectual future.

Yet even a layman can see the exciting

prospects for a great intellectual breakthrough.

Whoever it is that

formulates a general theory of organization would attain the same

lasting prominence in the biological sciences and in the study of human

organizations as have Newton and Einstein in physics.

The Boulding criteria, in the meantime, are a fair stand-in for

a theoretically rigorous definition of organic systems.

Among the

virtues of the Boulding criteria (the capacities for self-maintenance

and self-reproduction) is their direct applicability to the social

realm.

Talcott Parsons, for instance, has stated that the property of

pattern maintenance is a prerequisite to the functioning of all social

systems.

The match between Boulding's self-maintenance and Parson's

pattern maintenance is close, perhaps because both writers had the

biological example in mind.

System Control

If we turn our attention to problems of information and control,

it becomes apparent that the simplified systems typology of static

structures, simple machines, machines with input, and organic systems

is inadequate.

The distinction which Kenneth Boulding made, for

136

example, between clockwork and thermostatic closed systems is lost in

the simplified four level typology.

What is needed is a way of ampli-

fying the typology so as to allow for the categorization of different

kinds of systems control.

The control and direction of complex systems is a long-standing

concern in the field of organizational management.

In the literature

in this field, the best classification scheme of which I am aware is the

one which was developed by Robert N. Anthony of the Harvard Business

School.

Anthony and his research team read through the literature on

management and administration, extracted several hundred statements

about the nature of management, and sought to sort them out into various

typologies of management functions.

He experimented with, and rejected,

several standard methods of classification such as long range-short

range, top management-middle management, and planning-control.

The

categories and definitions which he finally settled upon are:

Qperational control is the process of assuring

that specific tasks are carried out effectively

and efficiently.

Management control is the process by which man-

agers assure that resources are obtained and

used effectively and efficiently in the accom-

plishment of the organization's objectives....

Strategic planning is the process of deciding

on objectives of the organization, on changes in

those objectives, on the resources used to

obtain those objectives, and on the policies

that are to goVern the vauisition, use, and

disposition of these resources.

An instructive comparison can be made between Anthony's,hierarchy

of management functions and the organizational stratification proposed

by Talcott Parsons.

In a chapter on "Some Ingredients of a General

Theory of Formal Organization,"l+1 Parsons proposed that there typically

137

are three levels in the hierarchical structure of organizations:

the

technical level, the managerial level, and the institutional level.

The primary function of the technical level is usually productive

activity; this is the level at which students are taught or automobiles

assembled.

The vauisition of resources and the disposal of products,

plus the management of personnel, are the major functions of the manager-

ial level.

At the institutional level, internal policies are set and

relationships of the organization to society at large are determined.

Parsons cites school boards and corporate boards of directors as

examples of subsystems at the institutional level.

The two uppermost levels of Parson's organizational classification

(the managerial level and the institutional level) bear a close resem-

blance to the top two categories in Anthony's typology (management

control and strategic planning).

There is no distinct place in the

Parsonian SCheme, however, for the lowest level of management, which

Anthony labelled the level of operational control.

In adopting

Anthony's, rather than Parson's, typology I am implicitly emphasizing

the importance of operational control, the sort of management exercised

by foreman in factories and by sergeants and lieutenants in the army.

‘A Complete Typology g; Systems

By laying out the four system levels of Figure 3-1 on the vertical

dimension, and Robert Anthony's categories of control as the horizontal

dimension, a complete typological matrix of systems can be created, as

is shown below:

138

Operating

Operational

management

Strategic

Plant

Control

Control

Planning

 

Static Structures

Simple Machines

Machines with Input  

Organic Systems

 

 

 

Figure 3~B:

A complete systems typology.

To test the validity of this typology, let us see how it matches

up with other prominent schemes for systems classification.

First, the

Boulding typology.

I have already mentioned the omission of machines

with input from Boulding's scheme: the oversight shows up clearly as a

blank row in Figure 3-9.

Operating

Operational

Management

Strategic

Plant

Control

Control

Planning

 

Static Structures

Frameworks

Simple Machines

Clockworks

Thermostats

................ t - - _ - - - - - - _ - - - - - - - _

Machines with Input  

Organic Systems

Cells

 

Plants

 

Animals

 

Humans

Figure 3-9:

Comparison with the Boulding typology.

The distinctions among cells, plants and animals implied by

Figure 3-9 are not entirely realistic.

Even organisms as simple as

cells exhibit some of the attributes of operational control, inasmuch

as they are capable of regulating their own metabolism.

Plants,

139

similarly, can exercise some control over the inflow and outflow of

resources (management control) through mechanisms such as the rolling

up of their leaves to reduce transpiration.

Part of the awkwardness of

the match between Boulding's typology and my own derives from the off-

handed way in which Boulding labelled his systemic categories.“2 He

indicates briefly, for example, that a distinguishing mark of the type

of system which he calls "plants" is a division of labor among the

parts.- A more appropriate label for this level of systemic organization

would be "organs."

The emergent significance of operational control at

this level of systemic complexity would then be more apparent, since

the separate tasks of different organs (leaves and roots, for example)

must be integrated into a purposeful whole.

Such is the function of

hormones in living systems, and of sergeants and lieutenants in an

army.

When it comes to his animal and human categories, Boulding's

criteria for classification focus on the information-handling capabilities

of the organism.

Nervous systems are not found in plants and reach

their highest development (on this planet, at least) in man.

Boulding

describes animals as responding to an image structure which intervenes

between stimulus and response, and he notes that man is capable of

using language and symbolism.

While these distinctions are valid in

their own right, Figure 3-9 suggests that animals may be distinguished

from plants (or from organs) and that man may be distinguished from

the other animals according to a different set of criteria.

What sets

animals apart is their ability to control their internal environments

and their ability to control their interactions, particularly

in

regards to the vauisition of food, with their external environments.

These are functions of management control.

By virtue of the refinement

iho

of his brain, man alone is capable of strategic planning.

Another instructive comparison can be made between the typological

matrix and different types of mathematical system models.

For the sake

of simplicity I have used linear continuous time examples in Figure

3-10, as in most of this chapter.

Operating

Operational

Management

Strategic

Plant

Control

Control

Planning

 

Static Structures

AX = K

Simple Machines

i = AX

gggisgik

................ h - - - - - - - - - - - - t - - - - -

Machines with Input

Organic Systems

 

i = AX + GU

gggiggfk

 

 

 

Figure 3—10: Comparison with mathematical models.

The absence of any entries in the last row of Figure 3-10 may

be more the consequence of my own ignorance than of a deficiency in

mathematics.

In his Nobel Prize lecture, Illya Prigogine develops his

ideas of non-equilibrium thermodynamics in mathematical terms (such

as Lyapounov functions, partial differential equations, and systemic

bifurcations) which are beyond my comprehension.

Prigogine makes a

tantalizing reference to applications of the theory of catastrophes

which has been developed by Rene Thom, a professor of mathematics at

the Paris Institute for Higher Scientific Studies.

Perhaps all this

does indeed constitute a mathematical language sufficient for the

description of organic systems.

My impression, though, is that the

mathematical techniques employed by Prigogine do not fall together

as one mathematical language, but are more nearly analogous to

141

isolated sentences.

To an outsider, at least, it looks as if the

mathematical methodology for modelling organic systems is still dis-

jointed.

Feedback and Operational Control

Kenneth Boulding distinguished his two categories of closed

dynamic systems, clockworks and thermostats, by the absence or’presence

of regulatory feedback.

Because they make use of feedback, simple machines

of the thermostatic type can adjust system variables toward an established

standard.

The same sort of distinction applies at the level of machines

with input, although feedback for systems of this order is usually in

reference to an input signal rather than to a fixed internal standard-

142

In the field of systems science as it has been elaborated by electrical

engineers, control systems are typically divided into two categories:

Open loop and closed loop.

Open loop systems would fall in the first

column of Figure 3-10, and closed 100p systems in the second column.

Feedback is a characteristic of closed loop control.

For definitions

of these terms, I rely upon Di Steffano gt al:

-- An open loop control system is one in which

the control action is independent of the out-

put.

-- A closed loop control system is one in which

the control action is somehow dependent on

the output.

-- Feedback is that property of a closed loop

system which permits the output (or some con—

trolled variable of the system) to be compared

with the input of the system (or an input to

some other internally situated component or

subsystem of the system) so that the appro-

priate control action may be formed as some

function of the output and input. 43

The idea of controlling a system without reference to its behavior

or output may seem odd.

Control in Open loop systems, according to

Di Steffano, is accomplished by calibrating the system so that the

desired input-output relationships obtain.

System instructions are all

pre—programmed, and cannot vary with changes in the systems environment.

A vivid example of how self-destructive pro-programmed instructions can

be has been provided by Henry A. Kissinger, who described the centrali-

zation of decision making in the Soviet armies in World War II.

Field regulations prescribed the exact location

'

of company commanders in the rear of their troops.

Field orders determined not only the direction of

the attack but the precise form it was to take.

The sphere of initiative of division commanders

could not have been more restricted.

Divisions

were prohibited from crossing divisional

143

boundaries, and commanders carried out this order

even if it meant the destruction of a neighboring

unit before their eyes.

Regiments advanced on

prescribed lines even into their own artillery

fire.

To deviate from orders was an offense

punishable by court-martial if it did not succeed.

To carry out even suicidal orders did not in-

volve any stigma444

Elementary Subsystems

At the beginning of this chapter I commented on the comparison

between Boulding's system typology and

Mendeleyev's chemical table

of the elements.

As it happens, the typological matrix I have constructed

does bear a resemblance to Mendeleyev's chart.

I would now like to

pursue that analogy further, and inquire whether there are some elemen—

tary system components whose number and arrangement determine the place-

ment of a system in the cells of the typological matrix.

In atoms, the

number of protons in the nucleus controls this placement:

one more

proton in the nucleus, and the element fits into the next cell in

Mendeleyev's table.

Several theorists have sought to analyze social systems into a

fundamental set of subsystems -- building blocks from which all social

systems are composed.

Talcott Parsons wrote of subsystems for adapta-

tion, integration, goal attainment, and pattern maintenance.

Shrode

and Voich have suggested that human organizations can be broken down

into five basic components:

structure, techniques and information,

people, purposes and objectives, and managementffl5

A similar paradigm

for the analysis of complex organizations has been proposed by Kast

and Rosenzweig,“6 who propose that organizations can be viewed as

composed of a structural subsystem, a technical subsystem, a psycho-

social subsystem, a goals and values subsystem, and a managerial

11m

. —-~‘- .

‘J‘

'

., ‘

‘

-

'

‘ .I..

'

‘i,

,.,.

.

V_

\

‘

-'

‘ ENVIRONMENTAL

'

% _'

a JUPRAJYJI'EM '

*

gE?

 

60"“ “’0 W2304’“ 7£CWICAL

‘43:; '-

'JUBJflTFM

 

\

sues/:72»:

\

l

1'

X

I

I]

l

\

I
/
/

\

 

 

-.

‘

”Va/050cm;

swoon/4.4

‘

‘

Joann/w

 

:uafirew

   

‘

‘

‘

's

g

.

'

'

"-‘,

.‘

.

y

.

>

-

~v

.v

.. .w

.

-~.‘x.

~-

.

r.

‘o—-

'

.

.

.

’

.

- h

‘

.-

.

‘

.

.

,‘s-

_

.

"
.

,

'

._

.

-
.

p

.

7

..

.

I

i

J‘

I:

.

._

'

'

u.

'

.

-

..

.

‘

‘

‘

.

.

'

-‘
’

.

'

I

'1

.

‘

.

_-.

,

.

‘

.

.

.

‘.

_

-

.

.1

.

V

,

..

,

i

.

.

.

.

,

'

.

.

..

r

.

.

_

‘

.«

..

'

_

.

a

.

.

MAN/16511.44. 505: first»: —i

<3

...  

C>‘ mpur- OUTPUT :40 W or

mun/,4 4:, amour, Ame

lNF’OR/MA r/cw.

#76 0!! 5. // M57 awn [OJeA/ZWEI 63 PARADmm

o; ORGAN/2.4 f/ONAL JUJS/thma‘.

 

1H5

subsystem.

The diagram in which they present their paradigm is

reproduced in Figure 3-11.

The diagram bears an eerie resemblance to

a science textbook portrait of an atomic nucleus.

The diagram of interlocking circles in Figure 3-11 is a simple one,

and excepting the arrows, might have been produced at random by a child

playing with a compass.

The virtue of the diagram, however, is twofold:

a) each of the circles (subsystems) can be separately linked to a

distinct school of thought in organization theory: and b) the circles

taken together allow for a fairly complete description of an open

organizational system.

There is cause for confidence, therefore, that

the five organizational subsystems in Kast and Rosenzweig's scheme

include a place for the consideration of most factors of importance in

organizational analysis.

The multi-circled paradigm can be used as Kast and Rosenzweig put

it, as

an aid to understanding the evolution of organi-

zation theory.

Traditional management theory

emphasized the structural and managerial sub-

systems and was concerned with developing

principles.

The human relationists and behavioral

scientists emphasized the psychosocial subsystem

and focused their attention on motivation, group

dynamics, and other related factors.

The manage-

ment science school emphasized the technical sube

system and methods for quantifying decision-making

and control processes.

Thus each approach to

organization and management has tended to emphasize

particular subsystems, with little recognition of

the importance of the others.

The modern approach

views the organization as an open, sociotechnical

system and considers all the primary subsystems

egg their interactionEZ-Tt?

Notable for its omission from this passage is reference to the

goals and values subsystem, which was shown in Figure 3-11 as a basic

component of the organizational system.

Indeed, throughout their text

146

Kast and Rosenzweig show an ambivalent attitude toward organizational

goals and values.

They sometimes lean toward the idea that organiza-

tional goals are established by management, and then hastily chastise

themselves with references to the work of Etzioni, Perrow, and other

modern writers who have asserted that official top-level organizational

goals may have little bearing on the actual behavior of an organization.

my inclination is to eliminate the circle representing the goals

and values subsystem from the paradigm shown in Figure 3-11.

Those

organizational goals which are in truth established by management could

then be reviewed in the context of the managerial subsystem.

Kast and

Rosenzweig themselves incline toward this analytic methodology, in that

they write of strategic, coordinative, and operating organizational

goals, each set at a different level of management.

Some of the actual

goals of an organization may well arise from the orientation of the

rank and file, and it seems to me that such goals could be studied

within the framework of the psychosocial subsystem.

The values of

organizational participants, which Kast and Rosenzweig lump together _

with goals, would fit quite readily into the psychosocial classification.

The cavalier manner in which I have subtracted a subsystem from

the Kast and Rosenzweig paradigm may trouble the thoughtful reader.

If goals and values were so important to Kast and Rosenzweig as to

merit the status of a separate subsystem, then how can such a signifi-

cant classification be abolished?

And is that all that a subsystem is,

anyway:

just a classification?

Then why not call it a group or a

family, in the manner of biological classification, rather than a

subsystem?

I define a subsystem as a set of two or more interdependent

elements of a system that interact according to definable rules.

A

147

”ll/C 72/314

     

3063/! 7‘47”

\\ //

      

/>‘\

\

/

    

/

\

    

/

\

  

\

\ , MAN/i case/AL

/

\

\

     

\/\ 5035/375M

p/

    \~___.-—’/\
  

  

I \

PSXCA/OJOCML

  

flail/W044

   

sass/57am

Juuflrem

     

A76 UR! 3- /z A area/VD We»! 0; 7w: ORGAN/:24 770A!

.45 corAPI/sto OF 54/5er 75/91:.

  

 

  

1H8

system, similarly, is an organized, unitary whole consisting of two or

more interdependent parts, components, or subSystems that interact

according to definable rules, and which can be delineated by identi-

fiable boundaries from its environment.

With the addition of the

phrase "that interact according to definable rules" this is the same

definition of a system as is offered by Kast and RosenzweingB

It was with these definitions in mind that I removed the goals and

values subsystem from Kast and Rosenzweig's five-circled paradigm.

From a reading of their text, it is not clear to me which are the inter-

dependent elements in what they describe as the goals and values sub-

system.

For that matter, Kast and Rosenzweig never explain what they

mean by the term "subsystem."

They use the word, apparently, more for

the sake of trendy embellishment than for clarification.

I shall try

to be more precise.

What, then, is the systemic substance of Kast and Rosenzweig's

psychosocial subsystem?

It turns out that the psychosocial subsystem is

as much a chimera as the goals and values subsystem.

Kast and Rosenzweig

describe the psychosocial subsystem in these terms:

Every organization has a psychosocial subsystem

which is composed of individuals and groups in

interaction.

It consists of individual behavior

and motivation, status and role relationships,

group dynamics, and influence systems.

It is

also affected by sentiments, values, attitudes,

expectations and aspirations of the pe0ple in

the organization.

Obviously, the psychosocial

subsystem is affected by external environmental

forces as well as by the tasks, technology. and

structure of the internal organization.49

It appears that the psychosocial subsystem in Kast and Rosenzweig's

book serves the function of a theoretical attic in which they stash all

of the factors which they could not place elsewhere in their schema for

149

organizational analysis.

If Kast and Rosenzweig's definition of the

psychosocial subsystem were accepted, it would be next to impossible

to fOrmulate the rules which govern the behavior of this heterogeneous

array.

In the absence of such rules, the concept of a subsystem does

not apply.

A more coherent definition of this subsystem can be derived from

the theory of organic systems which was introduced earlier.

The unique

ordering capabilities of organic systems, it was argued, lie in their

ability to sustain their systemic integrity and to reproduce them-

selves.

If that integrity were indeed reflected in a mathematical

model such as

X = AX + GU, then the ordering ability of the system

would be indicated by its capacity to keep its internal processes in

step with that model.

The system would sustain order and resist

entropy.

Many organization theorists have been misled by the classical

formulation of the August second law of thermodynamics, and have

therefore contended that human organizations resist entrOpy simply by

importing energy, or some relevant equivalent.

Kast and Rosenzweig,

for instance, assert that "...in the Open biological or social system,

entropy can be arrested and may even be transformed to negative

entropy -- a process of more complete organization and ability to

transform resources.

This is possible because in open systems the

resources (material, energy and information) utilized to arrest the

entropy process are imported from the external environment."50

Many of the ordering processes in biological systems and in human

organizations have little to do with system inputs.

Genetic instructions

encoded in molecules of DNA are quite internal to the organism.

Status

and doctrine, which are two common ordering factors in organizations,

150

can be independent of inputs.

There is no doubt that organizations can

use a portion of their throughput for the maintenance of order -- the

distribution of salaries being a good example.

But inputs are not the

sole source of order.

In the biological case order is indeed purchased through the

consumption of inputs of energy in the form of food.

Through the

transformation of food, which involves the dissipation of heat and a

net increase in entropy, the organism can construct new macromolecules,

cells, and organs to replace those which have been worn down, and can

thereby resist the tendancy toward degeneration.

Local entropy within

the organism is controlled at the expense of increasing entropy in

the organism plus its environment.

All of this is true, but it

represents only a small part of the story, not including the part

which is most relevant to the theory of human organizations.

The mechanisms which regulate the thermodynamic processes of

biological organisms are themselves almost independent of thermo-

dynamics.

Complex proteins such as enzymes or DNA alter their own

energy states only to a miniscule degree when they serve as catalysts

or templates for metabolic or reproductive reactions.

The change in

thermodynamic potential in the chemical process governed by an enzyme

may be large, despite the small expenditure of energy by the enzyme

itself.

Jvaues Monod has compared this situation to that of an

electronic relay which with a tiny throughput of energy can fire a

ballistic missile.

The ultimate instance of the independance of regulatory,processes

from systemic inputs is that of the genetic code inscribed in molecules

of deoxyribonucleic acid (DNA).

The code is completely, hermitically

isolated from outside influences.

Information is always transmitted

151

from DNA to the recipient protein, never in the opposite direction.

The sequence of nucleotides in a segment of DNA is utterly impervious

to external forces.

In this one-way system for the transmission of

genetic information -- which works entirely from the "inside" out,

and never from the "outside" in -- we see the explanation for the extra-

ordinary stability of species which have unalterably reproduced them-

selves for hundreds of millions of years.51

When it comes to the mechanisms for maintaining internal order,

there is a close analogy between biological organisms and human organ-

izations.

Organizations can maintain themselves through ideologies and

status systems which may be largely internal and independent of inputs.

An organization with a well-articulated and widely-shared system of

beliefs and expectations is like an organism with its genetic code.

The ideology can shape and control the behavior of the organization's

members and can be (although it does not have to be) pretty much

impervious to influences external to the organization.

And a system of

beliefs about the nature and purpose of an organization, like a genetic

code, can be the key to organizational reproduction.

The psychosocial subsystem as described by Kast and Rosenzweig

wOUld seem to contribute to the order of interpersonal relationships in

an organization.

They assert that the forces involved in the psycho-

social subsystem "set the 'organizational climate' within which the

human participants perfbrm their roles and activities."

In organiza-

52

tions where that climate is uniform, we might anticipate greater

152

uniformity, or order, in the behavior of the participants.

Yet here we

come upon a paradox.

For many of the attributes of the psychosocial

subsystem which are mentioned by Kast and Rosenzweig, such as status

and role relationships, can exert a destabilizing influence as much as

a stabilizing one.

We saw in the last chapter, for instance, that the

status of the big farmers in Comilla thana, and their traditional roles

as moneylenders, were both incompatible with the purposes of the

cooperatives.

Although the procedure is not entirely satisfactory, I will resolve

this dilemma by means of a definition.

In this volume: the psychosocial

subsystem will be defined as phgp_§pp_pfqp§;iefs and practices which

ppstain functional order ;p_the working relationships 9; the organiza-

tion's participants.

This may appear to be an excessively restricted,

almost a militaristic, definition.

Where in this construct is there

room for all the idiosyncracies of motivation and behavior which make

organizational psychology such an intriguing field of study?

Where,

indeed, is there room for the individual variations, from greed to

altruism, which can be identified in all people, and which are sometimes

accentuated in organizations?

By adopting a narrower view of the psychosocial subsystem than

that which Kast and Rosenzweig allow, we have effected a trade-off

between scope and rigor.

The finer the brush, the less canvas can be

painted.

The precise definition accords well with the general theory

of systems which has been advanced in this chapter.

The ordering

function of the psychosocial subsystem can readily be compared with the

metabolic regulation accomplished by enzymes.

On the other hand, there

is a great deal of obstructionism and maneuvering in any organization

which no longer seems to have a place in our paradigm.

153

E/V V/iO/V/‘A ENT 0" THE ORGAN/2A T/O/V

371A 7156/6 PLAN/VINO -

'

MNAGEM EN 7' C ON 73¢OL

095M 7/OA’44 CO/V7".004

 

mmuz:

Q

> paw,

L

04/7707”?

>

 

. I

I‘

\\

‘\

 

AOUNMR/a

IN raw/0N OF

5N V/A'cwmszv 7:44 FORCE:

F/GUIE .3 ‘ /3

77/5 OROAN/ZA 7'IO/V A: A CO/‘IIPOJ'H’! OF

Lil/€43 OF CONTROL

151+

Another intriguing diagram presented by Kast and Rosenzweig shows

the management of an organization as consisting of concentric layers

through which inputs and outputs pass.‘53

A modified version of that

diagram is shown in Figure 3-13.

Pursuing, for a while, the atomic

analogy, I would remark that Figure 3—13 shows a distant resemblance

to a science textbook drawing of the orbits of electrons about an atomic

nucleus.

In explanation of the similar diagram in their book, which they

say is an extension of the work of Talcott Parsons, Thomas.A. Petit,

and James D. Thompson, Kast and Rosenzweig comment on the differences

in the orientation of management at the different levels of their

onion-like structure.

The innermost layers of the organization are

protected from the vagaries of the environment so that the core pro-

ductive technology of the organization can function in a stable, rational,

programable setting.

Management at the outermost layer, that of

strategic planning, has a perspective that is flexible, judgemental,

and non-programable.

Strategic management has a long run view of time;

operational control a short run view.

Strategic managers emphasize

satisficing, while operational controllers concentrate on optimizing.

A superior paradigm for organizational analysis can be constructed

by combining Figures 3—12 and 3-13.

The explanatory power inherent in

both diagrams can thereby be compressed into one diagram.

The result

of such a combination is displayed in Figure 3—14.

Because the paradigm shown in Figure 3-14 is so inclusive -- incor-

porating as it does most of the ideas displayed by Kast and Rosenzweig

in two diagrams -- the gain in explanatory power is offset by an

increase in complexity.

The paradigm may be more readily understood

if it is seen in an exploded view, such as Figure 3-15.

As much for the

155

EN V/RON/‘AEN7 OF 7W5 ORGAN/24 7'/O/V

MANAGi/‘A exvr

com/7204

3055/: 75M

apex/4 rave

PLANT

  

  

INPU 7:!

0071007:

\

I

I

\

\

 

/

I

l

I

I

/

373,4 rte/c pun/MM;

Jugs/.9 7.5M

0,0564 T/O/VAL CONTROL Juasfifsm

Rife/{0:0 47.44

.5055)“: rem

rte/imam

Jun/175,14

15760:: 3- /4

77/! ORGAN/2'1 770A! A: A COMPOJ/ 7'!

a; SUI: 7373M!

156

.9770! 756/:

PLANNING

Jun: firs/M

-

 

I

IMAM/A 65M ewr \'

at.

ii’ZS’im—m \

 

I

I

 
.

I

OPERA 770M“.

CONTROL \\

" Juasflfem

Ne

 

‘

 

37:41:70,014

|
Juaswrem\ '

 

 

'i

l

_

 

 

 

'

,

 

l

l

 

\

 

I

|

 

J

 

: W

 

J

 

"ass/7:44”

5053/4761”:

”Mt/cc 5A: "’

305473771913

Aime/030cm;

.905: r: raw

TiCflNICAL

Just/dram

OPERA TIA/6

PLANT

”GU35" 345'

lXPLODED Wen! Of 77/5 ORGAN/2A T/ONAL

P4RADIG/‘A .

sake of amusement as of edification, I have carried the atomic analogy for-

157

ward. in Figure 3-15, and have labeled the structural subsystem, the

technical subsystem, and the psychosocial subsystem as "nuclear"

subsystems, and the control subsystems as "orbital" subsystems.

The

significance of this analogy should not be exaggerated.

Yet it does

suggest -- and I think correctly -- that the core components of an

organization are its structural, technical, and psychosocial subsystems.

And the very core of the core -- the operating plant of the organiza-

tion -- is to be found in the region where these three nuclear sub-

systems interact.

A conceptualization of the operating plant would be

incomplete if it did not take all three nuclear subsystems into

account.

In Figure 3-16 I have indicated what I conceive to be the relation-

ship between these multi-circled analytic paradigms and the general

typology of systems.

"S" in Figure 3—16 stands for structural subsystem,

"T" for technical subsystem, "E" for environmental subsystem, "P" for

psychosocial subsystem, "O" for operational control subsystem, "M" for

management control subsystem, and "R" for strategic planning subsystem.

Operating

Operational

Management

Strategic

Plant

Control

Control

Planning

 

Static Systems

Simple Machines

S

ST

Machines with Input

Organic Systems

STE

STEP

 

 

STO

STEO

STEPO

STEOM

STEPOM

 

 

STEPOMR

Figure 3-16:

Subsystem components of organizational systems.

Here at last is a firm link between organization theory and general

systems theory.

Like the chemical table of the elements, Figure 3-16

158

indicates that different kinds of systems are built up out of different

combinations of elementary components.

(The analogy which I drew above

between nuclear and orbital subsystems does not carry through in this

instance.)

I very much doubt that computers, for example, will ever

exhibit the characteristics of true strategic planning.

The farthest

that computers can go, I suspect, is to the STEOM classification.

So

the last column of the row for machines with input is blank.

On

similar grounds, I would doubt that any of the other blank cells in

Figure 3-16 could be filled in with actual examples.

Different combinations of elementary organizational subsystems

create different kinds of organizations.

The most complex and sophisti-

cated organizations, such as large industrial corporations or modern

governments, contain all of the elementary subsystems: they would fit

the STEPOMR classification.

I would venture to say that no collection

of human beings could be called an organization, however, if it were

not an open system, and if it did not include the basic nucleus of

structural, technical, and psychosocial subsystems.

To put the matter

another way, I would posit that all human organizations are organic

systems, and are appropriately classified in one of the four cells in

the bottom row of Figure 3-16.

For the purposes of this book, the utility of Figure 3-16 is in

the order which it lends to organizational analysis.

Just as Kast and

Rosenzweig's book is structured around the paradigm shown in Figure 3-11,

the remaining chapters of this book are structured around Figure 3-16.

As an ordering device, the typological matrix of Figure 3-16 is superior

to the multi-circled paradigm shown in Figures 3—14 and 3-15.

An

appropriate sequence of presentation, building up from simple to complex,

can be inferred from the matrix.

Choosing which circle (subsystem) in

159

Figure 3-14 to describe first, on the other hand, is quite arbitrary.

In terms of the systems typology, the fourth through eighth chapters of

this book can be related to one another as in Figure 3-17.

Operational

Operational

Management

Strategic

Plant

Control

Control

Planning

 

Static Systems

Simple Machines

Machines With Input

Organic Systems

 

4

5

6

7

 

8

 

8

 

8

Figure 3-17:

Chapters in the context of the systems typology.

I remarked in Chapter 2 that the territorial hierarchy of admini-

stration has been of fundamental importance in the design and operation

of rural organizations in South Asia, including those intended for

administration and those intended for development.

Whereas in complex

industrial organizations in developed countries geography is often used

as a basis for horizontal departmentalization -- the west coast branch

or the European division, for example-~in rural organizations in

developing countries, and especially in South Asia, geography is used

as the basis for the establishment of a vertical hierarchy of authority.

In defference to the significance of the territorial hierarchy, in my

analysis of the Comilla cooPeratives I will portray the three nuclear

subsystems shown in Figure 3-15 at the village and at the thana level.

When it comes to describing the three orbital subsystems, those for

operational control, management control, and strategic planning, I will

modify the strict division of the text into village level and thana

level aspects of the cooperative organization.

One further modification needs to be introduced so as to make the

160

 

 

  

 

 

771M!

#IOUIé' Jo /6’

were-mm oat/e4women/7'

161

systems analytic framework suitable for the purposes of this book.

I

remarked in Chapter 1 that I had set out to discuss the evolution of the

rural development organizations in Comilla. and especially of the rural

cooperatives, during the decade of the nineteen sixties.

Yet the analytic

paradigm displayed in Figures 3-13 and 3—14vis obviously inadequate for

such a purpose, because it is static.

Figure 3-14 resembles a cross-

section of the trunk of a many-holed cypress tree.

For an historical

study, what is needed is a way of representing how the tree has grown

and changed its shape and proportions over time.

By introducing the territorial hierarchy as a vertical dimension of

analysis, I have already added a third dimension to the two-dimensional

organizational cross-section of Figure 3—14.

To show changes in the

model over time, a fourth dimension would have to be embodied in the

model.

While a true four dimensional representation is of course

impossible on a two dimensional page, the influence of time can be

shown schematically as in Figure 3-18, which portrays the three dimen-

sional organizational model moving along a time line.

The processes of institution-building for rural deve10pment

typically takesmany years to unfold.

In a speech at Michigan State

University in 1966, Akhter Hameed Khan described the significance of

time.

...building institutions is not like erecting a

building where you just have a plan, raise the

walls, and then put the roof on.

Actually,

nothing goes according to the plan.

The analogy

would be much nearer to the planting of an or-

chard, where the seed is put in and watered, and

then you just wait.

You can't do a thing.

You

can't push the germination process, and you can't

push the growth of the trees.

Sometimes the seed

may not germinate at all: or the plant may be

killed by disease, insects, and so on.

So there

is a great deal of waste, and a great deal of

162

patience is required to see these institutions

grow up.

Building institutions is not simply a

matter of writing a constitution or by-laws.

It's really a question of creating loyalty, and

loyalty is created only by time and testing.

We

cannot claim loyalty, but it will come after

these institutions have been tested.

We have to

wait; we have to be very patient.

There has been

some progress in the last five or six years, but

what we really need is sixty years, not just

five or six years.54

A half century might, as Akhter Hameed Khan suggests, approximate

the appropriate span of time in which to evaluate a rural development

organization.

This study of the Camilla rural development organizations

is nonetheless limited to a much shorter period, to the ten years from

1960 to 1970.

The decade of the nineteen sixties included the golden

era of the Comilla cooperatives (the period 1964-67), but that is not

the reason for the selection of the sixties as the period for investi-

gation.

As we shall see, several unfavorable trends came to light in

the late sixties, so the period balances the bad with the good.

The

reason for limiting this study to the first decade of the Comilla experi-

ment is that during those ten years the background economic and political

conditions in East Pakistan were relatively stable, whereas they

changed markedly during and after the upheavals which created

Bangladesh.

Notes for Chapter 3

A Systems Approach to Organization Theory

1 Richard A. Johnson, Fremont E. Kast and James F. Rosenzweig,

The Theory _apd Management pf Systems, McGraw Hill, New Yofii, 1967,

p. 64.

Copyright 1974.

Used with permission Of McGraw Hill Book

Company.

2 William G. Scott, "Organization theory:

an overview and an

appraisal," in Frank Baker, ed., Organipgpionp;_Systemp, General Systems

Approaches pp Complex Orggngations, Richard D. Irwin, Homewood,

Illinois, 1973. P. 113.

3 James G. Miller, "The Nature of Living Systems," in Frank Baker,

ed. , Organizational Systems, General Systems Approaches pp Complex

Orgapizatiopp, Richard D. Irwin, Homewood, Illinois, 1973, p. 48.

Reprinted with permission.

14' Ludwig von Bertalanffy, "The Meaning Of General System Theory,"

reprinted as Chapter 2 of Ludwig von Bertalanffy, General System Theogy,

Foundations, Development, Appliggtions, George Braziller, New York,

1968, pp. 35, 36.

Reprinted with permission.

5 David Berlinski, 9;; Systems Analysis:

An_ Essfiay Concerning the

Lillitgtions 9: Some Mathematical Methods _ig the Social, Political, fl

Biological Sciences, The M.I.T. Press, Cambridge, Massachusetts, 1976.

6 von Bertalanffy, 9p pip, pp. 56—63.

7 Berlinski, gpfli, p. 4.

8 Kenneth Boulding, "General Systems Theory, the Skeleton of

Science," Management Science, April, 1956, pp. 197-208.

9 Fremont E. Kast and James E. Rosenzweig, Orgaggation _a_nd M__a._n_ag_-

ment, a Systems Approach, McGraw Hill, New York, 1974, second edition,

pp. 101, 102.

Copyright 1974 by McGraw Hill Book Company.

Used with

permission of McGraw Hill Book Company.

10 Boulding, pp_p;p, p. 205.

11 Ibid., p. 205.

12 Ibid., p. 206.

163

164

13 Ibid., p. 202.

1“ Ibid., p. 202.

15 The meaning of the term "simple machine" is the same as what

W. Ross Ashby intends by his phrase "determinate machine."

16 Boulding, 9p cit, pp. 202, 203.

17 John E. Hunter, "Mathematical Models and the Cybernetic

Equation," (mimeographed), p. 4.

18 John Hunter has posed a counter example to refute the univer-

sality of these observations about the reversibility of time in closed

system dynamic models.

It is possible for two trajectories Of such a

model to converge and form one trajectory, like the two branches of a Y

joining to fOrm the stem.

If time is reversed in such a configuration,

the trajectories split from the stem to the branches Of the Y.

The

implication of such a split is that the model is not state determined,

and therefore does not conform to the definition of a simple machine.

19 David.Layzer, "The Arrow of Time," Scientific American,

December, 1975’ PI 69.

20 Adapted from class notes for’PsychOlogy 820, "Mathematical

Models in Psychology," Michigan State University, Fall, 1975.

The for-

mat and symbolism in this definition have been altered to parallel

Hunter's definition of a simple machine, quoted above.

21 More technically, the expression eA(t- T) is a weighting

function.

It plays a role in the forced response analogous to that of

the expression eAt in the free response.

22 For a mathematical demonstration of this point, see Appendix A.

23 There are exceptions to this pronouncement.

The open system

model with an undelayed step input, illustrated as a phase plane in

Figure 3-4, is a good example of the exceptions.

2“ Walter F. Buckley, Sociology gag Modern Systems Theopy, Prentice

Hall, Englewood Cliffs, New Jersey, 1967, p. 50.

Reprinted with

permission.

25 Kast and Rosenzweig, 1974, pp pip, p. 102.

Used with permission

of McGraw Hill Book Company.

26 Francisco Jose Ayala and Theodosius Dobzhansky, eds.,

Studies

in the Philosophy of Biology, Reduction and Related Problems, University

 

of CalifOrnia Press, Berkeley, 1974.

27 Peter Medawar, "A Geometric Model Of Reduction and Emergence,"

in Ayala and Dobzhansky, eds., pp pip, pp. 57-63.

Medawar's paper

includes a layman's interpretation of tOpOlogy which I found illuminating.

If a closed figure such as a circle is drawn on a sheet of flexible

165

rubber, the figure will preserve a meaningful inside and outside, but

the regular shape of the circle will be destroyed.

NO matter how the

rubber is stretched, and so long as it is not torn, there will be a

continuous point-to-point transformation between the original circle and

the new shape on the surface of the rubber.

Topology, Medawar asserts,

is the study of such continuous one-to-one mappings in space.

8 Giuseppe MOntalenti, "From Aristotle to Democritus via Darwin,"

in Ayala and Dobzhansky, pp_cit, p. 12.

29 Jacques Monod, Chance and Necessity, Ag Essay pp the Natural

Philosophy pf Modern Biology, Vintage Books, New York, 1971.

3° 5. E. Luria, Life, the Unfinished Experiment, Charles Scribners and

Sons, New York, 1973.

31 Solomon Lefschetz, Differential Equationg, Geometrig Theopy,

Interscience Publishers, New York, 1957, p. 239.

32 von Bertalanffy, pp cit, p. 34.

Reprinted with permission.

33 Paul Chambodal, Paradoxes 9: Physics, Transworld Publishers,

London, 1971, p. 35.

34 Nicholas Wade, "Nicholas Georescu-Roegen:

Entropy the Measure

of Economic Man," Science, 31 October 1975. V. 190, p. 448.

Copyright 1975

by the American Association for the Advancement of Science.

Reprinted

with permission.

35 Quoted in Cynthia Eagle Russett, The Concept pfiqE uilibrium pp

American Social Thought, Yale University Press, New Haven, 19

, p. 39.

36 Layzer, pp cit.

37 Robert C. Cowen, "Challenges to the Second Law:

Nonsense,"

The Christian Science Monitor, June 20, 1979.

38 Ilya Prigogine, "Time, Structure, and Fluctuations," Science,

v. 201, NO. 4358, September 1, 1978. pp. 777-785.

39 MOnOd, pp g;p, pp. 68, 69.

40 Robert N. Anthony, Planning gag Control Systems, A Expmework for

Analysis, Harvard Graduate SChOOl Of Business Administration, Boston,

1956, pp. 16-18.

Reprinted with permission.

41 Talcott Parsons, Structure and Process ;p_MOdern Societies,

The Free Press, Glencoe, 1963, pp. 59-96.

2 In evolutionary terms, plants are conceived of as having

developed in a different direction from animals.

The way Boulding has

ordered his typology, by contrast, makes plants appear more primitive

than animals.

#3 Di Steffano, Qp_9ip, p. 3.

Used with permission Of McGraw Hill

Book Company.

166

41+ Henry A. Kissinger, Nuclear Weapons a_n_c_1_ Forei

Policy,

Doubleday Anchor Books, Garden City, 1958. p. 1 3.

[+5 William A. Schrode and Dan Voich, Jr. , Organization g3;

 

Management:

Basic System; Concepts, Richard D. Irwin, Homewood,

Illinois, 1974, p. 506.

46 Kast and Rosenzweig, 9p 9;, p. 112.

1+7 Ipfpd. , p. 113.

Used with permission of McGraw Hill Book Company.

“8 Ibid., p. 101.

1+9 Ibid., p. 111.

5° Ibid., p. 115.

Used with permission of McGraw Hill Book Company.

51

52

Monod, pp 313, p. 110, 111.

Kast and Rosenzweig, _O_p pip, p. 111.

53 _I_p;d_., p. 120.

54 Akhter Hameed Khan, Communipy and Agrimllturagl Development _i_r_1_

Pakistan, Asian Studies Center, Michigan State University, East Lansing,

1989! P0 310

CHAPTER 4

The Structural Subsystem

Operating

Operational

Management

Strategic

plant

control

control

planning

Static systems

S

Simple machines

Machines with input

Organic systems

Traditionally, organizations are described by

organization charts.

An organization chart

specifies the authority or reportorial structure

of the system.

Although it is subject to fre-

quent private jokes, considerable scorn on the

part of sophisticated Observers, and dubious

championing by archaic organizational architects,

the organization chart communicates some Of the

most important attributes of the system.

It

usually errs by not reflecting the nuances of

relationships within the organization: it usually

deals poorly with informal control and informal

authority, usually underestimates the role of

personality variables in molding the actual sys-

tem, and usually exaggerates the isomorphism

between the authority system and the communication

system.

Nevertheless, the organization chart

.

still provides a lot of information conveniently --

partly because the organization usually has come

to consider relationships in terms of the chart.1

167

Introduction

 

168

Systemic architecture, or the established pattern of relationships

among the component parts of an organization, is the proper focus of

investigation when we consider the structural subsystem.

Unlike

animals, which have well defined skeletons, organizations do not

typically have a single describable pattern Of internal structure.

Top

management may conceive of the significant internal patterns as follow-

ing an organization chart.

Middle level managers who get things done

may be more concerned with tracing patterns of personal influence --

patterns which as often as not cries-cross and short-circuit the formal

hierarchy.

Production personnel may think Of the organization in a more

linear, horizontal fashion, as a chain Of work stations.

Each of these

categories of personnel may be thought Of as having their own sociolo-

gical equivalent Of an x-ray; on the emulsion of their minds, they each

burn a different image of the bone structure of the organization.

My discussion will be biased toward the organization chart image

of systemic architecture.

For all its omissions and limitations, the

organization chart does summarize much that is crucial in the under-

standing Of formal organizations.

An organization chart can give clues,

for instance, about how activities are differentiated and coordinated,

and about channels Of authority.

For practical administrators, insti-

tution—builders, and organization theorists alike, organization charts

stand in the same fundamental relationship to the comprehension Of

modern organizations as does the Copernican model Of the solar system to

the understanding Of the movement of the planets around the'sun.

To the degree that an organization chart or some such model can

be fitted to an organization, moreover, the model defines the rules of

169

established relationships which allow us to characterize the structure

of the organization as a subsystem.

Without such rules, the relation-

ships among the members of the group, though they may have form at any

given time, cannot be said to comprise a subsystem.

Many significant

human groups, such as crowds and friendly gatherings, are of this

ameboid sort.

If a group does not have a structural subsystem,

it does not have the most elementary characteristic Of a system, much

less of an organization.

The study Of organizational structure, and Of the charts which

delineate that structure, is just as significant for the understanding

Of rural deve10pment organizations as it is for'any other class of

organizations.

The first of Norman Uphoff and Milton Esman's seven

design principles for rural development institutions, cited in

chapter two, was a structural prescription.

They recommended that

"Local institutions should have more than one leyp; pi organization,

probably a two-tier pattern, in which the lower tier performs functions

at the neighborhood or small group level, while the other undertakes

more complex business activities that require relatively large scale

Operations."

The idea of tiers of organization is not common in the literature

of organization theory, where the unified bureaucratic hierarchy is

the conceptual and empirical norm.

A concept which is frequently

encountered in the literature on organization theory is that of levels

in a hierarchy Of authority.

Levels should be distinguished, however,

from tiers.

In the quotation above Uphoff and Esman use the words

synonymously ("...more than one lpypl of organization, probably a two

pig; pattern..."), yet this practice can lead to confusion.

A level,

as the term is customarily employed in the organization theory

170

 

 

 

2

l
e
v
e
L

1
'

I

  

 

 

l

l

 

 

1

l
e
v
e
L

I

L

 

 

  

!
”
n
I

l
l
'
l
l

[

 

 

 

J
_
_
_
_
_
_
_
_
_
.
_
_
.
_
_
_
_
_
_
_
_
_
_
_
_
_
.
_
_
_
_
.
_
_
_
_
_
_
.
_
_
_
_
_
.
_
_
_
_
_
_

.
y
r
a
d
n
u
o
b

l
a
n
o
i
t
a
z
i
n
a
g
r
o

e
h
t

s
e
t
a
c
i
d
n
i

e
n
i
l

d
e
t
t
o
d

e
h
T

.
y
c
a
r
c
u
a
e
r
u
b

l
e
v
e
l

e
e
r
h
t

A

:
1
-
4

e
r
u
g
i
F

 
 
 

 
 

 
 
 

 
 

 

 

I

I
I

I

171

l
-
B

l
e
v
e
L

B

r
e
i
T

 

   

 

 

 

 

l
-
A

l
e
v
e
L

A

r
e
i
T

 

 

2
-
A

l
e
v
e
L

 

    

I

 

 

 

 
 
   
 

 

Z
-
B

l
e
v
e
L

 

 

 

 
 
 

 

 

 

 

.
r
e
i
t

h
c
a
e

t
a

s
l
e
v
e
l

o
w
t

e
r
a

e
r
e
h
T

.
n
o
i
t
a
z
i
n
a
g
r
o

d
e
r
e
i
t
-
o
w
t

A

:
2
-
4

e
r
u
g
i
F

172

literature, means a step in the vertical chain of command in a bureau—

cracy.

Organizations characterized by tiers, on the other hand, have

well-defined boundaries at each tier.

The units at each tier have

greater autonomy (in terms of decision making power and the acquisition

and use of resources) than do the units at each level of a bureaucracy.

The distinction is presented schematically in Figures 4—1 and 4-2.

It

would be misleading to assert that the organization whose structure is

diagrammed in Figure 4-1 has three levels and that the organization

shown in Figure 4-2 has four levels, and to leave the matter at that.

The structural representation in Figure 4-2 is not very different,

as it happens, from the basic form of the two-tiered Comilla cooperative

system.

The units at level B-2 can be thought Of’as farmers (about 30

per primary cooperative), the units at level B-1 as the managing

committees Of the village cooperatives, the units at level A-2 as the

sections (divisions, in the American phrase) of the thana cooperative

association, and the unit at level A-i as the director of the thana

cooperative association.

In the Expansion Program thanas where

OOOperatives of the Comilla type were set up during the nineteen sixties,

the only variation on this pattern was to form blocks in the field

supervision section, thereby introducing three levels into tier A.

In

Comilla Kotwali Thana structural experimentation was carried further,

to include the multiplication of tiers above the thana cooperative

federation.

Relevant Research

In an article on the general theory of growth, published in 1953,

Kenneth Boulding2 hypothesized that as an organization increases in

173

scale its internal relationships, which would be comparable to volume

in a three-dimensional pyramidal model Of an organization, increase as

the cube of changes along any of the height, width, or depth dimen-

sions.

Because the administrative portions of organization are charged

with cOordinating these internal relationships, the administration would

grow disproportionately in size as the organization expands.

Boulding's reasoning by geometric analogy has not held up well

under the brunt of subsequent research.

Some of the more recent

empirical research on organizational structure might not have been con-

ducted, however, if it had not been for Boulding's provocative conten-

tion.

An impressive investigation of the relationship between organiza-

tional size and organizational architecture was reported in 1970 by

Peter M. Blau, a sociologist who was then on the faculty of the

University of Chicago.3

In a cross-sectional study, Blau examined

the structure Of 1201 local branches of 53 employment security agencies

in the United States.

He plotted distinctly curvilinear relationships

between organizational size and a) the number of hierarchical levels in

the organizations, and b) the percentage of supervisory personnel.

(As

his index of size, Blau used the total number of employees in the

employment security agencies.)

Organizational architecture, in other

words, did not remain constant as the scale Of the organization changed.

Due to economies Of scale, larger organizations could make do with a

prOportionately smaller administrative staff.

This finding was precisely

contradictory to the law of swelling management which Kenneth Boulding

had proposed.

Blau interpreted the primary effect of an enlargement in organiza-

tional scale as being the increase in homogeneity within specialized

174

units in larger organizations; as a consequence of this specialization,

the organizational units find their tasks increasingly well—defined

and they therefore require less supervision.

In large employment

security agencies, Blau discovered, managers typically had a wider

span of control than in small agencies.

But a secondary, and opposite,

effect of increasing organizational size is to increase the diversity

among specialized units, and thereby to compound.problems of interbunit

coordination.

Blau proposed that this secondary effect (which was the

only one that Boulding had considered) puts a brake on the primary

effect, and explains the curvilinear relationship between size and

differentiation.

He summarized his findings with the assertion that

"Increasing size generates structural differentiation in organizations

along various dimensions at decreasing rates."

To anyone familiar with the calculus, the mention of decreasing

rates of change is an indication that a second derivative with a negative

sign is lurking in the shadowss

In an elegant demonstration for’which

Blau himself has admitted his admiration, Norman P. Hummon restated

the propositions contained in Blau's paper in mathematical terms.“

Hummon showed that most of Blau's nine propositions about organizational

structure were mathematically derivable from the preposition quoted

above.

Another sociologist who felt challenged to improve on Peter Blau's

research was Marshall w. Meyer.5 He pointed out that since Blau's

study had been cross—sectional in nature, it could not be used.as the

basis for causal inferences about the relationships of variables to

one another.

Using his own data from 19h departments of finance in

city, county, and state governments that were studied in 1966 and

again in 1971, Meyer claimed to have demonstrated, through the application

175

of the technique of path analysis, "that size is a cause of other

parameters of organizational structure, whereas other parameters

neither cause size nor are causally related to one another....

Once

size is taken into account, the other parameters show no causal effects;

organizational structure is thus the result of size."

6 As had

Peter Blau before him, Meyer took the number of full time employees in

a department of finance as the measure of size.

The structural variables

which Meyer contended that he had explained in terms of the longitudinal

influence of size were:

the number of horizontal divisions in the

organizations, the number of vertical levels, and the number of super-

visors.

Marshall Meyer's model of univariate causation in the structuring

of formal organizations is a crushing simplification.

His assertion of

the ubiquitous influence of size is unmodified and categorical.

Some

reinforcement of Meyer's views may be derived, it must nonetheless be

admitted, from the literature on the mathematical theory of organiza-

tions.

For example, Childers 23 g; have argued that if organizations

were randomly structured, then the expected value of a variable of

structural differentiation (defined as the number of occupations in an

organization) onld be %(S + 1), where S is the number of employees in

the organization.

According to this fOrmulation, structural differen-

tiation is a linear function of size alone.

Surprisingly enough,

Childers and his colleagues found.that the predictions of their'model

of random organizational differentiation matched the actual patterns

of differentiation found in units of the United States Coast Guard.7

David Specht strengthened the credibility of the Childers model by

snowing that the same expression for the expected value of structural

differentiation, E(D) = as + 1), could be derived from the assumption

176

that the distribution of the random variable D is binomial.8

Childers

had assumed that the distribution of random differentiation is uni-

form.

The equation 3(1)) = as + 1) is an algebraic expression.

In the

previous chapter, a resemblance between algebraic statements, frame-

works, and structural subsystems was suggested.

A review of Figures

3-9, 3—10, and 3-16 reveals that algebraic expressions such as AX = K,

the "frameworks" of Kenneth Boulding's typology of systems, and the

structural subsystem of Kast and Rosenzweig's analytic paradigm can

all be classified as fitting into the upper left hand cell of the

typological matrix of systems.

This congruence suggests that algebra

might be the usual mathematical mode for describing organizational

structure.

And such has generally been the case.

In addition to

Childers‘ and Specht's articles, in which mathematical descriptions of

organizational structure fell exactly into the anticipated category,

other algebraic models of organizational structure have been advocated

by Meyer,9 and by Mayhew.10

Some of the more abstruse connections

between algebra and organization charts were explored by MOrris

Friedell.11 who has noted that formal organizations can be viewed as

semi-lattices.

Since most studies of organizations have been cross-

sectional -- equivalent, in other words, to the study of organization

charts -- the predominant mode of discourse has been algebraic.

Four years after the publication of his mathematical reformulation

of Blau's research, Norman Hummon broke into the literature of organiza-

tion theory once more, this time as the principal author of a paper

which modelled the development of organizational structure.12

And

once again, Hummon took Peter Blau's formal theory of differentiation

in organizations as a point of departure.

The core of Blau's theory,

177

Hummon claimed, was the causal ordering:

size-—-differentiation.

Yet

the cross—sectional study which Blau conducted COUld equally well

support the hypothesis:

differentiation-—-size.

To probe more

deeply into causal relationships in the development of organizational

structure, Hummon and his colleagues created an ingenious differential

equation model which tied together actual and ideal forms of organiza—

tions.

Although I think that Hummon's presentation of this model is

flawed, I will nonetheless explain it in some detail because I will

use it as the take off point for'my own mathematical model of the

structural subsystem of thana cooperative organizations in East

Pakistan.

As the four basic variables whose relationship to one another he

wished to explain, Hummon took:

P, the number of employees primarily performing

the output tasks of the organization, e.g. the

production employees:

D, the number of divisions which functionally

differentiate the work force;

S, the number of supervisory employees in the

organization;

L, the mean number of hierarchical levels over

all divisions.13

Since I have difficulty remembering the significance of the symbol

P, and since I suspect that the reader may as well, I will substitute

the symbol W, for workers, in the following discussion.

All of these four variables describing organizational structure

are readily observable in most formal organizations.

Linked to these

observable variables Hummon postulated that there are four ideal and

unobservable variables W*, D*, S* and L*.

There is a linear algebraic

178

relationship, Hummon posited, between any one of these ideal variables

(called structural control variables in the article) and the other

three observable variables.

Thus,

w* =

a12

D + a

13

s + a14 L

+ r .

(u.1)

Change is introduced into the structural control model by means

of what John Hunter of Michigan State University calls the cybernetic

feedback equation.

The change in W, the number of workers, is governed

by the difference between the ideal number of workers and the actual

number of workers .

dW
2175‘ c1 (we - w)

(4.2)

This equation indicates that when the number of workers is below the

ideal level (W < W*), the change in W will be positive.

The size of

the coefficient c1 indicates how responsive the change in W will be to

this discrepancy.

The transition from equation 4.1 to 4.2 is equivalent to the shift

from the first to the second row of the typological matrix of systems,

that is:

from static systems to simple machines.

Time is a basic

part of simple machines; the solutions to such system models are

functions of time.

Despite the fact that Hummon reached into the realm

, of dynamic systems to fashion his complete model, it seems to me that

his work is not incompatable with the prOposition that the fundamentally

most fitting way to represent the structural subsystem of an organiza-

tion is as a static system.

The fact that equations of the form of

4.1 are at the core of Hummon's model lends credence to this proposi-

tion.

Equation 4.1 has all of the attributes that we might expect to

179

encounter in a mathematical representation of the structural subsystem,

or of what Kenneth Boulding called frameworks.

The equation is static.

Like an organization chart, the equation summarizes relationships in

which time plays no part.

Substituting the right hand side of equation 4.1 into equation

4.2, in place of the variable W*, gives

dW _

By applying identical reasoning to the other four observable

variables and to their unobservable counterparts, Hummon obtained a

dynamic linear translate mathematical model of the form x = AX + R.

x

in this instance is the vector of variables (W, D, S, L).

R is simply

a vector of sealers (c1 r1, c2 r2, 03 r3, cu rh).

Since there are

no inputs, the model is closed.

But because the vector of scalers

is appended, the model is not linear.14 Following John Hunter's

terminology once again, I have classified it as a linear trans1ate

mathematical model.

To test their model empirically, Hummon and his colleagues made use

of Marshall Meyer's data on government finance departments studied in

1966 and 1971.

Instead of the univariate pattern of causation which

Meyer claimed to have uncovered with his path analysis of the same

data, the Hummon group ascertained that their structural control model

showed the pattern of causation to be as shown in Figure 4-3.

(Those who

read the Hummon article itself will notice that I have rotated his diagram

ninety degrees.

It seems more appropriate to have the supervisors placed

above the workers, rather than next to them as in the original diagram.)

180

Size

Differentiation

Variables

Variables

Administrative

Variabl es

 

Production

Variables

 

_____ negative

impact

positive

impact

Figure 4-3:

Causal loop diagram15 showing the interaction

among structural variables in the Hummon model.

Meyer's argument, and Peter Blau's, was that the growth of organ-

izational structures is governed by the causal ordering size-_a—

differentiation, which would correspond to the rightward.pointing arrows

in Figure 4—3.

There are four of these:

the dashed arrows leading

from S to L and from W to D, and the solid arrows leading from S to D

and from W to L.

Even in the absence of the equations on which

Figure 4—3 is based, a glance at the diagram gives the impression that

the causal ordering size-—a-differentiation encompasses only a portion

of the interactions among the feur variables describing organizational

structure.

In the case of the daShed arrows pointing rightward from S

to L and from W to D, moreover, the diagram shows that the Hummon model

indicates that increasing size decreases differentiation, which is the

opposite of what Blau and Meyer had proposed.

Since the Hummon model describes a §y§tgm of interactions among

structural variables in organization, indirect effects of one variable

upon the change in another must also be taken into account.

Figure 4-3

Shows, for example, that a chain of positive influence leads from W to

 

181

S to D to L and then back to W.

TraCing indirect effects through

Figure 4—3 is no easy task.

To simplify the interpretation of systemic

patterns of causal interaction among the structural variables, the

Hummon group redrew the causal loop diagram shown in Figure 4-3 so as

to eliminate the less important mechanisms of interaction.

Their

simplified causal loop diagram is shown as Figure 4-4.

Size

Differentiation

Variables

Variables

Administrative

Variables

 

Production

Variables  

 

negative

impact

positive

impact

Figure 4-4:

Simplified causal loop diagram of

the Hummon model.

Although Hummon and his colleagues extracted only the sparsest of

conclusions from this simplified causal loop diagram, it would seem

to have devastating implications for the theories of organizational

structure proposed by Blau, Meyer, Childers, Specht, and Mayhew, all

of which were reviewed earlier in this chapter.

There is no direct

lg; indirect route of influence in Figure 4-4 that leads from ElEEEE

of the size variables, W and S, to either of the differentiation

variables, D and L.

The clear implication of Figure 4—4 is that

differentiation leads to size, but that size does not lead to differen-

tiation.

182

When a study contains implications which are so sharply at odds

with the bulk of the contemporary literature on organization theory,

the validity of the research needs to be examined with special rigor.

And when it is examined closely, the Hummon study reveals many defi-

ciencies.

The most obvious problem with the Hummon methodology is the

assumption, implicit in the mathematical model but not explicitly

addressed in the text, that the four organizational variables W, D, S

and L constitute a closed system.

To assert that the ideal number of

workers in an organization is determined only by the actual number of

supervisors, levels and divisions would seem to contradict the premises

of production economies and of managerial practice.

The assumption

shared by both disciplines is that the number of workers hired is

likely to be determined at least in part by the amount of work to be

done.

If an automobile manufacturer wishes to increase his production

of cars, he hires more workers -- subject to whatever labor saving

substitutions he might wiSh to make so as to increase productivity in

other ways.

And in a municipal, county or state department of

finance -- the type of organization from which Hummon, via Meyer, de-

rived.his test data -- the chief of the department would most likely

seek authority to hire more clerks as the number of taxpayers and the

volume of financial transactions grew.

AlthOUgh this line of argument is appealing from one direction it

is troublesome from another.

In the typology of systems presented in

Chapter 3 the structural subsystems of organizations were classified

as closed systems.

If an input variable representing workload were

added to the mathematical model of the structural subsystem, the doors

of the system would be thrown wide open:

it would become a machine

183

with input.

So to the extent that Hummon is correct in modeling the

structural subsystem as closed, to the same extent the validity of the

general systems typology is affirmed.

Another criticism of the Hummon paper is that it presents broad

conclusions about a dynamic system on the basis of only two data

points:

the observations of finance departments in 1966 and in 1971.

A test of the validity of the model would require a third set of

observations to test the predictions derived from the first two data

sets.

Since the Hummon model has four system variables, it can be said

to trace out a trajectory in a four dimensional phase space.

Through

the application of expgst statistical estimation procedures, Hummon

calculated parameters for his model so as to assure that the system

trajectory passed through the 1966 and 1971 data points in the four

dimensional phase space.

But did.the model trajectory pass near the

actual 1968 data point?

And as the system trajectory in the phase

space continues to, say, the 1976 data point, will it hit or miss the

target?

A final objection is that the Hummon paper does not link the

rationale for the concept of structural control to any of the existing

184

literature on organization theory.

The first step in the formulation of

the Hummon model is the postulation of the existence of an unobserved

"control structure," an image residing in the minds of organizational

directors which describes how the organizational structure shgulg_be

patterned.

The directors attempt to minimize the discrepancy between

this ideal pattern and the observed, realized organizational structure.

From this adjustment process the Hummon model derives its dynamics.

Once the vocabulary has been altered so as to put the proposition

in more familiar terms, the existence of such an adjustment process

becomes plausible.

The relevant phrase in the real world is not

"control structure" but "staffing pattern."

The military have

carried the idea of staffing patterns the farthest, making it into a

doctrine of organizational structure.

The organizational doctrine of

staffing patterns in the U.S. army decrees, for example, that there

should be 12 men in a squad, and that each squad Should be led by a

sergeant; that there should be three squads in a platoon, and that each

platoon should be led by a lieutenant; that there should be three

platoons in a company, and that each company should be led by a captain;

and so on up to the staffing patterns for divisions, corps, and armies.

When his company is depleted in battle, the company commander requests

replacements to bring his unit up to its full complement of troops, a

process of structural control resembling the one that Hummon and his

colleagues propose.

Organizational managers do attempt to adjust the

rank, number, and roles of their personnel so as to make them accord

with staffing patterns.

And interestingly enough, those staffing

patterns are often tied to rules of proportionality -- spans of control,

ratios of professors to secretaries, ratios of occupants of different

civil service grades.

The algebraic equation which Hummon proposes for

the ideal number of supervisors in an organization,

185

8* = a

W + a

D

+ a3“ Li+

(4,4)

31

32

r3'

is like a composite of several ratios.

If all variables on the right

hand side except W are zero, for example, the coefficient a31 indicates

the ideal ratio of supervisors to workers.

So the equation is plausible.

The Structural Subsystem at the Village Tier

Few of the preceding theoretical considerations can readily be

applied at the village tier of the Comilla cooperative organization.

The primary cooperatives in the villages were more primitive in organiza-

tional form than were the thana cooperative associations.

As we shall

see in later chapters, some organizational subsystems which were present

at the thana tier were absent at the village tier of organization.

The

village cooperatives, moreover, were less formal in their specifica-

tion of many aspects of their organization.

Furthermore, since most of

the theories of organizational structure reviewed above are numerical

in character, the testing of these theories at the village tier would

require the accumulation of survey research data on the structure of

primary cooperatives.

Such a survey has not been undertaken, either by

the author or by any other investigator.

The following remarks, then,

should be taken as general background information on the structure of

the village cooperatives.

The basic organizational unit at the village tier was the primary

cooperative.

Occasionally, farmers from two villages would join

together to form a primary cooperative, but ordinarily the members of

one COOperative were drawn from one village.

(The very concept of a

186

village is ill-defined in the Comilla area.

Houses are scattered in

small clumps around the delta's plain, rather than being gathered in

village clusters.

Several Bengali words, with different referents, have

been translated into the single English term "village.")16

At its annual general meeting, each primary cooperative was

required to select a third of the members of the cooperative's managing

committee, who served for overlapping three year terms.

As a rule, the

selection was by common consent rather than by vote.

These directors

17

then selected from among themselves a manager, a chairman, and a vice

chairman.

The manager (called the organizer until 1962) was the most important

officer in the cooperative.

His duties were numerous and critical.

He

collected and disbursed loan funds, served as financial agent for the

village cooperative in its dealings with the central association, over-

saw the keeping of accounts and records, arranged for the rental and use

of machinery owned by the central association, encouraged the members

to attend meetings, save, purchase shares, and repay loans, and acted

as agent for the cooperative in joint marketing ventures.

The central

association paid the manager several sorts of commissions as well as

a travel allowance for attending weekly training classes; he received

no salary from either the central association or the primary c00pera-

tive.'

_

The chairman's role appeared more honorary than operative.

His

major job was to chair the weekly cooperative meetings which were in

practice usually dominated by the manager.

Often the managing committee

selected a prestigious member of their group to serve as chairman.

A

good chairman could use his prestige in the village to lend respect-

ability and order to cooperative meetings.

He could be of help to the

187

manager in enforcing the attendance at weekly meetings, the deposit

of savings, or the repayment of loans.

He received no salary, com-

mission, or travel allowances.

The vice chairman and the other members of the managing committee

had, in the early sixties, no defined job in the village cooperative.

They were installed in office in conformance with the cooperative

regulations of East Pakistan, but the leader of the thana-wide c00pera-

tive project, Akhter Hameed Khan, opposed their assuming any regular

functions.

In his view, the failure of the Union Multi-Purpose Coopera-

tives, which had been active in the Comilla area for decades, resulted

from the assumption of operational authority by their managing committees.

Akhter Hameed Khan advocated conducting all business in the weekly

general meetings of the cooperative societies rather than in sessions

of their managing committees.

In the long run, Akhter Hameed Khan's philosophy of direct demo-

cracy through the weekly general meeting did not prevail.

The managing

committees assumed more and more responsibility, until in 1972 they

were described by Badruddin Ahmed as making most policy decisions on

matters of budget, membership, credit, production, supplies, and

marketing.

Ahmed described the weekly general meeting as a control

mechanism which approved or disapproved the decisions of the managing

committee, and kept the functionaires of the cooperative from deviating

too far from the general sentiment of the membership.18

In addition to the managing committee, the members of each village

cooperative also selected a model farmer.

The model farmer was expected

to attend monthly training classes arranged by the central cooperative

association.

In those classes he was taught improved methods of

agriculture, pisiculture, and animal husbandry.

He was expected to

188

adopt these methods himself and to use his farm, pond, and herd as

demonstration models for the cooperative members.

The model farmer

was paid only a travel allowance, which the central association gave

him for rickshaw fare to and from his training classes.

The Structural Subsystem at the Thana Tier

In the Comilla cooperative system, organizational structure was

more formal at the thana than at the village tier.

Constructs drawn

from organization theory could more readily be applied and tested at

the thana tier.

A study of the central cooperative association in

Kotwali Thana of Comilla would have been insufficient, though, as a

data base for such research.

To enlarge the number of cases, the

author surveyed all of the 24 thanas in which a central cooperative

association of the Comilla type was established during the decade of

the nineteen sixties.

Comilla Kotwali Thana was the first of these.

In 1963-64, experimental cooperative associations on the Comilla model

were launched in Gaibandha thana of Rangpur district, in Natore thana

of Rajshahi district, and in Gouripur thana of Mymensingh district.

Two years later seven more thanas of Comilla district were brought into

the expansion program, followed by the remaining 13 thanas in Comilla

district in 1967-68.

The project directors of the central cooperative

associations in all of these thanas were sent questionnaires in

September of 1977.

Eventually, all but one of them responded.

The

results of this survey of organizational structure in the thana coopera-

tive associations will be reported later in this chapter.

To introduce

that aggregate investigation, though, let us examine first the structural

evolution of the central association in Comilla Kotwali Thana as a case

139

 

n
o
i
t
c
e
S

e
-
m
a
r
g
o
r
P

n
o
i
s
n
e
t
x
E

r
e
t
a
W

;
m
o
n
e
v

I

'

I

l
a
r
u
t
l
u
c
i
r
g
A

I

e
n
i
h
c
a
M

n
o
i
t
a
t
S

 

g
n
i
s
i
d
n
m
l
d
fi

n
o
i
t
c
e
S

&

g
n
i
k
n
a
B

t
i
d
e
r
C

 

 

n
a
-
e
o
P

I

s
t
e
s

s
n
I

.
t
s
s
A

]
r
o
t
c
e
p
s
n
I

.
t
s
e
n

I

n
m
e
r
o
F

I

J

n
a
m
e
r
o
F

I

]
r
o
t
c
e
p
s
n
I

.
t
s
s
A
I

r
o
t
c
e
p
s
n
I

.
t
s
s
A
'

 

 

e
c
n
e
n
e
t
n
i
a
M

n
o
i
t
c
e
S

n
o
i
t
c
e
S

.
I

n
e
-
e
v
o
H

“
1
‘
“

s
g
n
i
v
e
S

k
n
a
B

.
A
.
C
.
C
.
‘
I
’
.
K

n
m
i
a
h
C

e
e
t
t
i
n
u
o
C

g
n
i
g
a
n
a
M

r
o
t
c
e
r
i
D

t
c
e
j
o
r
P

I

I

l

.

I

I

.

'

I
s
c
i
n
a
h
c
e
H
I

]
r
o
s
i
v
r
e
p
u
S
[

  

s
c
i
n
a
h
c
e
M

n
u
o
c
c
A

I

I

I

n

I

1

r
e
n
i
m
—
g
r
O

r
e
z
1
n
:
;

0

n
e
m
o
W

r
e
z
i
n
n
:

0

r
e
z
i
n
:
«
;
r
0

s
r
e
v
i
r
D

r
e
z
i
n
a
g
r
O

r
e
s
i
n
a
g
r
O

r
e
z
i
n
a
g
r
O

-
l
e
m
a
P

l
e
d
o
M

/

B
F
I
‘
I
I
E
I
W
S

E
G
A
L
L
I
V

.
“
9
1

f
o

s
a

)
m
m
(

n
o
i
t
a
i
c
o
s
s
A

e
v
i
t
a
r
e
p
o
o
C

l
a
r
t
n
e
C

e
n
a
h
T

n
a
m
o
R

e
h
t

f
o

.
t
r
a
m

n
o
i
t
-
c
i
n
a
g
r
O

:
5
—
h

e
r
u
g
i
F

study.

190

The Kotwali Thana Central Cooperative Association (KTCCA) was

registered as a cooperative society on January 26, 1962, nearly two

years after the first local cooperative had been organized in South

Rampur village on March 15, 1960.

Even though the village cooperatives

had a head start, the thana association quickly outstripped them in

influence and sophistication.

It could be said, indeed, that the

Comilla system did not significantly accelerate rural change until the

thana association got into gear.

From 1962 to 1966 the KTCCA expanded its operations into new sorts

of activity until it included nine sections:

field supervision, main-

tenance, movement, dairy, grain mill, cold storage, accounting, water

development, and marketing.

The purpose of the subsidiary sections was

to service the primary cooperatives, but the very volume of effort

required to manage and coordinate the sections distracted the administra-

tors of the KTCCA from the immediate problems of the individual coopera-

tives.

The field supervision section was lagging from administrative

neglect at the very time when the scale of its operations should have

called for more decisive attention.

In January of 1966 the chairman of the KTCCA, Akhter Hameed Khan,

decided that the credit program of the agricultural cooperatives would

benefit from an independent managing structure.

On the precedent of the

non-agricultural cooperatives, which had been divided out into a

separate federation in 1964, the Agricultural Cooperative Federation

(ACF) was set up.

According to ACF by-laws, the KTCCA was to serve

as its financing bank.

Shamsul Huq, who had been the KTCCA deputy

director in charge of field supervision, was deputed by the KTCCA to

serve as project director of the new federation.

 

  

s
r
e
g
a
n
a
M

0
1
-

191

I
I
I

'
I

e
g
a

I

o
t
c
e
p
s
n
I

r 

r
e
d
n
U

h
c
h
E

s
r
e
z
i
n
a
g
r
O

s
e
i
t
e
i
c
o
S

e
g
a
l
l
i
V

0
1
—

)
?
I
I
(

e
r
e
-
r
a
F

l
e
d
o
M

0
1
—

;
s
e
i
t
e
i
c
o
S

0
1
-
6

r
o
f

h
c
h

2
-
1

s
t
n
a
t
n
u
o
c
c
A

e
g
a
l
l
i
V

.
9
6
9
1

f
o

s
a

)
F
C
A
(

n
o
i
t
a
r
e
d
e
F

s
e
v
i
t
a
r
e
p
o
o
C

l
a
r
u
t
l
u
c
i
r
g
A

e
h
t

f
o

t
r
a
h
C

n
o
i
t
a
z
i
m
g
r
O

:
6
-
0
‘

e
r
u
g
i
F

w
s
n
o
r
c
s
r
s
u
l

)
?
(
s
n
o
r
c
u
P
S
N
I

)
U
S
R
O
T
n
E
P
S
N
I

)
7
(
s
n
o
T
m
P
S
N
I

)
7
(
s
a
o
T
c
a
P
S
N
I

 

   

  

V
I
-
(

v
-
x
c
w
n

C
O
L
B

.
a

  
I  

I
_
.

l

I
I
I
-
K
C
O
L
B

I
I
—
K
C
O
L
B

I
-
K
C
O
L
B

)
I

r
o
t
c
e
p
s
n
I

f
e
i
h
C

)
“
r
o
t
c
e
p
s
n
I

f
e
i
h
C

;
1

r
o
t
c
e
p
s
n
I

f
e
i
h
C

I
i
I
r
o
t
c
e
p
s
n
l

f
e
i
h
C

;
“
r
o
t
c
e
p
s
n
I

f
e
i
h
C

)
2

:
m
m

?
m
c

T
S
S
A

)
2
(
c
m
s
m

r
u
c

T
S
S
A

2

m
r
a
w

‘
i
m
c

T
S
S
A

2

c
s
P
S
N
I

r
u
c

T
S
S
A

2
(
C
E
P
S
N
I

r
u
o

T
S
S
A

e
e
t
t
i
m
o
C

  1

g
n
i
g
a
n
a
M

I

2

I

6

.
l
d
d
A

r
o
t
c
e
r
i
D

d
e
t
a
n
i
m
o
N

I   

r
o
t
c
e
r
i
D

,
d
e
t
c
e
l
E

a

I

I

I
U
I

n
o
r
c
e
m
n

"
w
e
n

n
o
m
t
n
m

Y
T
U
P
E
D

G
N
I
N
I
A
R
T

)
1
(

:
r
e
m
m
u
s

I

I

)
1
(

S
T
N
U
O
C
C
A

)
1
(

N
A
O
L

w
o
r
'
m
n
m

m
n
c
n

n
o
r
c
s
m
n

Y
T
U
P
E
D

I

192

Most of the author's on—site research on the Comilla cooperatives

was done in the winter and spring of 1966 while the new federation was

being organized.

It was a time of shake-up and reconsideration in the

Comilla cooperatives.

It was an excellent time for an inquiry into

fundamentals. since the bone structure of the old field supervision

section. the main support of the credit system. was at that time

revealed and rearranged.

Because the organizational upheaval of 1966 was so comprehensive,

the faculty at the Academy for Rural Development who were charged with

writing annual reports of the cooperative experiment devoted consider-

able amounts of text in the seventh (1966-67) through ninth (1968-69)

reports to explanations of the structural transformation.

It was as

if the academy faculty, and perhaps the leaders of the thana cooperative

organizations. were seeking to fix in their minds a comprehensible

image of the new organizational relationships.

The explanatory tech-

nique to which they often resorted was the organization chart.

The eighth annual report, covering the year 1967—68, opens with a

diagramatic explanation of the original. two-tiered organizational form

of the Kotwali Thana Central Cooperative Association.

The organization

chart shown as Figure 4—5. which shows the structure of the KTCCA in

1961, is displayed on page 1 of the report.

Even in its very first

year, evidently, the organizational structure of the KTCCA had been

horizontally differentiated, with distinct organizational sections for

banking and credit, the machine station, and so forth.

In the vertical

dimension, the original design entailed a careful layering of super-

vision.

Attention to formal organizational structure had characterized

the KTCCA from its beginning.

By the end of the decade, the KTCCA had been significantly

193

rearranged.

Figure 4—6, which shows the organization of the KTCCA in

June 1969,19 is taken from the ninth annual report.

The KTCCA had

become, in the words of that report. "an association of associations."

No longer did the KTCCA relate directly to the village cooperative

societies. but rather to federations of societies such as the Agricul-

tural Cooperatives Federation (ACF) and the Special Cooperative

Societies Federation (SCSF).

A few primary cooperatives, such as the

Comilla Housing Cooperative Society, were still linked directly to the

KTCCA, with no intervening federation, but these served a professional

clientele which was a class apart from the village population.

Indeed,

the ACF was the only one of the seven organizational members of the

KTCCA in 1969 which was predominantly rural in its orientation.

The

members of the other six c00perative associations lived mainly in

Comilla town or in the suburban enclave at Kotbari, where some of the

faculty and employees of the Academy for’Rural Development resided.

The roots of the KTCCA, evidently, had extended themselves as much into

the urban as into the rural areas of the thana.

Within the KTCCA. the most notable organizational change from

1961 to 1969 was the addition of a complement of upper-echelon managers.

the senior vice chairman and the two vice chairmen. whose positions on

the organization charts resembled that of vice presidents in a large

corporation.

Indeed. the similarities to the more intricate forms of

corporate organization were numerous -- so numerous, in fact, as to

lead one to suspect that the organizational image which Akhter Hameed

Khan had in his mind as he constructed the cooperative organization was

not that of the Chinese commune (which he claimed later to have been

imitating) but rather a multi-tiered corporation such as General Motors.

In this regard, the influence of the foreign. largely American, corps

194

of project advisors in Comilla should not be overlooked.

The Michigan

State advisors there during the sixties were typical of their country—

men in recommending American organizational solutions to management

problems encountered in the developing nations.

The most astounding trend in the organizational evolution of the

cooperatives was toward vertical differentiation in the organization.

At its start, the Comilla cooperative project adopted a two-tiered

organization. with primary cooperatives at the village level and with

the KTCCA at the thana level.

When the Special Cooperative Societies

Federation was introduced in 1964, followed by the Agricultural

Cooperatives Federation in 1966, the basic pattern became three-tiered.

At the top, the KTCCA came to be described as an "apex organization" and

as a "superstructure."

Its role became analogous to that of a holding

company in a corporate empire.

The ideological distinction between a

cooperative and a corporation, according to which the pattern of owner-

ship and control in the cooperative pyramided upwards rather than

downwards. was nearly moot in practice.

Farmers did. to be sure,

purchase shares in their primary cooperative societies. which

purchased shares in the Agricultural COOperatives Federation, which

purchased shares in the Kotwali Thana Central Cooperative Association.

The influence of the lowest upon the highest tier of the organization,

though, was diluted by these stages.

Three tiers was not the limit.

An effort was made, late in the

nineteen sixties, to bring together the thana central cooperative

associations of the Comilla type into one master cooperative associa-

tion, the Comilla Industrial Cooperative Society.

The KTCCA bought

shares in this organization, as did the cooperative associations in

nine other thanas.

The two cold storage warehouses (used mainly for

195

holding potatoes for sale at higher prices in the off-season) and the

creamery in Comilla were turned over to the Industrial Cooperative

Society.

The idea was that the management of these capital intensive.

industrial-type projects should not be limited by the production or

marketing potential of a single thana.

As the number of tiers of organization in the Comilla cooPeratives

grew in the nineteen sixties. so did the organizational distance

between the peasant farmers of the thana and the leaders of their

cooperatives.

This increasing distance contributed to the dry rot

which afflicted many of the primary societies. and even the lower levels

of the ACF, toward the end of the decade.

As the distractions of

horizontal and vertical coordination multiplied, it became increasingly

difficult for the incorruptible idealists at the tap of the cooperative

organization. acting through the filters of many layers and tiers of

management, to hold all their subordinates accountable.

Standing as gatekeepers between the primary cooperatives and the

thana level organization were the inspectors.

To anyone trained in the

reading of organization charts, the key role of the inspectors within

the Agricultural Cooperatives Federation is obvious from Figure 4-6.

In 1969, when the organization chart shown in Figure h—6 was drawn,

each inspector was expected to supervise 10 primary societies.

The

actual ratio was closer to 8 village cooperatives per inspector.

The key personnel in the Agricultural Cooperatives Federation were

the cadre of inspectors.

Three of the nine inspectors employed by the

ACF in 1966 had worked previously in the Village Agricultural and

Development (V-AID) program in the nineteen fifties.

Indeed the very

notion of the inspector's task could be traced conceptually to that of

the old village level workers in V-AID.

The village level workers, who

196

are still extant in India and in the Peace Corps, were expected to

catalyze

community organization and modernization in the villages.

By the middle of the decade, many of those exhortative community-

development functions were still apparent in the inspectors' own con-

ception of their duties.

In December of 1965, the author submitted a

questionnaire to each of the nine inspectors of the KTCCA, and followed

up with an individual interview with each inspector.

The inspectors

gave a moderately high ranking (fifth place out of 15 ranked functions)

to the relative amount of time they spent "urging society members to

adopt more improved agricultural practices."

The functions on which

the inspectors said they spent the most time throughout the year were.

in order from first to fourth ranking, "urging society members to save

more;" "urging society members to buy more share capital;" "organizing

new societies;" and "collecting loan repayments."20

By the mid-sixties it was already apparent that the ACF leadership

was using the inspectors as an echelon of middle level organizational

managers through whom it could communicate with the primary cooperative

societies.

And by 1973. when Harry W. Blair conducted his research on

the Comilla c00peratives, the role of the inspector seemed to have

become entirely that of the bureaucratic middleman.

Blair wrote that

the duties of an inspector were:

...to prepare loan preposals for each KSS (the

Bengali acronym for primary cooperative) to borrow

from the ACF, to make recommendations concerning

loans from the ACF to the K58, to keep data on

weekly meetings of the village societies, pur-

chase of shares and savings deposits. loan issue

and repayment, to assist and supervise the K58

manager in disbursing loans to individual

farmers. signing receipts. and so on.

In short,

he is the intermediary between the primary society

at the village level and the administration at the

thana 1eve1.41

197

In the author's 1966 survey of the ways in which inspectors

allocated their time, most of the duties mentioned by Blair ranked in

the bottom half of the list.

"Processing loan applications," "advising

primary societies how to apply for loans," and "recommending loan

limits; ranked 10, 11 and 15 in the 1966 survey, whereas Blair gives

the impression that overseeing the loan process was a very important

function of the inspectors in 1973.

Still more significant, however, is

that Blair's description of the inspector's role includes not a trace

of the idealistic V-AID concept of a village level worker who would

inspire and organize the villagers.

In the course of a decade, the

inspectors had been transformed from community development activists

into functionaries.

The need for some sort of intermediate managerial personnel is

apparent, nonetheless, from an examination of statistics about the

scope of the cooPerative movement in Comilla.

In the 107 square miles

of Comilla Kotwali Thana there were approximately 24,000 farm families

at the beginning of the nineteen sixties.

In the first five years of

the Camilla experiment, membership in the village cooperatives rose

above 6,000, implying that one fourth of the farm families had

associated themselves with the cooperatives.

One hundred sixty

primary c00perative societies had been organized by the middle of the

decade, indicating that farmers in a third of the #60 villages had

joined the movement.

It would have been impossible for the staff of the Central Associa-

tion to have kept track of the affairs of all 160 primary societies,

and of the patterns of influence of the 6,000 members upon them. if it

had not been for the intervening field staff of inspectors.

Even in

1966, the Comilla credit cooperatives formed a large, complex organization.

198

That organization could not have functioned without the middle level

management provided by the inspectors.

The activities of the Comilla cooperatives attained a certain

stability in the mid sixties. and at the level of operations then

attained things ticked along quite nicely.

But in 1966-67 the Rakistan

government decided, with Akhter Hameed Khan's approval, to push beyond

the 1964-66 plateau and to raise the memberShip and loan volume of the

cooperatives in Kotwali Thana.

From the plateau of 1964-66, a hundred

new c00peratives were organized in 1967-68, membership almost doubled,

and loan issue shot up six times.

This helter-skelter growth was achieved at the cost of the relaxa-

tion of many aspects of cooperative discipline.

The growth occured,

moreover, at the very time when many of the richer farmers were for

the first time showing a strong interest in joining the cooperatives

and turning them to their own ends, because of the newly apparent

benefits of tubewell irrigation and the Green Revolution.

The

opportunities for subversion and corruption were just too numerous to

handle, and the solidarity of many village societies disintegrated.

The Comilla cooperative movement struggled for at least five years

to recover from its attempt in 1967 to bite off more than it could

chew.

Efforts to retrieve the gusher of loans issued in 1967 were

complicated by political turmoil and civil war.

The origin of the

problem, nonetheless, had been a deliberate policy decision by the

cooperative leadership, so the bloated loan account could not be

excused by reference to external factors alone.

A campaign of retrench-

ment brought the annual loan issue down to about the level of the

1964-66 plateau by 1970-71.

The number of loan recipients fell back to

below the 1964—66 level, but since so many newly enrolled members were

199

still on the books, the rate of participation in the loan program

dropped to a mere 33% in 1970-71.

Even in the peak loan issue year of

1967-68, over sixty percent of the cooperative members had taken loans,

and the participation rate had been over eighty percent prior to the

1967 expansion.

The 1970-71 annual report of the Comilla cooperative

experiment states that ACF personnel explained the decreasing trend of

loan participation as being due to "(1) non-repayment of previous loans

which disqualified the members for further loan receipts, (2) political

turmoil leading to the War of Liberation, and (3) limitation imposed on

the types of loans issued by the ACF."2‘2

Investment in agriculture was

limited because of political and military turmoil from 1969 to 1971,

the report notes.

Morever, as part of its retrenchment strategy the

ACF cut back drastically on the number of purposes for which loans could

be issued.

By 1970-71 there were only 10 approved loan categories, all

directly related to agriculture, whereas in 1965-66 farmers had been

permitted to borrow for 24 categories of expense, some of which were

personal.

I

A bleak interpretation of what he terms the "administrative decline"

in the Comilla cooperatives in the late sixties is given by Harry W.

Blair.23 He avers that because of the pressure to issue more and

bigger loans in 1967, many suspect societies and bogus members were

enrolled, and many loans were rolled over.

To administer the rapidly

growing loan program, Blair contends that the central federation under-

took emergency recruitment of inspectors, and drew many from among

primary society managers who were themselves big borrowers.

These

ex—managers sometimes used their new positions to cover up their own

loan defaults.

Some of the inspectors in this period, Blair reports,

accepted bribes and arranged kickbacks, while others took a lax

attitude toward irregularities in the village cooperatives under their

supervision.

200

This unhappy set of events is ascribable to growth which was overly

rapid.

But it is conceivable as well that the Comilla form of coopera-

tive organization had attained a near-optimum size in the mid-sixties.

and that organizational breakdown resulted from the attempt to push

membership and the volume of loan transactions beyond that optimum.

In the literature on rural development organizations, little

guidance is to be found on the question of optimum scale.

Arthur R.

Mosher, then president of the Agricultural Development Council, proposed

in 1970 an optimum size for integrated rural development projects, but

the geographic area of 1,000 to 3,000 square miles which he advocated

was from 10 to 30 times bigger than Comilla Kotwali Thana.

Mosher

reasoned that the farming district to be included in an integrated rural

development project should be "large enough to include at least ten

farming localities and to include at least twenty to twenty-five

extension workers (since that is the minimum number for effective and

continuous in-service training).

The upper limit is set by convenience

in supervision and servicing from the district headquarters"?4 For an

integrated rural development project of the Comilla type, that upper

limit of convenience will be encountered, at least in a country as

densely populated as Bangladesh, in a region far smaller than 1,000

square miles.

Uphoff and Esman, in their analysis of rural deve10pment in Asia,

argue for the efficacy of a two—tiered organization, and they comment on

the size which is workable for each tier.

Unfortunately they remark

only on the minimum, and not on the maximum, desirable size for the

second tier.

201

The primary level of organization tends to be

the hamlet, the neighborhood or some other group

of from 30 to 100 families.

Organizations of

this scale can be effective for pooling labor,

regulating irrigation water at the field level,

controlling pests and weeds, harmonizing plant-

ing, weeding and harvesting schedules, and per-

forming some common services....

These organizations need in turn to be knit into

some larger organization which can provide the

benefits of scale -- 1,000 families or 5,000

persons as a minimum -- to afford the managerial

cadres and the physical facilities necessary to

operate complex structures handling large sums of

money as well as physical in uts, storage, and

even processing activities.

With an average primary cooperative membership during the nineteen

sixties of 34 -- each member commonly being presumed to represent one

family -- the lower tier of the Comilla cooperatives matched Uphoff and

Esman's prescription for size.

With 6,000 families covered by the mid-

sixties, though, the upper tier of the Comilla cooperative structure

was six times the minimum size which these authors recommend.

By 1967

the thana organization covered more than ten times that minimum number

of families.

At the size which it had attained in the mid sixties, had

the Comilla cooperative structure already reached its scale of maximum

efficiency, 50 that the 1967 expansion carried the cooperatives into a

region of diminishing absolute returns to growth?

Considerations of maximum organizational scale were prominent in

Akhter Hameed Khan's thoughts when he undertook his tour of twenty

thanas in December of 1970 and January of 1971.

The tour, which covered

the twenty thanas of Comilla District in which cooperatives of the

Comilla type had been established at the time, was undertaken on his

own initiative and as the result of his personal curiosity.

In his

remarks on the development projects in Laksam thana he wrote:

202

But the task in Laksam is indeed daunting.

In

the first place the thana is far too big:

209

sq. miles, 27 unions, 542 villages.

Ten years

ago the population was three hundred and sixteen

thousand; now it must be near four hundred

thousand.

The average thana in East Pakistan

is about half the size of Laksam.

A hundred

square miles of land with two thousand persons

living in one square mile can be managed with

much difficulty.

With much trouble and effort

planning and servicing institutions, local

government councils and central cooperative

associations can be built.

Fifty square miles

could be better organized; two hundred square

miles is an impossible unit.

Unless an energetic

government sheds the inertia of tradition and

rearranges reasonably the units of development

administration, thanas like Laksam will suffer

the consequences of haphazard planning and poor

coordination.25

§_Systems Interpretation

Peter Blau, it may be recalled, described two contradictory

relationships between the size of employment security agencies and

the efficiency of their supervision.

He observed positive returns to

scale for management, in that smaller organizations had more hierarchical

levels and more supervisors relative to their total size than did large

organizations.

In his view, these returns to scale resulted from the

greater internal homogeneity of specialized units in larger organiza-

tions.

On the other hand, Blau also reported that the increasing need

for coordination among units in a large organization had exactly the

opposite influence, in that it tended to narrow the span of administra-

tive control and increase the number of hierarchical levels of super»

vision.

Since this latter effect was the one propounded by Kenneth

Boulding in his seminal article "Toward a General Theory of Growth,"

I have labelled it the "Boulding Effect."

In Figure 4-7 below, I have

termed the first effect, which Peter Blau considered to be the primary

one, the "Blau effect."

203

I

Percent of
perso

e1

A.
   

Blau effect

who are

A

supervisors

 

.“_‘(;-combined effect

   
§  

~~

“

~

“

D

 

size (Number of employees)

Figure 4—7:

The Blau and Boulding effects.

If both the Blau and Boulding effects were linear algebraic

functions of size, and if neither took on negative values, the combina-

tion of the two effects might be as shown by the line ABCD in Figure 4-7.

Angular as it is, the ABC portion of the curve, at least, bears a

certain resemblance to the smooth descending curve, asymptotically

approaching the horizontal axis, which Peter Blau inferred from his data

on structural differentiation in employment security agencies.

If the

Blau and Boulding effects were themselves curvilinear, as in Figure

b—B, a "U" shaped combined effect curve would be generated.

The left

hand, descending, portion of this curve would resemble the one which

Blau traced through his data.

The existence of the right hand portion

of the combined effects curve, the part labelled BC in Figure 4-8, was

not even a topic of speculation in Blau's research.

Perhaps this was

because the largest of the 1201 local employment security offices in

his sample had only 225 employees.

 

204

c

  
   

,rcombined effect

Boulding effect

Percent of

h

personnel

who are

supervisors

 

Figure 4—8:

A curvilinear version of the Blau and Boulding effects.

size

Clearly point B in Figure 4—8 represents an organizational size of

maximum administrative efficiency, a size at which a minimal percentage

of supervisors could manage the enterprise.

(At this optimum, the

increasing returns to scale of the Blau effect are balanced by the

decreasing returns to scale of the Boulding effect.

The organizational scale yielding maximum administrative efficiency

is not directly related to, and could be quite different from, the scale

of operations yielding maximum_productive efficiency.

The economic

theory of production functions explains the potentially curvilinear

relationship between the size of the workforce and economic productivity.

The interaction of the Blau and Boulding effects, comparably, explains

the curvilinear relationship between the size of an organization and

its administrative efficiency.

Evidence from the literature for the existence of a U shaped

curve of administrative efficiency is admittedly slight.

Most empirical

studies of the relationship between size and the proportion of superb

visors in an organization have traced a curve such as AB in Figure

4-8, but have not shown a rise such as BC at the upper end of the

scale of size.

The only study I have read about which directly supports

the proposition summarized in Figure 4—8 was conducted by Haas, Hall,

 

205

and Johnson.

They sampled organizations of widely differing types,

and found that the proportion of support personnel (including super-

visors and all staff personnel not directly engaged in the primary

work of the organization) was greater for small organizations (0-700

employees) and for large organizations (more than 1,400 employees) than

for organizations of intermediate size.27

But the small size of their

sample and the fact that other factors were not controlled (the

organizations in their sample included an educational television

station, a municipal airport, and a railroad) cast doubt on the general-

ity of their findings.

Within the Agricultural Cooperatives Federation in Comilla Kotwali

Thana there clearly was an administrative optimum, as can be seen in

Figure 4-9.

In 1967, when the Agricultural Cooperatives Federation had

attained a size of 75 employees, the percentage of those employees who

were supervisors was at an all time minimum of eight percent.

As the

Agricultural Cooperatives Federation grew thereafter, it swelled

disproportionately at the upper levels, adding deputy directors and

chief inspectors faster than it added non-supervisory employees.

To find out whether the case of the Agricultural CooPeratives

Federation was a fluke or was representative of some more general

organizational principle, data were analyzed from questionnaires which

the author mailed to the directors of thana central cooperative

associations of the Comilla type that were active in the nineteen

sixties.

The questionnaire was to be filled out once for each year in

which a thana cooperative association was active.

Most of the project

directors complied with this request, although it entailed a good deal

of backtracking through their own records.

0f the possible total of

82 questionnaires for the expansion projects in Nator, Gaibandha, and

206

"
a
l
i
V
!
!
!
0
9
.

i
l
A

O
/
A
V
I

J
I
I
/
O
M
J
F
O

’
7
/
A
I
C
(
”

emaciawsowxaom/oa/onzcw

.02.:

13/603! 4' 9

.972! VINO! PIICIAO' OI oft/.053 W101!

//V 77/! 463/ 604 fulAL COOPé'MT/Vl’J

Ffplfll 770M

207

Gouripur thanas and in 20 thanas of Comilla district, 69 useable

questionnaires were returned, for a response rate of 8# percent.

Careful examination of the questionnaire responses caused me to

discard two groups of them for the analysis of the structural sub-

system of the thana cooperative associations.

Responses from the first

three expansion projects, those in Nator thana in Rajshahi district,

Gaibandha thana in Rangpur district, and Gouripur thana in Mymensingh

district, showed strange breaks in the evolution of those organizations.

Between 1969 and 1970, for example, the number of employees in the

Gaibandha Thana Central Cooperative Association fell from 87 to 48;

the irrigation section of the association was disbanded that year, and

the size of the transportation section was cut by 60 percent.

In

Gouripur thana, a machine station with 40 workers

was supervised fer

four years by one administrative officer, Mr. M. A. Kadir, whose span

of control was enormous by the standards of most organizations.

Several other factors bore on the decision to discard structural data

from the final 13 projects that were started up in Comilla district in

1968.

Funding for'that last increment of pre-independence expansion

projects was constricted due to the hostility toward the Comilla

academy of the Governor of East Pakistan, Abdul Monem Khan.

A maximum

of only two data points were collected from these 13 projects (the

questionnaires for 1969 and 1970), so their utility in time series

analysis was slight.

Also, some of the questionnaires for the 13

projects were completed impr0perly, so that data set was of doubtful

validity.

After the subtraction of data from the projects founded in

1963-64 and in 1968-69, the remaining sample of expansion project

questionnaires consists of 29 responses from the 7 thana cooperative

associations that were launched in Comilla district in 1965-66.

208

I7

I!

II

t

  0

a

a:

an x xx

a p

:0

.19

4a a a:

”
n
a
v
“
A
0
1

!
1
4
.

o
h
W

”
!
r
o
n
i
M
t
a

r
M
/
a
t
o
I

”603! 41'”

54477186001“ a! ’1“ watt! 19‘1“” OF
JUPO'R V410” IN 7"]! /“ flea/[€733

40::

 

209

Eighty-three percent of the potential total of 35 questionnaires were

received from this group of organizations.

Each questionnaire was to be filled out with data as of June 30,

the end of the fiscal year on which the cooperative projects operated.

My first data point for the cooperative projects founded in 1965-66,

therefore, is June 30, 1966.

Since my data set for these projects

begins in 1966, rather than in 1965, I will refer hereafter to the

projects that were launched in 1965-66 as the "1966 projects."

When Peter Blau analyzed the organizational structure of local

employment security offices, he found that the percentage of super-

visors was scattered around a curve resembling the segment AB in the

combined effects curve of Figure 4—8.

A similar scattergram

for the

1966 projects, with a curve traced through it according to the same

methods used by Blau,28 is shown here as Figure 4—10.

The resemblance

of the curve DEF in Figure 4-10 to the curve ABC in Figure 4—8 is

strong.

How much confidence can be placed in the U-shaped curve in

Figure 4-10?

Was the mean of percent supervisors in the middle, point

E, significantly lower than the two points on either side, D and F?

0r might the apparent minimum have been a random phenomenon, ascribable

to chance and a small sample?

Two sample, one-tailed T-tests were

made between point E and the points plotted on either side.

The

T-tests reveal that there was about a 20% chance that the data plotted

as a minimum were drawn from the same population as the data for the

neighboring points.

(See Table 4-1.)

In other words, the existence

of the administrative Optimum for the 1966 projects can be affirmed,

according to this test, with about 80 percent confidence.

Table u-l:

Estests for minimum lg percent'gfi supervisors lg Egg

1935 projects.

210

 

1966 Projects

 

Comparison

t

Freedom

of t

Degrees of

Probability

 

de

fe

.926

.826

22

15

.82

.79

All of Peter Blau's observations of U.S. employment security

agencies were made in the same year, in a cross-sectional survey, so

his data do not reveal growth patterns in the organizations he analyzed.

In the present study all the thana cooperative associations were

surveyed longitudinally.

As a consequence, a picture of their evolu-

tion can be drawn, as in Figure 4-11.

There was a change in the

growth pattern of these organizations, it appears, after 1968.

The

rate of their growth, as measured by the annual increase in their

employees, decelerated sharply.

1969-70 was the post-Ayub era in Pakistan.

During the previous

decade President Ayub Khan had been a vigorous supporter of the

Comilla experiment, and of its expansion into other thanas.

Without

the influence of Ayub Khan in the central government to hold him in

check after 1968, the conservative governor of East Pakistan, Abdul

Monem Khan, enjoyed a longbsought opportunity to put the Comilla

experiments in their place by restricting funds.

The ceteris paribus

condition was violated in the two years at the end of the decade, and

as a consequence the cooperative projects were driven by a different

211

  a

I“ ”OJ!6719

l“

”a

M

I?”

3‘

E3
«3

£3

\
33

2

'3».

u‘

‘4:
d:

a
/

a

//

D

9

a

””wflwfiflfl7/M/IO

J/Zl

l/Wfll 4 -/l lVOLUf/ON 0F paw7" O, JUP!‘ V1101:

VIIIUI II“ /N 7:41AM (”FIR/U Tl V6

WC/AT/ONJ. pO/NTJ M7760 MI

AV!“ 6! £644de

dynamic.

212

From a theoretical as well as a practical standpoint, this environ-

mental change was unfortunate.

Had it not been for the fall of Ayub

Khan and the rise of Abdul Monem Khan, the trajectory of percent of

supervisors versus size for the 1966 projects might not have turned

downwards after 1968, thereby casting doubt on the existence of a

U-shaped curve showing an administrative optimum for these thana

cooPerative associations.

Any extrapolation or generalization based on the points plotted

in Figures #—10 and 4—11 would be essentially an exercise in curve

fitting.

For a more theoretically rigorous interpretation of this

data on the structural subsystem of thana central cooperative associa-

tions, let us turn to the sort of dynamic model of organizational

structure preposed by Norman Hummon.

I have already mentioned several

points of disagreement with Hummon's structural control model.

The

dynamic model elaborated below overcomes some of those deficiencies.

In order to make further use of the thana cooperative association

data -- which, instead of the two annual data points in the Hummon

study, covers several years -- let us construct a mathematical model

based on difference equations rather than on differential equations.

Each annual data point can then be used in the estimation of system

213

parameters, and in testing the accuracy of the model.

For*the study of

public organizations, at least, difference equations seem more appro-

priate than differential equations anyway.

Changes in the staffing of

a publicly funded organization occur mainly in annual increments, not

as continuous and infinitely divisible alterations.

Let us begin by positing a set of linear algebraic equations

similar to those which Hummon proposed to describe ideal, unobservable

relations among system variables-

As the system variables themselves,

we take W for the number of workers (non-supervisory employees) in a

thana cooperative association, D for the number of divisions, S for the

number of supervisors, and L for the number of levels.

Precise descrip-

tions of how these variables were evaluated from the survey question-

naires are to be found in Appendix B.

Since I criticized the Hummon

model for not taking the influence of external factors into account,

let us include one input variable:

T fbr the total loan transactions

of the thana cooperative association.

Unlike the state (or level)

variables W, D, S and L, the variable T represents a rate.29 It is

the sum of the annual rate of loans issued by the thana c00perative

association, and the annual rate of loans repayed to the association.

T is an indicator of menoed. If we were modelling the growth of a

college, instead of a thana cooperative association, an analogous

input variable would be the annual rate of student credit hours taken.

As our set of structural control equations we have, then:

W*-_-

a12D+a138+a1LLL

+g1T

D*=

a21 W

+ a23 S + a2)+ L

+ g2 T

(4.5)

8* = a31 W + a32 D

+ a3“ L

+ 33 T

L*=au1w+a42D+au3S

+3413

214

where W*, S*, D* and L* are the ideal values for workers, divisions,

supervisors and levels, given the configurations described on the right

hand side of each equation.

Suppose, now, that the annual change in any one of the observable

systemic variables was governed by a cybernetic feedback process

directed at reducing the difference between the ideal and the actual

numbers of workers. divisions. supervisors, and levels.

Mathematically,

such a feedback adjustment could be described as:

AW = o1 (W* - w)

AD: o2 (D* - D)

As = c3 (3* - 3)

AL: CL} (L* - L).

(4.6)

By substituting the expressions on the right hand side of each

equation in #15 for the starred variables in #ué, we arrive at the

following open system dynamic model:

AW = -c1 W + a12 c1 D + a13 c1 S + a1# 01 L-+ c1 g1 T

AD

a21 c2 W - 02 D + a23 c2 S + a24 c2 L'+ 02 g2 T

(4.7 )

AS

a31

c

3

W + a

32 c3 D

- c

3 S + a3“ c3 Li+ c3 g3 T

AL = a“,1 04 W + a42 c4 D + a43 c,‘L S - c,+ Li+ cu gu T.

This is cumbersome, so we will summarize the model in the matrix

format ‘AX.= AX + GU, in which X, the vector of system variables, is

(w, D, s, L)T: U, the vector of input variables, is (T, o,

o, o)T, and

the matrices A and G are

215

-c1

+ a12 c1 + a13 c1 + a14 c1

-

a21 c2 - c2

+ a23 c2 + a2,+ 02

A -

a

c

+ a

c

- c

+ a

c

(““8)

31

3

32

3

3

34

3

a

c

+ a

c + a

c

- c

L_4i

4

42

4

43 4

4 __J

o:

01 g1

c

g

2

2

c3 g3

c

g

O

0

0

0

0

O

0

0

O

O

O

0

 

._.}.

 

_‘t "

 

.

(4.9)

What is the meaning of the expression AX?

It is a symbolic

representation of the change in system variables from one time to

another.

The usual way to designate the data sampling points is with

subscripts:

Xn denotes the value of the system variables at time n.

The expression

AX, then, can be rewritten as X

- Xn.

This yields

n + 1

the equation Kn + 1 - Kn}: AXn + GUn.

Some very simple matrix algebra

allows us to rewrite this as:

xn + 1 = (A + I) xn + GUn,

(4.10)

in which I is the identity matrix.

Still further simplification,

substituting B for (A-+ I), gives

X

n + 1 = an + GUn'

(4-11)

Equation 4.11 can be used as an estimating equation for this model

of structural change.

The set of variables on the left hand side,

represented by the vector Xn + 1, can be treated as dependent variables

and regressed against the independent variables on the right hand side.

The regression gives estimates of the parameters contained in the

matrices B and C.

One feature of this method is peculiar, though, from

216

the perspective of standard regression theory.

The regression must be

forced through the origin, since no vector of constants appears in

4.11.

While it is entirely possible to accomplish such a regression,

the results are hard to interpret.

The usual measure of the proportion

of variance explained by the regression, r2, does not apply because the

correlation coefficients have not been adjusted for the means of the

variables.

Rather than forcing the regression equations through the

origin, a safer and more general procedure would be to allow for the

possibility of regression constants, and to see if they did indeed turn

out to be zero.

Constants in the equations for w, 13*, 3* and L* would enter the

overall model as a vector R = (c1 r1, c2 r2, c3 r3, cu rh)T, which

would be the final term in the matrix equation

Xn + 1 = BXn-+ GUn + R.

(4.12)

I have used equation 4.12 as the basic estimating equation for a

mathematical model of the structural subsystem in thana c00perative

associations of the Comilla type.

The dependent variables in the

vector X

were regressed against independent system variables in

n + 1

the vector Xn and the input variable in the vector Un'

When stepwise regressions were performed on the structural sub-

system data from the 1966 expansion projects, the results in Table

4-2 were obtained.

Excluded from the table are all regression steps in

which an independent variable entered the equation with an F ratio which

was significant at the .25 level or above.

Even though this-exclusion

policy was based on a generous entry criterion, its consequence was to

render the coefficient matrix B sparce, with few entries off of the

main diagonal.

And because of the exclusion policy, the matrix G was

217

e
c
n
a
c
i
f
i
n
g
i
S

l
l
a
r
e
v
O

e
l
p
i
t
l
u
M

F
f
o

F

R

R
+

U
G
+

n

n

X
B

=

X

1
+
n

 

0
0
0
.

6
8
.
0
2

2
8
.

7
9
.
1
2

+

n
D
9
6
'
2

-

n
W
5
8
.

=

1
+
n
w

s
n
o
i
s
s
e
r
g
e
r

r
u
o
f

l
l
a

r
o
f

s
n
o
i
t
c
i
d
e
r
p

f
o

r
e
b
m
u
n

e
h
T

.
s
t
c
e
g
o
r
p

6
6
9
1

e
h
t

r
o
f

s
r
e
t
e
m
a
r
a
p

m
e
t
s
y
s
b
u
s

l
a
r
u
t
c
u
r
t
s

f
o

s
e
t
a
m
i
t
s
e

n
o
i
s
s
e
r
g
e
R

:
2
—
4

e
l
b
a
T

3
.
3
2

s
a
w

0
0
0
.

0
1
.
3
2

2
7
.

8
4
.
1

+

n
D
0
7
.

=

1
+
n
D

0
0
0
.

1
9
0
&
2

’
4
1
8
0

6
9
0
2

+

1
r
8
8
8
.

+

D
D
M
.

'
"

=

1
m
S

9
6
.

4
1
.

+

0
0
0
.

7
1
.
9
1

  

L
5
9
.

=

1
+
n
L

218

In order to compare these results with those reported for the

Hummon study of structural control in finance departments, one must bear

in mind that the matrix B in the recursive equation X

= BXn + GUh + R

n + 1

is not directly analogous to the A matrix in Hummon's differential

equation model X,= AX.+ R.

The closest analog to Hummon's differential

matrix A is the difference matrix A in the general equation AX.=

AX.n + GUn + B.

This matrix A and the B matrix in Table 4—1 have the

relation A = B - I.

Subtraction of the identity matrix gives:

K15

_

:1

'—

-,

-c2

A =

=

(4.13)

-.12

-c3

 

a_.

-005

 

._J

 

._.

“Cu

._.

 

This result is similar to the one Hummon obtained.

He reported

negative inertial effects of the variables W, D, S and L upon their

own rates of change.

His study showed the values of the diagonal

coefficients to be:

c

1

= 010’ C

= 0214‘, C

-

23, and. Cu: 023.

2

3 ' '

In

other respects, though, the matrix A given in 4.13 is

different from the A matrix reported in the Hummon study.

(Hummon's

A matrix was reproduced earlier in this chapter’as equation 4.4.)

EVen

the simplified version of Hummon's findings, shown earlier as Figure

4-4, does not have much resemblance to these results.

In his study of

state and local finance departments, Hummon found that several system

variables influenced the rate of change of other system variables.

His A matrix was dense with coefficients off of the main diagonal.

For the thana cooperative associations, the input matrix C is

empty, indicating that their evolution was not influenced by workload.

at least insofar as the rate of loan transactions can be taken as an

219

indicator of workload.

In this respect, Hummon's closed system model

would seem to have been vindicated.

Graphically, the structural subsystem model derived from these

regressions can be depicted as in Figure 4—12.

A comparison of this

diagram with the simplified causal loop diagram for the Hummon model

Shown as Figure 4—4 demonstrates that the only commonality between the

two sets of findings is in the negative inertial effect that each

structural variable has on its own rate of change.

While the Hummon

study of government finance departments in the United States snowed

that organizational divisions influenced workers and supervisors, the

effect was positive rather than negative.

The effect was, however, in

the same direction, leading from the differentiation variables toward

the size variables.

So in contradistinction to the published findings

of Peter Blau, MarShall Meyer, Grant Childers, Bruce Mayhew, and other

major researchers who have investigated the structure of formal

organizations, I am led,re1uctantly, to assert that this set of thana

cooperative association data appears to support the contention that

differentiation -d-size, albiet diminished size rather than expanded

size.

Input

Size

Differentiation

Variables

Variables

Variables

220

I 12

O 05

/"‘\\

/’\\

CID/I

\

\

\\

‘\<44

‘\

\\

CT)

'®__ ._.—34.6.2 __ _\2@

(

\ __ /

o 15

\

I

\./l

' 3

Administrative

Variables

Production

Variables

Figure 4—12:

Causal loop diagram for the structural subsystem in the

1966 projects.

Since there are so many features of Figure 4—12 which are contrary

to expectation, it would seem worthwhile to look into the possibility

that its anamolies -- such as the closed character of the system, or the

negative impact of divisions on workers -— might be due to a break in

the time series after 1968.

In Figure 4-11 we saw that the system

trajectory took a sharp turn downwards in 1968, after the fall of

President Ayub Khan.

By cutting out the predictions in the post-Ayub

era, and thereby reducing the sample of predictions from 23 to 12, we

obtain a data base for the regression estimates shown in Table 4—3.

A graphical presentation of these regression results, in Figure

4—13, makes it clear that in most respects they are not in accord with

the findings from the whole series of observations of the 1966 pro-

jects, running from 1966 to 1970.

The structural subsystem portrayed

in Figure 4—13 is open; it shows only one negative inertial effect of

a variable upon its own rate of change: and the interaction among state

variables supports the theory that size

differentiation.

221

 

2
0
0
0

}
1
8
.
2
1

6
8
0

1
5
0
1
1

'
"

T
3
0
0

"

“
“
6
0
0
2

ll

1
m

3

e
c
n
a
c
i
f
i
n
g
i
S

l
l
a
r
e
v
O

e
l
p
i
t
l
u
M

F

f
o

F

R

R

+

n
U
G

+

n
X
B

=

1
+
n
X

.
d
e
d
u
l
c
x
e

0
7
9
1

d
n
a

9
6
9
1

r
o
f

s
n
o
i
t
c
i
d
e
r
P

.
a
r
e

n
a
h
K

b
u
y
A

e
h
t

n
i

s
t
c
e
j
o
r
p

6
6
9
1

e
h
t

r
o
f

s
n
o
i
s
s
e
r
g
e
r

m
e
t
s
y
s
b
u
s

.
l
a
r
u
t
c
u
r
t
S

:
3
—
4

e
l
b
a
T

3
0
0
.

3
7
.
1
1

5
8
.

‘
6
1
.

-

n
D
4
8
.

+

n
W
5
0
.

II

1
+
n
C:

3
0
0
.

1
9
.
5
1

8
7
.

9
8
.

+

n
5
1
1
.
1

II

l
m

U)

s
i
h
t

n
i

n
L

e
g
n
a
r

t
n
a
t
s
n
o
C

  

=

1
+
n
L

222

Input

Size

Differentiation

.11

(:EEE>

<::)

Administrative

6).- __ 3.0.3.. -@— ' 05

 

:Q)\

Production

1.06

\_,/l

-.16

Figure 4—13:

Causal loop diagram of structural subsystem parameters

for the period 1966-1968.

Solid lines indicate positive

impact, dotted lines negative impact.

Where lies the truth of the matter?

Was the structural subsystem

of the thana cooperative associations influenced by workload or not?

And if it was, how could the causal relationship be such that the more

work there was to be done, in the form of c00perative loans to be

issued and collected, the smaller the workforce that would be required?

And does differentiation

size, or does size

differentiation?

From a small sample study such as this one, no clear resolution

of these questions can be obtained.

Inasmuch as the findings of this

study are contradictory from one sub-sample to the next, I cannot say

that I have reached any firm conclusions about the interaction of state

variables in the structural subsystem of the thana cooperative associa-

tions, or about the influence of workload on the evolution of these

organizations.

In the absence of well validated detail about the structural sub-

system of the thana cooperative associations, the most prudent course

seems to be to adopt a conservative standard, and pursue the implica-

tions only of those research findings for which the strongest evidence

is available.

To maximize the size of the sample on which our

223

conclusions would be based, let us work from the entire series of

observations, from 1966 through 1970.

Furthermore, let us impose a

more rigid criterion for permitting independent variables to enter the

regression equations:

instead of a .25 significance criterion, let us

say that no variable will be allowed into the regression unless it

enters with an F ratio which is significant at at least the .10 level.

Upon imposing these more conservative standards, we obtain the

regression results in Table 4-4.

This is a closed system model, in

which all four state variables are independent of each other.

Size

does not lead to differentiation, nor does differentiation lead to size.

Table 4—4:

A conservative interpretation of the structural subsystem

parameters for the 1966 projects.

F level for entry set at

significance of .10.

X

=

n+1

BX

n

+ GU

+ R

R

F

of F

n

Multiple

Overall

Significance

 

Dn+1 =

.70Dn

+

1.48

.72

23.10

.000

n+1 =

.898

+

1.32

.82

45.59

.000

  

To demonstrate the dynamic impliCations of these regression

results, the coefficients shown in Table 4-4 can be incorporated in the

solution of the mathematical model of the structural subsystem.

Since

the matrix C is empty, the model can be reduced to the form .

X

= BXn + R.

If such a system has an equilibrium'X*, a point in

n+1

phase space from which the solution does not deviate with the passage of

time, then the system solution has the form

224

xn = x* + 8n (xo - x*),

(4.16)

in which X0 is the vector of initial values of the system variables.31

Recall that the ideal number of workers in the thana cooperative

association was assumed to be W

*2

a12 D.+ a13 S + a14 L

+ $1 T +

r

1.

All of the subscripted "a" and "g" coefficients in this equation

have been found to be zero.

So the ideal number of workers is

W* = r1.

What is the value of r1?

With a little manipulation, r1 can

be derived from the regression results.

In Table 4—4 the equation for

the number of workers in the next time period was Wn + 1 =

.77 Wn + 15.37.

The remainder, 15.37, is equal to c

1 r1'

To find c

1’

we subtract 1 from 0.77, since c

is a diagonal element in the change

1

matrix A = B —'I.

Therefore,

W* = c1 r1 e c1 = 15.37 + .23 = 66.83.

(4.17)

According to this method, the full set of equilibrium, or ideal,

values of the state variables can be determined as W* = 66.83,

D*'= 4.93, 5* = 12, and L* = 2.8.

Given the ambiguities in interpreta-

tion of this data set, I wOUld not place much substantive significance

on these values.

Since the matrix B is in this case a diagonal matrix, any power

of the matrix can be found simply by raising the coefficients along the

diagonal to the desired power.

(Actually, matrix methods for solution

are not even needed, since the state variables are independent of each

other.)

For purposes, the most interesting predictions of this mathematical

model are those for W and S, since these variables can be used to form

the variable Size = W + S, and Percent of Employees who are

225

 

  

—-or~AM/c ANA“my

---- an: 77:: MAL/5.6

z

omaxaawvxwm

J/ZI

fill/ll 40 /4

OYNAM/G M0062. Plea/C7701“! COM/’4!” WITH

1718776 [VINO/~65 m: 71/! lm PIG/(67".

226

Supervisors = W—E—S'

A plot of the model's predictions for Size versus

Percent of Supervisors is shown as the solid line MNO in Figure 4—14.32

The dotted curve in the figure is the same as curve DEF in Figure 4—10.

The similarity of the two curves is remarkable, and may be taken as

further evidence in support of the contention that there was an

administrative optimum for the thana cooperative associations.

Point N,

the minimum of the dynamic curve, is at a size of 53 employees.

By

using the static investigative methodology employed by Peter Blau,

point E on the dotted curve was found to be at an organizational size

of 49 employees.

From the proximity of these two estimates of the

location of the administrative optimum, we can assert that thana

cooperative associations were most efficiently staffed at a scale of

about 50 employees.

Beyond that size, they became loaded down with

superVisors.

This chain of reasoning proceeds from the very narrow assumption

that a rural development organization should operate at the scale of

greatest administrative efficiency, as expressed in the ratio of

supervisors to size.

Such a criterion would have made sense if it had

been possible to serve the needs of the rural population in a thana by

establishing several thana cooperative associations (each of an optimum

size) covering the same region.

This was not ordinarily a feasible

alternative because the goVernment assumed that one thana should be

served by one thana cooperative association.

So long as that assumption

was respected, the only really satisfactory alternative would be to

subdivide the thanas, much as Akhter Hameed Khan advocated splitting

up Laksam into several thanas each of which wOUld be of more manageable

dimensions.

But his call for revamping the territorial hierarchy of

administration was not heeded, and probably never will be.

227

Short of the radical solutions of splitting the thana cooperative

associations or splitting the thanas, there was an intermediate way of

restraining the thana cooperative associations from the temptations of

administrative overload and inefficiency.

(These temptations present

themselves to the administrators of any organization, but they are

especially acute in the less developed countries, where self-serving

public bureaucracies are a principal source of employment for'the

middle class.)

Instead of attacking the problem through horizontal

segmentation, it could be resolved through vertical reorganization.

Such was the solution to which Akhter’Hameed Khan resorted in Comilla

Kotwali Thana, where he split off the field supervision section of the

KTCCA to form the compact subordinate entity known as the Agricultural

Cooperatives Federation.

Before I end this discussion of the structural subsystem, I

should like to report one more curious finding about the thana coopera-

tive associations.

In reviewing the literature about mathematical

interpretations of organizational structure, I encountered

Grant Childers' article on "System Size and Structural Differentiation,"

which I mentioned earlier in this chapter.

Childers developed an

argument for the equation 12(1)) = as + 1), in which E(D) is the expected

value of differentiation in the organization, and S is the size of the

organization.

(These variables should not be confused with D and s in

the Hummon model.)

Childers' measure of differentiation was the number

of ocCUpations in an organization, and his measure of size was total

personnel.

To see whether this theory applied to the thana cooperative

associations, I tested it against data from all of the thana coopera-

tive association questionnaires that were returned to me, including

those for the Agricultural Cooperatives Federation, for the projects

228

a

a

a o

3

\

‘r‘ "

Q

t x

3:

3
Q

2r

3 "

-

o

o

/

//

/

z /’

‘

,

‘

3

V
5‘?

r

§

i

If

,9

   

-—-.4C7UAL

°

'

’

.

5

'

——-P£¢0/67'£10

IV ' ‘9

ommaalaioéomabecmno/zo/x

4125

new?! 4- /5’ :12: any: DIFFIRMT/A 7/0N IN 77/: flaw

empeur/Vt Assoc/4770119.

'

229

founded in 1963-64, for the projects founded in 1965-66, and for the

projects founded in 1968-69.

If the actual number of occupational classifications in the thana

cooperative associations were the same as the expected value of differb

entiation, we would find that a regression of job classifications (as

the dependent variable) against size would have a slope of 0.5 and an

intercept on the job classification (vertical) axis of 0. 5.

When the

regression was carried out, with data from 69 questionnaires, the slope

turned out to be 0.23 and the intercept 6.62.

The anticipated.values

of these coefficients are not included in either 95% confidence interval;

for the slope, the 95% confidence interval extends from .19 to .26, and

for the intercept from 4.66 to 8.58.

The results do not support

Childers' hypothesis about the way differentiation is related to size.

Yet a relationship between size and differentiation clearly did

obtain in the thana cooperative associations, even if it was not the

one pIOposed by Grant Childers.

The regression of the number of job

classifications against size

explained.72% of the variance in job

classifications.

The F ratio for the regression was 168.78, which was

significant at the .000 level.

In the scattergram shown as Figure 4-21

one can see that only a handful of the 69 cases strayed from the dotted

regression line D = .235 + 6.62.

Since the one theoretical explanation which I know of for this

phenomenon does not seem to apply, I am at a loss to explain the very

strong relationship between size and differentiation in the thana

cooperative associations.

I can only speculate that different kinds of

organizations may have different "signatures" of differentiation.

Childers, for example, found that differentiation was related to size

according to three distinct regression equations in vessels,

230

headquarters offices and lifeboat stations in the U.S. Coast Guard.

The regression equation for lifeboat stations near Seattle

(D = .235 + 6.33) was just about identical to the one which applies to

the thana cooperative associations in East Pakistan.

The reader is

free to make his own inferences about the cause of this surprising

commonality.

Notes for Chapter 4

The Structural Subsystem

1 Richard M. Cyert and James G. March, A Behavioral Theory g§.th§

Firm, Prentice Hall, Englewood Cliffs, N.J., 1963, p. 289.

Copyright

1963 by Prentice Hall, Inc.

Reprinted with permission.

2 Kenneth E. Boulding, "Toward a General Theory of Growth,"

Canadian Journal of Economics and Political Science, V. 19, August,

fi—m_w

1953: PP' 326-40-

3 Peter M. Blau, "A Formal Theory of Differentiation in Organiza-

tions," American Sociological Review, V. 35, No. 2, April, 1970,

4 Norman P. Hummon, "A Mathematical Theory of Differentiation in

Organizations." American Sociological Review, V. 36, April, 1971,

PP. 297-303.

5 Marshall w. Meyer, "Size and the Structure of Organizations:

A Causal Analysis," American Sociological Review, V. 37, August, 1972,

pp. 434-441.

6 Ibid., p. 437.

7 Grant w. Childers, Bruce H. Mayhew, Jr., and Louis N. Gray,

"System Size and Structural Differentiation in Military Organizations:

Testing a Baseline Model of the DiVision of Labor," American Journal 9:

Sociology, V. 76, March, 1971, pp. 813-30.

8 David A. Specht, "System Size and Structural Differentiation in

Formal Organizations:

An Alternative Baseline Generator," American

Sociological Review, V. 38, August, 1973, pp. 479—80.

9 Marshall W. Meyer, "Some Constraints in Analyzing Data on

Organizational Structures:

A Comment on Blau's Paper," gmerican Socio-

logical Review, V. 36, April, 1971, pp. 294—297.

.

10 Bruce H. Mayhew, Roger L. Levinger, J. Miller McPherson, and

Thomas F. James, "System Size and Structural Differentiation in Formal

Organizations:

A Baseline Generator for Two Major Theoretical Proposi-

tions," American Sociological Review, V. 37. October, 1972, pp. 629-633.

11 Morris F. Friedell, "Organizations as Semilattices,” American

Sociological Review, V. 32, February, 1967, pp. 46-54.

231

232

12 Norman P. Hummon, Patrick Doreian, and Klaus Teuter, "A

Structural Control Model of Organizational Change," American Sociological

Review, V. 40, December, 1975, pp. 813—824.

13 Ibid. , p. 816.

In most of the research papers cited previously

in this chapter, including Peter Blau's and Marshall Meyer's, the size

of the organization was indicated by a single variable, usually designated

S, which corresponded to the total number of employees in the organiza-

tion.

In the Hummon paper, size is broken down into two components:

the number of production employees, P, and the number of supervisors,

S.

A“ In order for a mathematical model of a simple machine to be

linear, it must be true that if S is a solution and if T is a solution,

then s + T is a solution.

In the case of X = AX, the general solution

isX= eAtxo. IfT= ATandS= AS, thenT= eAtTo, s: eAt so, and

S + T = eAt To + eAt S0 = eAt (To + So), so the system meets the

conditions of linearity.

But for X = AX + R, X = eAt So + R.

If

'
T=AT+RandS=AS+R,thenT=e

At

TO+R,S=e SO+R,and

At

S+T=eMTQ+R+eAt

So+R—e (TO+SO)+2R. Thislast

_

At

equation does not have the form X = eAt XO + R, so S + T cannot be a

solution to the system.

The conclusion is that the system X = AX + R is

not linear.

15 For an explanation of causal loop diagrams and their signifi-

cance in the modelling of dynamic systems, see Thomas J. Manetsch and

Gerald L. Park, System Analysis _a_nd Simulation with Applications 1:9

Economic and Social Systems, Part _1_,

Department of Electrical Engineer-

ing and System Science, Michigan State University, 1974, pp. 16-24.

Causal loop diagrams are used extensively in Dennis Meadows, gt _a_l,

The Limits _t_o_ Growth, New American Library, 1972.

I have added circles

around the variables in Figures 4-1 and 4—2 to make them resemble

standard causal loop diagrams more closely.

Hummon seems to be unaware

of the generality of the type of diagram he has drawn, since he does

not refer to it either as a "causal loop diagram," or by the alternative

name, promoted by W. Ross Ashby in Desigg £11; 2 Brain, of an "effects

diagram. "

16 Peter J. Bertocci, Elusive Villages, Social Structure and

Community Organization _i_:_i Rural gst Pakistan, Ph.D. Thesis,

Michigan State University, 1970.

17 This method of decision-making pervaded.all village institutions

in Camilla.

Few issues were ever brought to a vote in the cooperative

weekly meetings or in the sessions of the local government bodies known

as the Union Councils.

When the voting method was required, as in the

election of Union Council members, the villagers tended to divide into

antagonistic factions which sometimes persisted for years.

233

18 Badruddin Ahmed, Manual p§_Comilla Cooperatives, Bangladesh

Academy for Rural Development, Comilla, 1972.

‘

19 As it is printed on page 4 of the report, the heading of the

organization chart contains a typographical error.

The printed heading

reads "Organization Chart of Kotwali Thana Central Cooperative Associa-

tion (KTCCA) as on June 1967."

The chart can be correctly dated at

June 1969, since it shows the number of members of the Agricultural

Cooperatives Federation as being 11,673, which was the membership

figure for June 1969.

20 The list of fifteen functions which the inspectors ranked was

prepared by the author and Shamsul Huq, who was at the time the director

of the field supervision section of the KTCCA.

When the Agricultural

Cooperatives Federation was formed, Shamsul Huq was appointed as its

director.

21 Harry N. Blair, The Elusivenepg pi; Equity:

Institutiona_.l_._

Approaches pg Develppment ip Bangladesh, Rural Development Committee of

Cornell University, Ithica, 1974, p. 51.

22 A_N 3 Rural Cooperative System for Camilla Thana, Eleventh

Annual Report, 1970:Z;, Bangladesh Academy for Rural Development,

Comilla. 1973, p. 40.

23 Harry W. Blair, pp pip, pp. 49-52.

24 Arthur T. Mosher, "Projects of Integrated Rural Development,"

paper presented at the Seminar on Small Farmer Development Strategies,

Ohio State University, September, 1970 (Mimeograph), pp. 16, 17.

25 Norman T. Uphoff and Milton J. Esman, Local gpganization for

Rural Development:

Analysis pf Egg Asian Experience, Rural Development

Committee, Cornell University, 1974, pp. 67-69.

26 Akhter Hameed Khan, Tour 93 Twenty Thanas, East Pakistan

Academy for Rural Development, 1971, p. 24.

27 Eugene Haas. Richard H. Hall, and Norman J. Johnson, "The Size

of Supportive Components in Organizations:

A Multi—Organizational

Analysis," Social Forces, V. 42, pp. 9-17.

28 Peter M. Blau, pp_gip, p. 205.

Blau segmented the size axis and

calculated the mean value of the percent of supervisors for each seg-

ment.

He plotted these values against the mean value of size within

each segment, and drew a smooth curve connecting the resultant points.

In constructing Figure 4-10, the horizontal axis was divided into three

equal segments, from 20 to 39, 40 to 59. and 60 to 79.

Blau's proce-

dure for tracing a curve was then followed.

.

29 The distinction between level and rate variables is clearly

described by Jay W. Forrester in Industrial Dypamics, John Wiley and

Sons, New York, 1961, pp. 68, 69.

234

30 The observant reader may have noticed the discrepancy between

N = 29 in Figure 4—10 and N = 23 in Table 4-2.

The sample size is

smaller for the regressions because they are based on the number of

predictions derivable from the data.

With two consecutive annual data

points in one thana, for example, only one prediction can be estimated.

31 This solution form is given in John M. Hunter, "Solving Linear

Translate Systems with Matrices," mimeographed handout for Psychology

4OOH, Michigan State University, p. 3.

Hunter's derivation is given in

Appendix A.

32 Point M, the beginning of the trajectory of model predictions,

represents the average size (27 employees) and percentage of super-

visors (13.6) of the 1966 projects at the end of their first year of

operation; that is, on June 30, 1966.

  

-‘

.-

'1‘I'111.1

181,311,111

,1. I"1‘12113".

11: '111‘,II'".‘11,,'11,;‘13-

   

1'

‘

1

     

1,

.4

.

a
.0

  

1.

111‘

.

II1§,1 'tha $I .4

111""1IIII1'II1III1‘I'I‘11'.'.‘11"{11112331-

  
. , ~-
  

1.-1'I‘"I11':":|1',1‘11 ‘

' I1,

,

‘

1.

4

,

.

   

   
     
  
     

"11"1‘,‘I1111"11,I,"i'1,'1,,..1

.y‘ I1III11 "1'

11'"1,,

   
  

  

1.1111 11111111111,

1111 "

..

,

    

1'

   

2.111,3u1"'f”111g4u3.
111%III‘I’EIII'II'.

~.

.,.

'

1'1'1’r‘1rkfiu 1322.1? .‘

‘

1?" .-

L.11""1 III'1111

'111111111,111

1

.1:

M1111? H

III1.1 1111,11111

1,,

11,.”11

11111111111111111111111113
'I1'I1I1" " "
1..“
I:.'
" "111,"11I"1,4'1I"'11,I'11I"'
,1,,1,

I."
1,1“111'.11I1‘}..':

III,I[1;'

‘
11"’

I.,,

.

'

.111”

V 4..."1'1 511:1." "11111'

.

".1,'."1"1|1] :1,,I I1'II

"'1""' 1111'1111171'1'1'1,“"1111,

  

  

1I‘ "I" 'I

I

. 1‘ -'.,;I-‘;.

I ’1‘,

1,11.

,

I,

'I

'

1 ,11',:‘.1

'

"

1": 11.14% I‘"

  

.11"? 'r'I

.1 I-

11.

',

""11

'

I‘ll 1,1111,

.'..' ”I"
,Vi11..: I.”

'1.1

1, 11

i

"A.

”"1111

')

III1

—‘

1’11‘

..-‘

"‘,I.,',.1,;H1-E;aw11:., 1

,I

‘

.111141 II.1 ,111

('11,

I

.

l

,

..

.1: 1

, 1'14", fig.1"45

"

-.

 

1111“11.. 1111111111, .,I:

  

:,'

'M1' 1

,I‘1

’-

J111II'1 I11.1

”11""1"11 1'

41,35:

11'1 1"" "'1I
.

:_

'

I

.

-.

.45.““ff-.11.,""""""1

11.1,. -.

I‘..',,,,.I f '..., ,1" I-“'.1'4' 1,. ,3.

~.—.i_,,.,.III-IIVI.,.
1" II1- "1'

0'1{1'5 I1"""1"“1-4-1.{1‘

~

-

  

1,.

)5;-

     

);v

x

V 5"; “ | "

Lb
.

  

161105!

  
“111’,,.,..1"1111‘53'1'

""E‘,‘,':35;

J54};';,-;_' '.
:1": #1117:

  

.

   
   
   

~2115‘ ‘

n

"1..

' "11111.

   

PI. "1.III.,1...IIIIIfiI'I-i'i';1...

  

'

'

,1.:“1l11'1

1'111'1'11,1l I" _

.‘.,'5,I.,11|1'

("'.Ol,,1.1‘1'1.""111;:'>'!jl") 1'111'1M111'I'J'11'1""1'1151"?%

.

‘II

:.II;.~.. I ‘ II III-:1 :

',"III.I . -"I “'.1'11:1.“ ‘

I1I

.

4.

1I

“11 ‘11'1'11II111 111111'1

',:I1111|, ' 1,11"I111.1,,., , 1.

111111, ,1' ':1

4"11': ""111111'13'

"I

'1II,11:.11‘,'I1'I. V1}

.''.‘II."I"..

'1111..I...II""II1' "111111111111"1'11'11I1

   

,.,I.

I

~

I."?1'III.

""

.

'I"""'I-'l

""1""1"11IT1,‘1‘:"' 111(111 I"

_

.,,II....I4..-I,l .

‘111,111,

...' 'I'.

1"..11" W1

1

l

. I

”1,, I

I...
 
 
 

“-I.I... III."III"IIIII"

- 111.’,,' 1'11111'111I11‘1111p11

 

 

 
 
 

.I-mm. Ilm. '1'11

" ''111'1111111.. .'

1,,HII,,11I‘

M —-

1'

" '

.III.

"

"

1|

 

 

 

'

.

mmmxmmmwmm .

3129 0009

 

 

F

_
‘
.

.
_
.
.
.
A
'
"
"
~

'

—
o
.
.
.
_
’

-
'

_

-
‘

.
.
_
§
-
»

,
-

—
—
—
t
-
—
—
—
A
.
—
”
‘
—
.
_
“
~
*
-
‘
I
~
"
"
_

‘

P

H

»

1

35¢ Per day per item

OVERDUE FINES:

J‘
7’ 33%“ A
“or/5....” .

'5 ' “3:3,?”-

RETURuzus LIBRARY MATERIALS.
Place in book return to remove
charge from circulation records

.

 

 

 

CHAPTER 5

The Technical Subsystem

Operating

Operational

management

Strategic

Plant

Control

Control

Planning

Static systems

Simple machines

S

ST

Machines with input

Organic systems

A social system is a structuring of events or

happenings rather than of physical parts and it

therefore has no structure apart from its func-

tioning.

Physical or biological systems such

as automobiles or organisms have anatomical

structures which can be identified even when

they are not functioning.

In other words. these

systems have both an anatomy and a physiology.

There is no anatomy to a social system in this

sense.

When the biological organism ceases to

function, the physical body is still present and

its anatomy can be examined in a post-mortem

analysis.

When a social system ceases to func-

tion, there is no longer an identifiable struc-

ture.1

Introduction

What does an organization do?

What are its functions?

One does

hear charges, usually political in character, that certain comatose

235

236

agencies do nothing.

Yet a defender of the bureau can always be found

who explains that the personnel of the unit do this or do that.

There

are always functions to be found.

Without them, as Daniel Katz and

Robert Kahn explain in the passage quoted above. the grouping could

not properly be described as an organization.

Anthropologists such as Bruno Malinowski2 and sociologists such

as Philip Selznick3 have built their professional reputations around

descriptions of the interplay of structure and function.

The conceptual

framework of the structural-functional school is that "functions deter—

mine the structure of organizations and that structures can be understood

by analyzing their functions."4

In anthropology, sociology, and political

science. structural functionalism had its heyday in the United States

in the nineteen fifties.

Structural functionalism is a powerful

analytic approach. but it has been displaced by more encompassing per-

spectives.

Modern organization theorists such as Amitai Etzion15 are

inclined to examine the three way interaction among structure, function

and goals.

Other contemporary pundits take the even broader view of

systems theory. which in the opinion of writers such as Fremont Kast and

James Rosenzweig6 is capable of interpreting the complicated relation-

ships among structure, function, goals, management, psychosocial

processes, and the organizational environment.

The price to be paid

for such inclusiveness should be intuitively obvious.

The more factors

are considered, the more diffuse and potentially incoherent is the

analysis.

Qualified credit is due to the systems theorists, though, for

attempting to sharpen our perception of organizational functioning.

Because of the commentary of the systems analysts. it is no longer so

fashionable to refer to the functions of an organization.

Instead,

237

analysts writing in the contemporary vein discuss the technology of the

organization.

As defined by Kast and Rosenzweig. organizational

technology is the technique used for the transformation of inputs into

outputs.7 For'a steel mill, this definition of technology would lead

to the consideration of open hearth furnaces and such -- the common

concept of technology.

For'a service organization like the Comilla

cooperatives, however, the usual ideas of inputs, transformational

processes. and outputs must be reexamined.

Technology needs to be

thought of as being embodied in more than a set of mechanical artifacts

(such as irrigation pumps).

Broadly speaking, the up-to—date idea of

technology encompasses the human interactions which result in providing

a service, and even the ideas, rules. and standards according to which

those productive interactions are patterned.

For this enlarged concept

of technology it might seem that we could just as well substitute the

fuzzy old idea of functions.

The same conundrum which confronted us in the last chapter'arises

here as well.

Is it possible to construct realistic closed system

models of the structural subsystem or of the technical subsystem of

organizations? For the structural subsystem, we found that the organi-

zation chart. which is a closed and static representation, has all the

characteristics that would be expected in the kind of model which

Kenneth Boulding termed a framework, and which I have called a static

system.

But when it came to the mathematical representation of the

structural subsystem. my presentation in Chapter 4 wavered back and

forth between the closed and Open systems points of view.

For the

technical subsystem this theoretical issue presents itself more force-

fully.

As was noted above, Kast and Rosenzweig have defined organiza-

tional technology in terms of inputs and outputs, which clearly

238

indicates an open system perspective.

How can I justify, then. the

classification of the technical subsystem in the row of the typological

matrix for simple machines?

The dotted horizontal line in that matrix,

which is reproduced at the head of this chapter, designates the division

between closed and Open systems.

The T which stands for technical sub-

systems is placed above that dotted line, in the region of closed

system models.

The breed of organization theorists who have been infected, as I

have, with general systems theory often show a fondness for the compari—

son of system structure with anatomy and of system function with physio-

logy.

An example of this predilection is to be found in the passage by

Daniel Katz and Robert Kahn at the head of the chapter.

Much of the

subject matter of physiology does consist of closed system models.

The

circulation of the blood, for example. can be modeled to a good approxi-

mation as a closed system. even though nutrients and oxygen are contin-

ually entering the bloodstream, and waste products are being extracted

from it.

If these inputs to and outputs from the bloodstream are

ignored. a great deal can still be explained about the nature of the

circulatory system.

The criterion of classification for an organizational subsystem,

I suggest. is that it

should be located in the cell of the typolo-

gical matrix which permits a minimally acceptable description of the

subsystem.

An organization chart, which is a static closed system

model. provides a minimally acceptable description of the structural

subsystem.

Therefore the S for structural subsystem is appropriately

placed in the upper left hand cell of the typological matrix.

If the

technical subsystem can in some minimally acceptable sense be described

as a closed system, then the placement of the T in the typological

239

matrix will be justified.

In the last section of this chapter. where

a simulation of the credit Operations of a thana cooperative association

is presented, we will. therefore, examine whether in some minimally

acceptable sense that model of the technical subsystem can be considered

a closed dynamic system. or simple machine.

One functional prescription from the Cornell canon deserves to be

heeded in this chapter.

As the third of their seven maxims on the

design of rural deve10pment organizations. Uphoff’and Esman asserted,

"Though specialization may be indicated for some activities, local

institutions should ordinarily be vested with multiple function; in

order to achieve benefits of scale and to facilitate the local integra-

tion of services."

Relevant Research

In the social sciences, the study of technology has been primarily

a concern of economists.

The classic example of such a study was

provided by Adam Smith, the first great economist, who described how

the specialization of labor could increase the production of pins.

The

economic theory of production functions (in what different ways can

units of labor and units of capital be combined to produce 10 widgets?)

is essentially a theory about technologies; for each production function

there is a distinctive underlying technology.

Contemporary organization theorists have paid relatively little

attention to technology 22£.§2-

Although a spate of studies have

recently appeared which have examined how technology influenced

organizational structure. on the one hand. or organizational management,

on the other, technology in and of itself has not been a topic of

interest to modern organization theorists.

ZhO

Such was not always the case.

The founder of what has come to be

known as the classical school of organization theory, Frederick W.

Taylor, was fascinated by the arrangement of work.

Writing with the

enthusiasm of an evangelist, Taylor once described how his associate,

Frank B. Gilbreth, investigated and restructured the job of a brick-

layer.8

By introducing such technological innovations as a frame for

holding pro-sorted bricks, and an adjustable scaffold which could keep

the supply of bricks at a convenient height (eliminating the need to

turn and stoop to pick up a new brick), Mr. Gilbreth accomplished a

threefold increase in the productivity of bricklayers.

In the spirit

of scientific management, a movement which he founded, Taylor praised

Gilbreth for designing this new configuration for the bricklayer's

workspace, a configuration which permitted the laborer to work with two

hands at once -- picking up a new brick with his left hand at the same

time as he sc00ped up a trowelful of mortar with his right hand.

The

organizational implication of this technological advance, in Mr. Taylor's

view, was that all the bricklayers working in a row along the wall of

a building under construction must be brought up to the same rate of

production, since the wall should rise along its whole length at an

even speed.

It therefore became the responsibility of the construction

management to teach and enforce the new techniques, and to discharge

any workmen who failed to cooperate.

(There were no trade unions to

speak of in that autocratic era at the beginning of the twentieth

century.)

The comparison of a scientifically managed workman to a

science fiction robot has been drawn, and it is appropriate.

By the middle of the twentieth century, complex organizations had

come to dominate the social and industrial landscape in the United

241

States and in Europe, and the attention of many organization theorists

Shifted to the internal functioning of these behemoths.

William Whyte's

exposition of the way that giant corporations enforce conformity and

supress individualism was typical of the mid-century literature.9 The

relationships of personnel to one another within the bureaucracy became

the tapic of interest.

The interface between management and the work

fOrce, which was the level of Frederick Taylor's concern, was generally

relegated to the industrial engineers. or to the experts on collective

bargaining and industrial relations.

That technology should have receeded from view as the size of

typical organizations grew is consistent with the findings of cross-

sectional empirical studies of the relationship of size, technology,

and organizational structure.

A group of researchers at the University

of Aston, in Birmingham, England, investigated these relationships in a

sample of 52 organizations in the Birmingham area ranging from factories

to stores.

They concluded that:

The smaller the organization, the wider the

structural effects of technology; the larger the

organization, the more such effects are confined

to particular variables. and size and dependence

and similar factors make the greater overall im-

pact.

In the smaller organizations, everyone is

closer to the "shOp floor," and structural re:

sponses to the problems of size (for example)

have not begun to show.

In larger organizations,

managers and.administrators are buffered from the

technology itself by the specialist departments,

standard procedures, and formalized paperwork

that size brings with it.10

This explanation by the Aston group is an intuitively reasonable

reconstruction of the relative influence of technology in small and

large organizations.

But this ray of common sense is a rarity in the

literature on technology and organizational structure.

Several

242

empirical studies of organization and technology have been published

in the last decade or so, but they seem to have generated more contro-

versy than agreement.11

The pioneering work in this field was that of Joan Woodward, who

reported in 1965 her findings from a study of some 100 industrial firms

in the English midlands.

Woodward and her’associates found a strong

correlation between the technology of a firm and such aspects of its

structure as the length of the line of command, the chief executive's

span of control, and the ratio of direct to indirect labor.

Woodward's

studies indicated, moreover, that there was an optimum form of organi-

zation for each technology, since the most successful companies within

each industry had similar structures.12

Woodward's research has sub-

sequently been referred to as the "technological imperative" school of

thought, which contends that for every technology there is an appro-

priate organizational structure.

Such a provocative contention is likely to stir other scholars to

challenge the original research methodology, or to find out if the

basic findings can be replicated or refined.

Several authors, for

example, have questioned Joan Woodward's concept of the way that the

technology of an industry should be measured.

She classified the firms

in her study along a scale of technical complexity which ranged

through (1) unit to small batch, (2) large batch to mass production,

and (3) continuous processing.

Examples of these three types of

industrial technologies would be shipbuilding, automobile manufacturing,

and petroleum refining.

Among the alternative scales for a.technological

variable which have since been proposed are that of Edward Harvey,13

who measured "technical diffuseness" by the number of product changes

recorded for the firm in the last 10 years, and that of

243

Van de van and Delbecq,14 who created a matrix taxonomy of work units

from the two dimensions of task

variability and task difficulty.

Each

such researcher, as might be expected, has presented elaborate statis-

tical evidence, in the form of correlations, factor analysis, analysis

of covariance, and so forth, which supports his own interpretation of

the validity or invalidity of the concept of the technological impera-

tive.

Pradip Khandwalla has summarized the findings of the many studies

which were inspired by Woodward's research:

1.

There is no unanimity that technology affects

organizational structure.

Some studies show

a pronounced effect; others show a very weak

effect.

2.

Measurement of operations technology is a

serious problem.

3.

The impact of technology on organizational

structure, if any, is likely to be quite selec-

tive.

In other words, it is likely to affect

some -- perhaps few -- dimensions of organi-

zational structure rather than all structural

dimensions.

4.

The search for empirical relationships between

technology and organizational structure may

best be described as serendipitous.

The major

researchers have not been

guided by proposi-

tions derived from an explicit model of how

technology might affect organizational structure.

5.

There is no clear evidence that relationships,

if any, between technology and organization

remain when size of the organization is con-

trolled.

6.

The research to date has virtually no prescrip-

tive implications.15

Another'major group of organization theorists have looked at the

influence of technology upon organizational management.

In the terms

of reference used in this book, this group of investigators can be said

to have examined the influence of the technical subsystem upon the

244

managerial subsystem.

The preeminent study in this field is that of

Burns and Stalker.16

They investigated the way that 20 English and

Scottish firms which had operated with stable technologies responded to

corporate moves into the rapidly changing electronics industry.

These

firms had previously been managed in a mechanistic way:

that is, the

tasks, methods, duties and powers attached to each role in their fixed

organizational structures had been rigidly prescribed.

Success in the

electronics field, by contrast, seemed to demand an organic system of

management, in which there was a continual adjustment and redefinition

of roles within a flexible network (as opposed to a hierarchy) of

control.

In a review of this research, East and Rosenzweig commented,

"Burns and Stalker*also emphasize the difficulties involved in making

the transformation from the mechanistic to the organic system by the

firms that were trying to move into the electronics industry.

The

unstructured and highly dynamic nature of the organic system often

created anxiety and insecurity on the part of the managers who had.been

used to working in the structured, mechanistic system."17

In a critique of Max Weber's abstract model of bureaucracy,

Eugene Litwak has made a similar argument about the appropriateness of

different methods of management.

The Weberian model is the classic

mechanistic administration.

In the hierarchical bureaucracy described

by Weber, the system of authority and obligations is specified.§.p;iq§i,

and is attached to the roles of the bureaucracy rather than to the

individuals who occupy those roles.

Litwak pointed out that such a

rational, impersonal ordering is suitable only if the tasks.which the

bureaucracy is to perform are uniform.

"In short," he wrote, "where

organizations deal with non-uniform events, a model of bureaucracy may

be more efficient which differs in degree from Weber's in at least six

245

characteristics:

horizontal patterns of authority, minimal speciali-

zation, mixture of decisions on policy and on administration, little

‘g‘priggi limitation of duty and privileges to a given office, personal

rather than impersonal relations, and a minimum of general rules."18

The electronics industry, Burns and Stalker would agree, is character-

ized by non uniformity of tasks.

The organic pattern of management

which Burns and Stalker found suitable in the electronics field, more-

over, is similar to the non-Weberian administrative model described by

Litwak.

It seems that there is relatively less disagreement about the

probable influence of technology on management than there is about the

influence of technology on structure.

The cause of this apparent

consensus, however, may well be that the definition and dimensions of

the variables describing organizational management are as yet so nebulous

as to preclude substantial contention.

In the following sections of this chapter I discuss three signi-

ficant technologies in the Comilla cooperatives:

dry-season irrigation,

tractor cultivation, and credit.

The first two of these are assessed

mainly at the village tier, while the credit system is evaluated pri-

marily at the thana tier.

At several points I have made inferences as

to how these technologies influenced the structure, management, and

costs of the cooperatives.

But I have made no attempt to substantiate

these anecdotal inferences concerning Comilla Kotwali Thana

with

statistical evidence drawn from survey research across thanas.

Toward

the end of the chapter I present a dynamic systems model of.the credit

operations of a thana cooperative association.

Most of the parameters

of this model were estimated from financial data from the Comilla

type cooperative associations founded in 1965-66.

The model is then

246

tested against data from that same group of organizations, much as was

done in Chapter 4.

The focus of the model, though, is exclusively on

the economic aspects of cooperative credit.

The model does not, in

its present state of development, allow for the investigation of the

implications of the credit technology for the structure and management

of the cooperatives.

The system model of cooperative credit has been

constructed, thus, more nearly in the tradition of economic theory

than of organization theory.

For organizational technology, there is

nothing comparable to Norman Hummon's structural control model to draw

upon, so I have built an economic model instead.

The Technical Subsystem at_the Village Tier

As was noted above, the concept of technology and the concept of

function are closely related.

Let us begin our investigation of the

technical subsystem, then, with a review of the functions which are

appropriate to rural development organizations.

In their summary volume for the Cornell study of rural local govern-

ment in Asia, Norman Uphoff and Milton Esman describe six major

functions which constitute, in their estimation, a taxonomy of the most

important things that rural development organizations should do.

To

one of these five, the provision of services to farmers, they alloted

fifty out of the hundred weighting points distributed across the six

functions.

These functions are:

(a

Planning and goal-setting (10 points)

(b Resource mobilization (10 points;

(c

Provision of services (50 points

(1

(2

(3

(4

water (10 points)

fertilizer (10 points)

credit (10 points)

extension information (10 points)

247

(5) marketing (10 points)

(d) Integration of services (10 points)

(e) Control over bureaucracy serving rural

development (10 points)

(f) Making claims for more and better support

of rural development (10 points).19

Although most of what has been published about the village coopera-

tives of the Comilla type emphasizes their participation in the provision

of services, in point of fact all of these functions were exercised by

the primary village cooperatives.

In most cases the primary coopera-

tive did not perform these functions alone; rather they did so in

collaboration with the thana level association.

Function e, control

over the bureaucracy serving rural development, is the most obvious

instance in which the village cooperative would have been unable to

fulfill one of these rural development functions without the support of

the thana organization.

The installation and repair of irrigation

pumps, to take one example, was the responsibility of’a section of the

thana cooperative association headquartered in Comilla town.

Electrifi-

cation of the motors of tubewell pumps, which resulted in economies of

pump operation and maintenance, required coordination at the provincial

level of government, in Dacca, between advocates of the Comilla coopera-

tive project and the leaders of the East Pakistan Water and Power

Development Authority, which had to be convinced to string power lines

to particular villages in Comilla thana.

At the thana tier, responsibility for each of four of the five

sub-functions listed above under "Provision of Services" was allocated

to a separate section of the thana cooperative organization.

The

pumping of irrigation water was the responsibility of the water section

of the machine station of the KTCCA,and the provision of credit was

the task of the Agricultural Cooperatives Federation, which had

248

separate sections for extension and marketing.

Only the provision of

fertilizer, among the five crucial services in Uphoff and Esman's

list, was not made the special responsibility of a distinct organiza-

tional section at the thana tier.

At the village tier, by contrast,

all of these functions were fused within the structurally undifferen-

tiated primary cooperative societies.

In accordance with Eugene Litwak's

theory, we might expect that the multiplicity of functions performed

by the village cooperatives would cause them to have horizontal patterns

of authority, minimal specialization, a mixture of decisions on policy

and administration, little a priori limitation of duties and privileges

to a given office, personal rather than impersonal relations, and a

minimum of general rules.

All of these attributes did, in fact, apply

to the village cooperatives.

Although the irrigation of dry season crops falls under category

c, the provision of services, in the above taxonomy of rural development

functions, the accomplishment of this objective entailed the fulfill-

ment of all of the other rural develOpment functions as well.

Or to

put the matter the other way around, in order to provide water for dry

season irrigation, a multi-functional organization with village roots

such as the Comilla cooperatives was a necessity.

Zaker Hussein, who

was the chairman of the KTCCA during much of the nineteen sixties and

who became the director of the Bangladesh Academy for Rural DevelOpment

at Comilla in the mid-seventies, graphically described the necessity

of village organization:

For making pump irrigation a success the organ-

ization of a village or'a pump group is a must.

A one-cusec pump can irrigate 60 to 70 acres.

Unless a compact block is organized the Opera-

tion is bound to be wasteful.

Turns of irri-

gation have to be decided upon in order that

249

each plot gets irrigation at regular intervals.

There will be complaints: there will be dis-

putes: there will be farmers who would like to

go their own ways.

All these problems have to

be tackled.

Unless there is an organization

to take care of all these problems there is

bound to be chaos and confusion leading to the

collapse of the programme.20

Among the most important functions of such a village organization

are planning and goal setting, and resource mobilization.

The signifi—

cance of these functions is illustrated in a passage from S. A. Rahim's

report on "Voluntary Group Adoption of Power Pump Irrigation in Five

East Pakistan Villagesz"

In the case of water distribution the groups, at

the initial stage, showed resistance in digging

water carrying channels over the entire crop

area.

In early 1960, when Tongipar'was visitedy

we found no arrangements for a water channel.

Water discharged from the mouth of the pipe was

simply overflowing the nearest plots.

Even

hours later the water did not reach the distant

plots.

In spite of repeated insistence by the

(pump) driver, no suitable action was taken.

Even our advice could not convince the group.

The problem was to determine who would take the

trouble of performing the manual work.

Soon,

however, the members of the group found that

without channels the pumps would be useless.

The group discussed the problem and.arrived at a

decision.

It was decided that the major channels

carrying water to blocks of plot would be dug

jointly.

From these main waterways each indivi-

dual would dig his tributaries to carry water to

his own plots.

This arrangement worked very

satisfactorily.

Within a week a network of

channels was dug.

When we made our next visit

the whole scene had changed.21

Irrigation of small farms, the Tongipar villagers had discovered,

is inherently a c00perative activity requiring planning of the location

of the irrigation channels, the setting of goals such as the number of

acres to be irrigated, and the mobilization of labor to dig and repair

250

the channels.

Technology, in other words. promoted cooperation and

organization.

The underlying reason why these activities had to be accomplished

in an organized group, rather than by individuals. was that the size of

the farmers' holdings was small, and that each farmer's holdings were

fragmented in turn into little plots each of which was often no bigger

than a backyard garden would be in the United States.

"Because of the

small fields and soattered land holdings," Levern Faidley remarked,

"the technically simple task of building a small water canal across a

quarter mile of level alluvial plain became a major organizational

problem.

Even a wealthy man could not afford to purchase the easements

and rights—of-way needed to irrigate all of his plots.

Therefore, the

only solution was for all of the owners to come together and.agree that

canals could be built across their lands."22

More was involved in growing a dry season, or 3939, crop of rice

or potatoes than delivering water to the otherwise sun-baked earth.

For the high yielding rice varieties from the International Rice

Research Institute which were introduced in Comilla in 1965. and which

were grown by 98% of Comilla farmers by 1970, successful agriculture

in the dry season required increased inputs of labor, fertilizer, and

insecticides, as well as new methods of sowing and threshing.

(The

traditional threshing method was to have bullocks tread over the hare

vested rice and straw.

This did not work well for the high yielding

rice varieties, so a pedal driven thresher with a rotating drum was

introduced as an alternative.)

Assuring the availability of these

supplies at the right time and in the right quantities was an exercise

in the integration of services. listed by Uphoff and Esman as function

d in their taxonomy of organizational activities for the promotion of

rural development.

251

Once the profitability of the new agricultural technology centering

on high yielding rice varieties and dry season irrigation had been

demonstrated, competition for the requisite inputs commenced.

Some of

these, such as fertilizer and.insecticides. were available on the market

as well as from the public sector, so the laws of supply and demand

governed much of their distribution.

The central cooperative associa-

tion, however, had a virtual monopoly on the distribution of means of

supplying irrigation water.

There were three of these:

a gravity flow

irrigation canal leading from the Gumti river to villages in the

Sonaichuri basin. low lift pumps which drew water from canals and

rivers and from village ponds (called tanks in the local.parlance),

and deep tubewells which drew up water from an aquifer about 250 feet

below the surface.

There was no prohibition on the private sale of

pumps and tubewells, which sold briskly in West Pakistan during the

same era.

No farmer in Comilla Kotwali Thana, though, held enough

contiguous acreage to justify the cost of buying his own irrigation well

or pump.

The local cooperative societies assumed the function, there-

fore, of lobbying the thana central association for the installation of

tubewells, for the rental of low lift pumps. and for the provision of

increased.agricu1tural credit to finance the more expensive, and more

profitable, modern agricultural inputs.

(Some villages which already

had one tubewell, such as Sreemontapur, lobbied for a second, while

other villages sought to emulate their prospering neighbors.)

This

claim-making function which the local cooperatives assumed is the last

of the six functional requisites of rural development mentioned in

Uphoff and Esman's taxonomy.

When the faculty of the Comilla Academy first broached the idea of

252

dry season irrigation in 1959 and 1960, many of the local farmers

laughed.

It seemed absurd to suggest that a seasonal shortage of water

was a problem in a land which was drenched during the monsoon by a

hundred inches of rainfall, and which was inundated annually to depths

of as much as twenty feet of floodwater.

But the villagers' scepticism

was overcome. and the irrigation of farmland in Comilla rose in a

steadily ascending curve to over 15,000 acres in 1970-71.

The introduction of another mechanized agricultural technology,

tractor cultivation, was more problematic.

One of the Peace Corps

volunteers who helped set up the Camilla tractor station recounted that

a farmer had told him, in a tone of disgust, that he would not try to

use a tractor to harrow his fields again.

What was the matter, the

young American inquired.

The farmer spat red beetle nut juice and

replied that the tractor had slipped into a ditch.

"When my bullock

falls down he gets up again, but that useless tractor just lies there

and doesn't stand up."

In his doctoral dissertation in agricultural engineering at Michigan

State University, Lavern Faidley tabulated the slight economic advan-

tages accruing to the farmers who employed a combination of tractor

and bullock power to prepare their land for planting a dry season crop

of high yielding rice.

Using bullock power alone, the farmer would

incur total tillage costs averaging 54.54 rupees per acre, while with a

combination of bullock and tractor power the farmer would spend an

average of 48.62 rupees per acre.

(Even when the land had been harrowed

by tractor, bullocks were still needed for the final field leveling

and puddling which preceeded planting.)

The cost differential of about

six rupees or one dollar per acre contrasted markedly with the net

return from irrigation of high yielding rice of as much as 1,200 rupees

253

per acre.

In proportion to total potential profit, the cost saving

from using tractors was insignificant.

As might be expected, the number

of acres tilled by tractor in Comilla Kotwali Thana did not increase

as rapidly as the number of acres irrigated.

By the late sixties. a

plateau of about 6,000 acres of tractor tillage in the thana was

reached?3

As was the case with irrigation, villagers who wished to avail

themselves of the tractor technology were compelled to cooperate.

To

Open up a block of land sufficiently large for’a 35 horsepower tractor

to cultivate efficiently. the owners of adjoining plots had to agree to

join in the tractor rental. and to demarcate their plot boundaries

(usually with stakes) so that ownership could be reestablished after

the tractor had tilled straight across the land.

In some instances

farmers who intended to hire a tractor from the central cooperative

association also had to synchronize their harvesting. so as to clear the

area for block cultivation prior to the planting of the next crop.

In the preceding section of this chapter. I reviewed research on

the influence of technology on the structure and management of organiza-

tions.

The Aston group of researchers, it will be recalled, asserted

that the impact of technology is more apparent in small organizations

than in large ones.

From that premise alone. it might be expected

that technology would have a strong impact on the village co0peratives,

all of which were small.

But the variety of technologies with which

the village cooperatives dealt tended to cancel out any strong pattern

of influence.

Judging from what I learned during the field research project

which took me to 45 village cooperatives in Camilla Kotwali Thana in

the 1966 dry season, I would say that in the village cooperatives the

254

influence of technology upon structure was less often noticeable than

was the influence of technology upon management.

The field research team, which was led by Zaker Hussain, observed

that in the village of Bordoil paid personnel were engaged by the local

cooPerative to supervise the distribution of irrigation water from the

tubewell.

many village cooperatives instituted water management

committees. another structural innovation, for the same purpose.

In

such cOOperatives, it could be said that the technical subsystem had

influenced the structural subsystem.

Given the fragmented pattern of landholdings. the technologies of

irrigation and tractor cultivation were management-intensive.

Water

channels and cultivation blocks had to be planned, fees had to be

assessed and collected, and accounts had to be maintained.

For irriga-

tion pumps. fuel and oil had to be purchased and repairs had to be

arranged.

Discipline in adhering to the plan for the method.and

sequence of distributing water was important, lest water be wasted and

irrigated acreage decline.

All of these activities called for strong

leadership and sound management in the village c00peratives.

The new

technologies thereby enhanced the authority of the manager of the

primary c00perative, and of its managing committee.

The Technical Subgystem §£_the Thana Tier

Technology in the broad sense -— to include not only the processes

by which machines turn raw materials into useable products but also

the procedures according to which human knowledge and effort are

transformed into services -- had a strong influence, perhaps a pre-

dominant influence, on the structure, management, and costs of the

255

thana cooperative associations in Comilla.

Some of the sections of the

KTCCA were technologically-based in the narrow, mechanical and commonly

accepted sense of technology.

The cold storage section, for example,

operated chilled warehouses where potatoes were held for sale at

optimum off-season prices.

The creamery section operated a milk

pasteurizing and butter and cheese manufacturing plant.

The machine

station ran the fleet of tractors and the tubewells.

On the other

hand, the main business of the KTCCA's two major subsidiary cooperative

federations -- the Agricultural Cooperatives Federation and the Special

CooPerative Societies Federation -- was to provide a service:

coopera-

tive credit.

If the procedures according to which that service was

provided can be considered a technology, as the broad definition of

technology assumes they should, then technology can be said to have

influenced those federations as well.

The impact of technology upon units such as the machine station

was easily discerned.

The cost of tractor repairs in Camilla was 100%

greater than in the United States.

Tractors broke down frequently in

the humid environment and flooded fields of'East Pakistan. which caused

deterioration of rubber and of insulation around electrical wiring.

Wear on the tractors was excessive, because fuels and lubricants were of

low standard and because tractor drivers were poorly trained.

Whereas

tractor

repairs in the U.S. are often accomplished in hours, in

Comilla days and even months were typically required to repair*a

tractor.

Spare parts for the Massey Ferguson 35 h.p. tractors used in

Comilla had to be imported, so the machine station stocked an unusually

large inventory of spares as a hedge.

Even so. broken tractors some-

times stood idle for 6 months to a year waiting for parts to be

imported.2u Because of these repair problems, the machine station had

256

to hire a large staff of mechanics. and had to allot an exceptionally

large portion of its budget to maintenance.

The c00perative creamery employed a continuous processing tech-

nology.

The equipment of the creamery. much of which was donated and

installed by the Danish government as part of their foreign aid program,

had a capacity in excess of the capability of local farmers to supply

it with milk, so the processing was in reality more intermitent than

continuous.

In the Western nations continuous processing industries

such as steel and petroleum have resorted to vertical integration --

ownership of their sources of raw materials and of their marketing

outlets -- in order to protect their core processing technologies from

the effects of fluctuations in raw materials or product markets.

The

corner gasoline station, which is linked within the same corporation all

the way back to oil fields in distant corners of the globe. is a symbol

of vertical integration.

The creamery section in Comilla reacted

much as other continuous processing industries have responded when

threatened with large oscillations in production.

To boost and sustain

the supply of raw milk. the creamery section sponsored special dairy

cooperatives in Comilla villages.

On the output side, the creamery

section personnel tended to bypass the marketing section of the KTCCA,

and to make their own marketing trips to Dacca and Chittagong to arrange

outlets for cheese and butter.

The impact of technology upon some of the other units of the thana

cooperative association was not as obvious.

I would argue that the

structure, management, and costs of the Agricultural Cooperatives

Federation were influenced by the technology of its credit operations.

But to assess the validity of that contention, the changing nature of

the credit technology must first be understood.

From the beginning, the leaders of the Comilla cooperatives believed

257

that credit should be supervised.

In the course of the first decade of

the operation of the cooperatives, they held to two successive concepts

of why and how that supervision should be effected.

Until July. 1965, the theory and practice of credit supervision in

Comilla were aligned with those of supervised credit programs in other

countries.

The American influence was notable, especially in the

requirement that every farmer applying for a loan draw up a production

plan.

Among county agents in the United States. a favorite extension

technique is to assist a farmer in drawing up such a plan, either for

his whole farm or for some segment of its operations.

The orthodox concept of supervised credit which prevailed during

the first five years in Comilla was closely linked to the idea that

agricultural extension was the key to rural development.

It was believed

that the techniques for improved agricultural production were known.

and that the problem was how to accelerate their adoption by peasant

farmers.

Supervised credit was a mildly coercive method for inducing

the adoption of modern farming techniques.

Under the orthodox supervised credit system, farmers wishing to

apply for loans would first hold a general discussion of agricultural

plans in one of the weekly meetings of their village cooperative.

Each

member would state how much land he intended to plant in the coming

crOp season, and how much credit he needed to finance each of his anti-

cipated categories of expenditure.

The manager and other members of the

village cooperative would discuss whether the loan request was justified,

and would try to persuade the member that he could finance his crop

with a smaller loan.

The member's request would be cut down if the

consensus was that he did not need so much credit for the agricultural

258

operations he had specified, or that he would be unable to repay the

loan on time.

The manager would then enter the mutually agreed-upon

total loan request on the production plan sheet of the village coopera-

tive, and the member would say how he thought he would divide up his

total loan for the finance of his different inputs and operations.

(Production planning, it might be noted. was thereby conducted in a

communal setting quite different from that of the individualistic

farmer in the United States.)

When the village production plan was complete, the manager would

take it to the inspector, who would check it over for inconsistencies

and errors of form.

If the inspector found such mistakes, he would

discuss them with the manager in the weekly class which all managers

attended at the Thana Training and Development Center.

The manager

would then either make the suggested changes himself, or take the plan

back to his village society for reconsideration.

When the inspector was satisfied that the production plan was

formally correct. he would refer it to the agricultural extension sec-

tion of the KTCCA.

There a trained agronomist would compare the figures

entered by the farmers with standard production formulae.

If the

agronomist recommended changes in items such as the planned dosages of

fertilizer'and insecticide. the inspector would consult again with the

manager who might take the plan back to his society for further cor-

rections.

Preconceptions of the foot-dragging traditional farmer'and the go-

getter modernizing agronomist might lead to the supposition.that the

agronomist, when he reviewed the production plan, would recommend a

doubling of fertilizer use or a tripling of the dosage of insecticide.

In practice, the process worked exactly the other way around.

The

259

farmers would ask for more fertilizer or insecticide than they needed,

or intended to use, in hopes of selling the surplus.

When the market

for these commodities was depressed. farmers would still typically

request extra credit so that they could divert funds to personal con-

sumption. or to pay off debts to moneylenders, or to other'purposes not

specified in the production plan.

The agronomist's task. then. was

usually to cut down credit requests to what would be needed to secure

the recommended combination of inputs.

What was the purpose of this rigamarole?

The orthodox program of

supervised credit made sense only as a supplement to agricultural

extension.

Supervised credit was supposed to act like an afterburner on

the jet engine of extension.

It was expected to give extra thrust and

speed up the movement of improved agricultural technology into the

countryside.

The advocates of orthodox supervised credit proposed that if the

farmers could actually be held to the implementation of the production

plan. rather than letting them treat it as an exercise in chicanery, they

could thereby be inveigled to try out new production techniques.

The low interest supervised credit program posed the farmer with

the following alternatives.

1.

Finance his consumption and traditional agricultural inputs by

taking a high interest. high collateral loan from the village money-

lenders: or

2.

Finance modern agricultural inputs by taking a low interest,

low collateral loan from the supervised credit agency.

If the credit had been unsupervised, the farmer might have taken

the lowbinterest, lowbcollateral government loan and used it to finance

his consumption and traditional agricultural inputs.

But supervision

260

was intended to remove this third possibility.

If he had to make a

choice between the first two alternatives, the farmer would probably

decide to take part in the supervised credit program, though all the

while he might try to subvert the program so as to achieve the third

course of action.

Several assumptions were implicit in the orthodox approach to

supervised credit.

1.

It was assumed that the profit to be had from adopting improved

practices of cultivation was not large enough, or sure enough, that

these practices would sell themselves.

Instead, the farmers had to be

coaxed into using methods which would yield a small margin of gain.

In the years prior to the Green Revolution, which were indeed the

years of the orthodox approach to supervised credit in Comilla, this

assumption may unfortunately have been valid.

In a faculty meeting at

the Pakistan Academy for Rural Development in August, 1966, Akhter

Hameed Khan related that in 1960. when the Academy formed its first

village cooperatives, the major improved practices being taught were low

lift pump irrigation. line sowing, and the use of fertilizer.

Although

the farmers did take to irrigation once the appropriate technology and

organization had been worked out, there was but a faltering adoption

of line sowing and chemical fertilization.

The acreage under line sow-

ing in 1962-63 was much less than in 1961-62.

It was apparent that the

farmers had rejected line sowing, and it seemed the reason was that the

effect of line sowing on yields was nominal.

There was no such specta-

cular reverse in the history of the adoption of fertilizer,.but it was

found that the farmers were for a long time wary of their use because

of the tendancy of local rice varieties to lodge (fall over) when too

great a dosage was applied.

261

Taking a lesson from these experiences, the leaders of the Comilla

experiment concluded that the route to agricultural modernization in

Kotwali Thana was not through the alteration of the techniques of culti-

vation of the existing §u§,and.§p9g_rice crops, but through the provi-

sion of irrigation water'and tractor services which would allow the

farmers to grow a third dry season crop of bng rice.

Because tubewells

and tractors are management-intensive inputs, the emphasis of the

c00perative program shifted from extension to organization.

2.

It was assumed that the cost of cooperative loans was consider-

ably lower than that of credit provided by local moneylenders.

In the

absence of such a cost differential, farmers would prefer to finance

their operations through familiar sources, and forgo the effort and

discipline required by the cooperatives.

What motivated the small farmer to put up with all the paperwork

and policing of supervised credit was fear of the moneylender.

The more

extortionate the moneylender's terms, the more willing were the farmers

to participate in the supervised credit program.

Shamsul Huq, the man who was in charge of the Comilla supervised

credit system for the first half decade, reported that interest rates

charged by moneylenders averaged 100 percent per year when he organized

his first cooperative in South Rampur village in 1960.

The interest

was charged in a curious fashion.

To make a loan of 100 rupees, the

moneylender would require the borrower to sell him a parcel of land

valued at 150 to 250 rupees.

The moneylender would pay the borrower

100 rupees for the land, which was turned over to the moneylender with

full legal certification of the sale.

When the debt came due, the

borrower would repurchase his land at its market value of 150 to 250

rupees.

Should the borrower be late in repaying his debt, the moneylender

could raise the price of resale.

262

From the moneylender's point of view, this system of selling and

reselling land had the advantage of absolute security for the loan

transaction.

In case of a default, there was no need to foreclose on

a mortgaged asset, since the creditor possessed the land already.

Ideologically, the villagers justified the system as conforming to the

Muslim religion, which forbids usury.

The moneylender would say that

the difference between the price he paid for the parcel of land.and the

price at which he resold it represented profit, not interest.

There is some evidence that fear of the moneylenders and the

desire to repay old debts to them was among the strongest initial moti-

vations of small farmers who joined the Gomilla cooperatives.

A study

conducted by the Academy for Rural Development in 1963 showed that the

largest category of loan funds requested (as stipulated in the produc-

tion plans). and the largest category of the actual utilization of loan

funds (as determined by a survey) was for the repayment of outstanding

debts.25

3.

The orthodox supervised credit program also assumed that

borrowers could be held accountable for the expenditure of their loan

in accordance with the production plan.

Without such accountability,

village cooperative members might use the loans for whatever purpose

they pleased, undermining the purpose of the orthodox supervised credit

program.

In the early stages of the Gomilla project, the inspectors were

asked to visit the villages to check on individual member's,loan

utilization to see that it conformed to the production plan.

This

supervisory technique proved impractical. though. because the inspectors

usually visited the villages at night to attend cooperative meetings,

263

whereas the farmers cultivated their fields by day.

In those cases in

which the production plan had called for services from the cooperative

association, such as tractor cultivation, the use of the credit could be

controlled.

In most cases. though, the cooperative association found

itself incapable of restricting loan diversion.

As might be suspected from the above discussion, the cooperative

association came to recognize that the orthodox approach to supervised

credit was unworkable.

The production plan requirement for loan appli-

cations was abandoned in 1965.

Akhter Hameed Khan put the case against

the production plan in the following terms:

The production plan was useless.

It was a

favorite device of the agricultural extension

people in the United States, who liked to have

a farm plan for each farmer that the extension

worker served.

It was also recommended by the

Ford Foundation group in their report on India's

Food Crisis egg Steps 39 Meet I3.

In our vil-

lages it didn't work.

We found that we had

nobody who could draw up a good production plan

for the farmers, because nobody knew what to

recommend.

There was so much variation in the

actual practices.

We were encouraging the

inspectors to participate in a deception in

drawing up these plans which we could not en-

force or check up on.

The purpose of the produc-

tion plan was to encourage the farmers to adopt

improved practices, but we had evidence from

our field research that they were adopting the

improved practices anyway.

So what was the

point of the production plan?2

As the rationale for the production plans crumbled, so likewise did

the justification for employing the staff required for their review.

The trained agronomist who had been in charge of production plan review

left the KTCCA in August of 1964. and was not replaced.

For a year

thereafter, the cooperative association continued the ritual of requir-

ing production plans, but with no one to check the plans the procedure

264

was a farce.

In July, 1965 the production plan requirement was dropped

altogether.

'

If all controls had been lifted in 1965 the inspectors might have

seen their own jobs emperilled.

For if credit had been granted without

restriction, there would have been no need for the inspectorate.

The

central association could have granted unsupervised credit in the same

manner as it ran the cold storage facility, by direct negotiation with

the managers of the primary cooperative societies.

But in the place of the orthodox system of supervised agricultural

practices. the central association introduced a new system of supervised

management.

As the production plan had been the keystone of the ortho-

dox system, the loan limit upheld the arch of the new regime.

And

the inspectors, it turned out, played a more influential role in the

new system than in the old.

As can be seen in Table 5-1, there was a sharp break in the ratio

of cooperatives per inspector between 1966 and 1967.

The ratio dropped

from 15.8 to 9.78 in one year.

Despite a push for expansion in the

number of primary c00peratives in Kotwali thana, the inspectorate grew

disproportionately.

It would be wrong to attribute all of this growth

to the changeover from the system of supervised agricultural practices

to the system of supervised management.

That change of policy had taken

effect two years previously. in 1965, so whatever effect it had was a

delayed one.

Two other factors were at play.

1966-67 was the first

full fiscal year in which the Agricultural Cooperatives Federation

operated as an independent entity, and the doubling of the inspectorate

between June of 1966 and June of 1967 could well have been a response

to that structural change.

Also. the volume of loan transactions was

surging ahead, and the increase in staffing may have been partly

265

5
7

-
3
7

7
1
.
7
9

8
8
.
3
7

0
0
.
3
4
1

0
6
.
5
4
1

1
9
.
4
1
1

6
4
.
7
4
1

4
8
.
2
4
1

5
7
.
4
1

3
3
.
8
1

5
2
.
5
1

0
2
.
5
1

0
8
.
5
1

8
7

.
9

6
3
.
6

4
1
.
8

7
9
.
8
0
1

8
8
.
9

4

6

8

0
1

0
1

3
2

1
4

7
3

2
3

5
9
2

3
8
5

1
9
5

0
3
4
1

6
5
4
1

3
4
6
2

6
4
0
6

5
8
2
5

7
8
4
3

9
5

0
1
1

2
2
1

2
5
1

8
5
1

5
2
2

1
6
2

1
0
3

2
6
-
1
6
9
1

3
6
-
2
6
9
1

4
6
-
3
6
9
1

5
6
-
4
6
9
1

6
6
-
5
6
9
1

7
6
-
6
6
9
1

8
6
-
7
6
9
1

9
6
-
8
6
9
1

6
1
3

0
7
-
9
6
9
1

 

s
n
o
i
t
c
a
s
n
a
r
T

s
e
v
i
t
a
r
e
p
o
o
C

b

a

n
a
o
L

e
g
a
l
l
i
V

r
o
t
c
e
p
s
n
i

r
e
p

r
o
t
c
e
p
s
n
i

r
e
p

s
r
o
t
c
e
p
s
n
I

s
n
o
i
t
c
a
s
n
a
r
T

s
e
v
i
t
a
r
e
p
o
o
C

r
a
e
Y

.
r
a
e
y

r
e
p

s
e
e
p
u
r

f
o

s
d
n
a
s
u
o
h
T

.
r
a
e
y

l
a
c
s
i
f

e
h
t

g
n
i
r
u
d

t
n
e
m
y
a
p
e
r

n
a
o
l

d
n
a

e
u
s
s
i

n
a
o
l

f
o

m
u
S

a

f
o

d
n
e

e
h
t

f
o

s
a

e
r
a

s
r
o
t
c
e
p
s
n
i

d
n
a

s
e
v
i
t
a
r
e
p
o
o
c

r
o
f

a
t
a
D

.
s
r
o
t
c
e
p
s
n
i

t
n
a
t
s
i
s
s
a

s
e
d
u
l
c
n
I

b

.
r
a
e
y

l
a
c
s
i
f

e
h
t

.
s
e
s
n
o
p
s
e
r

e
r
i
a
n
n
o
i
t
s
e
u
q

d
n
a

,
A
C
C
T
K

e
h
t
'
f
o
s
u
r
o
p
e
r

l
a
u
n
n
A

:
s
e
c
r
u
o
S

l
a
r
u
t
l
u
c
i
r
g
A

e
h
t

n
i

d
n
a

)
6
6
-
2
6
9
1
(

A
C
C
T
K

e
h
t

f
o

n
o
i
t
c
e
s

n
o
i
s
i
v
r
e
p
u
s

.
)
0
7
-
7
6
9
1
(

n
o
i
t
a
r
e
d
e
F

s
e
v
i
t
a
r
e
p
o
o
C

d
l
e
i
f

e
h
t

n
i

s
r
o
t
c
e
p
s
n
i

d
n
a

,
s
n
o
i
t
c
a
s
n
a
r
t

n
a
o
l

.
s
e
v
i
t
a
r
e
p
o
o
c

f
o

h
t
w
o
r
G

:
1
-
5

e
l
b
a
T

266

attributable to the increased workload.

Yet as I pointed out above.

the greater workload might have had no effect at all on the size of the

inspectorate if the abandonment of the policy of supervised agricultural

practices had not been followed by its replacement with the policy of

supervised management.

All in all, I think it is fair to say that the

change in loan policy did have an influence on the structure of the

Agricultural Cooperatives Federation.

The new loan policy provided

a justification for the perpetuation, and indeed for the continued

growth, of the inspectorate.

The technical subsystem, in sum, influenced

the structural subsystem.

The fundamental concern of the system of supervised agricultural

practices was loan demand.

The objective was to influence the pattern

of loan demand so as to promote agricultural modernization.

By contrast, the system of supervised management was concerned with

the pattern of loan supply.

The basic question was how much should be

loaned to each primary cooperative.

Since the supervised management

system was not built around an attempted control of credit demand, the

diversion of loans to unauthorized purposes ceased to be a major concern

of the thana c00perative association.

The categories of purposes for

which loans were issued remained fairly stable, as can be seen from

Table 5-2, but the inspectors of the central association were no longer

charged with checking to assure that every rupee was spent for the

purposes authorized.

Some controls were retained:

if a loan to a

village cooperative was for the purpose of financing the spring season,

or'gug. rice crop, then enough of the authorized loan to pay for seed

would be released first, an amount sufficient to pay for fertilizer

would be released later, and so forth.

But the major responsibility for

assuring that loan funds were spent properly was shifted from the thana

267

Table 5-2:

Distribution of Agricultural Cooperatives Federation LOans

by Purpose (percent of annual loan issue per purpose).

1964-65

1965-66

1966-67

1967-68

1968-69

Crop financing

Aus rice crop (spring)

20

32

33

Anon rice crop (fall)

Boro rice crop (winter)

Sugarcane

Potato

Vegetable

Sub total

Land

 

Release from moneylender

Lease

Purchase

Sub total

Aggiculturel inputsa

 

(Irrigation, fertilizer,

insecticides, fuel and

lubricants)

Animal husbgpq

 

(Milch cow, poultry

pisiculture, bullock)

7

7

11

11

__1_

57

11

5

_;’

16

6

3

11

9

__1_

62

7

3

‘_:

10

4

5

9

4

__.:

55

9

6

7_:

15

19

6

10

3

4

__:

52

17

17

‘_1

35

7

5

5

4

14

2

5

2

17

13

16

-

4

.2:

60

8

19

{_2

29

2

1

2

Personal

1

-

-

1

(Food. house repair,

house construction)

Other

5

21

20

6

6

Total percent

100

100

100

100

100

a Most agricultural inputs were paid for through crop financing

loans, rather than through special loans.

Source:

Annual reports of the cooperative experiment.

268

association to the village cooperatives. which had to collect from

individual borrowers to repay one loan from the central association

before another loan would be granted.

(In accordance with the multi-

tiered pattern of cooperative organization, individual farmers borrowed

from and repaid loans to the village cooperative. which borrowed in

turn from the thana association.)

Table 5-2 shows that in the period from 1967 to 1969, when the

Green Revolution in irrigated production of high yielding rice varieties

swept the hitherto barren dry season fields of Gomilla. loans for the

‘pqgg rice crop grew to the largest category of credit.

The high yield-

ing rice was more profitable than sugarcane or vegetables, and so

replaced them as a crop.

The increased profitability of farming is also

reflected in the jump in loan issue for the lease, and even for the

purchase. of land.

The change in the purpOSes for which loans were taken coincided in

time with a sharp increase in loan volume, and in the size of the

organizational base in the villages.

But this growth was not synchronized.

The rise in loan issue outstripped the increase in membership and in

cooperative societies, as can be seen in Table 5-3.

Through the decade

of the sixties, the average loan issue per member per year was 180

rupees.

Twice as much money, 370 rupees per member, was loaned out in

1967-68.

Because the average number of members per village cooperative

swelled as well in 1967-68, the spike in loan issue per c00perative in

that year was even more pronounced.

The intent behind this big push was to spread the<3reen Revolution

in rice cultivation as quickly as possible through the experimental

thana in Gomilla, so that it could become a model of technically

progressive farming.

That goal was largely achieved.

But as I have

269

remarked on several occasions in this book, some of the organizational

side effects of the big push were deleterious.

Table 5-3:

Growth in membership, cooperatives. and loan issue in

Kotwali thana.

Year

Members

Cooperatives

Issue*

Member

Cooperative

Loan

Loans per

Loans per

(a)

(b)

(c)

(c/ax1,000)

(c/bx1,000)

1961-62

1860

59

235

130

3980

1962-63

3156

110

366

120

3330

1963-64

3833

122

304

80

2490

1964-65

4910

152

878

180

5780

1965-66

5161

1966-67

8462

158

225

796

1671

150

200

5040

7430

1967-68

11518

261

4244

370

16260

1968-69

11673

301

2930

250

9730

1969-70

11151

316

1713

150

6420

 

Av. 61-60

6858

189

1460

180

6610

 

*Loan issue in thousands of rupees.

Source:

Annual reports of the KTCGA.

The Agricultural Cooperatives Federation was formed in 1966, a

year before the big push was launched.

My own field research into the

Gomilla COOperative system was conducted in 1965-66, during the period

when the ACF was created.

I had no inkling then of the impending drive

for a massive increase in loan issue.

However, my interviews with

officials of the thana cooperative organizations, and with farmers in

the 45 villages visited by the field research team,

27

revealed some of

the financial stresses with which the Agricultural Cooperatives

Federation would have to contend.

270

Separation from its parent organization, the KTCCA, impelled the

Agricultural Cooperatives Federation toward financial self-sufficiency.

Loanable funds could be obtained more cheaply from the village member-

ship -- in the form of savings and share purchases -- than they could

be borrowed from the KTCCA.

The Kotwali Thana Central Co0perative

Association assumed the role of a financing bank for the ACF, and

charged interest accordingly.

The effective interest rates on savings

and the effective dividend rates on shares were lower than the interest

rate on borrowed funds charged by the KTCCA, so the financial self-

interest of the ACF drove this new association to rely more heavily on

equity capital raised in the villages.

In this instance. it appeared

that a change in the structural subsystem of the thana cooperative

association induced a change in the technical subsystem.

When roll was called in the weekly nighttime meeting of the primary

cooperatives, each member was asked to make a savings deposit.

The

amounts deposited were usually less than a rupee per member per week.

but spread over thousands of cooperative members in the thana the

savings deposits could accumulate to substantial amounts.

The manager,

when he went to attend his weekly training class at the Thana Training

and Development Center, deposited the collected cash in his society's

account at the United Bank Limited, a private bank with a branch office

on the campus of the thana center.

(By engaging the services of the

United Bank as an intermediary in all financial transactions, the

central association kept actual cash out of the hands of all of its

officers, including the inspectors, and thereby reduced the potential

for corruption.)

Interest payments on the savings deposits of the primary societies

271

were calculated annually at the compounded rate of 3.5%.

In fact, those

interest payments did not reach the individual cooperative members

until 1966, because the managers and village accountants had not bothered

to enter interest earnings in the members' passbooks.

Only in August of

1966, one month after the Agricultural Cooperatives Federation was

registered as a cooperative organization distinct from the KTCCA, did

Shamsul Huq, the.ACF director, put enough pressure on the village

accountants to get them to calculate and enter the individual interest

payments.

Shamsul Huq's action was an excellent indication of the long-

standing need for autonomous management of the credit program.

In order

to raise equity capital, the new management of the ACF began to hustle

for savings.

According to the bylaws promulgated for each primary society, the

members were required to buy one share worth 10 rupees in their village

cooperative each year.

Once a year, the primary cooperatives were

granted shares in the central cooperative association in proportion to

the number of shares individual members had purchased in each primary

society.

Ehst Pakistan cooperative law required that share purchases

cover at least one tenth of the value of loans.

In the Gomilla coopera-

tives. this coverage requirement was raised, so that one fifth of the

value of loans had to be matched by the combined total of share pur-

chases and savings.

The cooperative association sometimes raised this

coverage requirement to as high as 33% for cooperatives whose management

was suspect or whose loan repayment record was poor.

As with savings, the cooperative association was dilatory in

according monetary rewards to its shareholders.

The first cash dividend

- on shares was declared only in 1965-66.

Shares. moreover. were not

saleable, so once money had been invested in them it could not be

withdrawn.

272

While savings deposits could be withdrawn, the procedures for doing

so were strewn with obstacles.

A cooperative member desiring to make a

savings withdrawal had to justify his plan to his fellow cooperative

members, and have his request incorporated in a written resolution of

the society.

The resolution would be approved by the society's inspector

only if the requested withdrawal amount was not pledged as coverage for

an outstanding loan.

The field research team in which the author participated in 1966

found that cooPerative savings had acquired no appeal as a way of

investing surplus funds in the villages.

Farmer after farmer, when

asked what he would do with any extra money that might come to hand,

said that he would first of all buy land, and secondly would stock paddy

to resell it at higher prices.

The farmers were aware of the high profit

investment opportunities in the village economy, and were reluctant to

put their money into no-interest blockable savings accounts, or into

the purchase of shares which paid no dividends.

As can be seen from Table 5-4, the Agricultural Cooperatives Federa-

tion Scored some success in raising equity capital from the members of

the village cooperatives.

The rate of capital formation per member

tripled between 1966-67 and 1967-68.

Despite the explanation of the newly-discovered financial.incen-

tives in the ACF, the data in Figure 5-4 is puzzling.

The very year in

which the rate of capital formation per member tripled, 1967-68, was the

year when the Green Revolution swept the thana.

Despite the fact that

the Agricultural Cooperatives Federation was finally offering interest

on savings and dividends on shares. the new rice growing technology had

opened up investment opportunities in the villages which were much more

273

-
r
o
f

l
a
t
i
p
a
C

-
r
o
f

l
a
t
i
p
a
C

—
r
u
p

e
r
a
h
S

s
g
n
i
v
a
S

a

d
r
e
b
m
e
m
/
n
o
i
t
a
m

c
e
t
a
r

n
o
i
t
a
m

b
e
t
a
r

e
s
a
h
c

b
e
t
a
R

s
r
e
b
m
e
M

r
a
e
Y

 

.
a
n
a
h
t

i
l
a
w
t
o
K

n
i

s
e
v
i
t
a
r
e
p
o
o
c

l
a
r
u
t
l
u
c
i
r
g
a

e
h
t

n
i

n
o
i
t
a
m
r
o
f

l
a
t
i
p
a
C

:
#
-
5

e
l
b
a
T

8
5
.
2
2

6
6
.
5
2

0
3
.
5
2

1
3
.
7
1

0
5
.
1
2

5
9
.
5
1

2
9
.
7
#

3
6
.
5
3

4
2
.
9
2

2
0

1
8

7
9

5
8

1
1
1

5
3
1

2
5
5

6
1
#

6
2
3

a
1

7
2

8
2

6
6

9
7

0
5

5
8
2

9
3
1

2
8
1

8
2

4
5

9
6

9
1

2
3

5
8

0
6
8
1

2
6
-
1
6
9
1

6
5
1
3

3
6
-
2
6
9
1

3
3
8
3

4
6
-
3
6
9
1

0
1
9
#

5
6
-
0
6
9
1

1
6
1
5

6
6
-
5
6
9
1

2
6
0
8

7
6
-
6
6
9
1

7
6
2

8
1
5
1
1

8
6
—
7
6
9
1

7
7
2

3
7
6
1
1

9
6
-
8
6
9
1

4
0
1

1
5
1
1
1

0
7
-
9
6
9
1

.
r
a
e
y

r
e
p

s
e
e
p
u
r

f
o

s
d
n
a
s
u
o
h
T

.
e
t
a
r

e
s
a
h
c
r
u
p

e
r
a
h
s

d
n
a

e
t
a
r

s
g
n
i
v
a
s

f
o

m
u
S

0

.
r
a
e
y

l
a
c
s
i
f

e
h
t

g
n
i
r
u
d

s
e
r
a
h
s

f
o

s
g
n
i
v
a
s

l
a
t
o
t

n
i

e
g
n
a
h
c

t
e
N

b

.
r
a
e
y

l
a
c
s
i
f

e
h
t

f
o

d
n
e

e
h
t

f
o

s
A

a

.
s
r
e
b
m
e
m

f

0
0
0
,
1

x

e
t
a
r

n
o
i
t
a
m
r
o
f

l
a
t
i
p
a
C

d

.
A
C
C
T
K

e
h
t

f
o

s
t
r
o
p
e
r

l
a
u
n
n
A

:
e
c
r
u
o
S

27#

lucrative than c00perative savings accounts yielding 3.5% annual

interest.

Why, then. was there such a dramatic spurt in capital forma-

tion by the membership in 1967-68?

To appreciate the cause of this

phenomenon, one must understand the policies for loan issue which pre-

vailed under the system of supervised management, which had replaced the

system of supervised agricultural practices.

In order to obtain loans under the supervised management system.

cooperative members would first have to participate in the determination

of loan limits for each cooperative member individually and for their

village cooperative collectively.

Every six months the members met to

discuss their loan needs for the coming half year.

Each member would

describe the cr0ps he was going to grow. estimate his credit needs. and

request a loan limit up to which he would be allowed to borrow in the

next six months.

If the other c00perative members thought that the

farmer could not repay the requested amount, because he was poor,

incompetent as a farmer, or had outstanding debts to the cooperative,

they would reduce the requested limit.

The manager would reduce the

limit still further if he found that it exceeded five times the sum of

the member's savings and shares.

When loan limits had been agreed upon

for'all members, the manager submitted them to the cooperative inspector.

If the inspector discovered any mistakes in the individual loan

limit requests. as of a member whose loan limit exceeded five times

the sum of shares and savings. he would correct the irregularities.

The

inspector then prepared a loan ceiling proposal for each of the

societies under his supervision.

In his proposal were mentioned

relevant facts about each society. including the rate of attendance of

the members at weekly meetings, the rate of loan repayment, the total

savings and shares pledged for the next six months as security against

275

loans,

and the loan limit

which

was approved

for the previous

six

months.

The draft

loan limit

had to clear

two more

hurdles

before

it was

officially

published.

It had to be reviewed

first

by a loan ceiling

committee

consisting

of the ACF management

and inspectors,

and then by

a loan committee

comprised

of ACF management

and elected

members

of the

ACF managing

committee.

.After

passing

all these

stages

of review,

the

six.month

loan ceilings

for'all

primary

cooperatives

were published

as

a loan gazette.

Akhter

Hameed

Khan,

who sat on the KTCCA

loan committee

before

the

credit

operations

were transferred

to the ACF, recalled

that the

committee

would

"go through

five or more societies

in a day to set the

loan ceilings.

Those

were open meetings.

You could

tell whether'a

society

was active

or not by seeing

how many of the members

came.

The

manager

and the society's

managing

committee

were supposed

to attend,

but in many

cases

they didn't."28

At each stage

of review,

the proposed

loan ceiling

for each society

could

be raised

or lowered.

In practice,

the society

ceilings

were

usually

reduced.

The manager’and

the inspector

were likely

to shave

the

limit

so that it would

accord

with the regulation

that total

loans

to a cooperative

society

must not exceed

five times

the sum of pledged

savings

plus shares.

At later

stages

of review,

the ACF management

was

likely

to cut the limit

if they thought

the society

was a poor

loan

risk.

The percentages

of attendance

and of loan repayment

indicated

to

them whether

the cooperative

manager

was able to discipline.his

members

and meet his society's

obligations.

To these

indications

of the quality

of a society's

management,

the members

of the loan committees

would

add

their own considerable knowledge of the society's history and current

situation.

276

As a guideline for setting loan limits, the ACF management

categorized the primary cooperatives into classes A, B, and C.

An "A"

class society could borrow up to five times its total of shares and

savings, a "B" class society could borrow four times this total, and a

"C" class society only three times.

The quality of a society's manage-

ment, as assessed by the indicators described above, was the primary

determinant of its classification.29

‘

By controlling the supply of agricultural credit through the loan

limit policy, the Agricultural Cooperatives Federation created a strong

incentive for the deposit of savings and the purchase of shares by the

cooperative members.

This incentive was much stronger and more

effective, I contend, than the attractions of interest from savings

accounts or dividends from shares.

When the cooperative members

deposited more funds, they could borrow more money.

‘A Systems Interpretation

The explicit relationship which the loan limit policy enforced

between loans and the sum of savings and share purchases suggests that

that relationship may be susceptible to mathematical formulation.

Since

the loan limit was based on the sum of savings deposits and share

purchases, that sum can be treated as one variable in the model.

Let

us call it C, for capital.

C represents a level, not a rate.

A second

major variable in the model would be OL, outstanding loans, which would

also be a level variable.

The loan limit policy was based on the

relationship between C and CL.

For an A class society 0L could grow

to 5G, whereas for a C class society 0L was not allowed to exceed 30.

277

Since they carried a lower interest rate than loans from the money-

lenders, cooperative loans were always in high demand.

After the advent

of the Green Revolution, low interest c00perative loans to finance

agricultural inputs were sought more vigorously than ever.

To obtain

more loans through their local c00perative, members could:

a) deposit

more savings and buy more shares, thereby raising their cooperative

society's loan limit, b) repay outstanding loans, thereby lowering their

society's total of loans outstanding below the society's loan limit, or

c) improve their society's record in attendance, repayment of overdue

loans and overall management, thereby obtaining a better credit rating.

(Note that all three of these options have to do with a village coopera-

tive society, suggesting that the appropriate level of analysis is not

the aggregate relationship between 0L and C in the thana as a whole, but

the relationship between 0L and C in an average village cooperative.

To bring the variables 0L and G into a scale appropriate to a single

local 000perative, we will measure these variables in rupees, rather

than in thousands of rupees.)

Let us concentrate on the first of the three behavioral options

mentioned above, and construct a model of the change in C, savings

deposits plus share purchases.

If the mixture of A, B and C class

societies were fairly constant,30 it would be sensible to discuss an

average village cooperative for which the loan limit was a fixed

multiple of C.

We will call this factor of multiplication f3.

Suppose,

now, that 0L was greater than

30.

Loans outstanding would then be

in excess of the loan limit, and the village cooperative would be cut

off from further borrowing.

What to do?

Deposit savings and buy

shares.

If, on the other hand, 0L was less than

BC, the members of

the cooperative could withdraw some of their savings.

(The cooperative

278

members could also, in this latter situation, take out more loans.

But

lthe present discussion focuses on the change in C, not in OL.

In

building mathematical models, it is standard practice to look at the

determinants of change in one variable at a time, pretending that all

other variables in the system are constant while each differential

or difference equation is being built.)

Restating these observations mathematically, we can postulate that

if the quantity (0L -E3C) were greater than zero then C would rise,

whereas if (0L -E3C) were less than zero C would fall.

When C rises

the change in capital, A C, is positive, and when G falls A C is nega-

tive.

So we have the model equation

AC = 01(OLn - eon)

(5.1)

in which the parameter a.indicates the responsiveness of A,C to the

discrepancy (0L - 8 C).

While the model summarized in equation 5.1 is a good beginning, it

is not wholly realistic.

Every village cooperative was required to

purchase a certain minimum number of shares in its central association

each year, and at the next level down every cooperative member was

required to purchase an annual minimum number of shares in his local

cooperative.

Also, some savings may have been deposited and some

shares may have been purchased for reasons other than the desire to

deposit more money.

Peer pressure to deposit money, loyalty to the

local cooperative and its needs, and even the attraction of collecting

three and a half percent annual interest could all have influenced

villagers to buy shares or deposit savings.

Such additional factors

can be incorporated in a constant term, R, which can be appended to the

model equation.

The complete equation for the generation of capital

in the local cooperatives thus becomes

279

A0 =

c(OLn - B On) + B.

(5-2)

or’equivalently,

Cn+1 =

aOLn + (1 - a8 ) Cu + R'

(503)

From the preceeding discussion we can surmise that the value of

would be 5 or less.

The assumptions of the model, according to which

C adjusts to a discrepancy between 0L and a multiple of C (not a

fraction of C), would seem to be violated if

B were less than unity.

So we may hypothesize that

8 would be greater than one.

Since R

represents additional capital formation not ascribable to the feedback

adjustment between 0L and C, we would expect R to be positive.

What

about a , the responsiveness parameter?

The theory of capital genera-

tion outlined above WOuld lead us to expect that a would be positive.

Beyond that, there is not much that can be said with confidence about

the expected value of'

a-

It might be surprising if ,3 were greater

than unity, since this would indicate that the villagers compensated

in less than one year fer the discrepancy (0Ln -

Cn), but if the

villagers were very eager to obtain more loans they might deposit

savings and buy Shares at a very rapid rate.31

Equation 5.2 was used as an estimating equation in a regression

analysis of financial data from Kotwali thana and the 1966 projects.32

Since the model summarized in equation 5.2 concerns the behavior of an

average local cooperative, the ideal data set against which to test the

model would have been collected from the hundreds of local cooperatives

280

in each of the experimental thanas in the nineteen sixties.

Such

cooperative by cooperative data was unavailable.

As a substitute, the

aggregate financial data for each thana and each year was divided by

the number of local cooperatives then extant in the thana, to yield

yearly averages of loans outstanding per cooperative, of savings plus

shares per cooperative, and of the change in capital per cooperative.33

Because the variable Cn enters into both sides of equation 5.2,

directly on the right hand side and indirectly on the left in

AC = C

- On, the coefficient of determination, r2, in regressions

n+1

based on equation 5.2 tends to be low.

The F ratios and their signifi-

cance also appear poor.

A more reliable measure of the accuracy of the

regressions can be obtained by testing 5.3, which is.algebraically

equivalent to 5.2.

(To obtain 5.3 from 5.2, simply add Cn to both

sides.)

The values of r2, F, and the significance of F in Table 5-5

were obtained, accordingly, from regressions of the form of equation 5.3;

it was reassuring to discover that in these regressions the values of

a and R were identical to those based on the A C regressions, while

the coefficient of C differed by exactly 1.

The values of 53 in

Table 5-5 were obtained by dividing the regression coefficient for 0L

into the regression coefficient for C.

Table 5-5.

Regression estimates for the capital generation equation

in Kotwali thana and in the 1966 projects.

AC: a(OLn—gcn)+R

N

r2

Overall

Significance

F

ofF

 

Kotwali thana

A 0 = .16 (0Ln - 2.91 on) + 1187

8

.90

21.79

'

.003

1266 projects

A c = .27 (0Ln - 4.17 on) + 2245

 

28

.60

39.15

.000

 

281

In Kotwali thana and in the 1966 projects the coefficients of the

capital generation equation came out as expected.

The values of

3

were in the range 1 <'

5 S 5, and the coefficients R and.a

were posi-

tive.

The estimated values of‘a , moreover, seem intuitively reasonable,

since they indicate that the villagers adjusted gradually to the dis-

crepancy (OLn - 8 On).

I was surprised to find that even with as few as four cases (data

for 1966, 1967, 1968 and 1969) in one thana, the capital generation

equation still applied, with most parameter values in the range expected.

Only in Chandina thana were the results of a stepwise regression based

on equation 5.2 inconsistent with the stepwise regression results based

on equation 5.3 -- in that case only loans outstanding per cooperative

entered one equation, and only total capital (savings and shares) per

cooperative entered the other regression equation.

In the other six

thanas the (AC regression results corresponded to the Cn+1 regression

results.

Three parameter values were larger than anticipated -- the (3

parameter in Laksam and Sarail and the (1 parameter in Brahman Bari

--

but otherwise the parameter estimates conformed to expectations.

Of the two state variables, 0L and C, in our elementary model of

cooperative finance, we have so far accounted for the change in one of

them.

What is needed next is a method for modeling the change in 0L,

total loans outstanding per cooperative.

It may occur to the reader

that the same sort of cybernetic feedback equation which was used to

estimate A 0 could be used to estimate A OL.

Such an equation would

have the form

AOL: e (eCn-OLn)+R,

(5.1+)

in which 9 would be a sensitivity parameter comparable to a in

282

Table 5-6.

Regression estimates for the capital generation equation in

seven thanas.

_

_

AC—d(OLn BCn)+R

N

2

r

Overall

F

Significance

ofF

 

Laksam

Ac = .22 (0Ln - 7.62 on) + 3737

4

.87

3.37

.360

Chandina

5.2 and 5.3 inconsistent

Sarail

AC = .4 (0Ln - 5.07 on) + 3613

4

.25

0.17

.866

Hajigailj

Ac = .84 (0Ln - 4.11 on) + 3156

4

.95

8.82

.232

Quasba

AC = .90 (0Ln - 2.40 on) + 1255

4

.65

.92

.594

 

50.53

.099

Brahman Baria

AC = 3.27 (0Ln - 2.74 on) + 2416

  

AC = .77 (0Ln - 3.68 on) + 3918

4

4

.99

Chandpur

.97

18.85

.161

283

equation 5.3, and i: would be a factor of multiplication determined by

the loan limit policy, as was 8 in 5.3.

There are several reasons,

though, why this is not a suitable mathematical model of the change in

OL.

First of all, local capital generation really was determined within

the village cooperatives, which is where the savings deposits and share

purchases originated.

A village-level model for 110 was therefore

appropriate.

On the other hand, even though loan demand originated in

the villages, loan funds were dispensed by the cooperative associations

at the thana tier.

Since decisions on loan issue were made by officials

of the thana cooPerative associations. a thana-level mathematical model

would be more appropriate for loans.

Secondly, there was a fundamental difference between the change in

local capital, 15C, and the change in loans outstanding;

Because sav-

ings could be either deposited or withdrawn, the level of local capital

could go up or down.

(In most thanas in most years C went up, but in

a few cases, as in Gaibandha thana in 1966, it declined.)

Positive and

negative changes in C were explicable in the same terms, and could be

modelled as one process.

The positive and negative components of the change in outstanding

loans were different from one another, and therefore cannot be modelled

as one process.

If we designate TVLOAN as the amount of loans out-

standing at the thana tier (measured in thousands of rupees), then

TVLCAN will go up when loans are issued to village cooperatives, and

TVLOAN will go down when loans are repaid by village cooperatives.

Let

us call the rate at which loans are issued to village 000peratives VLI,

and the rate at which village cooperatives repay their loans VLR; both

of these rate variables would be measured in thousands of rupees per

284

year.

There is a lag between VLI and VLR, so over time the positive

influence on TVLOAN, which is VLI, determines the negative influence on

TVLOAN, which is VLR.

Such is not the case with local capital genera-

tion.

Savings deposits do not determine savings withdrawals.

What

is needed, therefore, is a pair of models:

one for VLI and one for VLR.

First the model of loan issue.

Suppose for the sake of discussion that all loans issued to village

cooperatives were repaid in exactly one year.

And suppose further that

the officers of the thana cooperative association always issued loans

right up to the maximum loan limit of five times the total of savings

and shares.

Although that loan limitation was applied to particular

000peratives, its influence across all the local cooperatives in the

thana should still be evident in the aggregate data.

Let us designate

the thana-wide sum of savings and shares as TSAV.

Then if the thana

cooperative officials behaved as described, the ratio of TVLOAN to

TSAV would always be 5.

The ratio of loan issue to total capital

(VLI/TSAV) would also be 5.

If the delay in repaying village loans

increased to more than a year, the ratio TVLOAN/TSAV would remain at 5

but the ratio VLI/TSAV would decrease.

And conversely, if the delay in

repayment were shortened to less than a year, loans equal to more than

5 times total savings and shares could be issued during the year with—

out violating the loan limit defined by the ratio TVLOAN/TSAV = 5.

Reasonable as the preceding argument may appear, it contains one

flaw.

In actuality all loan requests by the village c00peratives were

not granted, even when they fell within the loan limit guidelines.

Also, there were some cases of village cooperatives which did not

request the loans to which they were entitled by virtue of their

accumulated savings and shares.

There were, consequently, some savings

285

and shares which were not balanced either against loans outstanding or

against loan issue.

Instead of a simple ratio, therefore, a more

complete model of the relationship cf VLI to TSAV would have the form

VLI =

g' TSAV - R'.

(5.5)

In this equation

8' is a parameter with an expected value in the

range 1 <

B' S. 5, similar to the parameter

8 in the capital generation

equation.

R' is a constant which represents village capital not used

as coverage for’loans.

In a test of equation 5.5 against the financial data from the

thana cooperative associations, the model held up poorly for Kotwali

thana, and excellently for the seven expansion projects founded in

1965-66.

For'Kotwali thana, the estimating equation was

VLI = 1.25 TSAV + 879.14 (r2 = .27, F = 1.88, significance of F = .229,

N = 7).

Although the

3' parameter was in the anticipated range between

1 and 5, the sign of the constant was contrary to predictions.

A look

at the raw data on loan issue in Kotwali thana (see Table 5-3) reveals

that there were fluctuations in loan issue from year to year.

These

irregularities, which were unmatched by fluctuations in total savings

and shares, probably are the cause of the poor fit of the loan issue

model in Kotwali thana.

For the 1966 projects, the estimating equation

for VLI ordained through regression analysis was VLI = 4.74 TSAV -

174.65.

The coefficient of determination for the equation, r2, was

.95.

The overall F ratio for the equation was 210.36, which was

significant at the .000 level.3h

The last component of our basic model of cooperative finance

which needs to be put in place is a means of estimating VLR. the rate

of repayment of village loans.

This is, as it happens, a classic

286

problem in dynamic systems analysis, and it can be solved by standard

methods.

The solution methods employed here«are those described by

Jay W. Forrester, of the Massachusetts Institute of Technology, in his

book Industrial Dynamics.

I am endebted to Douglas Franco, a doctoral

candidate in systems science at Michigan State University, for pointing

out to me the relevance to the cooperative credit problem of Forrester's

approach to the mathematical representation of delay processes.

There are two basic approaches to take in the mathematical model-

ing of delays. depending upon whether the delay is being described in

discrete time or in continuous time.

For reasons which will become

apparent, I chose to model the financial operations of the thana coopera-

tive associations in continuous time.

Although the sub-model of capital

generation embodied in equation 5.2 is founded in discrete time, that

equation can be enfolded, as I shall show, in a larger continuous time

model.

An estimation technique appropriate to continuous time was

employed in deriving the coefficients of the loan issue equation, 5.5.

Loan issue during the full course of a year was regressed against an

estimate of total savings and shares in the middle of the year.

The

implication of this procedure is that the prediction of VLI can be

continuously adjusted during the year as TSAV changes.

If the model

of VLI had been constructed in discrete time, the appropriate regression

would have been of loan issue during the year'against local capital at

the beginning of the year, a procedure which would have allowed for'a

single annual readjustment in the prediction for loan issue at the

start of each year.

The loan delay model is most easily explained by reference to the

diagram in Figure 5-1.

There are two stages to the basic delay, plus

an additional stage for the repayment of overdue loans.35

Storage

28?

I IL

Y
A
P
E
R

 

.
y
a
l
e
d

t
n
e
m
y
a
p
e
r

n
a
o
l

e
h
t

f
o

m
a
r
g
a
i
d

k
c
o
l
B

.
1
-
5

e
r
u
g
i
F

.
.
.
f

   i

 

 

 

k

 

)
2
.
3
5
3
1
(

 

)
2
/
I
L
E
D
(

S
T
B
E
D
B

Z
Y
A
P
E
R

E
U
D
R
V
O

Z
T
N
I
R

2
E
R
O
T
S

_
l
T
N
I
R

l
E
R
O
T
S

R
L
P

m

2

e
g
a
t
S

1

e
g
a
t
S

 

 

 

 

 

 

 

 

 

 

 

)
2
/
1
L
E
D
(

 
   

288

within the first two stages of the delay is labelled STOREi and STOREZ,

while storage in the overdue stage is called OVRDUE.

(The reader who

is schooled in computer languages doubtless recognizes that these

variable names have a maximum of 6 characters, and anticipates,

correctly, that a FORTRAN program is in the offing.)

The sum of these

three components of storage within the delay is TVLOAN, total village

loans outstanding.

There is some leakage from the delay.

A proportion

of the overdue loans, represented by the multiplicative parameter PLR,

are written off each year as bad debts.

From the evidence available

there was no way of ascertaining the proportionate loss rate empirically,

so I set PLR at what seemed to me a reasonable value of .33.

For the entire two stage basic delay there is an average delay time

of DEL1.

Following Forrester's example,36 this is split evenly between

the two stages.

Another average delay time, DELZ, characterizes the

repayment of loans which fall overdue.

Judging from what I know of

the slowness of the cooperative inspectorate and the East Pakistan courts

in pursuing and prosecuting loan defaulters, I set the value of DELZ at

2 years.

Through experimentation with the computer program which I

shall explain shortly, I determined that the most appropriate value for

DELl was 1.25 years.

Subsequently, I confirmed this experimental

estimate when I discovered a table in one of the annual reports of

the Gomilla cooperative experiment37 from which the average term of

loans issued in Kotwali thana could be calculated as 1.23 years.

Loans which are repaid on time are indicated by the variable

REPAYI, while loans which enter the overdue delay are indicated by

RINT2.

The relationship of these two rates is governed by the para-

meter PRLATE, the proportion of loans which are late.

The main output

of the overdue delay is another repayment rate, REPAYZ, which together

with REPAYl sums to VLR. the rate of loan repayment.

289

An algelraically simple relationship can be used to determine the

delay output rates. RINTl, RINT2, REPAYl and.REPAY2.

This relationship

can be expressed as

OUT (t) = §I§E§ifll ,

(5.6)

in which OUT(t) is the rate of output of the delay, STORE(t) is the

storage in the delay} and DEL is the average length of the delay.38

This delay-output relationship can be rephrased into the following

FORTRAN equations. in which the asterisks indicate multiplication, and

the slashes indicate division:

RINTl

STOREl/(.5*DEL1)

RINTZ

PRLATE * STOREZ/(.5*DEL1)

REPAYl

(1 - PRLATE) * STOREZ/(.5*DEL1)

REPAYZ

OVRDUE/DELZ.

(

)

5.7

Once these internal delay rate equations have been written, all

that remains in order to complete the sub-model of village loan repay-

ment (VLR) is to specify how the values of the internal storage variables.

STOREl, STOREZ and OVRDUE, change

over time.

Over’a time interval DT,

the following equation clearly applies

STORE (T + rm) = STORE (T) +

T + DT

f

T

dSTORE(t)

(5 8)

dt

'

To estimate the value of the integral on the right hand side of 5.8 we

will use the technique of Eulerian integration.

Despite its crudeness,

this numerical estimation technique is surprisingly accurate when DT is

290

small.

In Figure 5-2, the Eulerian integral of the function f (t) over

the interval from T to T + DI is shown by the shaded area.

It is a

rectangle obtained by multiplying the value of f (t) at time T by DT.

This extraordinarily simple numerical integration technique is used

throughout Forrester's Industrial Dynamics.

It is also fundamental to

Forrester's famous DYNAMO compiler, and to all of the simulation

studies which have been based on DYNAMO, such as Dennis Meadow's book,

_Tpe Limits _tp Growth.

f (t)

 

/

g

T

T + DT

_

v‘

t

Figure 5-2.

Eulerian integral of the function f (t).

The rate of change of storage, dSTgEE t , is determined by the law

of conservation of flow, which stipulates that

gigglil g

IN (t) — OUT (t).

(5.9)

The rate of change of storage, in other words, equals the input rate

minus the output rate.39 Following the convention employed previously,

in Chapters 3 and 1+, I have designated 3% as x.

The rate of change of

STOREl would thus be designated ST1DOT, and so forth.

So we‘ have

291

STlDOT = VLI - RINTl

ST2DOT = RINTi - RINTZ - REPAYZ

BDEBTS = PLR * OVRDUE

OVRDOT = RINTZ - BDEBTS - REPAYZ.

(5.10)

The corresponding integral equations are, in FORTRAN format,

STORE1 = STORE1 + DT * ST1DOT

STOREZ = STOREZ + DT * ST2DOT

(5.11)

OVRDUE = OVRDUE + DT * OVRDOT.

Taken altogether, the financial equations which have been developed

so far in this chapter describe a nearly closed system, as can be seen

from the causal loop diagram in Figure 5-3.

The only "input" to the

system is the number of village cooperatives. NCOOPS.

In the simulation

model constructed from these financial equations (see Appendix C) the

number of cooperatives enters the savings function, XSAV, where it

mediates between the thana and the village levels of analysis.

First

NCOOPS divides TSAV and TVLOANS to bring them into the proper units

for the capital generation equation to act upon.

The annual rate of

capital generation per cooperative is calculated as AC = a (0Ln - BCn) ,

after which NCOOPS multiplies the result to bring it back to the aggre-

gate scale appropriate to the other thana-level financial equations.

The actual FORTRAN statements are given in Figure 5-4.

It should be

apparent that NCOOPS does not influence the model in the classic sense

of an input variable.

NCOOPS cannot be identified as an element of an

input vector U in a model of the form X = AX + GU.

The influence of

NCOOPS upon the system is more like that of a variable parameter.

If it is accepted that NCOOPS does not really represent an input

292

@ s...

T...

—" positive

influence

_....._.. negative

influence

 
flvu

 

TVLOAN

BDEBTS

\

\\

 
1

//

/

‘

VLR

 

BDSM

Figure 5-3.

Causal loop diagram of the financial equations.

NCOOPS

- Number of village cooperatives.

SAV

- Rate of accumulation of savings and shares.a

TSAV

- Total accumulation of savings and shares.

VLI

- Village loans issue rate.

VLR

- Village loans repayment rate.

TVLOAN

- Total village loans outstanding

BDEIBTS

- Rate of accumulation of bad debts.

BDSM

- Integral of bad debts.

5‘ SAV =

c * NCOOPS/1000.

293

C

C

C

C

C

C

FUNCTION XSAV (TSAV, TVLOAN, NCOOPS, ALPHA, BETA)

THIS FUNCTION DETERMINES THE RATE OF CAPITAL

FORMATION (SAVINGS PLUS SHARE PURCHASE) BY

THE VILLAGE COOPERATIVES.

XSAV ASSIGNS A

VALUE TO SAV ONCE EACH YEAR.

THE UNITS OF

XSAV ARE THOUSANDS OF RUPEES PER YEAR.

0 = (TSAV/NCOOPS) * 1000.

OL = (TVLOAN/NCOOPS) * 1000.

R = 2245.

XSAV = (ALPHA * (OL - BETA * c) + R) * NCOOPS/1000.

RETURN

END

Figure 5-4.

The capital generation function, XSAV.

variable, then the system shown in the causal loop diagram in Figure 5-3

is truly closed.

I would argue that this causal loop diagram and its

accompanying equations constitute a minimally acceptable description of

the credit operations of a thana cooperative association.

Within the

technical subsystem of these organizations, cooperative credit opera-

tions were the dominant component.

Consequently, it does seem sensible

‘to classify the T for technical subsystem in the category of closed

systems, above the dotted line in the typological matrix at the head of

this chapter.

While the mathematical model of cooperative credit described so far

suffices as a minimally acceptable representation of the dynamics of

the system, it is far from complete.

A look at the complete computer

program of the cooperative credit simulation in Appendix C will reveal

that the equations which have been explained so far comprise only a

portion of the total model.

Other model components which are explained

below do include input variables.

So even within the special universe

of cooperative credit, I am not contending that the closed system

perspective is sufficient to explain all of the dynamics of the

technical subsystem.

In most organizations, though, a lot can be

294

learned about the essential dynamics of the technical subsystem throngh

closed system analysis.

Mathematically, such an analysis may consist

of supposing that the inputs to the system remain zero, and examining

how the system evolves in response to its initial conditions.

The main deficiency of the credit model depicted in Figure 5-3

is that it ignores one very real constraint:

whether the thana

cooperative association has enough money on hand to authorize all the

loans called for by the loan issue equation, VLI =

3' TSAV - R' .

It

was a shock to me, when I started probing the credit operations of the

cooperative association in Kotwali thana, to discover that every rupee

loaned to the village cooperatives had been" borrowed from the United

Bank, the commercial bank which located a branch office on the grounds

of the'Thana Training and Development Center at Abhoy Ashram.#o

I had

thought that all the money loaned out had come from savings and share

purchases, and from loan repayments.

That preconception, which

corresponds to the causal loop diagram in Figure 5-3, was incomplete.

In November and December of 1965 I visited several of the coopera-

tive expansion projects in Gomilla district.

Of the seven thana

cooperative associations that were founded in 1965-66, my tour took me

to four:

those in Chandina, Laksam, Hajigang, and Brahman Baria thanas.

In my interviews with the directorsof these thana cooperative associa-

tions, I concentrated on achieving an understanding of their financial

operations.

The thana cooperative associations maintained, typically, two

major accounts with the local branch of the United Bank, a current

account and a fixed deposit account.

In the simulation model, these

two accounts are designated CACCT and FIXDEP.

The project directors

said that they conducted almost all their business, including payroll

295

and cooperative

credit

transactions,

through

the current

account.

Abdul

Mannan,

the project

director

of the Laksam

Thana

Central

Coopera-

tive Association,

explained

that money

in the fixed

deposit

account

could

not be withdrawn

for four years.

The main purpose

of the fixed

deposit

account,

he said,

was to serve

as a security

against

which

the

central

association

could

borrow

an equal

amount

at six percent

interest,

one percent

more than the United

Bank paid for the amount

on fixed

deposits.

The bank's

loans

against

the fixed

deposit

security

took the

simple

form of an overdraft

on the current

account.l+1

Although

it was possible

for an overdraft

to swell

beyond

the size

of the fixed

deposit,

this practice

was discouraged,

both as a matter

of policy

and by a higher

loan interest

charge

by the United

Bank.

In

the simulation

model,

therefore,

the issuance

of loans

to village

cooperatives

is halted

whenever

the overdraft

in the current

account

exceeds

the amount

on fixed

deposit.

While

loan issue

is suspended

the

current

account

replenishes

itself,

through

sources

such as loan repay-

ments,

savings

deposits

and interest

earnings,

until

the overdraft

on

current

account

is once again

less than the fixed

deposit.

To

‘ represent

this loan issue

decision

process,

I created

a dummy

variable,

RLOANS.

which

was set equal

to the ideal

loan issue

rate defined

by the

expression

8' TSAV - R' .

So long as the current

account

overdraft

was

not excessive,

VLI, the actual

loan issue

rate,

was equated

to RLOANS.

Whenever

the overdraft

exceeded

the fixed

deposit,

VLI was set at zero.

The FORTRAN equations representing this procedure are in Figure 5-5.

296

RLOANS = 4.74 * TSAV - 174.15

VLI = XISSUE (RLOANS, CACCT, FIXDEP)

FUNCTION XISSUE (RLOANS, CACCT, FIXDEP)

THIS FUNCTION DETERMINES THE RATE AT

WHICH LOANS ARE ISSUED TO THE VILLAGE

COOPERATIVES.

THE UNITS OF XISSUE ARE

THOUSANDS OF RUPEES PER YEAR.

0
0
3
C
0
0
0

XISSUE = RLOANS

IF (XISSUE .LT. O.) XISSUE = 0.

IF (CACCT .LE. - FIXDEP) XISSUE = O.

RETURN

END

Figure 5-5.

The loan issue equations.

The advantages of a continuous time simulation are particularly

evident when it comes to implementing this loan issue procedure.

In

reality, a digital computer can simulate continuous time only by taking

many small steps in discrete time, but those steps can be made frequent

enough to approximate a continuous testing of the relationship of CACCT

and FIXDEP.

By contrast, in a discrete time model in which the incre-

ment of time was one year there would be no way to test during the year

for the size of overdrafts on current account.

Another'advantage of a

continuous as against a discrete time simulation is that the continuous

time approach allows for more accurate computation of interest charges

and interest income.

The scheme for financing the expansion projects in other thanas

of Gomilla district was set forth in a report issued in 1964 by

A. Z. M. Obaidullah Khan, a member of the elite Civil Service of

Pakistan -- the inheritor in Pakistan of the Indian Civil Service of

British days.

During a year which he spent at the Gomilla academy

Obaidullah Khan had studied the costs and financial requirements of

the Kotwali Thana Central Cooperative Association.

In his plan for

297

spreading the Comilla approach to all of the thanas of Gomilla district,

he stipulated that each new project should receive a capital loan of

200 thousand rupees per year for each of five years, after which no

more capital loans were to be issued and the amount loaned was to be

repaid at the rate of 50 thousand rupees per year.

Twenty-five years

after the expansion projects began, all the capital loan funds were to

be repaid. ‘These long term government loans, which were interest free,

were to be used as equity capital, providingta source of loanable funds

for the cooperative credit program.

An unstated hypothesis was implicit in Obaidullah Khan's financial

projections.

He assumed that after a five year*period of start up

funding, the new cooperative projects would be generating enough funds,

in the form of interest earnings and local capital, so as to be able

not only to continue their credit Operations but to repay the government

as well.

In actuality Obaidullah Khan's hypothesis was not tested,

because political agitation and the war for independence from Pakistan

interrupted the course of events in the fifth year of the twenty-five

year period.

The computerized simulation of cooperative credit allows

us to test Obaidullah Khan's projection without disruption.

Through

the simulation, we can freeze the background conditions -- such as

interest rates, wage rates, and political stability -- which prevailed

in East Pakistan in the mid nineteen sixties.

The feasibility of

Obaidullah Khan's long run proposal can then be examined within these

constant conditions.

In addition to capital loans for each new thana cooperative

association, Obaidullah Khan anticipated that a grant of 50 thousand

rupees per project per year would be required to defray salaries and

other’administrative costs.

No cut off date for the issuance of these

grants was specified.)+2

298

Obaidullah Khan's overall proposal was accepted by the Pakistan

government, with the limitation that it be implemented to start with in

only seven of the 21 thanas in Gomilla district.

Those seven projects,

which began business in 1965-66, are the ones I have referred to as the

1966 projects.

In the simulation program, I have called the annual rate at which

capital loans were issued to the thana COOperative associations CLI,

and the rate at which capital loans were repaid CLR.

The annual subsidy

for operating costs is called GRANT.

When I visited the new expansion projects in Gomilla district in

the last months of 1965, I found that the project directors had placed

half of their first years capital loan of 200 thousand rupees on fixed

deposit in the local branch of the United Bank.

The initial value of

FIXDEP is therefore set at 100 in the simulation program, which begins

running at the end of the start up year for the 1966 projects.

To estimate what the annual operating costs of a thana cOOperative

association might be, I relied Upon three sources of information:

a

copy of the 1968-69 budget of the Agricultural COOperatives Federation,1+3

a cost study of the field supervision section of the Kotwali Thana

Central Cooperatives Federation which I completed in 1966,44 and the

research findings on the structural dynamics of the thana COOperative

associations which were reported in Chapter 4.

Breaking down the employees budgeted for the Agricultural Coopera-

tives Federation in 1968-69 into workers and supervisors, according to

the same classification criteria employed in the structural research

described in Chapter 4, I ascertained that the budgeted salary costs

of the ACF could be summarized, in thousands of rupees per year, by

299

the expression 1.4 * WORKRS + 5.4 * SUPVIS.

Supervisors' salaries, in

other words, were nearly four times the level of workers' salaries.

While I was stationed in Gomilla in 1965-66 I looked into the admini-

strative costs of the COOperative credit program, and determined that

over the period 1962-63 to 1965-66 personnel costs comprised an

average of 86.2% of total administrative costs.“5 The administrative

cost equation, thus. is

OPCOST = (1.4 * WORKRS + 5.4 * SUPVIS)/.862

(5.12)

To project the number of workers and superVisors in an average

thana cOOperative association, the findings of the structural research

discussed in the last chapter were incorporated into the simulation via

the two equations

WORKRS

.766 * WORKRS + 15.374

SUPVIS

.894 * SUPVIS + 1.316.

(5-13)

These FORTRAN equations correspond to the equations forwm1 and Sn+1

for the 1966 projects in Table 4-4 in the last chapter.

Like the

discrete time equation for SAV, the rate of local capital formation,

these equations for WORKRS and SUPVIS are evaluated in the simulation

once each year, at the beginning of the year.

In the collections of Gomilla documents at Michigan State University,

I could not find a complete series of data on the actual administrative

costs of the seven thana cooperative projects founded in 1965-66.

The only evidence of actual costs is in a report published in 1967,

which gives the administrative costs of each of the seven expansion

projects for two years, 1965-66 and 1966-67.46

In Table 5-7 I have

compared these figures to the cost estimates derived from the simulation

300

Table 5-7.

Compgrison g; actual and estimated gperating costs

gfgag average thana cooperative association.

Agile;

Mean

1265-66

1966-62

64

73

(Standard deviation)

(14)

(12.7)

N

Estimated

53'

7

OPCOST

61

83

(Standard error of

estimate)

(19.0)

(18.6)

Percent error

 

-’+-7

13-7

a Data from Sarail and Chandpur projects excluded.

equations 5.12 and 5.13.

The approximation is good for 1965-66, and

reasonable for 1966-67.

(In the calculation of the statistiCs in the

first column of the table, the cost figures for Sarail and Chandpur

were omitted.

Each of these thana cooperative associations incurred

over one hundred thousand rupees in administrative expenses in 1965-66,

and cut back substantially in 1966-67.

It appears that they spent at

an exaggerated rate in the start up year, and were called to task for

their excesses.)

Since all financial transactions were handled through the current

account in the United Bank, a great many factors came into play in

determining the rate of change of the account.

These factors included

interest collection and payment, administrative expenses, the Operating

cost subsidy from the government, loan issue and loan repayment (llé a

Elé the government's central bank, and.gi§ g gig the village coopera-

tives), and savings deposits and share purchases.

This variety of

influences is summarized in the following three equations:

301

FMARG = VLIP - CIP - SAVIP + CACTIP + FIXDIP

(5.14)

PROFIT

FMARG - OPCOST + GRANT

(5.15)

CACDOT

(1 - DEP) * CLI - CLR - VLI + VLR + SAV + PROFIT.

(5.16)

The financial margin, FMARG, of the thana cooperative association

is determined by 5.14 to be related to collections of interest on

village loans (VLIP), payments of interest on capital loans (CIP),

interest payments on savings deposits and dividends on shares (SAVIP),

interest collections or payments on the current account (CACTIP), and

interest collections from the fixed deposit account (FIXDIP).

PROFIT is the difference between interest income of the thana

cooperative association and its operating costs, plus the annual grant

subsidy.

In keeping with the non-profit character of the cooperative,

any positive profit is deposited in the current account, where it adds

to the funds available for loan issue.

In case the COOperative associa-

tion is losing money (as would be the case if PROFIT were negative),

the deficiency is withdrawn from the current account.

With the exception of the parameter DEP, equation 5.16 describing

the rate of change of the capital account is self—explanatory.

DEP is

the proportion of capital loans issued each year which is placed in the

fixed deposit account.

Further details of the simulation program, such as initialization

procedures and the method of calculation of interest collections and

payments, are described in Appendix C.

The simulation program itself,

which is reproduced in Appendix C, also contains explanations of the

meaning of all of the program variables.

302

To test the fidelity of the model to the actual dynamics of the

credit cooperatives, the program was run for'a simulated period of

four years.

Initial conditions were calibrated to average variable

values for June of 1966, when the seven expansion projects in Gomilla

thana had been in operation for one year.

Program parameters. such as

ALPHA and BETA in the capital generation equation, were set to values

determined to be suitable for the 1966 projects, as reported above in

this chapter.

By making adjustments in the parameter PRLATE, the

proportion of loans which are not repaid on time, I forced one set of

variables, the sum of loans overdue (OVRDUE) and of loans which had

been written off (BDSM), to come out just about on target in each of

the four years.

Through experimentation, I discovered that the appro-

priate values for PRLATE in 1966-67, 1967-68, 1968-69, and 1969-70

were 0, .08, .09, and .3.

The sharp increase in PRLATE in 1969-70 is

reflective of the widespread political agitation which preceded the

war for independence.

In an interview in 1975. Akhter Hameed Khan

told me that "political defaults" had become quite numerous in 1969-70,

and he was helpless to counteract them.

"The political parties were

saying that the loans had been given by an oppressive government and

should not be repaid.

The politicians said that if they came to power

they would write off the loans," Akhter Hameed Khan said.

Results of the four year fidelity test run of the simulation are

given in Table 5-8.

While the model proves itself adept at imitating

the average performance of the 1966 projects, the high values for some

of the standard errors of estimate (particularly for'total village

loans outstanding) indicate that the model would not be likely to track

the financial history of a particular thana cooperative association with

great fidelity.

The variation in financial data among the thana

Table 5—8.

Compgrison 9; actual and estimated values for the

fidelity rug,

303

Total savings and shares (TSAV)

Village loans issued (VLISM)

Year

ggppgl,

Estimated

Pct. Error

ggppag,

Estimated

Pct. Error

 

 

1967

102

108

5.9

216

203

-6.0

1968

236

229

-3-0

599

619

3-3

1969

475

1970

632

418

642

-12.0

1179

1142

-3.1

1.6

1205

1196

-O.7

Standard error of estimate:

Standard error of estimate:

134.1

(N:28)

249.5

(N214)

 

Village loans repaid (VLRSM)

Vil. loans outstanding(TVLOAN+BDSM)

Actual

Estimated

Pct. Error

Actual

Estimated

Pct. Error

1967

87

83

-4.6

220

1968

225

211

-6.2

594

211

619

-4.1

4.2

1969

480

547

14.0

. 1293

1213

-6.2

1970

786

727

—7.5

1712

1663

-2.9

Standard error of estimate:

Standard error of estimate:

260.4

(N=14)

446.1

(Ne28)

 

Overdues and bad debts(OVRDUE+BDSM)

Loan interest collected (VLINT)

Actual

Estimated

Pct . Dror

1967

O

1968

15

1969

53

O

15

54

1970

264

269

O

O

1.9

1.9

Actual

Estimated

Pct. Error

12

31

89

12

34

84

O

9.7

-5.6

140

127

-9.3

 

304

cooperative associations was sizeable.

When it comes to predicting

average values, though, the model earns good marks for accuracy.

With

the exception of only 2 out of the 24 predictions tabulated, the

percentage errors are below ten percent.47

In the fidelity run the value of VLIR, the rate of interest

collected on village loans, was set at .09.

This was three percent

lower than the official interest rate of twelve percent.

The thana

cooperative associations actually collected less interest on loans

than was due to them.

Against all unrepaid village loans, including

overdues and bad debts, the actual rate of interest collections in the

four fiscal years from 1966-67 to 1969-70 can be calculated as 7.8%,

7.8%, 9.5%, and 9.3%.

(For details of these calculations. see

Appendix A.)

The under collection of loan interest probably resulted

from the practice of forgiving the interest but collecting the principle

in hardship cases arising from flood damage, insect infestation, or

other disasters.

At a nine percent effective loan interest rate, the model predicts

that the average thana cooperative association which is the subject of

the simulation would lose money in the first two years, and would make

a profit on its credit operations in the second two years.

This

pattern of initial losses and later profits is traced in Figure 5-6.

The implication of this trend is that in the initial years of the pro-

ject expenses would have been covered by drawing down the current

account, whereas in the later period the profit from credit operations

would have contributed to the current account, making more money

available for loan issue.

While there is no published data on the actual profitability of

the credit operations of the 1966 projects, the shape of the profit

y

v

 
s
 

 

V /a-a «or

77M! r

/

  

 

;

.5

om‘

 

fl‘Ufll J- 6

Jul/0‘7 [UN P101]? Pica-(cram ’03 77V!

J/I‘Ayum 7"!!4N4 COOPIRA 7/ VI WOO/A 7105’.

306

graph does correspond to what is known about profit trends in the

Agricultural Cooperatives Federation in Kotwali thana.

my own pre-

vious research into the income and expenses of the field supervision

section of the Kotwali Thana Central Cooperative Association (the

predecessor of the ACF) showed that the deficit of interest income

against administrative costs narrowed each year, until in 1965-66 the

section very nearly broke even.

If an annual grant of fifty thousand

rupees had been factored in, as it was in the expansion projects, the

field supervision section of the KTCCA would have shifted from the red

to the black in 1963-64, two years after it started operations.

Even

without such a subsidy, the Agricultural Cooperatives Federation

recorded its first annual profit of 41 thousand rupees in 1967-68.

The trends that are evident in Table 5-8 and in Figure 5-6 appear

quite favorable.

Local capital, loans to village cOOperatives, and

the profitability of the simulated cooperative association are all on

an upward course.

A simple extrapolation of these trends might lead to

the supposition that the financial prospects of the simulated COOpera-

tive association were good, and that the scheme for the expansion

projects proposed by Obaidullah Khan in 1964 was feasible over the long

run.

Was it?

In order to test Obaidullah Khan's assumptions over the full 25

years stipulated in his plan, the simulation program had to be changed

in one respect.

For each of the four years of the fidelity run, a

value of PRLATE was experimentally determined which would yield the

same amount of overdues and bad debts as were actually reported.

For

the extended run, a fixed value of .08 was chosen for the percentage

of loans not repaid on time.

Of the many variables which the simulation tracks to the end of the

307

2000

moo

IGOO

N
'w'
§.

/400

Q

K

0

s.

Q
S
3.

§

Q

E

IZOO

/OOO

500

goo

400

200

a

.

.'

K

50””

v

505M

New

/ z .3 4 5 6 7 a 9/0/1/2/3/45/6/7/5/92011152!

t

#76035 .5- 7 I10“: 770”: or LOAN! ourJrANo/Na ( Thaw),

54.0 on 7': (loam ) ANA mu“ cam 72:. c 7'.“ V)
Item m: I: nwoeo [UM

308

25 year period projected by Obaidullah Khan, four have been selected

for display.

Total loans outstanding, TVLOAN, bad debts, BDSM, and

the stock of local capital, TSAV, are shown as curves in Figure 5-7.

Annual profits, PROFSM, is displayed as a bar graph in Figure 5-8.

Because this long run simulation is a fanciful exercise, the time axis

in these graphs has been numbered with values of t, rather than specific

years.

For the sake of the literal reader, it may be noted that t = 0

corresponds to June 30. 1966, when the seven expansion projects in

Gomilla district were a year old, and that t = 24, the end of the 25

year period, corresponds to June 30, 1990.

The long run prediction of financial disaster for the thana

cooperative association is obvious.

Annual losses became so great

that after t = 20 they would have extended below the bottom of the graph

in Figure 5-8.

By the time the final 50 thousand rupees of the capital

loan are repaid to the government in the twenty fifth year (when

t.= 24, because t = 0 corresponds in this simulation to the end of the

first actual year), the cooperative association is running an annual

loss of 213 thousand rupees, and is getting by through massive over-

drafts on the current account in the local bank.

At the end of the

extended run, when t = 24, the bank overdraft exceeds a million rupees.

In the real world, either the United Bank, or the cooperative associa-

tion or the government would intervene to prevent the financial

catastrophe predicted by the simulation.

The predictions do merit atten-

tion, though, if only because they do not correspond to what any of the

local planners or American economic advisors anticipated when they

advocated the expansion of the Gomilla COOperative model to other thanas

in East Pakistan.

To provide more insight into the simulation predictions, I have

310

plotted them as a phase plane in Figure 5-9.

The variable on the

horizontal axis, TSAV, is the same as one of the variables in Figure

5-7.

But to make an appropriate comparison of predictions with actual

data, bad debts have been added to loans outstanding to form TOTLON =

TVLOAN + BDSM, the variable for the vertical axis.

This adjustment

is required because the actual data reported do not distinguish bad

debts from overdues.

The actual data for TOTLON and TSAV plot a curve on the phase

plane which is so close to a straight line that when a ruler is laid

along the points the deviation is hardly perceptible.

The actual

data conform closely to the equation TOTLON = 2.82 TSAV - 62.46.

The

coefficient of determination, r2, for this regression equation through

the five actual data points is .999.

It COUld be said.that the

actual system was in a state of moving eQuilibrium from 1965-66 to

1969-70.”8

The extended run of the simulation tracks the straight line

trajectory of moving equilibrium until t = 4, after which the simula-

tion trajectory bends to the right.

In a phase plane in which bad

debts are not included in the variable on the vertical axis, as in

Figure 5-10, a very similar pattern can be seen to prevail.

From the causal loop diagram of the credit system, Figure 5-3,

we know that total village loans Outstanding exerts a positive influence

on the rate of local capital generation, SAV, and thus indirectly on

the stock of local capital, TSAV.

We also know that TSAV has a positive

influence on VLI, and thus indirectly on total village loans outstand-

ing, TVLOAN.

This relationship of mutual reinforcement appears to

prevail over the range in which the system is in a condition of moving

equilibrium.

But after four years or so the cycle of mutual

311

N
O

4
'
7
0
'
7

 

--"- AC 7'0)“.

-— [IQ/(€710

zoommmmxmmmmomc

flout! 5- 9

PAM.“ PLAN: 0; ”44.44: 40AM! our: T'AND/NG

.0441: fix“? 0:: 7’: ( r0 non!) tutu/.9 104446:

CAP/7'44 ( 7.9% V ) mom 71/: ex ravage .ewV.

312

N
M
(
-
M
I
T

l‘OO

I400

/800

was

600

600

400

200

 

O (”man/“la!“

F/‘Uié' 540 MAJ! ”(TR/UT 0" TVLOAN may:

7’4"! V IROM 77/! (X7M0£0 RUN.

313

reinforcement (or positive feedback) is interrupted.

An increase in

village capital, such as is observed from t = 4 to t = 5 in Figure

5-10, still yields an increase in loans outstanding, but not in the

same degree as before.

The reason that the system trajectory departs from moving equili-

brium after t = 4 is that the simulated cooperative association runs

out of loanable funds.

There is no longer enough money in the current

account to issue loans at the annual rate defined by the loan issue

equation RLOANS = 4.74 * TSAV - 174.65.

That equation was determined

from a linear regression on data from June 1967 and June 1968, when

the Shortage of loanable funds had.not yet shown up.

The simulation

model shows that the current account becomes negative, indicating an

overdraft, between t = 2 and t = 3, an interval which corresponds to

the fiscal year 1968-69.

Even when the balance on current account is

negative, loans can still be issued at the prescribed rate so long as

the overdraft does not exceed the amount on fixed deposit.

The

simulation hits the loan shut-off switch in the XISSUE function with

increasing frequency after t = 4.

Loans are issued less and less

frequently each year, until after t = 19 not a single new loan is issued

by the thana cooperative association.

When farmers find that they can no longer obtain as many loans as

they think themselves entitled to because of their savings and share

purchases, they lose interest in the cooperative association.

They

withdraw their savings and purchase only the minimal number of shares.

The rate of local capital formation slows, and eventually becomes

negative after t = 9.

In the more complete version of the system

causal loop diagram shown as Figure 5‘11, the decline of TSA\fafter

t = 9 can be interpreted as due to the dominance of the dotted arrow

314

 

 

 

 

positive

influence

negative

influence

 

 

Figure 5-11.

A more complete causal loop diagram for the cooperative

credit system.

(indicating negative feedback) leading from TSAV to SAV.

315

The shift of the state variables TOTLON and TSAV away from the

trajectory of moving equilibrium after t = 4 is also explicable for a

related reason.

The last 200 thousand rupees of capital loan is

received by the cOOperative association during the fiscal year which

ends at t = 4.

Thereafter, the cooperative association starts repay-

ing 50 thousand rupees a year to the government.

Since the underlying

financial conditions change at t = 4, the system might be expected to

respond differently than it had before it passed the dividing line

between receipt and repayment of Capital loans.

Should we be led to conclude, therefore, that the financial scheme

proposed by Obaidullah Khan in 1964 was unrealistic, and that the only

way to support a viable thana cooperative association was to subsidize

it in perpetuity with annual infusions of loanable funds?

Even at a

zero interest rate, was it impractical for the government to expect

that the capital loans made to the cooperative association in its early

years could later be repaid?

Before giving up on the scheme proposed in Obaidullah Khan's

report, let us explore some policy alternatives, to see if there might

be something the model cooperative association could do to avert the

financial degeneration predicted in the extended run of the simulation.'

To investigate such corrective policies, one change was made in the

simulation program that was used for the extended run.

For the policy

experiments, it was assumed that the beta values of the capital

generation and loan issue equations were identical.

The value chosen

for the two betas was 4.46, the average of the

B and

B' coefficients

that had been determined through linear regression.“9 As a result of

this modification, the long run simulation results could more readily

316

/

--- amt/pea RUN (a w’)

   

-—,—-MJIA/Ne MN (I '4’)

:00 m we no moo mm mm

rsAV

N
O
L
T
O
T

2000

/300

no»

you

IMO

IOOO

§
§
§
§
.

F/WII 5' /Z

COMP/03470” Of WJJCTOR/G'J MR 734‘

(XTEVDIO RUN AND 756’! .5145!UN! RUM

317

be used as a baseline against which policy changes could be tested.

It is evident from the comparison of phase plane trajectories in

Figure 5-12 that the assumption

3 =

3' does not substantially alter

the overall dynamics of the system.

To set the stage for'a review of policy options for improving the

long run financial performance of the simulated thana cooperative

association, let us first take a look at a situation which is much

worse than the baseline run -- the closed system condition in which

the organization Operates with no inputs.

The system trajectory for

the closed system seen plotted in Figure 5-13 proceeds from initial

conditions alone.

The number of cOOperatives (NCOOPS) is held constant

at the initial level, and there is no infusion or repayment of capital

loans.

The closed system run was made under the baseline assumption

that

B =

8'.

The resultant numbers are small, so Figure 5-13 was

plotted on a scale magnified 10 times from that used in the other phase

planes in this chapter.

Direct comparison of the closed system and

baseline runs on the same graph is precluded by the change of scale.

However, the shapes of the trajectories in Figure 5-12 and 5-13 show

some similarities in the general behavior of the system.

With only the funds initially available in the current account,

the cooperative association manages

to issue loans and to raise

capital in the villages.

The rally is short lived, though.

The loan

shut—off switch in the XISSUE function restricts loan issue as early

as the second year, and by the end of that year loans outstanding have

reached their maximum.

Village capital reaches its peak the-year

after.

The last loan is issued by the cooperative association in the

fourth year.

An interesting finding which is evident in Figure 5-13 is that the

2':

Jan

.00

Z“

240

20

2x

/
4
0
4
7
0
'
7

t

£50

 

 

 

t
t
‘
k
fi
'
é
’
t
s

I
I
I
I
U
I

319

 
 

/¥

 

7'44 V

605M

rum”

I

z. 1416 71 cld/I/t/I/l/Jfifl/l/ezozlafifi

77M l

F7603! 544 .930ch 770M! 0’ 40A”: OUT’TAND/IVO, 510 “I 75,

MD VII-4.465 CAP/7344 {IO/'11

TA’I £40550 JIJT'IM

IUN.

320

cooperative credit system has a stationary equilibrium point.

In terms

of the variables in Figure 5-13, the equilibrium is located at

TSAV = 76, TOTLON = 10.

After t = 9 the system solution approaches

indistinguishably close to that equilibrium point.

From the curves

in Figure 5-14, it is apparent that virtually all of the unrepaid

loans after t = 9 consist of bad debts.

A glance back at some of the other graphs in this chapter reveals

that the extended and the baseline runs of the simulation also approach

a stationary equilibrium near the end of the 24 year period.

For the

extended run, the stationary equilibrium is at TSAV = 698, BDSM.= 576,

TVLOAN = 0, while the baseline run reaches a stationary equilibrium at

TSAV = 653, BDSM = 520, TVLOAN = 0.

In all three cases, then -- in

the extended.baseline and closed system runs -- these state variables

come to rest when there are no more village loans outstanding.

Although the exercise has something of the flavor of.a beauty

contest among cadavers, a comparison of the stationary equilibrra of

the closed and Open systems permits some inferences to be made about

the contribution of inputs to system growth.

The closed system comes

to rest with a stock of village capital valued at 76 thousand rupees.

In the baseline run, the final value of TSAV'is 653 thousand.

The

entire difference of 577 thousand rupees is due to two factors:

growth in the number of village cooperatives, and capital infusions.

To

separate these influences, the simulation was run a) with the regular

schedule of capital infusions and repayments, but with no growth in

village cooperatives, and b) with the no capital loans, but with the

standard growth pattern for the village cooperatives. ‘The resultant

three-way comparison among the no-growth run, the no-capital-loans

run, and the baseline run is shown in Figure 5-15.

All three runs were

321

row

I”

I”-"’)

has

.

/

/

/

I

I

Pager/wt gay

man

/

mama,

wv

/

,

N
A
O
L
V
T

mmmm/m/zx/m

7'54 V

Ila/t: Jo/J

71.441: 701 145 [at me No- 634»! 770 .60/v ,
M: Na - cop/7:44. 40.4w :wv, AND m:
swam: aw.

 

made under the baseline assumption that 8

==

8'.

322

Without any growth in the number of village coOperatives, the

system trajectory spurts upward, and then plummets to the same

stationary equilibrium point on the TSAV axis as was reached by the

closed system run.50 The effect of organizational growth at the village

tier without capital infusions to the thana cooperative association is

to swing the trajectory over toward the same resting point at

TSAV = 653, TVLOAN = 0 as was the target of the trajectory of the base-

line run.

After a while the two trajectories even follow the same path

toward that stationary equilibrium point.51 As with the comparison

between the closed system and the baseline runs, we may note that the

difference in the final level of village capital between the no-growth

and the no-capital run is 577 thousand rupees.

All of that difference,

we now know, is ascribable to the increase in the number of village

cooperatives.

If growth in the number of village COOperatives is a force for

financial expansion in the thana cooperative association, perhaps a

policy of stimulating growth would boost village capital and loan

volume, and pull the thana cooperative association out of its long

run depression.

This possibility was investigated. with results to be

described shortly.

During the period from 1966 to 1970, the growth in the number of

village cooperatives in the seven expansion projects in Comilla dis-

trict was nearly a linear function of time.

A regression of local

cooperatives against the simulation variable t, which has a value of

zero in 1966, yielded the equation NCOOPS = 42 * T + 41 (N = 35,

r? = .75, overall F = 98.66, significance of F = .000).

With two

modifications, this equation was incorporated into the simulation

~323

program in the XCOOPS function.

First, the parameter RGROW was

substituted for 42, so as to allow for variation in the rate of growth

of village cooperatives.

Second, a cap of 350 cooperatives was inserted

in the XCOOPS function.

Roughly 73% of the villages in an average

thana would have been organized by the time 350 local COOperatives had

been established.

The remaining 25% would represent those villages

which declined, for reasons of their own, to join the cooperative

program by organizing a local cooperative society.

The way that these

considerations are written into the simulation program is shown in

Figure 5-16.

C

C

C

C

C

MCOOPS = 350.

NCOOPS = XCOOPS (T, RGROW, MCOOPS)

FUNCTION XCOOPS (T, RGROW, MCOOPS)

THIS FUNCTION DETERMINES THE NUMBER OF VILLAGE

COOPERATIVES.

XCOOPS = RGROW * T + 41.

IF (XCOOPS .GE. MCOOPS) XCOOPS = MCOOPS

RETURN

END

Figure 5-16.

The XCOOPS function.

The documented experience of the COOperative movement in Gomilla

Kotwali Thana indicates that there is a price to be paid for rapid

growth in the number of village cooperatives.

The faster new local

cooperatives are organized and registered, the greater is the prOportion

of their members who are insincere.

In the rush for expansion,

marginal members will be picked up who will not follow through in

repaying their loans, or who may intend to subvert the organization by

corrupting it.

Moreover, the more attention the inspectors pay to the

enrollment of new cooperatives, the less attention they can pay to

324

M55411“!

N

mmmm/coo/mm

731V

“00

§
§
§
§
§
§
§
§
§
§
,

HOUR! I. /7

’A’AJl POI 7M/7' 0’ :7er [LIFO/VII 7'0
7"!!!“ M7”! 44’ “0W7?! IN V/“Adl COO/91R A77 VII.

325

supervising the management of those already registered.

As a conse-

quence of these factors, the proportion of loans not repaid on time

increases with the rate of growth of cooperative societies.

To give mathematical form to this assertion, two representative

cases were chosen.

When the rate of growth of the local cooperatives

is zero, the inspectors can concentrate all their attention on main-

taining management standards among the existing set of cooperatives.

In that instance, I presumed that the proportion of loans not repaid on

time, PRLATE. would have a minimum value of .03.

For the baseline run,

in which the rate of growth, RGROW, is 42, I presumed that PRLATE

would equal .08.

Matching up these cases, I derived the relationship

PRLATE = .0019 * RGROW + .03.

This equation was used in the baseline

run, the closed system run, the no-growth run, the no-capital loans

run, and in most of the runs reported hereafter in this chapter.

In Figure 5-17 we see the system response to rates of village

cooperative growth about one half and about double the rate in the

baseline run.

A faster rate of growth does indeed allow for'a higher

maximum volume of loans, but the denoument is identical for'all rates

of growth.

The faster the system grows, the faster it reaches the

ceiling of 350 local cooperatives, and thereafter the faster it falls

toward the stationary equilibrium point at TSAV = 653, TVLOAN = 0.

Forcing the rate of growth through a crash program to register village

cOOperatives amounts to a quick fix for the thana cooperative

association.

For a short while profits glisten, but the final ruin

is precipitous.

The sins of fevered expansion are punished swiftly.

Is there an optimum rate of growth for the village cooperatives?

To investigate this question, measures of the successfulness of

different growth policies are required.

For such measures I chose the

326

Table 5—9.

Year 10 values of village capital, loans outstanding, and

annual profits for different rates of growth in the number

of village cooperatives.

RGROW

TSAV

TVLOAN

PROFSM

O

10

20

30

40

50

60

70

80

90

100

277

506

728

950

1046

1049

1041

1035

1026

1016

1001

780

1085

1347

1563

1660

1647

1635

1598

1547

1498

1423

-46

-30

-15

- 3

5

5

1

- 2

- 5

- 9

-14

32?

values of’TSAV, TVLOAN and PROFSM at the end of’the tenth year of

the simulation run.

According to these measures, which are presented

in Table 5-9, the optimal rate of growth lay in the range of 40 to 50

new local cooperatives per year.

It was exactly in this range that the

actual average rate of growth of 42 was located.

We may conclude that

there would have been no grounds for'a policy of decelerating or

accelerating the average rate of growth of the number of village

COOperatives in the seven expansion projects in Gomilla district.52

If it were not for the loan shut-off switch in the XISSUE function,

the trajectory of the baseline run of the cooperative credit simulation

would continue its upward course without end.

The trajectory bends

downwards because the thana cooperative association runs out of loanable

funds.

Loans are shut off whenever overdrafts on the current account

exceed fixed deposits.

It occured to me, as perhaps it has occured to

the directors of the thana cooperative associations, that a remedy

might be to increase the amount of money in the fixed deposit account,

so as to allow more loans to be issued before the XISSUE function trips

the loan shut-off switch.

To increase fixed deposits, part of the

annual capital loan of 200 thousand rupees was diverted from the current

account to the fixed deposit account.

The value assigned to the

parameter DEP defines the proportion of capital loans diverted to

fixed deposits.

In the baseline run DEP has a value of zero.

To test

the hypothesis that long run financial performance would improve if

fixed deposits were increased, the simulation program was run with

DEP = .1, DEP = .2 and DEP = .3.

The results of this experiment, as

measured by the year 10 values of the state variables representing

village capital, loans outstanding, and annual profits, are displayed

in Table 5-10.

328

3500

N
A
O
L
V
T

 

zoo 4” can no macaque/w xawmfzoemm

TJAV

F1604! f-II

fill/L45 MITM/T 0" 3737471! llJflOA’J! 7v fill/(1!”

VII-Ml“ 0’ 6! TA

Table 5—10.

Year 10 state variable values for different rates of

deposit in the fixed deposit account.

329

DEP

TSAV

TVLOAN

PROFSM

O

.1

.2

-3

1045

1050

1042

1659

1679

1653

1037

1634

5

4

2

0

Between the baseline policy of no diversion of capital loans into

the fixed deposit account (DEP = 0) and the policy of diverting 10% of

each year's capital loan, the simulation yields an ambiguous comparison.

Whatever advantage there may be in such minimal diversion is small.

When larger proportions of the capital loan are entered in the fixed

deposit account the long run financial performance of the coOperative

association deteriorates, but not drastically.

More substantial changes in system behavior’are observed when the

BETA parameter is altered.

The BETA parameter influences the credit

system in two opposing ways.

In the equation for the annual rate of

loan issue BETA has a positive influence:

the larger the value of

BETA the more loans will be issued, so long as loanable funds are

available.

In the capital generation equation BETA exerts a negative

influence:

the larger the value of BETA the less capital will be

raised in the villages through savings deposits and share purchases.

Figure 5-18 shows how these counteracting influences play themselves

out for different values of the BETA parameter.

The dashed lines in Figure 5-18 demonstrate that the net effect

of BETA on outstanding loans reverses as the thana COOperative

association matures.

At t = 2 a change from a larger to a smaller

value of BETA causes the system trajectory to shift downward.

Eight

330

years later» at t = 10, a similar shift in BETA causes the system

trajectory to rise.

What is happening is that the greater'rate of

local capital formation caused by a smaller BETA allows the current

account to replenish itself faster at t = 10, resulting in less

frequent cancellations of loan issue.

For the first time, a long term remedy for the financial diffi-

culties of the thana COOperative association is apparent.

At a BETA

value of 2.0, the system trajectory escapes the downward pull of

declining loan issue, and holds to a steady upward course.

A more

detailed check of BETA values than is shown in Figure 5-17 reveals

that 2.0 is indeed the critical value for BETA, at which the trajectory

breaks away.

At BETA = 2.25 the trajectory begins a gradual descent

after t = 15.

BETA equals 2 is like the escape velocity of a rocket;

it is the parameter value at which there is no evidence of a tendancy

to descend throughout the entire 24 year course of'a simulation run.

Let us call this critical value of the BETA parameter the "breakaway

value."

Even though the simulation has revealed a long-term financial

solution for the thana cooperative association, it is unlikely that

the corresponding policy would be adOpted.

The BETA parameter is

determined by the coverage requirements for loans which are imposed

by the thana cooperative association.

The most liberal loan policy

of the thana cooperative associations allowed total loans outstanding

to attain five times the sum of savings and shares.

Often times the

credit rating of a village cooperative would be reduced, so that the

ceiling on loans outstanding would be three or four times the accumu-

lation of local capital.

Regression analyses reported earlier dis-

closed that the (3 parameter in the capital generation equation for

331

seven expansion projects was 4.17, whereas for Kotwali Thana it was

2.91.

My interpretation of the difference between these two beta

values is that average credit ratings were lower in Kbtwali Thana, due

to a stricter adherence there to the policy of supervised management.

In Kbtwali Thana, it will be recalled, the least credit worthy local

societies (the C class societies) were permitted to borrow only 3

times their capital contribution.53

From the evidence of the regression analyses of financial data

from Kotwali Thana and the 1966 projects, it seems that the parameters

8 and

8' typically assume values which are slightly less than the

average multiple of local capital defined by the loan issue policy of

the thana cooperative association.

A joint BETA value of 2.0, which

was determined above to be the breakaway value for the simulated thana

cooperative association, would correspond to a policy in which loans

outstanding were not allowed to exceed about two and a half times local

capital.

That would be a strict policy indeed, and one which the

farmers in the local cooperatives might not be willing to accept.

Certainly such a stringent policy would be at the opposite extreme

from the loan policy of the network of Union Multipurpose COOperatives

in East Pakistan; these chronically bankrupt cooperatives required

that only one tenth of loans outstanding be covered by local capital

contributions.

Such an easy-going policy would yield a BETA value of

about nine.

Extrapolation of the trends charted for BETA values up to

5 in Figure 5-18 would indicate that a cooperative which operated with

such a slim coverage requirement would quickly run out of loanable

funds, and could only continue its operations through massive and

unremitting injections of capital from the government.

Such was. in

reality, the experience of the Union Multipurpose Cooperatives.

332

In his report on his Tour g; Twenty Thanas Akhter'Hameed Khan

reflected on the ideology of self-sufficiency which was drummed into the

leaders and members of the Comilla-type cooperatives.

As the preceding

analysis has made clear, this ideology and its accompanying loan issue

policies were pivotal to the financial success of the cooperatives.

The cooperative programme is a hard programme.

It demands discipline and self-sacrifice.

It is

not a programme which temptingly offers facili-

ties without insisting on any reciprocal effort....

The cooperative project promises release from

the money lenders if the members would learn the

habits of thrift and honest repayment and bear

the exfiense of the organization.

Not other-

wise.5

Rather than raising loan coverage requirements to the austere

levels indicated by Figure 5-18, the thana cooperative associations

had another and more palatable policy option available.

They could

have collected the interest which was due to them at their stipulated

rate of twelve percent.

As can be seen in Figure 5-19, an effective

village loan interest rate of VLIR.= .12 would have allowed the simulated

cooperative association to break away to a permanently ascending

trajectory.

Figure 5-19 vindicates the financial plan for the new cooperative

projects which Obaidullah Khan prepared in 1964.

If only the expansion

projects had collected the total amount of loan interest which was

their due, they would have prospered to the end of the century and

beyond.

From the viewpoint of systems theory, the phase portrait in

Figure 5-19 is of interest because of the straight line which all the

system trajectories eventually follow, either downward if the value of

VLIR is low, or upward if VLIR is set at the breakaway value of .12 or

§
§
§
§
§
§
§
§
§
§
§

\
400

§
§
§
§
§
§
§
§
°

!
M
0
.
1
V
T

333

 

zoo 4006x400 macaw/max

7141/

FIJI/RI

5V9

par/Au Accra/7’ at mam» ”UPON“ 7v
alfituewr VAL at: 0: um, 77/: must
40sz xxvruur are.

334

above.

Somewhere between the eleven and twelve percent rates of

interest there must be a value of VLIR at which the system trajectory

comes to rest along the slightly tilted straight line of moving

equilibrium.

If the breakaway value of twelve percent interest is

comparable to the escape velocity of a rocket leaving the earth, then

that special interest rate between eleven and twelve percent is

analogous to the velocity required to launch a satellite into orbit.

Like a satellite circling the earth, the cooperative association which

collected the right rate of interest between eleven and twelve percent

would end up in a stationary condition, with neither village capital

nor loans outstanding changing from year to year.

Up to this point we have examined the response of the cooperative

credit system to variations in one parameter'at a time.

We have reviewed

the way that the system reacts to unilateral alterations in the rate of

growth of village cooperatives (RGROW) , to the proportion of capital

loans diverted to the fixed deposit account (DEP), to the loan coverage

requirements (BETA), and to the rate of interest on cooperative

loans (VLIR) .

The simulation program can just as well be used to look at the

implications of variation in more than one parameter;

The interaction

of parameter variation can thereby be demonstrated.

Without having

the explicit intention of doing so, we have already uncovered one such

parameter interaction.

From Figures 5-18 and 5-19 we know that the

breakaway value for the village loan interest rate, VLIR, is .09 when

BETA is set at 2, and is .12 when BETA has a value of 4.46. ,A more

formal parameter interaction study can readily be executed by the

simulation program.

The appropriate FORTRAN commands can be written

in a few lines as instructions for a set of nested DO loops.

One such

335

parameter interaction study is reported here.

Many others could be

devised.

To obtain a more complete view of the relationship between loan

limit policies, as reflected in the value of the parameter BETA, and

the breakaway value for the village loan interest rate, VLIR, a third

parameter was incorporated in the study.

The percentage of loans not

repaid on time, PRLATE, should have a bearing on the relationship

between BETA and the breakaway value for VLIR.

The greater the

proportion of loans which fall overdue, the higher the interest rate

that would have to be charged (for any constant value of BETA) so as

to sustain the financial health of the thana COOperative association.

The findings of this study of the interaction between loan coverage

requirements (BETA), the percentage of loans not repaid on time (PRLATE),

and the breakaway values of village loan interest rates (VLIR) are

reported.in Table 5-11.

As a point of incidental information, it may be

noted that the simulation program was run 126 times, for 24 years each

time, during the course of this investigation.

Thanks to the simpli-

city of FORTRAN instructions, this massive amount of computing was all

reported in one printout.

To control the cost of the study, the

increment of time, DT, was changed to .05 from the regular value of .01.

Two technologies, or methods. of providing rural credit were

discussed earlier in this chapter:

supervised agricultural credit and

supervised cooperative management.

The case for sUpervised management

could hardly be made more clearly than by the information presented in

Table 5-11.

Under the supervised management policy, those local

cooperative societies which had few overdue loans were rewarded with

higher loan limits.

As can be seen from the table, such a trade off

could be accomplished without sacrificing the financial health of the

335

Table 5-11:

Breakaway values of VLIR for different settings of PRLATE

and BETA.

 

2.5

BETA

3.5

4.5

PRLATE

.00

.05

.10

.15

.20

.25

.08

.10

.12

.12

.14

.14

.10

.12

.12

.14

.14

.16

.10

.12

.14

.14

.16

.16

 

337

thana cooperative association.

At an effective interest collection rate

of twelve percent, for example, the cOOperative association could afferd

to offer'a loan limit of 5 times village capital (corresponding to a

BETA value of about 4.5) to those local cOOperatives having five

percent overdues.

Village cooperatives in which ten percent of the

loans fell overdue should be allowed a loan limit of no more than 4

times their accumulation of savings and shares: the BETA value for such

cooperative societies would be about 3.5.

Finally, a loan limit of 3

times village capital, corresponding to BETA = 2.5, should be imposed

on local societies in which 15% of the loans fell overdue.

These three

recommended policies match up well with the loan limits for A, B, and

C class societies which were enforced in the mid-nineteen sixties in

Gomilla Kotwali Thana.

To their subsequent regret, the cooperative officials in Kotwali

thana abandoned these strict loan limit policies in 1967, when they

launched a drive for the issuance of a huge volume of cooperative loans.

The impact of that change of policy is apparent in Figure 5-20.

During

the period from 1963 to 1966, the cooperative credit system in Kotwali

thana had followed a straight line trajectory of moving equilibrium.

The four data points for those years all lie close to the equation

TOTLON = 1.85 TSAV + 147.55 The coefficient of determination, r2, for

this regression through the 1963 to 1966 data points is .98.

When the

loan drive was launched, the phase plane trajectory of TOTLON versus

TSAV for Kotwali Thana departed from its established course.

The

campaign of retrenchment, begun in 1969. turned the trajectory back

toward the course that it had followed in the early years.

In the Gomilla cooperative system the thana and the village tiers

of organization were interdependent.

The thana association depended on

338

H
H
H
H
H
H
.

 

zoo mmmxmxav/m/év/mm.

75.4 V

[I‘d/II 3" 20

III/'3‘ ’01 TM” Of VIM“ 641'] 7746 VII!“

‘01”: OWJflQIVO/IVC ’0‘ 77!! #071064 Ll

f/IAIVA COOPIIA 7'7 YIJ.

339

the village cOOperatives to generate equity capital through savings

and shares, and to cover part of operating costs through the payment

of loan interest.

The village societies depended upon the thana

association for much more:

for loans, for training, for marketing, for

the installation and maintenance of tubewells, for tractors, and for

many other services.

Without the thana association the village coopera-

tives would have withered and expired.

Within the administrative and

political context of East Pakistan, or of contemporary Bangladesh, it

would.have been impossible for villagers to start or to regenerate a

thana cooperative association.

So despite the interdependence of the

two tiers, I would say that the thana cooperative association, rather

than the village cooperatives, was in the long run the more important

institution, the one whose health should be most carefully safeguarded.

The supervised management policy protected the financial well-being

of the thana cooperative association.

The policy of supervised

agricultural credit did not.

The supervised management policy may

therefore be judged the superior of the two, both for the special case

of the Gomilla cooperatives and for cooperative institutions in other

less-developed countries.

Notes for Chapter 5

The Technical Subsystem

1 Daniel Katz and Robert L. Kahn, The Social Psychology 9;;

Orggpizations, John Wiley and Sons, New York, 1966, pp. 30, 31.

Re-

printed by permission of John Wiley and Sons.

2 Bruno Malinowski, A Scientific Theory 9;; Culture, The University

of North Carolina Press, 1944.

3 Philip Selznick, Leadership 1g Administration, Row, Peterson,

Evanston, Ill., 1957.

4 Charles Perrow, Complex Organizationg, A Critical Essay, Scott,

Foresman, Glenview, Ill., 1972, p. 177.

5 Amitai Etzioni, Modern Organizations, Prentice Hall, Englewood

Cliffs, New Jersey, 1964.

6 Fremont E. Kast and James F. Rosenzweig, Orggpization gag

Management:

‘A Systems A

roach, McGraw Hill, New York, 1974.

7 Ibid., p. 181.

8 Frederick Taylor, Scientific Management, Harper, New York, 1947.

Mr. Gilbreth's fame as the father of 12 children, one of whom wrote

Cheaper 21 the Dozen, has been more durable.

9 William H. Whyte, Jr., The Organization HEB: Simon and Schuster,

New York, 1956.

19 David J. Hickson, D. 5. Push. and Diana C. Pheysey, "Operations

Technology and Organization Structure:

An Empirical Reappraisal," Admini-

strative Science Quarterly, 1969, V. 14, p. 395.

Reprinted with permission.

11 One of the more insightful reviews of the discrepancies among

these studies is to be feund in Bernard C. Reimann's article, "Dimen-

sions of Organizational Technology and Structure:

An Exploratory

Study," in Human Relations, V. 30, No. 6, 1977. PP. 545-566..

12 Joan Woodward, Industrial Organization:

Theory and Practice,

Oxford University Press, London, 1965.

13 Edward Harvey, "Technology and the Structure of Organizations,"

American Sociological Review, V. 33, 1968, pp. 247-259.

340

341

14 Andrew H. Van de Ven and Andre L. Delbecq, "A Task-Contingent

Model of Work Unit Structure," Administrative Science gm ay, V. 19,

1974. pp- 183-197

15 Pradip N. Khandwalla,

"Mass Output Orientation of Operations

Technology and Organizational Structure," Agpinistrative Science

Quarterly, V. 19. 1974. pp. 74-97-

16 Tom Burns and G. M. Stalker, The Management pf Innovation,

Tavistock Publications, London. 1961.

17

Kast and Rosenzweig, _p, p;p., p. 197.

18 Eugene Litwak, "Models of Bureaucracy which Permit Conflict,"

in Fred D. Carver and Thomas J. Sergiovanni, eds., Organizations and

Human Behavior:

Focus on Schools, McGraw Hill, New York. 1969,

“”3—

19 Norman T. Uphoff and Milton J. Esman, Local Government for

Rural Development:

Analysis pf Asian EXperience, Rural Development

Committee, Cornell University, Ithica, 1974, p. 28.

20 M. Zaker Hussain, pr‘pp Orgppize Village Groups £95 Pump

Irrigetion, Pakistan Academy for Rural Development, Gomilla. 1966,

p. 4.

21 S. A. Rahim, Voluntary Group Adoption of Power Pump Irrigation

in Five East Pakistan Villages, Pakistan Academy forVillage Development,

Comilla, 1961,p. 12.

22 Levern W. Faidley, Computer Simulation epippe_Cooperative

Approach 39 Tractor Mechanization pp“; Developing Country, Ph.D. Thesis,

Michigan State University, 1974, p. 11.

23 Ibid., pp. 18, 61. 69.

2” _Il-icL. pp. 57. 136.

25 Nazmul Ashan, Loan Utilization p1 the Cooperative Members pf

Gomilla, Pakistan Academy for Rural Development, Comilla, 1966.

6 Conversation with the author, 1975.

27 The findings of the field research team, of which I was a

member, were reported in Zaker Hussain, A Field Investigation into ppe

Management of Village Cooperatives in GomillaExperimentalArea,

 

Pakistan Academyfor Rural Development,Gomilla. 1966.

8 Conversation with the author, 1975.

29 After the gazette of loan limits was published, cooperative

members had to go through a similar multi-staged procedure to obtain

their money.

Like the procedures for saving, share purchase, and

setting loan limits, the loan issue procedure required the individual

cooperative member to negotiate with the ACF through his village

342

cooperative.

Except when it prosecuted defaulters, the ACchonducted

. transactions only with its member cooperative societies, not with

individual farmers.

For a description of the loan issue procedure,

the reader is referred to Ali Akhter Khan, Rural Credit Program pf ppe

Agpicultural Cooperative§_Federation, Bangladesh Academy for Rural

Development, 1971. PP- 9-17.

Much of the text on those pages is

plagarized verbatim, and without attribution, from the author's un-

published manuscript, written in 1966, on Tpe,Comilla System pf_Spper-

vised Credit.

30 No copies of the loan gazettes could be found among the

collections of Gomilla documents at MSU, so I cannot estimate trends

in the proportions of primary cooperatives in the A, B and C categories.

31 Mathematically, the important issue is not the value of a by

itself, but the value of the product

as and the behavior over time

of the variable 0L.

If OL were constant (which it was not), then a

value of

GB greater than one would result in oscillations of increas-

ing amplitude in the value of C.

32

Financial data for the 196h’projects (Nator, Gaibandha, and

Gouripur) is unreliable.

The figures on total loans outstanding

presented in different annual reports of the Gomilla cooperative

experiment are inconsistent for these projects.

33 In case anyone is intent on replicating this research, or on

applying these analytic methods to present day cOOperative projects in

Bangladesh, let me warn against a trap in which I became ensnared dur-

ing several weeks and many computer runs.

The cOOperative projects

routinely report the rate of savings and the rate of share purchase

during the year.

I mistook this as A C, and regressed it against loans

outstanding (0L) and total savings and shares (C) at the close of the

fiscal year.

But if OL and C are given the index n for the end of the

fiscal year, then the

A C in the annual reports represents Cn - C

.

Equation 5.2 is not an historical equation, but a predictive equatgbh.

What is needed is a value for A C which represents Cn+

- Cn'

The

adjustment was easy to make, once I recognized the need for it.

3“ Assuring the comparability of data on the two sides of the

regression equation was important.

Figures on yearly loan issue in the

annual reports represent loans issued during the fiscal year.

Figures

on savings and shares are given as of the_end of the fiscal year.

The

local capital (savings and shares) figures for the end of the last

fiscal year and the end of the current fiscal year were averaged to

derive an estimate for TSAV which could properly be regressed against

loan issue during the fiscal year.

Because I did not have reliable

data on loan issue in each of the expansion projects in 1968-69 and in

1969-70, I had to eliminate those years from the data base.

1965-66

was also eliminated, because it was the start-up year.

The.regression

was run, then, on only 14 cases in the years 1966-67 and 1967-68.

Despite the slim data base, I have confidence in the regression results

for the loan issue equation in the 1966 projects for several reasons.

First, the statistical indicators r2 and F were exceptionally high.

Second, the model given in 5.5 came out much better than some reason-

able competitors.

An alternative explanation for loan issue is that it

343

is related to the number of borrowers, or to the number of local

cooperatives.

The more borrowers or cooperatives, the more loans

issued.

Neither of these variables entered the regression equation on

the first step.

When TSAV and COOperative members were tested as

independent variables, members entered after TSAV, and with an F value

which was significant at the .662 level.

The change in :2 which

resulted from adding members to the regression was only .0009?, which

is tiny by anyone's standards.

In a separate test in which TSAV and

the number of local cooperatives were given the opportunity to enter

the regression, COOperatives also entered on the second step, and

contributed only .00064 to the change in r2.

my third reason for hav-

ing faith in the loan issue equation VLI = 4.74 TSAV - 174.65 was that

it performed excellently in the simulation model which is explained

somewhat later in the text.

35 The two stage basic delay model is known technically as a

second order'delay.

The order of the delay is the same as the order

of a corresponding linear differential equation.

For the two stage

loan delay, the corresponding equation is

2

a2 Ql_1:iii.+ a.1 érgéiil + y (t) = x (t),

dt

in which y (t) is the lagged output, VLR, and x (t) is the unlagged

input, VLI.

In accordance with Forrester's approach, I have side-

stepped calculus in this presentation.

36 Jay W. Forrester, Industrial Dxpamics, John Wiley and Sons,

New York, 1961, p. 88.

37 Badaruddin Ahmed, A. m Rural Cooperative System _fpr Gomilla

Thana, Ninth Annual Report, 1968169, Pakistan Academy for Rural DeveIOp-

ment, Gomilla. 1970, p. 27.

38 Forrester, pp. p;p., p. 87.

39 To see how equation 5.9 demonstrates the conservation of flow,

take the integral of both sides.

The equation would then state that

the amount in storage equals the amount put into storage less the

amount taken out of storage.

40 My notes of an interview in 1965 with Mr. Shubash, loan

accountant for the Kotwali Thana Central Cooperative Association, have

a large question mark next to the statement that the KTCCA was "paying

5%% to United Bank to obtain money in order to loan it to primary

societies."

“1 Interview with Abdul Mannan, Project Director, Laksam Thana

Cooperative Association, November, 1965.

For overdrafts in excess of

fixed deposits, interest charges rose to 7%%.

This subtlety'is

ignored in my simulation model.

“2 A.Z.M. Obaidullah Khan, The Gomilla District Development Project,

Pakistan Academy for Rural Development, Gomilla, 1964.

In addition to

the capital loan of 200 thousand per year and the grant of 50 thousand

per year, the projects were to receive a loan of 200 thousand a year

3a).,

for each of five years for the purchase of tractors, pumps, and other

agricultural machinery.

I have ignored this additional loan, since it

was unrelated to credit activities.

#3 Included in The Kotwali Thana Central Cooperative Association,

m

Ltd, Old Abhoy Ashram, Gomilla, Budget for the Year 19

69, pp. 77-92.

#7 From my unpublished monograph on Tpe_Comilla System p§_Super-

vised Credit.

45 The category of administrative costs covers all staff salaries

of the thana cooperative association, including the salaries of village

accountants. _Travel allowances for the inspectors are included under

the category of non-personnel expenses.

Other non-personnel expenses

would be items such as court fees for prosecuting loan defaulters, and

office supplies.

In the standard accounting terminology for the cooper-

ative projects there was an additional category of costs for the credit

program, termed expenditures for training and extension.

These were

reported separately from administrative expenditures.

Expenses of

training and extension would include costs such as local travel allowe

ances for managers, chairmen and.model farmers from village coopera-

tives, who attended classes at the thana training and develoPment

center.

The OPCOST equation, 5.12, does not estimate these training

and extension expenses.

“6 Md. I. H. Khan and others, Ag Evaluation Report pp_the Progrese

pfippe Seven Thana Projects Under phe Gomilla District Integpgted.Rural

DevelOpment Programme, Pakistan Academy for’Rural Development, Gomilla,

1967, Table 20.

“7 Only 14 cases could be included in the calculation of the

standard error of estimate for VLISM and VLRSM because of the lack of

reliable thana by thana data on loan issue and loan repayment in

1968-69 and in 1970.

The actual figures cited for these two variables

for these years were calculated by dividing aggregate figures,

culled from the eleventh annual report of the cooperative experiment,

by seven.

The entire series of actual figures for overdues and bad

debts, and for loan interest collected, was calculated in the same way.

The eleventh annual report does not distinguish between overdues and

bad debts; the report groups both into the category of "overdues."

Since the simulation separated out bad debts, these had to be added

back in to match the reported figures or overdues and on total village

loans outstanding.

Hence, the variable additions indicated in the

table by (TVLOAN + BDSM) and (OVRDUE + BDSM).

”8 The constancy of the relationship between outstanding loans

and village capital has been noted by cooperative officials in East

Pakistan and Bangladesh.

In the Eleventh Annual Report of the Rural

Cooperative Pilot Experiment, A. K. M. Obaidullah commented that the

percentage of outstanding loan to own capital was similar’across the

years from 1966-67 to 1969-70 in the seven expansion projects in

Gomilla district.

Akhter Hameed Khan also called attention to this

ratio in his report on the Tour 9; Twenty Thanae.

These commentators

did not recognize that what they thought of as a ratio relationship

could be more precisely explained through an equation including a

constant term.

345

49 This revision in the parameters of the model was statistically

defensible, since the average value of 4.46 lies within the 93% con-

fidence interval for £3', the coefficient in the loan issue equation.

The lower limit of the confidence interval is 4.03.

50 Probably because of the approximations involved in Eulerian

integration, the system trajectory actually comes to rest at the

stationary equilibrium points.

An analytic solution to the equations

of the system would show that the trajectory approaches the stationary

equilibrium, but does not ever reach it within any finite span of time.

51 The trajectories of the closed system and no-growth runs also

lie on top of one another in their final stages, but that effect is not

apparent in Figure 5—15, which does not show the trajectory of the

closed system run.

Such overlapping of trajectories is proscribed in

true phase planes.

A graph such as Figure 5-15 is not a true phase

plane.' In a phase plane, all system parameters and inputs are held

constant and the different trajectories show the effect of varying the

initial conditions of the system.

In Figure 5-15 the initial

conditions are held constant and the different trajectories show the

effects of varying the parameters and inputs.

An appropriate term

for such a graph might be a phase portrait.

That term was used in

labelling many of the figures in this chapter.

52 As with many another assertion in social science, this conclu-

sion holds under ceteris paribus conditions.

The optimum.rate of growth

might be different for other values of key parameters such as the

village loan interest rate (VLIR), or the proportion of overdue loans

which are written off as bad debts (PLR).

In the runs reported, VLIR

was .09 and PLR was .33.

An alteration in the equation relating RGROW

and PRLATE could also change the results.

53 In 1964-65 the Kotwali Thana Central COOperative Association

experimented with a much stricter policy.

Several primary societies

which had not repaid their loans were black-listed.

These black-

listed societies were denied further loans or tubewell water until they

repaid their debt to the central association.

Such a harshly punitive '

policy was not attempted again.

The experiment with black-listing was

never voluntarily mentioned by the cooperative officials in Gomilla.

Perhaps the black-listing policy was a crude progenitor of the super-

vised management policy, with its gradations of A, B and C class

societies.

5“ Akhter'Hameed Khan, Tour pf Twenty Thanas, Pakistan Academy for

Rural Development, Gomilla, 1971, p. 20.

55 The slope of this regression line for the early Kotwali Thana

data, 1.85, is markedly less than the slope of the equation TOTLON =

2.82 TSAV - 62.46, which was reported above to fit the average data

series for the 1966 projects with a coefficient of determination of

.999.

The implication of the difference in slope is that the loan issue

policies in Kotwali Thana prior to 1966 were stricter than the loan

issue policies in the expansion projects in Gomilla district after 1966.

346

Since the Agricultural Cooperatives Federation relaxed its own loan

limit policies after 1966, the leaders of the pilot cooperative project

in Kotwali thana were hardly in a position to reprimand the new project

directors for their laxity.

CHAPTER 6

Linkage to the Environment

Operating

Operational

Management

Strategic

Plant

Control

Control

Planning

Static Systems

S

Simple Machines

ST

Machines with

Input

STE

Organic Systems

Social organizations are flagrantly open

systems in that the input of energies and the

conversion of output into further energic

input consists of transactions between the

organization and its environment.1

Introduction

In an otherwise excellent text on a systems approach to organiza-

tions and management, Fremont Kast and James Rosenzweig refer to the

setting of an organization as the "environmental suprasystem."

They

do not say, though, why it is that the organizational environment

should, or always does, have systemic characteristics.

They unfor-

tunately use the term "suprasystem" as something of a trendy catchword.

347

348

In this susceptability to pretentious systems jargon they are certainly

not alone.

Many a social scientist these days will refer to phenomena

of interest as comprising a system, and will make no attempt to specify

the nature of that system in rigorous terms.

In previous chapters I have examined the rules governing established

relationships among the components of the structural and technical

subsystems of the Gomilla cooperatives.

In the case of the structural

subsystem, those rules were found to be described in organization

charts, and in the recursive equations of the structural control model.

Rules for the technical subsystem were enbodied in the procedures for

obtaining loans, and in the mathematical model of cooperative credit.

There are some precedents in the social sciences for describing

the rules governing an organizational environment.

Economists, for

example, have built theories describing how many firms in competition,

or a few firms in an oligopoly, will behave relative to one another.

But in organization theory, the field of interorganizational ayalysis

is conceptually much poorer.

In the absence of theoretically sound

guidelines for describing systemic characteristics of the environment

of the Gomilla cooperatives, I have foregone the attempt.

So in

contrast to previous chapters, this one does not pretend to be concerned

with describing an organizational subsystem.

What distinguishes machines with input from simple machines is

not the nature of the environment in which the system operates

(i.e. whether the environment can be characterized as a system), but

the existence of linkages which permit the system to receive inputs

from the environment.

Machines with input may also have linkages

through which they can make outputs to their environment.

The focus

of this chapter, accordingly, is on the linkages through which the

349

Camilla cooperatives interacted with other groups and organizations

in their environment.

Before the term "linkages" became popular among social scientists,

conduits of interaction between a system and its environment were

commonly described as "channels."

The reference, obviously, was to

communication theory.

One of the basic theorems in communications

is that the probability of a message being accurately received is

higher when sender and receiver are united by several communication

channels instead of one.

This theorem reappeared in Norman Uphoff

and Milton Esman's summary of the findings of the Cornell survey of

rural local governments in Asia.

"Local communities should be linked

to higher level decision centers by multiple channels," Uphoff and

Esman asserted, "both to achieve the benefits of specialization in

communication and to enjoy alternative avenues of influence.

One

composite channel appears quite inadequate, being liable to overload or

blockage or interruption if all village-center communication must be

carried by it."

This was the second of Uphoff and Esman's seven

2

prescriptions for effective rural development organization.

Relevant Research

In the years before he joined Normal Uphoff as the co-director of

the Cornell study of rural local government in Asia, Milton Esman was

the prime intellectual mover in the Inter-University Research Program

in Institution Building, a project which began in the late sixties and

incorporated the research of scholars at Michigan State University,

Syracuse University, Indiana University, and the University of

Pittsburgh.

Neither Esman nor most of the other participating researchers

350

was steeped in systems theory or in organization theory.

Working from

ill-defined premises, Esman patched together a schema for analyzing

the imperatives of institution building.

That schema was used by

other scholars in the Inter-University Research Program as the frame-

work for their case studies and deliberations.

A central feature of

Esman's conceptual schema was the importance of linkages between

nascent institutions and their environments.

Esman's basic presumption was that significant change in develop-

ing countries must be accomplished mainly through institutions.

As he

put it:

"Induced innovations require the vehicle of complex formal

organizations.

As societies modernize, the numbers of functionally

specific formal organizations multiplies and only through such

organizations and networks of organizations can complex modern

technologies be managed and integrated."3

While the language may be more pallid, the assertion that formal

organization is the vehicle of modernization is similar to V. I. Lenin's

contention that communist organization is the vehicle of the revolution.

The institution builders and the communist revolutionaries share,

moreover, an activist stance toward social problems.

Scholars of the

institution-building school refer to their doctrine as social engineer-

ing.

Instead of professional revolutionaries, the institution builders

speak of change agents.

In a passage which carries faint and distorted

echoes of Lenin's "What is to be Done?" Milton Esman wrote, "The I B

model is an elitist theory with an explicit social engineering bias.

Changes occur from the top down, not from the bottom up, and they are

guided by persons enjoying a measure of official authority or sanction.

The vehicle of-change is a formal and probably bureaucratic organiza-

tion."4

Change agents intent on social engineering, in other words,

351

are helpless without an organizational vehicle in which to propel

themselves and the society along.

The institution builders break with the Leninists at several

points.

In sketching those points, we can limn the borders of the

institution-builders' philosophy of social change.

First of all, the

institution builders perceive themselves as consultants to the national

authority structure.

They do not see themselves as revolutionaries,

or even as mild-mannered critics of existing social conditions.

Hardly

anywhere in their writings is their mention of class struggle.

In

this respect, the Inter-University Research Program in Institution

Building, which was funded in part by the U.S. Agency for International

Development (USAID).exhibits a characteristic deficiency which has been

described by Akhter Hameed Khan:

I would say that one of the main weaknesses of

the development programs sponsored by the AID

agencies for the excolonial countries is that

they obstinately refused to take into account

the class struggle.

And the host governments

were only too willing to go along.

Neither

agricultural extension nor community develop-

ment takes the class struggle into account, and

so these programs prescribe solutions which are

supposed to apply to everybody.5

To Akhter Hameed Khan's list of USAID programs which are blind

to social tension we may add institution building.

A second point at which institution building and Leninism

diverge is in the acceptance of the use of force.

Joseph Eaton, the

editor of the principal book on institution building, remarks,

"Members of the Inter-University Research Program in Institution

Building consortium accepted the viewpoint that planners could use

such powers as are inherent in the persuasiveness of data-related

35?-

reasoning, economic incentives, political support, and national service

ideals.

Planners who think that the long run objectives of their

effort justify the threat or use of banishment, imprisonment, and

terror will find that these methodological alternatives are not con-

sidered....

The conceptual framework that was adopted had the function

of facilitating comparative study of planful institution building

within the restraints of a democratic political philosophy."6

For the most part, I accept these restraints as well in this book,

just as Akhter Hameed Khan accepted them in building his rural develop-

ment institutions in Gomilla.

When he was at Michigan State University

in the late nineteen seventies, Akhter Hameed Khan was prone to look

back in disillusionment on some of the projects which he fostered in

Gomilla; he toyed more and more with the idea of forceful retaliation

against those who perverted the Gomilla cooperatives for selfish ends.

"There should have been a purge," he would mutter.

He said that he

wanted to carry the enforcement of cooperative debt collection to the

stage of people's courts, in which the small farmers in the villages

would take action into their own hands so as to force big farmers to

repay their cooperative loans.

But either because the political

situation or his own philosophy did not warrant it, he seldom ventured

in practice into the realm of brute enforcement.

The third major distinction between institution building and

Leninism has to do with the purpose of the program.

Lenin saw the

communist party as nothing but a means toward his constant goal of

revolution.

The institution builders, on the other hand, tend to

confuse the achievement of a state of institutionality with whatever

the social purpose of the institution building project was in the

first place.

To quote Esman:

353

The objective is to achieve institutionality -

 

meaning that innovative norms and.action patterns

are valued within the organization and by the

larger society and are incorporated into the

behavior of linked organizations and groups.

The environment becomes supportive of the in-

novations; the organization and the innovations

it represents become valued and meaningful

elements in the surrounding society.7

By the time he collaborated with Norman Uphoff in appraising

the findings of the massive Cornell study of rural local governments

in Asia, Milton Esman seemed to have overcome this confusion over

whether the aim in view was to construct organizations or to change

society.

The Uphoff and Esman report states explicitly that the

research task

was "to examine and assess the contribution of'lggal

organization to rural development....

"8

Specific evaluative criteria

for rural development are proposed in the report, and the effectiveness

of rural local organization is assessed in terms of its influence in the

accomplishment of those overall goals, such as increases in agricul-

tural productivity, redistribution of income, and the enhancement of

rural welfare.

Organization, in Uphoff and Esman's terms, is viewed

as an independent variable.

Of the major components of the original institution building

schema, the one which was carried forward most fully into the Uphoff

and Esman report was the concept of supportive organizational linkages.

As Melvin G. Blase described it, Esman's initial institution building

schema proposed that

Every institution is dependent upon other organ-

izations for its authority and resources; hence,

its linkages with other entities are Vitally

important.

These linkages also include an in-

stitution's dependency on complementary produce

tion of other institutions and on the ability

of the environment to use these resources.

354

Finally, linkages are also concerned with and

subject to the norms of the society.

Through

these linkages the institution maintains ex-

change relationships with its environment, an

interdependent complex of functionally related

organizations.

In the Uphoff and Esman report, the concept of linkage was

defined in operational, scalable terms.

Uphoff and Esman interpreted

linkage as meaning "the extent and effectiveness of communication and

influence between levels of organization (such as district and sub-

district). "10 They asked the members of their interdisciplinary working

group on rural local government to assign ordinal weights from O to

5 to the linkage upwards from and downwards to the different levels of

sub-national organization.

The weightings were discussed among members

of the working group, all of whom had expert knowledge of rural organi-

zations derived from their professional backgrounds and their conduct

of the case studies.

Discussion continued until consensus weightings

were arrived at for each case, and until there was agreement on weight-

ings for similar cases.

The case studies were then grouped into more-

organized and less-organized cases in accordance with their total

linkage weightings, plus another numerical estimate of the importance

of various channels for serving the functions of rural development.

The more organized cases, it turned out, had performed better on

almost every indicator of rural develoPment than had the less organized

cases.

Summarizing their findings, Uphoff and Esman wrote:

Organization for rural development must be seen

as a system of institutions performing various

functions in the rural sector of a particular

country.

We found no case where only one insti-

tution was carrying the full responsibility for

rural development or where complementarities

among institutions were not as important as

what the institutions themselves did.

Of key

355

significance was the extent and effectiveness of

linkages between and among the institutions,

horizontally with other organizations at the

same level and especially vertically between

local organizations and structures at the center

of government which set policy and allocate

resources essential to success in rural develop-

ment.

Local institutions which are separated and iso-

lated from the other levels are likely to be

impotent deve10pmentally.

Local autonomy by

itself provides little leverage for development.

What makes the most difference are systems or

networks of organization that make local develop-

ment more than an enclave phenomenon.

In this chapter, I have sought to apply these principles to the

analysis of the Gomilla cooperatives.

In the investigation of linkages

to the environment at the village and thana tiers, I have looked

especially at the patterns of horizontal and vertical ties between the

cooperatives and other groups.

Through these horizontal and vertical

linkages, the Gomilla cooperatives were joined to a social and organi-

zational network in the rural areas, and in the government of East

Pakistan.

Lgnkages With the Environment at the Village Tier

Within a village in Comilla the primary cooperative society did

not stand alone.

There was a preexisting social network -- a control

system of ancient origin -- with which the village cooperative might

compete, or, as Uphoff and Esman indicate is more desirable, with

which it could deve10p a complementary relationship.

It seems that in

most villages, and particularly in those where the most successful

cooperatives were to be found, the traditional social institutions and

the modern cooperative were mutually supportive, with the leaders of

356

each respecting the others' sphere of influence and authority.

Yet

quarreling and competition are always just beneath the surface in a

Bengali village, and in some instances this took the form of strife

between the leaders of the village cooperative and the leaders of the

traditional social institutions.

The traditional social network was imperfectly understood by the

westernized, often urban-reared faculty of the academy for rural

development.

The study of the clash and coalescence of traditionalism

and modernity in the villages, moreover, was not easily cast into the

survey methodology upon which most of the academy research publications

were based.

The standard technique for an academy research project was

for a faculty member to create a questionnaire, and then issue it to a

staff of village enumerators to take round to the villagers so as to

gather the requested data.

Questionnaire answers were then tabulated

in the academy research section, and the faculty member wrote his

report by examining and interpreting the tables.

This methodology

worked admirably for studies of the costs and returns of potatoes,

watermelons and other crops, but it was ill suited to describing, for

example, the rise and fall of a family's influence in a local coopera-

tive.

Beyond the family itself.the fundamental social grouping in the

Bengali villages was the Bari, a cluster of four to ten houses, together

with kitchens and cow sheds, surrounding a rectangular compound.

The

architecture of the houses was indicative of the wealth of the

occupants.

Richer villagers, belonging perhaps to the class-of

surplus farmers described in Chapter 1, would have houses of dried mud

covered with clay or cowdung.

The only ventilation and illumination

for the house, which might be 20 feet long by 10 feet wide, would be

357

provided by two small heavily shuttered windows.

Some of the bedrooms

in the huts would have no windows at all.

At the cost of what a

westerner would consider unbearable stuffiness in the tropical night,

the villagers thereby attained security from thieves who might try to

break into their house through the windows, and from enemies who might

hurl a spear into the house while they were sleeping.12 The poorer

villagers could not afford mud plaster houses, and slept in the breezy

insecurity of houses walled with woven bamboo.

The social significance of the compound of houses comprising a bag;

was lost on most of the men, whether Bengalis, Americans, or Europeans,

who have written about the social structure of Muslim Bengali villagers.

An American woman, Florence McCarthy, has portrayed the importance of

the bar; vividly.13 For the Gomilla peasant women, the bag; of their

husband in one village and the bar; of their father in another village

comprised the perimeter of their universe.

In accordance with the

Islamic custom of purdha, the village women were not to be seen by

strangers -- which in practice meant that they had to shield themselves

from the view of everyone beyond their husbands' close relatives in

the bag; compound.

Women were not even to befriend other women in the

village beyond their bari.

They did not go to market to sell the

produce of their husbands' farms or to shop for groceries or clothing;

all marketing was done by the men.

When they returned to their

fathers' bag; for a once or twice yearly visit, strictest custom called

for the women to be transported there at night.

Except in earliest girlhood, the female villagers were referred

to as "so-and-so's sister," or "so-and-so's mother," or simply as

"hey you" (ggHggu).

Many older women found it difficult to recall their

given names, since their individual identity had been ignored for so

358

long.

The greatest decision affecting a woman's life -— the choice of

her husband -- was entirely out of her hands.

Parents arranged

marriages for the girls when they attained puberty, usually selecting

a man about ten years older as the husband.

While the man's reaction

to the proposed bride was sometimes taken into account, the girl's

views of her prospective groom were not solicited.

The dearth of challenge and recognition in the women's world might

be expected to stunt their intellectual and social development, and so

it did.

Florence McCarthy found that she had to coax her women

interviewees along so as to teach them the simple skill of answering '

questions.

When the Academy for Rural Developnent, acting via the KTI'CA,

sought to train village women as organizers of a cooperative women's

program, similar difficulties were encountered.

The village women who

came to the Abhoy Ashram campus had to be taught, for example, how to

sit in chairs.

There was puzzlement at first as to why the village

women came to the weekly training classes so early, until it was noted

that the first to arrive could claim the corners of the classroom,

where they could hunker down, covering their faces with a veil and

hOping no one would speak with them.

In their training classes at Abhoy Ashram the peasant women were

taught practical skills, such as how to read and write, cultivate a

garden, spin cotton, or raise chickens.

On returning to their villages,

or more correctly to their Earls, the organizers taught other women

these same skills.

Despite their instructors' admonition that they

should involve women from throughout their village in the program, the

female organizers usually taught first the women from their own basis,

since the social injunction against befriending a woman in an adjacent

bari was stronger than the censure resulting from taking a trip six

miles or so to Abhoy Ashram.

359

Such censure was abundant.

The women who accepted the proferred

training at Abhoy Ashram were told that the thana training center was

part of the end of the age, that it would destroy the custom of

purdha_(reclusion of women) and even the Islamic relhgion.

The train-

ing center was a dangerous place where there were foreign gahibg.

The

women who travel there must be prostitutes.

Most intimidating of

all, neighbors told the women's program organizers that they would not

be buried until three or four days after they died.

In Islam it is

mandatory that the dead should be buried within 24 hours.

In linking the cooperatives to the emancipation of village women,

the leaders of the Gomilla rural development experiment precipitated

social turmoil.

In the short run this was a hinderance to the coopera-

tive movement.

The support of some men in the villages diminished

as a consequence of what they perceived as the radical implications of

the women's program.

Yet in the long run the cooperatives in the

villages where the women's program took root were strengthened by the

participation of the women.

With encouragement from the women's

program organizers, and from the less tradition-bound men who often

gravitated to the leadership positions in the local cooperatives, the

women joined the village cooperatives in rising numbers.

Like the

male cooperative members, the women made savings deposits and took out

loans.

By 1968, over'a thousand women in 67 villages had joined

cooperatives.

They had invested, on the average, a fifth as much as

the male members in the purchase of cooPerative shares and in savings

deposits, but their participation was still financially significant.

Over five hundred village women attended the annual cooperative rally

in 1969.

A village midwife chaired the rally.

360

Resistance to the women's program slowly eroded in the villages,

largely because of the perceptible impact it had on the financial

condition of the women involved, and on their families.

Florence McCarthy has described the process by which the program

brought the organizers out of utter dependency in their dark houses

and into the mainstream of the village economy.

A similar process

could be traced among the peasant women who were trained in turn by

these organizers.

Most of the organizers mentioned that support

for their work developed as people saw the "im-

provement in their condition."

This was

crucial particularly for the organizers.

Com-

ing from economic conditions involving great

need, and from generally lower social status

positions in the village, any change in the

conditions of these women was quite apparent.

At first many of the women saved part of their

travel allowance and with the extra money bought

more rice for the family.

Gradually through

persistent savings, they saved enough to buy a

goat.

From the sale of the kids of that one

goat, more food was purchased:

such items as

meat, fish, eggs and vegetables besides the

main staple of rice.

Chickens were acquired,

and from the sale of eggs and chicks, more goats

were purchased, or a cow.

Small gardens near

the bari were planted, or as women become more

ambitious, potatoes are grown in ricefields

lying fallow.

The income from the gardens or

potatoes goes for additional items for the

bari, usually food, then livestock, clothes,

and finally the entering of a child in school,

and the first down payment on land.

When a

woman, largely by her own actions, can bring

this much improvement to her family, it is no

surprise that opposition to her work diminishes

in the village.

This has been the experience

for all of the organizers, not just the excep-

tional few.14

It is very much to Akhter Hameed Khan's credit that he understood

the deadening injustice of the women's traditional status, and that he

gave his vigorous endorsement to the women's program.

Speaking in

Bengali in 1963, he said:

361

One cause of our misery and poverty is that we

keep our women-folk at home, guarded over con-

stantly.

We keep them indoors; We do not

educate them, and because they are confined they

cannot educate themselves: so they are nearly

all illiterate: they are timid.

And so long as

the women are uneducated, development can hardly

be expected in our country.

If the mother is illiterate, if she has no

courage, how can one expect courage in her

children?

The mother is the teacher....

I think our country can never progress until we

can emancipate the women.15

If the bari was the well-defined place of confinement of the women

in the traditional Muslim village in Bengal, there was no such clarity

in the social world of the unconfined men.

The world of the village

men has been characterized as diffuse and atomized.16

The village as a unit was not a coherent focus of identification.

Just what constitutes a village is a matter of contention, not only

for the "villager" but also for the government official or'anthropolo-

gical observer.

There are several words for village, and each corres-

ponds to a different grouping of people, houses, ponds and ricefields.

The Bengal plain is carved up into revenue villages (mauza), census

villages, and what Peter Bertocci has labelled "native villages"

(33am).

Whatever loyalty the residents felt toward their village

adhered to the gram.

There were #63 such gram in Gomilla Kotwali

Thana, averaging 135 acres in size and 340 persons in population.17

The ggam was the social unit within which the primary cooperatives

were organized, each cooperative taking the name of its ggam.

The

village cooperative was the sole salient institution, whether tradi-

tional or imposed, which encompassed one gram.

Family structures,

362

traditional hierarchies of social authority, religious congregations

and marketing patterns all were established without much attention to

village borders.

There was no village-level government, the lowest

level of local government being the union council, which was

responsible for an average of 36 villages.18

A significant social unit intermediate between the‘ggag and the

35;; was the neighborhood, or'paga,

Each pa;§_in a village was geo-

graphically distinctive, being set off from the other parts of the

village by paddy fields, bamboo groves, garden plots, or ponds (called

tanks by the English-speaking Bengalis).

In many villages local

politics could be sorted out as the struggle for influence and prestige

between the male residents of different REEEE-

An unusually tumultuous illustration of neighborhood politics

was the conflict over cooperative leadership between residents of the

north and the south pagag in JangLa village.

The Jangla Agricultural

Cooperative Society was established with its headquarters in the north

papa of the village.

However, within a few months of the cooperative's

registration two of its leaders arranged to shift the headquarters to

the southnggg, where they resided.

A dispute over this and other

matters arose.

The inspector from the thana cooperative association

arranged a public meeting in Jangla to resolve the issues.

During the

meeting some men in the audience jumped up and grabbed the records of

the village cooperative from the table in front of the inspector.

The

meeting became chaotic and was dispersed.

Leaders from the south E33;

filed a case with the police, who issued warrants against several of

the north pa;a_dissidents.

The case was settled out of court with the

north paga_residents agreeing to take back their savings deposits and

leave the cooperative.

The north para faction later sought to organize

363

their own cooperative society, but their*application was denied by the

central association, which maintained that only one cooperative should

be established in one village.19

Associated with the men's loyalty to their neighborhood, or‘pgrg,

was their loyalty to their class, or ggyai,

A large proportion of the

members of a given clan typically lived in the same pggg of the village.

The p§;a_may have had its origin, hundreds of years ago, as the locus of

settlement of a kinship group.

There remained some convergence, in

the modern village, between kinship loyalty and neighborhood loyalty.

At the head of each.;§yai_was an elderly man known as a gaggag;who

was one of the most important leaders of the traditional social system

in the Muslim Bengali villages.

The sardars of several adjacent

villages comprised a council, known as the gamaj.

The most important

function of the gagaj was to sit as a local court, to which the

villagers could bring disputes of all kinds.

The gag§j_also maintained

social discipline, offered advice on marriage proposals, and arranged

village rites and feasts.20

Such, at least, were among the ideal

functions of the gagaj,

In actuality, as Peter Bertocci discovered

when he took a close look at Hajipur and Tinpara villages in Gomilla

Kotwali Thana in 1967, the gamgj had lost much of its potency as an

instrument for maintaining social order.

During the year that Bertocci

lived in these villages, thquapaj met only once to mediate a dispute,

and this was dispite the prolific disputatiousness of the Hajipur

and Tinpara villagers, who were forever battling one another for the

crumbs of status or economic advancement that were to be found in that

overcrowded landscape.21

Like much of anthropological research, Peter Bertocci's study of

the two villages in which he resided for a year is at its best in the

364

telling of tales.

He recounts, for example, the story of Ershak Mia,

a young man who invited his fellow members of the Hajipur cooperative

to his wedding.

Upon hearing that people from outside of their

family lineage (Egyai) had been asked to the wedding, Ershak.Mia's

father was incensed.

The father asked for the advice of a counsin who

was a gagdégg a representative on the local ggmaj.

The unanimous

opinion of the members of the gémaj was that Ershak Mia should with-

draw those wedding invitations which had been sent to persons outside

his own lineage.

Chagrined but obedient, Ershak Mia retracted the

offensive wedding invitations, and thereby affronted most of the

members of the Hajipur cooperative.

In the next election for officers

of the cooperative, the candidates from Ershak Mia's lineage were

defeated. 22

The system of lineage leadership culminating in the gagaj was the

most important traditional social network with which the village

cooperatives were impelled to establish a relationship.

In Jangla

village, for example, the sardars on the §gg§j_enforced.a social disci-

pline of extreme conservatism.

The older men in the village had long

black beards.

There was one mgllah, or Islamic religious functionary,

for every 32 persons in the village.

Listening to radios was prohibited

in Jangla.

The leaders of the Jangla cooperative, who were younger

by a generation than the sardars, were often at odds with these

traditional social leaders.

The Opposite situation obtained in Monogram

village.

The two sardars in Monogram fully supported the modernizing

efforts of the cooperative, and sometimes applied their influence to

assure that cooperative loans were repaid.

There was common acknowledge-

ment of the responsibility of the sardars for the arrangement of

ceremonies and festivals and for the settlement of disputes, and of the

365

model farmer, for example, in matters of agricultural improvement.23

The Jangla cooperative, whose relationships with the surrounding

social system were full of friction, was among the least successful

in promoting economic improvement or in repaying loans.

The Menogram

cooperative, on the other hand, was often cited by officers of the

Agricultural Cooperatives Federation as one of the most progressive in

the thana, tripling crop yields in a decade and repaying all its loans

on time.

In addition to writing the case studies of Jangla and Monogram

cooperatives, two faculty at the Academy for Rural Development in

Comilla have attempted more systematic investigations of the relation-

ship between the cooperatives and the traditional social leadership in

the villages.

Unfortunately, both of these studies are flawed.

With the support of a Ford Foundation grant, Abdul Muyeed sampled

35 agricultural and nonagricultural cooperatives in Gomilla thana in

1968, and employed five enumerators and four research assistants for

six months in filling out and tabulating a questionnaire administered

to the manager and two members of each of these cooperatives.

Muyeed

obtained a 100% response rate from his sample of 35 x 3:: 105

villagers by employing the expedient of paying each of them five

rupees to come to the academy campus to be interviewed.

The researchers,

it seems, barely bestirred themselves to go into the villages.

Data

in hand, Muyeed returned to Michigan State University in 1969 to write

up his study as a doctoral dissertation?)+ The thesis was a methodolo-

gical disaster.

The bulk of it consists of a bombastic recitation of

the literature on economic and community development, and an uncon-

sequential presentation of a factor analysis of the relationships

among 90 of the 183 variables extracted from the questionnaire.

Simple

366

statistical techniques such as cross-tabulation, correlation, and

analysis of variance were ignored, much to the grief of anyone attempt-

ing to decipher the findings.

Seventy—seven of the respondents in Muyeed's survey said that

the traditional leaders were still active in their villages, and only

two said that they were not.

The Jangla situation of hostility between

the sardars and the cooperative officers was apparently uncommon.

Eleven of the respondents said that relations between the traditional

leaders and the COOperative were aggressive, while 62 said that the

relationship was very cordial.

One of the most interesting questions

which Muyeed posed -- "Has the cooperative given rise to a new power

structure in the village?" -- was phrased, sadly enough, in the jargon

of American social science, and was not explored in terms of the back-

grounds of office holders, or the play of influence in arriving at

important decisions.

Forty-nine of the 105 respondents in Muyeed's

sample said that a new power structure had been created, 55 said there

had been no change, and only one was undecided.

The respondents were

about evenly divided on this key question.

A much less expensive and less pretentious study of the relation—

ship between the leaders of the primary cooperative and the traditional

social leaders was conducted in 1971 by Badruddin Ahmed of the faculty

of the Bangladesh Academy for Rural Development.

Ahmed took the class

of traditional local leaders to include sardars as well as matabbars,

the latter being a non-hereditary and somewhat less prestigious

category of traditional leader.

He surveyed 250 farmers in 29 villages

in Gomilla Kotwali Thana, randomly selecting his samples of villages

and farmers.

He found that when it came to the selection of a pump

driver for an irrigation well, 45% of his respondents said that the

367

decision would be made by the general membership of the local coopera-

tive, 8% said that the decision would be made by the cooperative's

managing committee, h2% said it would be made by the chairman and

manager of the cooperative, and the rest, most of whom were not coopera—

tive members, didn't know how the decision would be made.

None of the

villagers said that the sardars and matab‘rars would participate in the

selection of the pump driver.

The survey revealed similar estimates

of how decisions would be made to purchase

rice mill shares, enroll

new members in the cooperative, receive cOOperative loans, distribute

the loans, or select a site for an irrigation pump.

The only coopera-

tive decision in which the villagers surmised that the traditional

leaders would play a part was whether the cooperative should participate

in the women's program.

Four percent of those surveyed thought that

that decision would be made by the cooperative chairman, the cooperative

manager, and the traditional leaders in concert, and another M% thought

the decision would be made by the traditional leaders alone?5

In addition to the insight which these responses give about the

minimal influence of the sardars and matabbars on most decisions con-

cerning the operation of the cooperatives, the answers also call

attention to the real participatory democracy in the local cooperatives.

Not only was the traditional elite excluded from vital decisions

affecting the COOperatives, but the decision making power of the

elected officers of the local cooperatives was restricted as well.

Across the seven indicator decisions pertaining to cooperative affairs

(selecting a pump driver, purchasing rice mill shares, enrolling new

members, etc.) in his survey, Ahmed found that an average of 46% of

the farmers surveyed thought that the general membership of the

cOOperatives would make the decisions.

368

A confusing aspect of Ahmed's monograph is that the influence of

sardars and matabbars who joined the cooperatives is not differentiated

from the influence of those who stayed aloof.

Almost half of the 125

sardars and matabbars in the 29 villages in Ahmed's sample had joined

a cooperative.

And approximately half of those who had joined a

cooperative had been elected to the managing committee of the organiza-

tion.

Eight of these had been chosen by their fellow directors as

cooperative chairmen, five as vice chairmen, and four'as managers.

Another 10, for a total of 27, had remained.as directors at large.

While the wisdom of co-opting the traditional leaders into the cOOpera-

tives seems to have been widely recognized, the sardars and matabbars

were generally shunted to the more ceremonial positions, such as chair-

man, rather than to the functionally critical position of manager.

It

seems that when Badruddin Ahmed reports that the traditional leadership

took no part in decisions such as the distribution of cooperative loans,

he was referring to the lack of influence of the traditional leaders

who had not joined the cooperatives and been elected to their managing

committees.

For four indicator decisions affecting village society (solution

of disputes, punishment of culprits, prevention of an epidemic, and

management of the maktab, or religious school), Ahmed reported a

pattern of influence which was the reverse of that for decisions

affecting the cooperatives.

A majority of the villagers thought that

these decisions would be made by the sardars and matabbars alone, while

a quarter thought that the traditional leaders would be joined by

the cooperative chairman and manager in making these decisions.

The

general membership of the local cooperative was expected to play

hardly any role in these affairs.

369

Good relations between the cooperative leaders and the traditional

leaders seemed to hinge on a mutual recognition of the sphere of

influence of the other group.

The division of responsibility was not

complete, however, since some of the sardars and matabbars had assumed

leading positions in the cooperatives, and since the cooperative

leaders were sometimes consulted about social decisions.

Badruddin

Ahmed commented that another reason for the lack of apparent conflict

between the two groups was that many of the kinsmen of the traditional

leaders had been elected to leadership positions in the cooperatives.

This observation squares with Peter Bertocci's findings in the villages

of Hajipur and Tinpara, where electoral politics in the primary coopera-

tives was largely a matter of competition among locally prominent

families.

So if new leaders were brought forth as a consequence of the

coming of the cooperatives, they may frequently have been younger

relatives of persons of established influence.

Like democratic insti-

tutions in most countries, the Gomilla cooperatives mostly reflected,

and only slightly reshaped, the status qu_.

The linkages at the village tier which have been discussed so

far -- linkages with the women's program, with the village neighbor-

hoods, and with the traditional social leadership -- were horizontal

in nature.

Collaboration with groups such as these in the villages

could be important to the success of the local cOOperatives, but was

not vital.

The linkage that was essential led upwards, to the thana

cooperative association.

Anwarul Hoque put the matter well in his

study of the Jangla cooperative:

All of the respondents maintained that it would

make a great difference if the village institu-

tion had not been a member of the Central Asso—

ciation for, in that case, the facilities

p r o v i d e d
b y i t c o u l d
T h e y a l s o m a i n t a i n e d
C e n t r a l
t i v e w o u l d

A s s o c i a t i o n

n o t s u r v i v e . . . .

3 7 0

n o t b e o b t a i n e d
t h a t i f r e l a t i o n s

e l s e w h e r e .
w i t h t h e

w e r e b r o k e n ,

t h e C o o p e r a -

g o v e r n m e n t

d e v e l o p m e n t

a g e n c i e s

w i t h t h e C e n t r a l

A s s o c i a t i o n .

w o r k
V i l l a g e r s

d i s t i n g u i s h

A s s o c i a t i o n

t h e m a s s e p a r a t e

s i n c e t h e c o o r d i n -

u s u a l l y

T h e o t h e r
a l o n g
c a n n o t
f r o m t h e C e n t r a l
a t i o n
s o c i e t i e s
m a i n l y

t h r o u g h

r e c e i v e

i s d o n e a t t h e t h a n a l e v e l .

s e r v i c e s
t h e C e n t r a l

T h e v i l l a g e
f r o m t h e s e a e n c i e s
A s s o c i a t i o n .

6

T h e p e r s o n a l

l i n k a g e s

e n u m e r a t e d

b y H o q u e i n c l u d e d

t h e c l a s s e s

a n d

m e e t i n g s

a t t h e A b h o y

A s h r a m

c a m p u s

w h i c h

t h e c h a i r m a n ,

m a n a g e r

a n d m o d e l

f a r m e r

o f t h e J a n g l a

c o o p e r a t i v e

a t t e n d e d

r o u t i n e l y .

A n i n s p e c t o r

a n d

a v i l l a g e

a c c o u n t a n t ,

b o t h e m p l o y e e s

o f t h e t h a n a c o O p e r a t i v e

a s s o c i a -

t i o n , v i s i t e d

t h e v i l l a g e

r e g u l a r l y .

S o f i v e p e r s o n s ,

t h r e e f r o m t h e

v i l l a g e

c o o p e r a t i v e

a n d t w o f r o m t h e t h a n a a s s o c i a t i o n ,

s e r v e d

a s

d e p e n d a b l e

l i n k s

b e t w e e n

t h e t w o t i e r s o f c o o p e r a t i v e

o r g a n i z a t i o n .

T h e y p r o v i d e d

t h e m u l t i p l e

c h a n n e l s

o f v e r t i c a l

c o m m u n i c a t i o n

w h i c h

U p h o f f

a n d E s m a n ,

i t w i l l b e r e c a l l e d ,

a s s e r t e d

w e r e c r i t i c a l

t o a n

e f f i c a c i o u s

r e l a t i o n s h i p .

O u t s i d e

o f t h e c o o p e r a t i v e

s t r u c t u r e ,

v e r t i c a l

l i n k s t o t h e v i l l a g e

w e r e f e e b l e .

W i t h i n

t h e B a s i c D e m o c r a c i e s

s y s t e m ,

t h e r e w e r e f i t f u l

a t t e m p t s

t o s t r e n g t h e n

t h e e l e c t o r a l

w a r d s

o f t h e u n i o n c o u n c i l s

a s

f U n d a m e n t a l

u n i t s

o f g o v e r n m e n t

w h i c h w o u l d

b e c l o s e r

t o t h e l e v e l a t

w h i c h

t h e v i l l a g e r s

a c t u a l l y

l i v e d .

I n s o m e i n s t a n c e s

t h e u n i o n

c o u n c i l

r e p r e s e n t a t i v e

f r o m t h e w a r d s e r v e d

a s a n e f f e c t i v e

l i n k t o

h i g h e r

l e v e l s

o f g o v e r n m e n t ,

b u t t h i s w a s a m a t t e r

o f t h e p e r s o n a l

s t y l e o f t h e r e p r e s e n t a t i v e

r a t h e r

t h a n a n e n f o r c e d

r e q u i r e m e n t .

P o l i t i c a l

p a r t i e s

d i d n o t r e a c h

d o w n t o t h e v i l l a g e

l e v e l .

T h e c o n g r e -

g a t i o n s

a n d p r i e s t h o o d

o f I s l a m

w e r e n o t l i n k e d

t o o n e a n o t h e r

i n a n y

formal hierarchy, such as is commonly found in Christian churches.

371

Summing up what has been said about the horizontal and vertical

linkages of the village cooperatives, I would assert that with the

partial exception of the link to the women horizontal relationships

within the village were a matter of happenstance (though usually

harmonious), and that vertical communication with the Thana Training

and Development Center was typically constricted to the one channel of

the cooperatives.

The insufficiency of formal linkages in both direc-

tions placed limits on the extent to which the villagers could be

mobilized for the purposes of rural development.

I make this assertion

not so much on the basis of the Cornell scriptures as on the basis of

my own exposure to another example of the use of organization for the

mobilization of the Asian masses.

When my year at Gomilla was finished in the summer of 1966, I

returned to the United States to study the language of a country where

the political and military mobilization of the rural populace had

become an issue of international significance.

I was a young man, and

the choice before me was evident:

either go to Vietnam as a foot

soldier, or go to Vietnam as a civilian.

I chose the latter.

For

four years I worked as a refugee and local government advisor in the

heart of the Mekong delta, as an advisor at a training camp on the

Vung Tau peninsula, and as a correspondent based in Saigon.

It has

taken me a decade to rearrange in a sensible order the memories of the

Pakistan Academy for Rural Development at Gomilla.

It will take a

lifetime to sort through the exhileration, the terror, and the beauty

of Vietnam.

The mechanism for mass mobilization which was employed by the

ultimately victorious communists in South Vietnam was the National

Liberation Front.

The idea and the name of such a liberation front

372

have been crudely copied by dissidents in many nations.

Nowhere has a

liberation front been so finely designed as was the NLF in rural

South Vietnam in the nineteen sixties.

The organizational experiment

conducted during the same decade in the villages of Comilla, a

thousand miles to the west, was a pallid venture by comparison.

When the great military force of the United States was thrown

against it, and when the American shield provided a space for non-

communist political and intelligence forces within Vietnamese society

to muster themselves against it, the National Liberation Front of

South Vietnam gradually unraveled.

The civil war was won by the

communists, in the end, through the might of their conventional army

and through the spiritless incompetence of the army serving the Thieu

regime.

I came to Vietnam in 1967, when the National Liberation Front

was past its heyday, and when it was preparing, in secret, to sacrifice

the last of its best in the spasm of the 1968 Tet offensive.

For a description of the NLF at its peak in about 1964 or 1965,

I have relied on a book which I read in 1966 in Honolulu, where I was

then studying the Vietnamese language.

Douglas Pike's Zigfi Qggg27

has been criticized as a mechanistic portrayal of the political

struggle that was waged in the villages of Vietnam.

But so also is

organization theory a mechanistic abstraction, so the reference seems

fitting for this volume.

There might, in an ideally organized Vietnamese village, be three

functional liberation associations linked together into the village

National Liberation Front Association.

These were:

the Farmers'

Liberation Association, the Women's Liberation Association, and the

Youth Liberation Association.

Representatives of local religious

groups and political parties would also be represented in the village

373

associations.

A central committee of the village National Liberation

Front Association controlled the constituent liberation associations,

and was in turn controlled, either overtly or covertly, by cadres

who were members of the Peoples' Revolutionary Party.

A platoon of

guerillas might also be organized in the village.

From this thumbmail sketch of the organization of the National

Liberation Front at the village tier one can properly infer that

horizontal linkages to other’groups in the village were more formal

and more fully developed in the NLF than they were in the Gomilla

cooperatives.

Like other totalitarian movements, the NLF sought the

total mobilization of every segment of society.

The NLF was distinctive

in that it initiated that mobilization at a very low level of the

organizational hierarchy, within the rural village.

Yet in the view of the leaders of the Vietnamese revolution the

elaborate NLF network at the village tier was insignificant in and of

itself.

The dedicated revolutionaries in the Peoples' Revolutionary

Party entertained no romanticism about village socialist republics

comparable to the mysticism about village panchayats in India.

The

liberation associations provided what Douglas Pike has called an

"organizational weapon" for the mobilization and control of the village

people.

The village level structure of the NLF was useful as a means

for providing access to the villagers by higher level officials and

military forces requiring the support of the village for the pursuit

of the larger revolutionary war.

Once horizontal linkages had been

established within a village, the vertical linkages were developed.

The vertical structure provided avenues of access to the village for

agit-prop teams, tax collectors, cadres, and military forces -- for

the whole panoply of revolutionary

government at the district and

371+

provincial tiers.

(The district and province in Vietnam are comparable

to the thana and district in East Pakistan and Bangladesh.)

Upon the

horizontal foundation of the village-level National Liberation Front,

a multi-channeled vertical structure was erected.

All of the channels

were controlled, to be sure, by the Peoples' Revolutionary Party.

That probably was not the model which Norman Thomas Uphoff, who is a

steadfast democratic pluralist, had in mind when he advocated a

multiplicity of vertical linkages in rural development organizations.

Yet it worked sufficiently well, after its fashion.

The relationship between the tiers of revolutionary organizations

in Vietnam should not be misconstrued as having been that of a hier-

archical bureaucratic structure of superiors and subordinates.

In the

revolutionary government of Vietnam there was great centralization of

policy making, with the crucial directives coming from the central

committee of the Peoples' Revolutionary Party or from COSVN, the

Central Office for South Vietnam.

This was accompanied by decentrali-

zation of implementation, so that units all the way down to the village

tier had great flexibility in interpreting policy to fit their cir-

cumstances.

There was a connection between events in South Vietnam and in

Gomilla in the late fifties and early sixties, but it was not obvious.

An apparent, but misleading, connection derived from Michigan State

University's involvement in both places.

From an acquaintance which

an assistant professor of political science, Wesley Fishel, struck

up in Tokyo in 1950 with a Vietnamese exile, Ngo Dinh Diem, there grew,

after Diem became President of South Vietnam, the largest foreign

technical assistance program undertaken in that era by any American

university.

The MSU group in Vietnam included, at its height, 51

375

members appointed by the University's Board of Trustees and 151 Viet-

namese and other locally hired employees.28 The MSU advisory project

in Vietnam focused on strengthening public administration and police

services.

The project was sponsored jointly by the Government of the

Republic of Vietnam and by the United States Operations Mission in

Vietnam.

MSU was phasing out of Vietnam by 1961, and.at the same

time was becoming heavily involved in the Ford Foundation funded

project to assist the rural development academies at Peshawar'and at

Gomilla.

There was no transfer of personnel from the one project to

the other, however.

The MSU faculty who went to Vietnam came mainly

from the departments of political science, criminal justice, and

economics, whereas those involved in the Pakistan project held appoint—

ments in education, agricultural economics, and sociology.

The two

faculty empires for providing foreign assistance did not overlap.

One of the men who was appointed as MSU senioradvisor at Gomilla,

Nicolaas Luykx, had been to Vietnam before he came to East Pakistan, but

in the capacity of a graduate student doing field research for his

doctoral thesis at Cornell University,29 not as a faculty member at

MSU.

The connection between the turmoil in South Vietnam and the

rural development schemes in Comilla was not one of personnel going

from one place to the other.

It was, rather, a bond of anti-communist

ideology which was shared by most Americans engaged in foreign assis-

tance work in that era.

The best exemplar of that ideology

was one of the first advisors to the

Gomilla project.

Henry w. Fairchild.

He was stationed at

Gomilla as the senior advisor for four crucial years, from 1959 to

1963.

Before coming to Comilla, he had worked as a rural development

3 7 6

a d v i s o r

i n S o u t h

K o r e a .

m e s s a i n i c

A m e r i c a n

F a i r c h i l d
a n t i - c o m m u n i s t s

w a s o n e o f a g e n e r a t i o n

o f

w h o s a w t h e i r

m i s s i o n

a s a r r a n g i n g

a n a l t e r n a t i v e
t u n a t e l y . . . i n t e r n a t i o n a l

t o r a d i c a l

r e v o l u t i o n

i n p o s t - c o l o n i a l

A s i a .

" U n f o r -

c o m m u n i s m

h a s n o c o m p u n c t i o n s

a t a l l a b o u t

a s s i s t i n g

o r a g i t a t i n g

i n a n a t i o n

o u t s i d e

o f e s t a b l i s h e d

g o v e r n m e n t

c h a n n e l s , "

F a i r c h i l d

w r o t e .

" T h e r e a l p r o b l e m

f a c e d

b y t h e U n i t e d

S t a t e s

G o v e r n m e n t

a n d t h e r e s p o n s i b l e

l o c a l

p o l i t i c i a n s

o f t h e u n d e r -

d e v e l o p e d

n a t i o n

i s t h i s .

I s t h e r e a n y b e t t e r

a l t e r n a t i v e

f o r ' a p e o p l e

i n t h i s c h a o t i c

s t a t e t h a n I n t e r n a t i o n a l

C o m m u n i s m ?

A s I h a v e t h o u g h t

a b o u t

t h e p r o b l e m

I h a v e c o m e t o t h e c o n c l u s i o n

t h a t t h e r e

i s , b u t

t i m e i s a g a i n s t

i t a n d I n t e r n a t i o n a l

C o m m u n i s m

h a s a w e l l t e s t e d

s y s t e m .

N o n e t h e l e s s

t h e r e

i s s t i l l

a c h a n c e

a n d t h a t c h a n c e

s h o u l d

n o t

b e l o s t . " 3 0

F a i r c h i l d

h a s r e c o r d e d

a t o u c h i n g

c o n v e r s a t i o n

w h i c h

h e h a d w i t h

A l i A k b a r ,

h i s f r i e n d

f r o m t h e G o m i l l a

v i l l a g e

o f L a m p u r .

A l i A k b a r

h a d l e a r n e d

E n g l i s h

w h i l e

s e r v i n g

t h e B r i t i s h

m i l i t a r y

i n C a l c u t t a

a n d

i n G o m i l l a ,

w h e r e a m a j o r

B r i t i s h

h e a d q u a r t e r s

h a d b e e n l o c a t e d

i n

W o r l d

W a r I I .

O n e i m a g i n e s

H e n r y

F a i r c h i l d

a n d A l i A k b a r

s i t t i n g

u n d e r

a c o c o a n u t

t r e e i n L a m p u r .

A f t e r

c o n v e r s i n g

a b o u t

c r o p s

a n d c o o P e r a -

t i v e s ,

F a i r c h i l d

f i n a l l y

a s k s h i s w i z e n e d ,

E n g l i s h

s p e a k i n g

f r i e n d

a b o u t

t h e s p o o k i s h

b e i n g s

w h i c h

i n h a b i t

t h e i n n e r

r e g i o n s

o f h i s m i n d .

I a s k e d h i m o n c e i f h e h a d e v e r h e a r d o f c o m -
m u n i s m .

Y e s , h e s a i d , h e h a d h e a r d o f i t i n B u r m a .

" A r e t h e r e

a n y c o m m u n i s t s

h e r e ? "

I a s k e d .

"Not in this village," he said, "but there

are some around. "

I asked him what he thought communism stood

for.

It stood for everyone being equal, ran

377

the answer.

"What is being equal?" I asked.

He said, "The communists say this means tak-

ing land from the bigger farmers and giving

it to those with no land."

"Do you want this?" I asked.

"I need more land," he said.

I asked him, "If you take land from the bigger

farmers and give it to the ones without land

would you get any land?"

"No," he said, "not in this village.

There

are about thirty families without land and

only four or five families who have more

than five acres."

I asked him then, "Do you know anything else

about communism?"

"No," he said.

"There was a man in Burma

when I was there.

He was a communist.

Every-

body said he was crazy.

He was always trying

to talk to us."

"What do other people of Lampur think of

communi em?" I asked.

"Moulana Ali Ashraf says its against Allah."

"Would you be against Allah?" I asked.

"No," he said, "I am a believer."31

Linkages With the Environment BE the Thana_Tier

3 7 8

The basic weakness of the pre-existant governmental structure in

East Pakistan, in Akhter Hameed Khan's view, was its incapacity for

productive collaboration with the villagers.

The civil service worked

efficiently at the provincial and district tiers, but between the

district and village tiers good government plans were lost in imple-

mentation.

"Good plans and.poor implementation really indicate the

absence of some links in the administrative chain," Akhter Hameed Khan

observed.32

"It was this weak link with the village which had to be

forged.

This new approach sets up powerful institutions at the thana

level to support the village."

The concept of a Thana Training and.Development Center, the

aggregation of institutions which the rural development academy

advocated establishing at the thana level, was introduced to the reader

in Chapters 1 and 2.

To review:

its major components were a thana

central COOperative association, a thana council, which was a part of

the structure of Basic Democracies, and representatives of the nation

building departments such as the Thana Agricultural Officer and the Thana

Irrigation Officer.

The complex was to include facilities for training

379

many kinds of villagers, such as model farmers, managers and chairmen

from the primary cooperatives, village accountants, midwives, and

school teachers.

Horizontal linkages of the thana cooperative associa-

tion were to be established mainly within the framework of the Thana

Training and Development Center (TTDC).

Through the TTDC the thana

cooperative association could call upon the Thana Agricultural Officer,

for example, for assistance in training model farmers.

Or, to take

another example, the TTDC could facilitate the coordination of plans

for low lift pump or tubewell irrigation among the thana cooperative

association, the Thana Irrigation Officer, and the rural public works

program conducted by the thana and union councils.

The departments, councils, and associations which were to be

coordinated within the Thana Training and Development Center did not

readily collaborate with one another.

"Activating the departmental

officers," A. K. M. Mohsen remarked, "seemed to be more difficult than

mobilizing the people."33 .As late as 1969, when the Thana Training

and Development Center at Gomilla had been active for six years,

there were reports of slack attendance by departmental officers in the

meetings of the Kotwali Thana Council.

Leadership of the institutional

complex was a source of continuing disputes.

The project director

of the thana cooperative association and the Circle Officer of the

thana (the lowest level officer in the system of general administration

for the thanas, districts, and provinces of Pakistan) jockeyed for

recognition as the commanding officer of the organizational complex.

Not until 1970 was that contention officially resolved through the

designation of the Circle Officer as the leader of the Thana Training

and Development Center.

His authority as the chief administrative

officer of the thana was not automatically accepted.

The provincial

380

Department of Agriculture, in particular, resisted the subordination

of their staff to the Circle Officer.

Senior officials in the Depart-

ment of Agriculture were unsympathetic to the concept of a Thana

Training and Development center, and sought to expand, instead, the

standard type of centrally controlled extension system.

Ironically enough, the original model of an agricultural extension

system was the Cooperative Extension Service of the first land grant

college in the United States, Michigan State University.

That model

had been accepted as dogma by the midwestern schools which dominated

American agricultural assistance to the less developed countries.

American advisors and foreign graduates from these schools preached

the dogma of extension, which fitted well with the paternalistic

orientation of government leaders in post-colonial countries.3u So

when the MSU advisors in Comilla joined with the faculty of the Academy

for Rural Development in advocating the training of model farmers as a

more appropriate method of transmitting information to the farmers,

they were doing battle with the ghost of their own university's

success.

In the opinion of the Michigan State faculty

who were associated

with the Gomilla project, the linkages which were most important to

its eventual fruition were those which led from Gomilla upwards to

the power centers of the provincial and national governments.

Professor Cole Brembeck, who was a member of the team of MSU faculty

who recommended Akhter Hameed Khan for the directorship of the new

academy in Gomilla, recalls that Akhter Hameed Khan "had a peculiar

power over the ministers of the national government.

He would ring

them up and say,

' That piece of paper is on your desk, is it not?

We need your signature by 4 o'clock this afternoon.’ and they would

381

sign.

The national leaders were afraid that Akhter Hameed Khan would

enter national politics, and that he would be a winner.

They were

just as happy to keep him in office in Gomilla rather than frustrate

him into becoming active in national politics."35

Despite the academy director's prestige as a popular figure and

as a former member of the revered elite of the Indian Civil Service,

there were many occasions when his personal and ideological enemies

nearly succeeded in conspiring to halt the experiment in Gomilla.

Richard O. Niehoff, a professor of education who was the chief MSU

advisor for the Pakistan academies from 1962 to 1972, remembers the

Deviationism Affair as "absolutely the most critical point in the

evolution of the Gomilla project."36’ The matter came to a climax at

the first meeting of the Joint Board of Governors of the Pakistan

Academies for Village Development.

To comprehend the origins of the

dispute which was resolved at that meeting, one must probe some of the

very early history of the Gomilla project.

According to a mimeographed pamphlet dated March, 1960, which was

found in a box of Pakistan Project files in the MSU archives, "The

beginnings of the academies can be traced to a conference on

14 October 1955 between Mr. Chaudhury Mohammad Ali, then Prime

Minister of Pakistan, and Dr. George F. Cant, then representative of

the Ford Foundation in Pakistan.

The prime minister expressed

particular interest in being advised as to ways in which officials and

employees of the Government could be trained or retrained in concepts

and techniques of modern administration in accordance with the aims

of the government -- with particular reference to nation-building

activities. "3 7

Two years after this conversation between Chaudhury and Gant took

382

place, the scheme for the academies which in the meantime had been

deveIOped, with Ford.Foundation support, by Dr. Floyd Reeves and his

team of Michigan State faculty was submitted to the Pakistan cabinet

for its approval.

In July of 1957 the cabinet passed a resolution

endorsing the scheme for the academies.

The dispute which was to

baloon into the Deviationism Affair centered on the issue of whether

the Gomilla academy had adhered to or deviated from the intent of

the July 1957 cabinet resolution.

Several charges were made, such as

that the academy was not training as many government officials as

originally had been promised, and that the academy was overspending.

The nub of the dispute, though, was that the Gomilla Academy, under

the leadership of Akhter Hameed Khan, was looking with a critical eye

at the effectiveness of the government establishment in the field of

rural development, and was seeking to invent new ways of solving rural

deve10pment problems.

The irreverence of this attitude did not sit

well with several high-level Pakistani officials.

Richard Niehoff recalls that the leader of the clique of civil

servants who were attempting to chastize the Comilla academy was

Shafi Niaz, chief of the agriculture section of the Pakistan Planning

Commission.

One of Niaz's favorite complaints was that the Gomilla

academy was starting its own dairy, poultry farm, and agricultural

demonstration plots, rather than relying on the demonstrations con-

ducted by the technical departments of the Pakistan government.

Niehoff

prepared a voluminous report to defend the rural deve10pment experiments

underway at the Comilla academy, in preparation for the decisive first

meeting of the Joint Board of Governors for the Gomilla and Peshawar

(East and West Pakistan) academies which was held in the committee

room of the Pakistan secretariat, building number one, in Rawalpindi

383

on October 29, 1960.

Niehoff and Akhter Hameed Khan attended the

meeting, which was chaired by Lt. General W. A. Burki, Pakistan

Minister of Health, Labor, and Social Welfare.

After hearing some

of the charges of deviationism at that meeting, Akhter Hameed Khan

retorted angrily that if the Government of Pakistan was not prepared

to support the sort of experiment that was underway in Comilla, then

they could have his resignation.

"I will give you gentlemen an annual

report, and if at any time you disagree with it, you can have my

resignation," Khan said.

General Burki, who had been listening silently to the argument,

intervened at this point.

If he had not done so, Richard Niehoff

believes that the whole Gomilla experiment would have collapsed at

that very early stage.

-Burki said firmly, "If this is deviation, we

have to have more of it."

Burki told everyone present that the issue

was closed, and that they should proceed with other business.

Niehoff and Burki became good friends thereafter.

On each of his

many subsequent trips to Pakistan, Niehoff tried to make a point of

calling on General Burki.

The general confided that he had been primed

before the Rawalpindi meeting by the chief secretary of East Pakistan,

who had sent him a note ahead of time saying that the new institution

in Gomilla was serving the needs of East Pakistan very well, thank

you, and please don't change it. 38

The sociology of the Deviationism Affair is intriguing.

At the

very top levels of the Pakistan government, among ministers such as

General Burki and senior civil servants such as the chief secretary

in Decca, the Gomilla experiment was strongly supported.

The most

influential supporter of all was the president of Pakistan, General

Mohammad Ayub Khan.

But among civil servants who were close to the

384

center of power, but who were not members of the innermost circle,

resentment of the Comilla academy and of its stiff-necked director,

Akhter Hameed Khan, was often the predominant emotion.

The nature of the relationship between President Ayub Khan and

Akhter Hameed Khan is subject to dispute.

Lawrence Ziring, in his book

on the Ayub era, goes so far as to characterize Akhter’Hameed Khan as

"President Ayub's closest advisor on village affairs and rural

development."39

When I asked Akhter Hameed.Khan, in early 1977, for

his assessment of Ziring's statement, he replied, "That is rubbish!

I always kept my distance from Ayub Khan's regime.

I consider Ayub

Khan a good man, but I was always aware of his weaknesses.

He was a

kind of Colonel Blimp.

Once I suggested to him that he should come to

East Pakistan during the dry season and go to one of the rural works

construction sites, to identify himself with the laborers.

He asked

me, 'But aren't these village pe0ple aware already that I am doing

all this for them?‘

I told him that is not the way things go in

politics.

Ayub Khan asked me at one point to come to Rawalpindi

and be his organizer for village affairs, but I refused." Akhter

Hameed Khan disagreed with Ayub Khan's concept of using the union

council members as an electoral college, but it seems that he did not

make that disagreement known to the president.

Certainly the two men had a lot in common.

They were both Pathans.

They were of similar'age, Ayub Khan being seven years older.

Both

had trained in England for service in the British imperial regime:

Akhter Hameed Khan in Cambridge University in preparation for the

Indian Civil Service and Ayub Khan at the Royal Military College at

Sandhurst prior to his service in the British army in India.

They

both had held positions of significant responsibility under the

385

British:

Ayub Khan as the commander of an army battalion in World War

II, and Akhter Hameed Khan as a sub-divisional and debt settlement

officer in Bengal in the nineteen thirties.

They were, in essence,

members of the same first generation of national leaders who had

learned the skills of administration under British tutelage.

Akhter

Hameed Khan seems to have exploited this similarity in order to gain

access to President Ayub Khan, thereby assuring top-level support for

the experiments in Gomilla.

The best that can be said for the

reciprocal relaticnship is that Akhter Hameed Khan made.Ayub Khan's

regime look good because it permitted a project which the villagers

perceived as worthwhile to go forward.

Access to the higher levels of government in Pakistan was facili-

tated by Akhter Hameed Khan's status and connections, but another

parallel channel was sometimes employed, and in the interest of

historical accuracy deserves recognition.

In Pakistan in the nineteen

sixties there was a well established network of American advisors,

some in the MSU Pakistan Project field staff, some in the Harvard

Advisory Group, some directly connected to the Ford Foundation, and

some in the U.S. Agency for International Development.

These American

advisors had counterparts and connections in the ministries and planning

commissions of the central government and of the provinces of East and

West Pakistan.

Richard Patten, for example, was a member of the

Harvard Advisory Group who was very well connected with the Department

of Basic Democracies and Local Government in Dacca, and with its

energetic and imaginative secretary, Musa Ahmed.

Patten's intervention

on behalf of the rural development concepts emerging from the Gomilla

experiment was crucial in several instances, as in the formulation of

the Rural Public Works Program, and in the campaign for dry season

irrigation with low lift pumps.

386

Richard Niehoff, the director of the MSU Pakistan project, had

his home base in Pakistan in Rawalpindi, in West Pakistan.

He campaigned

vigorously to get the approval of the Pakistan government for’the

expansion of the Gomilla project from a simple training academy to an

experiment in rural development encompassing Comilla Kotwali Thana.

In

the course of that campaign, he recalled that he made so many trips over

to Dacca that the government officials in the Dacca office of the Central

Secretary of Agriculture came to call him Mr. Gomilla.

Finally he said,

"Look, I will just get a sleeping bag and some K rations and camp out

here until this thing is signed."

The document authorizing the Academy

to take control of Kotwali Thana was signed, due largely to the combined

persuasion of Richard Niehoff and Akhter Hameed Khan.

The accomplishment which Niehoff recalls as being the most signi-

ficant, especially in terms of the potential for replicating the rural

development institutions in Gomilla in other countries, was the formation

of the Board of Governors of the Academy for Rural Development.

Neihoff

laid much of the foundation for the establishment of the Board, and wrote

some of the government documents from which it took its charge.

He

argued that the academy should be accountable to its clients, the

ministries which were to send their personnel to Gomilla for training

in the techniques of rural development.

Accordingly, the membership of

the academy's Board of Governors came to consist mainly of representa-

tives of these client ministries.

Because it was controlled by its

clients, via the Board of Governors, the Gomilla academy attained a

degree of institutional acceptance within the Government of East Pakistan.

The government came to embrace, one might say, its own gadfly.

387

The link to the Ford.Foundation and its financial support provided

the academy with much greater flexibility in experimentation than it

would have had in dealing with the Pakistan government alone.

In addi-

tion to the funds which it channeled to Michigan State University and to

the Academy for Rural Development, the Ford Foundation provided a substan-

tial grant of $810,000 to the Government of Pakistan for the agricultural

modernization experiments being conducted by the Kbtwali Thana Central

Cooperative Association; much of this was to cover the foreign exchange

costs of importing tractors and pumps, and furnishing them with spare parts.

The national politics of Pakistan and the provincial politics of

East Pakistan had a definite influence on the course of events in

Gomilla.

In an unusually candid interview in MaICh of 1975, Akhter

Hameed Khan discussed the political winds which had blown upon him

during his tenure as the leader of the Gomilla project.

I have

reconstructed his comments from my notes-

"The businessmen and the contractors were the financiers of the

Muslim League.

They financed almost an annual campaign against the

388

cooperatives in Comilla.

They organized demonstrations against me

from 1965 onwards.

Ayub Khan himself told me that the rightists had

brought two deputations against me.

"The Muslim League leaders in the town of Gomilla were the

truckers and the contractors.

Those were the most affluent groups.

They used to use the butchers in Gomilla as their muscle men, because

the butchers were accustomed to bloodshed and could intimidate people

by walking in front of a procession waving their knives.

Then I

organized the butchers into their own cooperative.

The butchers were

pacified: they no longer helped to terrorize the people.

So the Muslim

League leaders opposed me.

They also resented my having organized the

rickshaw drivers, who used to be the mainstay of their political

demonstrations.

Once they joined the cooperatives, the rickshaw

drivers lost interest in the sham trade unions that had been set up for

them by the Muslim League.

"The Maoist underground moved against me in 1967.

They came to

Gomilla to urge the people to burn the irrigation pumps and tubewells.

They attacked me as a non-Bengali, and threatened violence against me.

But they couldn't make any progress in the villages at all.

The only

toehold they got was in the machine section of the KTTCA, where my

organization was captured by the leftists.

They agitated for the sheer

joy of destruction.

Parts were taken, and I finally had to dismiss

several of the tractor station employees.

"The Awami League people also decided that they would destroy

the Comilla project.

They joined hands with the registrar of coopera-

tives in Dacca to oppose our cooperatives.

When I was in Princeton

in 1969, it came out in the Awami League newspaper that I had stolen

20 million rupees and had bought a hotel in Switzerland.

When I

389

returned from the United States I toured all the village cooperatives

from the first of February to the first of May.

I wrote a letter to

the papers stating that the charges against me were false, and that I

wanted to live in peace.

There was a rally in Gomilla of about 50,000

people in support of me.

When the Awami Leaguers went out to the

villages they learned that the line of attacking me and the cooperatives

was not popular.

In July of 1970 the Awami Leaguers switched their

position, and thereafter they gave me their support.

"The disintegration of government became severe in the late

sixties, during the long fall of Ayub Khan and afterwards.

It got to

the point that there was no government.

The police, for example,

would give us absolutely no help.

When I fired the employees from

the tractor station they came around to the creamery section and

threatened to burn it.

The creamery and cold storage staff asked me

what they should do.

I said we should not beg the police for help

because the police were aware of the problem and did nothing about it.

So I advised them that the next time those ruffians came around they

should defend themselves.

They did.

They took iron reinforcing

rods and threatened to beat the fellows up.

After that there was no

more trouble.

"During the early years we could impose discipline in the coopera-

tives because there was a government.

Ayub Khan wanted this work done,

so the administration supported us.

After 1967 there was no govern-

ment.

From 1968 on all the politicians were saying that the loans

issued by the cooperatives should not be repaid.

"It has been said that the Comilla program was not political, and

mostly that is true.

But to keep it alive I kept in touch with the

national government, particularly under Ayub Khan, and with the

390

Opposition brain trusters Of the Awami League.

And if that is not

politics I don't know what is.

When the Awami League came into power

in the new government Of Bangladesh, their rural development program

was essentially the Comilla program."

Notes for Chapter 6

Linkages to the Environment

1 Daniel Katz and Robert L. Kahn, The Socig; Psychology Q§_Q£g§gr

izations, John Wiley and Sons, New York, 1966, p. 1?.

Reprinted by

permission of John Wiley and Sons.

2 Norman T. Uphoff and Milton J. Esman, Local Organization :25

Rural Development:

Apalysis.pf_ppg.Asian Egperience, Rural DevelOpment

Committee, Cornell University, 1974, p. 100.

3 Milton J. Esman, "The Elements of Institution-Building," in

Joseph W. Eaton, ed., Institution Building gpd Development, From Con-

cepts pg_Applicatiop§, Sage Publications, Beverly Hills, 1972, p. 25.

“ Ibid., p. 26.

5 Lecture notes. Agricultural Economics 865, Michigan State

University, 1975.

6 Joseph W. Eaton, "Institution Building as Planned Change," in

Joseph W. Eaton, pp pip, p. 13.

Reprinted with permission.

7 Esman, 9p pip, p. 25.

8 Uphoff and Esman, pp.2ip, p. 3.

Italics in the original.

9 Melvin G. Blase, Institution Buildin :

A_Source Book, Midwest

Universities Consortium, East Lansing, 1973, pp. 6, 7.

Reprinted with

permission.

1° Uphoff and Esman, pp_p;p, p. 111.

11 Ibid., pp. x1, XII.

12 Ramakrishna Mukherjee, Six Villgges pf Bengal, Pbpular Prakoshan,

Bombay, 1971. P- 31-

13 Florence E. McCarthy, nggali Village Women:

Medigtors Between

Tradition gag Development, Masters Thesis, Michigan State University, 1967.

1“ Ibis. pp. 132. 133.

15 Akhter Hameed Khan, The Role of Women in a Country's Develop-

ment, Pakistan Academy for Rural Development, Gomilla, 1963, pp. 2, 3.

391

392

16

Elliot L. Tepper, "The Administration of Rural Reform:

Structural Constraints and Political Dilemmas," in Robert D. Stevens

and Peter J. Bertocci, eds., Rural Development pp Bangladesh and

Pakistan, The University Press of Hawaii, Honolulu, 1976, p. 32.

17 Peter J. Bertocci, "Social Organization and Agricultural

Development in Bangladesh," in Stevens and Bertocci, pp_p;p, p. 161.

The figures are for 1963.

18 The absence of a village focus was more marked in Bengal than

in other regions of the Indian sub-continent.

The reason for this, it

has been suggested, was that the principle of individual land owner-

ship was entrenched in Bengal.

In some other parts of ancient India

village lands were held in common, and therefore had to be managed

in common.

(See S. A. Qadir, Village Danishawar, Pakistan Academy for

Rural Development, Gomilla, 1960, p. 21:)

19 Anwarul Hoque, "Cooperation under Extreme Traditionalism and

Unfavorable Physical Conditions:

A Study of a Cooperative in an East

Pakistan Village," in Inayatullah, ed., Cooperapaves and Planned

Change ;p_A§ian Rugal,Communitie§, United Nations Research Institute

for Social Development, Geneva, 1970, p. 67.

20 Abdul Muyeed, Strategies Evolved la a Development System pf

Planned.§gcial Changa pp Rural East Pakistan:

A Stugy pimppg Process

pfilnstitution-Building and its Integration pp pap Politico -

Administrative Structure, Ph.D. Thesis, Michigan State University,

T969. P- 38-

21 Peter J. Bertocci, Elusive Villages:

Social Structure and

Community Organization _ip Rural East Pakistan, Ph.D. Thesis, Michigan

State University, 1970, p. 150.

22 Ibid., PP. 159' 160.

23 M. Nurul Huq, Village Development pp Bangladesh (A_Study p:

Monogpaa Village), Bangladesh Academy for Rural DevelOpment, Gomilla,

1973. P- 92-

24

Abdul Muyeed, pp gap.

25 Badruddin Ahmed, Leadership pp Villaga Cooperatives, Bangladesh

Academy for Rural Development, Gomilla, 1972, P. 15.

26 Anwarul Hoque, pp gap, pp. 81, 82.

27 Douglas Pike, Viet Cong, The M.I.T. Press, Cambridge, Massa-

chusetts, 1966.

28 Guy H. Fox, Final Report Covering Activities pf the Michigan

State University Vietnam Advisopy Grogp for the Period May pp, 1955 -

June 39, 1262, MSU Group, Saigon, 1967, p. 1.

393

29 Nicholaas G. M. Luykx, III, Some Comparative A§pects pf Rural

Public Institutions pp Thailand, pap Philippines, apg Vietnam, Ph.D.

Thesis, Cornell University, 19 2.

-

30 Henry W. Fairchild, "Stopover at Comilla, 1959-1963’"

unpublished manuscript in the Asia Collection of the MSU Library.

31 Ibid.

32 Akhter Hameed Khan, Community apg Agricultural DevelOpaent pa

Pakistan, Asian Studies Center, Michigan State University, 1969, p. 30.

33 A. K. M. Mohsen, Report pp a Rural Public Works Prograp,

Pakistan Academy fOr Rural Development, Gomilla, 1962, p. 18.

34 Akhter Hameed Khan, Transfer 9_I_‘_ the Americaa Mgel p: g.gr_icul-

tural Exten§ion pp_Ex-Colonial Countriga, Rural Development Project,

Michigan State University, 1978.

35 Conversation with the author, 1977.

36 Conversation with the author, 1976.

37 "Summary Of Essential Background Facts on the Pakistan Academies

for Village Development as of 21 March 1960," authorship not noted.

38 The note from the Chief Secretary had been drafted, at his

request, by Akhter Hameed Khan and Richard Niehoff.

39 Lawrence Ziring, The Ayub Khan Epa:

Politics pp Pakistan,

1258-1962, Syracuse University Press, Syracuse, New York, 1971, p. 156.

CHAPTER 7

The Psychosocial Subsystem

Operating

Operational

Management

Strategic

Plant

Control

Control

Planning

 

Static structures

Simple machines

S

ST

................ L - - - - _ - _ - - - - - i - - - - - -

Machines with input

Organic systems

 

STE

STEP

 

 

 

Social structures are essentially contrived sys-

tems.

They are made by men and are imperfect

systems.

They can come apart at the seams over-

night, but they can also out last by centuries

the biological organisms which originally created

them.

The cement which holds them together is

essentially psychological rather than biological.

Social systems are anchored in the attitudes,

perceptions, beliefs, motivations, habits and

expectations of human beings.

Such systems

represent patterns of relationships in which the

constancy Of the individual units involved in

the relationships can be very low.

An organiza-

tion can have a very high rate of turnover Of

personnel and still persist.

The relationships

Of items rather than the items themselves provide

the constancy.

394

Introduction

395

In the third chapter, a visual paradigm for organizational analysis

was presented as Figure 3-16.

Since the diagram incorporated the

contents Of two figures in Kast and Rosenzweig's text, it was rather

complex.

TO simplify interpretation Of the paradigm, an exploded view

Of it was also displayed.

That view is reproduced in Figure 7—1.

The labelling of three of the subsystems in Figure 7-1 as nuclear,

and Of three as ortdtal subsystems should not be taken tOO seriously.

The analogy with the structure Of an atom, which has several particles

clumped together at the center and has shells Of orbiting electrons, was

made to give visual emphasis to the idea that the three most critical

subsystems in an organization are the structural subsystem, the

technical subsystem, and the psychosocial subsystem.

Together with the

pattern Of linkages with the environment (including input and output

flows), these three subsystems define the essential Operating character-

istics of the organization.

Another heuristic virtue Of the paradigm shown in Figure 7-1 is that

it calls attention to the areas in which the three nuclear subsystems

overlap.

In chapter five, accordingly, the interaction of the technical

and the structural subsystems was examined.

In this chapter, which

treats the psychosocial subsystem, the interdependency of the psycho-

social subsystem, the structural subsystem, and the technical subsystem

will be investigated.

The central region in which these three sub-

systems converge has been labelled the operating plant in Figure 7—1.

From the interaction of these three subsystems in an organization's

operating plant, the most essential qualities Of the organization

emerge.

396

£7.64 756/6

-

-

MAW/N6

Jun: ’6' row:

 

I
MAMAGEM cwr \

3229399,: \

 

I

 

i

ope-RA 770M“

cameo; \\

Juslfirew:

\e

'

 

i

emu: rug/:4

JUJSflfE/‘A\

 

b

 

:

 

I

 

I

A

 

 

 

I

 

I
I

 

”0.2517714”

Jabs/”em:

\

 

J

 
.
n

 

l

l R

 
 

$ ”Mt/c4 5A.: ”

30506794:

 

J

PsycHosocmL

.7063YJ7'EM

7£C#N/CAL

305.97.375.21

OPERA TIA/6

PLANT

F/GURE 7- /

exp; 00:0 V/e'u/ or 7796' OR GAIN/2A T/ OM44

mun/6M.

397

The concept Of a psychosocial subsystem advanced by Fremont Kast

and James Rosenzweig was also discussed in the third chapter, and a

substitute definition of the psychosocial subsystem was proposed.

The

which sus_t§_i_a functional 93133 i_n_ the workiag relation§h_i_p_s_. _O_f_‘ 3:113

orgapization's participanta.

Although this concept was touted as being

more precise than Kast and Rosenzweig's, it does require clarification.

The first question that needs to be answered before this definition

can be applied is:

What is functional order?

A method of addressing

that question can be found in Figure 7-1.

The psychosocial subsystem

contributes to the maintenance of’order in:

1) the structural subsystem,

2) the technical subsystem, and 3) the internal consequences of an

organization's linkages with its environment.

The structural and

technical subsystems of an organization are bonded to the human psyche

by a configuration of status and role relationships.

These are part of

the organization's psychOSOCial subsystem.

Fremont Kast and James Rosenzweig, who invented the term "psycho-

social subsystem," do not explicitly consider the interaction between

the psychosocial subsystem and the other subsystems in an organization.

In a few passages, such as those quOted below, they demonstrate a

flickering awareness of the interdependence Of the psychosocial subsystem

and other subsystems in an organization, but they do not develop the

argument in a formal way.

IndiViduals in social relationships constitute

the psychosocial subsystem in organizations.

The .

general "atmosphere" is affected by many variables,

some integral, others peripheral.

Societal cul-

ture sets an overall framework; industry mores

and practices have an impact; and many variables

are peculiar to specific organizations.

Tech-

nology and structure affect organizational

398

climate, as do employee attitudes and morale.

Status systems serve to structure social relation-

ships and provide a framework within which group

endeavor can be coordinated toward Objectives-

Role systems are integrally related with status

systems.

Status concerns the relative prestige

Of a position in a structural relationship within

organizations: role relates to the behavior

patterns identified or expected for a given

position.2

There was a brief mention in the third chapter of the

restrictiveness of the definition of the psychosocial subsystem adopted

for the purposes of this volume.

The maneuvering and.obstructionism

which so Often are to be found in organizations did not seem to have

a place in the functional concept of a psychosocial subsystem.

Posi-

tive attitudes which reinforce and stabilize the structural and technical

subsystems could be treated within the context of the proposed definition.

Yet obtuse or self-serving attitudes, though they might be prevalent

among the organization's members, would seem not to be a part of the

psychosocial subsystem.

Political struggle within the organization

might also seem out Of place.

Very few organization theorists are well schooled in political

science.

The tumult Of politics is beyond their recognition.

Yet

politics is an aspect of personal interaction which can be Of over-

whelming importance for a rural deve10pment organization, as Uphoff

and Esman indicate in two Of their seven propositions about the

desiderata for successful rural development.

issue whenever raised....

More equitable distribution appears to be

a necessary if not sufficient condition for extensive rural develop-

ment, though enclave development can occur in its absence."

- "Pblitics, the competition among groups for influence and

399

resources, must be accepted as inevitable and legitimate in rural local

organization.

We must recognize that there will always be politics

and that it can distort and corrupt instead Of reconcile and uplift.

The basic objective should be that all_people be included in a share

of benefits....

What must be avoided is the politics Of exclusion or

monopoly, where practices of patronage out certain persons or groups

Off from the benefits of government.

This is devastating to an

organization's legitimacy and ultimately to its effectiveness."3

Are these political factors, since they may serve to disrupt

rather than to stabilize the structural and technical subsystems, to

be considered as outside Of the compass of the psychosocial subsystem?

If no way were to be found to accomodate the review of these forces,

then the analytic paradigm presented in this book would have to be

judged deficient.

A study of the physiology of the human body wOUld be incomplete if

the disruptions of discare were ignored.

The analogy is not entirely

fair, since Uphoff and Esman pointed out that politics can be just as much

a positive as a negative influence in rural development organizations.

All politics should not be equated with organizational disease.

An

inclusive political process may, in fact, contribute significantly to

the legitimacy of the organization, reinforcing the credibility and

status of its leaders, and thereby strengthening the structural sub-

system.

Relevant Research

400

To prepare a theoretical foundation for the consideration of the

psychosocial subsystem in the Gomilla cooperatives, strands of research

must be brought together from the fields of institution-building,

political science, and non-formal education.

Because of the intellectual

walls which usually divide these disciplines from one another, their

union may seem strange, perhaps forced.

The reader's patience is

solicited.

-

The InterbUniversity Research Program in Institution Building was

introduced in the last chapter.

There the notion of linkages, which is

401

one aspect Of the institution building perspective, proved helpful

in explaining the relationship of the Gomilla cooperatives to their

environment.

The overall institution-building schema has been

summarized by Milton Esman, chief theoretician of the inter-university

research program, in the diagram shown as Figure 7-2.

 

 

Institution

 

Linkages

Institution variables

leadership

doctrine

program

Transactions

Normative linkages

Enabling linkages

Functional linkages

 

“

"’ Diffused linkages

resources

 
    

internal structure  

 

Figure 7—2:

The Institution-Building Universe.t¥

Scholars in the institution-building research program were dis-

appointed by the inutility of their fourfold taxonomy of linkages, so

the discussion in Chapter 6 disregarded the distinctions displayed on

the right hand side of Figure 7-2 among enabling, functional, normative

and diffused linkages.

A simple distinction between horizontal and

vertical linkages was employed instead.

The discussion in this chapter focuses on the contents Of the left

hand side Of Figure 7-2.

The five significant components Of an

institution are there described as leadership, doctrine, program,

resources, and internal structure.

At least one of these phrases,

internal structure, should seem familiar to the reader.

Chapter 4 Of

this book dealt with the structural subsystem of the Gomilla coopera—

tives.

The combination of Esman's two variables Of program and resources

can be compared to the technical subsystem of Chapter 5.

Another

association can be made between leadership and the management

402

subsystems which will be treated in Chapter 8.

From these comparisons,

it is simple to make the inference that doctrine is the subject matter

of this chapter.'5

Of all the institutional attributes highlighted by the institution-

building model, Milton Esman thought that doctrine was the most elusive

and one Of the most important.

In a retrospective essay he commented,

"The original institution building definition Of doctrine as the

'specification of values, Objectives, and Operational methods...the

stable reference point...‘ has been proved by several researchers to be

too static and too monolithic....

Institutional doctrine is a combina-

tion of these themes which can be and.are manipulated by institutional

leadership to enhance internal cohesion, and to make it more acceptable

in the external environment.

Management Of values is not a trivial

function of leadership, especially in an organization committed to

innovative purposes in an environment where neither the organization

nor*its innovations have yet been accepted."6

There is considerable overlap, evidently, between the attribute

of doctrine, which Esman contends can "enhance internal cohesion," and

the concept of a psychosocial subsystem, which was defined above as

sustaining "a functional order in the working relationships Of an

organization's participants."

Are the concepts identical, or does one

contain some elements not found in the other?

If the two concepts

were really identical, why should we not relabel the psychosocial

subsystem the "doctrinal subsystem," and be done with it?

In their own paradigm for organizational analysis, Fremont Kast

and James Rosenzweig include what they call a "goals and values sub-

system."

Values and doctrine are such similar constructs that Esman

treats them as synonymous.

In the discussion Of Kast and Rosenzweig's

403

paradigm in the third chapter, I argued that the goals and values sub-

system could be abolished as an analytic category;

As described by

Kast and Rosenzweig, it was not a subsystem.

values could be addressed,

I contended, in the context of the psychosocial subsystem.

And it is

here, indeed, that we have come across them under the rubric Of doctrine.

The addition of doctrine, or organizational values, to the psycho-

social subsystem gives no cause for concluding that doctrine comprises

the sole contents of the subsystem.

All is still present that resided

in the psychosocial subsystem before the addition Of the subject Of

doctrine.

Status and role relationships, mentioned earlier, are good

examples Of other factors which belong in the general category Of the

psychosocial subsystem.

Yet there is no heading in the institution

building paradigm under which status and role relationships can be

discussed.

This logic demonstrates that the institution building

paradigm displayed in Figure 7-2 is less comprehensive than the para-

digm shown as Figure 7—1.

Since Milton Esman admits that he found the concept of doctrine

to be elusive, I will turn to two political scientists, Gabriel Almond

and Sidney Verba, who have executed a brilliant empirical study of the

related construct of political culture.

In their classic book, Tag

Civic Culture, Almond and Verba make use of the concept of culture in

the sense Of "a psychological orientation toward social-Objects.“7

The political culture, they say, consists of "attitudes toward the

political system and its various parts, and attitudes toward the role

of the self in the system."8 This is a more diffuse and fundamental

concept than that of political doctrine, which would consist Of ideol-

ogical beliefs about the political system, founded upon the attitudes

incorporated in the political culture.

404

Previous to Almond and Verba's monumental research project,

which was conducted in 1959-60, many investigations had been under-

taken Of the influence of family patterns and child rearing practices

upon national politics.

Most such studies, such as those which linked

the authoritarian German family to the fragility Of democracy in the

Weimar republic, had been centered on one country.

Almond and Verba

compared the political cultures of five countries -- the United States,

Great Britain, Germany, Italy and Mexico -- on the basis of 1,000 inter-

views in each country.

From such a huge data base, any quick-minded

social scientist could derive a host of conclusions.

The main thing

that saved Almond and Verba's book from the appearance of a statistical

digest is that they did not analyze their data by computer, theirs

being one of the last great examples of survey research in the pre-

computer era;9

The wealth of findings reported in the book is not

susceptible to a quick summary, so I shall forgo the attempt.

For our purposes, the most useful chapter in Almond and Verba's

book is the last, which contains no tables at all.

In that summary

chapter, the authors describe the sort of political culture which they

found, from their studies of the five nations, to be most suitable to

a stable democracy.

They described the civic culture as a mixed

culture, in which

There is political activity, but not so much as

to destroy governmental authority; there is in-

volvement and commitment, but they are moderated:

there is political cleavage, but it is held in

check.

Above all, the political orientations

that make up the civic culture are closely re-

lated to general social and interpersonal orien-

tations.

Within the civic culture the norms of

interpersonal relationships, of general trust

and confidence in ones social environment, pfne-

trate political attitudes and temper them.1

405

Trust and confidence.

Those are not the terms in which the

typical Bengali village has been described.

Atomized and suspicious

would be closer to the mark.

Muslim Bengali women, as we have seen,

are confined by tradition to their pappa, and must overcome powerful

social injunctions if they seek to build trustful relationships with

women from other quarters Of their village.

The men congregate in their

mosques and in their clans, but they do not have strong communal

loyalties.

Yet if attitudes of trust and confidence are fundamental

to the civic culture, and if the civic culture is the basis of demo-

cracy, how could it be expected that a democratic organization such as

a COOperative would thrive in rural East Pakistan or in Bangladesh?

Part of the answer; as we shall see in the next chapter, is that

some of the most successful village cooperatives were more authoritarian

than democratic.

Still, the official management structure Of the

village cooperatives was democratic in form, having elected Officers

and relying on frequent participation by the membership in decision

making.

If the attitudes which were appropriate to this form of organ-

ization were not present to begin with among the villagers, could it be

that the cooperatives in some manner instilled them?

Almond and Verba ponder the same sort of issues in the last pages

of their book.

They wonder how democracy can be expected to take root

in the developing countries, where the prevalent political culture is

typically very different from that found in the United States and in

Great Britain.

The sense of involvement with the democratic political

system, and of commitment to it, is often weak in such countries.

Political cleavages predominate over political bonds.

The atmosphere

is not right for democracy.

How, then, can the civic culture be

created, or approximated, out of alienation and hostility?

Almond and

406

Verba's response is brief, and not well thought out.

They suggest,

in the next to the last paragraph Of their book, that in the developing

countries education and industrialization will enhance the attributes

which make up the civic culture.

This was not a consistent position for Almond and Verba to assume,

since in earlier passages they make it clear that the civic culture

has not been taught in the schools in the United States.

The political

culture which has been advocated in high school civics classes is a

rational-activist culture, in which the citizen is expected to be for-

ever calculating his advantages and involving himself in all aspects

of the political process.

In the civic culture, by contrast, the citizen

believes he is capable of influencing the government, but is actually

passive most Of the time.

The possitivity of the citizenry allows the

official elites the latitude necessary to govern.

The civic culture

is learned in the United States, Almond and Verba contend, through

informal out-Of-school observation Of the political behavior of family

members, friends, and respected acquaintances.

In the early nineteen sixties, when Almond and Verba wrote their

book, education was still usually equated with schooling.

Many

deficiencies of this equation have since come to the fore, such as:

that it neglects the learning potential Of adults, that it is too

expensive as a means Of teaching all of the burgeoning population of

developing countries, and that it biases education toward a narrow band

of academic subjects.

Enlightened policy makers have since come to

recognize that education which takes place out of school and which is

more closely focused on specific skills, from literacy to line-sowing,

is more suited to many of the needs of developing countries.

Non-

fbrmal education is the label which is generally applied to such modes

of instruction.

407

The chain of thought which has been developed above leads to the

conclusion that non-formal education is the technique Of preference

for inculcating the attitudes which would bind a rural development

organization together.

The skills required for effective participation

in and management of a rural development organization could also be

imparted through non—formal education.

These conclusions may seem

Obvious here, but I have been amazed to discover that they are almost

entirely absent from the literature on non-formal education, apg from

the literature on institution-building.

Some examples of this oversight

are cited below.

The late Frederick H. Harbison of Princeton University was one of

the most distinguished advocates Of the importance Of non-formal

education in developing countries.

In his book.prap_Resources aa gag

Egalpp 9: Nations, Harbison advocated an alternative to the Gross

National Product, or GNP, measure of national economic growth.

The GNP

approach measures progress in terms Of the production of goods and

services.

In order tO accelerate such production, economic development

policies based on the GNP approach may favor capital-intensive modern

economic enclaves, usually in urban areas, at the cost of neglecting

the persistent unemployment of the mass of the country's labor force.

The human resources approach, instead, holds that the ultimate basis

Of wealth is the realization of human potential rather than the accumu-

lation of material Objects.

Under a development policy guided by the

precepts of the human resources approach, the employment and improve-

ment of the entire labor force, including the residents of rural areas,

becomes a central goal.

Both formal and nonfOrmal education contribute

to the employability, enrichment, and productivity of the labor force.

408

Harbison admits that organization is the most perplexing of all

human resources problems.

"Organization is required to mobilize the

energies of the labor force.

It is necessary to create a state, to

build economic and political institutions, to propagate ideologies,

and thus to carry forward every aspect of national development....

In the human resources perspective, organization-building is the most

critical of all tasks for national development.

Good organization makes

it possible to maximize employment and learning Opportunities: poor

organization can perpetuate the underdevelopment and underutilization

Of the capacities of man."11

The proposition is grand, but the sequel

is incredibly feeble.

In the three sentences which follow the ones just

quoted, Harbison says that nobody knows how to design organizations

that are suitable to the needs Of developing countries, or to produce

the organization-builders who will construct them.

In sad contrast

to his populist faith in the productive potential of the labor force

in other spheres, Harbison implies that organization building is an

elite activity.

He expresses the hope that organization-builders will

arise out of the welter of learning Opportunities provided by a human

resources apprdach to development.

The issue is dropped there, and never

raised again in his book.

Nonformal educational endeavors in rural areas of the developing

countries are uncoordinated, unevaluated, and inefficient.

Such is

the complaint Of Philip Coombs and Manzoor Ahmed, authors of a World

Bank sponsored study on the potential contribution Of nonformal educa-

tion to the alleviation Of rural poverty.

They say that what has been

missing is a unifying doctrine, "an all encompassing vision Of what

the basic social mission of agricultural education and research should

be and Of what different elements, in conjunction with one another,

409

would be required in order to accomplish that mission."12 How might

such a doctrine be elaborated and.imparted?

Coombs and Ahmed's

suggestion is a pallid one, but characteristic of their professional

outlook as international consultants.

Each developing country should

"undertake a comprehensive and dispassionate appraisal of its agricul-

tural knowledge system.

This effort should be led by one or more

broad gauged senior analysts who have no axe to grind either for or

against any particular component, and should involve the participation

and close cooperation of all interested government organizations."13

Imagine the benighted fisheries officer and plant protection officer

in some remote riceland seat of government seeking mutual inspiration

from the doctrine resulting from this dispassionate, bureaucratic

review!

As a part of their investigation of rural local government in

Asia, the Rural Development Committee at Cornell commissioned a study

of training and research related to rural development.

The report of

the working group on Rural DevelOpment Training and Research makes no

mention of the role of training as an instrument for institution

building.

The strategy Of extended rural development, the report

asserts, "very puch gapends pppp_ap aapanded and well-articulated s s-

pap 9; organization down pg and including the local level."mu

But even

in the passages of the report that deal with institution-building, the

contribution which training can make to the crystallization of such

local organizations is neglected.

In 1976 the U.S. Agency for International Development underwrote

a conference on Non Formal Education and the Rural Poor which was held

at Michigan State University.

Syed A. Rahim Of the East-West Communica-

tion Institute, in Honolulu, presented a paper at the conference on the

410

non-formal education aspects Of the Gomilla project.

Rahim describes

the many training sessions for model farmers, village COOperative

managers, and others that were held at the Thana Training and Develop-

ment Center at Abhoy Ashram in Gomilla.

He also mentions the in-service

training of government Officials conducted at the Academy fOr Rural

Development campus at Kotbari, a few miles away.

The purpose Of all

this training, as Rahim describes it, was to encourage acceptance of

proposed innovations.

Rahim's major example of an innovation is tech-

nical, the cultivation of rice according to the methods advocated by

the Japanese agricultural experts at Comilla.

Rahim showed no perception

of the contribution of training to obtaining acceptance of organizational

innovations.15

The scholars who participated in the InterbUniversity Research

Program in Institution Building were caught up in the exposition and

critique of their analytic paradigm, and gave out very little advice

as to how someone engaged in the task should go about building an

institution.

Despite their academic origins, these pundits did not

recognize the significance of learning in building institutions.

Yet it

is inescapably apparent, from deductive logic alone, that roles, status

and doctrine are learned.

The people who participate in organizations

may change from year to year, but there is a discernable constancy in

the way the participants view roles, status, and doctrine in the organ—

ization.

They must have learned these things.

And where there is

learning there must be education, even if it is the sort of informal

education which comes from Observing and conversing with other workers

on the job.16

For models of how education can be mobilized in the service of

institution-building let us turn to the revolutionary periods in China

411

and Vietnam.

In the nineteen thirties in China and in the nineteen

sixties in Vietnam revolutionary organizations of all sorts - armies,

parties, peasant associations -- were constructed by communist insure

gents out of the same sort Of quarrelsome, tradition-bound, and seem-

ingly intractable people as must form the basic material Of rural

development organizations around the world.

These revolutionary

organizations proved to be supple and powerful.

They were built, in

large part, through non-formal education.

First an example Of fairly fermal education for revolutionary

institution-building.

After the Long March which took them 6,000 miles

around the periphery of China, past the snowy borders of Tibet, and to

the security of a new base in the empoverished northern province of

Shensi, the Chinese communists set about rebuilding their forces in

preparation for a protracted war.

Mao Tse Tung himself had admitted

that in the southern base area which the reds had been forced to abandon

insufficient

time had been devoted to training the communist army.

He

had Observed in 1928, six years before the start of the Long March,

that "The average soldier needs six months' or a year's training before

he can fight, but our soldiers, though recruited only yesterday, have

to fight today with practically no training to speak of.

Exceedingly

poor in military technique, they fight by courage alone."17 The

mistake Of neglecting training was not to be repeated in the Shensi

base area.

In 1936 a young.American reporter, Edgar Snow, penetrated to the

rebel stronghold in Shensi and spent four months there interviewing

the communist leaders and soldiers.

He visited the Red Army Univer-

sity, which was led by a brilliant 28 year Old army commander named

Lin Piao, who was reputed never to have lost a battle.

With some

412

modifications, the Red.Army University was recognizable as an Officers

training academy such as all modern armies maintain.

Classes of the

Red Army University were held in caves, and fOr'paper students used

the blank sides of nationalist propaganda leaflets.

For every two years

of active service, an army commander or commissar was required to

spend four months in training.

This was a higher ratio of training

time to service time than is to be found in most armies, certainly more

than would be the case in other’armies at war.

The content of the

training, also, was distinctive.

Political and doctrinal matters were

given equal emphasis with military skills, as Edgar Snow discovered to

his embarassment when he Offered to answer questions from cadets on

current affairs.

He was showered with pointed, knowledgeable inquiries

such as:

"What is the future Of the international student movement, which

has its center in Paris?"

"In your Opinion, can Leith-Ross's visit to Japan result in

Anglo-Japanese agreement on policies toward China?"18

Classroom training was not the only, or the most important, method

of education employed in the Chinese red army.

The most common and

potent method was education by example.

This was deemed especially

prOper in the teaching of social status relationships, which were

intended to be egalitarian rather than hierarchical.

In a report to

the central committee of the Chinese Community Party, Mao Tse Tung

wrote:

All alike share the same hardships; everybody

from the army commander down to the cook lives

on a daily fare worth five cents, apart from

grain.

In the matter Of pocket money, if two

dimes are alloted, it is two dimes for everybody;

if four dimes are alloted, it is four dimes for

413

everybody.

Apart from the role played by the Party, the

reason why the Red Army can sustain itself with-

out collapse in spite of such a poor standard Of

material life and such incessant engagement is

its practice of democracy.

The officers do not

beat the men: officers and men receive equal

treatment; soldiers enjoy freedom of assembly

and speech; cumbersome formalities and ceremonies

are done away with: and the account books are

open to all....

The newly captured soldiers in

particular feel that our'army and the Kuomintang's

army are worlds apart.

They feel that, though in

material life they are worse Off in the Red army

than in the White army, spiritually they are

liberated.19

These two methods Of teaching -- didactically and by example --

were combined and exploited to an unprecedented degree by the insurgents

in south Vietnam in the nineteen sixties.

In the secure base areas an

extraordinary number Of people were trained for’all sorts of roles --

as officers, as medics, as nurses, as political cadre.

When I was

stationed in Sadec, in the middle of the Mekong delta, in 1967 and 1968

I would talk now and then to the revolutionary defectors, or fig; Chanh,

at the provincial reeducation center.

It was astonishing how many Of

them had travelled to the mangrove swamps of Camau, at the tip of the

delta, to attend training sessions.

In the locale where the guerilla

war was being conducted, moreover, the insurgent troops would spend

most of their time training, planning, preparing themselves for battle,

conducting propaganda among the villagers, and farming.

Douglas Pike

contends that in the period of 1962-63, when the intensity of the

guerilla war was only moderate, a typical Main Force unit would spend

only one day a month in combat.20

In organizational sophistication, the communist revolutionaries

in Vietnam surpassed by far their mentors in China.

For Mao Tse Tung,

414

the primary instrument of revolution was the red army, which he maneuvered

brilliantly around the huge territory Of China.

Mao had no exact

counterpart in Vietnam, since as an intellectual and a theoretician he

stood head and shoulders above the Vietnamese communists.

The

Vietnamese political leader most analogous to Mao Tse Tung was

Ho Chi Minh.

Ho was once asked why he didn't write books in the manner

of Mao Tse Tung.

He laughed and retorted, "If there is a subject

Chairman Mao hasn't written about, tell me and I'll try to fill in the

gap. .21

The gap that Ho Chi Minh filled in actuality was one that he did

not expound upon in writing.

Ho was a master inventor of front organi-

zations.

The Viet Minh, under whose banner the Vietnamese fought for

their independence from the French, was the most renowned front organi-

zation founded by HO Chi Minh.

Parties as well were maneuvered by him

like front organizations.

He would found them, infiltrate them, dis-

band them, and change their names, always with the aim Of promoting

revolution, independence, and national unification.

For Mao Tse Tung

the revolution was war; for Ho Chi Minh it was a political struggle

conducted by parties and front organizations, and assisted when need be

by military force.

The Leninist doctrine Of the organizational weapon

was clearly imprinted on HO Chi Minh, and only vaguely perceptible in

Mao Tse Tung.22 HO Chi Minh refined the organizational weapon.

Lenin

had wielded it like a cudgel.

Ho Chi Minh parried and thrust with it

like a rapier.

It is likely that HO Chi Minh had a hand in the formation of the

South Vietnamese National Liberation Front.

The plan for creating such

a front was first announced in North Vietnam at the third congress of

the Lao Dong (workers) party in September Of 1960.

The National

 

415

Liberation Front was Officially fOrmed two months later, in the south.

For the purposes of this book, the significance of the National

Liberation Front lay in its deployment as an institution-building

instrument.

Through the NLF, the rural people of South Vietnam were

cajoled, persuaded, and intimidated into becoming active, well-organized

revolutionaries.

In the National Liberation Front they learned the

doctrine, roles, and skills appropriate to the revolution.

This was

non-fOrmal education with a vengance.

Douglas Pike, who has character-

ized the NLF as "a Sputnik in the political sphere Of the Cold War,"2:3

describes the indoctrination carried on by the front in terms of

communication.

He could just as easily have employed the concept Of

non-formal education.

Pike explains that the main channels of communication employed by

the National Liberation Front were the functional liberation associa-

tions, such as the Farmers' Liberation Association and the Women's

Liberation Association, which were formed by the National Liberation

Front in each village.

With the social organization as a communication

device we reach the heart and the power of the

NLF.

Here lay the solution to the mystery that

for so long puzzled knowledgeable and thinking

Americans:

How could the NLF achieve success

in the face of overwhelming GVN* military super-

iority and massive inputs of American material

resources for civic action programs to alleviate

economic grievances?

Not superior ideology,

not more dedicated personnel, not because the

voluntary support Of the villager'had been won,

but the social movement shaped into a self-

contained, self-supporting channel of com ica-

tion -- that was the NLF's secret weapon.2

* Government of Vietnam, the south Vietnamese regime.

 

416

The range Of techniques employed by the National Liberation Front

for the indoctrination and mobilization of the villagers - agit-prop

teams, struggle movements, mutual criticism, and the like -- need not

distract us here.

The relevant point was that the NLF built its

organization by reshaping the villagers' attitudes.

The front played

upon latent hostilities toward the government, and upon ever-present

Vietnamese nationalism, tO elicit the villagers' emotional commitment

to its cause;

Once the villagers were engaged by the social movement

of the NLF, they were molded into dedicated, effective supporters Of

the organizations through which the revolution was pursued.

The follow-

ing text, from a document captured in 1961, describes the symbiosis

between organization-building and psychological transformation in the

National Liberation Front.

It could just as well have been written

about the COOperatives in Gomilla, though the struggle in Comilla was

against poverty, not against the government.

An enlightened people if unorganized cannot be a

force to deal with the enemy....

Therefore organ-

ization of the masses is essential; it facilitates

our cause in all ways....

The (social movement)

provides a strong force to Oppose the enemy; it

makes the Party's task much easier, (and) it both

provides an audience for the agit-prop cadre and

facilitates further agit-prop work.

The (social

movement) is a measure of our physical and moral

strength: it is a practical way of both serving

the people's interests and guaranteeing Party leader-

ship among the peopge; it is the decisive element

in the Revolution.

The Psychosocial Subsystem ap_pp§_Village Tier

A prodigious amount of training of village personnel was conducted

in the Gomilla cooperatives.

Statistics give a glimpse of the scale

of this non-formal educational activity.

As of the end of 1965, weekly

417

classes at the Thana Training and Development Center were being attended

by 183 managers of village cooperatives, 139 model farmers, 51 village

accountants, 91 Imams learning to teach literacy in their mosques, and

by 34 village midwives.

Together with the village health workers (27

in weekly training classes) schoolteachers (95 trainees), research

enumerators (20), family planning workers (20), women teachers (26),

other women trainees (159), and other miscellaneous trainees (125), a

total of 970 villagers were coming to weekly training classes in 1965.

Altogether, 2,244 trainees from the villages of Comilla Kotwali Thana

came to Abhoy Ashram to attend classes in 1965.

Nine hundred seventy

came to weekly classes, 80 to fortnightly classes, 268 to monthly

classes, 253 to short courses, and 673 to intermittent training sessions?6

This barrage of figures confirms that training was not a casual endeavor

in the Comilla cooperatives.

The overt purpose of all Of this training was to impart technical

information.

Village midwives, or ga_i_s_, learned about sanitation and

techniques of home delivery.

The village accountants learned to keep

ledgers.

The chairmen of village cooperatives, who came to Abhoy Ashram

for monthly classes, learned how to conduct meetings.

And the model

farmers, who were most commonly mentioned in the Gomilla literature,

Observed the demonstration farm run by the Japanese agricultural experts

at Abhoy Ashram, and learned about plant spacing and fertilizer require-

ments for improved rice variaties.

In the various classes at the

Abhoy Ashram training center, the instructors were usually either the

thana officers Of the nation-building departments (i.e. agriculture,

fisheries, health), or personnel Of the thana cooperative association,

such as the inspectors.

In most cases the trainees were expected to

return to their village and impart the newly acquired knowledge to

418

the other members of the local organization (cooperative, school,

mosque) from which they came.

This staged strategy of training -- in

which the local organization selected someone to be trained, so that

he could return to the community to disseminate what he learned -- has

been cited as a more effective alternative to the extension approach,

in which a knowledgeable Official goes out to the villages to teach.27

Commenting on the difference between the Gomilla approach to

training and the orthodox extension approach, Akhter’Hameed Khan has

written:

We replaced the government village level worker,

who in reality supervised five villages, by a

manager and a model farmer for each village,

chosen by the group itself:

Each manager'and

model farmer was trained directly by the Thana

Training and Development Center experts.

In a

manner of speaking, an inefficient middle man in

the delivery of advice and supplies was eliminated.

Thus a major step was taken toward the peasants'

self-management and self-reliance.

However, the

prospect of fewer government "workers" did not at

all please the departments.

Instinctively they

hated decentralization, delegation and autonomy.

And they ardently desired their own battalions

of assistants.28

The foregoing passage appears in a paper which Akhter Hameed Khan

delivered to an international convocation of social communications

experts, held at the East-West Center in Honolulu.

When he was talking

to the peasants of Gomilla, as he often did at cOOperative rallies or in

individual conversations, Akhter Hameed Khan did not dwell on conflicts

between different methodologies for the dissemination of agricultural

information.

What he sought to communicate to the villagers, for the

most part, was a doctrine Of cooperative labor toward common goals.

References to the brotherhood Of all Muslims, to the need for solidarity

among the exploited, and to the necessity of communal self-help laced

419

these speeches, as can be seen in the excerpts below.

Causes of

disunity, such as factionalism and envy, were consistently denounced.

If you are hoping for your situation to be improved

by the efforts of outsiders, you are mistaken.

That will not happen.

You must make the efforts

yourselves.

Those Of you who are Muslims can read

in the Quran that Allah says:

"Allah does not

better the condition of any nation until that

nation makes the change in its heart."

That is,

if there is envy, lack of faith, and factionalism

in men's hearts, then you will see that their con-

dition will not improve.

Gradually it will worsen.

But if the state Of mind and heart changes, the

external state will also change.

There is a verse in the Quran, addressed to those

who have just become Muslim.

Allah, who knew of

their condition, said:

"Remember how you were

separate, each group hostile to the others, and

how Allah united your hearts.

In Allah's goodness

and kindness you all became like brothers."29

I think the root of all progress in our thana -—

scientific irrigation and introduction of improved

methods -- is your discipline and organization.

In

this thana of ours, due to your concerted efforts,

a new mentality and a new society has emerged.

You

have given respect in place Of jealousy....

You

have given up quarrels and started to work with

unity and cooperation, and learned to save instead

Of wasting money.

You have realized the value Of

labor, of industriousness in place of indolence.

This is the main foundation.

If we have such a

character and such a society where there are no

quarrels and no disunity, where all want to work

cooperatively, where there is no jealousy but

respect, and where there is will to help one another:

people of such a society will not remain poor'and

their destitution will not persist.30

Was Akhter Hameed Khan's Islamic ideology Of cooperation communi-

cated to the village trainees at Abhoy Ashram by the inspectors and

departmental officers who served as instructors?

NO record-of what

was said in those classes is available, so there is no way of telling

for sure.

Probably the fidelity with which the "party line" was

communicated was poorer than in countries such as China and Vietnam,

420

since there was no party organization dedicated tO maintaining the

ideological purity Of the cadres who convinced the populace.

It can

be said, though, that a greal deal of attention was given to training

the trainers, both in Comilla and in the expansion program thanas where

the cooperative model was taken up in the nineteen sixties.

Top-level

personnel were brought to workshops and conferences at the Academy for

Rural DevelOpment at Kotbari, and lower level personnel were indoc-

trinated and admonished in staff meetings.

A sense of solidarity was engendered, also, in the ritual of

travelling from even the most remote villages to the Thana Training

and Development Center for weekly classes.

These regular meetings at

the training center were among the few Opportunities which the

peasants had for'moving outside of their usual orbits of village,

market, and mosque.

Within each village, similarly, attendance at the

weekly cooperative society meeting fulfilled a deep hunger'among the

atomized Bengali villagers for identification with one another and with

a larger movement.

In the and meetings and ideology proved to be insufficient instru-

ments for the maintenance of solidarity among the cooperative member-

ship.

SO long as the entrenched rural elites disregarded the village

cooperatives -- as they did, by and large, until the advent Of the

Green Revolution transformed their attitudes toward the profitability

Of farming -- the small farmers within the cooperative fold made

significant progress toward mutual betterment.

However, the bulwark

of small farmer solidarity was not strong enough to confine.the machin-

ations of the large farmers and moneylenders.

In the two organizational precepts quoted in the introductory

section Of this chapter, Norman Uphoff and Milton Esman bespoke their

421

comprehension of the disruptive potential Of favoritism and inequality.

They called for’a politics Of inclusiveness in rural development

organizations, so that all could share in the benefits, and they asserted

that organizational solidarity was more easily attained in societies

where the underlying distribution Of assets and income is more egali-

tarian.

I have explained already, in chapters one and two, how the

democracy of poverty in East Pakistan allowed nonetheless for differen-

tiation of the rural society into classes Of big, or surplus.farmers,

small farmers, land poor farmers, and landless laborers.

All of these

classes might be fitted into a range of landholdings Of from zero to

ten acres.

Even in the fertile alluvial land Of East Pakistan, a farmer

could not garner the wherewithal from ten acres to carry on like a

landed aristocrat.

But he could acquire enough leverage to exploit

his neighbors.

Left to itself, the politics of the village cooperatives

gravitated toward control by these surplus farmers, once those farmers

saw the advantage Of obtaining cheap credit to finance their'agricultural

operations.

A politics of equality, and an ideology of solidarity,

required active organizational intervention from above to balance

the local power of the landed elite.

In China and Vietnam, such inter-

vention from above was accomplished via the communist party and its

front organizations, supported by military force.

NO analog of

such a party existed in East Pakistan, or exists in Bangladesh.

The

frail, and unfortunately corruptable, reed of the cooperative

inspectorate was all that could take its place.

But now we are getting

ahead Of our story, since the managerial supervision exercised through

the inspectorate is a topic for the next chapter.

The major purpose of the cooperatives was to help the small

farmers, not the big ones.

The primary cooperatives would have been

#22

easier prey for the local landlords if the ideology of solidarity had

not been communicated to the members.

Up to a point, a powerful

organizational doctrine could hold in check the disruptive machinations

of the village elites, and thereby hold the cooperatives true to

their purposes.

When a local cooperative was captured by landlords,

participation by the general membership fell off, savings deposits and

loan repayments declined, and the organization disintegrated.

So in

safeguarding the integrity of the village cooperatives, the doctrine of

solidarity helped to preserve functional order in the organization,

much as a component of the psychosocial subsystem would be expected to

do.

The contribution of doctrine to the maintenance of organizational

order was even greater in the National Liberation Front of South

Vietnam.

The threat to organizational integrity there was posed not

so much by local landlords, who had fled to the cities decades ago, but

by the army of an opposing political system.

Villagers who had been

indoctrinated by the NLF were amazingly resistant to the blows which

the Government of Vietnam and its American ally directed against them.

In the end, as I have said above, the National Liberation Front crumbled

under the pressure, leaving the final battles to be fought by main

force North Vietnamese divisions.

But considering the enemies it was

faced with, the NLF put up an heroic struggle.

The politicking between

small and big farmers in Gomilla seems

petty in contrast.

The

resiliance of the NLF was due in large part to the strength of its

revolutionary doctrine.

The psychosocial subsystem of the NLF was

nearly unbeatable.

Non-formal education in Gomilla communicated more than values and

doctrine.

Through training, the clay of the village people was molded

423

into competent

managers, exemplary model farmers, and sagacious chair-

men.

Such, at least, was the ideal, and in a surprising number of

cases it was achieved.

management skills were learned, and so was the

capability to perform other roles in the village cooperatives.

The managers, model farmers and chairmen learned.a great deal from

each other when they gathered for their classes at the Thana Training

and Development Center, Abhoy Ashram.

Self-confidence through identi-

fication with a "guild" of local managers (or model farmers, or chair-

men, or womens program organizers) may have been one of the most import-

ant things learned at Abhoy Ashram.

Additionally, the inspectors and

department officers who taught the Abhoy Ashram classes served as non-

formal educators imparting job-related skills.

There were classes on

how to conduct a meeting, classes on how to cultivate fish, and classes

on how to prepare loan applications.

The teaching and learning of

organizational roles and of job related skills contributed enormously

to the viability of the structural and technical subsystems in the

village cooperatives.

Not until the officers of the village coopera-

tives learned to work together within the prescribed organizational

framework could the new cooperative organizations function in the

villages.

During the nineteen sixties the Gomilla project favored the train-

ing of cooperative managers over the training of model farmers.

The

Gomilla experiment in rural development had had its genesis in the

collapse of the Village Agricultural and Industrial Development, or

V-AID, program.

V-AID had been born out of the same assumptions as

the community development movement in India.

As in the Indian prOgram,

the front line activist in V-AID was a village level worker.

During the

early nineteen sixties the administrators of the Gomilla project were

#2#

ambivalent as to whether the functions of the village level workers

should devolve, in the new scheme, upon two or upon four new persons.

In the four-way contest, the managers of the village cooperatives and

the field inspectors of the central cooperative association were

favored, while the model farmers and the extension staff were occassion-

ally given attention and more frequently neglected.

From July 1964 to

April 1965, the position of model farmer was actually abolished, and

in that period the managers were expected to double as model farmers.

The ambivalence of the Gomilla project administration toward the

two types of village workers resulted in part from the disappointments

of extension in the V-AID program and in the early years of the Gomilla

experiment.

In 1960, when the academy formed its first village coopera-

tives, the major improved practices being taught were low lift pump

irrigation, line sowing, and the use of fertilizer.

Although the

farmers did take to irrigation once the technology and organization

of that practice had been worked out by the staff of the Academy for

Rural Development, there was but a faltering adoption of line sowing

and chemical fertilization.

According to Akhter Hameed Khan, the

acreage under line sowing in 1962-63 was much less than in 1961-62.

It

was apparent that the farmers had rejected line sowing, and it seemed

the reason was that the effect of line sowing on yields was nominal.

There was no such spectacular reverse in the history of the adoption

of fertilizer, but it was found that the farmers were for a long time

wary of their use because of the tendancy of local rice variaties to

lodge (fall over) when too great a dosage was applied.

Taking a lesson from these experiences, Akhter Hameed Khan concluded

that the route to agricultural modernization was not through the alter-

ation of the techniques of cultivation of the existing aus and amon

#25

rice crops, but through the provision of irrigation water’and tractor

services which would allow the farmers to grow a third crop of dry-

season 3939 rice.

Because tubewells and tractors are management-

intensive inputs, the emphasis of the program shifted from extension

to organization.

From 1964 to 1968, the slogan of the Gomilla cooperatives might

have been "All power to the Managers!"

For'a time, the managers even

assumed the duties of model farmers.

All commissions from the central

association were given to the managers, who collected:

1.

A 1% commission on all loans realized on time.

2.

a.

50 paisa per member per month if attendance at weekly

meetings averaged 81-100%.

b.

37 paisa per member per month if attendance at weekly

meetings averaged 61-8Q%.

c.

25 paisa per member per month if attendance at weekly

meetings averaged 51-6Q%.

3.

15 rupees per month if the village cooperative had an irriga-

tion tubewell which was irrigating at least 25 acres of land.

a.

5 rupees for every eight-hour day that a tractor worked in

the village.

5.

5O paisa per acre under improved practices of cultivation.

Although three of these five types of commissions relate to the

utilization of improved practices, none of them was allocated to the

model farmers.

In the Gomilla-type rural development projects which

had been organized in neighboring Gaibandha thana there was.a much

stronger extension emphasis.

The model farmers in the Gaibandha project

were given the commission for acreage under improved practices.

Asked to explain the Gomilla project's discrimination against the

model farmers, Akhter Hameed Khan said:

426

I came down squarely on the side of the managers

in this dispute over the roles of the manager'and

the model farmer.

I didn't want to have two

people in the village running the local COOpera-

tive.

The model farmer‘s function should be to

demonstrate on his own farm the improved prac-

tices he had learned, and that is all.

He should

not challenge the manager's authority.31

In the last two years of the decade, in 1968-69 and in 1969-70,

the pendulum of attention swung back to a more intermediate position

between the managers and the model farmers.

By then there was such a

flood of new technical information to be imparted -- about the high

yielding Green Revolution rice varieties.about potato and vegetable

cultivation, and so forth -- that the contribution of the model farmer

to the cooperatives' success attained greater importance.

This uptick

of interest in the model farmers is evident in the last row of

Table 7-1.

During the period from 1962-63 to 1967-68, the proportion of

village COOperatives whose model farmers attended training classes

hovered in the neighborhood of 63%.

The proportion shot up to the

ninety percent range in the last two years.

The annual reports of

the Kotwali Thana Central COOperative Association, from which these

figures were extracted, do not explain why more than one manager per

village COOperative was in training at Abhoy Ashram each year.

When the managers, model farmers, and others who had attended

classes at the Abhoy Ashram training center returned to their villages,

they were expected to educate their fellow cooperative members.

The

major forum for this stage of the non-formal education process in

the Gomilla cooperatives was the weekly general meeting in each village

cooperative society.

Badruddin Ahmed puts it very well:

 

#27

 

 

 

1
9
0

#
5
.

7
6
.

  

8
5
-

5
7
.

2
7
.

“
5
.

 

 

 

 

 

)
a
/
c
(

s
p
o
o
c
/
s
r
e
m
r
a
f

l
e
d
o
M

7
6
—
6
6

8
6
-
7
6

j
6
-
8
6

7
7
9
6

0 

6
6
-
5
6

5
6
-
4
6

4
6
-
3
6

3
6
-
2
6

f
o

s
e
v
i
t
a
r
e
p
o
o
c

l
a
r
u
t
l
u
c
i
r
g
a

e
h
t

r
o
f

r
a
e
y

h
c
a
e

d
e
n
i
a
r
t

s
r
e
m
r
a
f

l
e
d
o
m

d
n
a

s
r
e
g
a
n
a
M

:
1
—
7

e
l
b
a
T

.
a
n
a
h
T

i
l
a
w
t
o
K

a
l
l
i
m
o
G

6
1
3

1
0
3

1
6
2

2
0
#

8
8
2

9
8
3

8
7
2

0
3
3

0
4
1

7
2
.
1

9
2
.
1

6
2
.
1

5
2
2

6
9
2

8
5
1

0
2
2

2
5
1

7
0
2

2
2
1

0
1
1

)
a
(

s
e
v
i
t
a
r
e
p
o
o
c

e
g
a
l
l
i
V

2
6
1

2
2
1

)
b
(

s
r
e
g
a
n
a
M

0
5
1

5
1
1

5
1
1

1
7

9
5

)
0
(

s
r
e
m
r
a
f

l
e
d
o
M

1
3
.
1

9
3
-
1

6
3
.
1

3
3
.
1

1
1
.
1

)
c
/
b
(

s
e
v
i
t
a
r
e
p
o
o
c
/
s
r
e
g
a
n
a
M

#28

The weekly cooperative meeting is not just a

meeting held to conduct a little fermal business.

It is an educational experience intended to help

the member learn how to become a modern man.

It

is as unthinkable to try to run a village coopera—

tive without a weekly meeting as it would be to

run a school without attendance at classes.

And

this is precisely because the village cooperative

itself is a special kind of school.

This school

is linked to a communication centre which feeds

it through the monitors in the class -- the mana-

ger, model farmer, chairman and those who receive

training at the Thana Training and Development

Centre.32

As a'bpecial kind of school" in which non-formal education for

the cooperative membership was carried on, the weekly meeting was

expected to inculcate apprOpriate attitudes in the villagers.

One

of the most important attitudes was that of a p0sitive orientation

toward savings and loan repayment.

When the cooperative experiment

began, the scheme was derided by many seasoned officials and foreign

experts on the grounds that Gomilla villagers were too poor to save.

But the poor farmers showed that they could save, and could do so

consistently.

Peer pressure was the main method for teaching the habit

of savings.

When roll was called in the weekly meeting, each member

was expected to come forward and deposit a small amount of money.

The

loan limit alloted to the village cooperative society was dependent

upon the society's record in savings and loan repayment, so each member

recognized that he had a stake in the compliance of every other member.

Judging by the cooperatives' good performance in savings and in

loan repayment, both in Gomilla Kotwali Thana and in the expansion

projects, the peer-pressure technique succeeded.

In the older system of

Union Multipurpose Cooperatives, where the membership was much less

involved in the running of the organization, the record of capital

accumulation by the membership and of loan repayment was much worse.

429

As a consequence the Union Multipurpose Cooperatives were incapable of

attaining financial self-sufficiency, they served in most cases as

conduits for lightly disguised grants from the government to the rural

upper classes.

Cooperatives of the Gomilla type, on the other hand,

proved themselves capable of covering a large portion of their’admini-

strative costs, and of financing most of their own loan operations.

We can conclude that the psychosocial subsystem in the village coopera-

tives, acting through peer pressure on the membership, came to the

rescue of the technical subsystem at the thana tier.

Did the peer pressure technique transform attitudes, or only

behavior?

I mentioned earlier Gabriel Almond and Sidney Verba's con-

tention that democracy works best where the political attitudes of the

citizenry approximate those of the civic culture.

A parallel case

could be made, to argue that the credit program in the village coopera-

tives functioned best where the membership had the financial attitudes

of good Swiss citizens, saving money and repaying loans as a matter of

high moral obligation.33 were such attitudes inculcated by the psycho-

social subsystem of the village cooperatives? There is no way of

knowing, since no social psychologist ever took to the field in Gomilla

in the nineteen sixties to investigate that question.

Neither is it known how the personal interactions within the

village cooperatives influenced the political attitudes of the member-

ship.

Did the cooPerative members really acquire the attitudes of the

civic culture?

The fractious tumult of coOperative polities which

Peter Bertocci observed in the Gomilla villages of Hajipur and Tinpara

would not appear to have been grounded in the mutual trust that is

characteristic of the civic culture.

Yet it is conceivable that

political attitudes were slowly and positively transformed in the local

Leo

cooperatives, so that they could eventually have become democratic in

actuality as well as in legal form.

In the second chapter I mentioned

Ayub Khan's quip to the effect that after a period of administrative

tutelage the Basic Democracies system would become more democratic and

less basic.

A process of gradual democratization may have taken place

in the Gomilla COOperatives.

Such is the impression of long time

observers such as Akhter Hameed Khan.

But solid evidence, gathered

across many villages and over a span of many years, is lacking.

I caught a brief glimpse of the politics of Gomilla village coopera-

tives when I participated in a field research team, led by Zaker Hussain,

which investigated management in #5 primary cooperatives in Kotwali

Thana in 1966.

In the unpublished manuscript of my monograph on The

Gomilla System of Supervised Credit, which I wrote after our field team

had finished its work, I reflected as follows on democracy in the

village cooperatives:

The fundamental concept of the Gomilla coopera-

tives is that of democratic village organization.

It is expected that if the members chose their

own managing committee, and if the managers dis-

cuss all of the activities of the cooperative in

the weekly meeting, then the cooperative members

will themselves assure that the manager stays

honest and continues to represent their interests.

But the villagers of East Pakistan, even of

Kotwali Thana, are not always the best of demo-

cratic citizens.

They may not comprehend what

is in their best interest or how their cooperative

should be run.

Being unaccustomed to cooperative

organization, they may not care to reform their

cooperative, even if they know it is going wrong.

And the managers are not always the best of public

servants.

Some of the managers are rich farmers,

more closely allied with the money lending class

than with the cooperative membership.

Some of

them are outright thieves.

#31

The conclusion reached by Zaker'Hussain, Akhter’Hameed Khan, and

the other leaders of the Gomilla cooperative organizations was that

village democracy could not be trusted to function on its own.

It must

be closely supervised through the cooperative inspectorate.

The

inspectors were encouraged to mobilize public Opinion against an

irredeemable manager, or in extreme instances to file a criminal case

against those who could not otherwise be dislodged.

The approach of

supervised democracy in the village cooperatives was more cautious and

paternalistic than Ayub Khan's Basic Democracy system for governing

rural Pakistan.

Ayub Khan's regime did not exercise administrative

sanctions against elected union councilors who were corrupt.

The

Gomilla cooperatives did so, and by winning popular respect proved

themselves more durable than the system of Basic Democracies.

The Psychosocial Subsystem gt the Thana Tier

At the thana as at the village tier, training and indoctrination

were significant in sustaining organizational order, and in restraining

the temptation to corruption.

Harry W. Blair has summed up the situation

in pithy terms.

"In the ethos of bureaucratic venality and peculation

that is so pervasive in the subcontinent, it was considered absolutely

essential to have a credible program that the individual peasant would

have the confidence to invest in."3”

Akhter Hameed Khan sought to indoctrinate the employees of the

thana cooperative associations with the ideology of service, solidarity,

and probity.

Speaking to the employees of the Kotwali Thana Central

Cooperative Association in 1969, for example, he said,

An organization is damaged by three causes:

—.

/'

.

432

thefts, fraud, and factionalism....

The meaning

of theft is:

to get a thing on which you have

no legitimate right and to take advantage there-

from....

The meaning of deception is:

I was

given an assignment, but I didn't discharge it

properly, or just set it aside, or I did not make

any attempt to learn the work....

The next is

factionalism -- internal dissension and disputes....

Our organization has been growing only because

these three things did not exist here.

We have

always thought that we are not working for the

benefit of any particular person.

The entire

work is for the benefit of all, specially for the

poor people in the villages and the town.$

Exhortations of the brotherhood of all Muslims are to be found in

Khan's speeches to the thana employees, but they are muted compared to

the trumpet blasts in his speeches to the villagers.

It is as if the

distinction between the thana and village tiers was matched in Akhter

Hameed Khan's mind by the differentiation between Western and non-

Western society.

His vision of a smoothly coordinated thana government

was abstract and quite Western in character.

At the thana level was to

be found a set of organizations sufficiently formal, and sufficiently

sophisticated, so that he could bring to bear upon it his training in

the Indian Civil Service and the advice of his cohort of foreign

experts.

The struggle to reconcile modernity and traditionalism, or Islamic

and European values, is an issue of acute sensitivity throughout the

Muslim world.

Failure to achieve such a reconciliation led to an

Islamic revolution in Iran in 1979.

For Akhter Hameed Khan, as for

many officials in Pakistan and in Bangladesh, the solution to this

dilemma lay in a segregation of the extremes.

The Western influence upon the thana cooperative organizations was

direct.

Robert Havener, a.senior advisor in Comilla in the mid-

sixties, used to conduct weekly classes in accounting and business

management for the staff of the KTCCA.

A few years later'Henry Larzelere,

433

an MSU professor of agricultural economics who was stationed at Comilla,

did the same.

In an interview in June of 1976, Larzelere recalled that

he ran management training exercises for the section chiefs in the

Kotwali Thana Central Cooperative Association (such as the head of the

dairy and of the cold storage), encouraging all of them to propose

ways of handling administrative problems in each section.

"The key

decisions in the KTCCA were made by the section chiefs," Larzelere

said.

"There were no farmers in that group.

Those people were all

bureaucrats running their own shops.

It worked so long as there was a

person such as Akhter Hameed who would call them up short to remind them

of’the people they were really working for."

The problem of maintaining responsiveness and honesty in public

agencies is not unique to Gomilla.

It is faced equally by the govern-

ment of New York City.

The distinction between the two locales was

that the quota of in-service training, exhortation and indoctrination

per employee was probably greater in Gomilla than in New York.

The

integrity of the structural and technical subsystems in the Comilla

cooperatives was thereby protected, and the responsiveness of the system

enhanced.

In the course of the theoretical discussion in the third chapter,

Kenneth Boulding's criteria for distinguishing living from non-living

systems were described, and deemed sufficient for discriminating, in the

typology which I proposed, between machines with input and organic

systems.

Organic systems, it was proposed, are capable of self-

maintenance and reproduction.

In this chapter, most of the discussion has centered on ways in

u—

#34

which formal organizations such as the cooperatives can maintain their

internal integrity.

The contribution of ideology to organizational

maintenance was given special emphasis.

It bears mentioning that ideas

and values are also critical in organizational replication.

A well-

articulated concept, or doctrine if you will, of how an organization

should be structured and of how it should function can play the same

sort of role in organizational replication as does the genetic code in

biological reproduction.

In the case of Gomilla, the genetic instructions for the structural

and technical subsystems of an ideal Thana Central COOperative Associa-

tion were clarified in Kotwali Thana and in the three earliest expansion

projects during the experimental period lasting from about 1960 to 1965.

By the mid-sixties the doctrine of what such an organization should

look like and how it should work had crystallized in the minds of people

such as Akhter'Hameed Khan, Zaker Hussain, and some of the faculty at

the Academy for Rural Development.

The molecule of DNA containing

the genetic code had been completed.

To spawn other'Thana Central

Cooperative Associations thereafter, the new doctrine had to be trans-

mitted to the personnel, and especially to the leaders, of the unhatched

cooperative associations.

The genetic code had to be passed on to the

next generation.

This task of transmission was accomplished, by and

large, through non-formal education.

Through classes at the academy and

through field trips to the Abhoy Ashram training center and around the

thana, the directors of the expansion projects were indoctrinated.

And

just as it is a miracle of nature that each newborn creatureis not an

absolute mutant, but instead bears a strong resemblance to its

parents and ancestors, so the cooperative associations that sprung up

in the thanas of Gomilla district, and later in many other thanas of

435

East Pakistan and of Bangladesh, looked rather like the original

organization, the Kotwali Thana Central Cooperative Association, as

it had appeared in about 1965, before it altered its own form by adding

the extra tier of the Agricultural Cooperatives Federation.

The analogy between biological and organizational reproduction is

imperfect, because the instructions for replication are not inculcated

solely in the personnel of the next generation of organizations.

In

the organizational universe, a supervisory official or agency, operating

from a predetermined budget and staffing pattern, frequently manages the

process of organizational duplication.

If the newly founded organizations

stray from this manager's conception, he brings them into compliance

with his standards.

No such super-organic manager is to be found in

the biological world, unless, that is, one credits the reality of

Mother Nature .

Notes for Chapter 7

The Psychosocial Subsystem

1 Daniel Katz and Robert L. Kahn, _T_h_e_ Social Psychology 9; 9_rga_n_-

izations, John Wiley and Sons, New York, 1963, p. 33.

Reprinted by

permission of John Wiley and Sons.

2 Fremont E. Kast and James R. Rosenzweig, Orgggization gag

Management:

A Systemg Approach, Second edition, MCGraw Hill, New York,

1974, pp. 245-247.

Copyright 197” by the McGraw Hill Book Company.

Used with permission of McGraw Hill Book Company.

3 Norman T. Uphoff and Milton J. Esman, Local Orggization :2;

Rural Development:

Analysis 9; Asian Egperience, Rural Development

Committee, Cornell University, Ithica, 1974, pp. 101, 103.

u Milton J. Esman, "The Elements of Institution Building," in

Joseph W. Eaton, ed., Institution Building egg Development, from

Concepts 39 Application, Sage Publications, Beverly Hills, 1972, p. 22.

5 This closeness of fit between the institution building paradigm

and my own is fortuitous.

The scholars who participated in the Inter-

University Research Program in Institution Building knew next to nothing

about organization theory.

In an essay written near the end of the

research project, which terminated when its Ford.Foundation and USAID

funding expired, William J. Siffin rubbed his eyes and observed, "The

IB perspective sits on top of a veritable midden heap of what is loosely

called 'organization theory.'"

(William J. Siffin, "The Institution

Building Perspective:

Properties, Problems, and Promise," in Ingtitu-

tion-Buildigg:

A Model f9; Applied Social thggg, edited by D. Woods

Thomas, Harry R. Potter, William L. Miller and Adrian F. Aveni,

Schenkman Publishing Company, Cambridge, Massachusetts, 1972, p. 122.)

6 Milton J. Esman, "Some Issues in Institution Building Theory,"

in Thomas, 9p git, p. 79.

Reprinted with permission.

7 Gabriel A. Almond and Sidney Verba, The Civic Culture:

Political

Attitudes gag Democracy ig_Five Nations, Little, Brown and 00., Boston,

1985: P' 13-

8 Ibid., P0 120

436

437

9 Peter Almond, Gabriel's son, was employed by his father through-

out a summer in sorting the 5,000 questionnaires so as to prepare the

tables for the book.

Almond and Verta's quaint methodologr of data

analysis had its drawbacks, since they did not make use of correlation,

chi-square, tests of statistical significance, or other techniques

which would have enriched the book.

The credibility of the study

rests mainly on the size of the sample.

10 Almond and Verba, gp_gi§, p. 360.

11

Nations, New York, Oxford University Press, 1973? 57'133.

"-

12 Philip H. Coombs and Manzur Ahmed, Attacking'Rurgl Poverty,

How Nonformal Education Can Help, The Johns HOpkins University Press,

Baltimore, 1974, p. 134.

13 Ibid., p. 135.

14 E. Patrick Alleyne, et al, Training and Research for Extended

Rural Development ig_Asia, Rural DevelOpment Committee, Cornell Univer-

sity, Ithica, 197b, p. 8.

Italics in original.

15 Syed A. Rahim, "Non Formal Aspects of the Gomilla (Bangladesh)

Project," in Richard O. Niehoff, editoru Egg Formal Education §3g_thg

Rural Poor, Program of Studies in Non Formal Education, Michigan State

University, East Lansing, 1977.

6 Coombs and Ahmed distinguish among informal education, non-

formal education, and formal education.

Both non-formal and formal

education are structured and systematic, but non-formal education is not

school-bound.

Informal education, in their definition, takes in all

education that is not structured or systematic, such as the education

that results from interactions in the work environment.

17 Mao Tse Tung, "The Struggle in the Chinkang Mbuntains," in

Selected Works 9: Mac Tse Tung, Volume one, Lawrence and Wishart,

London, 1954, p. 81.

18 Edgar Snow, Red Star Over China, first revised and enlarged

edition, Victor Bollancz, Ltd., London, 1968, p. 116.

19 Mao Tse Tung, "The Struggle in the Chinkang MOuntains,"

.92 at! PP° 82: 83-

20 Douglas Pike, Viet Cong, The M.I.T. Press, Cambridge, Massa-

chusetts, 1966, p. 238.

21 Jean Lacouture, Hg Chi Minh:

_A_ Political Biography, Random

House, New York, 1968, p. 2&7.

22

438

Ho was the more senior of the two revolutionaries, and had

learned his Leninism at first hand, in Mescow in 1927’and 1924.

Ho

never met Lenin himself, but he was an intimate colleague of Lenin's

closest desciples.

Ho was a strong supporter of Lenin's third inter-

national.

He once wrote:

"There is a legend, in our country as well

as in China, on the miraculous 'Book of the Wise.‘

When facing great

difficulties, one opens it and finds a way out.

Leninism is not only a

miraculous 'book of the wise,‘ a compass for us Vietnamese revolution-

aries and people:

it is also the radiant sun illuminating our’path

to final victory, to Socialism and Communism."

(From Lacouture, pp pip,

p. 32.)

mac Tse Tung, in his formative years, had no international

exposure.

He first set foot outside of China after the revolution had

triumphed, when he travelled to Moscow in 1949.

23Douglas Pike, 9;; c_i_t_, p. 111.

24

25

Ibid., p. 124w

Reprinted with permission.

Ibid., pp. 125, 126.

26These figures are from Table C of Appendix XI in Arthur F.

Raper, Harry L. Case, Richard O. Niehoff, William T. Ross, and Edgar

A. Schuler, Huge; DevelOpment pp Action, The Comprehengive Experiment

;p_Comilla, East Pakistan, Cornell University Press, Ithica, 1970.

27

Nurul Islam, ed., _gricultura; Policy ip Developing

Countries, Halstead Press, New York, 1974, p. XVI.

28 Akhter Hameed Khan, "My Lessons in Communication," paper

presented at the Communication Institute Seminar, East-West Center,

Hawaii, January 1975 (mimeographed), p. 10.

29 Akhter Hameed Khan, "A Light in the Darkness," speech at Gomilla

Cooperative Societies rally, October 14, 1962, translated from the

Bengali by Marianna Tax Choldin.

30 Akhter Hameed Khan, Address to the Gomilla Cooperative Societies

rally, November 26, 1967, translated from the Bengali by Khandker M.

Rahman.

31 Conversation with the author, 1975.

32 Badruddin Ahmed, Manual pp Comilla Cooperatives, Bangladesh

 

Academy for Rural Development, Comilla, 1972, p. 41.

33 The Raiffeisen cooperatives were sometimes cited.as progenitors

of the Gomilla cooperatives.

Could the success of those nineteenth

century German cooperatives have been due in part to the Calvinist

beliefs of the member farmers?

Fear of punishment by a wrathful God

would have been a considerable incentive to loan repayment.

Ju'Harry W. Blair, The Elusivene§§.2£ Equity:

Institutional

Approaches pp_Rural Development ip_Bangladesh, Rural Development

Committee, Cornell University, Ithica, 1974, p.

'42.

3‘5 Akhter Hameed Khan, "Address Before the Employees of the

KTCCA," June, 1969 (mimeographed).

439

CHAPTER 8

The Management Subsystems

Operational

Operational

Management

Strategic

Plant

Control

Control

Planning

Static structures

Simple machines

S

ST

Machines with input

STE

Organic systems

STEP

STEPO

STEPOM

STEPOMR

The near universality of hierarchy in the compo-

sition of complex systems suggests that there is

something fundamental in this structural prin-

ciple that goes beyond the peculiarities of human

organization....

There are strong reasons for

believing that almost any system of sufficient

complexity would have the rooms within rooms

structure that we observe in actual human organi-

zations.

The reasons for hierarchy go far beyond

the need for unity of command or other considera-

tions related to authority.1

Introduction

 

In his insightful little book on planning and control systems, in

which he sets out the typology of operational control, management

'control, and strategic planning, Robert N. Anthony compares the process

440

441

of operational control to the action of the autonomic nervous system,

and the process of management control to the activity of the human

cerebral cortex.

He flatly asserts, however, that "Biological analogies

are not useful for studying the process of strategic planning."2

Through the exercise of the strategic planning function, an organiza-

tion can change its form and function, as by adding a new product

division.

Human beings, Anthony points out, are incapable of such

conscious self-alteration.

PeOple cannot will themselves an extra

arm or leg.

The capability of human organizations consciously to

transform themselves, Anthony says, is unique in the universe.

Anthony puts the case too strongly.

Human beings do exhibit the

ability to transform many aspects of themselves, or at least of their

relationship to the world around them, through the exercise of strategic

planning.

True enough, pe0ple don't change their biological bodies

(except in minor ways, such as exercising to keep fit), but they can

alter their jobs, their marital status, and.many other'aspects of

their personal reality.

The capacity for foresight in humans is thought to reside in the

frontal lobes of the brain, the section immediately behind the fore-

head.

Anxiety resides there also.

Extirpation of the frontal lobes

through a lobotomy does not entirely eliminate a patient's ability to

ponder events which may lie ahead, but does effectively remove his

ability to picture himself as an actor who could influence those impend-

ing events.3 The patient glides serenely into the oncoming future,

taking everything as it happens.

At the other end of Robert Anthony's spectrum of management

functions, it may be reasonable to compare the function of operational

control to the operation of the autonomic nervous system, which

442

controls basic body functions such as digestion, breathing, and blood

flow.

Anthony states that operational control is concerned with the

effective completion of tasks in an organization.

This is the realm

of foremen and sergeants.

Though they might not be flattered by the

comparison, to an observer who is detached from the personal realities

of the organization their work might sensibly be compared to that of

the medulla oblongata, the spinal cord, and the efferent visceral

nerves in a human being.

The common mission is to keep fundamental

processes going at the right pace, and.in the right relation to one

another.

The common methodology is that of closed loop feedback

control, which is classically illustrated in the sympathetic and para-

sympathetic components of the autonomic nervous system.

The sympathetic

nerves speed up the functioning of the heart and other organs to which

they are connected, and the parasympathetic nerves slow them down; too

strong a message from the one will induce a restraining message from

the other.

Anthony's comparison of management control to the functions of the

cerebral cortex requires some refinement.

The neocortex of the human

brain is subdivided into hemispheres, lobes, and folds.

The frontal

lobes, as I have mentioned, are the locality of foresight and of

rational intervention, functions which are comparable to strategic

planning in organizations.

Various other portions of the convoluted

neocortex correspond to different sensory and muscular systems in the

body.

There are folds which assimilate the information sent in from

the sense of touch in the hand, and folds which transmit instructions

to the muscles of the hand.

Inasmuch as obtaining and assuring the

prOper use of resources (or inputs) is one aspect of management control

in organizations, and inasmuch as human beings accomplish these same

“*3

functions via the sensory and motor systems controlled by the neocortex,

then the neocortex can be considered the location of neurological

activities which resemble management control.

In the previous chapter I stated that the operating plant of an

organization is the locus of interaction among the structural, tech-

nical and psychosocial subsystems.

These three subsystems taken

together*are what the skeleton, muscles and organs, and hormones are to

a human being.

The structural, technical and psychosocial subsystems

encompass the fundamental processes of the organization, just as the

basic activities that make a human being live and move are carried out

by the skeleton, muscles and organs, and hormones.

Just as a man would be incapacitated without a brain, and dead

without a nervous system, so an organization requires management sub-

systems to provide it with control, competence and direction.

But not

all parts of the brain are vital.

A man can live, may even appear

normal, if his frontal lobes have been excised.

So also an organization

can get along handily without a strategic planning subsystem.

Robert

Anthony cites a 1962 Stanford Research Institute study in which it was

estimated

that no more than 25 manufacturing companies in the United

States had.a long range planning program which was fully developed and

staffed.

Especially for an organization which performs repetitive

tasks in a predictable environment, a strategic planning subsystem is

a luxury.

At the other end of the management scale, we might ask if an

organization could "survive" without an operational control.subsystem.

Perhaps it could.

In Chapter 3 I quoted Henry Kissinger's description

of soviet army battalions that advanced into their own artillery fire,

carrying out orders without adjustment to circumstances.

The soviet

444

battalions behaved like systems with open loop control; they operated

according to pre-calibrated instructions, and were incapable of feed-

back.

They carried on through the strength of their organizational

doctrine, which held that they should move on and execute orders at

all costs.

The soviet army example brings to light an interesting rule of

organizational design:

doctrine can be substituted for'management.

In Gomilla, the need for managerial sanctions to control peculation

was diminished to the extent that the cooperative employees believed in

Akhter Hameed Khan's ideology of probity and public service.

In

another sphere, the trade off between ideology and administrative

control can be seen as the nub of the ideological disagreement between

Chinese and Russian communists.

The Chinese believe that the populace

should control itself through the acceptance of doctrine.

Because mass

conformity has been attained in China, the communist party there rules

with a relatively light hand, although it is quite prepared to snuff

out any deviation which challenges the basic precepts of communism.

The Russian communists have inherited the dark suspicion of the common

people which typified the czars.

They rule through draconian autocracy.

What about the intermediate range of management?

Is the possession

of a management control subsystem a luxury or a necessity for human

organizations?

In biological organisms, the acquisition of resources

such as food, water'and air are vital functions.

In Robert Anthony's

typology, control over the acquisition and use of resources are functions

of management control.

Can we conclude that an organization must have

a management control subsystem to survive?

Not necessarily.

The

example of plants -- which have no nervous systems -— makes it clear

that the functions of acquiring and dispensing resources can be

445

accomplished without much management control.

Indeed, when a stroke

reduces a human being to endless coma, due to the loss of function in

the neocortex, we commonly say that that person has become a vegetable.

He lives, but in a terribly reduced state.

So likewise an organization

might get by without a management control subsystem, just as the soviet

battalions carried on without an operational control subsystem.

An

organization thus deprived of management would hardly be agile, but it

might function.

There may appear to be a problem with plants, because in Figure

3—11 the category of plants in the Boulding typology was matched with

the category of organic systems with operational control.

The analogy

of the operational control subsystem in an organization to the autonomic

nervous system in man was posed above.

Yet plants have no nervous

systems.

How, then, can plants perform functions of operational

control?

It seems that they do so, in an inefficient but satisfactory

way, by means of hydraulic and chemical feedback.

Two of Uphoff and Esman's maxims about rural development organi-

zation are pertinent to the study of management subsystems in the

Comilla cooperatives.

- "Sanctions to control the acts of leaders of local organizations

should be applied both from above and below to get best performance.

- "Decentralizapipp of operating decisions within a system of

centrally determined policies provides the best way of combining

authority with the most relevant information and most informed judge-

ment."’+

Additionally, we will be looking again, but from a different angle,

at a proposition which was reviewed earlier, in the chapter on organi-

zational structure.

There the multi-tiered pattern of rural development

446

organization was discussed in terms of architectural desiderata.

In

this chapter we will consider the implications for management of the

multi-tiered arrangement.

Relevant Research

The largest proportion of those who write seriously about modern

organizations are business school professors, business consultants,

retired businessmen, or some combination of these types.

Management

advice, as a consequence, is the core topic of most of the

literature

about organizations.

The number of books and articles on the managerial

process, business decision making and such topics is huge, and far

beyond my capability to review them sensibly.

It will better serve

the purposes of this chapter to concentrate on some further commentary

on Robert Anthony's management typology, and to mention some examples

of management systems in COOperatives in other areas of the world.

The seeds of Robert Anthony's generalizations seem to have fallen

on infertile ground.

Hardly anyone among the great cohort of manage-

ment writers has sought to make use of his ideas.

The taxanomic urge

is a highly unusual characteristic among the members of the intellectual

circle of business school professors such as Robert Anthony.

He knew

as much when he published his book.

In the preface, he quotes the

commentary of one of his colleagues about the manuscript of his book on

planning and control systems:

I am a patient man, I believe, but my tolerance

of the sort of Medieval scholasticism that spec-

ulated on 'angels per pin point' or on orders of

classification is not great.

I, frankly,

couldn't care less.

Go ahead and publish, but

don't expect me to read it.5

447

Perhaps it was because the faculty of the Harvard Business School

are accustomed to dealing with the top management of American corpora-

tions that Anthony said he found it so difficult to define the lowest

level of management, the level which he eventually termed operational

control.

In retrospect, the two ends of his scale, strategic planning

and operational control, appear to be more sharply defined than the

middle management role which he described as that of management control.

A slight variation on Anthony's typology of management has been

set forth by Fremont Kast and James Rosenzweig.

They admit the polarity

of the concepts of strategic planning and operational control, and

term the intermediate function coordinative management.

"The coordinative

manager operates between the operating and the strategic levels and

serves to mediate and coordinate the two.

This level transforms the

uncertainty of the environment into the economic-technical rationality

necessary for input into the operating subsystem."6

Kast and Rosenzweig

have developed an excellent diagram to summarize the differences among

these types of management.

I have reproduced it as Figure 8—1.

I hope that Figure 8-1, taken together with an exhibit from

Anthony's book which I have shown as Figure 8-2, will suffice to

explain the distinctions among strategic planning, management control.

and operational control.

The reader should note that most of the

activities listed in the first column of Figure 8-2 are purely

planning activities, that most of the entries in the third column,

under operational control, are purely control activities, and that the

entries in the middle column include a mixture of planning and control.

This observation should not mislead the reader into thinking that a

superior classification could be devised by separating management

functions into those of planning on the one hand and control on the

 

 
 

 

g
n
i
k
a
m
-
n
o
i
s
i
c
e
D

l
a
r
e
n
e
G

e
m
i
T

l
a
t
n
e
m
n
o
r
i
v
n
E

l
a
i
r
e
g
a
n
a
m

m
e
t
s
y
s
b
u
s

s
e
u
q
i
n
h
c
e
t

s
e
s
s
e
c
o
r
p

t
n
i
o
p
w
e
i
V

e
v
i
t
c
e
p
s
r
e
p

s
m
e
t
s
y
s

k
s
a
t

l
e
v
e
l

r
o

y
r
a
m
i
r
P

l
a
n
o
i
t
a
z
i
n
a
g
r
O

l
b
a
m
m
a
r
g

1

n
e
m
g
d
u
J

—
o
r
p
n
o
W

g
n

i
f
s
i
t
a
S

g
n
o
L

n
e
p
O

 

 

 

 

 

-
z
i
m
i
t
p
O

g
n
i

 

 

 

 

 

l
a
n
o
i
t

e
l
b
a
m
m
a
r
g

e
t
u
p
m
o
C

n
u
r

t
r
o
h
S

d
e
s
o
l
C

 
 
 
 
 

n
o
i
t
a
z
i
n
a
g
r
o

o
t

e
v
i
s
n
e
h
e
r
p
m
o
c

d
n
a

s
m
e
t
s
y
s

;
t
n
e
m
n
o
r
i
v
n
e

n
g
i
s
e
d

s
n
a
l
p

e
t
a
l
e
R

c
i
g
e
t
a
r
t
S

s
e
i
t
i
v
i
t
c
a

s
e
v
i
t
c
e
j
b
o

h
s
i
l
p
m
o
c
c
A

e
t
a
r
g
e
t
n
I

l
a
n
r
e
t
n
i

e
v
i
t
a
n
i
d
r
o
o
C

y
l
t
n
e
i
c
i
f
f
e

d
n
a

y
l
e
v
i
t
c
e
f
f
e

g
n
i
t
a
r
e
p
O

7

.
s
m
e
t
s
y
s
b
u
s

g
n
i
t
a
r
e
p
o

d
n
a

,
e
v
i
t
a
n
i
d
r
o
o
c

,
c
i
g
e
t
a
r
t
s

:
k
s
a
t

l
a
i
r
e
g
a
n
a
m

e
h
T

:
1
-
8

e
r
u
g
i
F

4’49

Strategic Planning

Management Contrpl

Operational Control

Choosing company

Formulating budgets

objectives

Planning the

Planning staff levels

Controlling hiring

organization

Setting personnel

Formulating personnel

Implementing policies

policies

practices

Setting financial

Working capital

Controlling credit

policies

planning

extension

Setting marketing

Formulating'adver-

Controlling placement

policies

tising programs

of advertisements

Setting research

Deciding on

policies

research projects

Choosing new

Choosing product

product lines

improvements

Acquiring a new

Deciding on plant

Scheduling production

division

rearrangement

Deciding on non-

Deciding on routine

routine capital

capital expenditures

expenditures

Formulating decision

Controlling inventory

rules for operational

control

Measuring, appraising,

Measuring, appraising,

and-improving man-

and improving workers'

agement performance

efficiency

Figure 8-2:

Examples of activities in a Business Organization Included

in Major Framework Headings.8

450

other.

This is, in fact, the sort of classification which is employed

by many authors of texts on business management.

Anthony contends,

though, that "this is ppp_a useful breakdown.

The trouble, essentially

is that, although planning and control are definable abstractions and

are easily understood as calling for different types of peppg; activity,

they do not relate to separable major categories of activities actually

carried on in an organization, either*at different times, or by

different people, or for different situations."9

Strategic planning and operational control can readily be distin-

guished from one another because in almost all complex organizations

these activities gravitate to different categories of people.

In a

business corporation the board of directors and the president would do

the strategic planning, while production foremen and clerical super-

visors would perform the duties of operational control.

In an army

strategic planning would be the responsibility of the generals and the

general staff (plus, perhaps, the political leaders of the country),

while the duties of operational control would fall to sergeants and

lieutenants.

In a communist party the central committee would take

charge of strategic planning, and operational control would be accom-

plished by cadres and cell leaders.

Management control, as Kast and

Rosenzweig imply by their term "coordinative management," is more or

less the set of management activities which are performed by the

managers in the middle, who mediate between the strategic planners

above and the operational controllers below.

As these examples demonstrate, the most successful kinds of

organizations in the modern world have tended to segregate the functions

of strategic planning, management control and operational control, and

to array the personnel charged with these activities into a vertical

451

hierarchy.

Rural cooperative organizations in several developing

countries have adopted a hierarchical, multi-tiered pattern of organi-

zation, but this has not been clearly linked to a hierarchy of manage-

ment functions.

Therein has lain a major weakness, I believe, of the

cooperative movement in most developing countries.

As early as 1895, when Frederick Nicholson published a highly

influential report on the possibilities for cooperatives in India, a

two-tiered structure for Asian rural cooperatives was proposed.

Nicholson had been deputed by the Madras Presidency to tour Europe and

study rural credit.

He returned to India enthused by the rural credit

cooperatives in Germany which had been established by Friedrich Wilhelm

Raiffeisen.

(More than half a century after'Nicholson wrote his

report, Akhter Hameed Khan, who is an astute historian, was to cite the

Raiffeisen cOOperatives as one of his inspirational models.)

Nicholson

proposed that village cooperatives should be serviced by district

cooperative banks.

The Indian Cooperative Credit Societies Act of

1904, which drew upon Nicholson's recommendations, was later transplanted

to other countries in the eastern hemisphere by the British colonial

service.10 As a consequence, a multi-tiered cooperative structure is

now the norm in many Asian countries.

Gordon Hugh Ward, a contemporary international consultant on

cOOperative organizations, advocates a

pyramidal system of cooperative

federations as the best structure for most developing countries.

At

the base of such a structure might be joint liability savings and loan

societies of 10-50 members each, organized at the village level.

The

second tier, at the county, township or market town level, would consist

of multi purpose farmers cooperatives of SOD-3,000 members.

These

two tiers would be supported by

provincial cooperative marketing and

452

purchasing federations,

and by a cooperative purchasing and marketing

organization at the national level.11

The hierarchy which Ward has in

mind, clearly, is based on a layering Of operational.functions rather

than a layering Of management functions.

Management pp the Village Tier

Robert Anthony defined Operational control as "the process Of

assuring that specific tasks are carried out effectively and

efficiently."12

That phrase describes the minimal function of a manager

of a village cooperative in the Gomilla system.

The local cooperative

manager was expected to chastise and cajole members who did not attend

weekly meetings, did not make savings deposits, or did not repay

their loans.

He was expected to see to it that the society's paperwork

was kept in order, including especially the prompt issuance of receipts

for all cash collected.

In some respects the manager of the village

cooperative was not so much an Operational controler as an Operative.

He was the trusted courier, for example, who carried money to and from

the bank at the thana training center, and he collected and disbursed

funds within his village society.

In those village cooperatives in

which the manager failed to perform his minimal functions -- sometimes

due to his disinterest but more often due to his corruption -- the

society typically fell apart and its relations with the central

association went sour.

The following two case studies, which are condensed from M. Zaker

Hussain's report on A Field Investigation into the Management pprillage

Cooperativep, illustrate the disintegration of solidarity that generally

followed a manager's deriliction Of duty.

In the first case the

453

manager was simply inattentive to administrative detail, while in the

second case the manager was corrupt.

The cooperative in South Rampur

village

was the very first to be organized in Gomilla Kotwali Thana,

and was regarded with special affection by the leaders of the coopera-

tive experiment.

Perhaps they were deluded by their nostalgia when

they made the curious decision to employ the society's upright but

incompetent manager at the thana level.

The black-listed societies,

mentioned at the end Of the South Rampur case study, were a group of 21

primary cooperatives in which management had fallen into such a bad

state by 1965 that the central association denied them irrigation water

and credit until they could bring their affairs into order.

South Rampur

The manager Of this society was respected by all

the members.

Nobody ever suspected his inte-

grity.

Members were complacent that things

would be alright so long as he was in charge.

They never bothered about receipts or entries in

their pass books.

Members repaid their loan to

him without asking for receipts.

He did not

realize the necessity Of placing the accounts

before the members so that they could discuss

them in meetings.

Members seemed to be happy

with him.

But he was withdrawn from the society

and given employment in the central association

on 28-7-62.

Trouble started from that time on-

ward.

The new manager was not given all the

records, particularly the pass books, by the Old

manager.

He could not explain the society's ac-

counts tO the members.

Accounts began to fall

into arrears.

The members now became suspicious

and they gradually took less and less interest in

the society.

The society reached a state Of

collapse.

It fell into heavy arrears with the

central association and declined to the status Of

a black-listed society.13

Bhubanghar

The Bhubanghar society was organized on 16-10-61

with 37 members.

Mr. Wahidur Rahman was its mana-

ger from the start.

He comes from a respectable

1.51,

family in the village.

It was therefore nothing

unusual that Wahidur Rahman was selected as the

manager of the society, though he does not command

as much respect as his forefathers.

People say

he married thirteen times and at present has four

wives.

The society started working fairly well.

Atten-

dance and savings were fairly regular.

The society

got two loans -- one in February '62 for the pur-

chase of fertilizer for demonstration plots in

the village and the other of Rs. 5,000 in March '62

for the production Of an aus crop.

The manager

took the entire fertilizer loan and got two shares

of the aus loan -- one in his own name and the

other in the name of his wife.

The third loan

which was issued in November '62 was also taken

by the manager for the installation Of a private

tubewell in his house.

From the beginning the

manager'attempted to derive the maximum personal

benefit from the society's loan transactions.

It was no wonder that the members lost interest

in the society.

The society reached its collap-

sing stage when Mr. Wahidur Rahman juggled the

accounts Of the society, failed to issue receipts

for monetary transactions, issued unauthorized

loans to his favorites without knowledge or

sanction Of the society, and himself defaulted in

loan repayments.

The inspector, by great persuasive efforts, did

manage to realize all the society's loan, but it

was impossible for him to revive the society

until the members made up their minds to remove

the manager’and elect a new one.

The regular

visit Of the inspector and his probe into the

irregularities committed by the manager brought

the society's problems to the surface.

The inspec-

tor educated and influenced public Opinion in the

society.

Gradually the members decided that they

could not permit Wahidur Rahman to run their

affairs.

Once the members were mobilized, they

not only removed Wahidur Rahman from Office but

resisted the disruptive tactics and family in-

fluence which he brought tp bear’against the

reconstituted cooperative. 4

In contrast to these two cases, the Rajashpur cooperative was well

managed, functioning smoothly under the methodical leadership of a

retired civil servant.

455

Rajashpur

Members are happy with the performance of the

society.

They feel they are benefiting by it.

Many members, for instance, do not have enough

land.

They take annual leases Of land with money

borrowed from the central association; they can

repay the loan with the proceeds Of one crop and

can save the second crop for their own family

consumption.

There is good leadership in the society.

Nobody

can remain absent from the weekly meeting without

permission.

If any member fails to attend due to

unavoidable circumstances he must come to the

manager the next morning and report the cause of

his absence.

In case of failure Of the absentee

member to report, the manager himself pays him a

visit.

Mr. Hasan Ali Amin, the manager of this

societm.feels strongly that a society will go bad

if the manager does not remind members to attend

meetings and does not follow up those who show

signs of indiscipline.

The manager worked as a surveyor under the Settle-

ment Department of the Government of East Pakistan.

He is a competent person, systematic and methodi-

cal in his record keeping.

He is respected by the

members.

He is considered to be an honest man.15

Zaker Hussain concluded that the best cooperative societies, such

as Rajashpur, were typified by a harsh, self-imposed discipline which

kept wavering members in line.

The members of such societies were will-

ing to undergo such discipline because their cooperative brought them

economic prosperity, and served as a forum for settling their disputes.

A necessary precondition for such solidarity and discipline was that

the society have an honest manager who was conscientious in his

adherance to recommended COOperative procedures.

The core functions Of a village cooperative manager -- such as

assuring that paperwork was completed preperly, and that cooperative

precedures were followed -- were in the realm of Operational control.

The managerial tasks of creating group solidarity and maintaining

456

discipline could also be classified as Operational control functions.

These are tasks in which military sergeants, after>all, are also expert.

In some instances the village cOOperative managers were challenged

to assume responsibilities that lay in the higher realm of management

control.

These instances arose, usually, when the village COOperative

decided to extend itself beyond participation in the credit program,

and to go into ventures such as tubewell irrigation or brick manufactur-

ing.

In chapter five I discussed the way that tubewells placed demands

upon village cooperative management to plan and construct irrigation

channels, set water rates and policies, and arrange for the supply Of

fuel.

TO be sure, none of these particular tasks appear in the middle

column of Figure 8-2, where Robert Anthony lists examples of management

control in an American business corporation.

He defined management

control as "the process by which managers assure that resources are

Obtained and used effectively and efficiently in the accomplishment of

the organization's Objectives."16 Assuring the fuel for the pump is

secured, and that pumped water is used effectively, are activities

which legitimately belong in the category Of management control.

The impact of good local management on the capability of a coopera-

tive to irrigate the village acreage is indicated by the data in

Table 8-1.

The table was constructed by combining information from

two sources.

At the conclusion of a four man field investigation into

the management of 45 village cooperatives in Gomilla thana, Zaker Hussain,

the vice chairman Of the Kotwali Thana Central Cooperative Association

and the leader Of the field research team, rated the management of the

cooperatives as good, medium, and bad.

The quality of management in

several of these local cooperative societies had previously been rated

as good or bad by Shamsul Huq, the director of the Agricultural

457

Cooperatives Federation.

Shamsul Huq did not stipulate his criteria

for judging a society's management as good or bad, while Zaker’Hussain

did.

Use of irrigation is just one Of 17 features which Zaker'Hussain

stated can be used to distinguish a good from a bad society.

In

sixteen of the villages that had been visited by the field team,

irrigation tubewells had been installed.

For two of these 16 tubewell

societies, Zaker*Hussain and Shamsul HUq disagreed in their ratings,17

so those two cooperatives, Chowara and Sreemantapur, have been excluded

from the data reported in Table 8-1.

The acreage irrigated in each of

the other 14 management-rated societies in 1965-66 was taken from the

sixth annual report of the Gomilla cOOperative experiment.18

Table 8-1:

Irrigated acreage for cooperatives with good, average and

 

pgd,management.

Average acreage irrigated for:

 

Cooperatives with:

members

members

Total

 

Cooperative

Non-

A. Good management

40.17

20.9

61.1

(N = 5)

B. Average management

24.08

18.71

42.79

(N := 6)

C. Bad management

.

17.02

5.7

22.7

(N = 3)

 

One-Tailed TyTests

 

AB: 9 degrees of freedom

BC: 7 degrees Of freedom

AC: 6 degrees Of freedom

T value

fiarobability)

1.32 (.110)

1.40 (.102)

1.32 (.118)

 

 

.28 (.393)

1.47 (.093)

2.00 (.046)

 

1.05 (.161)

1.60 (.077)

1.73 (.068)

The uniformly increasing trend from the bottom to the top of each

column of Table 8-1 suggests that the better the local cooperative

458

management, the more acreage can be irrigated by a tubewell.

The

strongest distinction shown in the table is that between the mean

irrigated acreage for non-members of cooperatives with good and bad

management (comparison AC in the middle column).

On first glance, it

might appear Odd that those who benefit most from good management inside

a village cooperative are those outside the cooperative.

But on

reflection it does make sense that a well managed cooperative would

minimize the cost of tubewell water to its members by selling as much

as possible to outsiders.

A six inch tubewell could command an irrigated

area Of 120 acres or more, and since the highest average total irrigated

acreage shown in Table 8-1 was half that, it can be inferred that the

supply of water was not the limiting factor in tubewell utilization.

The limiting factor was the quality Of cOOperative management.

The importance of good local management for village irrigation was

noted by the field research team which Zaker'Hussain led in 1966.

In my

unpublished monograph on The Gomilla System of Supervised Credit, written

a few months after our team had concluded its work, I remarked:

The research team discovered that in many coopera-

tives, managerial dishonesty and.inefficiency were

the limiting factors for the acreage covered by

tubewell irrigation.

Where the manager was incom-

petent or distrusted, there were numerous disputes

about the allocation of water, cases of water

theft by illegally cutting distributory channels,

and instances of refusal to pay for water.

Where

either the manager or'a body delegated by him had

set up a plan for water distribution, authorized

one or two men to carry out the plan by digging

channels and releasing water into the fields, and

announced a schedule for the payment of water rates,

there were few disputes and a large acreage was

brought under irrigation.

Although the manager was the most important figure in the manage-

ment of a village COOperative, others were involved.

Jockeying among

459

management personnel could paralyze an honest manager, as was the case

in Shishpur cooperative.

The chairman of the Shishpur society, accord-

ing to Zaker'Hussain's account, started formenting trouble when he grew

jealous Of the manager's success in placing two of his brothers in jobs

with the central association.

The chairman enlisted the support of

members of the rai to which the manager did not belong, and began a

campaign to unseat the manager.

As a consequence Of the prolonged

quarrel which ensued, the society's tubewell closed down altogether.19

In the second chapter I described how electoral sanctions and the

pressure of public Opinion were exerted from below against the managing

committee Of each village cOOperative.

The frequency of cooperative

meetings, which were supposed to be held in each village society once a

week, contributed to the effectiveness of these mechanisms for accounta-

bility.

Badruddin Ahmed explains:

Every functionary has to come prepared to the

weekly meeting so that he can answer to the

queries Of any member.

A manager, for example,

has to keep the account up-to-date so that any

member wishing to compare his account can do so

in the weekly meeting.

Similarly a model farmer

has to give the members the solution to a pro-

blem given him in the previous week, if he has

got it from the Thana Training and DevelOpment

Center.20

As the history Of the Shishpur society illustrates, though, pressure

from below could be a cause of disruption in cooperative management.

Sanctions exerted from above, as through the recommendations Of the

inspectors on loan limits for each society, were also subject to distor-

tion if the next higher level of management became corrupt.

Accounta-

bility was the key problem in implementing the Gomilla plan of decentral-

izating operational control of the cooperatives to the village level.

Manggement pp the Thana Tier

 

460

In their review of the Asian experience in rural development

organization, Uphoff and Esman remarked that two contradictory philo-

SOphies have been prevalent in the field Of rural development.

Accord-

ing to the paternalistic approach, the passive, fatalistic villagers

are incapable of improving their lot, and must be organized and guided

from above.

The populistic approach, on the other hand, assumes that the

people in the villages can manage their own affairs and transform their

own communities.

Students of organization theory will recognize the

similarity of these approaches to Douglas McGregor's theories of manage-

ment, theory X and theory Y.

The strength Of the Gomilla cooperatives resulted in part from their

having found an organizational home for each of these philosophies.

The village cooperatives were expected to function according to the

populistic philosophy, while the thana cooperative association saw

itself as the instigator of change in the villages and stood ready to

chastize -- sometimes even to blacklist -- errant village cooperatives

and to help them to straighten out their’affairs.

In the intermediate echelon of the inspectorate the two philosophies

met head on.

The inspectors stood midway between the village level and

the thana level cooperative organizations, and their sympathies were

Often divided between the two.

Was the inspector an organizer and

inspirational leader in the community development tradition?

Or was he

a disciplinarian and a petty bureaucrat in the tradition of South Asian

civil servants?

The cooperative association never clearly prescribed

either the proper functions or the proper allegiance of the inspectors.

Perhaps it was natural that toward the end of the decade of the sixties

461

some inspectors succumbed, and took advantage of their positions for

personal gain.

During the year that I was stationed at the Academy for Rural

Development in Gomilla, 1965-66, I paid a good deal of attention to the

inspectors.

I befriended a few inspectors and traveled with them by

night to cooperative meetings in the villages.

To Obtain a more system-

atic assessment Of the inspectors' work, I submitted a written question-

naire to all the Kotwali Thana inspectors in December of 1965, and followed

up with an interview with each inspector.

On the basis of that pilot

study of the inspectors in Kotwali Thana, I prepared a more extensive

questionnaire and asked the inspectors in Kotwali Thana and in the

expansiOn project thanas to fill it in.

The second questionnaire was

completed and returned by the inspectors in two of the three expansion

projects that had been founded in 1963-64 (those in Natore and Gaibandha

thanas) and by five Of the seven expansion projects dating from 1965-66

(those in Hajigonj, Brahman Baria, Serail, Chandina, and Laksam thanas

in Comilla district).

Taking into account the number Of inspectors who

were employed in each thana at the time, the overall response rates for

the survey were:

Gomilla Kotwali Thana, 75%; 1964 projects, 67%; and

1966 projects, 58%.

At the time of the survey, the Kotwali Thana

Central Cooperative Association was five years Old.

The Agricultural

Cooperatives Federation in Kotwali Thana had not yet been created.

The projects in Natore Thana of Rajshai District and in Gaibandha

Thana of Rangpur District (1964 projects) were three years old, and the

expansion projects in Gomilla district had been running for,nine months.

In Table 8-2 I have summarized some of the demographic characteris-

tics Of the cooperative inspectors, as determined by my survey.

The

Kotwali Thana inspectors differed in several respects from those that had

462

Table 8-2:

Background characteristics Of inspectors.

Kotwali

Thana

1964 projects

1966 projects

Inspectors

All

32 yrs.

26 yrs.

28 yrs.

28 yrs.

Mean age

Education

Up to Class IV

100%

Up to Class VII

100%

Up to Class X

100%

Matriculate

Intermediate

Graduate

Training in

Agriculture

11%

0%

0%

22%

100%

100%

100%

83%

8%

0%

100%

100%

100%

100%

42%

3%

5%

46%

100%

100%

100%

§
s
a

50%

Special training

67%

92%

60%

75%

for job of

inspector

Occupation 9:

Father

Farmer

Educator

Civil servant

Doctor

22299292232

Rural area

Urban area

Residence

Rural area

Urban area

56%

22%

11%

11%

100%

0%

67%

33%

67%

8%

17%

8%

83%

17%

42%

53%

79%

5%

16%

0%

93%

5%

74%

26%

12

19

70%

10%

13%

3%

93%

7%

63%

37%

4O

463

been hired in the newer projects.

The Kotwali thana inspectors were

Older, somewhat less educated, less likely to be trained in agriculture

or in specific preparation for their job, and.more likely to have been

born in a rural area.

These differences arose because recruitment

policies in Kotwali thana favored the hiring Of talented managers from

village cooperatives, whereas the expansion projects favored hiring

personnel from other government agencies, or young people taking their

first job.

The Kotwali thana policy had the advantage of tying the

village and the thana tiers Of the cOOperative organization more closely

together.

on the other hand, inspectors who were recruited from the

villages may have been more likely to succumb to the temptations of

favoritism or outright corruption.

Akhter Hameed Khan always opposed the local recruitment Of high

level Officers of the thana COOperative associations, whether in

Kotwali thana or elsewhere.

For fear Of the corrupting influence Of

friends and relatives upon these Officers, they were never posted to

their home thana.

Similar precautions might have been advisable for

the inspectors in Kotwali thana.

After I had completed my pilot interviews with the inspectors in

Kotwali thana, I consulted with Mr. Shamsul Huq, who was then Deputy

Project Director Of the KTCCA in charge of field supervision, to

prepare a list Of fifteen functions which the inspectors fulfilled.

On

the second questionnaire which I circulated, the inspectors were asked

to rank these fifteen functions in terms Of the amount Of time they

devoted to each function throughout the year.

An aggregate.task

ranking, together with a rank order correlation coefficient (W) was

then computed for each group of inspectors.

The low rank order correlation Of .38 for the inSpectors in the

464

Table 8-3:

Inspectors' ranking of their'activities.

Kotwali

1964

1966

thana

projects

projects

 

1

2

3

4

5

6

7

8

9

10

11

2

4,5

15

11

1

8

2

1

3

Urging society members to save more

Urging society members to buy more

share capital

Organizing new societies

14

Collecting loan repayments

9

Urging society members to adopt more

improved agricultural practices

10

Checking on loan utilization

10

4

Inspecting the books of the societies

4,5

11

Urging tubewell societies to use

more irrigation water

3

7

6

5

Urging societies to hire more tractor

services

12

Processing loan applications

7

Advising primary societies how to

apply for loans (including advice on

production plans)

12

12

13

Urging societies to engage in joint

marketing through the Central

Cooperative Association

13

14

6

Reporting grievances and demands of

the primary societies to the Officers

Of the Central Cooperative Associa-

tion

14

13

8

Organizational Paperwork (drawing up

tables, writing letters and reports,

etc.) and meetings of the Central

Cooperative Association

15

9

15

Recommending loan limits

W = 064

W = 081

W = 038

465

1966 projects was probably due to their newness on the job; having been

at work for only nine months, these inapectors had not settled into

a regular work pattern.

The inspectors in the 1964 projects, who

showed strong agreement on the ranking of their functions, emphasized

agricultural improvement more than the Kotwali inspectors.

The 1964

project inspectors spent most Of their time urging the adoption Of

improved agricultural practices, and they gave relatively high rankings

to urging cooperative members to hire more tractor services and to use

more tubewell irrigation water.

The bulk of the work of the Kotwali

inspectors, by contrast, was centered on the administration of the loan

program.

The similarity between the jobs of the inspectors and the coopera-

tive managers in Kotwali thana was striking.

The inspector seemed,

in many instances, like a back up manager for troubled cooperative

societies.

In my interviews with the Kotwali inspectors, I asked

whether they would spend more time with a society which had a bad

manager or with a society whiCh had a good manager.

.All of the nine

Kotwali thana inspectors said they would spend more time helping the

society with the bad manager.

As one inspector'put it, "I have to

spend more time with the bad manager because he cannot do everything

regularly and systematically without the help of the inspector."

In a 1964 survey, a staff researcher of the Pakistan Academy for

Rural Development asked all of the managers of the primary cooperative

societies to list their duties and responsibilities.21

Some Of the

managers' responses to this Open ended question are listed in the left

hand side of Table 8-4.

The comparable activities on the right hand

side of the table are listed as they were ranked by the Kotwali inspec-

tors in the author's survey.

466

Table 8-4:

Comparison of Managers' and Inspectors' activities in

Kotwali thana.

Managers' Activity

Description Ranking

Activity

Place in

Frequency

Order in

Inspectors'

To collect weekly savings

and/or deposit them in the

Central Bank

To teach the members improved

agricultural practices

To keep accounts and records

of the societies

To procure loans and distribute

them

TO prepare plans and programs

Of the societies

1

2

3

6

8

1

Urging society

members to save

more

5

Urging society

members to adopt

more improved

agricultural -

practices

7

Inspecting the

books of the

societies

10

Processing loan

applications

11

Advising primary

societies on how

to apply for'loans

TO maintain liaison between

9

’ 13

Reporting griev-

the Central Association and

the primary societies

ances and demands

of the primary

societies to the

Central Coopera-

tive Association

The similarity between the two lists is remarkable.

On the basis

of these surveys, one could almost characterize an inspector as a

glorified manager who was not authorized to carry cash.

An inspector

was both a teacher and a trouble shooter.

As a field worker, he was

needed mainly when irregularities arose.

If all the managers had

carried out in their village cooperatives the tasks in which they were

instructed in their weekly training classes, there would have been

little need for the inspectors to go on tour.

Good managers were

substitutes for good inspectors.

Good inspectors were also substitutes for bad managers.

As

467

disciplinary figures, the inspectors would recommend that the central

association file charges in court against managers who had stolen

money or against borrowers who had defaulted.

The inspectors did not

Often bring this residual authority into play, however, since they

preferred to mobilize group Opinion in a cOOperative against wrong-

doers.

A strong dosage of democratic correction by the cooperative

members -— the eviction of a criminal manager from his office -- could

cure the worst cases of corruption.

But if the inspector did not

influence public Opinion in the society, inciting the members to over-

throw their manager, the cooperative was likely to languish in distrust.

Given the overall similarity of the duties Of’the managers and

the inspectors, it seems reasonable to assert that the major managerial

responsibilities Of the inspectors, like those of the managers, were

in the category Of Operational control.

In accordance with our general

theory, we might anticipate that closed loop information feedback would

be an important aspect of the managerial control exercised by the

managers and inspectors.

Such was indeed the case, though more markedly

for the inspectors than for the managers.

An inspector in the Comilla cooperatives worked at the confluence

of several information feedback loops.

Many of these loops were

described in the section of chapter five which dealt with loan issue and

repayment procedures.

In drawing up draft loan limits for each society

under his supervision, for example, the inspector would lower his

proposed credit limit for societies where attendance, loan repayment,

and savings and share purchases were poor.

The resultant restriction

of cooperative credit —- which was in especially high demand in the

late sixties because of the costly input requirements of the high

yielding varieties of rice seeds -- would stimulate the malfunctioning

societies to bring themselves up to standard.

468

Audit procedures Of the central association were another example of

managerial control through information feedback, although because these

procedures were so slow-paced the inspectors tended not to be involved

in them in an on going way.

The accounts Of all the primary societies

were audited for each fiscal year by trained auditors deputed to the

Gomilla project from the COOperative department.

Since so few auditors

were deputed, and since the accounting system was itself so Byzantine,

there was usually about a one-year lag in the completion of these audits.

An audit report was written up for each society, and the manager

was required to write an audit rectification report which would show

how he had corrected each of the errors noted in the audit report.

At

the annual general meeting of the village cooperative, the inspector

for the society would read aloud both the audit report and the audit

rectification report.

Management control, as defined by Anthony, is distinguished from

Operational control in two critical respects:

a) management control

involves the acquisition and use Of resources, and b) management control

is tuned to the accomplishment of the organization's Objectives.

The upper-echelon management Of the Agricultural Cooperatives

Federation displayed both of these characteristics Of management control.

As one Of the autonomously managed profit centers in the KTCCA, the ACF

was responsible for matching costs with expenditures with the Objective

of running the rural credit Operations at as small a deficit as

possible.

Net savings, loan issue and repayment, and borrowed equity

capital were to the managers Of the ACF the essence of their organiza-

tion's cash flow.

They tried to maximize savings, share purchases, and

loan repayments, and to minimize the need for borrowing from the

469

KTCCA, in order to hold their deficit in bounds.

The quality of

management in the village societies, similarly, was to the Agricultural

Cooperatives Federation a problem Of more than philosophic importance;

a poorly managed village cooperative was a poor loan risk.

The income-earning orientation in the AgriCUltural Cooperatives

Federation had its drawbacks.

During the frantic growth year Of 1967-68,

an inordinately large quantity Of loans were issued to the primary

cOOperatives.

Badruddin Ahmed comments:

"To implement this policy

the inspectors were given immense responsibilities and were asked to

invest more to justify the viability of the Agricultural COOperatives

Federation as an independent organization capable of maintaining its

own staff.

Higher investment means higher revenue income and higher

income means higher salary."23

Aside from the Objective of achieving financial viability, the

Agricultural Cooperatives Federation shared in many of the larger

purposes of economic and social uplift Of the whole Gomilla experiment.

The women's program and the youth program, for instance, were administered

through the Agricultural Cooperatives Federation.

The major assignment

Of the ACF, however, was the administration of the rural credit pro-

gram, and in that sector the pressure for self-sufficiency was paramount.

The resemblance of the concept of management control to that of

middle management was touched upon previously in this chapter.

Nhnagement controllers usually report upwards; they are not typically

the top brass in an organization.

The major function Of the execu-

tives (the directors and deputy directors, not the inspectors) in

the Agricultural Cooperatives Federation was management control over

the rural credit program.

As with middle managers anywhere, the

inclination of these executives was to report upwards in a bureaucratic

470

hierarchy.

By and large, the administrators in the Agricultural

Cooperatives Federation saw themselves as answering to the higher level

decision-makers in the Kotwali Thana Central Cooperative Association.

Yet according to the formal organizational structure and by-laws of

the Agricultural Cooperatives Federation, this inclination was incorrect.

Because the ACF was registered as a cooperative organization, it possessed,

in accordance with East Pakistan cooperative law, a managing committee.

The majority of the members Of the managing committee were elected by

the membership.

Primary cooperatives, not farmers, were the members

of the ACF, and each village cooperative was entitled to one vote in

electing the managing committee.

(For a few years one third Of the

members of the ACF managing committee were appointed by the registrar

Of cooperative societies, who usually chose employees of the ACF or Of

the KTCCA, or faculty from the Academy for Rural Development.

The by-

laws Of the ACF were later changed to provide for the election of all

members of the managing committee.)

According to the theory of separation of powers embodied in the

cooperative law, the managing committee was to set policies for the

Agricultural Cooperatives Federation, and the ACF project director'and

his staff were responsible for implementing those policies.

That was

not the way it worked.

In a study that spanned the years from 1967 to

1970, Badruddin Ahmed, a faculty member Of the rural development

academy, reported that the managing committee rubber-stamped the

decisions of the ACF administration concerning the distribution of

loans, the allotment of new tubewells, the preparation Of the annual

budget, and the appointment of new employees in the federation.2u

According to one school of thought, it was just as well tO keep

the ACF managing committee at arms length when it came to substantive

471

decision making.

Even more than the managing committee of the primary

cooperatives, the ACF managing committee came to be dominated by larger

farmers, and by persons jointly engaged in farming and in business.

NO landless laborer or near landless farmer was ever elected to the

ACF managing committee between 1966 and 1970.

The proportion Of small

farmers owning from one to two acres -- the class of farmers whom the

cooperative system was intended to serve, in the manner of a rural

trade union protecting them from exploitation -- was never greater

than one quarter of the members of the managing committee, and fell to

17% in the committee elected for 1969-70.

In the period of the early nineteen seventies, when administrative

discipline was especially shaky throughout Bangladesh under the regime

of Sheik Majibur Rahman, the ACF managing committee exerted its influence

over the running of the federation more strongly than it had before.

The committee's tendancy was not to set general policies -- which was

the proper role envisioned for it in the cooperative law -- but to delve

into the detailed operation of the credit program.

An instance of

the resultant abuses was reported by Ali Akhtar*Khan:

There was an influential and.active manager....

He was a member Of the ACF managing committee

and of its loan committee.

He used to visit

the ACF at least once a day.

From there he

could manage a big flow Of loans to his neigh-

boring societies, tO his own society, and ul-

timately to himself and other relatives.25

As the apex organization Of the Gomilla cooperatives, the Kotwali

Thana Central Cooperative Association engaged in strategic planning,

the highest of Anthony's three levels Of management.

Policy decisions

about the scale of credit Operations, and about expansion into new

areas of economic activity, were made at the KTCCA level.

And it was

472

among the leaders Of the KTCCA, and between them and the faculty of

the Academy for Rural Development, that the great debates took place

as to where the cooperative system should turn so as to accelerate rural

development in the thana.

The contigent of foreign experts and

advisors in Gomilla, who perforce saw the larger outlines of affairs

rather than the fine details, also made their impact mainly at the

level of the KTCCA.

The Kotwali Thana Central Cooperative Association did not specialize

in strategic planning.

As the financing bank for its member federations

such as the Agricultural Cooperatives Federation and the Special

Cooperative Societies Federation, the KTCCA was mainly engaged in

management control over its loan funds.

But of all the three tiers in

the Gomilla cooperative organization, the KTCCA partook much more than

the other two in strategic planning activities.

The Academy for Rural DevelOpment at Kotbari, some seven miles

away from the COOperative headquarters complex at Abhoy Ashram, had

an important role in strategic planning which should not be overlooked.

The academy, after all, was the organization which first gave birth to

the cooperative experiment in the early nineteen sixties.

Through

their field research in the villages OfaCOmilla Kotwali Thana, the

academy faculty monitored the impact Of the cooperatives on many

aspects of rural life.

And in weekly staff meetings and yearly planning

conferences in which academy faculty and cooperative association

Officials participated, the progress and shortcomings Of the coopera-

tive organizations at each tier Of the Gomilla system were reviewed,

and proposals for improvements were set forth.

Notes for Chapter 8

The Management Subsystems

1 Herbert A. Simon, The New Science pf Management Decision, Harper

and Row) New York, 1960, p. 10.

Reprinted by permission of Herbert A.

Simon.

2 Robert N. Anthony, Planning ppg_Control Systems:

§_Framework for

Analysis, Graduate School of Business Administration, Harvard University,

BOSton, 1965, PO 550

3 Carl Sagan, The Dragon§.pp_Eden:

Speculation pp_the Evolution

prHuman Intelligence, Random House, New York, 1977, p. 70.

4 Norman T. Uphoff and Milton J. Esman, Local Orggpization pp;

Rupgl Development:

Analysis pf Asian Experience, Rural Development

Committee, Cornell University, Ithica, 1974, p. 102.

5 Anthony, 22 pip, p. vii.

Reprinted with permission.

6 Fremont E. Kast and James E. Rosenzweig, Organization egg

Manegement:

‘5 Systems Approach, second edition, MCGraw Hill Book

 

Company, New York, 1974, p. 120.

7 Ibid., p. 121.

8 Anthony, op_§ip, p. 19.

9 Ibid., pp. 10, 11.

Reprinted with permission.

10 For concise histories of the Indian COOperative movement, see

Pakistan, Asian Studies Center, Michigan State University, East Lansing,

1966, and Gordon Hugh Ward, "The Structure and Organization of

COOperatives in Developing Nations," in Mary J. McGrath, ed., Guide-

lines for Cooperppives ip_Developing Economies, International Coopera-

tive Training Center, The University Of Wisconsin, Madison, 1969.

11 Ward, opgg, p. 14.

12 Anthony, gp_p;p, p. 18.

473

1.71,

13 M. Zaker*Hussain, A Field.;nve§pigation into the Management'pp

Village Cooperatives‘gp the Gomilla Egperimental Area, Pakistan Academy

for Rural DevelOpment, Comilla, 1966, pp. 25, 2 .

Zaker'Hussain led a

four-man investigative team which visited 45 village cooperatives in

Gomilla Kotwali Thana to gather’material for this report.

The author

was one of the members of that team.

1” Ibid., pp. 60, 61. 65.

15 Ibid., pp. 47, 49.

16 Mthony, 92 92;, P' 170

17 The reader who pursues this matter and checks Chapter 2 of

Hussain's report will find that the comparison Of ratings is a bit

more complicated than the text portrays it to be.

Shamsul Huq only

rated societies as having good and.bad.management.

The list which he

gave to the field team at the start of their investigation did not

include any cOOperatives with average management.

Chowara was rated

by Shamsul Huq as having bad management, whereas Zaker Hussain placed

it in the medium category.

Sreemontopur got a good rating from Shamsul

Huq, and a medium rating from Zaker'Hussain.

18 A. Aziz Khan, A New Rur§;_COoperative System for Comilla Thana,

Sixth Annual Report, Pakistan Academy for Rural Development, Gomilla,

1967! P0 90.

19

20 Badruddin Ahmed, Manual pp_Comilla Cooperatives, Bangladesh

Academy for Rural Development, Comilla, 1972, p. 44.

21 Md. Ameerul HUq, The Characteristics Of t_h_eMana

Manasger O_f th___e_

Aggiefltural Cooperatives in Gomilla, PakistanAcademy for Rural—Develop-

ment, Gomilla, 1965.

22 Managers responding to Md. Ameerul Huq's survey could make as

many responses as they liked.

Their responses were then grouped and

ordered according to frequency.

In the author's survey, the inspectors

were asked to rank 15 tasks on a predetermined list.

.

23 Badruddin Ahmed, Who Decidep?

Role pf Managing Committee ip

5%, Bangladesh Academy for Rural Development, Gomilla, 1972. PF- 23, 24.

24 Ibid.

25 Ali Akhter Khan, Rural CredithogmeOfAAgpgcultm10o-

Opemmtive Federation, Bangladesh Academy for Rural Development, Gomilla,

1971, p. 50.

CHAPTER 9

Conclusions

The commentary in this volume has covered a fair range Of tepics,

from the musings of astrophysicist David Layzer‘about the conditions

which prevailed during the first microsecond after the expansion of the

universe began to the remarks of cooperative administratdr Zaker'Hussain

about the necessity of village organization for pump irrigation.

The

scope of the inquiry was characteristic of essays in general systems

theory, the purpose of which is to construct universal principles within

which investigations of all phenomena can be encompassed.

The main contribution to general systems theory which has been

made in this volume is a typological one.

Three major typologies were

presented and integrated:

a typological matrix of systems, a classifi-

cation Of mathematical systems, and a paradigm for the analysis-Of

organizations as a set of subsystems.

To the extent that the integra-

tion of these typologies was successful, the overall coherence of the

widespread discussion in these pages should have been apparent.

In the third chapter I referred to David Berlinski's reservations

about the explanatory power of general systems theory.

Berlinski

levelled some of his strongest criticisms against mathematical systems

theory, particularly of the sort which takes the fOrm of systems Of

differential equations.

Social systems, Berlinski contended, are far

too complicated to be modelled mathematically.

475

476

The typological innovations which were described in the third

chapter permit the fOrmulation of a well-calibrated response to critics

Of systems theory such as David Berlinski.

It is quite true that

mathematical systems theory as it presently stands is inadequate for

the representation of important aspects of organic systems.

In formal

organizations, these aspects would be the psychosocial subsystem and

the managerial subsystems.

In chapters seven and eight of this volume,

accordingly, I did not venture to present mathematical models Of the

psychosocial subsystem or of the Operational control, management control

and strategic planning subsystems in the Gomilla cOOperatives.

The

absence of mathematical models from chapter six, which dealt with

linkages between the cOOperatives and their environment, was not due

to the general unfeasability of creating such models, but to the

deficiencies in my own education.

Since the idea of linkages is derived

from the concept of channels in communications theory, I suspect that

reasonable linkage models could be created from the mathematical

material of the theory of information and communication.

Since I am

not versed in that body Of theory, I resorted to less rigorous methods

for describing and explaining organizational linkages.

The sufficiency of mathematical systems theory for modelling the

structural and technical subsystems of organizations was demonstrated

in chapters three and four.

Both the structural and the technical

subsystem models showed themselves capable of tracking actual average

data fer the cooperative expansion projects.

Some of the most specific

conclusions emerging from this study were derived from those models.

Among them were the following:

1.

Economies of scale in administration can be reversed if the

organization grows beyond a Certain size.

For the Gomilla-type thana

477

cooperative associations in East Pakistan in the nineteen sixties, the

optimum organizational size was between 50 and 70 employees, including

‘both workers and supervisors.

When thana cooperative associations grew

beyond that size, they tended to become top heavy with supervisory

personnel.

2.

The optimum rate of growth for cooperatives of the Comilla-

type was between 40 and 50 new village cooperative societies per thana

per year.

3.

A continuation of the pattern of operations of the cooperative

associations founded in Gomilla District in 1965-66 would have led to

financial disaster.

Annual profits would have ended.in 1977, after

which annual losses would have become greater each year.

(This con-

clusion, like all others recorded here, assumes a history of environ-

mental shability for the cooperative associations.

In fact their

environment became very unstable, due to the struggle for independence

and its aftermath.

The precision of the conclusion is therefore

questionable.

However, I believe that the general prediction of long-

run financial peril is valid, even in the light of environmental

turmoil.)

4.

In order to achieve a trajectory which would have led to long-

run financial health, the thana cooperative associations could have:

a) Increased their loan coverage requirements, so that more than

one fifth of the value of loans outstanding would have had to be

matched by savings and shares.

b) Held down the percentage of loans not repaid on time.

c) Increased their collection of interest on village loans to at

least twelve percent.

Such policies could also have been mixed, in combinations specified

in Table 5-11 in Chapter 5.

478

Other conclusions Of a less numerical character were derived

from the non-mathematical analyses in the preceding chapters.

Among

these conclusions were the fOllowing:

5.

The validity of the seven principles of sound rural develop-

ment organization which were pronounced by Norman Uphoff and

Milton Esman in their summary of the Cornell study of rural local govern-

ment in Asia was confirmed in the instances Of the Gomilla-type

cooperatives, and Of the Rural Public Works Program in East Pakistan in

the nineteen sixties (Chapter 2).

6.

The doomsday proposition, based on the second law of thermo-

dynamics, that closed systems tend toward increasing entrOpy was shown

to have limited applicability to biological systems and to human

organizations (Chapter 3).

7.

The dangers of excessive vertical proliferation in a rural

development organization were pointed out.

When more than two tiers Of

organizational structure are present in a local development agency such

as a cooperative network in a thana, the organizational distance between

the top leadership and the village membership may make it difficult to

control corruption at the lower tiers (Chapter 4).

8.

Several of the rural development technologies advocated by

the Comilla experimenters required cooperation and organization for

their attainment.

This was especially true Of the technologies of

dry-season irrigation and tractor cultivation (Chapter 5).

479

9.

During the nineteen sixties two credit technologies were

employed in the cooperatives in Gomilla Kotwali Thana.

These were the

technologies of supervised agricultural credit and of supervised

cooperative management.

Of these two technologies, the second is

probably more suitable for replication in other develOping countries,

since it tended to promote the long-run financial health of the parent

institution, the thana COOperative association (Chapter 5).

10.

At the village tier, the Gomilla cooperatives drew much of

their propulsive energy from the previously untapped aspirations of

small farmers, and, to a much lesser extent, of women.

Both groups

had been trapped by the pre-existent social systems, and saw the

cooperatives as a means to improve their lot (Chapters 1, 2 and 6).

11.

Successful leadership of the village cooperatives called for

treading a narrow middle ground between promoting harmonious relation-

ships with prestigious persons and established leadership groups (such

as the ppm and _s_a_m_aj_) on the one hand, and avoiding capture on the

other hand by those local economic elites whose exploitation the

COOperatives sought to displace (Chapters 2 and 6).

12.

The capacity of the Gomilla-type rural cooperatives for the

mobilization of the rural populace was limited by the lack of formali-

zation Of horizontal linkages with social groups in the villages, and

by the constriction of vertical linkages leading upwards to the thana

cooperative association (Chapter 6).

13.

The establishment of horizontal linkages with other'agencies

operating at the thana tier was the essence of the concept of the

Thana Training and Development Center.

However, the accomplishment

Of this horizontal coordination was dependent upon support from the

provincial and central governments (Chapter 6).

480

14.

By imparting organizational skills and ideology, non-fOrmal

education can play a crucial role in building rural deve10pment organ-

izations (Chapter 7).

15.

At a minimum, rural development organizations require an

operational control subsystem at the village tier’and.a management

control subsystem at the equivalent of the thana tier (Chapter 8).

16.

Those rural development organizations which include a strate-

gic planning subsystem are likely to be more flexible, creative and

successful than those which do not (Chapter 8).

Throughout this volume I have sought to weave together theory and

reality.

To those readers who are or who have been immersed in the

daily complexities of running a rural deve10pment organization, the

discourses on general systems theory and the second law of thermo-

dynamics may have seemed irrelevant.

For an organization theorist,

by contrast, the discussions Of loan issue procedures in village cooper-

atives may have seemed inapplicable to the study of the industrial

corporations which comprise his customary universe.

If this research

project has succeeded, it will have broadened the vision of both kinds

of people.

when I began this research project, I described to my father some

of the theoretical concepts which I was seeking to test against the

subject matter of the rural cooperative organizations that I had

observed in Comilla in 1965-66.

With the disguised impatience of a

parent who is trying to sound interested in his offspring's work, my

father asked, "Will the study be of any use to anyone?"

If indeed

this study will ever be of any use to anyone, it will probably be by

481

Virtue of the sixteen

conclusions listed above.

As for the utility

Of the theoretical insights reported in earlier chapters, I can only

apologize, like-any social scientist, and say that you can't build an

atomic bomb from these equations.

APPENDIX A

Mathematical Derivations

Chapter):

Reinforcement model 9: dyadic interaction

The reinforcement model of dyadic interaction presented as 3.5 can

be solved as follows.

(A.1)

Take the derivative with respect to time of the second equation,

and substitute the expression for r from the first equation, to Obtain

.0

C

2

C.

2

Y= ax= C(ay)= a

y. y-o y=0.

The characteristic polynomial of this system is seen to be

AZ- a2: 0, whichfactors to( Ate )(A - a): O.

The rootsof

the characteristic polynomial, are A = -a and A

= a .

The general

solution for x must have the form

x — C18

+ 02 e

.

a‘t

-cxt

(A.2)

c1 and 02 can be expressed in terms of the initial values of the two

state variables.

Evaluating A.2 at t = O, we find

x0 = 01 e

- c2 e

= 01 + 02, so 01 = x0 - c

2.

TO Obtain an

00

‘00

expression for 02, differentiate A.2 to obtain it = 01a eat — cza e-at.

At t = 0, this becomes r0 = clcx

- czcx.

From A.1 we know that

482

483

xo=a yo.

So wehave dye: 01a -c2Cl , or 02= 01-yo'

Substituting into A.3, c1 = x

- 01 + yo, or c1 =,--Er—-u

O

X+

0

yo

Since c2 = c1 - yb, c2 = --§-——- -y0 =

2

==——= -—-Er——n

xol+ yo

xo‘+ yo - 2yo

xO - y

The general solution for x is therefore seen to be

x + y

x

- y

_

x = -Q-2-9 out + ~74 e 0M".

A similar procedure can be used to

determine the equation for y in 3.7.

_ngg M m2_d_el_.§_ 9f: dyadic interaction

The three open system models of dyadic interaction, with step,

delayed step, and ramp input signals, can readily be solved through the

use of La Place transforms.

In the general case, a linear dynamic

system subjected to an input signal, U (t), can be restated in terms Of

La Place transforms as follows:

k.= AX (t) + GU (t) becomes

sx (s) - x (O) = AX (s) + GU (s).

(A-3)

In the second equation S is an Operator indicating the La Place trans-

form L {f (t) }= x (s) =

7 x (t) e'St dt.

Rewriting A.3 to bring

similar terms together yields:

0

sx (s) - AX (s) = x (O)+ GU (s). or

(A.4)

(SI - A) x (s) = x (o)+ GU (5).

Since A.4 is a matrix equation, it cannot be solved by dividing through

by (SI - A).

Instead, the equation may be multiplied on both sides by

(SI - A)'1, the inverse of (SI - A), to yield

484

x (s) = (SI - A)'1 x (o) + (SI - A)'1 GU (s).

(A.5)

In the three open system dyadic interaction models shown in Table 3-2,

0

CL

S

'0'

the matrix A is

'] , so (SI - A) =

It can easily be shown

a

o

'0

S

that the inVerse of this matrix is

1

Sz_a2

s

'

S

a

To evaluate the first expression on the right hand side of A. 5,

(SI - A).1 X0, all that needs to be done is to find the inverse

La Place transform of (SI - A)-1.

There are two components to this

problem:

finding the inverse La Place transform of

S

, and

s

..a

finding the inVerse La Place transform of

<1

.

Each of these

S2_a2

problems can be solved by a partial fraction expansion.

s

_

s

_

k

1

k

2

 

   

SZ_GZ-(S-a)(S+a)_S-a+s+a'

(A°6)

Multiplication of A.6 by (s +o )(s -o ) yields

S=k1(S+a)+k2(S-a).

(11.7)

To find k1, set S equal to the root with which it is associated in

A.6,S=0L.

a=k1(a+a)+k2(a-a)=2ak1,sok1=%.

Similarly, for k2, - a: k1 (-o+a ) + k2 (-a-a ) a —2 akz, so k2 = g.

Therefore,

2

2=%[si +Sia]'

 

I-

(A'B)

From a table of inverse La Place transforms, A.8 can be rewritten

485

in the time domain as L.1

S

= % (e at + e- at).

S2 _ G2

The same procedure yields

L.1

a

] =

% (eat - e-at).

(A.9)

S2

a2

Putting A.8 and A.9 together, we see that

L"1

-————~*

= f e

e

e

' e

.

(n.1o)

In the examples in Chapter 3, the parameter cxin the dyadic interaction

model was assigned the value ln2, so A.10 can be evaluated as

P

‘

2t + 2't

2t - 2't

2

2

2t - Z-t

2t + Z-t

2

2

 

_

8

. 

which is the expression for eAt in the reinfOrcement model of dyadic

interaction which was given in the text of Chapter 3 in equation 3.9.

Because the input signal Of strength 3 is applied equally to both

parties in the dyadic interaction experiments described in Chapter 3,

the coefficient matrix C in the expression (SI - A)'1 GU (s) in

equation A.5 is in all three instances the vector g = (Z .

(SI - A)‘1 G can therefore be written as

£5,176:

L::) (S) = (SB- one”) [3: 3]: zg-‘E'aym' (“1) 

due to the cancellation of the factor (S +a ).

486

La Place transfbrms Of the three input signals illustrated in the

third chapter are tabulated below.

Table A-1:

‘La Place Transforms Q§_Three Inppt Signals

 

Signal

Time Function

Lg_Place Transform

Step

U (t)

Delayed Step

u (t - a)

Ramp

r (t)

i;

e-as

s

lé

The mechanics of finding the inverse La Place transform for these

three input signals are similar, so only one case will be demonstrated,

that of a step function input.

From A.11 we can write

-

B

1

1

(SI-A)1GU(S)=S_G(1)%=STS—E—J(1).

 

(A.12)

Partial fractions may be used to expand this as follows.

k

w

S(S-a)= SE0.

+8-2-

(A'13)

g=k15+k2(S-a)

Let S

“ fie kia'kf I:—

Lets=o::>s

4‘2“ ' k2 "'

“E—

Entering these values for k1 and k2

into A.13, we find

  

Taking the inverse La Place transform,

487

L'1[ (SI - A)'1 GU (3)]

=

3—
a

eat "
e

- 1

.

To check the validity of these derivations, we may use the

reduction method which was illustrated above for the solution of the

closed system model of dyadic interaction.

As an example, let us take

the solution for the ramp input signal.

In Table 3-1 the forced

x _
solution for the ram input was given as(y) -

B

e

- at - 1

—a2 (eat _ at _ 1) .

at

Combining this with the expression for eAt derived above, the general

solution for the variable x is seen to be:

x=%(e t+eat)Xo+%(eat—eat)yo+—E—Z-(eat-at-1).

O.

..

-

(A44)

To test this solution via the reduction method, let us reduce the Open

system model with a Mp input into an equation from which the variable

y has been eliminated.

The model is

X

CY+ 8’6

y: ax+ 8t.

Take the derivative with respect to time of the first equation, and

substitute the second equation into the result in order to eliminate y.

x: ay+8=a(ax+5t)+s =a2x+ a3t+g.

35..

(fr-_- aet+8.

(n.15)

In equation A.14 we have an expression for x derived from the La Place

transform solution of the model.

That expression, and its second

derivative with respect to time, can be substituted into the left hand

side of equation A.15.

After many cancellations, the value of the

emression 'x' - a2 x turns out indeed to be

a8 t + 8' confirming the

La Place transform solution.

488

Chapter _4

Solution 39H; closed system difference equation model

th pg gguilibrium

 

point

In matrix notation, a closed system difference equation model has

the fOrm Xh+1 = B Xn + R.

A comparable change equation fbrm of the

same model is A X = AX.n + R, in which A.= B - I.

If’the model is in

a state of static equilibrium A X = 0.

At that equilibrium point in

the phase space Of the system, therefore, AX: + R.= 0, so AX;:= -R.

It

is easy to conclude that at the equilibrium point xnit = -A'

1 R.

(The

star notation indicates equilibrium.)

Consider, now, a vector comprised Of differences between the

state variables Xn, and the equilibrium vector X*.

Call this vector of

deviation scores St.

in = xn - X*.

By definition,

Al = 3cn+1 - 3tn

= (Xn+1 - X*) - (Xn - X*)

= Xn+1 _ Xn'

= A X

= AXn + R

A(’X+X*)+R

Akb+ (AX*+ R).

It was determined above that the equilibrium point X* is the point

at which AX* + R = 0, so we have

Aky= Al.

The conclusion is that the

deviation scores constitute a linear system.

That system can be solved

through application of the standard solution formula k; =BN X0, which

489

is the equivalent for difference equations Of the formula X = eAt X

o

for differential equations.

But Xn = Xn - X*. and in particular

N

’b

X

0

= X

— X*.

Therefore,

o

”
Xn=Xn-X*=B (Xo-X*), or

N

x =X*+B (x -X*).

n

0

N

Chapter 5

Calculation pf effective interest 3:31;;

,The problem is to determine the effective, or prevailing interest

rate on loans outstanding.

Let x = the effective interest rate, an

unknown.

Let y = the known amount of interest collected during the

course of a fiscal year.

Let z (t) = the total of loans outstanding

at time t.

For example,

z (t1) is the value Of loans outstanding at

time t.

Measurements Of z (t) are reported annually, so let us look at

the time interval from t

to t2, where t2 — t1 = 1, representing one

1

year.

To solve this problem it is necessary to assume that z (t) is a

function of time, i.e. ,

z = f (t).

The simplest function is a

straight line, 2: a + bt.

Over the interval from t

to t2, then,

1

t2

t2

y=tf

xz(t)dt= { x(a+bt)dt

1

1

t

t

=xa f2 dt+xb fztdt

t1

t1

=xat|2 +3_c_b_t2

t

t

1

2

t
I2

t

1

490

_
-xa(t2

_

t1)+xb(t2

2_

2

t

).

1

The expression (t22 - tlz) can be factored to (t2 - t1)(t2 + t1).

We

assumed above that t2 - t1 = 1, so we have

4
t

I

xa+ 2 (t2+ t1)

xb

x (a + 5 (t2 + t1)).

This last expression is simply x times the average of z (t1) = a + bt1

and 2 (t2) = a + bt2.

Therefore,

  x =
z(t1) + z (t2)

z—(t1)+ thz)

23’

L

=

.

 

2

To illustrate the application of this formula, values for annual

interest collections (y) and loans outstanding (z) may be extracted

from Table 5.8,xwhich shows annual average values for financial

variables in the 1966 expansion projects.

On June 30, 1967, for

instance, total loans outstanding (z) stood at 220 thousand rupees.

A

year later the figure was 594 thousand rupees.

Interest collections

during 1967-68 were 31 thousand rupees.

Plugging in the values

z (t1) = 220, 2 (t2) = 594 and y’= 31, the prevailing interest rate, x,

can easily be calculated as 7.6%.

APPENDIX B

Cooperative Association Questionnaire

To gather data for the analysis of organizational structure

presented in Chapter 4, a questionnaire was circulated to the project

directors of all of the thana cooperative associations that were

functioning in the nineteen sixties.

The text of the questionnaire

is reproduced at the end Of this appendix.

In preparing the question-

naire, the author protested and refined it by filling it in with infor-

mation recorded in the 1968-69 budget of the Kotwali Thana Central

Cooperative Association.

For Kotwali thana, the actual questionnaire responses reflect

the staffing of the field supervision section of the KTCCA prior to

1966, and Of the Agricultural Cooperatives Federation thereafter.

For

the expansion projects, the responses reflect the total staffing Of the

thana cooperative association, including that of organizational units

not directly involved in cooperative credit activities, such as the

machine section.

The four most important variables which were extracted from this

questionnaire were Workers, Divisions, Supervisors, and Levels, the

four variables of the structural control model discussed in Chapter 4.

For definitions Of these variables, and for rules as to how they should

be evaluated, I relied principally on two organization theorists:

Marshall W. Meyer and Peter M. Blau.

(See Marshall W. Meyer, "Size and

491

492

the Structure of Organizations:

A Causal Analysis," American §ggigr

logical figxigg, V. 37, August, 1972, pp. 434-441, and Peter M. Blau

and Richard A. Schroenherr, The Strpgpure‘gf Organizationg, Basic Books,

New York, 1971, especially Appendix C.)

Although the structural control

model was drawn from the work of Norman Hummon, he had taken his data

from Meyer, so I worked from Meyer's variable definitions.

Workers were defined as all non-supervisory personnel in the

thana COOperative association, including, for example, inspectors,

village accountants, drivers. mechanics. typists, and peons.

In many

cases all workers were listed in column e of question 7 of the question-

naire.

Sometimes village accountants, assistant inspectors and

inspectors had to be added in from the responses to question 4 and 5.

Peter Blau defined an organizational division as a major organi-

zational unit whose chief reports to the director or deputy director

of the overall organization.

He stipulated that in order to be

classified as a division a unit must include at least five people.

Because the thana cooperative associations were smaller than the employ-

ment security agencies which Blau studied, I modified his definition

by asserting that a division must include at least 3 employees, includ-

ing a section chief.

In the questionnaires, the divisions were usually

listed in column a Of the responses to question 7.

Supervisors were defined as including the project director, the

deputy project directors (sometimes there were several), chief

inspectors and section chiefs.

Generally, these Officers were listed

in the responses to question 3, and in columns b and d of question 7.

The hardest variable to score was levels.

Meyer counted the levels

within each organizational section, summed this count across sections,

and added one for the organizational chief to Obtain the number of

493

organizational levels.

This procedure did not seem appropriate for

the thana cooperative associations. because the only organizational

section which showed any indication Of having more than two levels

.was the field supervision section.

If the field supervision section

had three levels the whole organization was scored as having three

levels.

Otherwise, the organization was scored as having two levels.

A decision table was drawn up to clarify the scoring procedure

for levels.

Some field supervision sections were divided up into

blocks, as was the case in the Agricultural Cooperatives Federation in

Kotwali Thana.

For those thana COOperative associations and years in

which the block pattern of field organization prevailed, the blocks

were listed in response to question 8 of the questionnaire.

Two other

characteristics of the field supervision section were incorporated into

the decision table: whether the section had a designated chief, and

whether there were assistant inspectors employed in the section.

In

the decision table, a Y indicates that the characteristic was observed,

and a N indicates it was not observed.

Table B-1:

Decision table for scoring levels.

Field

Section has

Blocks

Chief

Asst Insp

Y

Y

Y

Y

N

Y

Y

Y

N

Y

N

N

N

Y

Y

N

Y

N

N

N

Y

N

N

N

 

Levels

3

3

3

2

3

2

2

2

 

 

 

 

 

 

 

 

The actual questionnaire follows.

Thana:

Information as of June 30,

 

494

Cooperative Association Questionnaire

This questionnaire has been prepared by Blake W. H. Smith, a

graduate student at Michigan State University, in connection with his

doctoral thesis research.

The research study focuses on the evaluation

of the Gomilla-type COOperative associations during the nineteen sixties.

Although a great deal of information is available at Michigan State

University on the finances Of the cooperative associations, the avail-

able literature does not describe their internal organizational

structure.

The major purpose of this questionnaire is to elicit that

infOrmation.

So as to provide a picture of how the organization of

each cooperative association changed over time, you are asked to fill in

this questionnaire 0006 for every year in which your cooperative assoc-

iation was functioning during the nineteen sixties.

With the exception

Of question 6, all answers are to be given as of June 30 of the year

indicated above.

Your cooperation in completing this questionnaire will

be appreciated.

Please try to return the completed questionnaire to

the address below before November 1, 1977.

Thank you.

Blake W. H. Smith

1312-D University Village

East Lansing

Michigan

48823

USA

1.

Person completing this questionnaire.

Name:

Title:

2.

Sources used in completing this questionnaire.

3.

Who was the Officer in charge of the thana cooperative association?

Name:

Title:

4.

How many inspectors were employed?

Chief inspectors:

Inspectors:

Assistant inspectors:

5.

How many village accountants were employed?

495

For the amon and boro seasons of the fiscal year ending on the date

entered above, approximately what prOportion of loan applications

were approved for issue by the thana cooperative association?

Answers should be for rice production loans only.

Reasonable

estimates will be sufficient.

Amon

Boro

Percentage of applications:

Percentage of loan amount requested:

496

 

 

,
0
3

e
n
u
J

f
o

s
a

n
o
i
t
a
m
r
o
f
n
I

:
a
n
a
h
T

o
D

.
a
n
a
h
T

i
l
a
w
t
o
K

a
l
l
i
m
o
G

n
i

.
F
.
C
.
A

e
h
t

n
i

e
s
a
c

e
h
t

s
a
w

s
a

,
m
i
h

y
b

d
e
d
a
e
h

n
o
i
t
c
e
s

e
v
i
t
a
r
t
s
i
n
i
m
d
a

n
a

s
a
w

o
t

s
r
e
w
s
n
a

e
h
t

n
i

l
e
n
n
o
s
r
e
p

l
e
v
e
l
-
e
g
a
l
l
i
v

r
e
h
t
o

r
o

s
e
i
t
e
i
c
o
s

e
v
i
t
a
r
e
p
o
o
c

y
r
a
m
i
r
p

f
o

s
r
e
g
a
n
a
m

e
d
u
l
c
n
i

t
o
n

e
r
e
h
t

f
i

,
l
l
e
w

s
a

d
e
d
u
l
c
n
i

e
b

d
l
u
o
w

n
o
i
t
a
i
c
o
s
s
a

e
v
i
t
a
r
e
p
o
o
c

e
h
t

f
O

r
o
t
c
e
r
i
d

e
h
T

.
s
n
o
e
p

o
t

s
r
o
t
c
e
r
i
d

.
r
a
e
y

d
n
a

a
n
a
h
t

o
t

s
a

m
e
h
t

k
r
a
m

e
s
a
e
l
p

t
u
b

,
y
r
a
s
s
e
c
e
n

f
i

s
e
g
a
p

a
r
t
x
e

e
s
U

.
9
-
7

s
n
o
i
t
s
e
u
q

y
t
u
p
e
d

m
o
r
f

y
d
o
b
y
r
e
v
e

g
n
i
d
u
l
c
n
i

,
n
o
i
t
a
i
c
o
s
s
a

e
v
i
t
a
r
e
p
o
o
c

e
h
t

f
o

s
e
e
y
o
l
p
m
e

d
n
a

s
r
e
c
i
f
f
o

l
e
v
e
l

a
n
a
h
t

e
m
i
t

-
l
l
u
f

f
o

r
e
b
m
u
n

l
a
t
o
t

e
h
t

l
a
u
q
e

d
l
u
o
h
s

"
f
"

n
m
u
l
o
c

n
i

s
e
i
r
t
n
e

e
h
t

f
o

l
a
t
o
t

e
h
T

.
3

n
o
i
t
s
e
u
q

o
t

e
s
n
o
p
s
e
r

n
i

d
e
m
a
n

r
e
c
i
f
f
o

e
h
t

o
t

y
l
t
c
e
r
i
d

d
e
t
r
o
p
e
r

h
c
i
h
w

s
n
o
i
t
c
e
s

l
a
n
o
i
t
a
z
i
n
a
g
r
o

e
h
t

e
b
i
r
c
s
e
d

d
n
a

t
s
i
l

e
s
a
e
l
P

.
7

n
o
i
t
c
e
s

n
i

n
o
i
t
c
e
s

n
i

l
e
n
n
o
s
r
e
p

y
r
o
s
i
v

l
e
n
n
o
s
r
e
p

y
r
o
s
i
v

)
.
s
R
(

f
e
i
h
c

6

+
.
d

+
>
t

=

f

e

d

c

b

l
e
n
n
o
s
r
p
p

l
a
t
o
T

 

s
e
l
t
i
t

d
'
h
a

r
e
b
m
u
N

s
e
l
t
i
t

d
n
a

r
e
b
m
u
N

y
r
a
l
a
s
—
y
l
h
t
n
o
M

f
e
i
h
c

n
o
i
t
c
e
S

 

n
o
i
t
c
e
s

n
i

-
r
e
p
u
s
-
n
o
n

f
o

-
r
e
p
u
s

r
e
h
t
o

f
o

n
o
i
t
c
e
s

f
o

e
l
t
i
T

e
m
a
N

a

n
o
i
t
c
e
S

497

 

,
0
3

e
n
u
J

f
O

s
a

n
o
i
t
a
m
r
o
f
n
I

 

:
a
n
a
h
T

f
o

y
n
a

f
I

.
7

n
o
i
t
s
e
u
q

f
o

d

n
m
u
l
o
c

n
i

d
e
t
s
i
l

e
s
o
h
t

g
n
o
m
a

n
e
e
b

e
v
a
h

d
l
u
o
h
s

s
f
e
i
h
c

n
o
i
t
c
e
s
b
u
s

e
h
T

.
e
v
o
b
a

d
e
m
a
n

s
f
e
i
h
c

n
o
i
t
c
e
s

e
h
t

f
o

y
n
a

o
t

y
l
t
c
e
r
i
d

d
e
t
r
o
p
e
r

h
c
i
h
w

s
n
o
i
t
c
e
s
b
u
s

e
h
t

e
b
i
r
c
s
e
d

d
n
a

t
s
i
l

e
s
a
e
l
p

s
n
m
u
l
o
c

n
i

s
i
h
t

n
i
a
l
p
x
e

e
s
a
e
l
p

,
s
t
i
n
u

l
a
n
o
i
t
c
n
u
f

r
e
l
l
a
m
s

o
t
n
i

d
e
d
i
v
i
d

r
e
h
t
r
u
f

e
r
e
w

s
n
o
i
t
c
e
s
b
u
s

e
s
e
h
t

t
I

.
f
"

d
n
a

l
l
e
H

,
o
s

f
I

?
s
n
o
i
t
c
e
s
b
u
s

o
t
n
i

r
e
h
t
r
u
f

d
e
d
i
v
i
d

7

n
o
i
t
s
e
u
q

o
t

e
s
n
o
p
s
e
r

n
i

d
e
t
s
i
l

s
n
o
i
t
c
e
s

e
h
t

f
O

y
n
a

e
r
e
W

.
8

s
f
e
i
h
c

s
f
e
i
h
c

n
o
i
t
c
e
s

t
i
n
u

o
t

g
n
i
t
r
o
p
e
r

~
b
u
s

o
t

g
n
i
t
r
o
p
e
r

)
.
s
R
(

f
e
i
h
c

s
f
e
i
h
c

t
i
n
u
b
u
s

f
O

s
f
e
i
h
c

t
i
n
u

f
o

n
o
i
t
c
e
s
b
u
s

f
o

e
l
t
i
T

e
m
a
N

s
e
l
t
i
t

d
n
a

s
e
m
a
N

s
e
l
t
i
t

d
n
a

s
e
m
a
N

y
r
a
l
a
s

y
l
h
t
n
o
M

f
e
i
h
c

g
g
i
t
c
e
s
b
u
S

n
o
i
t
c
e
s
b
u
S

n
o
i
t
c
e
S

f

.
9

d

2

b

a

e
b

t
o
n

d
l
u
o
c

e
r
o
f
e
r
e
h
t

d
n
a

,
n
o
i
s
i
v
r
e
p
u
s

f
O

s
l
e
v
e
l

r
u
o
f

n
a
h
t

e
r
o
m

d
e
n
i
a
t
n
o
c

n
o
i
t
c
e
s

l
a
n
o
i
t
a
z
i
n
a
g
r
o

y
n
a

f
I

.
9

.
w
o
l
e
b

n
i
a
l
p
x
e

e
s
a
e
l
p

,
8

d
n
a

7

s
n
o
i
t
s
e
u
q

o
t

s
r
e
w
s
n
a

e
h
t

y
b

d
e
b
i
r
c
s
e
d

y
l
e
t
a
u
q
e
d
a

APPENDIX C

Cooperative Credit Simulgtion

The essentials of the computer program used to simulate the

financial Operations of a thana cooperative association were explained

in Chapter 5.

A few more details and conceptual issues are discussed

here.

The program itself is then listed.

The overall sequence Of the program is that recommended by

Thomas Manetsch in systems science courses at Michigan State Univer-

sity.

(See Thomas J. Manetsch and Gerald L. Park, §y§§§p Analysis qu

Simulation pgph_Applig§pions pg Eggnomic and Social Systemg, Depart-

ment of Electrical Engineering and Systems Science, Michigan State

University, East Lansing, 1977, p. 11-14.)

Walues are first assigned

to model parameters: state variables are initialized; time is initialized

at T = 0; rate variables are initialized; output characteristics are

specified; time is updated as T = T-+ DT; state variables are computed

for T = T + DT; rate mariables are computed for T = T-+ DT; output is

printed; and the last three stages are repeated (by use of a DO lOOp)

until the desired

run length has been attained.

In this instance, all

of these steps are included in a subroutine, labelled CREDIT, so that

various parameter'values may readily be changed from run to run in the

main program.

Most of the program initialization statements have been justified

previously, in Chapter 5.

Additionally, it might be noted that the

498

499

initial values of the variables for workers and supervisors are the

same as those that were assigned at n = O in the mathematical model

presented in Chapter 4.

The initial value of the capital account,

CACCT = 61. YES estimated by taking the capital loan for 1965-66 (200),

subtracting from it the amount placed on fixed deposit (100), adding

the amount of village capital generated in 1965-66 (52), subtracting

loans issued (105), and adding loans repaid (14).

This initialization

procedure ignores operating costs and interest earnings and payments

during 1965-66, because of gaps in the available data.

Probably the

error entailed in neglecting these factors is within the range of-t 20

for the initial value of the current account.

The long run effect of

a modest change in the initial value of CACCT should be minimal so the

error, if it exists, is not of much significance.

Quite a few interest payments and collections are included in the

simulation program,

One of these, interest payments on capital loans

from the government's central bank, was set to zero in all the runs

reported in this volume.

The implications of interest charges on

capital loans could easily be explored, though, by assigning positive

values

to the capital loan interest rate, CIR.

A simplification was involved in programming interest charges and

collections on the current account.

When the current account of a thana

cooperative association was negative, indicating an overdraft, the

interest charge was as the program indicates, at 6%.

When the current

account was positive, the project directors usually shifted any excess

above immediate needs to a savings account earning 3.5% interest.

The program ignores this point, and assumes that a positive balance on

the current account earns 8% interest.

The consequences of this

simplification are minor, because even under the most favorable

500

conditions discussed in Chapter 5, such as a village loan interest

rate of 14%, the current account in the simulation was always over-

drawn after T = 3.

One way to check on the internal consistency of the model is to

see if it conserves flow.

Conservation of flow is an important systems

prOperty.

The idea is that whatever is stored in a process such as a

delay should equal what was there

initially, plus what has come in

less what has gone out.

The actual financial data fOr each of the

expansion projects that were started in Gomilla district in 1965-66,

as well as the average data for the set of seven cooperative associations,

exhibit this prOperty perfectly.

In Table C-1, as in the actual data

reported, bad debts are included in total loans outstanding (TOTLON).

In the table, VLISM is the sum of loans issued during the year; and

VLRSM is the sum of loans repaid during the year.

Table C-1:

Conservation of flow in the actual data and in the fidelity

run.

(TOTLON)n

.—.

(TOTLON)n_1

+

(VLISM)n

-

(VLRSM)n +

error

Actual average data

 

1966

91

1967

220

1968

594

1969

1293

1970

1712

=

=

=

=

=

o

91

220

594

1293

+

+

+

+

+

105

216

599

1179

1205

Simplgtion repplpg frpp the fidelipy 323

 

 

t = 1

211

t = 2

619

t = 3

1213

t: L:

1712

=

=

=

=

91

211

619

1213

+

+

+

+

203

619

1142

1196

-

-

—

-

-

-

-

-

-

14

87

225

480

786

+

+

+

+

+

o

O

o

O

O

83

+ .

O

211

+

547

727

-

-

O

1

20

501

For the first two years, the model checks out as yielding per-

fectly consistent results.

Small errors appear in the last two years,

but the largest of these is no greater than 1.2% of loans outstanding.

The simulation program itself is listed on the fOllowing pages.

C

DECK LISTING FOR THE FIDELITY RUN.

502

FTN.

LGO.

*EOR

PROGRAM COOPS(INPUT,OUTPUT)

COMMON/AA/DEP, RGROW , BETA, VLIR

COMMON/BB/NIPP, RLGTH, KPRT

0
0
0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0

PARAMETERS WHICH MAY BE CHANGED FROM RUN TO RUN.

DEP IS THE PROPORTION OF CAPITAL LOANS ISSUED WHICH IS PLACED IN

THE FIXED DEPOSIT ACCOUNT IN THE LOCAL BANK.

RGROW IS THE RATE OF GROWTH OF THE VILLAGE COOPERATIVES.

THE UNITS

OF RGROW ARE COOPERATIVES PER YEAR.

BETA IS A PARAMETER OF THE SAVINGS FUNCTION, XSAV, AND OF THE LOAN

ISSUE RATE, RLOANS.

VLIR IS THE INTEREST RATE PAID BY THE VILLAGE COOPERATIVES ON LOANS

FROM THE THANA COOPERATIVE ASSOCIATION.

DEP = 0.

RGROW = ’42 o

BETA

4. 1 7

VLIR

.09

NIPP IS THE NUMBER OF ITERATIONS (CYCLES) PER PRINT.

RLGTH IS THE LENGTH OF THE RUN, AS MEASURED IN YEARS.

ICPRT IS A PARAMETER WHICH CONTROLS PRINTING.

NIPP a 25

RLCTH = u.

1 FORMAT(1H ,*STATE VARIABLES*//)

KPRT = 1

PRINT 1

CALL CREDIT

PRINT 10

PRINT 10

2 FORMAT(1H ,*RATE VARIABLES*//)

ICPRT = 2

PRINT 2

CALL CREDIT

PRINT 10

PRINT 1o

3 FORMAT(1H ,*INTERNAL DELAY RATEs*//)

KPRT = 3

PRINT 3

CALL CREDIT

10 FORMAT(*

*)

STOP

503

END

SUBROUTINE CREDIT

COMMON/AA/DEP , RGROW, BETA, VLIR

COMMON/BB/NIPP , RLCTH, ICPRT

0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0

0
0
0
0
0
0

0
0

DT IS THE AMOUNT BY WHICH TIME IS INCREMENTED ON EACH LOOP OF THE

PROGRAM.

IN THIS MODEL THE UNIT OF TIME IS ONE YEAR.

DT== .01,

FOR EXAMPLE, INDICATES THAT THE PROGRAM CYCLES 100 TIMES IN ONE

YEAR.

DT‘= .01

INITIALIZE PARAMETERS FOR THE DELAY IN REPAYING VILLAGE LOANS.

DELI IS THE AVERAGE DELAY IN REPAYING LOANS ON TIME.

PRLATE IS THE PROPORTION OF LOANS THAT FALL OVERDUE.

DELZ IS THE AVERAGE ADDITIONAL DELAY IN REPAYING OVERDUE LOANS.

PLR IS THE PROPORTION OF OVERDUE LOANS WHICH MUST BE WRITTEN OFF.

DELI = 1.25

PRLATE = O.

DELZ = 2.

PLR=.33

INITIALIZE INTEREST RATE PARAMETERS.

CIR IS THE INTEREST RATE PAID BY THE THANA COOPERATIVE ASSOCIATION ON

LOANS FROM CENTRAL BANKS.

FIXDIR IS THE INTEREST RATE EARNED ON FIXED DEPOSITS IN THE LOCAL BANK.

CACTIR IS THE INTEREST RATE EITHER EARNED OR PAID ON THE CURRENT

ACCOUNT IN THE LOCAL BANK.

SAVIR IS THE COMPOSITE INTEREST RATE WHICH THE THANA COOPERATIVE

ASSOCIATION PAYS TO THE VILLAGE COOPERATIVES FOR SAVINGS DEPOSITS,

AND AS DIVIDENDS ON SHARES.

CIR = O.

FIXDIR = .05

CACTIR = .06

SAVIR = .035

INITIALIZE MISCELLANEOUS PARAMETERS.

MCOOPS IS THE MAXIMUM NUMBER OF VILLAGE COOPERATIVES.

ALPHA IS THE RESPONSIVENESS COEFFICIENT IN THE SAVINGS FUNCTION,XSAV.

MCOOPS = 350.

ALPHA = .27

INITIALIZE STATE VARIABLES

504

NCOOPS IS THE NUMBER OF VILLAGE COOPERATIVES IN THE THANA.

WORKRS IS THE NUMBER OF WORKERS EMPLOYED BY THE THANA COOPERATIVE

ASSOCIATION.

SUPVIS IS THE NUMBER OF SUPERVISORS EMPLOYED BY THE THANA COOPERATIVE

ASSOCIATION.

WORKRS AND SUPVIS ARE INCREMENTED ONCE EACH YEAR, AT THE BEGINNING

OF THE YEAR.

CACCT IS THE CURRENT ACCOUNT OF THE THANA COOPERATIVE ASSOCIATION

WITH THE LOCAL BANK.

FIXDEP IS THE FIXED DEPOSIT OF THE THANA COOPERATIVE ASSOCIATION IN

THE LOCAL BANK.

TSAV IS THE INTEGRAL, OR RUNNING TOTAL, OF TOTAL SAVINGS AND SHARES.

THE VARIABLES STORE1 AND STORE2 ARE INTERNAL To THE DELAY IN REPAYMENT

OF VILLAGE LOANS.

PENDNG IS THE AMOUNT OF VILLAGE LOANS NOT YET DUE.

OVRDUE Is THE AMOUNT OF VILLAGE LOANS WHICH ARE OVERDUE.

TVLOAN IS THE TOTAL OF LOANS OUTSTANDING TO VILLAGE COOPERATIVES.

BDSM IS THE INTEGRAL OF BAD DEBTS.

_

TBLOAN IS THE TOTAL OF LOANS OUTSTANDING FROM CENTRAL BANKS.

CLISM IS THE INTEGRAL, OR RUNNING TOTAL, OF CAPITAL LOANS ISSUED BY

CENTRAL BANKS.

CLRSM IS THE INTEGRAL OF CAPITAL LOANS REPAID TO CENTRAL BANKS.

VLISM IS THE INTEGRAL OF LOANS ISSUED TO VILLAGE COOPERATIVES.

VLRSM IS THE INTEGRAL OF LOANS REPAID BY VILLAGE COOPERATIVES.

VLINT IS THE INTEGRAL OF INTEREST PAYMENTS ON LOANS TO VILLAGE

COOPERATIVES.

FMSM IS THE INTEGRAL OF THE FINANCIAL MARGIN FROM THE CREDIT OPERATIONS

OF THE THANA COOPERATIVE ASSOCIATION.

OPCSM IS THE INTEGRAL OF THE OPERATING COSTS OF THE CREDIT SYSTEM.

PROFSM IS THE INTEGRAL OF PROFITS.

THE VARIABLES CLISM THROUGH PROFSM ARE SET TO ZERO AT THE BEGINNING

OF EACH YEAR.

OUTPUT FOR THESE VARIABLES REPRESENTS THE VALUE

ACCUMULATED DURING THE YEAR.

A NEW YEAR IS CONSIDERED TO HAVE

BEGUN WHENEVER THE VALUE OF T PASSES AN INTEGER VALUE.

THE UNITS OF VARIABLES CACCT THROUGH PROFSM ARE THOUSANDS OF RUPEES.

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

NCOOPS = (+0

WORKRS = 23. 33

SUPVIS = 3. 67

CACCT = 61.

FIXDEP = 100.

TSAV = 52.

STORE1 = 60.

STOREZ = 31.

PENDNG = STORE1 + STOREZ

OVRDUE = 0.

TVLOAN = PENDNG + OVRDUE

BDSM = O.

TBLOAN = 200.

CLISM = O.

CLRSM = O.

VLISM = 0.

VLRSM = O.

0
0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

505

VLINT = O.

FMSM = 0.

OPCSM = O.

PROFSM = O .

INITIALIZE TIME

T=0.

INITIALIZE RATE VARIABLES

CLI IS THE ANNUAL RATE AT WHICH CAPITAL LOANS ARE ISSUED TO THE

THANA COOPERATIVE ASSOCIATION BY CENTRAL BANKS.

CLR IS THE ANNUAL RATE AT WHICH CAPITAL LOANS ARE REPAID TO CENTRAL

BANKS.

GRANT IS THE ANNUAL GOVERNMENT SUBSIDY FOR THE ADMINISTRATIVE COSTS

OF THE THANA COOPERATIVE ASSOCIATION.

'

OPCOST IS THE ANNUAL OPERATING COST OF THE THANA COOPERATIVE

ASSOCIATION.

RLOANS IS THE AVERAGE ANNUAL RATE AT WHICH LOANS ARE ISSUED TO

VILLAGE COOPERATIVES.

VLI IS THE CURRENT RATE AT WHICH LOANS ARE ISSUED TO VILLAGE

COOPERATIVES.

THE VARIABLES RINT1, RINT2, REPAYl, AND REPAY2 RE INTERNAL TO THE

DELAY IN LOAN REPAYMENT BY VILLAGE COOPERATIVES.

VLR IS THE RATE AT WHICH VILLAGE COOPERATIVES REPAY THEIR LOANS.

ST1DOT AND STZDOT ARE USED IN CALCULATING THE RATE OF CHANGE OF THE

TWO COMPONENTS OF LOANS NOT YET DUE, STORE1 AND STOREZ.

BDEBTS IS THE RATE AT WHICH OVERDUE LOANS ARE WRITTEN OFF.

OVRDOT IS THE RATE OF CHANGE OF LOANS WHICH ARE OVERDUE.

SAV IS THE RATE OF CAPITAL FORMATION (SAVINGS PLUS SHARE PURCHASE) BY

THE VILLAGE COOPERATIVES.

CIP IS THE INTEREST PAYMENTS MADE ON LOANS FROM CENTRAL BANKS

CACTIP IS THE INTEREST EITHER EARNED OR PAID ON THE CURRENT ACCOUNT

IN THE LOCAL BANK.

FIXDIP IS THE INTEREST EARNED ON FIXED DEPOSITS IN THE LOCAL BANK.

SAVIP IS THE INTEREST PAYMENT ON SAVINGS DEPOSITS, PLUS THE DIVIDEND

ON SHARES.

’

VLIP IS THE INTEREST PAID TO THE COOPERATIVE ASSOCIATION BY VILLAGE

COOPERATIVE BORROWERS.

FMARG IS THE FINANCIAL MARGIN OF THE COOPERATIVE ASSOCIATION.

CACDOT IS THE RATE OF CHANGE OF THE CURRENT ACCOUNT IN THE LOCAL BANK.

THE UNITS OF VARIABLES CLI THROUGH CACDOT ARE THOUSANDS OF RUPEES

PER YEAR.

CLI = XCLI(T)

CLR = XCLR(T

GRANT = 50.

OPCOST = (1.4*WORKRS + 5.4*SUPVIS)/.862

RLOANS = 4.74*TSAV - 174.65

VLI = XISSUE(RLOANS,CACCT,FIXDEP)

RINT1 = STORE1/(.5*DEL1)

RINT2 = PRLATE*STORE2/(.5*DEL1)

REPAYi

(1. - PRLATE)*STORE2/(.5*DEL1)

REPAY2

OVRDUE/DEL2

506

VLR = REPAYi + REPAY2

ST1DOT = VLI - RINT1

ST2DOT = RINT1 - RINT2 - REPAYi

BDEBTS = PLR*OVRDUE

OVRDOT = RINT2 - BDEBTS - REPAY2

SAV = XSAV(TSAV,TVLOAN,NCOOPS,ALPHA,BETA)

CIP = CIR*TBLOAN

CACTIP = CACTIR*CACCT

FIXDIP = FIXDIR*FIXDEP

SAVIP = SAVIR*TSAV

VLIP = VLIR*TVL0AN

0
0
0
0
0
0
0

FMARG = VLIP - CIP - SAVIP + CACTIP + FIXDIP

PROFIT = FMARG - OPCOST + GRANT

CACDOT a (1 - DEP)*CLI - CLR - VLI + VLR + SAV-+ PROFIT

SPECIFY OUTPUT CHARACTERISTICS AND PRINT VARIABLE VALUES FOR T = 0

NIT IS THE NUMBER OF ITERATIONS OF THE PROGRAM.

NIOL IS THE NUMBER OF ITERATIONS OF THE OUTER LOOP OF THE PROGRAM,

WHICH CONTROLS PRINTING.

NIT=RLGTH DT + .00001

NIOL=NIT NIPP

Go To (19,29,39),ICPRT

19 CONTINUE

20 FORMAT (1H ,T6,*T*,T10,*NCOOPS*,T20,*TBLOAN*,T31,*CACCT*,T42,

1*TSAV* ,T51 ,*VLISM* , T61 ,*VLRSM*, T72 ,*BDSM* , T80 ,*TVLOAN* , T90,

2*PENDNG*,T100,*OVRDUE*,T111,*VLINT*,T122,*FMSM*,T130,*PROFSM*)

21 FORMAT(F6.2,I9,12F10.2)

PRINT 20

22 PRINT 21,T,NCOOPS,TBLOAN,CACCT,TSAV,VLISM,VLRSM.BDSM,TVLOAN,

1PENDNG, OVRDUE, VLINT , FMSM. PROFSM

GO TO 15

29 CONTINUE

3o FORMAT(1H ,T6,*T*,T13,*CLI*,T23,*CLR*,T30,*CACDOT*,T43,*SAV*,

1T53,*VLI*,T63,*VLR*,T70,*BDEBTS*,T82,*VLIP*,T90,*FIXDIP*,T100,

2*CACTIP*,T111,*FMARG*,T120,*OPCOST*,T130,*PROFIT*)

31 FORMAT(F6.2,F9.O,12Fio.2)

PRINT 3o

32 PRINT 31,T,CLI,CLR,CACDOT,SAV,VLI,VLR,BDEBTS,VLIP,FIXDIP,CACTIP,

1FMARG,OPCOST,PROFIT

GO TO 15

39 CONTINUE

4o FORMAT(1H ,T6,*T*,T10,*STORE1*,T20,*STORE2*,T31,*RINT1*,T41,

1*RINT2*T50,*REPAY1*.T60,*REPAY2*)

41 FORMAT(F6.2,F9.2,5F10.2)

PRINT 4o

42 PRINT 41,T,STORE1,STORE2,RINT1,RINT2,REPAY1,REPAY2

15 CONTINUE

100 FORMAT(*

*)

C

C

C

EXECUTION PHASE

ICOUNT = 1.

0
0
0

0
0
0

C

C

C

507

WORKRS

.766*WORKRS + 15.374

SUPVIS

.894*SUPVIS + 1.316

DO 11 IOL=1,NIOL

DO 10 IIL=1.NIPP

UPDATE TIME

T=T+DT

COMPUTE STATE VARIABLES FOR TIME T + DT

TBLOAN = TBLOAN + DT*CLI - DT*CLR

NCOOPS = XCOOPS (T, RGROW , MCOOPS)

TSAV = TSAV + DT*SAV

STORE1

STORE1 + DT*STlDOT

STOREZ

STOREZ + DT*STZDOT

PENDNG = STORE1 + STOREZ

OVRDUE = OVRDUE + DT*OVRDOT

TVLOAN = PENDNG + OVRDUE

BDSM = BDSM + DT*BDEBTS

CLISM = CLISM + DT*CLI

CLRSM = CLRSM + DT*CLR

VLISM = VLISM + DT*VLI

VLRSM = VLRSM + DT*VLR

VLINT = VLINT + DT*VLIP

FMSM = FPBM + DT*FMARG

OPCSM = OPCSM + DT*OPCOST

PROFSM = PROFSM + DT*PROFIT

CACCT = CACCT + DT*CACDOT

FIXDEP = FIXDEP + DT*DEP*CLI

UPDATE RATE VARIABLES FOR TIME T + DT

CLI -.- XCLI (T)

CLR = XCLR(T)

GRANT = 50.

OPCOST = (1.4*WORKRS + 5.4*SUPVIS )/ .862

PRLATE = XPER(T)

RLOANS

4. 74*TSAV - 174.65

VLI = XISSUE(RLOANS,CACCT,FIXDEP)

RINT1 = STORE1/(.5*DEL1)

RINT2 = PRLATE*STORE2/(.5*DEL1)

REPAY1 a (1. - PRLATE)*STORE2/(.5*DEL1)

REPAY2 = OVRDUE/DELZ

VLR = REPAYi + REPAY2

ST1DOT = VLI - RINT1

ST2DOT =: RINT1 - RINT2 - REPAYi

BDEBTS -.- PLR*OVRDUE

OVRDOT = RINT2 - BDEBTS - REPAY2

CIP = CIR*TBLOAN

CACTIP = CACTIR*CACCT

FIXDIP = FIXDIR*FIXDEP

SAVIP = SAVIR*TSAV

VLIP : VLIR*TVLOAN

FMARG = VLIP - CIP - SAVIP + CACTIP + FIXDIP

C

C

C

0

0
0
0
0
0

508

PROFIT

FMARG - OPCOST + GRANT

CACDOT

(1 - DEP)*CLI - CLR - VLI + VLR + SAV + PROFIT

MAKE START OF THE YEAR ADJUSTMENTS AND PRINT.

IF(T .LE. ICOUNT) GO TO 10

ICOUNT = ICOUNT + 1.

WORKRS = .766*WORKRS + 15.374

SUPVIS = .894*SUPVIS + 1.316

OPCOST = (1.4*WORKRS + 5.4*SUPVIS)/ .862

SAV = XSAV (TSAV, TVLOAN, NCOOPS , ALPHA, BETA)

CLISM = o.

CLRSM== 0.

VLISM = O.

VLRSM = o.

VLINT = o.

FMSM== 0.

OPCSM = o.

PROFSM = o.

PRINT 100

CLISM = CLISM + DT*CLI

CLRSM a CLRSM + DT*CLR

VLISM = VLISM + DT*VLI

VLRSM = VLRSM + DT*VLR

VLINT -.- VLINT + DT*VLIP

FMSM = FMSM + DT*FMARG

OPCSM = OPCSM + DT*OPCOST

PROFSM = PROFSM + DT*PROFIT

10 CONTINUE

GO TO (24,34,44),KPRT

24 CONTINUE

25 PRINT 21 ,T , NCOOPS , TBLOAN, CACCT, TSAV, VLISM, VLRSM, BDSM, TVLOAN,

1PENDNG, OVRDUE, VLINT , FMSM, PROFSM

Go TO 11

34 CONTINUE

35 PRINT 31,T,CLI,CLR,CACDOT,SAV,VLI,VLR,BDEBTS,VLIP,FIXDIP,CACTIP,

1FMARG, OPCOST , PROFIT

GO TO 11

44 CONTINUE

45 PRINT 41 ,T,STORE1,STORE2,RINT1 ,RINT2 ,REPAY1 ,REPAY2

11 CONTINUE

RETURN

END

FUNCTION XCLI(T)

THIS FUNCTION IS USED IN DETERMINING THE VALUE OF CLI, CAPITAL LOANS

TO THE THANA COOPERATIVE ASSOCIATION BY CENTRAL BANKS.

THE UNITS

OF XCLI ARE THOUSANDS OF RUPEES PER YEAR.

XCLI = o.

IF(T .LE. 4.) XCLI a 200.

RETURN

END

0
0
0
0
0

0
0
0

0

0
0
0
0
0

0

0
0
0
0
0
0

FUNCTION XCLR(T)

509

THIS FUNCTION DETERMINES THE RATE AT WHICH CAPITAL LOANS ARE REPAID

TO CENTRAL BANKS.

THE UNITS OF XCLR ARE THOUSANDS OF RUPEES PER

YEAR.

XCLR = o.

IF(T .GE. 4. .AND. T .LE. 24.) XCLR = 50.

RETURN

END

FUNCTION XCOOPS(T,RGROW,MCOOPS)

THIS FUNCTION DETERMINES THE NUMBER OF VILLAGE COOPERATIVES.

‘XCOOPS = RGROW*T-+ 41.

IF(XCOOPS .GE. MCOOPS) XCOOPS== MCOOPS

RETURN

END

FUNCTION XISSUE(RLOANS,CACCT,FIXDEP)

THIS FUNCTION DETERMINES THE RATE AT WHICH LOANS ARE ISSUED TO THE

VILLAGE COOPERATIVES.

THE UNITS OF XISSUE ARE THOUSANDS OF RUPEES

PER YEAR.

XISSUE -_- RLOANS

IF(XISSUE .LT. o.) XISSUE = o.

IF(CACCT .LE. -FIXDEP) XISSUE = o.

RETURN

END

FUNCTION XSAV(TSAV,TVLOAN,NCOOPS,ALPHA,BETA)

THIS FUNCTION DETERMINES THE RATE OF CAPITAL FORMATION (SAVINGS PLUS

SHARE PURCHASE) BY THE VILLAGE COOPERATIVES.

XSAV ASSIGNS A

VALUE TO SAV ONCE EACH YEAR.

THE UNITS OF XSAV ARE THOUSANDS

OF RUPEES PER YEAR.

-(TSAV/NCOOPS)*1000.

OL= (TVLOAN/NCOOPS)*1000.

R = 2245.

XSAV = (ALPHA*(OL - BETA*C) + R)*NCOOPS/1000.

RETURN

END

FUNCTION XPER(T)

C

THIS FUNCTION DETERMINES THE PROPORTION OF LOANS WHICH FALL OVERDUE.

IF(T .LE. 1.) XPER : o.

IF(T .GT. 1. .AND. T .LE. 2.) XPER a .08

IF(T .GT. 2. .AND. T .LE. 3.) KPm = .09

IF(T .GT. 3.) XPER = .3

RETURN

END

*EOF

BIBLIOGRAPHY

Since several fields of intellectual inquiry have been drawn

together in this book, I have made no attempt in this bibliography to

indicate all of the works which I consulted in each field, or to compile

a good list of references for each.

One Of the best bibliographies on

the Gomilla projects is to be found in Harry W. Blair's book, ng

Elusiveness o£_§pp;_y.

The volume by Manetsch and Park includes a good

list of readings on systems science.

For a bibliography on organization

theory, I would recommend Kast and Rosenzweig's text on Orgppgzations

§£9;!éfléfiélfifli-

The bibliographic listing which follows is simply a reordering,

in alphabetical sequence, Of the books and articles cited in footnotes

at the end of the chapters of this book.

I.

Books

A_New Rural Cooperative System for Gomilla Thana. Annual Reports of the

Rural Cooperative Pilot Experiment, 1961-1971.

Ahmed, Badruddin, Leadership iinillgge Coopepptiveg, Bangladesh Academy

for Rural DevelOpment, Gomilla. 1972.

Ahmed, Badruddin, Manual op_Comilla Cooperatives, Bangladesh Academy for

Rural Development, Gomilla, 1972.

Ahmed, Badruddin, Who Decides?

Role o§_Managgng Committee ip_AQE,

Bangladesh Academy for Rural Development, Comilla, 1972.

Alleyne, E. Patrick, op gl,, Trainipg gpg_Research pop_Extended Rural

Development pp Asia, Rural DevelOpment Committee, Cornell Univer-

Sity , Ithica, 9 1975.

510

511

Almond, Gabriel and Sidney Verba, The Civic Culture:

Political Attitudes

gpg Democracy pp Five Nations, Little, Brown and Company, Boston,

1965.

Anthony, Robert N. , Plannarming and Control §I_Stems: A Framework f_9_r

Mysis, Graduate Schoolof Business Administration, Harvard

University, Boston, 1965.

Ashan, Nazmul, L__q_an Utilization ‘py the Cooperative Members o__f Com___i_l__la,

Pakistan Academy for Rural Development,Gomilla. 1966.

Baker, Frank, ed. , Orgap_izational Systems, Gener§l_ Systems Approacm _to

Complex Orwzations, Richard D. Irwin, Homewood, Illinois. 1973.

Berlinski, David, On System Analsiy g: AnE8%! Concerning _t__he L__im_i__ta-

tions o__f _S___0me—Ethematical Models in—the S_o_____cial.POlitica1,and

BiologicalSciences, The M. I. T. Press,Cambridge, Mass. , 1976.

Bertalanffy, Ludwig von, General Systemg Theory, Foundations, Develop-

ment, Applications, George Brazilier, New York, 1968.

Bertocci, Peter J., Elusive Villgges:

Social Structure gpg_Community

Orgailization 1;; Rural Fast Paki tan, Ph.D. thesis. Michigan State

University, 1970.

Blair, Harry W. , The Elusiveness of Equity:

Institutional Approaches pg

Rural Development Ln Bangladesh, Rural Development Committee,

Cornell University, Ithica, 1974.

Blase, Melvin G. , Institution Buildipg:

A_ Source Book, Midwest Univer-

sities Consortium, East Lansing, 1973.

Blau. Peter M. and Richard Schoenherr, The Structure of Orggizations,

Basic Books, New York, 1971.

Carver, Fred D. and Thomas J. Sergiovanni, eds. , Or anizations and. Human

Behavior:

Focus 9_n_ Schools, McGraw Hill, New York, 1969.

Coombs, Philip H. and Manzur Ahmed, Attacking Rural Pover_ty, How Non-

Formal Education Can Help, The John Hopkins University Press.

Baltimore, 1974.

Cyert, Richard M. and James G. March, _A Behavioral Theory _fO phi Firm,

Prentice Hall, Englewood Cliffs. New Jersey,43.963

Eaton, Joseph W. , ed. , Institution Building Q5}; Development, From

Concopts 39 Applications, Sage Publications, Beverly Hills, 1972.

Etzioni, Amitai, Modern Organioations, Prentice Hall, Englewood Cliffs,

New Jersey, 1964.

Faidley, Levern W. , Computer Simulation of the Cooperative Apgoach 3o

Tractor Mechanization _ip g Developing Country, Ph.D. thesis,

Michigan State University, 1974.

 

512

Forrestzr, Jay W. , Industrial _Dyn_a_.t_n_ig§, John Wiley and Sons, New York,

19 1.

Forrester, J. W. , Principles of; Systemg, Wright Allen, Cambridge,

Mass. , 1968.

Fox, Guy H. , Final Report Covering Activities 22:. 3139, Michigan State

University Vietnam Advisory Group for the Period _Méy go, A255 -

June go, 1262, MSU Group, Saigon, F237.

Griffin, Keith, _T_h_e_ Political Economy of; Ag_ra_1_‘i_a_.r_1 Change, An_ Essay 29..

33113 green Revolution, Mamillan, London, 1971+.

Oxford University Press, New York, 1973.

**——_-

Hewes, Lawrence, Rural Development, World Frontiers, The Iowa State

University Press, Ames, Iowa, 1974.

Huq, Md. Ammerul, fig Characteristics of; 3132 Managers o_£_‘_ _th_e Agricul-

tural COO eratives _i_r_1 Gomilla, Pakistan Academy for Rural Develop-

ment, Gomilla, 19 5.

Huq, M. Nurul, Village Development ya Bangadesh (A Study 9; Monogga‘m

Village), Bangladesh Academy for Rural Development, Gomilla, 1973.

Hussain, M. Zaker, A Field Irnvestigation into the Ma___n_agement g: Village

Cooperativog _i_n_ Gomilla Egerimental Area, Pakistan Academy for

Rural Development, Gomilla, 1966.

Hussain, M. Zaker, HOJ to Orga_r_1_ize Village _G_roups for Pump Irrigation,

Pakistan Academy for Rural Development, Comilla, 1966.

Inayatullah, ed. , Cooperatives Ag Planned 9% AA Asian Rural

Communities, United Nations Research Institute for Social Develop-

ment, Geneva, 1970.

Islam, Nurul, ed. , Aggcultural Policy _iA Developing Nations, Halstead

Press, New York, 1974.

Kast, Fremont E. and James F. Rosenzweig, Orgflization And Management:

A Systems Approach, McGraw Hill, New York, 1974.

Katz, Daniel and Robert L. Kahn, Lho Social Psychology 9: Organizations,

John Wiley and Sons, New York, 1966.

Khan, Akhter Hameed, The Role Of Women in a Country'g Development,

Pakistan Academy for Rural Development, Comilla, 1963.

Khan, Akhter Hameed, Community _apn_c_1_ Aggcultural Development _i_n_ Pakistan,

Asian Studies Center, Michigan State University, East Lansing,

1969-

Khan, Akhter Hameed, Tour of; Twenty Thanas, Pakistan Academy for Rural

Development , Gomilla , 1971 .

513

Khan, Akhter Hameed, Reflectiong pg the Comillg Rural Development

 

Projects, Overseas Liaison Committee, American Council on

Education, Washington, D. C., 1974.

Michigan State University, East Lansing, 1978.

Khan, Ali Akhter, Rural_CredAp_ProgE§pme o: the Agricultural,Cooperativeg

Federation, Bangladesh Academy for Rural Development, Gomilla, 1971.

Khan, A. Z. M. Obaidullah, Ego Gomilla District Development Project,

Pakistan Academy for Rural DevelOpment, Gomilla, 196E.

Khan, Md. I. H., .932 A1,, An Evaluation Report or; the Progrogg _; 1.112

Seven Thana Projectg Under the Gomilla District Intggggted Rural

 

.Development Proggag, Pakistan Academy fOr Rural Development,

Gomilla. 1967.

Lacouture, Jean, Ho QAA_Minh:

A Political Biogpap y, Random House,

New York ' 19$ 0

Lefschetz, Solomon, Differential Eguations:

Geometric Theory, Inter-

science Publishers, New York, 1957.

Luykx, Nicolaas G. M. III, Some Comparative Aspects o§_Public Institu-

tions AA_Thailand, the Philippiooo, and Vietnam, Ph.D. thesis,

Cornell University, 1962.

McCarthy, Florence E., EEBEELL Villago Women:

Mediato£§_Between

  

Tradition AAQ_Development, Masters Thesis, Michigan State Univer-

 

sity, 1967.

McGrath, Mary J., ed., Guidelines :9; Cooperatives AA Developing

Economieg, International Cooperative Training Center, The

University of Wisconsin, Madison, 1969.

1223, International Bank for Reconstruction and Development,

__fi—

Washington, D. C., 1973.

‘

Manetsch, Thomas J. and Gerald L. Park, System Analysis god Simulation

With Applications 39 Economic ggd_Social Systems, Department of

Electrical Engineering and System Science, Michigan State

University, 197h.

Mao Tse Tung, Selected Works o: Maofggg Tung, Lawrence and Wishart,

London, 1954.

Mohsen, A. K. M., 11113 Gomilla Rural Administration Experiment, History

and Annual Report, 1962-63, Pakistan Academy for Rural Development,

Gomilla, 1963.

Mohsen, A. K. M., Report og_§ Rural Public Works Progrgm, Pakistan Academy

for Rural Development, Gomilla, 1962.

514

Mukherjee, Ramakrishna. §l§ Villgges‘o§_Benga;, Popular Prakoshan,

Bombay, 1971.

Muyeed, Abdul, Stratogies Evolved Ln a Development System Lf Planned

Social ChangeLn Rural East Pakistan: A Study Lf theProcessLf

In§titution-Builfl ng and its Integration Ln thePolitico-

Adginistrative Structure, Ph. D. Thesis, Michigan State University,

199

Niehoff, Richard 0., eds, NonsFormal Education and the Rural_POor,

Program Of Studies in Non-Formal Education, Michigan State

University, East Lansing, 1977.

Owens, Edgar and Robert Shaw, Developnont Reconsidered:

Bringing the

‘Qgp Between Government £29 People, D. C. Heath and Co., Lexington,

Mass., 1972.

Parsons, Talcott, Structure and Proce§__Ln Modern Societies, The Free

Press , (Ilencoe , 1963

Perrow, Charles, Complex Orggnizations, A Critical Essay, Scott Fresman,

Glenview, Ill., 1972.

Pike, Douglas, VLet Conng, The M.I.T. Press, Cambridge, Massachusetts,

1966.

Qadir, S. A., Villnge Danishawar, Pakistan Academy for Rural Develop-

ment, Gomilla, 19

Prome, East Pakistan 1 62:18', Pakistan Academy for Rural

——_~~

Development, Comilla, 19 3.

Raper, Arthur E., Harry L. Case, Richard.0. Niehoff, William T. Ross

and Edgar A. Schuler, Rural Developnont in Action, The Compre-

 

hensive Eyperimgnt in Gomilla, East Pakistan, Cornell University

Press, Ithica, 1970.

Sagan, Carl, The Drggons Lf Edmm: Spoculations Ln the Evolution Lf

 

Human Intelli ence, Random House, New York, 1977.

Samuelson, Paul A., FoundationglogiEconomic Anglysig, Harvard Univerh

sity Press, Cambridge, Mass., 1947.

Saucedo, Robert and Earl E. Schiring, Introduction po_Continuous And

Digital Control Systems. MacMillan, New York, 196B.

Schram, Stuart R.,‘Tng Political Thopgnt oleao Tse Tung, Frederick A.

Prager, New York, 19 3.

,

Sen, Sudhir, A Richer Harvest:

New Horizons for Developing Countries,

 

Orbis Books, Maryknoll, New York, 197#.

Simon, Herbert A., The New Science Lf Management Decision, Harper and

Row, New York, 1960.

 

515

Snow, Edgar, 5.9.4 Star Over China, Victor Bollancz, Ltd. , London, 1968.

Sobhan, Rahman, Basic Democracies, Works Proger ing Rural Develppment

 

$.11 East Pakistan, Oxford University hess, Dacca, 1968.

Spear, Percival, India, A Modern History, The University of Michigan

Press, Ann Arbor, 19 1.

Stevens, Robert D. , Hamza Alavi and Peter J. Bertocci, eds. , Rural

DevelOpment _i_n_ Eggadesh and Pakistan, The University of Hawaii,

Honolulu, 1976.

Taylor, Frederick, Scientific Management, Harper, New York, 1947.

Tepper, Eliot, Qhéggipg Patterns 9; Administpation ;g_Rural East

Pakistan, Asian Studies Center, Michigan State University, East

Lansing, 1966.

'

Thomas, D. Woods, Harry R. Potter, William L. Miller and Adrian F. AVeni,

Institution Buildipg: A Model £93 Applied Social Change,

Schenkman Publishing Company, Cambridge, Mass. , 1972.

Thomas, John W. , Rural Public Works fig East Pakistan's Develgpment,

  

Development Advisory Service. Harvard University, Cambridge,

Mass. , 19$.

Uphoff, Norman T. and Milton J. Esman, Local Organizapions for Rural

Development:

Analysis Li; p_h_e Asian Experience, Rural Development

Committee, Cornell University, Ithica, New York, 1974.

Woodward, Joan, Indugtrial Organization:

Theory and Practice, Oxford

University Press, London, 1965.

World Bank, Rural Development Sector Policy Paper, World Bank,

Washington, D. C., 1975.

Ziring, Lawrence, 113g Aflb Khan Q: Politics Ln Pakistan, 1258-1262,

Syracuse UniVersity Press, Syracuse, New York, 1971.

II . flicles

Blau, Peter M. , "A Formal Theory Of Differentiation in Organizations,"

American Sociological Review, V. 35. No. 2, April, 1970,

PP. 201-218.

Boulding, Kenneth E. , "Toward a General Theory of Growth," Canadian

Journal 9; Economic _a_._n_g Social Science, V. 19, August, 1953.

pp. 326-40.

Boulding, Kenneth E. , "General Systems Theory, the Skeleton of Science,"

Management Science, April, 1956.

516

Buckley, Walter F. , Sociology _afl Modern Systems Theog, Prentice Hall,

Englewood Cliffs. New Jersey. 1967.

Chambodal, Paul, Paradoxes 9: Physics, Transworld Publishers, London,

1971.

Childers. Grant W., Bruce H. Mayhew. Jr., and Louis N. Gray, "System

Size and Structural Differentiation in Military Organizations:

Testing a Baseline Model of the Division of Labor," American

Journal 9: Sociology, V. 76, March. 1971. pp. 813-30.

Gomilla - USAID Conference Report, Authorship and Publisher not

recorded, 1953.

Di Staffano, Joseph J., III. Allen R. Stubberud. and Ivan J. Williams.

Feedback gag Control Systems, Schaum's Outline Series, McGraw

Hill. New York. 1967.

Fairchild, Henry W., "Stopover at Gomilla, 1959-1963," Unpublished

Manuscript in the Asia Collection of the MSU Library.

Friedell. Morris F., "Organizations as Semi-lattices," American

Sociological Review, V. 32, February. 1967. pp. 46-54.

Gilbert, Richard V.. "The Works Program in East Pakistan." International

Labor Review, V. 9, No. 3, MarCh. 196k.

Haas. Eugene, Richard H. Hall. and Norman J. Johnson, "The Size of

Supportive Components in Organizations:

A Multi-Organizational

Analysis." Social Forces, V. #2, pp. 9-17.

Hagen, Everett E., "Analytical Medels for the Study of Social Systems,"

American Journal 9: Sociology, V. 67, 1961.

Harvey, Edward, "Technology and the Structure of Organizations."

American Sociological Review, V. 33. 1968. pp. 247-259.

Hickson, David J., D. S. Pugh and Diana C. Pheysey, "Operations

Technology and Organization Structure:

An Empirical Reappraisal."

Administrative Science Quarterly, V. 14, 1969.

Holdcroft, Lane E., "The Rise and Fall of Community Development,"

Department of Agricultural Economics. Michigan State UniVersity,

EaSt Lansing, 1976.

Hummon, Norman P., "A Mathematical Theory of Differentiation in

Organizations." American Sociological Review, V. 36, April, 1971.

PP} 297-303-

Hummon, Norman P., Patrick Doreian. and Klaus Teuter, "A Structural

Control Medel of Organizational Change," American Sociological

Review, V. 40, December. 1975. pp. 813~82#.

Khan, Akhter Hameed, "Presidential Address to the East Bengal Teachers'

Conference," Comilla, 20 November 1952.

517

Khan. Akhter Hameed, "Principles of Rural Development," speech

sponsored by the Asian Studies Center, Michigan State University,

June 11 9 1961".

Khan, Akhter'Hameed, "The Essence of the Gomilla Program," speech

delivered at the MIT Conference on Productivity and Innovation in

Agriculture in the Underdeveloped Countries, July, 1964.

Khan. Akhter Hameed, "Address Before the Employees of the KTCCA,"

June, 1969.

Khan. Akhter Hameed, "Learning from China, A Pakistani Experience,"

speech delivered at the International Seminar on Rural Develop-

ment, Michigan State University, 1975.

Khan, Akhter Hameed, "My Lessons in Communication," unpublished paper

presented at the Communications Institute, East West Center,

Hawaii, January, 1975.

Khandwalla. Pradip N., "Mass Output Orientation of Operations Tech-

nology and Organizational Structure," Administrative Seience

Layzer, David, "The Arrow of Time," §gientific Americgg, December, 1975.

Lyman, Princeton, "Issues in Rural DevelOpment in East Pakistan,"

Journal 9§_Compar§tive Administration, May. 1971.

Mayhew, Bruce H., Roger L. Levinger, J. Miller McPherson and Thomas F.

James, "System Size and Structural Differentiation in Formal

Organizations:

A Baseline Generator for Two Major Theoretical

Propositions," American Sociologiggl Review, V. 37: October, 1972,

pp- 629-633-

Meyer, Marshall W., "Some Constraints in Analyzing Data on Organizational

Structures:

A Comment on Blau's Paper," American Sociologiggl

Review, V. 36, April, 1971. pp. 294-297.

Meyer, Marshall W., "Size and the Structure of Organizations:

A

Causal Analysis." American Sociologigal Revigg, V. 37. August,

1972, pp. u3u-441.

Mosher, Arthur T., "Projects of Integrated Rural Development," paper

presented at the Seminar on Small Farmer DevelOpment Strategies,

Ohio State University, September, 1970.

Reimann, Bernard C., "Dimensions of Organizational Technology and

Structure:

An Exploratory Study," in Human Relations, V. 30,

NO. 6, 1977, PP. $5-566.

o

Russett, Cynthia Eagle, The Concept 9: Equilibrium in American Social

Thogght, Yale University Press, New Haven, 196M.

 

 

518

Schroede, William A. and Dan Voich, Jr., Organization and Management:

Basic Systems Concepts, Richard D. Irwin, Homewood, Illinois, 1974.

Specht, David A. , "System Size and Structural Differentiation in

Formal Organizations:

An Alternative Baseline Generator,"

American Sociological Review, V. 38, August, 1973. pp. h79-80.

Vande Ven, Andraw H. and Andre L. Delbqu, "A Task Contingent Model

of Work Unit Structure," Administrative Science Q_uarterly,

V. 19. 1974. pp. 183-197-