rt...”-
.v.”.._....

._ . . w.
“h“ “-__~ ; u
.mme-MA-Q —-- o. - 1.»;- X. If“
..-. —.- - -
A—w.:.: «v.77 v... . .-<C..._ . up. N 1 .. f-r ‘ ._ ‘ " '7‘ .T
“:1... .135. . ‘-_:u. ~ -¥ .- ‘! ; v - - - » ‘A.. R! q " ‘u 4. . ‘5! ‘ (m
~0- '4'-" -. , . . ' _ ‘ .. .— ‘1. Fun. “nth” WIN;—
”3 7~-\ - . — ._< - ; 9‘ c. A f a‘ . .2‘12. —‘
.- > ‘
:77"

2:.
5-5‘.‘.~:u‘~”“:
7. 7. _._ ...
$33.22;: ..

\ ._‘“'"

1w“
by. ‘5‘}? -hwnr~ax7fl”'i: "1 ‘ 3. .0 «- .3
- n..u' “" £.:.:“.:..ui3"‘u.ul..1.~.:;c_.;..
. .m; “'.:':~ “.1
J

 

:f “:7
A: Vdsxv' \wh cur-u -» . ~ 1
a». an “I“ v- v . awn V

l‘4:3::~ q} ‘ 4 .n» . - '7":
- .4...- m - ww 1:: ugh—r‘ a;
bu.:..::_,". .-‘ ‘-
.. ,.- _

I: §:~--

n':

)nlnn}

.. ..... .. .0“... ...... .31“:'....-~-..'...m-m ..
J u~.‘-' ~-.—o«u«h..o -5. ‘. _.\I ~.
$301252: “?:M.":":“"’ “J...“
- . , fl. .
w“..-;“‘a.'5:..:~*.t:;t,f.w,. *“‘“ ’ 7:1,...“*-' ........“““‘"... ‘Lmq . :31”. . . - ..
- ..- A» “a a T4- «4 .w- u.‘ » ~...... .4. u.» 1 a 3‘. WLAJ~ 5.... n—
. w» ... m. a. E . h-

“m“ ..

.0
" c .‘ '
. .
fluv- ~ ‘
“:4 *,..fi. .__‘ m“ -‘.

 

a. “A
9—. L.~..J.._.L‘.~.,._J--
-.~_. “41 --..~.. ; o-

m J

4
“ “u‘ztczfl‘w —
u 2-; .. ; mo.-
‘*‘~”*:*“;:, .., .. t;- .. w; .. $4.: 2...“:433‘ " ‘"
- . . - , 9- ~ Am.
.4 1210‘ . .M "g § -. u J-‘r-fi. .u- nu-u-EuAE- ;1-‘:~h”——
.- . ‘ q w‘l- .,.I'N:

-~‘c Ah.»

3.362“ M

Eh...“
~71 ;?¢::?i. ( 3% 4eufi‘.‘

s»...-
. . 6 .aui-él-n D”
:27“ fink?“ .:..::. 4.0.“ . . .31.»:
’ 25::er . . .
I. mg}
......'u‘~i.1.'
M}. .‘K
.A 3- 72:31? M5
...: a”, .
b. .-

 

 

I F" J
.u. .A .-

q.,..,x..,.-..._.

“gm...”

1‘ . .7....

w ..
::-.:v. .1. e. ..
-vv. 7'" ‘ -Lv.’ "
£315.!" 1‘ ,
Tq?~:cgr

- v -r-"
..'_.'. "EA; 3a,: :{av

 

,
O
'} Inf-tt-w- l<n:!o go?” -
.Ill“. ' ... (21,,':'.',‘r.:.!x"m l-pr
,l.l""1'a‘3"" u} hfiog-a-pwwtg
{(9.4% 3‘55?

‘ . w a g I
Hfh‘fly. .

 

Date

uuwsnsrrv uanAnles

lllllllW J \ "llllllll"‘ll"‘ J J l

 

 

 

 

 

 

This is to certify that the

thesis entitled

An Assessment of Selected Land Redistribution

Programs in Latin America, Asia and Africa and

Implications for the Analysis of Land ‘
Resettlemeent in Zimbabwe

presented by

Martha Jane Sullins

has been accepted towards fulfillment
of the requirements for

M . Sc degree in _Ag_.__EQQrL._

 

    

Major professor

5/31/91

 

0-7 639

MS U is an Affirmative Action/Equal Opportunity Institution

 

LIBRARY J
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.

TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU to An Affirmative Action/Equal Opportunity institution
7 emails-p1

LU

AN ASSESSMENT OF SELECTED LAND REDISTRIBUTION PROGRAMS IN LATIN
AMERICA, ASIA AND AFRICA AND IMPLICATIONS

FOR THE ANALYSIS OF LAND RESE'ITLEMENT IN ZIMBABWE

By

Martha Jane Sullins

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Agricultural Economics

1991

ABSTRACT
AN ASSESSMENT OF SELECTED LAND REDIS'TRIBUTION PROGRAMS IN LATIN

AMERICA, ASIA AND AFRICA AND IMPLICATIONS
FOR THE ANALYSIS OF LAND RESETTLEMENT IN ZIMBABWE

By
Martha Jane Sullins

Many Latin American, Asian and African countries have redistributed large farms
to individual smallholders to increase agricultural production, redistribute incomes and
generate employment in rural areas. Case studies from Peru, Bolivia, Indonesia and Kenya
illustrated that the following factors significantly influence the impact of land redistribution
programs: 1) site selection; 2) settler selection; 3) institutional and administrative support;
4) land acquisition and financing; 5) land tenure policy; 6) agricultural and macroeconomic
policies; and 7) scheme-level monitoring and evaluation.

In Zimbabwe, although land reSettlement has not met the initial planning targets, the
government succeeded in increasing smallholder production and redistributing income-
generating opportunities. However, resettlement’s impact varies significantly across schemes
and data indicate that it may not have improved living standards for resource-poor families.
Where success is contingent on effective planning and implementation, decisionmakers have
received little timely information on scheme-level performance, underlining the need to

reorganize and improve current monitoring and evaluation activities.

sen‘e.
metic

since:

ACKNOWLEDGEMENTS

I would like to acknowledge the contributions of my committee members; Michael
Bratton and Eric Crawford. They provided considerable insights, both in the conceptual and
final stages of this paper. I express the greatest appreciation to Richard Bemsten, who
served as my advisor, and has been an invaluable resource. He read through each draft
meticulously and spent many hours discussing the development of this paper with me. I am
sincerely grateful for his outstanding contribution.

Furthermore, I must genuinely thank my house/ apartment mates, friends and
colleagues who have offered their support, energy and ideas at every step of the way. You
are the greatest! To my parents and sister, I thank you for your never-ending love and
support. Finally, to Ben who, with the assistance of MCI and Northwest, really got me
through all of this. To a great beginning!

22

TABLE OF CONTENTS

PAGE
LIST OF TABLES ................................................... xi
CHAPTER ONE .................................................... 1
INTRODUCTION ................................................... 1
1.1 Problem Statement ................................................ 2
1.2 General Objectives ................................................ 3
1.3 Specific Research Objectives ........................................ 3
1.4 Research Hypotheses .............................................. 4
1.5 Organization of the Study ........................................... 5
CHAPTER TWO .................................................... 7
LAND REDISTRIBUTION POLICIES: THEORY AND EMPIRICAL
EVIDENCE .................................................. 7
2.1 Land Reform, Resettlement and Redistribution .......................... 8
2.1.1 The Concept of Land Reform ................................ 8
2.1.1.1 Property Rights and Land Tenure Systems ................ 9
2.1.1.2 The Relationship Between Land Reform and Land Tenure
Policies .......................................... 1 1
2.1.1.3 Goals of Distributivist Land Reform Policies ............. 12
2.1.1.4 Land Redistribution and Resettlement Policies ........... 13
2.1.1.5 Ancillary Land Policies .............................. 15
2.1.2 Land Reform and Institutional Change ........................ 16
2.1.2.2 The Impetus for Land Reform ........................ 16
2.1.2.3 Impact of Land Reform on Agriculture: The Economies of
Land Reform ..................................... 17
2.1.2.3.1 Impact on Agricultural Production .............. 19
2.1.2.3.2 Impact on Income Distribution ................ 22
2.1.2.3.3 Impact on Employment ...................... 23
2.1.3 A Conceptual Framework for the Analysis of Land Redistribution
Policies ................................................ 23
2.2 Land Redistribution Policies in Latin America, Asia and Africa ............. 24
2.2.1 Land Tenure and Agrarian Structure in Latin America ............ 27
2.2.1.1 Land Redistribution in Peru ......................... 29
2.2.1.2 Land Redistribution in Bolivia ........................ 31
2.2.2 Land Tenure and Agrarian Structure in Asia . . : ................. 33

2.2.2.1 Land Redistribution in Indonesia: Transmigration Policy . . . 34

CHAPU

D~J

23 K5'

LiCb
CHAF
RCRA
31 H}

32 AJ

33 ]

35c

CHAPTER TWO Continued

2.2.3 Land Tenure and Agrarian Structure in Africa ................... 39
2.2.3. 1 Land Redistribution in Kenya ........................ 43
2.3 Key Factors Influencing the Success of Land Redistribution Programs ........ 47
2.3.1 Site Selection ............................................. 48
2.3.2 Settler Selection .......................................... 49
2.3.3 Provision of Institutional and Administrative Support ............. 50
2.3.4 Land Acquisition and Beneficiary Payment for Land .............. 5 1
2.3.5 Land Tenure Policy and Institutional Implications ................ 53
2.3.6 Scheme-Level Monitoring and Evaluation ...................... 55
2.4 Conclusions . . .................................................. 57
CHAPTER THREE ................................................. 61
RURAL RESETTLEMENT POLICY IN ZIMBABWE ...................... 61
3.1 History of Land Policy in Zimbabwe During the Pre-Independence Period . . . . 62
3.2 Agrarian Structure and Land Distribution in Zimbabwe ................... 65
3.2.1 Large-Scale Commercial Farms .............................. 66
3.2.2 Communal Areas ......................................... 68
3.2.3 Small-Scale Commercial Farms .............................. 69
3.3.4 State Farms ............................................. 69
3.3 Post-Independence Land Policy: The Rural Resettlement Program: 1980-
1990 ........................................................ 70
3.3.1 Objectives of the 1980 Resettlement Program ................... 71
3.3.2 Resettlement Financing .................................... 73
3.3.3 Organizational Arrangement for Resettlement within the National
Government ............................................ 74
3.3.4 Criteria for Land Acquisition ................................ 76
3.3.5 Scheme-level Planning and Appraisal .......................... 77
3.3.6 Settler Selection .......................................... 78
3.3.7 Resettlement Scheme Models .......... ' ...................... 80
3.3.8 Land Tenure Policy in the Resettlement Areas .................. 83
3.4 Assessment of Land Resettlement Policy in Zimbabwe ................... 84
3.4.1 Achievement of Stated Targets ............................... 84
3.4.2 Impact on Land Distribution and Agrarian Structure ............. 87
3.4.2 Impact on Agricultural Production ............................ 91
3.4.2.1 Cropping Patterns and Aggregate Production of Major
Crops ........................................... 91
3.4.2.2 Aggregate Livestock Production ....................... 95
3.4.2.3 Regional Agricultural Crop Production ................. 97
3.4.2.4 Regional Livestock Production ........................ 101
3.4.3 Impact on Employment .................................... 103
3.4.4 Impact on Income Distribution ............................... 104
3.5 Other Key Issues ................................................ 106
3.5.1 Land Tenure ............................................ 106
3.5.2 Condition of Women: Access to Land and Production
Opportunities ........................................... 108
3.5.3 Incorporating the Landless into Resettlement Schemes ............ 110

V

CM?

3.6 St:
3.7 C01

EVAL
4.} Ex
42 C;

4.4 1
45 c

Ciikl
SUM I

(.1. u. m
'4'. L... be
’71 f'" ’17

APPE
CASE

APP};
AGR

BIBL

CHAPTER THREE Continued

3.5.4 Land Degradation and Conservation . . .' ....................... 112
3.6 Summary and Implications for Further Evaluation ....................... 113
3.7 Conclusions ..................................................... 115
CHAPTER FOUR .................................................. 1 17
EVALUATING LAND RESETTLEMENT POLICY IN ZIMBABWE .......... 1 17
4.1 Evaluating Zimbabwe’s Resettlement Program .......................... 1 17
4.2 Conceptualizing Evaluation ......................................... 1 18
4.2.1 Information-Users and Levels of Analysis ...................... 122
4.2.2 Assessing Program Performance .............................. 123
4.2.3 Assessing Program Impact .................................. 126
4.2.4 Assessing Program Efficiency ................................ 129
4.3 Developing Evaluation Techniques for Resettlement ..................... 132
4.3.1 Monitoring and Evaluation ................................. 133
4.3.2 Strengthening Monitoring and Evaluation in Zimbabwe ............ 134
4.3.2.1 Organizing the Evaluation Method ..................... 135
4.3.2.2 Data Collection ................................... 136

4.3.2.3 Incorporating Issue-Oriented Studies into the Evaluation
Methodology ...................................... 138
4.3.2.4 Decentralizing Data Collection ........................ 140

4.4 Implications for the Future Planning and Implementation of Resettlement
Policy in Zimbabwe ............................................ 141
4.5 Conclusions .................................................... 141
CHAPTER FIVE ................................................... 143
SUMMARY AND CONCLUSIONS ..................................... 143
5.1 Summary ...................................................... 143
5.2 Policy Implications ............................................... 151
5.3 Limitations of the Study ........................................... 156
5.4 Future Research ................................................. 157
APPENDIX A ...................................................... 162
CASE STUDIES .................................................... 162
APPENDIX B ...................................................... 196
AGRICULTURE SECTOR DATA FOR ZIMBABWE ...................... 196
BIBLIOGRAPHY .................................... J ............... 206

 

 

4.1
42
43

A1

A4

Re“

312

LIST OF TABLES

TABLE PAGE
2.1 Primary Characteristics of Case Studies; Peru, Bolivia, Indonesia and Kenya. . . 26

3.1 Targeted vs. Actual Number of Families Resettled on All Schemes and Land

Area Settled in Zimbabwe, 1980- 1990. .............................. 85
3.2 Land Distribution in Zimbabwe by Farm Sector and Natural Region: 1969 and

1988 ........................................................ 88
3.3 Resettlement Area Locations by Natural Region (000s ha), 1988. ........... 90
3.4 Maize and Sorghum Retention Per Resettlement Household by Natural Region

in Zimbabwe, 1983/84 - 1985/86 ................................... 99
3.5 Average Area under Cultivation for Schemes in Natural Regions II-V in

Zimbabwe, 1985 / 86 ............................................ 101
4.1 Phases in the Evaluation Process. ................................... 121
4.2 Indicators of Program Performance for Resettlement Schemes. ............ 124
4.3 Indicators of Program Impact for Resettlement Schemes. ................. 127

A1 Land Area Expropriated, Distributed and Number of Families Benefitted
During Peru’s Land Redistribution, 1964-1978. ....................... 165

A2 Land Area Distributed and Number of Families Benefitted from Bolivian
Land Redistribution, 1955-1975 .................................... 171

A3 Indonesia’s Sponsored Transmigration Program: Targeted vs. Achieved
Settlement Rates Through 1989. .................................. 179

A4 Regional Contribution of Transmigration to Incremental Rice Production by
Region in Indonesia, 1980-1984 .................................... 181

A5 Comparison of Transmigrant, Javanese and Outer Island Incomes in
Indonesia, 1984-1985. ........................................... 182

 

 

TABLE

A6 Bilge
(ch

Bl Sim-ct
193

Bi Gm

B3 Maiz
an

8.4 Sorg.‘
an

35 Flue
CI

3.6 Cor

Kl
8.7 5,,
B3 LIV!
8.9 A“

B10 A

3.11 A

TABLE PAGE

A.6 Migrant Perceptions of Welfare by Transmigration Site in Indonesia, 1985

(Percentage of Respondents). .................................... 184
B.1 Structural Change in Zimbabwe’s Large-Scale Commercial Farm Sector, 1979-

1988. ....................................................... 196
B2 Growth Rates of National Crop Production in Zimbabwe (%), 1973-1988. . . . . 197

B3 Maize Production on Zimbabwe’s Large-Scale Commercial, Communal Area
and Resettlement Area Farms, 1978-1989. ........................... 198

BA Sorghum Production on Zimbabwe’s Large-Scale Commercial, Communal Area
and Resettlement Area Farms, 1978-1989. ........................... 199

B5 Flue and Burley Tobacco Production on Zimbabwe’s Large-Scale Commercial,
Communal Area and Resettlement Area Farms, 1978-1988. ............. 200

3.6 Cotton Production on Zimbabwe’s Large-Scale Commercial, Communal Area
and Resettlement Area Farms, 1978-1989. ........................... 201

B7 Groundnut Production on Zimbabwe’s Large-Scale Commercial, Communal
Area and Resettlement Area Farms, 1978-1989. ...................... 202

B8 Livestock Production on Zimbabwe’s Large-Scale Commercial and Communal
Area Farms (0003 head), 1978-1989. ............................... 203

B9 Average Area Planted, Yield and Gross Margins for Resettlement Area Maize
Production, 1983 / 84 - 1986/86. ................................... 204

B. 10 Average Area Planted, Yield and Gross Margins for Resettlement Area
Sorghum Production, 1983 / 84 - 1986 / 86. ............................ 204

B.11 Average Area Planted, Yield and Gross Margins for Resettlement Area
Cotton Production, 1983 / 84 - 1986 / 86. ............................. 205

viii

Of

dc‘

IE

CHAPTER ONE

INTRODUCTION

Latin American and Asian countries have used land redistribution extensively in
order to establish more equitable landholding systems. The redefinition of landholding
systems through land redistribution creates new social and political power structures
designed to provide greater economic opportunity for the rural population. In Africa,
several land redistribution programs have been carried out to achieve many of the same
basic goals as Latin American and Asian reforms, but by transforming different sets of land
tenure and agricultural policy institutions.1

Overall, reforms which redistributed land resources have achieved disparate results
in terms of their impact on transforming political, economic and social conditions in
developing countries. These different results are mostly a function of the social, political
and economic context within which reforms take place, as well as varying degrees of
government commitment to create more equitable agrarian systems. In addition, the
potential economic impacts of land reform policies are often misunderstood and misstated.

This study analyzes the planning and implementation of Zimbabwe’s Rural
Resettlement policy using economic theory and empirical evidence from case studies in

Latin America, Asia and Africa regarding the potential impacts of land redistribution on

 

1 See Robert Chambers, WW (1969) for an historical

overview of early settlement schemes in Africa.

1

2

agricultural production, income redistribution and employment generation for the
resettlement sector in Zimbabwe. From this analysis, recommendations for institutionalizing
an evaluation framework for Zimbabwe’s resettlement prog'am will be made, focusing on

assessing scheme- and farm-level incentives and constraints for smallholders.

1.1 Problem Statement

The conflict between European settlers and black farmers over the allocation and
use of agicultural land has resulted in the evolution of a dual agrarian structure that still
characterizes Zimbabwe’s land distribution. From the early days of colonialism until the
Lancaster House Agreement was signed in 1980, access to prime ag'icultural land was
gradually reallocated from black farmers to a minority of large-scale commercial farmers of
European descent. These farmers also benefitted from better access to production inputs
and markets for their products than the black peasant farmers. Until 1980, the majority of
Zimbabwean farmers were restricted by law to cultivating more marginal lands in areas
designated as Communal Areas (CA’s). To redress this inequity, the government of
Zimbabwe, under the 1980 Rural Resettlement Policy, began to purchase parcels of land
from large-scale commercial farmers for redistribution to peasant smallholders, war victims
and landless people residing in the overpopulated CA’s.

Today, ten years later, critics argue that the resettlement policy has failed to meet
the government’s stated social, economic and political goals. It is important to critically
assess these claims in order to draw the insights needed to improve the future desigi and
implementation of Zimbabwe’s resettlement policy. Desigi improvements would not only
have an impact on the resettlement sector, but also on other sectors into which the country’s

scarce resources could potentially be allocated.

 

 

12 i

thee:
Acne
told
[66'
broa

003

that
pm:
on i
dez
rec
17161

C'Va

13

0f;

ant

Prc

1.2 General Objectives

This study has two general objectives. First, this study will review the economic
theory underlying land redistribution policies and the experiences of selected Asian, Latin
American and African countries that have carried out land redistribution programs, in order
to identify recommendations to improve the planning, implementation and evaluation of
resettlement in Zimbabwe. Based on this review, the study will examine some of the
broader political, social and economic issues that influence the impact of land redistribution
on agricultural production, income distribution and employment generation.

Furthermore, focusing on the Zimbabwe experience, this study will examine the
manner in which resettlement policy has been evaluated to date. First, it will show that
previous analyses are insufficient to draw reliable conclusions regarding the policy’s effects
on income distribution, improvements in social welfare, agicultural production levels, the
deg'ee of profitability for resettlement farmers and possible alternative uses for the
resources invested into the program. Second, it will show that there are more effective
methods for evaluating resettlement in Zimbabwe, and propose an alternative method of

evaluation to analyze change in the resettlement sector.

13 Specific Research Objectives

The specific objectives of this study are to:

1. Clarify the role of land resettlement and redistribution policies within the context
of agrarian reform; highlight the general goals and objectives associated with these policies;
and examine the theory regarding economic impacts of land redistribution on ag'icultural

production, income distribution and employment generation.

4

2. Examine the performance of land resettlement and redistribution policies in
selected countries in Latin America, Asia and Africa, and highlight the findings from these
studies that are relevant for planning, implementing and evaluating the impact of
resettlement in Zimbabwe.

3. Review the history of land policy in Zimbabwe; the goals, development and
implementation of the rural resettlement program; the ag'arian structure these policies have
created; and examine the national and farm-level impacts of the resettlement program.

4. Survey various methods of project evaluation and examine their advantages,
limitations, data requirements and relevance for evaluating resettlement programs.

5. Propose specific recommendations for the monitoring and evaluation of current
schemes and for improving the planning and implementation of Zimbabwe’s land

resettlement program.

1.4 Research Hypotheses

The following hypotheses will be tested in this study:

1. Land reform, carried out through the redistribution of land and property rights,
involves changing the institutional arrangements governing people’s access to land
resources.2 The manner in which these institutions should be altered depends on the
current distribution of resources (benefits) to all members of society and how alternative
arrangements will affect this distribution according to desired land reform goals.

2. A resettlement program’s success or failure should be measured by evaluating its

contribution in achieving the goals and objectives established for that progam. This implies

 

2 'Institutions are sets of ordered relationships among people which define their rights,
exposure to the rights of others, privileges, and responsibilities" (Schmid, 1972).

5

that a program’s performance cannot be compared using a uniform common denominator
since each mom is carried out under diverse social, economic and political conditions.
Rather, program performance must be analyzed within the specific context of that country.

3. The manner in which Zimbabwe’s land resettlement policy has been analyzed to
date has provided insufficient insight into the policy’s impact on ag'icultural production,

income distribution, employment generation and settler welfare.

1.5 Organlntion of the Study

This study is organized into five chapters. Chapter One introduces the study
objectives, hypotheses and organization of the thesis. Chapter Two contains four sections.
The first section defines the concepts used in this paper, and discusses the role of property
rights in determining income distribution. The second section outlines the theoretical
arguments regarding the potential impacts of redistributive reforms on relevant indicators
of performance. The third section examines four case studies; two land redistribution
programs from Latin America (Bolivia and Peru), a colonization program from Asia
(Indonesia) and one land redistribution progam from Africa (Kenya). A summary of each
study is presented according to the format given above and a more detailed description of
each case is located in Appendix I. The conclusion to this chapter outlines the most
relevant issues for Zimbabwe’s resettlement progam. These issues are examined in depth
in the third section of Chapter Three.

Chapter Three is divided into three sections. The first section examines the
historical evolution of land policy in Zimbabwe and the objectives, goals and subsequent
implementation of resettlement policy. The second section analyzes the impact of

resettlement on land distribution, crop and livestock production, income distribution and

6

employnnent generation and examines other important issues related to the implementation
of resettlement policy. The third section evaluates the success of the resettlement policy in
meeting its objectives, and examines the degree to which current evaluation methods provide
policymakers with accurate interim information on scheme-level problems and progress.

Chapter Four examines the elements involved in evaluating rural development
programs; discusses the importance of monitoring and evaluation systems in assessing the
long-term performance of resettlement programs; and addresses the need to improve the
evaluation of Zimbabwe’s resettlement prog'am by making specific recommendations.

Chapter Five contains a summary of the study, recommendations to resolve
identified constraints for Zimbabwe’ resettlement program, limitations of the study and

issues warranting further research.

 

IAVD i

Cha
reform. lan
of prepen}
economy
impacts of
PTOduction
Cacti of wt
Cacti refer
from this]

CHAPTER TWO

LAND REDISTRIBUTION POLICIES: THEORY AND EMPIRICAL EVIDENCE

Chapter Two comprises four sections. The first section defines the concepts of land
reform, land resettlement and redistribution, as used in this paper, and discusses the role
of property rights in determining the pre- and post-reform distribution of income in the
economy. The second section outlines the theoretical arguments regarding the potential
impacts of redistributive reforms on relevant indicators of performance (i.e., ag'icultural
production, income distribution and employment). The third section presents case studies,
each of which is prefaced by a description of the general institutional factors influencing
each reform (i.e., land tenure and other relevant policies and the agrarian structure resulting
from this policy mix). These studies examine land redistribution policies in Peru, Bolivia,
Indonesia and Kenya, according to the framework developed in the second section. Each
case is summarized in this chapter and a detailed study is located in Appendix A. The final
section analyzes the primary factors influencing the performance of land redistribution
programs, and synthesizes the relevant points from each study in order to examine how well
the empirical results support theoretical arguments regarding the potential impacts of land
distn'bution and land tenure policy changes.

8
2.1 Land Reform, Resettlement and Redistribution

2.1.1 The Concept of Land Reform

In order to study land reform policies and the potential benefits and costs realized
from their implementation, one must first establish a working definition of land reform.
Since there are many definitions of land reform, the following sections develop a definition
to both clarify the meaning of land reform and its relationship to resettlement and
redistribution policies, and to lay the foundation for the conceptual framework used to

analyze the case studies.

Historically, the term land reform referred to the redistribution of land from large
landholders to individual small farmers or to cooperative groups. Over time, however, the
concept has been broadened to include reforms in land tenure and other policies to provide
smallholders with greater access to resources such as credit, education and marketing within
the ag'icultural sector (King, 1977).

Raup (1967) attributes the evolution of the concept of land reform to changes in the
forces affecting agriculture in this century. For example, prior to the early 1900s when
farmers employed relatively uniform production techniques, land redistribution was used to
redress political and social inequities, with little regard for the effects of land transfer on
agricultural production. However, growing population pressure on land resources has
shifted the emphasis of land reform towards improving economic efficiency, often with
concurrent emphasis on social equity considerations (Raup, 1967).

Fundamentally, the purpose of land reform is to alter or transform a country’s
agarian structure when an undesirable imbalance or inequity is perceived. An agrarian
structure is defined as "a system of social relations (modes of production and their

corresponding social class composition) and a system of land tenure (ownership and usufruct

I‘D

9
of land and water by farm sizes)” (de Janvry, 1984: 264). Following this definition, a land

reform could alter social structure, ag'icultural production relations and rights to land
resources depending on the type of landholding system and how it influences the modes of

ag'icultural production and the social structure in a specific country.

2.1.1.1 Property Rights and Land Tenure Systems

Fundamentally, the specification of property rights determines the respective
opportunity sets of resource users since these rights influence the extent of an individual’s
control over productive resources and his/ her ability to generate income and other values
from those resources (Schmid, 1987). Similarly, each tenure system is governed by sets of
property rights which influence people’s access to resources and the distribution of income
in an economy.

The distinction between use and exchange rights in property institutions is critical
in analyzing the potential distributional impacts resulting from changes in tenure
arrangements. Use rights permit an individual to utilize a resource, but provide no legal
means of transferring that right to another individual. Exchange rights allow individuals to
transfer or withhold resources from others. This distinction of rights has implications for
the distribution of wealth in an economy. For example, in a nontransfer (nonmarket)
economy, differences in wealth are limited because individuals cannot gain market income
by withholding or transferring resources, and wealth creation becomes more a function of
individual skill and knowledge than economic advantage. In an exchange (market) economy,
however, there can be vast differences in wealth among individuals (Schmid, 1987).

The distinction between use and exchange rights is important because under certain

conditions it may be possible to redistribute land without extending exchange rights to

10
reform beneficiaries, particularly given empirical findings which suggest that individualized
tenure systems do not always lead to g'eater efficiency in agriculture (Cohen, 1980; Hunter
and Mabbs-Zeno, 1986; Barrows and Roth, 1989). In other words, changing land use rights
through land redistribution may be sufficient to accomplish desired political, social and
economic objectives without redefining exchange rights. These findings will be further
examined in section 2.2.3 of this chapter.

Each tenure system is defined by the manner in which rights to land are allocated
to individuals by institutions. Those individuals’ opportunity sets (or potential options to
use land resources) are formed by the rules of that particular tenure system. Okoth-Ogendo
proposes the following:

“a tenure system... does not merely describe an isolated aspect of the economy of a

society it prescribes the degree of control that may be exercised over land

resources and consequently, circumscribes the manner in which they may be used
and the manner in which the benefits accruing therefrom may be distributed.

Further, a tenure system summarizes the set of relations which emerge through the

power processes of society.“ (1976: 152).

Dorner further asserts that ”the dimensions and future security of opportunities are critically
affected by labor, capital, and product markets thus, the land tenure system interrelates
with a wide range of other institutions” (1971: 15).

Although this paper discusses various tenure systems, including communal, freehold,
tenant-farming and cooperative arrangements, it focuses on freehold and communal tenure
systems, their effect on individual incentives and the degree to which they influence the
impact of land redistribution on agricultural production, income distribution and
employment. Brief definitions are given below to familiarize the reader with the general
characteristics of communal and freehold systems.

Communal tenure systems allocate land use rights to individuals, but no right of

exchange. Bruce describes many different communal landholding patterns, including: 1)

11

"common ownership that implies common exploitation and management;" 2) “the right of
each member of a goup to use independently the full extent of certain land of the group;"
or 3) "significant group control, reflecting some group interest, over land that is apportioned
for the relatively exclusive use of individuals or families of the group“ (1988: 24-25). Several
authors argue that the lack of tenure security associated with communal systems reduces
farmers’ incentives to invest in agiculture, as well as limits their ability to obtain credit
(Barber, 1970; King, 1977; Feder and Noronha, 1987).

In contrast, freehold or individualized tenure provides both use and exchange rights
to individuals, either through leasehold arrangements with the state or the registration and
allocation of full land ownership to the individual. The most common form of freehold
ownership is the small farm unit where the means of production and rights to the output
belong uniquely to the family. Mellor (1967) finds variants of the freehold system where
the landowner controls the production process but does not contribute the labor inputs. For
example, in some Latin American and Asian tenure systems, large landowners rent land to
or make output-sharing agreements with tenant farmers. Freehold tenure systems are
presumed to offer the greatest incentive to the owner to invest in his/ her operation and
produce the most efficient allocation of resources because rights to land are clearly defined
and legally enforceable and the costs and rewards of production are internalized by the

owner (Barrows and Roth, 1989).

2.1.1.2 The Relationship Between Land Reform and Land Tenure Policies
King (1977) defines land reform as "land tenure reform,” which involves either land
redistribution (a change in size of holdings and land use rights) or tenancy reform

(improvements in exchange rights, but no change in land distribution). King maintains that

12

while measures other than redistribution or tenancy reform (such as consolidation of
landholdings or land tax reform) may affect one aspect of the agrarian structure, they do not
have a significant, long-term effect on the whole structure because there is no direct change
in individuals’ rights to land.

As previously asserted, the purpose of a land reform is to change a country’s
agrarian structure in response to perceived inequities in the landholding system. Therefore,
only a change in use or exchange riglnts would alter the landholding system. Although land
may be redistributed either to individual smallholders (distributivist reform) or to collective
groups (collectivist reform) (Lipton, 1974), this study will examine only distributivist reforms,

where individuals’ use or exchange rights to land are changed.

2.1.1.3 Goals of Distributivist Land Reform Policies

The goals of land reform policies shape the redistribution of resources and the
reorganization of agricultural production relations within the agrarian sector. Policy goals
also influence the degree to which accompanying institutional changes are carried out (for
example, reforms to marketing, credit or extension institutions). The case studies in
Chapter Two reveal that although each reform program was based on a broad range of
objectives, these diverse objectives can be g'ouped into three categories leading to the
attainment of political, social equity and economic efficiency goals.

A common political goal is to create a more stable political environment, as in
Bolivia’s attempt to reduce rural political instability by destroying the large landowning class.
Increasing social equity emerges an important goal in every land redistribution case studied.
Particularly in the cases of Bolivia, Kenya, Indonesia and Zimbabwe, land was redistributed

to smallholders in order to improve their access to land resources and redistribute incomes

 

 

eflijc

hddi

hveh
cfikic
redst

i0 COL

lLlA

inlan
lesir
Place.
avafia
hndh
thee;
lbr b
infll

devek

[algae

13
in rural areas. This goal was sometimes reinforced by policies to improve rural dwellers’

access to social services and infrastructure. In addition, improving or maintaining economic
efl'iciency (in terms of aggregate output) is often of primary importance, particularly when
holdings in the large farm sector are redistributed to individual smallholders.

These three goals are not mutually exclusive, and more often than not, land reforms
involve multiple goals. Although, concurrent improvements in social equity and economic
efficiency are often interpreted as conflicting goals because it is often assumed that
redistributing assets to smallholders will reduce aggregate agricultural production, evidence

to counter this proposition is presented in section 2.1.23.1.

2.1.1.4 Land Redistribution and Resettlement Policies

Land redistribution and land resettlement are similar because both result in a change
in land use rights, but neither requires a change in exchange rights to land. The difference
lies in the fact that resettlement does not necessarily require land redistribution to take
place. For example, settlers may be moved onto new, unoccupied lands, if these are
available for settlement. Redistribution, on the other hand, implies that after existing
landholdings are dissolved, new farmers will be resettled onto the available land (except in
the case of tenancy reform). For example, Zimbabwe’s Rural Resettlement Policy provides
for both the subdivision and the redistribution of large commercial farms to individual
smallholders or cooperatives, and the resettlement of Communal Area farmers onto newly
developed resettlement schemes.

Land redistribution programs involve the acquisition of land, with or without
compensation to the owner, and its division among smallholders, the landless or other

targeted populations. Often land redistribution occurs within the context of revolution

 

 

 

i650

the '
not
but
the

[rag

l4
(Bolivia, for example), or through the confiscation of foreign-owned property (King, 1977).

The goal of a redistn'bution policy is to improve smaller farmers’ access to production
resources and change the distribution of income among members of a population.

Jacoby (1971: 170-71) argues that "any purposeful redistribution of land must entail
the redistribution of wealth, income“, status, capacity for saving and political influence which
not only will provide incentives for increased agicultural production and labor productivity
but will strengthen the socio-economic position of the peasant population.” He also stresses
the importance of accompanying institutional changes since "experience has proved the often
tragic consequences of redistribution programmes that are confined to the mere distribution
of public domain land and which completely ignore the vital problems of agicultural credit,
infrastructures, soil improvement and access to markets” (1971: 170).

Land settlement (or resettlement)3 and colonization programs involve the
relocation of selected farmers onto new, unsettled areas or less densely populated, already
settled areas, as in Indonesia’s transmigration program. Where governments are hesitant
to make the political changes necessary for land redistribution, land colonization programs
are often carried out instead (Oberai, 1988). Colonization is designed to reduce population
pressure and resource depletion in one area and generate a more equitable distribution of
land and labor for participating colonists (Jacoby, 1971). When farmers must relocate in
order to benefit from a land redistribution or land consolidation program, governments
often establish centrally-planned settlement prog'ams with the aim of providing the farmer

with new economic and political opportunities (J acoby, 1971). Oberai (1988) stresses,

 

3 Jacoby (1971) suggests that the term settlement could apply strictly to the
sedentarization of nomadic groups, and resettlement to the movement of sedentary people
from one area to another. In this paper, however, settlement and resettlement are used
synonomously.

ho

’J

15

however, that "land settlement should not be considered either as an end in itself or as a
convenient means of redistributing the population" since a settlement policy which supports
neither the long-term development of the area nor is responsive to scheme-level problems
as they arise, may provide settlers with little ability to sustain themselves independently of

continued government intervention.

2.1.1.5 Ancillary Land Policies

Other land policies, often carried out as substitutes for land reform, influence land
use rights, but have no direct redistributive effects. Land tax reform and consolidation
policies improve the incentive structure for ag'icultural producers through changes in use
rights to land, but do not change exchange rights.

Land tax reform is often implemented as a lower cost alternative to a land
redistribution program. The purpose of a land tax reform is to: 1) increase landholding
costs to larger landowners, forcing them to reduce the size of their estates and/ or increase
the intensity of their land use; and 2) raise revenue for the government. Governments enact
land tax reform to induce agricultural productivity and indirectly redistribute wealth and
income. In many countries, however, landlords hold sufficient political power to block or
evade tax reforms. Furthermore, King (1977) argues that in countries where they have been
enacted, government revenues have only increased by an additional 1- 10%.

Land consolidation refers to the reorganization of farms with widely dispersed
individual plots, in order to improve productivity and production efficiency. Fragnentation
is eliminated either by bringing the disaggregated plots together through land exchange or
reducing the total number of farms by combining them. Although fragnentation is often

the result of a rational farnning system (to reduce crop risk or avoid adverse variations in

16

soil types), critics argue that it limits the potential for mechanization and increases the time
allocated to farming activities. Once implemented, consolidation may be difficult to enforce
due to traditional methods of intergenerational land allocation (i.e., inheritance by family
members) or farmers’ unwillingness to change their existing farming system. This often

results in parcels becoming fragnented again, regardless of legislation (King, 1977).

2.1.2 Land Reform and Institutional Change

It was stated above that land reform involves changing the institutional arrangements
governing rights to land and other productive resources, which ultimately determine income
distribution among individuals. It is necessary to analyze these institutional arrangements
and the interdependencies they generate in order to assess the impact of a change in these

rights on the agarian structure and on desired political, social and econonnic goals. The
following sections examine factors that influence changes in land distribution and tenure in

an agrarian system, and the potential impacts of these changes on agricultural production,

income distn'bution and employment generation.

2.1.2.2 The Impetus for Land Reform
Ruttan and Hayami assert that economic forces often drive land reform (1990). For

examPle, the induced innovation model of agricultural development hypothesizes that
"Changes in market prices and technological opportunities introduce disequilibrium in
emEisting institutional arrangements by creating profitable opportunities for institutional
iIlllovaltions" (Ruttan and Hayami, 1990: 106). Forces such as population pressures or new
techn ology would cause the value of land to increase relative to labor or other inputs, and

induce the need for land reform (Stevens and Jabara, 1988: 269)-

 

17
Yet, this model has limited applicability for the study of land reforms since there are

examples of LDC tenure systems in nonmarket economies where the institutional rules only
provide for use rights. For example under communal tenure or tenant-farming
arrangements, individuals have no exchange rights to land and so there are no market
economy prices for their land. This model assumes, however, that markets exist through
which land can be bought and sold.
Cohen argues that land markets do exist in communal tenure systems, even though
they are not based on economic exchanges. While land is not directly salable, ”it can be
passed to others through a variety of actions, often with a profit” (1980: 360). Nonmarket
transfers may increase security of ”ownership” for communal land farmers (i.e., land rights
transfers through gifts, loans or inheritance), but do not necessarily lead to the development
of private property rights through the emergence of a land market. Therefore, in communal
tenure systems, economic forces may not induce changes in land tenure.
De Janvry concludes that if an agarian system comprises "a system of social
relations and a system of land tenure” (1984: 264), then it is influenced more by rules of the
state than by rules of the market. Ghose explains that "...because land relations tend to be
linked to political power relations, the process of agrarian reform is inherently a political
Process“ (1983: 6). Therefore, land reform constitutes a political instrument for
transforming an agrarian system when the current system’s production and consumption

relationships cannot be successfully altered by alternative policy strategies (Ghose, 1983).

2-1.2.3 Impact of Land Reform on Agriculture: The Economics of Land Reform
After years of experimenting with different economic growth strategies for

developing countries, development economists have observed that the returns from growth

18
did not always “trickle down" to lower income groups (Eicher and Staatz, 1984: 13-14).

Lower income groups often remain the most disadvantaged because they lack adequate
agricultural production and employment opportunities, even to achieve subsistence living
standards. Rural incomes are often skewed because an inequitable landholding system
restricts access to ag'icultural resources for a large percentage of the population. Thus,
assuming that land is the primary scarce resource ”and hence the main source of rural
inequality and power,” land reform can improve the distribution of resources, as well as
stimulate economic growth (Lipton, 1974: 271).

The following sections analyze the implications of land redistribution for agricultural
production levels, irncome distribution and employment generation, in countries where
economic development is constrained by inequitable landholding systems.

Reform programs that redistribute land to smallholders assume that: 1) small
farmers use land more efficiently than larger farmers; 2) post-reform smallholders are able
to contribute to aggregate marketed output; 3) smallholder ag'iculture creates more
employment opportunities than large-scale agriculture; and 4) the reallocation of land

rmurces to smallholders improves income distribution.

In most developing countries capital is a scarce resource, while labor is more

abundant. This is apparent in high underemployment and unemployment rates, particularly
in rural areas (Dovring, 1974). However, large-scale agriculture tends to employ more
capital-intensive methods of production to take advantage of economies of scale inherent
to mechanization and to reduce hired labor costs. Consequently, when land is redistributed

fi‘om large landowners to smallholders, the labor-intensive methods practiced on small farms

19
would absorb more labor resources.4 Meyer (1989) argues that in addition, land

redistribution to smallholders reduces the demand for scarce capital resources formerly used
in the more capital-intensive processes found on large farms. Further, small farm yields
would increase and these farmers would earn higher incomes.

Although most analysts agree that redistributing land to individual smallholders
would also redistribute wealth, there is less agreement regarding the potential impact of land
redistribution on agricultural production and employment generation. The following three

sections review the empirical evidence on this subject in greater detail.

2.1.23.1 Impact on Agricultural Production

In making the case for land reform in Kenya, Hunt (1984) summarizes the criticisms
against the subdivision of large farms into smallholdings. First, since large farms use labor
and machinery more efficiently, they can capture economies of scale in production,
marketing and in the dissemination of information. Second, large farms are more likely to
adopt technical innovations because they can both obtain credit more readily than smaller
fararers and are able to bear more risk. Third, large-scale agriculture makes a greater
contribution to overall economic growth because larger farmers have both a greater
Propensity to save and are more able to supply surplus agricultural goods for domestic use

and for export.
Hunt stresses, however, that a critical examination of these arguments reveals that
land redistribution can occur without decreasing aggregate agricultural production. First,

VVith reSpect to input efficiency, small farms are often able to vary the proportions of inputs

‘

large and small farms are relative terms whose definitions depend on the quantity
and qufllity of available land. The size of farms discussed in each case study will be defined
according to their respective context.

20

they use (i.e., labor for capital) such that the more abundant resources are fully employed;
and resources that cannot be fully utilized may be shared or hired out to others (see also
Binswanger and Elgin, 1990). Further, smallholders can realize economies of scale if they
act collectively. For example, farmers can purchase inputs, market their output and
participate in extension activities as a group (Hunt, 1984).

In assessing the relative efficiency of small and large farms, different performance
criteria may be used, depending on the objective to be maximized. For example,
considerable evidence supports the thesis that yield is inversely related to farm size and that
monetary expenditures per land unit are positively related to farm size, implying that small

farms produce greater output at lower cash costs (Berry and Cline, 1979; Hunt, 1984;
Binswanger and Elgin, 1990). Binswanger and Elgin (1990) attribute the higher cash costs
per unit of output on large farms to the fact that they employ more hired labor than smaller

farms. Assuming that the opportunity cost of family labor is less than the wage rate, family
labor costs less than hired labor. In addition, family workers have more incentive than hired
labor to be productive because they share the risk-bearing, as well as the farm output.
However, small farmers’ nonmonetary costs (e.g., family labor) may indicate that
total labor use is greater on small farms than on larger ones. Therefore, if the objective
b6ing maximized is to increase labor absorption in agriculture, a more relevant criterion
would consider person/days of labor employed per hectare on small versus large farms. In
the case where foreign exchange constraints exist and the objective is to minimize the use
of inputs with substantial foreign exchange components, then smallholder production may

be considered more efficient.

21

Second, if an institutional capacity already exists, credit can often be provided to
small farmers at a relatively low cost. With improved access to credit, smaller farmers can
adapt technical innovations and bear risk in proportion to their capital (Hunt, 1984).

Third, the potential contribution of large-scale agicultural production to national
economic growth depends on the type of crops gown and whether their production
generates foreign exchange or marketable surplus for domestic consumption. If small farms
produce higher yields than larger farms, they may actually transfer a geater amount of
aggregate production to other sectors of the economy than larger farms (Hunt, 1984), if

larger farms produce a geater proportion of crops for export. The relative contribution of
small farms is contingent on available family labor and the total area cultivated in food
versus cash crop production.

Lastly, small farm households have shown a high propensity to save where there are
viable investment alternatives (Hunt, 1984). Furthermore, even if larger farmers have a
higher: propensity to save than smaller farmers, these savings may not contribute to
economic gowth because large landowners have a higher propensity to purchase luxury
items and to use more foreign exchange per unit of output than smallholders (Hunt, 1984).

According to Berry and Cline (1979), the relationship between farm size and land
PIOdinity changes as significant off-farm employment opportunities for small farmers
emerge. As the small farmer’s opportunity cost of labor increases, the inverse relationship
between farm size and land productivity disappears because there is less incentive for small
fame“ to use labor intensively on the small farm. Clearly, however, in the develOping

9‘38“ °f§n economy, available evidence suggests that the redistribution of large-scale farms

into smelt farms stimulates aggegate agricultural production and economic gowth.

2.12.3.2 Impact on Income Distribution
Generally, the redistribution of landholdings to smallholders increases their access
to income-generating opportunities through employment creation in agriculture. Yet, the
impact of land reform on income redistribution in rural areas depends on the income-
generating potential (a function of yield) of the newly created small farms, which is highly
negatively correlated with the size of the holding distributed to the beneficiaries.5
Available evidence indicates that land redistribution to smallholders increases employment
per unit of land area (Sternberg, 1971). Furthermore, as argued above, smallholders achieve
higher yields than larger farmers. Depending on the level of inputs required for smallholder
production, higher yields may result in higher incomes for labor absorbed into the small
farm sector. In addition, in the presence of strong consumption linkages, increased farm
incomes may lead to increased consumer demand for agicultural goods and services, and
for goods and services produced in other sectors of the economy, depending on the
wmpOsition of the goods and services purchased by smallholders (PinstrupoAndersen and
Hazel], 1987). Therefore, the extent to which land reform redistributes incomes to rural
smallholders and the landless will depend on the total number of small farms to be
distributed, the size of holding created and on smallholders’ capacity to generate income and

Employment opportunities for both family and hired labor.

 

 

5 77" size of holding distributed determines the number of people who will benefit
from d‘ reform. The smaller the post-reform holding, the geater the number of
benefifim'ec. Yet, if the holdings are too small to meet household food needs, the condition
of many farmers could worsen (Lipton, 1974). The holding size must, therefore, take into
account the typical farm family’s management abilities, available technology, agoecological
conditions and the potential of the holding to produce an acceptable income level.

2.1.233 Impact on Employment

Land reform has been used in primarily agarian countries to combat high
unemployment and underemployment rates which are exacerbated by high rates of
population gowth and landholding systems that constrain the expansion of rural
employment possibilities. As a rural development strategy, employment creation through
land reform contributes to economic gowth and development, helps to redistribute incomes
(Sternberg, 1971) by increasing the amount of family and hired labor applied per land unit
(Lipton, 1974) and creates employment opportunities in smallholder agiculture.

Sternberg (1971) contends that although creating jobs is essential, it is equally
important to create W jobs, especially given that per capita incomes are generally
lower in agiculture than in other sectors of developing economies. He advocates not only
changing the existing allocation of resources through land reform, but also expanding
cultivated land area where possible and improving rural infrastructure and services. He
concludes that "the possibility of creating remunerative employment in agiculture is thus
related to the over-all availability of non-human resources in agiculture, the level of
technology and investment sources and uses, as well as to their redistribution and

redeployment in a more rational manner“ (1971: 4).

2.13 A Conceptual Framework for the Analysis of Land Redistribution Policies
It is difficult to assess the impact of land redistribution policies in developing
countries because there are generally insufficient pre- and post-reform data to rigorously
W changes in the agicultural sector. The lack of data is particularly critical for this

study which is based entirely on secondary data.

24

Therefore, due to insufficient time series data for each land redistribution case
examined, this study focuses on a descriptive analysis of the impacts of redistribution on
selected indicators of performance, based on case studies and other available information.
Each land redistribution case will be assessed by:

1) describing the political, economic and social context of the reform, including the
predominant pre-reform tenure systems, the distribution of land and incomes, and the
motivation for reform;

2) outlining each reform policy in terms of its stated goals, implementation and

development of supplementary policies;

3) analyzing the impact of redistribution on: a) smallholder agicultural production;
b) income redistribution, based on the number of beneficiaries and redistribution’s effect
on productivity; c) employment generation, for both family and hired labor, and the type of
employment created; and d) other case-specific indicators;

4) assessing the effects of redistribution on gowth in the agicultural sector, where
causality can be established; and

5) examining the effectiveness of monitoring and evaluation in improving the

Pedomance of the redistribution progam, where relevant.
Relevant lessons from these case studies and empirical data from Zimbabwe
concerning the desigt, implementation and evaluation of redistribution progams are

synthCSich in Chapter Three.

2-2 Lent Redistribution Policies in Latin America, Asia and Africa
Sweeted land redistribution progams in Latin America, Asia and Sub-saharan Africa

were chosen to illustrate the wide variety of conditions under which distributivist land

25

reforms occur (Table 2.1). Although each reform was implemented in response to different
political, social and economic pressures in order to alter the institutional structure governing
rights to land, each offers important insights for the analysis of land resettlement policy in
Zimbabwe.

First, the studies of land reform in Peru and Bolivia illustrate policies developed to
change tenancy and sharecropping systems to cooperative and individualized tenure for
smallholders. In each case, although the government emphasized cooperative development,
smallholder agiculture predominated. Furthermore, as policy makers had limited
experience with communal landholding systems (except in areas with Indian farming
communities), they assumed that freehold tenure provided the geatest incentive and
security to the individual landholder. The primary goal of these Latin American reforms
was to immediately abolish the latifundia (large landholding system). Hence, given the
short-term focus of these reforms, they failed to develop the complementary infrastructure
and services essential for a reform to have positive, long-term impacts on agricultural
production and income redistribution.

Second, the Indonesian transmigation progam illustrates a large-scale resettlement
progam designed to provide farmers with increased economic opportunities. This case is
especially useful in highlighting scheme-level planning, regional development and progam
evaluation issues.

Third, Kenya’s resettlement progam is most comparable to that of Zimbabwe
because it occurred in a similar political context and had similar policy objectives. For these
reasons, particular attention is given to analyzing the implementation of Kenya’s progam
and its impacts on smallholder agicultural production, income distribution (hence the

opportunity to test the gowth with equity hypothesis), and employment generation.

26

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—
Table 2.1: Primary Characteristics of Case Studies; Peru, Bolivia, Indonesia and Kenya.
Country: Perua Boliviaa Indonesiab Kenyac
Type of reform redistn. to redistn. to coops. smallholder redistn. to I

cooperatives and smallholders colonization smallholders

Period of 1964-1975 1953-1975 1907- 1962-1971
reform
Amt. of land 8.6 18.0 6.9 1.7
transferred
(millions of ha)
No. of families 370,000 477,000 939, 71,000
benefitted (% of (6.8) (30.0) (3.8) (4.0)“
rural pop.)
Average holding 23 30 1-2 10-13

M l _

a Data as of 1975.

b Data as of 1987.

c Data as of 1978.

d Transmigrants as a percentage of Java’ s total population.
c Number of beneficiaries as a percentage of the total number of smallholders, 1978.

 

Sources: Peru: Tom Alberts, . .: .:_ ’ ‘ ' ' . i ' I (1981);
Russell King, WM (1977). Bolivia: Schlomo Eckstein et al., mm
mm (1978); Thomas F Carroll. WW (1961)
Indonesia: World Bank, WW (1986): JM HardionO.
WM (19”); A S Oberai. When
Was (1988) Kenya “Him 1 House and Tony Killick. “Social
Justice and Development Policy m Kenya’ s Rural Economy. in WWW

Akin (1983)

 

27
2.2.1 Land Tenure and Agrarian Structure in latin America

The basic structural imbalances that characterize Latin American agiculture today
have their roots in the landholding systems established during the Spanish and Portuguese
conquests in the sixteenth century. The resulting political, economic and social inequalities
have led many countries to implement agarian reforms. Carrie Meyer (1989) points to the
fact that of 20 Latin American countries, nine have implemented significant reforms
aflecting over 20% of their rural populations. Several - like those in Chile and Guatemala -
- were short-term, but others - such as the revolutionary reforms carried out in Cuba, Haiti,
Bolivia and Mexico - have had a profound impact on the agarian structure of each country.
This section first describes the predominant landholding systems in Latin America today;
and then analyzes the reform progams carried out in Peru and Bolivia, which are indicative
of the redistribution and resettlement policies undertaken to alter this structure that neither
"corresponds to the aspirations of the rural population nor to the requirements of
technological progess" (Carroll, 1961: 162).

Four important landholding systems form the basis of Latin American agarian
structure; the latifundio, the minifundio, the communidad and the colono systems. The
latifundio system, which accounts for most of the land, is made up of large individually-
owned estates (either extensively cultivated haciendas or intensively worked farms) that are
cultivated by many tenant farmers. For example, King (1977) found that for the continent
as a whole, 90% of the agicultural land is in the hands of 10% of the landowners. As
Alexander points out, ”the traditional large landholding system has seriously hampered the
modernization of Latin America. It has hindered economic progess, made for widespread
illiteracy and ignorance, and prevented a large proportion of the inhabitants from

participating actively in the economic and political life of their countries“ (1974: 17).

28

The minifundia make up the majority of farm units. These small, individually-owned
family farms range from 2 to 20 hectares. While the latifundia are located on the best farm
land, the minifundia are generally found on the poorest soils. Population gowth,
compounded by in-migation, has resulted in severe fragnentation of plots and deterioration

of their productive capacity.

The communidad is a collective farming system operated by several extended
families, whose origins lie in early Indian agriculture. Although cooperative in nature, it is
considered restrictive because it does not provide incentives for individuals to invest human
or capital resources in the production process (Carroll, 1961). The ejido is a modern
version of the comunidad which resulted from the Mexican land reform, but is much more
integrated into the social and economic fabric of the country (Carroll, 1961).

The colono system serves as an agicultural labor reserve for the large estate sector.
Since most farmers do not own land, large landowners typically compensate them for their
labor with a cash payment, or more frequently permit them to temporarily cultivate a parcel
of land. To earn this compensation, the colono (tenant farmer) must work a specified
number of days, as well as fulfill other obligations to the estate. The colono system often
exists in conjunction with cash-rent tenancy or sharecropping.

These four systems share some of the same characteristics. All are generally
inefficient, and concentrate resources into the hands of a very few. Among the latifundios,
the hacienda system is the least efficient because landowners maintain a sigtificant amount
of idle land, resulting in low output per worker and per land unit. Although plantations
produce geater output per land unit, they still perpetuate a skewed distribution of
resources. The minifundios suffer from the increasing depletion of an already small

resource base, with no available incentives or opportunity to invest in better production

29

methods. The minifundios, similar to the colonos, lack resources, investment opportunities
and institutional support such as extension, credit, medical and school facilities.
The following two case studies examine land redistribution programs implemented

to counter the extreme concentration of land and incomes in Peru and Bolivia.

2.2.1.1 Land Redistribution in Peru

Peru’s land reform progam, while nonrevolutionary, was designed to abolish laws
favoring tenant-landlord relationships and redress the skewness in rural incomes.
Recognizing that the skewed rural landholding pattern was responsible for Peru’s linnited
agicultural gowth in the 19503 and early 1960s, the government proposed to redistribute
8.6 million hectares of agicultural land to approxirnnately 370,000 smallholder families (6.8%
of the rural population). The land reform policy unfolded over portions of two political
regimes; from 196468 and from 1969-75.

Initially, the government’s primary stated objectives were to improve the country’s
social structure and productive capacity, and later, to improve income distribution among
the population (Alberts, 1981). The reform was implemented by expropriating all but the
most efficient large farm operations and turning the land over to cooperatives and individual
smallholders (King, 1977; Kay, 1983). As compensation to the former property owners, the
government provided long-term bonds and cash. Major obstacles limiting the progam’s
success included the government’s ability to accommodate the numerous squatters who
applied for land titles, and the lengthy expropriation procedures which limited the total
amount of land transferred (Alberts, 1981).

By redistributing land to cooperatives and pockets of peasant insurgency, the reform

had no commitment to support smallholder agriculture and create the agarian structure

30

desired by the government. The primary beneficiaries were permanent workers on sugar,
cotton or rice plantations who received higher wages, improved working conditions and
membership in the newly formed cooperatives (Alberts, 1981). Although the number of
smallholdings increased, their average size decreased. Thus, the number of nninifundistas
actually increased as a result of land redistribution. Furthermore, Alberts (1981) found that
income distribution did not improve among Peru’s rural population and, in fact, worsened
among the lowest income goup.

Many smallholders received no benefits from the reform because they cultivated
remote parcels that did not command government attention, even though the government
had stated that it would also improve service provision to marginal rural people. In
retrospect, Kay (1983: 232-33) pointed out that, “within the context of a market economic
system and given the characteristics of the pre-existent agarian system, it is likely that a
redistributive type of reform which assigned land to peasant farmers and peasant
communities would have been more successful [than the government’s focus on the
cooperative sectors] ." It also seems likely that the reform increased income concentration
in the higher income groups (Kay, 1983). For example, census data show that during the
reform period, the percentage of the population working in agiculture declined by
approximately 8%, the number of self-employed agricultural workers increased by 25% and
the number of wage laborers fell by about 10% (due to the subdivision of the large farms
that had previously employed large numbers of wage laborers) (Alberts, 1981).

Although, official data do not show a decline in overall agicultural production
associated with the reform during these years, production stagnated and eventually
decreased per mm from the 1960s to the middle of the 1970’s. Alberts attributes this

primarily to a declining resource base, the cultivation of increasingly less productive land,

31

low private investment in agiculture and price controls and import subsidies that
discriminated against agiculture (1981).

Without question, the government succeeded in meeting its less publicized objectives
of increasing the food supply to urban areas by supporting large-scale cooperative
production levels. Although the government placed subsidies primarily on imported foods
to prevent urban food prices from rising (which amounted to 87% from 1973 to 1976), this
policy resulted in reduced rural incomes. In addition, Kay (1983: 232) observed that “the
government’s cheap food policy not only negatively affected rural incomes but also by
reducing agiculture’s profitability, indirectly affected investment, production and
employment in this sector.” The government, however, failed to mobilize rural political
support because most smallholders opposed the government’s large-scale and cooperative
enterprise models for the reform sector. Thus, ”in the absence of a genuine mass
participation, the collective cooperative model had to be imposed from above and was

perceived as a constraint by the very people it was supposed to benefit” (Kay, 1983: 235).

2.2.1.2 Land Redistribution in Bolivia

The Bolivian land reform, although the product of a revolution, introduced individual
smallholder agiculture into an agarian system historically based on sharecropping and
indigenous farming collectives. The government’s primary refornn goals were to destroy the
latifundia and decrease land concentration among large landowners. Specific objectives of
the 1953 land reform decree were to improve agricultural development, conserve natural
resources and stimulate internal migation (Carroll, 1961). As in the case of Peru, all large

landholdings except highly capital-intensive operations were expropriated. The government

32

provided 25-year bonds as compensation to former landowners and reform beneficiaries
were expected to repay the government for land received (Carroll, 1961).

Due to the violent nature of the peasant revolution and the political instability that
followed, the government faced difficulties in establishing titles for new beneficiaries and
thus farmers had insecure rights to land. 25 % of the land was initially distributed under
collective title, but by the 1970s, most of this land had been redistributed to smallholders
(Eckstein et al, 1978). Administrative, technical and financial resource constraints limited
the government’s ability to establish a technical basis for reform planning, and effectively
carry out the land redistribution progam, including the provision of complementary services
to new landowners (Carroll, 1961). As of 1960, few landlords had received their
compensation and the government had not required beneficiaries to pay for their holdings
(Carroll, 1961).

The reform had a positive impact in areas suffering from serious social conflict. In
those areas, the standard of living improved for farmers who received land and individual
title. On a limited regional basis, the reform increased peasant participation in the market
economy and expanded farmers’ access to metal plows, chemical fertilizers and hired tractor
services (Eckstein et al, 1978). While agricultural production decreased following the initial
land redistribution, subsequently, staple food production increased significantly as a result
of the reform - primarily due to an expansion in area cultivated and increased access to
inputs (Eckstein et al, 1978). Although there are no data on the impact of the reform on
income distribution, if marketed production increased and consumer spendings tripled at
one time (King, 1977), then it must have increased incomes of the land reform beneficiaries.

Since the reform was implemented as a reaction to a peasant revolution, the

government made no effort to systematically restructure Bolivia’s agarian structure. It is

33

clear that more comprehensive legislation and a geater commitment of resources would
have ensured a broader impact and a geater distribution of benefits. The reform affected
30% of the rural population, achieving some of the government’s objectives. For example,
a significant amount of land was distributed, agicultural labor relations were reformed, and
internal migation was stimulated through colonization progams. However, the Indian
communities experienced little or no development, simply land restitution (Garcia, 1970).
In addition, the reform had little impact on agicultural development outside of the reform
sector and it probably did not strengthen resource conservation, especially since the
conditions of the minifundista were not improved. Finally, the reform was not totally
successful is calming rural conflicts and some smallholders lost their new titles to large

landowners who wanted to regain their properties (King, 1977).

2.2.2 Land Tenure and Agrarian Structure in Asia

Agarian structure in Asian countries differs from that of Latin American countries
because farms sizes are smaller and the landholding pattern is less skewed. A large
proportion of land is individually owned (for example, 92% in India, 75% in Indonesia, 42%
in Pakistan, 58% in the Philippines).6 However, current landholding patterns result from
land reforms carried out in these countries. Prior to land reform, high rates of tenancy
existed in many countries (for example 50% in the Philippines (Ahmed, 1975), where
tenants paid a large proportion of their output to landlords and lacked security of tenure
(King. 1977). Even today, tenant farming exists, but to a much lesser extent. For example,
in India tenants farm only 4% of the land, and similarly 3.2% in Indonesia, while in the
Philippines they farm 29%, and 35% in Pakistan (Bell, 1990). Furthermore, tenants farm

¥

5 FAQ (1981), based on 1970 data.

34

much smaller plots than owner-operators. For example, in India, tenants cultivate about 1.5
hectares while owner-operators cultivate 3 hectares. Individual smallholders farm small plots
ranging from 0.9 hectares in the Republic of Korea, to 3.6 hectares in the Philippines and
53 hectares in Pakistan (Bell, 1990).

While there is still some inequality in resource distribution, rapid population gowth
constitutes the most severe problem facing most Asian countries today. The population in
many countries increases by at least 2% annually, while the labor force increases by 2-3%
(World Resources Institute, 1990). Population densities in rural areas average from 500
persons per square kilometer, to 1,500 per square kilometer on Java in Indonesia (King,
1977). Increasing population gowth results in small, fragnented plots for farmers,
overcrowding in urban areas, and increases in the labor force that often surpass gowth in
national food production. Much of the Asian labor force is employed in agiculture (for
example, 55% in Pakistan; 57% in Indonesia; 52% in the Philippines; and 70% in India).7
Thus, governments are faced with accelerating population pressures on a limited resource
base. In the past, many countries responded to these pressures by expropriating land from
large landowners for subdivision among smallholders, tenants and the landless (King, 1977).
Indonesia’s strategy, the transmigation progam, represents one such effort to alleviate
population pressure on Java, Madura, Bali and Lombok, create employment and reduce

poverty through land resettlement.

2.2.2.1 Land Redistribution in Indonesia: Transmigration Policy
Indonesia’s transnnigation policy evolved from early Dutch attempts to alleviate

population pressure on Java by moving voluntary nnigants to the outer islands (Oberai,

 

7 From the World Development Report, 1988.

35
1988). In the late 1960s, when average population densities had reached 600 persons per

square kilometer, the Indonesian government began to formulate a large-scale
transmigation progam under the first Five-Year plan (Repelita I). These plans
emphasized alleviating population pressure from Java, as well as increasing social welfare
and promoting regional development on the outer islands. Since revenues from oil exports
financed much of the progam, the transmigation budget varied with oil prices, fluctuating
over time from 0.7 to 6.1% of the total development budget.

The Indonesian government (GOI) selected transmigants according to specific
criteria (i.e., the head of household must be between 20-40 years old, married with a small
family and have some agicultural experience)8 and provided them with transportation to
the site, a house, land and a subsistence input package for one year. However, upon
arriving at the sites, settlers often encountered poorly defined land boundaries and an
absence of legal records, leading to conflicts between local residents and transmigants over
land titles. In addition, the indigenous people of the outer islands tended to resent the
migrants who have better access to facilities and services (Oberai, 1988). For example, a
1981 UNDP/ OPE Management and Monitoring survey found a higlner ratio of agicultural
extension workers to settlers in transnnigation areas than in the local outer island farming
communities (Babcock, 1986). Transrrnigration authorities have attempted to mitigate this
problem by settling some local people onto transmigation sites, improving local land
acquisition methods, and encouraging indigenous entrepreneurs to establish commercial

relations with the new settlers (Oberai, 1988).

 

8 The government also resettles flood victims, the landless, farmers whose land is
needed for development purposes, and the urban poor (Suratman and Guiness, 1977).

36
The government finances primarily rainfed and irrigated upland sites, tidal

cultivation sites and tree crop plantations. Critics argue that in many resettlement areas,
site planning and selection has resulted in crop failure, declining soil fertility and land
disputes with local farmers that disrupted farming activities (Suratman and Guiness, 1977).
Not all site-specific problems can be attributed to oversights in planning since part of the
site selection criteria includes national security considerations and the colonization of less
populated areas (Suratman and Guiness, 1977).

The GOI provides no compensation to outer island residents for land ceded to the
transnnigation progam since it feels that they should relinquish land as a contribution to
national development (World Bank, 1988). Although the government required earlier
transmigants to pay for the initial services they received, the government discontinued this
policy to allow transmigants to achieve an adequate standard of living more rapidly
(Hardjono, 1977). The 601 intended to transfer scheme-level management responsibility
to the provincial (district) level after five years of transmigation authority supervision. The
districts, however, often encounter difficulties in financing infrastructure maintenance and
scheme-level personnel after the 601 withdraws its support.

According to World Bank studies, transnnigants produce approximately 2 million
tons of unmilled rice, which is equivalent to 17% of total outer island production and 5%
of Indonesia’s total rice production (World Bank, 1988). On a regional level, the World
Bank (1988) estimates that since 1950, transmigants have contributed from 45-96% of
incremental rice production, depending on the area. However, due to decreasing soil
fertility (particularly on upland, rainfed sites), declining rice yields and thereafter, declining

agricultural incomes, have forced farmers to turn increasingly to non-farm sources of

employment.

37

Transnnigrant incomes vary by both farm model and length of settlement. More
recent transmigants generally earn lower incomes than farmers in the rural sending and
receiving areas, although more established farmers earn slightly higher monthly incomes
than rural Javanese (World Bank, 1988). Data show that farmers on older upland sites earn
low agicultural incomes, attributable to declining soil fertility, which they supplement with
off-farm income. Recent settlers in swamp reclamation areas earn the lowest total incomes
(farm and non-farm), while farmers with tree crops and those on older tidal sites earn have
the highest agicultural incomes. The proportion of total household income derived from
off-farm employment ranges from 40-80% (again depending on farm model and period of
settlement), indicating that some transmigants find off-farm employment more
remunerative than work in agiculture (World Bank, 1988).

While some new transmigants earn lower incomes, most of those surveyed believed
that their living conditions have improved after transmigration (World Bank, 1988). In
contrast, they reported that transportation services were worse in the transnnigation areas
than in Java (World Bank, 1988).

During Repelita III (1979-84), the transmigation progam created between 500,000-
600,000 jobs on both rainfed food crop schemes and on tree crop settlements (World Bank,
1988). This translates to 1.3-1.6 jobs created per transmigrant household, or a total of 12-
15% of the incremental labor force for Java and Bali (World Bank, 1988). Critics argue that
transmigation is very expensive, yet it costs less to create employment for sponsored
nnigants than in the industrial sector (810,000-20,000 per job), although more than in service
industries in Indonesia (World Bank, 1988). For example, on upland schemes, the
government pays approximately $3,300-4,100 to create one permanent job. On tidal

schemes, the cost increases to $4,500-5,500. On more remote sites, the cost per job

38
increases by approximately 25-50%, depending upon the farm model (World Bank, 1988).

Althougln associated rural works progams create some temporary employment, and
transmigation projects offer limited opportunities for low-level professionals, they generate
no important secondary industry (Babcock, 1986).

Critics also argue that transmigation has led to increased environmental
degadation. For example, the World Bank found that during Repelita 111, between 30-50%
of the total land cleared for sponsored rrnigants was previously forested. The relative
proportion of forested land cleared for each transmigation site differs depending upon the
province, and ranges from 0.2% in Sulawesi to 4.0% in Sumatra (World Bank, 1988). At
the regional level, transnnigation contributed to infrastructure development, but Oberai
(1988) observed a marginal impact on regional integration in terms of developing industry
and trade, exploiting and processing natural resources and improving transport and
communications.

To the degee that the progam has absorbed a small amount of Java’s gowing
population, transmigation has been successful. Between the 19408 and 1987, the
government settled almost 950,000 families on the outer islands, in addition to an unknown
number of spontaneous nnigants who received no government assistance. Even with this
tremendous effort, Oberai (1988) projects that the transmigation progam can only hope
to remove 20% of the annual increase in Java’s population gowth.

The program’s impacts on regional development are more difficult to discern.
Although many families achieved higher standards of living after the first several years of
acclimation, environmental degadation, a lack of human resources to administer the
transmigration sites, and gowing land scarcity are becoming increasingly important

problems. In addition, the government has made a very limited attempt to monitor the

39

progess of transmigration in any systematic fashion, which inhibits its ability to respond to
local-level problems as they arise. The economic rates of return to transmigation are very
low; from negative values to approximately 4% for the most common settlement scheme
(the upland model), to 11-13% for coconut, oil palm and rubber tree crop schemes. Critics
charge that the progam spends large sums of national funds on a small percentage of the
population, with excessively low rates of return (Babcock, 1986). However, these same

critics have yet to define any alternative policies to meet transnnigation progam objectives.

2.23 Land Tenure and Agarian Structure in Africa

Tenure systems in Africa differ significantly from those in Latin America and Asia.
Communal tenure initially formed the basis of African landholding systems, although
colonial governments altered traditional tenure systems by introducing individualized
ownership, which subsequently changed the agarian structure. A study by Clive Bell (1990)
examined average holding sizes and the distribution of landholdings by tenure system for

18 countries.9

His study, unfortunately, provides sparse data on the distribution of
landholdings in Africa by type of tenure. The only complete data set is for Cameroon,
which indicates that approximately 59% of the land is held under communal tenure, 25%
under mixed tenure and 2% privately owned. This is in contrast to some Asian countries
where 66% of the holdings are individually owned, and Latin America where 62% are
individually owned. Communal tenure is a much less important form of landholding in Asia
(less than 0.1%), as well as in Latin America (6%). Average landholding sizes in Africa are

small - 2.5 hectares - compared to 3.4 in Asia and 47.1 in Latin America.

 

9 FAO (1981), based on 1970 data.

40

Since landholding systems determine incentives to farmers and income distribution,
the choice of tenure system is an important policy issue, particularly in Africa. Neoclassical
economists argue that individualized tenure systems: 1) increase tenure security and reduce
the transactions costs involved in land disputes; 2) lead to geater investment and increased
demand for credit; and 3) create a land market whereby land is transferred to those who
are able to extract the highest value from it (Barrows and Roth, 1989: 6). The neoclassical
model thus concludes that traditional tenure systems are inefficient because “property rights
are not clearly defined, costs and rewards are not internalized and contracts are not legal
or enforceable" (Barrows and Roth, 1989: 2). Economic efficiency, it is thought, will guide
the evolution of land tenure systems.

As discussed in previous sections, property riglnts define the distn’bution of land and
determine how rents from land will be distributed. Bell states that the "incomes of the rural
poor depend heavily not only on the efficiency with which land is used, but also on the
distribution of the rents generated by its use" (1990: 144). Traditional African land tenure
systems, based on communal landholding, limited the degee of skewness in individual
incomes (Schmid, 1987: 144). However, during the colonial period, governments advocated
individual tenure because “land tenure [was] expected to evolve from the simple form of
tribal ownership to individual ownership” (F eder and Noronha 1987: 148). Consequently,
colonial governments developed different administrative rules to restrict indigenous farming,
including the delineation of districts which ultimately reinforced ethnic differences and
restricted the movement of people outside their districts (Feder and Noronha, 1987).

Following independence, African governments adopted different tenure systems
which permitted varying degees of communal and individualized tenure. Feder and

Noronha (1987) classify these systems into three categories: 1) those which permit

41

individual titles (exchange rights, freehold status); 2) those which recognize different types
of tenure (use and exchange rights systems); and 3) those which vest title to land in the state
(use rights only, leasehold status). They suggest that although colonial governments
distorted tenure systems in Africa, there has always been some degee of individualization
in African tenure systems because it offers security and a form of collateral to farmers.
Feder and Noronha (1987) conclude that to ensure increased land use efficiency,
governments must promote some form of individualized tenure (either long-term leases or
titles), depending on the costs of instituting such a system.

In contrast, Cohen (1980) contends that traditional tenure systems offer more
security and flexibility in land use than critics usually assume (also see Hunter and Mabbs-
Zeno, 1986). He states, ”proponents of reforms converting corporate [communal] patterns
to freehold rarely consider whether there might be advantages in corporate tenures that
should be preserved..., or whether credit systems and improvements in land use can be
promoted by strategies other than complete transformation to freehold" (1980: 355).

Similarly, Barrows and Roth (1989) argue that other institutional variables, in
addition to land tenure laws, affect investment levels and agicultural incentives. They found
no evidence in their three-country study (Kenya, Zimbabwe and Uganda) to support the
thesis that conversion to a freehold system increases the demand for capital and the
aggregate supply of credit, as presumed by the neoclassical model. 10

"In the presence of economic opportunities in agriculture, the institutions governing

control of land can constrain development if inflexible rules of tenure prevent
movement of resources among individuals or if tenure insecurity lowers investment

¥

1° Barrows and Roth (1939) did find, however, that individualized tenure in the
Prmnce of a land market may transfer land to those who are able to extract a higher value
if?!“ the resource. It should be noted that this result would be inconsistent with income
'bution objectives since those who were unable to obtain an equally high value from
Would have no access to the resource.

42

demand... However, it is clear that, in the absence of profitable technological

options, registration [for freehold tenure] will have little effect on investment and

productivity in agiculture” (1989: 40).

Therefore, although changing a tenure system from communal to freehold produces new
economic opportunities, registration of freehold titles may be a costly and unnecessary
undertaking, unless other institutional constraints which constrain agicultural production
are first removed.

Given the diversity of the institutional conditions shaping agicultural development
in African countries, there is certainly no one prescription for institutional reform in tenure
relations (Cohen, 1980). Hunter and Mabbs-Zeno point out that "policies which facilitate
adaptation are likely to be more beneficial than widespread upheaval in tenure relations...
Resources would better be spent in facilitating the adaptability of existing institutions and
improving pricing, credit, marketing, etc., policies” (1986: 118-19).

The question that arises then is not whether traditional tenure systems should be
maintained or replaced with individualized tenure systems. Rather, if there is to be a shift
away from traditional systems in response to new economic opportunities, given the inherent
flexibility of traditional tenure systems in adapting to institutional and technological change
(Cohen, 1980; Hunter and Mabbs-Zeno, 1986), policymakers must consider how institutional
reform (with or without land tenure reform) will promote desired goals of increased
agricultural production and increased equity.

Several researchers suggest the need for empirical research to explore the dynamics
of institutional change in indigenous tenure systems in Africa (Cohen, 1980; Feder and
Noronha, 1987; Barrows and Roth, 1989). This study attempts to identify and predict the
impact of some of these institutional variables (i.e., land tenure, marketing, credit, extension

and other rural development policies) on the performance of land resettlement in

43

Zimbabwe, using relevant information from the cases presented in this study and from other
empirical research. During Kenya’s land redistribution progam, which was accompanied
by the institution of individualized tenure for former communal land farmers (freehold
status), the government combined land tenure reform with other institutional changes, to
provide resettlement farmers with incentives to increase agicultural production and to

redistribute land more equitably among European and African farmers.

2.23.1 Land Redistribution in Kenya

Land redistribution in Kenya represents a political response to population pressure
in black farming areas and to smallholders’ desire for land from which the colonial
government had excluded them (House and Killick, 1983). The colonial system of land
ownership established separate status and land rights for Europeans, Africans and Asians.
Therefore, resettlement policy in Kenya sought to induce political stability by reducing
inequality in land ownership between European and African landholders and creating
geater employment opportunities for Africans. In addition, the government hoped to
increase the incomes of settlers who moved to the schemes.

To achieve these objectives, the Kenyan government instituted the Million Acre
progam to redistribute more than 400,000 hectares to smallholders, over a five-year period.
Thus, the government created high-density schemes to accommodate the landless and
unemployed, and low-density schemes on higher quality land for incoming farmers with
more agricultural experience. Priority was often given to the previous employees of a farm
whose land was to be redistributed.

Commercial farmers sold land to the Land Development and Settlement Board on

a willing buyer, willing seller basis. The government valued land at 1959 prices (the last year

44

in which a significant number of land market transactions occurred) and paid sellers a
proportion in cash and the balance in annual installments backed by a 5% promissory note.
The government ganted freehold tenure status to smallholders on the settlement schemes,
contingent upon repayment of their land and development loans, and compliance with
recommended farming procedures and government regulations (Harbeson, 1984).
Settlement regulations stipulated that farmers could neither subdivide, transfer, nor improve
their plots without government consent (I-Iarbeson, 1984).

The government provided roads, soil conservation works, water supplies and trading
centers for each scheme (de Wilde, 1967). As de Wilde notes, however, sometimes water
supplies for livestock and household use were inadequate at the time of initial settlement.
In addition, settlers could use credit to obtain seed, fertilizer, livestock, fencing and other
inputs, but Hazlewood (1985) attributes shortages of some inputs to settlers’ inability to
achieve target incomes, and thus repay their loans. The government developed scheme
administration differently than traditional project management so that external management
would decrease gadually as the scheme reached maturity (Clayton, 1978). The Department
of Settlement prOposed to transfer scheme-level management from its personnel to the
scheme cooperative in a four-stage process, but in many cases, the Department of
Settlement preferred to retain more control in order to ensure geater marketed surplus
(Clayton, 1978).

Studies present mixed findings regarding resettlement’s ability to generate gowth
in “mutual output and farm incomes. Von Haugwitz (1972) reported that neither the
high nor the low-density farm model was superior, because at low altitudes the low-density
farms Performed better but at high altitudes the inverse was true. Leo (1989) cites data

(1964-68) from the Government of Kenya’s Economic Appraisal of the Settlement Schemes

45
which indicated that output and farm profits per acre increased more rapidly on high than

on low-density schemes. Clayton (1978) points out that an expansion of the area cultivated
on the scheme explained much of this rapid increase in farm output. Clayton found higher
cash surpluses on the high-density schemes, although overall, a geater percentage of low-
density farms achieved positive cash surpluses. Hunt’s comparative study for settlement
schemes (1967 / 68) and for large farms (1970/71) shows that in both sectors, the smaller size
of farms“ produced significantly higher yields than larger sized farms (1984).

It is virtually impossible to quantify the eflects of Kenya’s land redistribution on the
poverty and inequality under which a large percentage of the rural pOpulation lived. Even
though approximately 71,000 families benefitted from resettlement, the benefits were not
distributed equally. For example, the landless did not obtain the same opportunities
accorded to other beneficiaries and they were often allocated marginal lands. Hazlewood
(1985) found that the need to ”make settlement pay" influenced the selection of settlers to
the point that many landless were excluded from the progam because they showed less
ability to repay loans for their land. Furthermore, income concentration increased because
as Migot-Adholla observed, settlement officers encouraged absentee urban-dwellers with
sufficient income to purchase land from smallholders who defaulted on their loan payments
(Collier and Lal, 1986). Leo (1989) states that throughout the planning and implementation
phases of Kenya’s resettlement program, the government established high-density
settlements to meet political objectives (i.e., to quell discontent in the African areas), while
establishing the low-density settlements for more viable smallholder agiculture.

Repayment rates were low, and consequently, indebtedness was high among farmers

 

11 Smaller refers to those farms under 20 acres (8 hectares) in the settlement sector and
under 250 acres (100 hectares) in the large farm sector. Many settlement plots are from
8 to 24 times larger than the typical Kenyan smallholding of 0.5 to 1.5 hectares (Hunt, 1984).

46

on the resettlement schemes. There are several reasons for high indebtedness levels, which
were for the most part, attributable to oversights in planning. First, the budgets developed
for the schemes provided insufficient income to meet a family’s cash expenditures (von
Haugwitz, 1972). Second, the government anticipated that a farm would achieve full
production in four years. However, the initial loan installments were due in full at the end
of the first six months and settlers often had difficulties generating sufficient income to
begin payment on these installments (von Haugwitz, 1972). Third, the farm budgets
provided little allowance for crop failures due to drought, payment delays, heavy plowing
charges levied by private firms or the obligation of cooperative members to purchase any
capital equipment left on a farm (Harbeson, 1984). Heyer (1976) cites that low repayment
rates on settler loans made the resettlement progam seem like a less viable development
program and limited the integation of the settlement schemes into the national progam
for smallholder development.

The settlement farmers employed more labor-intensive methods of production than
the capital-intensive, highly mechanized large-scale farmers. As a result, although wage
employment decreased as the large-scale farms were broken into smallholdings (Maitha,
1976), resettlement created geater employment opportunities for smallholders and their
families. Compared to other types of settlement schemes, rainfed smallholder settlement
employed more workers than the large farms (568 adults/ 1,000 hectares versus 319
adults/ 1.000 hectares), while the more capital-intensive irrigation-based resettlement
schemes employed nine times as many workers (2,807 adults/1,000 hectares) (Clayton,
1971).

During the years 1964-72, when most resettlement was carried out, goss domestic

pr 0‘1““ increased annually by 6.5%. Marketed agricultural output increased by 6.7%

47
annually, while subsistence agicultural output gew by 3.7% (population gowth, however,

averaged almost 4% annually), and manufacturing increased by 12% per annum (Hunt,
1984). After 1973, the Kenyan economy began to show signs of instability and the average
gowth rate began to decrease (Hunt, 1984).

Since resettlement, the most significant changes in the agarian structure have taken
place in the small farm sector, leaving the large-scale farming sector basically intact. This
implies that there is still a skewed distribution of land ownership in Kenya. For example,
one tlnird of all agicultural land is still held in large farms, where more than 60% are
holdings over 100 hectares (Cohen, 1980). Within the smallholder sector, 52% of the
smallholders possess less than 15% of the total agicultural land, in holdings that average
two hectares or less (Cohen, 1980). In addition, as of 1980 there were still an estimated
300,000 landless, 19% of Kenya’s 1.7 million rural households, who could no longer afford
to purchase land since title security caused land values to increase significantly (Cohen,

1980).

23 Key Factors Influencing the Success of Land Redistribution Progams

Although different political, economic and social forces shaped the land
redistribution progams discussed in the previous sections, important lessons can be drawn
from these studies. These lessons can then be applied to an analysis of resettlement policy
in Zimbabwe.

Each phase of land redistribution - planning, implementation and evaluation --
influences the progam’s impact on agicultural production, income distribution and

employment generation. The following sections outline the factors which, based on the case

48

studies reviewed in this paper, play the most significant roles in determining land

redistribution’s success as a rural development policy.12

23.1 Site Selection

The selection of individual sites for resettlement schemes, based on technical studies
of soil type, carrying capacity, access to water and other factors, determines the quantity and
quality of land that settlers will have to gow subsistence production and marketable outputs.
During the planning and implementation phases of Peru and Bolivia’s land reform progams
technical, administrative and financial constraints limited those governments’ ability to select
land with adequate short and long-term agicultural potential. As a result, many Peruvian
smallholders did not benefit from reforms biased toward the cooperative sector and received
small plots on poor soils, thus joining the gowing number of minifundistas. In Kenya,
technical studies of resettlement land provided planners with insufficient information on the
income-generating potential of farming systems, resulting in high indebtedness levels among
settlers. In Indonesia, inadequate soil testing during the planning phase led to variable soil
conditions on some sites which could not support intensive agricultural production, given
existing farming practices.

Clearly, the technical analysis of proposed resettlement sites influences not only the
Viability of agriculture in certain areas, but ultimately the distribution of incomes among

those resettled (Christodoulou, 1965).

k

. 12 Arthur Lewis first raised some of these issues in 1954 (subsequently published in
Etcher and Will. 1964; see also Christodoulou, 1965). It is evident from the case studies
Presented here that the lessons from earlier settlement attempts have not always been

mo°l'porated into more recent progams.

49
2.3.2 Settler Selection

Settler selection often influences the success of scheme-level cooperation and
management, as well as settlers’ success in farming (Lewis, 1964). In Kenya, settlers
originating from different communities encountered difficulties in forming scheme-level
cooperatives that would ensure the marketing of their produce, because they had not first
developed a community among themselves. In both Indonesia and Kenya, settlers initially
had difficulties associating with the neighboring communities because the local people
thought that settlers had access to better services and facilities.

Although the Kenyan and Indonesian authorities settled landless and less
experienced farmers, in addition to skilled agiculturalists, there is no indication in either
case that farmers with varying levels of expertise could not maintain viable farms given equal
access to services and technical support. 13 Clearly, however, more experienced farmers
will have the skills and resources to produce higher levels of output and make geater
investments in their farms than less experienced farmers.

The issue of settler selection ultimately becomes a policy question of who should
benefit from land redistribution and how the agicultural sector should develop to support
economic growth in other sectors. A welfare-oriented policy would emphasize the
settlement of resource-poor individuals, in order to help them improve their living standards.
A gowth-oriented policy would focus on the settlement of farmers with more advanced
agricultural skills to stimulate more rapid growth in agicultural output.

 

13 In Kenya, for example, many farmers settled on high-density schemes (primarily the
resource-poor - the unemployed and landless) achieved more rapid increases in output and
farm profits per acre than did the more experienced farmers on low-density farms (Von
Haugwitz, 1972; Leo, 1989).

50
This paper does not advocate one policy option above the other, but merely

recognizes the fact that, regardless of the strategy chosen, there must be some way of
incorporating resource-poor individuals into the development process. This implies either
providing these individuals with increased agiculture or non-farm opportunities in rural
areas, or providing them with job-training progams and employment opportunities in urban
areas. Without a coherent policy to incorporate resource-poor individuals into development
progams, nnigation into urban areas and conflicts over land use in rural areas will
eventually overwhelm governments’ administrative and financial capacity to mitigate these

problems.

233 Provision of Institutional and Administrative Support

Intensive institutional and administrative support are most important during the
early years of settlement, when farmers are unfamiliar with the new farming system and
have not yet developed community-level management skills. Resettlement planners should
emphasize the development and initial maintenance of scheme infrastructure to aid settlers
in establishing linkages with neighboring communities in the region. As settlements become
integated into regional networks for input supplies, transportation and marketing, they do
not need to rely on intensive government support. For example, in both Indonesia and
Kenya, the governments intended to transfer administrative responsibility from national
authorities to the scheme-level. A lack of scheme personnel has impeded the transfer of
management authority to the local level in Indonesia, while the Kenyan government
withdrew its authority more slowly to maintain control over the marketing of scheme

production.

51

The Indonesian government provided new settlers with essential inputs and
subsistence supplies until they harvested their first crops. The Kenyan government made
inputs available to settlers on credit, but this credit, in conjunction with other loan
repayments, only contributed to their eventual indebtedness. The Bolivian and Peruvian
governments lacked the resources and foresight necessary to extend input supply, marketing
and extension services to most of the reform beneficiaries. In Bolivia, however, where
extensive land redistribution occurred, studies show that where some farmers obtained
better access to existing services, they increased their production of some crops.

Macroeconomic policies also determine the impact of land redistribution progams.
For example, neither the Peruvian nor the Bolivian governments created the necessary
conditions for economic gowth following land redistribution. In both cases, price controls
and import subsidies discriminated against long-term gowth in the agiculture sector. As
Kay explains in Mosley (1985: 444), "much of the good done to agriculture by the [Peruvian]
agarian reform was undone by complementary policies which were unfavourable to
agiculture, including price controls on many home-gown food gains and subsidies on

imported foodstuffs.”

23.4 Land Acquisition and Beneficiary Payment for Land

The manner in which governments acquire land and exact beneficiary repayment
determines the impact of the progam on the number of beneficiaries and on the
distribution of incomes among those beneficiaries.

Governments often disguise expropriation as acquisition by using long-term bonds
Whose value depreciates geatly before maturity (the Peruvian and Bolivian governments

both purchased land in this manner) (Binswanger and Elgin, 1990). While such action

52

decreases government-level expenditures for land, it also decreases confidence in the
government and makes landowners less willing to sell their property voluntarily. Kenya
acquired land on a willing-buyer, willing-seller basis, accepting voluntary land sales at market
prices. The political motivation for this strategy outweighed its costs in terms of national
resources because it strengthened foreign investor and large-scale farmer confidence in the
government’s stability.

The costs of acquiring land for redistribution are very high and governments usually
finance these costs through loans and external gants. Most governments, therefore, try to
exact some payment from the reform beneficiaries to finance part of the progam. The
Indonesian government used its extensive oil reserves to finance the transmigation progam
and thus did not require transmigants to pay for their land or any services they received.
In addition, the government believed that transmigants could achieve adequate living
standards more quickly if they were not burdened with debt repayment.

The Kenyan government, on the other hand, had more limited financial resources,
and attempted to rapidly finance resettlement by accelerating settlers’ debt repayment
schedules. When farmers were unable to make their loan repayments because of crop
failures, they fell in arrears with their repayment schedules and accrued interest charges.
In fact, many subsequently lost their land titles to wealthier urban purchasers. The Kenyan
repayment plans could have been more successful had the government based these plans on
farm budgets which allowed for some of the risks that new settlers face during their initial

years.

53
23.5 land Tenure Policy and Institutional Implications

In each case examined in the previous sections, the pre-reform distribution of
property rights had produced an unacceptable distribution of land and incomes for the
landless and for those farmers who were either tenant-farmers or farmed small, marginal
holdings with limited rights to transfer or expand their landholdings. In the cases of Peru,
Bolivia and Kenya, the impetus for changing the specification of property rights originated
with those farmers whose rights to land restricted their access to more productive economic
opportunities. The Indonesian government, when faced with Java’s rapidly increasing
population and the inability of many inhabitants to generate sufficient incomes from their
landholdings, promoted settlement of the outer islands to reduce overcrowding in Java.

Institutional change in each case involved modifying property rights to increase the
flow of benefits that farmers could earn from smallholdings under individual proprietorship.
This was accomplished through: 1) the appropriation of land from larger titled landowners
and its redistribution to smallholders (Peru, Bolivia and Kenya) and cooperatives (Peru and
Bolivia); or 2) the appropriation of land from holdings under indigenous tenure law and its
subdivision to smallholders (Indonesia). In all four case studies, the property rights of the
targeted beneficiaries were changed from use rights in land to exchange rights.14
Exchange rights were specified differently in each reform. For example, in Kenya farmers
were accorded freehold status contingent upon their ability to repay loans received for land
and agicultural input purchases. In addition, exchange rights were defined such that settlers

were not allowed to transfer their land without government approval.

E

1.4 Some Indonesian transmigants held exchange fights ‘0 holdings on Java before
EMS to the outer islands, thus their access to more productive land was increased. but not
e Specrfication of their land rights-

54
Smallholders in Peru and Bolivia were given more broadly specified rights to land,

including the right to transfer or subdivide their plots. There is evidence, however, that
when the new titleholders could not pay for their land or became overburdened with debt,
the former landowners seized the opportunity to repurchase the land from them. Even
though the provision of individual ownership was an important departure from the previous
tenant-farming system, new owners were not protected by legislation ensuring that their land
could not be repurchased by the former owners - legislation which would enforce the
reform’s redistributive effects.

In Indonesia, the lack of a legal mechanism to settle land claims has resulted in
disputes between indigenous outer islanders and incoming transmigrants. In some cases, the
transmigrants have been forced to pay for their land (especially land to which improvements
have been made), or otherwise face expropriation of their property by local officials. Thus.
it is clear that tenure laws must often be reinforced by additional legislation to ensure the
long-term redistributive effects of the reform and limit any anticipated exposure of the new
landholders to the actions of other individuals. 15 The absence of such legislation in Peru
and Bolivia underlines a lack of government commitment to reform and the perceived
conflict between increasing equity through land redistribution and maintaining agricultural

production in the cooperative sector.16

. 15 Property rights define the individual’s opportunity set, or his/ her available lines of
action. In addition, property rights define the resource owner’s "opportunity to create costs
for others who are affected by the owner’s acts and to create benefits for the owner through
use or exchange" (Schmid, 1987: 6-7). The inverse also applies in that the specification of
property rights determines the exposure of the individual to others’ actions.

16 The benefits of creating additional tenure legislation must be weighed against the
fa.“ thf" the cost and complexity of legal procedures involved might mitigate the potentially
Wider “Pam of redistribution, as in the cases of Peru and Bolivia. Yet, Barrows and Roth
(1989: 41) point out, in certain cases "registration may be used to avoid negative equity
effects when farmers are threatened with loss of land to politically powerful individuals."

55

These case studies show that while land tenure systems determine who receives the
rents (and other benefits) and who pays the costs of land resource utilization, other
institutional variables influence the impact of land tenure laws on land use. For example,
adverse pricing policies stifled investment in agriculture following land redistribution in
Bolivia. Although smallholders gained freehold rights and larger parcels, the subsidies
placed on urban food prices reduced rural incomes and the smallholders’ ability to increase
their benefits from agricultural production. Incomplete specification of land tenure laws in
the outer islands resulted in conflicts over land use in Indonesia. Poorly planned loan
repayment schedules limited or negated the operating surplus earned by settlement farmers
in Kenya, and hence diminished their welfare after settlement.

Clearly, these case studies underline the importance of a thorough analysis of
projected primary and secondary benefits and costs, given available resources for

redistribution and the anticipated policy environment.

23.6 Scheme-Level Monitoring and Evaluation

Two important elements allow policymakers to influence the direction of land
redistribution programs; planning (i.e., site and settler selection, administrative and
institutional support, land acquisition and tenure policies) and evaluation. The case for
comprehensive advance planning has already been made. Once the program is underway,
evaluation provides policymakers with information on the weaker and successful aspects of
the program, and aids in future planning. Unfortunately, many evaluation efforts are
conducted mid-term or upon program completion, which leaves policymakers with little
information on interim scheme development and the appropriateness of the initial plans

(such as farm size, farm budgets, loan repayment schedules and institutional support). In

56

addition, uncertainty or lack of information about the expected performance of reform laws
with respect to desired policy goals, limits the ability of policymakers to respond to
unpredicted or unwanted outcomes as they arise.

Therefore, it is imperative for policymakers to effectively plan, monitor and evaluate
the performance of reform programs. For example, in Indonesia, weak monitoring and
evaluation of transmigration projects did not reveal problems of decreasing net returns to
agriculture and other site-specific problems in a timely manner. Clayton (1978) noted that
in Kenya, periodic surveys of critical schemes would have provided the management with
the ability to monitor scheme-level response to the agricultural system, and to evaluate levels
of service provision, thus avoiding long-term production problems. Bolivian and Peruvian
policymakers did not commit resources to interim and ex-post evaluations which would have
signaled the limited impact of these reforms, as well as the existence of noncomplementary

policies.

The implications for the study of distributivist land reforms are two-fold. First, a

broad set of institutional variables may influence the outcome of land redistribution. A

change in land tenure laws, for example, may not have the anticipated effects on agricultural

production, income distribution and employment generation if other policy measures provide

disinccntives to settlers, or limit the redistributive impact of the reform. These

contradictory policy outcomes may or may not be predictable in the pre-reform planning
Stages, but should be investigated by policymakers.

Second, interim monitoring and evaluation of land redistribution programs increases

the efl‘iciency of resource allocation by revealing areas where the program is less successful

(i.e., site selection or planning, settler selection, the provision of administrative and

57

institutional support or tenure system development), thus allowing policymakers to redirect

these resources and increase the quality of the program.

2.4 ConcluSions

Several lessons can be learned from these case studies and applied to the analysis
of Zimbabwe’s resettlement policy, which is presented in Chapter Three. First, governments
can offer incentives to increase smallholder agricultural production through land
redistribution, without a long-term decrease in aggregate output. However, there are
generally insufficient data from which to draw conclusions regarding the relative efficiency
of the newly created smallholdings versus commercial estates. A notable exception is the
Kenyan land resettlement program where data indicate that small farms generated higher
yields per hectare and used more labor than large farms, both important reform objectives.
Extensive pre-reform planning, as well as on-going monitoring and evaluation, are essential
components in ensuring that settlers can meet and maintain target production levels, and
will have the ability to make future investments and expand their farming operations.

Second, the redistribution of incomes and productive opportunities (where
redistributive equity is an important goal) is highly contingent upon the breadth of the
reform program and the degree of government commitment to ensuring that benefits are
realized by the most disadvantaged groups. In addition, in order to maintain adequate
income levels in the reform sector, successful credit programs and debt repayment plans
should be based on the income-generating potential of the farming system and on the
scttler’s anticipated cash-flow pattern.

Third, the value of employment generation through land redistribution depends on

the amount of remuneration to be derived from agriculture, as compared to alternative

 

58

employment possibilities (previous employment and present non-farm opportunities). In
Peru, Bolivia and Kenya, many wage employment opportunities were eliminated by with the
subdivision of large farms. Although an increase in smallholder agriculture creates new
employment opportunities, their value to the reform beneficiaries may be limited if these
opportunities do not remain profitable.

Finally, country-specific research and policy analysis on land tenure systems is
essential to understanding the incentives and impediments provided to smallholders,
particularly in the face of distributivist reform programs that may involve changing
traditional tenure systems. In addition, other institutional variables affect the ability of
smallholders to respond to a change in either use or exchange rights in land, and ultimately
determine the degree to which reform goals are reached. Site-specific monitoring and
evaluation programs provide scheme-level management personnel with information
regarding the impact of these different institutional variables on settlers’ ability to adapt to
a new farming system, achieve an adequate standard of living, and benefit from the
redistributive nature of the reform. The collection and analysis of such information allows
management personnel and policymakers to respond more readily to problems that might
impede the long-term performance of the program.

Chapter Three will examine the achievements of Zimbabwe’s Rural Resettlement
program to-date. The program’s implementation and evaluation will be assessed in the
context of: 1) the theoretical arguments regarding land redistribution and land tenure

issues, 2) lessons from the case studies presented in this chapter, and 3) empirical studies

from Zimbabwe.

 

CELKPTIHKTIflREII

RURAL RESETTLEMENT POLICY IN ZIMBABWE

The first section in this chapter examines the historical evolution of land policy in
Zimbabwe which led to an inequitable land distribution between large landowners and small-
scale black farmers; the objectives, goals and subsequent implementation of the resettlement
policy designed to reallocate land and incentives to small farmers. One of the principal
debates centers on the ability of the land reform policy to redistribute land while sustaining
or increasing current agricultural production levels.

The second section focuses on the impact of resettlement on key issues of land
distribution, crop and livestock production, income distribution and employment generation.
It also examines issues related to the implementation of resettlement policy; including land
tenure policy, women, the landless, soil conservation, and the potential for regional
development.

The third section evaluates the success of the resettlement policy in meeting its
objectives, and examines the degree to which current evaluation methods provide

policymakers with accurate interim information on scheme-level problems and progress.

3.1 History of Land Policy in Zimbabwe During the Pre-Independence Period
Since the early colonial period, black and white farmers have been locked in a

political and economic struggle for land. Initially, colonists entered Southern Rhodesia (now

59

60
Zimbabwe) from South Africa in hopes of exploiting the country’s mineral resources. As

it became clear that the country’s mineral wealth was not equivalent to a Second Rand
(named after a gold-vein first discovered in the South African Transvaal), the white settlers
turned to agriculture. At this time, they found a thriving agricultural economy among the
Africans who grew a variety of produce and engaged in local and regional trade (Palmer,
1977).

In 1894 the colonial government established the Native Reserves to contain Ndebele
and Shona farmers in Matabeleland. The Reserves served to: 1) to suppress blacks after
the Ndebele uprising; and 2) provide white settlers with a source of labor for mining and
farming. By 1910, white settlers had claimed approximately 23% of the land in Rhodesia.
while allocating only 26% to blacks as Native Reserves (Herbst, 1987b).

In 1914, the colonial government established the Native Reserves Commission to
oversee the Reserve areas and plan their future development (Parliament of the
Commonwealth of Australia (hereafter designated PCA), 1980). At this time, the colonial
government established more than 104 separate Reserves, ranging in size from 2,100 to
625,000 hectares (Akwabi-Ameyaw, 1988).

White agriculture came to play an increasingly important role in Rhodesia’s
economy, and with its growth came the decline of African farming and trade (Palmer, 1977).
To further protect their agricultural power from competition with the black farmers, the
emerging white agriculturalists pressured the government to draw up separate land purchase
areas for black and white farmers to further segregate the two races. As a result, the Morris

Carter Land Commission (1925) created Native Purchase Areas. This legislation allowed

61
Master Farmers 17 with sufficient resources to purchase land with freehold title. Although

white farmers developed the Purchase Areas to segregate blacks into their own designated
areas, black associations supported the legislation which would also secure more land for
black farmers (PCA, 1980).

The Land Apportionment Act (1930) tightened the colonialists hold on agricultural
land by prohibiting Africans from owning any land outside of the Purchase Areas, legalizing
the separation of the two races (Moyana, 1984). By 1930, white settlers had appropriated
approximately 50% of Rhodesia’s land, leaving black farmers with about 30% of the
allocated land (Herbst, 1987b). The Act formally delineated the Native Purchase Areas
where black farmers could purchase and own land, if they had the requisite agricultural skills
and resources. The Act divided Rhodesia’s land area as follows: 8.8 million hectares held
in Natives Reserves; 3 million hectares held in Native Purchase Areas; 19.9 million hectares
reserved for white farming and urban areas and a remaining 7.2 million hectares of
unallocated land (PCA, 1980). The remaining shreds of African independence were
destroyed during this period by the white farmers who received credit and inputs from the
government, thus undermining the competitiveness of the African farmers and driving them
to subsistence cultivation on marginal lands (Palmer, 1977).

In addition to implementing land policies that encouraged black farmers to move
into the Reserves, or buy land in the Purchase Areas, the colonial government developed
agricultural policies to reduce competition from black farmers (PCA, 1980). For example,
the Maize Control Act limited access to maize marketing outlets for black farmers and the

Cattle Levy Act levied a tax on all cattle in order to subsidize the settlers’ cattle industry

k

. 17 Master farmers are certified by extension agents as having passed field courses in
unproved CI'Op production methods.

62
(Moyana, 1984). By 1935, as a result of both overcrowding and overstocking, many black

farmers were forced to work for wages in the mines and on commercial farms because they
could no longer meet their expenses with earnings from gain and cattle in the Native
Reserves (Moyana, 1984).

As land degadation continued in the Reserves, government enacted legislation to
modify land use in the Reserves, believing that indigenous farming practices were the cause
of degradation .. and not severe overcrowding and land scarcity. The African Land
Husbandry Act (1951) attempted to change the traditional communal tenure system into a
private landholding system, and ”to provide for the control of the utilisation and allocation
of land occupied by natives and to ensure its efficient use for agricultural purposes, [and]
to require natives to perform labour for conserving natural resources..." (Wekwete,
forthcoming: 6). This attempt at changing land use failed because the individual holdings
were too small to promote the desired farming practices. In addition, government assigned
insufficient staff and services to support the measures in the Reserves (Moyana, 1984). As
a result, land deterioration continued and segegation exacerbated racial and political
tensions (Moyana, 1984).

The Land Tenure Act (1969) legally defined the races as "European” and ”African"
and enacted stricter prohibitions against the lease, ownership and occupation of land in
white farming areas by black farmers. In essence, Rhodesia’s land area was now divided in
half between African areas and European areas, with the remaining portion designated as
national land. This ”equal allocation“ of land overlooked the fact that by this time, whites
made up about 5% of the population and blacks 95% (PCA, 1980).

While the Land Tenure Amendment Act (1977) removed racial restrictions for all

lands mat for the Tribal Trust Lands (PCA, 1980). The wheels of change turned very

63
slowly. Even though the 1979 Land Tenure Repeal Act eliminated all existing racial

restrictions on land in non-rural areas, the distribution of land ownership remained relatively

unchanged until the institution of the Rural Resettlement Policy in 1980 (PCA, 1980).

3.2 Agrarian Structure and Land Distribution in Zimbabwe

Historically, Zimbabwe has depended on its agiculture sector as a catalyst for
economic gowth and it still plays a significant role in providing employment and generating
foreign exchange earnings, particularly the large-scale commercial farm sector.

Today, agiculture is the most important source of employment. Commercial
agriculture employed about 35 % of the workforce in the mid-1970s, and 25% in 1983 (C80.
1987 : 44-52). Overall, approximately 70% of the population derives its main source of
livelihood from smallholder agiculture (Bratton, 1990).

In addition, agiculture finances gowth in other sectors through its foreign exchange
earnings, representing between 53% and 60% of all foreign exchange earned during 1978-
1984 (C80, 1987: 170). Through its successful agicultural gowth strategy, Zimbabwe has
become the largest exporter of food and cash crops in the Southern African region. playing
a significant role in regional food security (Bratton, 1990).

Zimbabwe is divided into five Natural Regions (agoclimatic) which have served as
the basis for Zimbabwe’s agicultural policy and land use planning for the last 20 years
(Stoneman, 1982). Although 85% of Zimbabwe’s land area is used for agiculture (forest,
parks and national lands comprising the rest of the country’s land area), extreme variations
in rainfall render some areas of the country suitable only for extensive livestock gazing.

R‘igion 1 receives high rainfall (over 1,500 mm annually) and supports specialized

farming such as tea, coffee and intensive livestock production. Region 11 receives

64
moderately high rainfall (750-1,000 mm annually) and is suitable for intensive maize,

tobacco, cotton and wheat production, as well as livestock. Region III receives moderate
rainfall (650-800 mm annually) accompanied by nnid-season dry periods, and is best suited
for livestock and some types of crop production. Region IV experiences periodic drought
(450-650 mm of annual rainfall) and is suitable only for semi-extensive livestock production
and drought-resistant crop production. Region V is dry with erratic, low rainfall (below 600
mm annually) and suitable only for extensive livestock gazing (Stoneman, 1982; Weiner et
al, 1985).

Until 1980, four primary farming systems characterized agiculture in Zimbabwe;
large-scale commercial farms (LSCF), small-scale commercial farms (SSCF), communal area
farms (CAs) and state farms. With the 1980 land resettlement policy, the government
created the resettlement areas. land tenure laws, production organization, capital and
technological investments and management practices differentiate these four systems (Maya.
1986). For the purposes of classification, the following section describes the SSCF and state
farms. However, future sections examine only the LSCF and the CA3 which have more

relevance for the subsequent analysis.

3.2.1 large-Scale Commercial Farms

Prior to independence, there were approximately 6,000 large commercial farms,
ranging from irrigated farms of 100-200 hectares to ranches covering 1 million hectares in
more arid areas (Stoneman and Cliffe, 1989). LSCF farmers hold freehold rights to their
land, Which lies primarily in Natural Regions I and II. Before 1980, this sector constituted
15.3 million hectares or 39% of the total land in Zimbabwe. Even in the LSCF sector, land

is very unevenly distributed. For example, Herbst reports that 10.5% of the farmers own

65
60% of the farmland, while 55 % of these farmers own less than 10% of the commercial

farmland (Herbst, 1987b). Less successful farmers live in the poorer agoecological zones;
43% of the poorest farmers produce less than 10% of the commercial output (Herbst,
1987b). On the other hand, the LSCF sector also contains large, higlnly capitalized estates
owned by multinational corporations, particularly ago-processing enterprises. (Mayo, 1986)
Compared to communal area farmers, commercial farmers employ more advanced
technology and farming methods. Yet, some critics argue that they underutilize large
portions of their land (Stoneman and Cliffe, 1989). On average, only 5% of commercial
land is cropped, while gazing land constitutes more than 60% (Bruce, 1990). Riddell (1978)
contended that geater access to direct and indirect assistance (loans, price supports, capital
gants, low employment costs and artificial land prices) allowed farmers with unused or
underutilized land to continue farming in an inefficient manner.18
. The LSCF sector produces valuable export crops and provides employment for both
seasonal and permanent workers. For example, prior to resettlement, the LSCF sector
produced 90% of the marketed maize and most of the other export crops (tobacco, wheat,
coffee, tea, sugar). Similarly, as of 1974, the LSCFs employed a total of 336,000 workers

(Stoneman and Cliffe, 1989).

18 There are many estimates of land availability in Zimbabwe and land use intensity
on commercial farms. For example, the Whitsun Foundation determined that only 7% of
the country was suitable for intensive rainfed crop production, and that the ratio of crap
to total area in the LSCF in 1974 was only 4.1%. Riddell claimed that only 15% of arable
land m the LSCF was cultivated in 1976. Finally, Weiner et al. found that 35-52% of
Zimbabwe’s arable agicultural land may be underutilized (Weiner et al., 1985).

66
3.2.2 Communal Areas

The CAs are located primarily in Natural Regions III, IV and V and encompass
16.4 million hectares of land. Prior to independence, they supported 750,000 families, in
addition to a large nnigant population. Rapid population gowth (both human and animal)
in the CAs has resulted in severe land degadation. Mayo (1986) reports that between 1961
and 1977, the number of farmers increased from 349,000 to 675,000 (88% increase) and the
total area cultivated increased from 1.1 to 2.2 million hectares (91% increase). For this
same period, cattle herds increased by 70% from 2 to 3.4 million head, resulting in a net
reduction in gazing area (Moyo, 1986).19 Moyo estimated that in 1980, 57% of
Zimbabwe’s population lived in the CA3, cultivating arable plots of 1 to 10 hectares. Until
the mid-1980s, when the government improved marketing and extension services for maize
and cotton, low productivity and rainfed ox-plow cultivation characterized this sector (Bruce.
1990:1).

Under the landholding system in the CAs, local officials allocate use rights to
households for both arable and gazing lands. These rights extend to fuelwood, thatch.
water, and wild fruits and vegetables, and can be transferred through inheritance (Cusworth
and Walker, 1988). In addition, livestock owners share responsibility for communal gazing
land, but increased population pressure has resulted in their deterioration, as well as a
reduction in crop fallowing and pastureland (Stoneman and Cliffe, 1989).

' Bruce (1990) questions whether traditional land use patterns still exist today,
especially given colonial intervention and extreme population pressures. Stoneman and

Cliffe point out that ”the communal authorities responsible for land allocation were

§

19 The sudden increase in small and large livestock holdings in the communal areas was
a result of the lifting of compulsory destocking laws in 1961 (WhithWr 1935)-

67
weakened through successive [administrative law] changes. The chiefs’ and headmen’s

authority was first removed then reinstated by the colonial authorities... In the 1980s they
have clawed back some de facto influence over land..." (1989: 133). Changes in the tenure
system, as well as a lack of resources have contributed to underdevelopment in the CA3.
In fact, gowth in the LSCF sector has often been linked to underdevelopment in the

Communal Areas (Stoneman and Cliffe, 1989).

3.2.3 Small-Scale Commercial Farms

Small-scale commercial farms were originally designated as African Purchase lands
by the Iand Apportionment Act (1930). Prior to independence, approximately 8,500
farmers held freehold land rights for units ranging in size from 50 to 200 hectares (Maya.
1986). Located primarily in the poorer natural regions, these farms cover 2.5% of the total
land and hold 2.5% of Zimbabwe’s population. One-tenth the size of the LSCF, and much
less productive (Akwabi-Ameyaw, 1988), the SSCF typically contributes only a small amount
to national marketed surplus (Herbst, 1987b). The crops gown in the SSCF are similar to
those gown in the communal areas (i.e., maize, cotton, goundnuts and sunflower, with
some soyabeans, rapoko and tobacco). Land utilization rates are low and the sector tends
to produce below potential (Bruce, 1990). Bruce attributes the sector’s poor agicultural
performance to inadequate service provision (particularly credit) and the admission of less-

skilled communal area farmers onto SSCF land.

3.3.4 State Farms

In the 19603, the government created large, heavily mechanized and irrigated estates

to serve as growth points in relatively arid areas, intending to eventually privatize them

68
(Moyo, 1986; Bruce, 1990). After independence the government integated these farms into

the Agriculture and Rural Development Authority (ARDA). ARDA, which produces
cotton, wheat, and tea and other crops in the more marginal areas, is now responsible for
Model C resettlement schemes (core estate schemes). Although the estates produce export
crops, they also produce wheat and rice to substitute for imported products (Mayo, 1986).
Currently, ARDA supports 28 estates which generate $20 million annually in foreigl

exchange, and employ 4,000 permanent staff and 17,000 casual laborers (Bruce, 1990).

3.3 Post-Independence Land Policy: The Rural Resettlement Program: 1980-1990

Prior to independence, Zimbabwe’s agicultural strategy focused on channeling
resources into the commercial sector, in order to generate gowth by creating wage
employment and enqnort earnings. The 1980 Rural Resettlement progam represents a large
and long-term diversion of financial and human resources into the development of
smallholder agiculture.

In 1980, Zimbabwe gained independence from Great Britain as a result of the
Iancaster House Ageement. Through the Ageement, the new socialist government - led
by Robert Mugabe and the Zimbabwean African National Union (ZANU) - developed a
land policy reform designed to dissolve the strong racial lines dividing the country. The
Ageement addressed the land issue by placing extensive conditions on the reallocation of
Property rights. These conditions prohibited the government from altering the existing
ownership patterns, except througln the purchase of land at market prices in exchangeable
currency on a willing-seller, willing-buyer basis (Ministry of Finance, Economic Planning and
Development, 1986).

69
The Iancaster House Ageement also ensured that property could not be

expropriated without adequate compensation to the owner, which meant that the
government had to allocate more resources to land purchase (Ministry of Finance, Economic
Planning and Development, 1986). The clause was valid for 10 years (until 1990) and could
only be changed by unanimous ageement of the House of Assembly and a 2/3 vote from
the Senate (PCA, 1986). Although the government enacted this condition to reassure white
settlers about the security of their property and assets under the new government, it also
increased the financial and physical resources needed to carry out the resettlement progam.

The Constitutional Amendment Act (1991), developed after the Iancaster House
Ageement expired, tried to improve upon the limited success of the 1980 Resettlement
Policy by introducing legislation to make sufficient land available for resettlement at lower
market prices for government purchase. The new policy statement instituted price controls
on land to reduce its market value, allows compensation for land in domestic currency, and
makes provision to regulate the minimum and maximum size of farms. The law also

prohibits absentee landlords and land sales to foreigners (Morna, 1990; St. Clair, 1990).

33.1 Objectives of the 1980 Resettlement Program

The stated objectives of Zimbabwe’s resettlement policy are to improve rural living
conditions, particularly for communal area farmers, the landless and the unemployed.
Equally important, however, is the integation of the resettlement sector into national
agricultural development. Therefore, resettlement should not be a deficit sector which

continually absorbs national resources, but should produce a marketable surplus and

contribute to national development.

70
The government outlined the principal objectives of Zimbabwe’s Rural Resettlement

progam as:
0 alleviating population pressure in communal areas;
0 extending and improving the base for productive agiculture in the peasant farming

sector through both individuals and cooperatives;
0 improving the standard of living of the largest and poorest sector of the population;
0 ameliorating the plight of people who were adversely affected by the war and
rehabilitating them;

0 providing, at the lower end of the scale, opportunities for the landless, unemployed
and the destitute;

0 bringing abandoned or under-utilized land into full production as one facet of
implementing an equitable policy of land redistribution;

‘ expanding or improving the infrastructure and services needed to promote the well-
being and economic productivity of rural inhabitants; and

0 achieving national stability and progess in a country that has only recently emerged

from the turmoil of war (Zimbabwe National F armer’s Union, 1987 : 1).

The government’s objectives for economic gowth and agicultural development
provided policy direction for the resettlement progam. For example, the 1989 National
Land Policy document states that it "is recognized that without increasing agicultural
production, there would be no point in changing the land ownership pattern in the
economy... An increase in agicultural production will assist in the achievement of other
indirect objectives like food self-sufficiency, forex earnings and improvement in the balance
of payment position“ (MLARR, 1989: 5).20 Therefore, although resettlement policy is
intended to redress the inequitable distribution of land and create employment, to be
pOIitically acceptable, increased agicultural production must accompany land redistribution,
to meet both domestic and external priorities.

g

.20 Additional agicultural policy objectives include: 1) maintaining a balance between
equity, productivity and sustainability, 2) promoting emergent black large-scale farmers, 3)
creating employment and 4) land tenure reform (MLARR, 1989).

71
3.3.2 Resettlement Financing

Resettlement receives both government and external funding. The government of
Zimbabwe meets approximately 30% of the overall expenditures for resettlement, which
includes funds spent on land acquisition and development, planning, administration, and all
of the recurrent progam costs (Cusworth and Walker, 1988). While it is difficult to
estimate the total amount of domestic resources committed to resettlement, Cusworth and
Walker place the cost of resettling 40,000 households at approximately ZS 146.1 million,
including 2380.8 million in land acquisition and initial development costs, and an additional
2565 million for the eventual completion of land development work.

The Department of Agicultural and Extension Services (AGRITEX) spends an
estimated Z$500,000 annually on resettlement planning, and the Department of Rural
Development (DERUDE) spends Z35.0 million annually on resettlement scheme
administration, bringing the total annual recurrent costs to Z355 million, or 23137 per
settler (Cusworth and Walker, 1988).21 Annual government outlays on local government
administration alone in the Communal Areas total 2325 per household (Cusworth and
Walker, 1988).

External donors provide additional funding for resettlement, particularly to cover the
costs of land purchases, including; a United Kingdom gant in 1980 (332 million); the
African Development Bank/Fund (2327.2 million); Kuwait (257.8 million); the European

Economic Community (236.3 million) - a total of Z3733 million. Since 1986-87, no other

21 1982-83 budget estimate from the Overseas Development Administration evaluation
report (Cusworth and Walker, 1988). This calculation does not include the expenditure of
DECODE (Department of Cooperative Development) on services for the Model B schemes.
DECODE was transferred out of MLARR in the 1985 reorganization.

72
external sources of funding have been allocated to resettlement (Cusworth and Walker,
1988).

333 Organizational Arrangement for Resettlement within the National Government

A myriad of different ministries plan and implement resettlement, coordinating all
activities from site selection to scheme evaluation. A scarcity of trained personnel, as well
as bureaucratic confusion following the 1985 reorganization of the Ministry of lands,
Resettlement and Rural Development (MLRRD), caused delays in resettlement
implementation and problems in ministerial coordination.

After independence, the government created MLRRD to guide the resettlement
progam. Several departments within the Ministry of Agiculture, as well as other
ministries, implement resettlement on the gonad. For example, AGRITEX provides
planning and technical assistance. The Agicultural Finance Corporation (AFC) supplies
scheme-level credit, including short- and medium-term credit, and credit management
education (Wekwete, forthcoming). The Agricultural and Rural Development Authority
(ARDA) administers all of the state farm operations and Model C schemes.

During the initial years of the progam, the Ministry of Local Government and Town
Planning (MLGTP) maintained control over the rural and urban local authorities, which in
turn restricted MLRRD’s participation in rural development planning (Wekwete,
forthcoming). Within the Ministry of Lands, the government established a monitoring and
evaluation unit for resettlement schemes (M&E Section). The M&E unit, however, suffered

from a lack of supporting resources and did not produce a systematic assessment until 1986

(Cusworth and Walker, 1988).

73
In 1985, the MLRRD was merged into the Ministry of Agiculture to form the

Ministry of Lands, Agiculture and Rural Resettlement, formerly the Ministry of Local
Government and Town Flaming (Cusworth and Walker, 1988). Thus, the MLARR and the
Ministry of Local Government, Rural and Urban Development (MLGRUD) now share the
administration of the resettlement progam. The Department of Rural Development
(DERUDE) was reorganized and incorporated into MLGRUD, becoming the principal
implementor of the progam, particularly at the scheme-level.

At the provincial level, Provincial Rural Development Officers coordinate the
activities of the scheme-level resettlement officers. Resettlement officers, employed by
DERUDE, register settlers, allocate land, issue permits, supervise and monitor scheme
activities and collect socio-economic data on the settlers. Since resettlement areas do not
fall within any local government authority, DERUDE alone administers them (Cusworth and
Walker, 1988). Wekwete contends that because of the scattered nature of the resettlement
schemes and lack of personnel, it is difficult for DERUDE to develop regional strategies.
Instead, the government develops schemes according to available funding and land.

Other ministries providing assistance to the resettlement progam include; the
Ministries of Energy, Water Resources and Development; Education; Health; Construction
and National Housing; Community Development and Women’s Affairs; Cooperative
Development; Transport and the Natural Resources Board (Akwabi-Ameyaw, 1988).

Although the government hoped that ministerial reorganization would improve
resource coordination between the resettlement and communal areas, it has significantly
impeded the flow of resources into the resettlement progam (Wekwete, forthcoming). The
reorganization spread the limited number of trained personnel too thinly to ensure

coordination among ministries or commitment to resettlement’s success.

74
33.4 Criteria for Land Acquisition
The government based the initial criteria for land acquisition on a set of optimal
conditions for locating and developing resettlement schemes, drawn up during the Lancaster
House Ageement. Several years after independence, the land market withered as
commercial farmers were assured that the government had no plans to expropriate their

property, and most of the remaining land for sale was in smaller parcels in less desirable

areas.

The 1980 resettlement policy statement allowed the government to purchase land for

resettlement according to the following guidelines:

a) Iand suitable for resettlement which is near or adjacent to communal areas,
particularly those under heavy population pressure, is preferred.

b) The land should not be presently farmed and preferably is already on offer for
Government purchase.

c) Blocks of land available for purchase should be large enough to allow for
economic provisions of schools, clirnies, marketing facilities, etc., even if this means
the purchase of some occupied farms on a willing-seller basis.

d) In the initial stages of the progamme, preference has been given to areas
adequately served with basic infrastructure such as roads and water supplies.

e) To reduce delays to a minimum, preference is given to areas where planning

information is available (Ministry of Finance, Economic Planning and Development,
1986: 125).

Initially, the government appropriated land abandoned during the independence war,
or purchased it from whites who were unsure about the new government’s intentions
(Bratton, 1990). Several years later, however, the amount of land available for purchase did
not meet the demand for resettlement scheme development. Thus, in 1985 the Land
Ac‘luisition Bill modified the conditions on land acquisition imposed by the Iancaster
House Constitution. The bill gave the government the right of first refusal (anyone wishing

to sell land must offer it to the government before the private market) and ”the right to

75
designate blocks of farms as land that the State would like to use for public purpose, in

order to ensure that farmers in the designated blocks of farms sell their land to
Government” (Ministry of Finance, Economic Planning and Development, 1986: 125). The
1990 Policy Statement (described in section 33) gave the government even geater latitude

in planning and implementing resettlement.

33.5 Scheme-level Planning and Appraisal

The MLARR conducts the initial land selection, valuation and purchase for
resettlement schemes. AGRIT'EX provides the technical, financial and economic appraisal,
as well as planning and design of the physical scheme layout and proposed agicultural
production system (Cusworth and Walker, 1988). These analyses, however, may be
unreliable in projecting the benefits and costs for resettlement smallholders, and in
estimating their land management abilities.

MLARR plans resettlement schemes using information on soil type, availability of
arable land and rainfall gathered on commercial farms before independence. The level of
crop and livestock production necessary to reach the prescribed target income for each
Natural Region determines the scheme pattern, as well as the amount of land needed to
carry a cattle lnerd for draft power and manure, and maintain good conservation practices
(Cusworth and Walker, 1988). Commercial farm data might, however, overestimate the
amount of land a smallholder needs to maintain a viable operation, since commercial
farmers typically underutilize some portion of their farms.

AGRITEX developed farm modeling techniques to determine the costs and benefits
MS to the settler household from agiculture and livestock production, using data

drawn from the commercial farm sector (Cusworth and Walker, 1988). Scheme-level

76

economic analysis consists of subtracting the capital costs, including land purchases but not
infrastructure costs (such schools or clinics) and annual operating costs, from the total farm
benefits (income from crop and livestock production). After discounting the net benefits
over a twenty-five year period, AGRIT‘EX conducts a sensitivity analysis for each scheme
by varying the scheme-level costs and benefits to determine the effects of cost increases and
benefit decreases on scheme profitability (Cusworth and Walker, 1988).

Analyses based on commercial agiculture assume geater management skill and
resources than smallholders usually possess and thus, overestimate smallholder production
levels and investment capability. Thus, current planning and appraisal techniques give an
optimistic picture of the potential for smallholder crop and livestock production in the

resettlement areas.

33.6 Settler Selection

During the early years of the resettlement progam, the government adhered to strict
criteria to choose individuals for resettlement, emphasizing the settlement of displaced war
victims or refugees returning to the country. 602 did not, however, choose settlers in
groups in order to establish communities, but selected them for individual characteristics
(i.e., farming skills or employment status). The increasing number of squatters, settling
without authorization, indicates that the government insufficiently accommodates landless
pe0ple with the resettlement progam.

Initial criteria for settler selection stated that settlers should be:
0 efi‘ectively landless, and
0 not employed (nor is his spouse); and
' poor; and
married or widowed with dependents; and

77

0 aged 18 to 55 years, physically fit and potentially able to make productive use of the
land allocated; and

prepared to give up all land and gazing rights in the communal area of origin; 9;
a returned Zimbabwean refugee who is given special consideration; 5;;

0 an experienced or master farmer willing to give up all land rights in the communal
areas and wage employment elsewhere (ZNFU, 1987 : 2).

Since many of those first settled lacked significant agicultural experience, in 1982
the government added a clause permitting the settlement of master farmers. GOZ hoped
that more experienced farmers would provide a ”demonstration effect" for less experienced
farmers (Herbst, 1987b).

Some settlers, particularly squatters from the communal areas and refugees from
other countries, have been self-selected by illegally settling on the schemes (Herbst, 1987a).
Although resettlement officials have forcibly evicted some squatters, they have allowed
approximately 50% to remain on the scheme. Herbst attributes their success at
circumventing officials and gaining access to land on the schemes to their political
mobilization and determination. Furthermore, the resettlement bureaucracy is weak and
inexperienced in confronting certain situations, which allows many squatters to pressure local
officials and take advantage of resettlement’s administrative disorganization (Herbst, 1987 3)-

Bruce notes that the selection criteria have brought together goups of people "from
different backgounds and localities [that] may have made it more difficult to develop 3
sense of community and to establish useful practices such as sharing of draft animals" (1990:

33). This poses a potential limitation to organizing settlers to achieve political cohesiveness.

78
33.7 Resettlement Scheme Models22

The 602 has designed and implemented four different resettlement farm models:
a fannily farm (model A); a cooperative (model B); a cooperatively owned core-estate
(model C) and an extensive ranch model for the more arid areas in natural regions IV and
V (model D). The government based each model on a different rural development strategy,
and therefore, each receives different types of infrastructure development. Since Model A
schemes comprise 95 % of the schemes developed to date, this paper focuses more on the
performance of Model A schemes in resettlement planning and implementation.

There are two model A schemes, the Normal Intensive and Accelerated models. The
fundamental difference between these models lies in funding their development, and
providing scheme-level services. Both of these models consist of goups of individual family
holdings formed into nucleated villages. Each family receives a 5 hectare arable plot, 2500
square meters of land for a residential plot and a communal gazing right for 5 to 15
livestock units, depending on the carrying capacity of the natural region.23 Thus, the
average total landholding ranges from 29 to 88 hectares. A typical scheme contains 500
families in approximately 15 villages, serviced by 3 to 4 schools, 6 diptanks, a service center.
and a support staff of one resettlement ofiicer and several other government employeeS-

GOZ funds the intensive schemes with the assistance of donors such as Great
Britain, the European Economic Community, the African Development Bank, the
Netherlands and several N 603. After settlement in an intensive scheme, settlers receive

one-half of a hectare ploughed at no cost, along with any necessary agicultural inputs. In

 

22 The following section is compiled from Kinsey, 1982 and 1983; Bratton, 1990;
Wekwete, forthcoming.

23 One livestock unit is equivalent to 1.5 cattle, 1.5 donkeys, 5 goats or 4 sheep.

79

the second year, the government expects settlers to obtain seasonal credit with the AFC to
pay for tillage and input purchases. In the third year, settlers can apply for medium-term
loans for larger purchases such as draft animals, a plow or an ox-cart.

The GOZ fnmds the entire development of the accelerated schemes, without external
assistance. The government designed accelerated schemes to resettle qualified farmers more
rapidly following land purchase, without waiting for DERUDE and AGRITEX to carry out
preliminary planning and appraisal procedures. The government only provides settlers with
access to drinking water at the time of settlement, but eventually upgades Accelerated
schemes into Normal Intensive schemes. In the interim period, settlers use the services and
infrastructure in neighboring communal areas, if such an opportunity exists.

Resettlement, however, has progessed more quickly than anticipated, and the
government often cannot meet the demand for services, particularly water (Ministry of
Finance, Economic Planning and Development, 1986; Herbst, 1987b). Herbst found that the
Ministry of Water Resources has only met one-third of the demand for water on the
schemes. This is a critical because 91% of the resettlement schemes have insufficient
surface water for both agiculture and household use (Herbst, 1987b).

Model B resettlement schemes are cooperatively organized and managed goup
settlements, designed to eventually form the basis of Zimbabwe’s agiculture sector. GOZ,
with some NGO financing, funds these schemes. The government develops Model B
schemes around existing infrastructure on the purchase site, and thus must pay more to
purchase the land and existing assets than for other schemes. Most of the cooperatives
purchased in this manner continue to cultivate the management-intensive specialized crops
gown prior to the land purchase (i.e., coffee, fruit or irrigated wheat). Cooperative

members own and control the agicultural inputs and outputs, and agee on the distribution

80

of any earnings from the enterprise. On some schemes, members may own livestock and
a small subsistence plot privately, but must hold all other property cooperatively.

The Model C schemes combine individual settlement with cooperatively owned core-
estates. The estate provides technical and managerial services to the individual settlers, who
in turn work for some portion of the cooperative’s proceeds. Farmers may, however, keep
their own land and livestock apart from the estate venture. The Model C scheme is
appropriate for crops that require large-scale post-harvest processing. As of 1989, however,
there were only three functioning Model C farms - producing flue-cured tobacco, dairy and
cotton (Bratton, 1990).

Model D schemes include a goup ranching and an extensive gazing model. These
models are more appropriate for areas where intensive crop production is too risky due to
lack of rainfall and poorer soils. The goup ranching scheme entitles each family to a
residential stand in a nucleated village, one hectare of arable land, a gardening plot and the
right to hold a maximum of 20 livestock units on the farm. Model D farmers may use local
off-farm employment to supplement their incomes from low-intensity ranching. To create
the latter Model D scheme, the government purchases farms near existing communal areas
and uses this land in a rotation to relieve pressure on CA land. This model focuses on
reorganizing gazing land without resettling communal area farmers.

MLARR assumes that each scheme will be integated into the district and eventually
achieve political representation through the area district council (Harbeson, 1981). Critics
argue, however, that MLARR plans each scheme autonomously, without considering the
local capacity that might already exist in a neighboring communal area (Kinsey, 1982;

Ministry of Finance, Economic Planning and Development, 1986).

81
33.8 Land Tenure Policy in the Resettlement Areas

Some authors suggest that while settlers receive larger landholdings in the
resettlement areas than farmers cultivate in the CAs, their rights are more restricted on the
schemes. For example, the present land tenure laws prohibit land transfers and constraining
farmers’ abilities to improve their living standards through nonfarm employment.

While resettlement schemes extend leasehold ownership, ultimate control of land use
and allocation resides with the state. Settlers receive three permits which regulate residence
on the scheme, crop cultivation and livestock. These permits only give use rights to land;
prohibiting settlers from subdividing, selling or transferring their land through inheritance.
They also cannot not engage in any nonfarm activities on their residence site. Furthermore,
they have no legal ”recourse against governmental actions which are supposed to be taken
in the interests of effective land management” (Harbeson, 1981: 1 1).

Bruce observes that these permits mandate “broad rights on the part of the Ministry
but remarkably few rights for the permit holder” (1990: 35), altering settlers’ traditional
methods of land use. The government may, at any time, change or terminate the permit
without notice to the permit holder. Also, the permit specifies no time period of validity
(Bruce, 1990) and requires settlers to relinquish any landholdings in the communal areas
for the period during which they hold the permit. Harbeson notes that settlers are expected
to cultivate individual plots, but gaze their cattle in common areas. In addition, to promote
cooperation and econonnies of scale in service provision, the government requires settlers
to live closer to one another than in the CAs.

The degee to which tenure uncertainty influences both the settlers’ willingness to
invest in their farming operations and their ability to obtain credit is unknown. However,

potentially insecure land rights have not prevented many settlers from moving to

82
resettlement areas. This is because settlers receive larger and better quality holdings in the

resettlement areas than they held in the CAs. In a case study carried out in regions IV and
V, Campbell et al. (undated mirneogaph: 17) found that the "allocation of land for both
cultivation and gazing...represents a significant improvement in access to resources for the
settlers.“ Villagers did state, however, that in the future their landholdings would not be
large enough for their children. Campbell’s study indicates that, for settlers, the benefits of
relatively large arable landholdings may outweigh the costs of the uncertainty in the tenure
laws, at least in the short run. However, Bruce (1986) argues that the government must
make some provision for land inheritance on the schemes, particularly since attempts to
reform indigenous inheritance systems and prevent land fragnentation are very rare. A
system that limits subdivision to a certain size, if enforced, would not only prevent parcel

fragnentation, but would increase intergenerational tenure security.

3.4 Assessment of Land Resettlement Policy in Zimbabwe
3.4.1 Achievement of Stated Targets

As of 1981, the government planned to resettle 18,000 families on 1.1 million
hectares of land each year (Table 3.1). In 1982, however, the government stated in its
“Transitional National Development Plan” that it would resettle at least 162,000 families
from 1982-1984, after which it would resettle 15,000 families per year between 1986 and
1990 (Herbst, 1987b). Yet, by the end of 1985, the government had resettled only 36,000
families on approximately 2 million hectares of land (Herbst, 1987b). As of mid-1989,

52,000 households occupied 2.64 million hectares of land (Palmer, 1990).

83

Table 3.1: Targeted vs. Actual Number of Families Resettled on All
Schemes and Iand Area Settled in Zimbabwe, 1980-1990.

 

Target Number Actual Number of Total Iand Area
of Families‘1 Families Settled (000s ha)

18,000 NA NA
72,000 25,000 360
126,000 NA 1,100
180,000 31,000 2,200
195,000 36,000 2,300
255,000 52,000 2,640

 

 

 

 

 

 

 

 

 

 

a All column figures are cumulative.

Sources: B. Kinsey, "Forever Gained: Resettlement and Iand Policy in the Context of
National Development in Zimbabwe," (1982: 96-99) and “Emerging Policy Issues in
Zimbabwe’s Land Resettlement Progammes,” (1983: 179); S. Moyo, “The Iand Question,“
(1986: 192-3); J. Herbst, 'How the Weak Succeed: Tactics, Political Goods, and Institutions
in the Struggle over Land in Zimbabwe," (1987b: 8-9); R. Palmer, 'Iand Reform in
Zimbabwe: 1980-1990,” (1990: 169).

84

Clearly, with respect to the number of families settled, the resettlement progam has
not progessed as rapidly as initially planned. Financial constraints, organizational
difficulties at the national level, insufficient land for sale and political pressures to decelerate
resettlement influenced the pace and quality of resettlement scheme development. Funding
constraints, which developed after the 1983 budgetary reforms, were compounded by a less
than anticipated inflow of donor funds and the reallocation of some resettlement funding
to communal area development (Ministry of Finance, Economic Planning an Development,
1986; Bratton, 1990). Furthermore, land prices skyrocketed in the mid-19805, significantly
increasing the costs of resettlement, as commercial farmers became assured of future
political and economic stability in the country (Palmer, 1990).

Political pressures exerted by interest goups geatly influence the availability of land
for sale. Once the government purchased most of the initially available commercial farms,
the remaining sellers held a bargaining advantage over the government which allowed them
to curb the pace of resettlement by withholding their parcels from the market (Bratton,
1990). In addition, the CFU intensively lobbied the government to reduce the rate of
resettlement, arguing that the subdivision of large farms would adversely affect aggegate
agicultural production and stifle foreign investment (Palmer, 1990; Bratton, 1990).

Finally, while smallholders have received much of the redistributed land, government
officials and other elite Zimbabweans have purchased large tracts of land, thus concentrating
some landholding power in the hands of larger black landowners. Palmer notes that in
recent years, 500 black citizens joined the CFU, many of them politically powerful, changing
the power structure underlying the debate over the land question. He contends that “the
debate which emerged [over the land question]...was certainly complicated by the fact that

the white farmers now potentially have more powerful local allies than they did ten years

85

earlier. For some members of government, indeed, land distribution might now signify
taking land for themselves, rather than giving it to peasants" (1990: 175).

Although various administrative and technical problems prevented 602 from
settling the targeted 270,000 families by 1990, the resettlement effort is still impressive. As
of mid-1989, GOZ had resettled over 400,000 individuals, 94% of them on model A
schemes. These figures do not, however, disclose information on the types of individuals
resettled, nor on their success as resettlement farmers. Thus, the following sections
examine: 1) the impact of resettlement on redressing some of the remaining inequities in
land distribution; 2) the impact of resettlement on national agicultural production; 3) the
distribution of benefits from resettlement; and 4) the long-term viability of the resettlement

schemes.

3.4.2 Impact on Land Distribution and Agarian Structure

Since 1979, the LSCF sector decreased in size by 3.85 million hectares or 25.6% of
its pre-independence land area (Appendix B, Table B.1)24, however, the land transfers
have done little to alter the basic distributional imbalances in landholdings between
commercial and small farmers. Although land transfers occurred in each farm size category,
transfers involving blocks over 1,000 hectares, accounted for 95.5% of the total land
transferred. This is a function of the government’s desire to purchase only large tracts of
land for resettlement to achieve econonnies of scale in service and infrastructure

development on the schemes.

 

24 In addition to commercial land purchases, 504,491 ha of state land have been used
for resettlement, for a total of 3,284,000 ha of resettlement land (Roth, 1990).

     
 

86

‘ Table 3.2: Iand Distribution in Zimbabwe by Farm Sector and Natural Region:
7 1969 and 1988. ‘

 

    
 

 

 
  

 

  
  

 

  
 

 

  
 
 

 

 

  
   

a State farms and some unaccounted land missing from the total.

 

 

 

 

 

 

 

Sector 1969 1988 Natural Region (%) and Land Area (000s ha)
(%) (%)a I II III IV V
j National & 14.4 12.7 50.1 25.0 545.9 2541.1 1843.0
l unreserved (1.0) (0.5) (11.0) (503) (37.2)
1 land
? Communal 41.8 41.9 135.0 1270.0 2820.0 7340.0 4790.0
areas (0.8) (7.8) (17.2) (44.9) (29.3)
‘ Iarge-scale 40.0 28.7 202.2 3687.0 2405.4 2429.1 2489.7
commercial (1.8) (32.8) (21.5) (21.7) (22.2)
;. farms
?' Small-scale 3.8 3.2 7.3 222.2 438.3 4733 97.6
1 commercial (0.6) (17.9) (35.4) (38.2) (7.9)
; farms I
i Resettle- 0.0 8.4 30.0 590.0 1240.0 810.0 620.0
‘ ment areas (0.9) (17.9) (37.8) (24.6) (18.8)
, r- - ~ - - - _, - .— _

 

Sources: S. Moyo, "The Iand Question," (1986: 185); M. Roth, ”Analysis of Agarian
Structure and Iand Use Patterns in Zimbabwe,” (1990: 22).

87
Resettlement did not significantly change the national distribution of agicultural

land (Table 3.2). Large commercial farms still comprise almost 30% of the total
landholdings, 59% of which are located in the more fertile ago-ecological zones. In 1988,
the communal areas and the small-scale commercial farm sector comprised the same land
area as in 1969, even though the population in the former more than doubled since 1969.
While communal area reform is not an objective of Zimbabwe’s resettlement policy,
relieving population pressure in the CAs was stated as an objective of resettlement in 1980,
and clearly must be addressed by GOZ, either through increasing the redistribution of land
to CA farmers or through other interventions.

Much of the LSCF land sold to the government lies in the poorer ago-ecological
zones; in fact, 43% lies in regions IV and V (Table 3.2). Model A schemes clearly make up
the majority of land purchased for resettlement, with 58.6% of this land located in regions
III and IV (Table 3.3). As of 1987, 78% of the registered settlers lived on Model A schemes
(Roth, 1990: 127), and 1.5% of which lived on Model A accelerated schemes. Model B
cooperative schemes lie almost exclusively in region 1, comprising 12.4% of the resettlement
land area, and only 12% of the resettlement population. Model C and D schemes represent
less than 3% of the resettlement land area and less than 9% of the population.

The government’s focus on Model A schemes results partly from the U.K.’s funding
preferences for the capitalist model of land ownership, and partly from farmers’ preferences
for individual holdings. The fact that many of these settlers are intensively farming areas
suitable only for extensive livestock gazing or semi-extensive crop production leaves them

vulnerable to periodic crop loss due to drought, and hence potential food security problems.

 

1 Table 33: Resettlement Area Locations by Natural Region (000s ha), 1988.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Natural Region I
II III IV v Total
431.6 9752 5892 122.8 2137.1 l
(16.2) (36.5) (221) (4.6) (80.1)
17.2 12.9 26.9 735 130.5
(0.6) (05) (1.0) (2.8) (4.9)
43.2 54.7 3.1 0.0 332.1
(1.6) (2.0) (0.1) (0.0) (12.4)
0.5 5.4 0.0 0.0 11.9
(0.02) (0.2) (0.0) (0.0) (0.4)
0.0 0.0 0.0 58.0 58.0
(0.0) (0.0) (0.0) (22) (2.2) I
492.5 1048.2 619.2 254.3 2669.6 I
,, (18.4) (393) (23.2) (9.5) (100.0)

a Model A Intensive (infrastructure and services provided at settlement); Model A
Accelerated (rapid settlement, services extended later); Model B (cooperative); Model C
(estates combined with individual settlement); Model D (goup ranching and extensive

gazing). ‘
b Expressed as a percentage of total resettlement land area.

Source: MLARR M&E Unit 1986-87 Annual Household Survey (August, 1988).

89
3.4.2 Impact on Agricultural Production

3.4.2.1 Cropping Patterns and Aggregate Production of Major Crops

Since independence both commercial farmers and smallholders have changed their
cropping patterns of many food and industrial crops, in response to farm-level incentives
(i.e., price changes, availability of inputs, marketing outlets and extension services). The
following section illustrates these trends for five major crops; gowth rates for national crop
production appear in Table B2 in Appendix B.

Maize represents 30.2% of the total area under crop production in the commercial
sector and 45.3% in the CA3 (Roth, 1990). Since 1981, large-scale commercial production
of maize decreased by 61%, while total area planted decreased by 48% (Appendix B, Table
B3). Roth (1990) cites several factors contributing to the decline in area planted, including
the CFU’s policy promoting crop diversification and the government’s imposition of
production controls in 1986. In addition, real producer prices for maize have steadily
declined since the early 19803. Maize yields in the commercial sector currently average 4.2
tons per hectare, while yields in the communal areas average approximately 0.9 tons/ ha.

Communal area production of maize increased slightly during the 19803, while total
area planted remained relatively stable. Roth (1990) attributes yield increases in the
communal sector to increases in hybrid seed deliveries, fertilizer use and the number of
gain depots in communal areas.25 Stoneman and Cliffe (1989) explain that CA farmers
benefitted from increased access to credit, marketing channels and extension services, and

the elimination of discriminatory farm prices after independence. An estimated 90% of all

—*

25 Rohrbach (1988) states that, in the communal areas, as much of 30% of the post-war
(as of late 1979) area expansion in maize cultivation is a result of farmers reclaiming fields
abandonned during the war and opening gazing areas for crop cultivation.

90

CA farmers now plant hybrid maize and account for more than 50% of total marketing
board deliveries (Stoneman and Cliffe, 1989).

Resettlement area maize yields remain somewhat higher than those found in
communal areas and average 1.47 tons/ ha. Maize production in the resettlement areas
continues to increase, primarily due an expansion in area cultivated. In 1983, resettlement
area production represented 0.72% of total production, but gew to 19.2% by 1987. Area
planted in maize increases by an average of 31% annually (calculated from Appendix B,
Table B3).

Sorghum production in the LSCF sector accounts for only 0.8% of the total area
under cultivation, but 10.3% of the total cropped area in the CA3 (Roth, 1990). Total
production and area planted in the LSCF declined during the 19803, while yields continued
to increase from 2.4 tons/ha in 1980 to 3.0 tons/ha in 1988 (Appendix B, Table B.4).
Although breweries demand a significant amount of sorghum deliveries annually, its
importance in the LSCF sector is declining as farmers switch to higher value export crops
(Roth, 1990).

Communal area sorghum production fluctuated geatly during the 19803, whereas
total area planted increased by an average of 1.3% per year during this same period. While
yields averaged 0.4 tons/ ha - one-sixth the level of commercial yields - total production
averaged two to three times geater than commercial production. Resettlement area
production has gown from 65 tons in 1983, to 2,600 tons in 1987. The resettlement sector
now contributes 4.7% to national sorghum production, increasing from 0.1% in 1983.

Total LSCF production of tobacco (both flue and burley) fluctuated during the 19803,
and total area planted increased by 38%, due partly to an increase in nominal and real

tobacco prices since 1984 (Roth, 1990). LSCF tobacco yields have increased by over 30%

91

since the late 19703, to over 2 tons/ha, while communal farmers average a little over 0.6
tons/ha (Appendix B, Table B5). In 1988, tobacco represented 10.9% of the total area
under cultivation in the LSCF, but a relatively small proportion of total cropped area in the
communal areas.

Communal area tobacco production has increased significantly since the 19703, and
since 1981, production increased by 510%; an average annual increase of 68%. Total area
planted increased by 363% over this same period (an average of 50% annually), while yields
increased by an average of 6.9% annually (calculated from Appendix B, Table B.5).

Resettlement area tobacco production contributes an insignificant amount to
aggegate production in Zimbabwe (from 0.05% in 1983 to 0.1% in 1985), yet total
production did increase from 52 tons in 1983 to 347 tons in 1987. In 1988, relatively few
farmers produced tobacco on resettlement schemes, although it appears that the number of
farmers gowing tobacco increases yearly, particularly for burley tobacco (Cusworth and
Walker, 1988). Several factors currently constrain increases in resettlement’s contribution
to total tobacco production, including; farmer management skills, labor, lack of credit to
purchase necessary inputs and insufficient extension advice and training (Cusworth and
Walker, 1988). Data for tobacco production in resettlement areas indicate that yields may
be twice those achieved in the communal areas, but are insufficient to draw firm
conclusions.

Cotton represents 12% of the total area under cultivation in the LSCF sector and
7.2% in the communal areas (Roth, 1990). LSCF production of cotton fluctuated in the
19803, tending to decline after 1985 (Appendix B, Table B.6). In addition, prices paid to

producers have steadily declined since 1981 (Roth, 1990). Yields, however, have been

92

consistently high - twice as high as those achieved in the communal areas - which Roth
(1990) attributes to improved varieties, management, research and extension.

Communal area cotton production increased annually by 24% since 1981, while total
area under production increased by 15 %. Increased Cotton Marketing Board purchases, as
well as improved extension and training for cotton farmers in communal areas have
stimulated cotton production (Roth, 1990), however, yields appear unchanged in the 19803,
averaging 0.73 tons/ha.

Total resettlement area cotton production increased sigrificantly each year since
1982, primarily attributable to an increase in area planted; but some data suggest that yields
may be decreasing. Resettlement area cotton production increased from 0.5% of total
national production in 1983 to 7% in 1988. Yields are slightly higher than those achieved
in the communal areas, averaging 0.96 tons/ ha.

Groundnut production represents only 1% of the total area under cultivation in the
LSCF sector, and goundnuts are usually gown in rotation with tobacco (Roth, 1990).
LSCF production of goundnuts increased by 1.4%, yields gew by 0.9% and area planted
by 0.6% during the 19803 (Roth, 1990). Commercial area yields for goundnut production
are 2 to 3 times higher than yields realized in communal areas.

Groundnut production comprises 4.7 % of total crop production in the communal
areas (Appendix B, Table B.7). Production has increased annually by 24% since 1981, while
total area planted decreased by 5.5% per year from 1981 to 1989. Roth (1990) notes that
although communal area goundnut production constitutes 97% of national production, CA
farmers deliver only 3% of this production to the Grain Marketing Board. Farmers sell

most of their production on the private market or use it for home consumption.

93

Resettlement area goundnut yields, however, are higher than communal area yields
and production increased from 0.1% of national production in 1982 to 3.4% in 1988.
Cusworth and Walker (1988) attribute limited production of goundnuts in the early 19803

to initial seed shortages in the resettlement areas.

3.4.2.2 Aggregate Livestock Production

In the commercial sector, total cattle inventories have declined since 1978 (Appendix
B, Table B.8). Beef cattle inventories decreased by 21% since the 19703, due primarily to
drought, foot-and-mouth disease and a 24% decline in real cattle prices since 1982 (Roth,
1990). Dairy cattle inventories, on the other hand, have increased since independence, in
response to increased demand for dairy products and government price incentives, and also
as a function of improved management and breeding. On the other hand, LSCF small
livestock production (total holdings of sheep, goats and pigs) has decreased since 1980.

Small and large livestock holdings in the communal areas have increased markedly
since 1980. Although periodic droughts occurred in Masvingo and Matabeleland, national
cattle herd numbers gew by 2.0% per year (Roth, 1990). Communal area cattle inventories
have gown in importance from 54.0% of the national herd in 1980 to 66.0% in 1987, but
their commercial importance is rather low. In the communal areas, cattle ownership
signifies wealth and status, as well as a source of draft power, income and manure
(Gesellschaft Fur Agarprojekt in Ubersee (GFA), 1987; Cusworth and Walker, 1988). Goat
and sheep production almost doubled from 1980 to 1987, due most likely to increased
demand for meat and animal by—products such as mohair and wool (Roth, 1990). The only

data on resettlement livestock production available to the author originate from M&E

94

surveys of several resettlement schemes (ODA evaluation report, 1988), which contain no

aggegated information. Thus, they are discussed in section 3.4.2.4.

One can conclude that even though the total land area in the LSCF declined by
25.6% from 1980 to 1988, resettlement has not adversely affected aggegate crop production.
Data on area, production and yield gowth rates (Appendix B, Table B.2) show that, from
the early 19703 to 1988, a transition occurred in national crop production. In particular,
large commercial farms decreased their total area planted in cereal crops (maize, sorghum,
wheat, barley), shifting instead to industrial crop production (tobacco, coffee, cotton,
goundnuts, soyabeans, sunflower). In the early 19703, industrial crops averaged only 34.6%
of total cropped area, while in 1988 they averaged 46.6% (Roth, 1990). This shift in the
composition of crop production is due to government price controls on cereals crops and
shortages of capital-intensive inputs. Consequently, commercial farmers have shifted to
producing crops which require more labor (particularly seasonal labor which is unregulated
by wage laws), because input price controls make labor cheaper relative to other inputs. In
addition, labor-intensive crop production is generally not subject to output price controls.

It appears that macroeconomic policies have been more important determinants of
commercial agicultural production than the diversion of land resources to the resettlement
progam. Furthermore, the data reveal the importance of price incentives and market
access in stimulating production of cotton and maize in the communal and resettlement
areas. Overall, as commercial farmers respond to changing government restrictions (i.e.,
restrictions on maize production) and world market demand (for example, increased
demand for horticultural products and decreased demand for maize), small farmers, who

have a lesser investment capability to change their farming operations in response to short

95

and medium run price changes, fill the gaps in domestic production. Thus, once land
became available and the proper incentives were offered to small farmers, it is evident that
they significantly increased their marketed output.

Price policy also influenced the composition and size of livestock holdings in the
LSCF, encouraging farmers to produce fewer beef cattle, but increase their holdings of dairy
cattle. In contrast, since small farmers in communal and resettlement areas retain livestock
as wealth, insurance and an input into crop production, their inventories are influenced
more by climatic vagaries than by price levels.

Aggegate, national level data showed that crop and livestock production gew
steadily during the 19803. Therefore, to assess the distributional impacts of resettlement it
is necessary to analyze regional and household level data on crop and livestock production,
incomes and employment opportunities. Since little data exist on the distributional impact
of resettlement in Zimbabwe, this study will incorporate information from various case
studies, in an attempt to isolate some of resettlement’s impacts on farmers in different ago-

ecological regions.

3.4.23 Regional Agricultural Crop Production

Comparing communal and resettlement area agicultural production provides a basis
for analyzing the distributional impact of resettlement. Since most resettlement farmers
formerly lived in the communal areas, they have developed land management and farming
techniques that were appropriate under a communal tenure system, and allowed them
flexibility to alter the proportion of cropped and gazing land in response to climatic changes

and cattle stocking levels. In addition, when agiculture could not support the farm

96
household, families had the option of earning off-farm income locally or by nnigating out

of the communal areas.

Several government policies restricted resettlement farmers from utilizing their
communal coping strategies. In the resettlement areas, the MLARR prohibits settlers from
stocking more cattle than their allotted number of livestock units, removing some degee of
farmers’ flexibility in responding to changes in their environment. Furthermore, the
government bans settlers from engaging in off-farm activities, except Model D residents who
may earn insufficient incomes by ranching in the lower natural regions. The government
limits farmers’ crop and livestock inventories and restricts their access to off-farm income,
in an effort to reduce land degadation in the resettlement areas and encourage farmers to
concentrate their resources in agiculture. These limitations have two implications for
agicultural development in the resettlement areas. First, farmers must develop new coping
strategies during periods of drought, since they may have fewer cattle to sell when required.
Second, when agiculture is not remunerative, particularly in the event of crop failure,
farmers have no secondary source of income to draw upon.

To assess the household food security situation in the resettlement areas, Cusworth

and Walker (1988) examined 1985 M&E survey data26 from selected households in each

 

26 These data should be interpreted with care, and used only to highlight potential
problems. While the M&E unit collected data in 1985 and 1986, they also used recall
methods to collect data for 1984. All data pertain to Model A schemes of which 11 were
selected. Villages within each scheme were stratified by soil type but the settlers chosen in
those villages were often among the best farmers. Settlers were then interviewed to
determine land preparation methods, and were characterized according mechanical or ox
tillage practices. Finally, from each subset, two or tlnree settlers were chosen using random
number tables. Furthermore, no data were collected from Matabeleland due to security
reasons, therefore, the data for region IV originate from Masvingo only.

97 .
region.27 The data, which included only maize and sorghum retentions (which excludes
the degee to which vegetable crops and other gains contribute to household nutrition and
income), indicated that maize production appears to satisfy household food requirements
in all regions except V (Table 3.4). In fact, in the higher natural regions, farmers used only
40% of their arable land to attain these production levels for food crops and the rest of

their land for cash crops (Cusworth and Walker, 1988).

Table 3.4: Maize and Sorghum Retention Per Resettlement Household by Natural Region in E
Zimbabwe, 1983/84 - 1985/86.a ‘

 

 

 

 

 

 

: Natural Region 1
; n 111 IV v
Year Maize Sorghum Maize Sorghum Maize Sorghum Maize Sorghum
1983/84b 16.2 0.0 129 0.0 14.0 1.8 1.6 2.6
1' 1984/85 18.7 0.0 192 0.0 17.6 2.7 11.6 145
I 1985/86 17.4 1.7c 16.1 1.0 15.6 4.1 2.3 1.4 :

 

 

 

 

 

 

a Average number of bags retained after harvest, where average bag size = 90 kg.
Drought year.

 

c Estimate based on limited number of cases.

Source: Cusworth and Walker, "Land Resettlement in Zimbabwe, A Preliminary Evaluation," (1988:
104).

 

27 Riddell estimated that each individual in the CA3 requires 385 lbs or 175 kg annually
for consumption. If the average household size is 8.5 people (Cusworth and Walker, 1988),
then a family requires approximately 1488 lbs of maize or 7.4 bags annually.

98

An examination of average area, yield and goss margins (goss income minus
variable cash expenses for each crop) for maize, sorghum and cotton, again reveals the
degee of variability across regions in the resettlement sector (Appendix B, Tables B.9-B.1 1).
Maize is the primary food crop in all the resettlement areas, while farmers gow sorghum
in the lower natural regions (IV and V) as insurance for drought periods and cotton in
higher regions (II and III) as a cash cr0p. Although yields vary from year-to-year, the
geatest variation for all crops occurs in region V, in addition to the lowest goss margins.

For example, maize yields are high in regions 11, III and IV (2.49, 1.89, and 1.88
tons/ ha, respectively) compared to region V where yields are extremely variable (ranging
from 1 ton to 0.1 tons/ha). The data also indicate that farmers in regions 11, III and IV
receive higher goss margins per hectare than in region V, yet figures for all regions show
extreme year-to-year variability. Furthermore, sorghum yields average 1 ton / ha in regions
III and IV, yet under 0.5 tons/ ha in region V. Gross margins for sorghum indicate the same
variability as those for maize. Cotton, produced only in regions 11 and III, generates less

variable yields and goss margins.

i Table 3.5: Average Area under Cultivation for Schemes in
Natural Regions II-V in Zimbabwe, 1985 / 86.

 

Natural Region Average Area under Of Total Arable

Cultivation (ha) Iand Available (%)
3.0 60
3.6 71

 

 

 

 

 

 

 

Source: Cusworth and Walker, ”Iand Resettlement in Zimbabwe, A Preliminary Evaluation,"
(1988: 1 18).

These data further highlight the risk involved in crop production in the lowest
regions, especially since almost 100% of arable land is committed to crop production in
some areas (Table 3.5). In other high-potential areas, households meet their food needs
using 60-70% of their allotted land, implying that if labor is fully employed at current
production levels, then these plots may be too large for the average resettlement household

to manage.

3.4.2.4 Regional Livestock Production

Livestock holdings play a vital role in the communal area farming system, as well as
in the resettlement areas. In addition to serving as a source of draft power, livestock
provide an alternative source of income when crop production is insufficient in meeting
household needs. Therefore, unless farmers have other options through which to generate

income (i.e., off-farm income, or more reliable cropping systems in the lower natural

100

regions), livestock production will continue to be important to resettlement area farmers,
leading to potentially serious overgazing and soil erosion problems.

A communal area livestock survey conducted by the Gesellschaft fur Agarprojekte
in Ubersee (GFA) in 1987 revealed that households with larger herd sizes have higher
maize production; often four times geater than those of households without cattle (GFA,
1987). Increased production results, in part, from increased yields due to more intensive
planting and manure application, in addition to improved timing from using draft power.
Farmers in natural regions IV and V, however, experience more frequent crop failure and
spend more time in off-farm employment, which leads to neglect of cropping activities.

In addition to livestock’s general importance as an input to production (GFA, 1987;
Cusworth and Walker, 1988), small livestock also play an important role. They serve as an
intermediary step to large animal acquisition, and often allow women to make investments
that requires less labor and capital than cattle (Cusworth and Walker, 1988).

Therefore, given the role of livestock in both the farming system and the culture,
farmers may attempt to accumulate very large holdings of both small and large animals.
Although the government issues permits to control the number of livestock per holding. it
does not enforce these stocking rates. In fact, data show that some farmers have already
surpassed permitted stocking levels, particularly in the higher potential natural regions
(Cusworth and Walker, 1988). Furthermore, where overstocking occurs on adjacent
communal areas, communal area farmers trespass onto resettlement lands to gaze their
herds. A partial cause of the overgazing problem lies in the specification of the tenure
system, which permits individual holdings for crop cultivation but mandates communal

gazing. Thus, some farmers’ livestock holding patterns, especially those with very large

101
herds, may adversely affect other farmers with smaller herds, influencing their ability to

maintain viable farming systems.

3.43 Impact on Employment

LSCF farmers employ a large number of seasonal and permanent employees; in 1989
they employed 247,154 individuals, of which 75 % were male and 25 % female (Roth,
1990: 167). During the initial resettlement years, some critics voiced concerns that the
subdivision of large farms would decrease total wage employment opportunities in the
economy (Kinsey, 1982; 1983). These critics, however, did not anticipate the slower pace
of resettlement, the effect of transforming underutilized parcels with low labor/ output into
fully productive land with high labor requirements, or the shift of many commercial farms
into more labor intensive crops (i.e., tobacco, cotton, horticultural crops and vegetables)
(Roth, 1990).

Althougln the number of LSCF employees decreased by 9% from 1980 to 1989, it is
unlikely that the decline in wage employment is due primarily to land resettlement policy.
For example, during this period, the government enacted various policies which increased
the costs of labor for LSCF employers, including minimum wage laws and regulations
expanding workers’ rights. Therefore, declining wage employment could be part of an
overall trend in agiculture toward geater investment in labor-saving technology (Roth,
1990). Although some of the decline in the labor force can be attributed to the subdivision
of farms in the LSCF subsector, Roth contends that some of the land transferred was
underutilized and mechanization may have increased the ratio of cropped area /worker from
2.1 irn 1980 to 2.5 in 1985. Furthermore, foreign exchange shortages, gowing unemployment

and increased numbers of refugees entering the work force encouraged LSCF managers to

102

hire less permanent wage labor but more seasonal labor which is not controlled by
government regulations.

Evidence from Kenya and Peru suggested that small farms achieve higher
labor/ output than large farms. Therefore, policies designed to emand the small farm sector
will absorb surplus labor in agiculture. In the case of Zimbabwe, although total wage
employment in commercial agiculture decreased by 9% during the first 10 years of the
resettlement progam (from 271,291 to 247,154 employees, or a loss of approximately 24,000
jobs), there are now 52,000 families in the resettlement sector. However, the net gain in
employment in the smallholder sector (both in terms of family and hired labor) is difficult
to ascertain because many settlers were previously communal area smallholders or employed
on commercial farms. Thus, the net impact on employment is much smaller than the actual

number of adults settled.

3.4.4 Impact on Income Distribution

Several factors affect the distribution of incomes resulting from land resettlement
in Zimbabwe, including the natural region in which the settlement is located (site selection),
the size of the holdings and farmers’ access to services and infrastructure.

First, the location of the settlement determines the intensity of cr0p cultivation, and
hence settlers’ ability to earn sufficient household incomes and repay loans acquired for
input purchases. Cusworth and Walker (1988) found that farmers in natural region V
earned the lowest goss margins from crop production, but the highest average value for

their cattle herds.28’29 Farmers in region II earned the highest total goss margins for

 

28 Ofl'take is calculated as the ratio of the number of cattle sold, slaughtered or
disposed of during the year to the total herd at the beginning of the year (Cusworth and
Walker, 1988).

103
crop production and cattle offtake (1983 / 84 - 1985 / 86), yet also the highest levels of

indebtedness; averaging over 231,500. Farmers in regions 1]] and IV also accumulated
debts of almost Z3700. Since crop production and cattle offtake rates are highly contingent
upon climatic conditions, farmers with high levels of indebtedness could, after a year of
drought, find themselves unable to repay their debts and thus be inelig'ble for further credit.

Furthermore, low loan repayment rates to date indicate highly variable farm incomes
and debt accumulation. For example, Cusworth and Walker found higlner repayment rates
following a successful cropping season, and repayment rates from several schemes indicate
that farmers more often repay short-term than medium-term loans.30 Repayment rates
for short-term loans average from 4-83% and from 0-56% for medium-term loans. In
addition to repayment problems, settlers complained about late loan processing and input
delivery, as well as costly and insufficient transport (Cusworth and Walker, 1988).

The study by Campbell et al. indicated that families with a larger holding than in the
communal areas are able to build up livestock inventories and reinforce their coping
strategies. However, evidence also suggests that resettlement area holdings may be too
large, given available family labor and other production inputs, implying that resettlement
land is not accommodating the maximum number of beneficiaries. As Lipton pointed out

(Chapter Two, section 2.1.23.2), the size of the holding determines not only its income-

 

29 These figures may underestimate levels of indebtedness among settlers since
Cusworth and Walker compare the target net farm income of Z3400 with goss margin
calculations. Thus, farmers’ incomes are not adjusted for any fixed costs they incur (ie.,
maintenance and depreciation on farm equipment and structures), nor are they adjusted for
any remittances or off-farm income earned. They assume that the two additional sources
of income and expenditure net out, but this is an unusual assumption, given that during the
first few years of settlement expenses on plows, carts and hand tools are probably high.

30 Farmers generally use short-term loans to acquire seasonal inputs (fertilizer and

seed), whereas medium-term loans (four years) purchase oxen, carts or farm implements
(Cusworth and Walker, 1988).

104
generating potential, but the number of beneficiaries served by the reform. This implies

that if holding sizes are indeed too large in some areas (Table 3.5), then reducing plot size
will allow the government to settle more households, thus redistributing incomes more
widely.

Access to services and infrastructure also determines the farm household’s ability to
earn sufficient incomes. A lack or untimely delivery of inputs and certain services such as
credit, marketing and transport limits farmers’ short- and long-term potential to generate
income and invest in improved seed and farming equipment (Kinsey, 1982; Cusworth and
Walker, 1988). Thus, services and infrastructure ultimately play an important role in

determining who benefits from land redistribution, and by how much.

3.5 Other Key Issues
3.5.1 Land Tenure

Many analysts of Zimbabwe’s resettlement policy have criticized the current tenure
policy in the resettlement areas, arguing that it perpetuates the system found in the
communal areas, which is blamed for extensive land degadation (Whitlow and Campbell,
1989). There is little understanding of the incentive structure that the communal
landholding system provides for investment and conservation, and therefore, the
resettlement areas should be studied to determine how people will respond to a different
and expanded opportunity set. Bruce (1990) notes that many people recognize that the
permit system currently used in the resettlement areas is an inadequate long run solution
to establishing land rights because it does not address problems of future land subdivision

for inheritance, or potential abuses by administrators.

105

In recent years, several parties have proposed land reform initiatives. However, the
viability of each proposal depends on who will bear the costs and who will reap the benefits,
and whether the eventual distribution of benefits and costs is consistent with the
government’s long-term objectives for rural and agicultural development.

First, the Communal Area Development Plan recommends issuing five year permits,
which would subsequently be converted to 25 year leases, and include an option for
extension to 90 years. This proposal still embodies some uncertainty (and therefore
potential costs), since settlers could be subject to administrative changes after the initial five
year permit expires. It is also a costly option for the government because each land permit
would have to be evaluated and processed three times, as opposed to once if the
government accords long-term titles or leases. The government would gain little through
this process since there must be a less costly way to evaluate settlers’ landholdings.

Second, Lionel Cliffe proposes "a system of community rather than state control of
land, urging that it can allow for reallocations of land over time to cope with new households
and reallocations which may be required in attempts to intensify land use" (Bruce, 1990: 36).
There is some doubt as to whether a community-based system would function in the same
manner as a traditional landholding system where local officials allocate land and spatially
control land use. While Cliffe’s option involves fewer administrative costs for the
government, it is doubtful that a communal system could be successful in resettlement areas.
Since settlers typically originate from different ethnic goups, they would not recognize the
same local authorities, unless the government allows each scheme to select its own
management. Thus, this option has a high potential for land use conflicts whose
confrontation and resolution would involve the costs of creating and sustaining another

administrative agency.

106
Michael Bratton (1990) suggests that according exchange rights (either limited or

fully specified) will provide sufficient incentives for resettlement farmers to make
investments to increase agicultural output.31 Since policymakers are hesitant to instate
freehold tenure in the resettlement areas, he advocates establishing some form of leasehold
tenure with either protective legal measures to prevent subdivision and sale or an extended
lease. Either of these options would provide increased security,32 guarantee more
enforceable rights in land for landholders, and would require fewer administrative costs than
registering freehold titles or reissuing permits at specified time intervals.

The case studies in Chapter Two underlined the need for policymakers to provide
for institutional flexibility in designing tenure systems. Often, other institutional variables
affect the distribution of benefits and costs from land redistribution to a geater extent than
the specification of the tenure system itself. Therefore, in designing tenurial reforms, it is
necessary to study the opportunities created for farmers in the new resettlement areas and
judge whether a change in tenure will indeed create geater incentives to smallholder

agiculture, or whether other policy interventions would achieve the desired outcomes.

3.5.2 Condition of Women: Access to Land and Production Opportnmities
Sweral authors suggest that allocating land by permit in the resettlement areas

changes the traditional access and control women have over some parcels of land

 

31 The Commercial Farmers’ Union also advocates individualized tenure in the form
of a three-year lease during which farrrners’ progess is monitored, and if they are considered
successful farmers, they have the option to gain full title to their land (CFU, 1991).

3’2 Barrows and Roth define tenure security as " the landholder’s perception of the
probability of losing land within some time period...[or] losing a specific right in land such
as the right to cultivate, gaze, fallow, transfer, or mortgage" (1989: 35). They further
emphasize that titling procedures are not homogenous and, depending on the country and
institutional context, a title may convey varying degees of security of ownership.

107
(Chimedza, 1988; Gaidzanwa, 1988).33 In addition, because of the emphasis on cash-

cropping. women may lose rights to cultivate their own fields which often provide nutritional
supplements to the household diet, as well as generate additional income.

The government typically issues permits to male heads of household, except in the
case of female-headed households, where the woman receives the permit. To date only 7%
of the permits issued have been allocated to women (Chimedza, 1988). In addition,
although both the husband and wife can be designated as co-pernnitholders, Chimedza
reports that the MLARR discourages the practice. She also found that women’s lack of
knowledge about resettlement procedures and laws prevented them from seeking jointly-held
permits.

Lack of explicit rights to land prevents women from obtaining credit, and exposes
them to eviction in cases where the husband is convicted of a crime (Chimedza, 1988).
Furthermore, in moving from communal areas to resettlement schemes, women lose
traditional ties and assistance from extended family members. Often, the only available
family labor consists of the husband and wife, and the limited contribution of school-age
children (Chimedza, 1988).34

Although women lose some traditional rights to land on the resettlement schemes,
access to larger holdings enables them to gow their own crops, and co-wives can even have

individual plots (Chimedza, 1988). In Chimedza’s survey, she found women interpianted

 

33 See also Achola Pala Okeya’s description of the impact of land tenure reform on
women in Kenya as summarized by Joy K. Green in "Evaluating the Impact of Consolidation
of Holdings, Individualization of Tenure, and Registration of Title: Lessons from Kenya.“

34 More drastic labor shortages occur in areas where services are so inadequate that
parents must send their children back to communal area schools (Kinsey, 1982).

 

then

epic

taste
then
resu

KJE

legafl

gene

3£L3

the 1
05k“
and,
Natic
the h

urbar

108

their crops with maize but not with higher value cash crops (i.e., tobacco and cotton),
typically controlled by men.

Gaidzanwa suggests that prices could significantly affect women’s roles in the farming
system (1988). For example, if groundnut prices increase (traditionally a woman’s crop),
then demands for female labor will grow as they intensify groundnut production. As a
result, women could lose control over their crops since higher prices might encourage men
to exert control over the production of these higher value crops (Gaidzanwa, 1988).

Therefore, the legal framework created for the resettlement areas limits women’s
legal control over land, exposing them to eviction in some cases, and increasing the
possibility of them losing control over their labor input and own source of revenue

generation.

3.53 Incorporating the Landless into Resettlement Schemes

Resettlement’s impact on landlessness in Zimbabwe is unknown, primarily because
the number of landless individuals also remains unknown. Yet, one of the primary
objectives of resettlement policy is to provide ”opportunities for people who have no land
and who are without employment and may therefore be claimed as destitute...” (Zimbabwe
National Farmer’s Union, 1987 : 1). Providing productive opportunities in agriculture for
the landless improves their standard of living and reduces the influx of unemployed into
urban areas.

Estimating the number of landless is difficult since many people migrate to find
employment or work permanently on commercial farms, often with no holdings of their own.
In addition, many communal area residents are essentially landless, either because they have

no land or because they farm small holdings on marginal soils (Bratton, 1990). Various

109

estimates project that landless families make up 6-12% of communal area households
(Bruce, 1990), and may number more than 330,000; 42% of the total agricultural population
and 32% of Zimbabwe’s population (Bratton, 1990).35 Although resettlement represents
an opportunity to improve conditions for the landless, the problem is twofold.

First, even if the government could meet its goal of settling 110,000 households on
15 million hectares over the next ten years, redistribution would still not reach a substantial
number of landless people. Sixty percent of those currently settled on schemes came from
the communal areas (Moyo, 1986); another 28% were landless (Bratton, 1990). Therefore,
if the government did settle an additional 110,000 families, all landless, it could
accommodate only 30% of those needing land.

Second, the landless need more than just land. Bratton (1990) asserts that the
landless and near-landless also suffer from shortages of labor, agricultural inputs and
livestock which further constrain their ability to establish viable farms. As crities argue that
the government cannot always provide services and infrastructure for settlers in a timely
manner, this could adversely affect the productivity of the landless or near-landless who
move to resettlement schemes with little or no capital and farming equipment.

Resettlement has had some impact on reducing landlessness in Zimbabwe only
because some landless families have been resettled. However, the number of landless
continues to increase yearly. If the policy’s goal is to have a wide distributive impact, then
the government must provide more than simply land to new settlers. They also need access
to support services and infrastructure which would enable them to become productive

farmers, particularly those who move to schemes with few resources.

 

35 Estimates for 1980 include 180,000 truly landless households and 150,000 near-
landless households (Bratton, 1990). Bratton also estimates that by 1990 there will be
370,000 households in need of land.

110
3.5.4 Iand Degradation and Conservation

Land degradation in the communal areas has progressed to the point where it limits
the potential for agricultural development (GFA, 1987), and some fear similar consequences
in the resettlement areas, particularly if farmers subdivide their holdings for transfer to
other family members. Two factors exacerbate land degradation; human and livestock
population pressure, and lack of information and inputs to practice appropriate conservation
methods. In addition, inadequate income-generating opportunities in the resettlement areas
could accelerate degradation by forcing farmers to cultivate their fragile land more
intensively.

Surveying 30 communal areas, the Whitsun Foundation (1983, cited in CPA, 1987)
found that the long run carrying capacity of the land would support only 36% of the existing
population. Actual cattle stocking rates sometimes surpass the recommended rates by 100%
(particularly in natural region V), leading to overgrazing and exposed soils.36 High rates
of soil erosion - five to eight times the natural rate of erosion -- further reduce the land’s
carrying capacity (GFA, 1987). Cusworth and Walker (1988) reported that some
resettlement areas have already approached overstocking, especially in the higher natural
regions where crop yields are more variable. Thus, the potential for land degradation in the
resettlement areas already exists, primarily in areas where farmers rely on cattle to
supplement their income in case of crop failure.

Although the Technical Services division of AGRITEX conducts soil conservation,
land use and veld management courses, the division is severely understaffed in terms of

branch personnel and extension workers (GFA, 1987). Whitlow and Campbell (1989)

 

36 Whitlow and Campbell found that over 60.0% of all eroded land in the communal
areas was located in regions IV and V (1989).

 

advoca
come;
and c0
technil
profita

mainta
3.6 St

fesour
Purchz
(Cusw
(2.336.
dome5
agricu.
Seetor,
prOdUl

accele
landhc
and PC

SOme

inadeq

111

advocate making materials available to extension workers and education institutions that
convey the significance of erosion problems and the importance of proper land management
and conservation. While education is important in providing information on appropriate
techniques to farmers, perhaps the key solution lies in addressing the problem of variable
profitability in some resettlement area farming systems, and hence farmers’ need to

maintain more intensive stocking levels.

3.6 Summary and Implications for Further Evaluation

Zimbabwe’s land resettlement program demands extensive personnel and financial
resources. Since 1980, the government has spent approximately 231 17.5 million on land
purchases, development and recurrent costs, excluding donor-financed land purchases
(Cusworth and Walker, 1988; Palmer, 1990), or 252256 / household settled
(ZSB669/household including donor expenditures). This translates roughly to 0.4% of gross
domestic product per year. In the short run, resettlement has not caused aggregate
agricultural production levels to plummet, as many anticipated. In fact, the resettlement
sector, as well as the LSCFs and CAs, contribute incremental increases to national
production each year. Thus, as argued in Chapter Two, resettlement has a mm to
accelerate economic growth.

On the equity front, it appears that some households benefitted from gaining
landholdings in the resettlement areas, particularly former communal area farmers, landless
families, war victims and the unemployed. The question that remains concerns the extent
and permanency of those benefits. Available data indicate, not surprisingly, that settlers in
some resettlement areas experience highly variable returns to agriculture and have

inadequate access to services and infrastructure. In these areas, where even the short-term

 

 

benefit
only a

receive
small I:
landles
of the l
are big

SUfflcie

112

benefits from resettlement are tenuous - the long-term prospects are dim. Furthermore,
only a small percentage of those most in need of land; 4% of the landless and near-landless,
received land. The majority of those settled were communal area farmers who left their
small parcels for larger, better quality landholdings. The condition of the resource-poor; the
landless, farmers in marginal areas and women, may not have improved as a consequence
of the resettlement program. Therefore, although the government increased equity, benefits
are highly skewed toward the settlers in better natural regions and those with access to
sufficient inputs, services and infrastructure.

Currently, the Rural Resettlement program is one of many rural development
programs in Zimbabwe. For example, during 1984 the government committed 23786 million
to other programs, including; irrigation schemes, urban infrastructure in rural areas, small-
scale industry, rural credit, agricultural services, conservation and reclamation works, forest
development, parks and wildlife development, rural health facilities and schools, community
courts and rural infrastructure (GOZ, 1981). Resettlement represented approximately 13%
of the total funds allocated to rural development. Yet, the government, even with a
substantial monetary commitment to the resettlement program, has little indication of its
short-term performance, much less its longer run prospects. The eventual goal of
resettlement is to integrate these individual schemes into regional economic development,
by encouraging schemes to become more self-sufficient and less reliant on government
support. Yet, it is difficult to ascertain the extent to which this long-term goal may be
achieved when there is little effort to monitor and evaluate scheme-level progress and
difficulties. Although the government established an M&E unit, the Policies and Procedures
document for resettlement does not even indicate the unit’s source of financing (Cusworth

and Walker, 1988).

 

AlthOl

goven

a stak
Finall
acqui:
only l
to M
statec
YEgio:
the g.
by ex
Socio
Impr.

fUndS

3.7 c

113

Unfortunately, political forces constrain the success of the resettlement program.
Although significant amounts of underutilized land still exist in the commercial sector, the
government will not force farmers to sell their land for fear of driving them out of
commercial agriculture. In addition, an increasing number of government officials now have
a stake in decelerating the pace of resettlement in order to keep more land for themselves.
Finally, pressure from the British government, which provides large grants for land
acquisition, shapes both the past and the future of resettlement. For example, the British
only endorsed Model A schemes which provide for individualized tenure, a form contrary
to Mugabe’s socialist vision for Zimbabwe. Also, the British government more recently
stated that it will discontinue its funding unless the 602 purchases land only in natural
regions 1V and V (Palmer, 1990). Recognizing that these important parameters constrain
the government’s actions, there is still room to increase the quality of the current program
by expending efforts to more systematically collect and analyze agricultural production and
socio-economic data regarding farm-level performance on various resettlement schemes.
Improving monitoring and evaluation could allow the government to more efficiently direct

funds, and thus increase the breadth of the resettlement program.

3.7 Conclusions

Clearly, resettlement policy in Zimbabwe is guided not only by current political,
economic and socio-cultural conditions, but also by its historical legacy of land segregation
which led to the development of the overcrowded, degraded Communal Areas and the
powerful commercial farming sector. Today’s policies attempt to rectify some of these

problems by redistributing land to smallholders who have long been denied access.

114

Since 1980, resettlement has contributed to increased aggregate agricultural
production and served to redistribute incomes to some smallholders, as well as expand rural
employment opportunities. The policy’s impact is very limited, however, compared to the
government’s initial objectives. To date, the government has failed to adequately address
several important issues, including; options for resettling a greater number of landless and
near-landless, ensuring the legal security of women on resettlement schemes, developing
strategies to prevent potential overgrazing and soil erosion problems, and seriously
investigating land tenure reform for smallholders.

The four case studies presented in Chapter Two underlined the importance of pre-
reform planning and periodic evaluation to ensure that the goals of land redistribution are
met. Similarly, the impacts of Zimbabwe’s land resettlement policy on agricultural
production, income distribution and employment generation are highly contingent on site
and settler selection, timely provision of inputs and services, and the effects of other
agricultural and macroeconomic policies in providing incentives for smallholder production.

Monitoring and evaluation contributes to policymakers’ understanding of how diverse
micro- and macro-level policies influence the impacts of resettlement on smallholder
agriculture and provides a medium for developing interventions to make resettlement more
effective. The case studies, as well as evidence from Zimbabwe, highlight the need to
institutionalize an appropriate evaluation strategy that will improve planning and

implementation of land redistribution programs, as discussed in Chapter Four.

 

 

de’s’elc

the 10
the :4

plann

4.1 E

agfiel
aw e‘

80the

its CL
i“Clio;
(agric
shOul

1a“(ill

CHAPTER FOUR

EVALUATING LAND RESETTLEMENT POLICY IN ZIMBABWE

The first section of Chapter Four examines the elements involved in evaluating rural
development programs; focusing on performance, impact and efficiency analyses. The
second section discusses the importance of monitoring and evaluation systems in assessing
the long-term performance of resettlement programs, and addresses the need to improve
the evaluation of Zimbabwe’s resettlement program by more clearly defining its role in

planning and implementation and expanding the current methodology.

4.1 Evaluating Zimbabwe’s Resettlement Program

Chapter Three outlined the impact of Zimbabwe’s resettlement program on
agricultural production, income distribution and employment generation, both in terms of
aggregate indicators (i.e., total production, average yield, number of families resettled), and
some disaggregate information derived from case studies and evaluation reports.

Evidence presented in Chapter Three showed that the government should improve
its current evaluation methodology by systematically collecting information on broad
indicators across all schemes relating to resettlement’s impact at the scheme-level
(agricultural production, income distribution and employment). In addition, the government
should allocate resources to examining critical issues, including: 1) the ability of formerly

landless individuals to become productive resettlement farmers; 2) the development of soil

115

 

 

COHSt'

forhf

prod

and:
the l
servi
bran.
CEO
Hort

activ

116

conservation policies; 3) the effect of leasehold tenure on resettlement farmers’ incentives
for investment in accelerating agricultural production; 4) the viability of establishing schemes
based on intensive crop production in regions IV and V; and 5) women’s access to land and
production opportunities.

Although the government has established a monitoring and evaluation unit to collect
and analyze data on resettlement area activities, Cusworth and Walker (1988) found that
the unit is under-financed and lacking in both organization and continuity of personnel
service. Furthermore, the M&E unit (which operates through AGRITEX) is not the only
branch collecting data on resettlement scheme performance; MLGRUD, DERUDE, and the
CSO, among others, also collect information of use to their particular agencies. Wekwete
(forthcoming) argued that there is little inter-ministerial coordination of resettlement
activities, and therefore, there is most likely little coordination of data collection efforts.

Given the government’s increasing emphasis on creating a highly productive
resettlement sector (MLARR, 1989; GOZ, 1991), the government should establish a more
systematic method of evaluating resettlement to ensure that redistribution does lead to
growth in agricultural production. In fact, two main elements for evaluation emerge from
the government’s focus on increasing agricultural productivity. First, one must determine
the relative influence of selected scheme-level variables (i.e., input, service, management and
agro-ecological conditions) on farm performance. Second, one must isolate the settler types
that have the highest potential for contributing to growth in agricultural production. With
more detailed information about specific scheme-level performance variables, policymakers
can redirect future resettlement efforts towards providing requisite conditions and settler
types to encourage growth in the resettlement sector.

4.2 Conceptualizing Evaluation

 

 

dfide
(Unitl
pedol
arei

mdud

maria}

mukh

“fins

Slams

rate Q

117

Evaluation “is a process for determining systematically and objectively the relevance,
efficiency, effectiveness and impact" of a program’s activities in the context of its objectives
(United Nations, 1984). Casley and Kumar describe three specific types of evaluation;
performance, impact and efficiency assessment (1987). Performance assessment ”includes
a review of all the activities undertaken by the project to achieve its stated objectives"
including: 1) planning and appraisal, 2) specification of objectives, activities and targets, 3)
timing of implementation, 4) services and inputs provided, 5) beneficiary coverage, 6)
managerial performance, and 7) financial performance (1987, 101-2).

Impact assessment focuses on both the primary and secondary effects, including
anticipated and unanticipated results of the program. These effects are often measured in
terms of their impact on agricultural practices, or the health, nutrition and socio-economic
status of the target group.

Program efficiency assessment refers to estimating the financial and / or economic
rate of return generated by the government’s investment in the program, at the scheme,
regional or national levels of analysis.

Performance, impact and efficiency assessment are not mutually exclusive, but
evaluators often focus on one aspect more than another, particularly in the face of financial,
personnel or time constraints. Whatever the emphasis, evaluations should incorporate
theoretical and empirical knowledge from the relevant social science in order to derive
alternative programs and to test causal relationships (Freeman et al. 1979).

Evaluations may be conducted: 1) ex-ante for project appraisal; 2) periodically
during project implementation; and / or 3) ex-post, after project completion. The type of
evaluation system adopted must reflect program objectives and available resources, as well

as various information-users’ priorities for data collection and analysis.

118

Table 4.1 presents the three stages of evaluation - specification, design and

implementation - and highlights (‘) those discussed in this study.

      

~ Specification of evaluation
topic
‘Specifysmpjgg

evaluation

      
     
   
   
   
   
     

‘Specify mess of
evaluation

Design of evaluation

 

 

Wed“;

‘Specify type of ‘Select evaluationgmgm Designate W of?
evaluation evaluation

‘Select comparative
tandard

Specifydggmgn ‘Select evaluation
easures
Finalize evaluation topic
*Selectsa
procedures

 

‘Specify dataa nalysis
procedures

Specify datar Lem

procedures

‘Specify_____ dataco 99m
procedures

— Finalize evaluation design: _

   

Implementation of
evaluation
Check feasibm ' of

evaluation

 

 

 

__lfl

Test and refine evaluation
rocep___d___ures
Collect, analyze and
reportres esul

Complete findings and
recommendations

 

_i_

 

-4- ____ .1 A ___

 

a Those steps noted with an asterisk are discussed in this chapter.
b For example, identify who makes decisions regarding the evaluation, any relevant
timelines, constraints, and so forth.

Source: H. Freeman et al- WW (1979: 44).

120
4.2.1 Information-Users and Levels of Analysis

There are three primary clients who provide services or inputs for resettlement; each
with a different role in planning and implementation. Therefore, they require different sets
of information from which to make decisions relevant to their level of involvement. First,
ministry officials require information on the return to government resources invested in the
resettlement program, in order to judge whether these resources could be more efiiciently
allocated to another project. They also need to convey this information to donor agencies
who fund the program. In addition, they may want information on ministerial and scheme-
level administrative activities. Furthermore, some ministries and collaborating agencies
require information on the maintenance and use of specific inputs they provide to
resettlement schemes, such as water, roads, building equipment, veterinary supplies and so
forth.

Second, scheme administrators need information on program inputs and settlers’
farm-level performance, in order to inform government officials of ways in which services
and management can be improved, as well as to evaluate their own performance. Finally,
farmers need information on and access to inputs and services in order to improve their
farm management practices and decisionmaking tools.

Performance, impact and efficiency assessment may be conducted at the scheme,
regional and national levels. Successful evaluation requires that information-users establish
both evaluation priorities (i.e., identify a hierarchy for evaluating program objectives) and
determine the relevant level of aggregation for the analysis. Casley and Kumar argue that
the hierarchy must specify ”target populations, critical activities, and tasks to be monitored

against agreed targets," both in terms of short- and long-term objectives (1987: 23). This

121
hierarchy will determine the level of aggregation which is conditioned by the purpose of the

analysis (i.e., whether the evaluation will assess farm-level, regional or national impacts).
Thus, the specific evaluation methodology chosen is a function of the characteristics
of the program being evaluated, as well as the information needs of decisionmakers and any

constraints they face (i.e., political, financial, time and personnel constraints).

4.2.2 Assessing Program Performance

Program performance assessment measures the degree to which program activities
achieved the stated objectives for planning and implementation. The process of evaluating
program performance induces accountability, allows planners to link implementation
problems to shortfalls in impact and provides a basis for modifying program procedures
(Freeman et al., 1979). Table 4.2 lists some indicators of program performance that could
be used to determine how well a resettlement program meets its objectives, in terms of
general information on resettlement beneficiaries, scheme-level inputs and the performance

of scheme managers.

122

‘ Table 4.2: Indicators of Program Performance for Resettlement Schemes.

 

Performance variable Performance indicator Data source
‘ Characteristics of - No. of households settled - Settler registration
3 beneficiaries - Area of origin records
- Size of livestock holdings
- No. and type of farm
implements
- Size of farm parcel
Characteristics of - Price/ ha - Ministry of Agriculture
resettlement land (by agro- - Location data I
. ecological location and - Total area purchased
scheme type) - Total area developed
Service or input to scheme: - Timely delivery - Service records at
- Water - Maintenance ministry or scheme-
"- - Health center - Adequate staff level
- School - Adequate supply/ - Observation
- Seed, fertilizer, herbicide, distribution - Interviews with scheme
grain bags - Quality participants

 

 

 

“ - Ag. implements

- Credit
- Extension
- Transport

I - Housing

 

Managerial performance
(scheme and / or inter-

‘ ministerial)

Source: Compiled by author.

 

- Budget allocations
- Equipment and input
supplies and condition

(field and office)
- Enforcement of scheme

regulations

 

- Personnel time sheets
and records (written
and oral
communications)

- Interviews with scheme

participants

123

Performance evaluation requires the collection of relevant data in a cost-effective
manner. Some performance variables are directly measurable (i.e., number of households
resettled, total land purchases), while others are more qualitative (i.e., quality of housing or
services, staff performance) and require some subjective interpretation. Furthermore, some
variables must be evaluated more frequently than others, perhaps entailing a greater cost.
For example, it is relatively inexpensive to more frequently collect data on the number of
families settled in a certain region. On the other hand, obtaining information on the timing
of input deliveries entails contacting the ministries or organizations involved, as well as
inquiring about beneficiary satisfaction - a costly process which may require collecting data
several times a year, depending on the service in question. Yet, some knowledge of the
timeliness of input delivery is essential to understanding the conditions influencing
agricultural production and settler welfare at the scheme-level. This implies that in order
to monitor project performance in a cost-effective manner, evaluators must identify the set
of variables (objective indicators) that can be collected at minimal cost from readily
available administrative data, and supplement the data with information gathered through
informal interviews or case studies.

Data to assess performance objectives can be collected through various methods,
including observation, reviewing service records, and interviews with service providers and
settlers (key informants). The appropriate data collection method depends not only on
available resources, but also on the degree of precision desired in the analysis, as well as the
importance accorded to performance assessment. However, since it is necessary to collect
performance data over a wide range of variables, evaluators must devise a systematic and
relatively uniform method of collecting this information (Freeman et al., 1979), and establish

a baseline or target against which to evaluate program performance.

124
4.2.3 Assessing Program Impact

Impact assessment measures the value and magnitude of the program’s effects on
the target population (Freeman et al., 1979). In its most rigorous form, impact assessment
must show that measurable impacts on the target population are a function of that particular
program, in order to establish linkages between program inputs and outputs (Freeman et
al., 1979). However, it is often difficult to establish direct causal linkages because many
intervening factors influence the production of outputs. For example, increases in
agricultural production are a function of settlers’ improved farming conditions (such as more
arable land and better soil quality), as well as timely access to inputs, credit, marketing
opportunities and weather. Thus, program inputs are typically viewed as a bundle of factors
which appear associated with the output or impact. These factors are generally measured
in terms of indicators or "objectively verifiable measures of changes or results brought about
by an activity...used as markers of progress towards reaching intermediate or long-term
objectives” (United Nations, 1984: 37), either directly or indirectly. Direct indicators
measure the progress of an activity against a stated target, whereas indirect indicators
provide a measure of impact when direct indicators are too expensive or difficult to develop
(Freeman et al., 1979; United Nations, 1984).

The implications for evaluating resettlement are that 1) several indicators must be
used to measure changes in anticipated outputs, and 2) these indicators must be evaluated
at intervals frequent enough to provide relevant information. Table 4.3 presents some
indicators that could be used in measuring the impacts of resettlement on agricultural

production, income distribution and employment generation for the target population.

125

‘ Table 4.3: Indicators of Program Impact for Resettlement Schemes.

 

. Impact variable Outcome indicator Data source(s)

 

1 Agricultural production - area cultivated -Sample surveys

- crop composition (cash + -Marketing board
staples) and total prodn. records

- yield variability from year-
to-year

- proportion marketed vs.
consumed at home

 

‘ Income distribution 7 - total household income -Sample surveys
(farm income + non-farm -Participant interviews
income)
- livestock holdings
- housing

 

Employment generation - full-time farming jobs -Sample surveys
created -Participant interviews

- part-time farming jobs
created

- non-farm jobs created
(rural works, services)

Source: Compiled by author.

 

 

 

126

Program impact can be measured at the aggegate level (national or regional) to
gauge whether the progam has had a negative or positive aggegate effect on the economy,
or at a disaggregated level (individual scheme or farm). To estimate aggegate impact, the
analyst must first develop assumptions regarding program impact at the scheme or farm
level by conducting a more diagnostic analysis to identify micro-level impacts.

National statistics may provide information on aggregate-level impacts, and are
usually based on projections derived from a small sample of individuals. However,
aggregation disguises regional and local-level fluctuations which signal possible or actual
problems for a subset of program beneficiaries. Thus, if there are potentially undesired
outcomes resulting from progam implementation, then evaluators must examine micro-level
impacts more closely.

Freeman et al. (1979) divide impact assessment methodologies into two categories;
experimental and judgnental. Experimental methods involve rigorous research designs
intended to isolate the net impact of a program on its beneficiaries. These methods include
using control groups in either randomized or nonrandomized experiments and cross-
sectional studies with statistical controls. In general, these methods are costly, resource-
intensive and must be developed prior to progam implementation to ensure appropriate
experimental design.

Judgnental methods are less rigorous than experimental approaches but more
effective when resource constraints exist. These methods include formal and informal
surveys of the target goup, case studies, participant assessment of the outcomes, as well as
external evaluator and project adnninistrator assessments. Judgnental approaches can also
be used diagnostically to 1) identify problems, 2) determine contributing factors, 3) review

constraints on decisionmaking, 4) develop and appraise alternative solutions, 5) recommend

127
the optimal course of action and 6) monitor feedback (Casley and Kumar, 1987).

Nonetheless, judgnental approaches should be validated by more precise assessment

methods whenever possible (Freeman et al., 1979).

4.2.4 Assessing Program Efficiency

Ex-post financial and economic analyses provide decisionmakers with information
on the financial or economic rates of return to a program; information which could then be
compared to proposed alternatives. Both types of analyses extend the scope of impact
assessment since one must define the extent of all quantifiable impacts prior to conducting
rate of return analysis.

Financial rate of return analysis (private accounting method) assesses the progam’s
financial viability for various participants involved (Gittinger, 1982: 16). This includes
analyzing the financial profitability of resettlement for resettled farmers or calculating the
program’s revenue-generating capacity for the national funds invested.

Economic rate of return analysis (public and social accounting methods) helps the
policymaker determine if a project’s contribution to economic development justifies using
the scarce resources required for its implementation (Gittinger, 1982). Public accounting
methods allow one to select the alternative that contributes the most to national income (in
an ex post comparison), but do not determine how progam participants benefit differently.
On the other hand, social accounting methods for rate of return analysis incorporate the use
of distributional weights through which the analyst can assign varying degrees of importance

to income received by certain socio-economic groups, or the use of different numeraires.”

 

37 A numeraire is a "common measurement used as the unit of account that
measures the objective being maximized" by the analysis (Gittinger, 1982: 488). For
example, one could disaggregate the income stream generated by the project into streams

128

Both financial and econonnic rate of return analyses involve benefit maximization or
cost minimization with respect to the progam objective, which is expressed in terms of a
common denominator or numeraire. These analyses rely on the simplifying assumption that
all individuals or firms maximize the same objective through their actions. While this is
rarely true, this modification makes it possible to develop a yardstick measure from which
the analyst can then interpret the results for different goups of individuals. To appraise
the profitability of resettlement farming, one usually assumes that farmers’ common
objective is to maximize total income earned on the scheme, where benefits and costs are
expressed in units of real income in domestic currency. To conduct an economic analysis
of a resettlement program, where one considers the costs and returns of the investment to
society as a whole, the numeraire may be expressed in units of national income”. In
economic analysis, the values of the benefits and costs are adjusted for price distortions
(caused by government intervention or other institutional factors), and expressed in shadow
or efficiency prices to correct for these biases.

Nelson (1973) argues that economic objectives are often noncomplementary. For
example, economic objectives might "dictate that mechanization and high-technology
agiculture be applied at the expense of beneficiary participation, income distribution, and
employment generation objectives... [or] the need to attract entrepreneurial ability and
private capital may dictate subdivision into blocks larger than...generally established, thus

sacrificing some of the income distribution objective" (1973: 59). In the examples given

 

received by different individuals at different income levels, and weight these streams to
reflect a social premium on the components (Duloy, 1974).

38 National income may be defined as ”the total net earnings of labor and property
employed in the production of goods and services in a nation during a period, usually a
year.” This can be measured in terms of Gross Domestic Product (GDP) or Gross National
Product (GNP) (Gittinger, 1982: 486). However, other numeraires may be constructed.

129

above, where economic efficiency is not the sole performance criterion, theoretically the
other criteria should be weighted according to their importance as determined by national
policy (Nelson, 1973). However, in practice any weighting method is open to subjective
interpretation by the evaluator, implying that the outcome might not accurately represent
either the implicitly or explicitly stated national policy.

Many critics voice reservations about the use of efficiency analysis in program
evaluation. Carruthers and Clayton (1977) state that public policy objectives are often
difficult to determine precisely, particularly due to the politically sensitive nature of some
progam objectives. Hardaker et al. suggest that the partial equilibrium approach to
efficiency analysis neglects general equilibrium feedback loops through which technical
change may increase incomes, and hence stimulate demand (1984). This is particularly true
for resettlement areas where the establishment of large population centers affects the
regional and national economy in a relatively short period of time.

Furthermore, the methodology itself can be costly and time-consuming (particularly
those methods involving econometric estimation), and present a false impression of
precision regarding the outcome of the analysis. In addition, processes such as shadow-
pricing, which often require extensive time and resources, are less related to project success
or failure than other aspects of planning and evaluation (Carruthers and Clayton, 1977).
Finally, in order to establish reliable relationships between program inputs and outputs,
evaluators must incorporate uncertainties due to weather, pests, input supplies and markets
(Carruthers and Clayton, 1977). Just as these inherent uncertainties pose constraints for
small farmers and condition their economic behavior in agriculture, they also introduce a

certain degee of unreliability regarding long-term projections in an analysis.

130
4.3 Developing Evaluation Techniques for Resettlement

In order to appropriately assess the effectiveness of land resettlement in Zimbabwe,
analysts must employ an evaluation methodology that operates well in a dynamic context.
Dynamic evaluation of resettlement is essential for several reasons. First, the level of
organizational and managerial coordination required for resettlement programs changes over
the life of the progam. While intensive management is necessary during the initial years
of scheme development, tine government can witlndraw from direct involvement as scheme
members develop their own organizational capacity and become more integated into
regional administration. Therefore, the objectives by which one assesses management
performance also change.

Second, progam impacts are not always apparent during the first few years. In
many cases, although land redistribution programs result in short-term decreases in
agricultural production as settlers adapt to a new farming environment, production increases
after several years (for example in Bolivia). Conversely, a particular farming system may
become less sustainable after several years of intensive farming, such as the upland
transmigation sites in Indonesia. Furthermore, the distribution of resources may change
over the course of the progam as certain farmers accumulate more resources, or as other
productive activities arise (i.e., non-farm employment opportunities), creating a geater
concentration of incomes (as observed in each case study in Chapter Two).

Third, the rate of return to government’s investment in resettlement changes
depending on when the calculation is made during the progam’s net benefit stream. For
example, since expenditures are most likely to decrease over the life of the scheme, the
returns to the government’s investment increase during the later years of the program,

generating a geater rate of return. Therefore, rate of return analysis conducted in the later

131

years of a resettlement progam is more favorable than if conducted earlier when project
costs are higher and the net benefit stream is lower.

However, the dynamic aspects of performance, impact and efficiency assessment are
not incorporated into traditional mid-term or ex post analyses. Therefore, even an
extensive, one-shot evaluation will contribute little to understanding scheme-level
development and performance. On the other hand, a more systematic and timely evaluation
method could be developed in conjunction with the M&E Unit’s current activities by more
precisely defining the evaluation task, streamlining the data collection effort and producing

more timely and widely diffused syntheses of resettlement findings, as discussed below.

4.3.1 Monitoring and Evaluation

Monitoring and evaluation involve two different types of program assessment.
Monitoring refers to "the continuous or periodic review and surveillance by management...to
ensure that input deliveries, work schedules, targeted outputs and other required actions are
proceeding according to plan..., [and whose purpose] is to achieve efficient and efi'ective
project performance by providing feedback to project management at all levels” (United
Nations, 1984: 13-14). Evaluation methods incorporate information gathered during the
monitoring process into ongoing, mid-term and / or ex post analyses.

An essential element of monitoring is that it ”should focus on the interaction
between the project activities and reactions of the target population if it is to meet the
needs of management” (Casley and Kumar, 1987: 7). To accomplish this, monitoring should
be based on gathering information which reflects the link between progam implementation
and its impact on the target population, and should be well integrated into project

management structures, at all levels (Casley and Kumar, 1987).

132

A study of an irrigation project monitoring system (Ng and Letham, 1983) indicates
that most monitoring systems are inadequate for decisionmaking because the information
provided to decisionmakers: 1) is not current enough to assist in timely decisionmaking; 2)
does not assist in making operational or strategic decisions; 3) is too complex and difficult
to use; 4) does not reflect interactions between the project and its beneficiaries; and 5) is
seldom diagnostically compared with established targets. Ng and Letham stress that a
deficient system ”cannot provide early warning of impending problems so that they can be
dealt with before becoming acute" (1983:31).

The government of Zimbabwe has a strong political commitment to continuing its
investment in resettlement, both to meet its pre-independence comnnitrnent to redistribute
resources to smallholders and to assure commercial farmers and external donors that
redistribution can occur without undermining the economic health of the country.
Therefore, the government should place the greatest priority for evaluation on performance
and impact assessment, in order to determine how shortfalls in planning and implementation
can be rectified to make resettlement more effective. In Zimbabwe, strengthening the
current monitoring and evaluation system could increase the management and diagnostic
capability of both policymakers and administrators, and thus improve the success of

resettlement in increasing smallholder agicultural production and redistributing incomes.

4.3.2 Strengthening Monitoring and Evaluation in Zimbabwe

In order to institutionalize an ongoing evaluation system for resettlement, one must
first study how the current system of data collection, analysis and diffusion of results
provides information on program performance and progam impact on beneficiaries, and

identifies existing and potential problems. The monitoring and evaluation unit should be

133

responsible for coordinating all evaluation activities, in order to decrease repetition in data
collection and synthesize the findings into a format that provides useful information to
identified users. The following sections will examine the process of organizing and
implementing an ongoing evaluation of resettlement in Zimbabwe, including: 1) structuring
the evaluation method; 2) systematically collecting indicator data at both the scheme and
farm levels; 3) initiating issue-oriented studies to better understand local constraints and

potential problems; and 4) decentralizing data collection activities.

43.2.1 Organizing the Evaluation Method

The monitoring and evaluation unit, as the central agency responsible for evaluating
resettlement, must first identify the different information-users and their needs. These users
might include ministry officials, resettlement officers and other scheme-level personnel,
whose needs will determine the type and the frequency of data to be collected. Typically,
these data will support the analysis of broad indicators, as well as identified problems in
resettlement’s implementation. Once the unit identifies the types of data required by the
information-users, M&E personnel can then investigate existing data sources, including the
type of data currently collected, the quality and frequency of collection, and the form in
which it is collected. Gaps in existing information can then be filled by collecting additional
data.

Second, the M&E unit must aggregate information-users’ needs to ensure cost-
effective and complete data collection, specify the procedures for data collection, and
designate the responsibility for implementation to the various ministries and organizations
involved. Once the data are collected, the M&E unit is responsible for overseeing the

analysis and reporting of all results, both to the central and local levels.

134
4.3.2.2 Data Collection for Program Indicators

Impact and performance assessment of the resettlement progam entails collecting
information on homogeneous indicators across all schemes, in order to measure scheme
progess against baseline targets, compare outcomes on different schemes, and assess the
aggegate impact of resettlement on the economy. To collect information on performance
and impact indicators with limited resources, it will be necessary to employ relatively simple
assessment techniques.

Performance is assessed by measuring the progam’s objectives against its
achievements, both to gather one-time and periodic data. One-time data collection would
involve compiling information on the characteristics of land purchases and development and
on progam beneficiaries. Periodic data collection includes gathering information on the
types of services and inputs required for scheme development and the outputs of scheme
administrators (see Table 4.2). Much of this information can be obtained from settler
registration and scheme administration records, supplemented by interviews with
administrators and farmers. For example, records might indicate that the scheme depot
received seed and fertilizer for the planting season, but interviews could reveal that the
supplies arrived after the rains began, or in insufficient quantity to meet demand.

Generally, the assessment of resettlement progam impact focuses on evaluating
changes in agicultural production, income distribution and employment on progam
beneficiaries (see Table 4.3). For example, agricultural production can be measured using
marketing records or farmer estimates, and if area cultivated is known for each marketed
crop (either through scheme records or farmer knowledge), then yields can be calculated
from this information. However, if marketing records are used, it will be necessary to make

some adjustment for home consumption in calculating total production.

135

Income is much more difficult to measure, inn part because farmers prefer not to
disclose the amount or sources of their income for legal reasons (resettlement regulations
prohibit households from earning any income off-farm). Furthermore, since income data
are collected only periodically, farmers often cannot correctly recall the time and amount
of their earnings. Yet, some approximations can be made through cattle and crop
marketings, although these data will typically underestimate the amount of household
earnings and give no indication of the household’s cash flow after the recorded sales.

Indicators of employment generation include the number of adults engaged in full-
time farming, those with permanent service jobs and those engaged in temporary activities
such as rural works. This information could be obtained through scheme records of
resettled households, and the number of individuals employed by the government for
permanent and temporary positions. Employment data are primarily useful for examining
the aggegate impact of resettlement on labor absorption in agriculture, and have less
relevance for scheme-level analysis.

The data gathered by monitoring resettlement’s performance and impact at the
scheme-level can then be compared directly to initial targets established for production,
annual household income, per capita service provision (i.e., the number of people served by
health clinics, schools, extension workers, and so forth), credit use and the development of
housing and infrastructure. However, since targets frequently underestimate settlers’ ability
to achieve target incomes or provide enough labor to cultivate the recommended parcel size,
monitoring and evaluation data can be used to revise targets and provide policymakers with
information to improve future planning. Furthermore, these data also indicate where
differences in various schemes’ performance lie, signalling the need to study certain issues

more intensively.

136
4.3.2.3 Incorporating Issue-Oriented Studies into the Evaluation Methodology

Although broad impact and performance indicators provide policymakers with an
overview of resettlement’s progess, there are some issues not captured in these indicators.
These issues include existing problems identified from ongoing monitoring activities, and
potential problems identified from similar experiences in other countries or in other areas
of Zimbabwe (issues for future research). Thus, in order to address pertinent issues as they
arise, it is also necessary to incorporate the use of selected issue-oriented studies into the
evaluation of resettlement.

Issue-oriented studies represent a more flexible evaluation approach than monitoring
broad indicators across all schemes. In issue-oriented studies, the analyst can adjust the
data requirements, information sources and the number of schemes to fit the specific issue
of interest. For example, to assess the overall impact of resettlement on settler welfare, one
needs to identify groups among the settler population whose needs or response to
resettlement might differ from others. Several vulnerable, or at-risk goups can be
identified, including formerly landless and resource-poor settlers, women, and settlers in
lower potential areas. Therefore, one might target several schemes containing a high
percentage of individuals in these goups for each region and identify a number of cases for
study among the total population.

Indicators of settler welfare should be developed to measure the benefits that settlers
gain from living in resettlement areas. The critical values for these indicators would be
established according to the study region. These indicators might include: measuring health
and nutrition (i.e., vital statistics, anthropometric measurements, or food consumption),
housing (by type of construction materials), livestock inventories (by number and herd

composition), education (participation in formal and informal education, for children and

 

 

(1'.

(er

role
area
Cliffe
crop:
studi.
based

SUIVei

inVojve

DICSe l

137
adults), water source (type, variability of supply and distance from household), transport

(i.e., cost and availability for both passengers and commerce), and political representation
(either through the Zimbabwe National Farmers’ Union or local government).

Data sources for these studies will vary, depending on the issue in question. For
example, in exploring the impact of scheme regulations on women’s legal rights and on their
role as food producers, one would need to examine women’s legal riglnts in the communal
areas in comparison with the resettlement areas. Also, in order to capture the effect of
different cropping systems on demands for female labor (since the types of cash and food
crops gown vary in the higher and lower rainfall areas), it is necessary to select several
studies in each natural region (based on purposive sampling). The studies could then be
based on interviews with women in several villages on each selected scheme, and include
surveys of crop production according to gender and labor requirements.

Another issue warranting research is the environmental impact resulting from
resettlement. For example, to isolate some of the factors leading to land degradation and
pinpoint potential solutions, one could first identify areas facing a higher risk of soil erosion
(i.e., areas with high population densities, significant slopes, erodl'ble soils or rock outcrops).
After clustering these areas, one could stratify the clusters by soil type, agicultural practices
and livestock inventories, and select a sample for study from the clusters. This would allow
researchers to examine the links between soil erosion, household labor availability, and
access to agicultural inputs and extension information.

Finally, there are options for conducting these intensive studies to reduce the
involvement of AGRITEX personnel, who may already suffer from heavy work demands.
These might include contracting with University of Zimbabwe personnel or with local NGOs

138

who rrnight have area expertise and more available resources with which to undertake these

studies.

43.2.4 Decentralizing Data Collection

Decentralizing data collection activities decreases overall admirnistrative costs by
combinning data collection with other activities conducted by scheme personnel, and allows
participants to gain management and problem-solving skills, particularly farmers. For
example, scheme administrators and extension workers can be involved in collecting impact
and performance indicator data, and in providing support to researchers investigating
specific issues identified through the ongoing monitoring activities. They could record
information on the availability, quality and timeliness of service provision to the scheme, and
gather information from farmers regarding their perceptions of resettlement. In addition,
with training they could participate in scheme-level monitoring by signalling input shortages
or other problems to the appropriate agency. Furthermore, literate farmers can record yield
or production estimates, and information on the end uses of their food crop production.

However, the principal issue involved in promoting this component of evaluation is
finding incentives to motivate individuals to participate and provide reliable information.
This is particularly crucial for resettlement officers and other scheme personnel who may
perceive no advantage in extending their work into additional activities if they receive no
benefits (monetary benefits or nonmonetary benefits such as recognition, land or special

privileges).

139

4.4 Implications for the Future Planning and Implementation of Resettlement Policy in
Zimbabwe

Strengthening the monitoring and evaluation of resettlement schemes in Zimbabwe
will improve future resettlement planning and implementation by providing policymakers
with more timely and relevant information on program and scheme-level performance. For
example, coordinated and timely data collection efforts will provide policymakers and
scheme-level administrators with early warning information, enabling them to resolve
problems as they arise. Furthermore, in developing future resettlement sites, planners can
generate more realistic projections of actual smallholder returns to farming in the
resettlement sector by basing farm budgeting on communal area and not commercial farm
data. Finally, by extending the monitoring and evaluation system to examine the
cooperative, core estate and goup ranching schemes (Model B, C and D schemes), the
government could determine which factors have impeded their development into viable
models for farmers in Zimbabwe, and thus where appropriate, redirect the resources

necessary to make these models more successful.

45 Conclusions

Due to organizational and financial constraints, current activities for evaluating
Zimbabwe’s resettlement progam fail to provide policymakers and administrators with
timely information on resettlement’s performance; and its impact on the national economy
and on program beneficiaries. The M&E unit does not systematically compare recognized
baseline indicators with farm-level performance in different natural regions and among
different types of farmers. Finally, they have made no systematic effort to assess the effects

of institutional and policy factors (i.e., tenure policy, resettlement regulations) on farm-level

140

incentives for increasing agricultural production and on the distribution of resources -
factors previously shown to influence the outcome of land redistribution policies.

To fill this gap in planning and implementation and to better address the needs of
the more vulnerable goups among the target population, the government should reorganize
its current evaluation method to make the M&E unit responsible for the collection, analysis
and synthesis of all data for resettlement. The unit should develop a set of broad indicators
to assess the impact and performance of resettlement policy in Zimbabwe, and initiate issue-
oriented studies to explore region-specific problems. By expanding the evaluation
methodology to include these studies, policymakers could begin to examine the impacts of
existing or potential problems raised through program-wide data collection and analysis. An
ongoing, more diagnostic approach to assessing resettlement could ensure that the
redistribution of resources through resettlement also results in increasing agicultural

production and economic growth.

 

 

SJ

land
gene
redh
subd
prod
agrk
0th
agric

incor

in 3g
ITORA
than
dfiTer
land

inch“

CHAPTERFIVE

SUMMARY AND CONCLUSIONS

5.1 Summary

This paper examined theoretical and empirical evidence regarding the impacts of
land redistribution policies on agicultural production, income distribution and employment
generation in the smallholder reform sector. Evidence from an analysis of land
redistribution programs in Peru, Bolivia, Indonesia and Kenya supports the proposition that
subdividing large farms into smallholdings has often led to a net increase in agricultural
production, as well as wider income distribution and geater employment opportunities in
agiculture. However, factors outside of the reform sector will strongly influence the
outcome of land redistribution policies, including institutional, macroeconomic and
agricultural policies. The case studies suggest several lessons that policymakers should
incorporate into planning land redistribution policies.

First, subdividing large farms into smallholdings has not led to a long-term decline
in aggregate agicultural production, as many critics of distributivist land reforms argue.
However, it is difficult to assess whether these new smallholdin gs operate more efficiently
than the former large farms, primarily due to insufficient data with which to analyze
different efficiency criteria. In Peru, although national production did not decline following
land redistribution, per capita production subsequently decreased due to many factors,

including low agricultural investment, price controls and import subsidies. Although

141

142
agicultural production initially decreased following Bolivia’s land reform, thereafter food

production increased significantly in some regions as smallholders benefitted from greater
access to production inputs. Indonesia’s transmigration program is credited with increasing
national rice production by 5% during the 1980s, although the farm-level impacts varied
according to the location and time of settlement. Finally, in Kenya land resettlement also
led to increases in national agricultural production, with variations noted among different
settlement types.

Second, redistributing land to smallholders has improved income distribution, where
the newly created smallholdings have provided the settlers with adequate income-generating
possibilities. The potential for post-reform improvements in income distribution depend on
governmental policies, as well as the size of the holding, which must take into account the
typical farm family’s management abilities and available technology, agoecological
conditions and the potential of the holding to produce an acceptable income level. For
example, in both Peru and Bolivia, the governments’ priority to quell peasant uprisings
skewed the reform benefits primarily toward those regions where unrest was greatest. As
a result, there is evidence in Peru that income concentration increased among some higher
income rural families. Since the early 1900s, Indonesia’s transmigation progam has served
to distribute incomes and other benefits (i.e., improved housing and health care) to almost
one million families of landless laborers and Javanese farmers with very small subsistence
holdings - approximately 4% of Java’s population. Kenya’s resettlement program
distributed land to 71,000 smallholders (4% of all smalllnolders). However, the program was
less successful in generating adequate incomes for settlers in the long run because high

indebtedness among settlers forced some smallholders to default on their loans and

 

 

 

 

CVCl'liUZ

experie

surplus
increasi
creation
for the
(smalli
P0pulat
titles to
been on
transmi
Kenya,
For em
08th l
the Ere:

adults/1

143
eventually lose their land (particularly on the high-density schemes designed for less-

experienced farmers).

Third, the case studies illustrated that land redistribution to smallholders absorbs
surplus labor in agriculture, increases the amount of labor used per land unit, and thus
increases employment in smallholder agriculture. However, the potential impact on job
creation also depends on the long- and short-term profitability of the new farming system
for the settlers. For example, in Peru the reform created a large number of minifundistas
(small farmers with marginal, subsistence holdings). In Bolivia, although much of the rural
population benefitted from employment in agiculture, many subsequently lost their land
titles to returning landowners who asserted their prior claims. The Indonesian reform has
been more successful in creating both on-farm, off-farm and temporary employment for
transmigants, totalling approximately 500,000-600,000 jobs on rainfed schemes alone. In
Kenya, resettlement was very successful in creating smallholder employment in agiculture.
For example, although rainfed schemes absorbed more workers than large commercial
estates (568 adults/1,000 hectares versus 319 adults/1,000 hectares), irrigated schemes had
the geatest impact on job creation - employing nine times as many workers (2,807
adults/1,000 hectares) as commercial estates.

In each case study examined, smallholders received broader rights to land as pre-
reform land tenure policies were changed to stimulate investment and increases agricultural
production in the reform sector, and to redistribute income more widely in the economy.
Land tenure policies comprise sets of property rights which influence individuals’ access to
land and other resources, and ultimately, the distribution of income throughout the

economy.

 

 

ar
is

en

0f .
of

dis

dEI

Sele

144
Access to land may be defined in terms of exchange rights (the individual may

transfer or withhold his/ her right to the land) and use rights (the individual may utilize the
land with no right to transfer it to others). Neoclassical economists argue that tenure
systems which allocate only use rights to individuals (communal landholding systems in
particular) are inefficient because they discourage investment and demand for credit,
increase transactions costs in land disputes, and impede the development of a land market
which would encourage more efficient land use. However, there is much debate, especially
among scholars of African land tenure systems, that a significant degree of individualization
is inherent to communal landholding systems, providing the tenure security necessary to
encourage investment and use of credit (Cohen, 1980; Hunter and Mabbs-Zeno, 1986;
Barrows and Roth, 1989). These critics suggest that other institutional constraints (such as
pricing, credit and marketing policies) often prevent investment and increases in agicultural
production, arguing that resources should first be applied toward adapting existing
institutions to provide proper incentives rather than transforming tenure institutions.

The case studies illustrated that while ganting smallholders individualized rights to
land may provide some benefits and incentives to increase agicultural production, the
impacts are often tenuous and short-term. This results from two factors, including the effect
of other institutional variables may which skew the impact of the land tenure law, and a lack
of additional legislation to protect the rights of the new landholder and reinforce the
distributive impact of the reform.

The analysis presented in Chapter Two concludes that the most significant factors
determining the impact of land redistribution progams were: 1) site selection; 2) settler
selection; 3) government provision of institutional and administrative support to new

landholders; 4) land acquisition and financing; 5) land tenure policy adopted for the reform

145

sector; 6) agicultural and macroeconomic policies affecting the reform sector; and 7)
scheme-level monitoring and evaluation.

Thus, it is important for policymakers to recognize that a broad set of variables
affects the outcome of land redistribution programs on the target population. For example,
in Bolivia adverse pricing policies stifled investment in agiculture following land
redistribution. Although smallholders gained freehold rights and larger parcels, the subsidies
placed on urban food prices reduced rural incomes and the smallholders’ ability to increase
their incomes from expanding agicultural production. In Indonesia, incomplete
specification of land tenure laws in the outer islands resulted in conflicts over land use. In
Kenya, unrealistically high loan repayment schedules limited or negated the profits earned
by settlement farmers, and hence reduced their welfare.

The diversity of institutional variables influencing the outcome of land redistribution
progams makes it difficult for policymakers to anticipate all possible impacts during the
planning stage. Therefore, these case studies underline the importance of establishing
interim monitoring and evaluation activities as a component of land redistribution progams.
By revealing those areas where the progam is less successful (i.e., site selection or planning,
settler selection, the provision of administrative and institutional support or tenure system
development), monitoring and evaluation provides policymakers with information needed
to increase resource efficiency by redirecting resources and progam components to
maximize their impact.

These lessons from the four area case studies regarding the potential impacts of land
redistribution progams have direct relevance for land resettlement policy in Zimbabwe.
The government of Zimbabwe implemented the Rural Resettlement Program in 1980 to

redistribute large tracts of underutilized commercial farmland to smallholders who had

 

 

Shun
on 2
form

total

gove
prog
proc

aPPI

imp.

that

IEgjc
equa
then
Prod
large

holdj

146
gadually been relocated into low potential agicultural regions during the colonial rule.

Since 1980, the government has settled approximately 52,000 families (416,000 individuals)
on 2.6 million hectares of land, primarily in natural regions III and IV. To date, 18% of
former commercial farmland has been redistributed to smallholders, although purchases
total 25.6% of all commercial lands (3.8 million hectares).

Although the current pace of resettlement does not meet the targets set by the
government in 1980 (resettling 255,000 families by 1989), the accomplishments of the
progam are still significant. First, resettlement’s contribution to national food crop
production increases yearly, particularly in maize and goundnut production. Second,
approximately 6% of the rural population has benefitted from resettlement, compared to
4% benefitted during the Kenyan resettlement program. Furthermore, there is evidence
improved price incentives and market access have stimulated resettlement area cotton and
maize production.

However, a closer examination of scheme-level crop production in different natural
regions reveals significant differences in productivity, suggesting that not all farmers benefit
equally from resettlement. Farm families in the higher potential areas are able to satisfy
their household food requirements using only 40% of their arable land for food crop
production and the remainder for cash crops, perhaps indicating that holding sizes are too
large for available family labor. However, farmers in regions IV and V use 75-96% of their
holdings primarily for food crap production and may not satisfy their household food needs
in some years.

Resettlement has served to increase employment opportunities in smallholder
agiculture, during a period of declining wage employment in the commercial sector.

Decreases in commercial sector employment are due primarily to post-independence

147

minimum wage regulations and a shortage of foreign exchange, rather than the subdivision
of commercial farms. This is evident in the decreasing ratio of permanent to seasonal labor
used in commercial agiculture (from 4.0 in 1979 to 1.2 in 1989) which has occurred since
minimum wage government regulations took effect.

Finally, income distribution on resettlement schemes appears to be skewed toward
settlers in better natural regions and those with geater access to production inputs, services
and infrastructure. In a survey of selected schemes, farmers in natural region V earned the
lowest gross margins for crop production, while farmers in region II earned the highest total
goss margins for crop production and cattle offtake. Contrary to government objectives,
only 4% of the total pOpulation resettled were originally landless or near-landless families;
the majority were communal area farmers who left small parcels for larger, better quality
holding in the resettlement areas. Furthermore, data indicate that although the
resettlement program was supposed to address the needs of the resource-poor (the landless,
farmers in marginal areas and women), their condition may not have improved.

Overall, while Zimbabwe’s resettlement policy has been successful in creating an
increasingly productive smallholder sector and generating employment in agriculture, its
impact on income distribution has varied significantly across resettlement schemes. Success
has been contingent on effective planning and implementation, but decisionmakers (from
ministry officials to resettlement officers) have received little timely information on the
effectiveness of site and settler selection, the provision of inputs and services, and the effects
of other agricultural and macroeconomic policies in providing incentives for smallholder
production.

Policymakers do not receive timely and appropriate information because the

government has placed little priority on evaluating the national and scheme-level impacts

148

of resettlement, even though this information could improve resource allocation. Although
the government established a monitoring and evaluation unit responsible for collecting and
analyzing data on resettlement schemes, to date the unit has received insufficient financial
and human resource support to conduct a more comprehensive evaluation of resettlement.
Furthermore, its evaluation activities are poorly coordinated with those of other agencies
who provide inputs and services for schemes and collect data pertaining to their own
activities. Strengthening the current monitoring and evaluation effort for Zimbabwe’s
resettlement progam would contribute to policymakers’ knowledge of how resettlement
planning and implementation influence program performance, and its impact on smallholder
agriculture production, income distribution and employment.

Generally, in order to provide useful and timely information, an evaluation must
involve the collection and analysis of data identified as important to the various users of the
evaluation output. The type of evaluation chosen to assess a particular progam is then a
function of the characteristics of the program being evaluated, as well as the information
needs of decisionmakers and any constraints they face (i.e., political, financial, time and
personnel constraints).

Three types of general evaluation techniques are described in the literature:
performance, impact and efficiency assessment. Performance assessment measures the
degee to which progam activities achieved the stated planning and implementation
objectives. Impact assessment focuses on both the primary and secondary effects of the
progam on its beneficiaries, including anticipated and unanticipated results. Program
efficiency assessment refers to estimating the financial and / or economic rate of return
generated by the government’s investment in the progam, at the scheme, regional or

national levels of analysis.

149
Evaluation of resettlement programs typically requires a long time-frame for analysis,

particularly since the level of government resource commitment and the program impacts
on its beneficiaries change over the life of the program. Deficiencies in planning and
implementation, as well as variable impacts on participants on different resettlement
schemes in Zimbabwe, signal the need for reorganizing and improving the monitoring and
evaluation unit’s current evaluation methods. This could be accomplished by increasing the
flow of information to policymakers and scheme administrators by extending the current
monitoring and evaluation activities to include: 1) centralizing the responsibility for
monitoring and evaluation such that the M&E unit could reassess how the existing
evaluation data meet the needs of information-users, in addition to coordinating the
collection, analysis and diffusion of necessary information; 2) systematically collecting
indicator data at both the scheme and farm levels; 3) initiating issue-oriented studies to
better understand local constraints and potential problems; and 4) decentralizing data
collection activities to decrease total government costs for evaluation and increase

participation in scheme-level assessment.

5.2 Policy Implications

Resettlement progams do not yield predictable outcomes for progam beneficiaries.
Due to their size and extended time horizon, resettlement programs are often geatly
influenced by other policies not specifically directed at the reform sector (i.e., foreign
exchange controls, price policies or labor laws). Therefore, in order to direct the progam
toward desired outcomes in terms of agricultural production, income distribution or
employment generation, policymakers must carefully monitor both progam impact and

performance, adjusting its implementation where shortfalls are observed. Although these

150

observations on the importance of planning and implementation for resettlement progams
have already been documented (Lewis, 1964; Christodoulou, 1965; Lewis, 1973; Oberai,
1988), it seems as though the lessons learned from other progams are overlooked, or not
widely known. In particular, policymakers tend to place little priority on establishing
effective monitoring and evaluation systems to improve planning and implementation for
resettlement programs.

One can identify several critical planning and implementation issues for Zimbabwe’s
resettlement policy which could be addressed by improving monitoring and evaluation and

more intensively investigating potential and identified problems in the resettlement areas.

These include:

 

The Kenyan land resettlement progam illustrated the importance of correctly
estimating settlers’ ability to generate sufficient income to repay loans obtained for land and
agicultural inputs. Repayment rates for loans which are based on poor technical data or
unrealistic expectations in the face of uncertain climatic conditions lead to low repayment
rates and settler indebtedness, thus undermining the redistributive equity gained through the
reform. .

Currently in Zimbabwe, many settlers have accrued annual debts more than one and
one-half times greater than their annual incomes. The implications for planning and
evaluation of resettlement in Zimbabwe are two-fold. First, in establishing repayment
schedules for loans (for land or agricultural inputs), planners should develop farm budgets
based on projected smallholders’ yields and input costs. Data for both farm budget
projections and loan repayment schedules are currently derived from pre-reform commercial

farming data which may overestimate smalllnolders’ profits. Therefore, future planning of

151
repayment schedules should be based on actual smallholder costs and returns, allowing for

extended schedules where uncertain climate conditions may affect settlers’ ability to repay
their debts and still have an Operating surplus for future investment. Second, credit records
should be carefully monitored to verify that settlers can indeed meet the planned repayment
schedules. If repayment rates are low for certain types of settlers, or for settlers in specific
areas, then the problem can be addressed by either revising the target rates or examining

the viability of the farming system.

29W

Land tenure policy defines who receives the benefits and who pays the costs of using
land. Therefore, tenure laws can be formulated to restrict or encourage access to land. In
each case study cited above, land redistribution improved smallholders’ access to land
through exchange rights (freehold title, sometimes with restrictions), which could be
transferred or withheld from other individuals. However, in Peru, Bolivia and Kenya many
smallholders who received individual titles to land subsequently lost their rights, either
because they defaulted on their loans or because returning large landowners pressured them
to relinquish their land. In Zimbabwe, settlers’ rights to land are embodied in permits of
unspecified duration. Critics currently debate the type of tenure system that should be
instituted in the resettlement areas; either communal land rights or some degee of freehold
ownership.

Ultimately, the choice of tenure policy depends on three factors. First, government
policy objectives determine whether use riglnts will be ganted to maintain a wider income
distribution among the population or whether farmers will receive exchange rights as an
incentive to expand and invest in their farms, resulting in some degee of income

concentration. Second, the length of farmers’ potential investment horizon determines the

 

 

 

tenure sec
made. Th
extension

In

productior

l

U638. whi
dlSlIlbUllV
freehold Ill
10 CXpandI
sector. Fal
thoroughlil
use in lhll
resettlemgl
irrigallon
deVelopm,
of the lOn
prices and
prodUCtiol

Th

Sustaining

 

 

 

institution

Smaflhmm

 

152

tenure security necessary to ensure that they will capture the benefits from investments
made. Third, the availability of institutional support (such as input and output markets,
extension services and credit) shapes farmers’ ability to make and sustain investments.

In Zimbabwe, the government’s increasing emphasis on increasing smallholder
production signals the need to reevaluate the current permit system in the resettlement
areas, which poorly defines settlers’ rights to land. Full freehold ownership could limit
distributive equity as illustrated by the case studies. However, it is feasible to specify
freehold rights to land such that settlers with more resources can purchase land from others
to expand their holding, with limited implications for reducing equity in the resettlement
sector. Farmers’ perceptions of tenure security in the resettlement areas have not yet been
thoroughly investigated. However, perceptions of tenure security will influence future land
use in the resettlement areas. For clample, intensification will soon be important in
resettlement agriculture as population pressure increases on the schemes. Small-scale
irrigation systems would allow settlers to intensify crop production in many areas, but their
development is contingent on settlers’ ability to obtain long-term credit and their perceptions
of the long run security of their land improvements. Incentives such as favorable output
prices and improved market access have stimulated communal and resettlement area crop
production, leading to area expansion but not intensification of production.

Therefore, it is necessary to investigate which factors are most important in
sustaining growth in smallholder production, including perceptions of tenure security and
institutional incentives. Empirical evidence regarding the actual constraints facing

smallholders would suggest policy options for land tenure.

l’CSl

I68!

fan

for

int

Fo

ma

TEE

 

alt

tra

M

EV;

tin

Dr:

of

ag

 

153

es e e

Increasing pressure from human and cattle populations on communal area land has
resulted in severe land degradation in these areas. There is already evidence that some
resettlement area farmers are surpassing recommended stocking rates for cattle, suggesting
that overgazing and soil erosion could be imminent in the resettlement areas. Resettlement
farmers, who have larger landholding, also have more options than communal area farmers
for expanding their grazing area to accommodate larger herds or developing rotational
grazing systems to prevent overgazing. However, there are limits to gazing area expansion,
and thus, farmers need viable options for developing new livestock management techniques.
Therefore, further research is necessary in order to identify sustainable alternatives for
intensive or extensive livestock production that are suitable to the various natural regions.
For example, in the higher natural regions it may be possible to improve pasture
management by experimenting with fertilizers and new fodder crops, while in the lower
regions options for extensive rangeland management must be explored. Any proposed
alternative must provide farmers with low-cost options, and access to the information and
training necessary to implement them.

l l . . ! l .

Due to organizational and financial constraints, Zimbabwe’s current monitoring and
evaluation activities for resettlement fail to provide policymakers and administrators with
timely information on its performance; and its impact on the national economy and on
progam beneficiaries. Moreover, current methods do not permit a systematic investigation
of the effects of institutional and policy factors on farm-level incentives for increasing

agricultural production and on the distribution of resources. Therefore, there is a critical

 

 

 

need to

follow-in

unit con

in addit

scheme

Compari

the nati

Potentia

154

need to institutionalize a more effective evaluation strategy, which would include the
following components:

1) centralizing the responsibility for monitoring and evaluation such that the M&E
unit could reassess how the existing evaluation data meet the needs of information-users,
in addition to coordinating the collection, analysis and diffusion of necessary information;

2) systematically collecting performance and impact indicator data at both the
scheme and farm levels which could then be compared to planned targets, used to draw
comparisons among other schemes or settler groups, or quantify resettlement’s impact on
the national economy;

3) initiating issue-oriented studies to better understand local constraints and
potential problems based on issues identified though the collection of indicator data; and

4) decentralizing data collection and selected data analysis activities to decrease total
government costs for evaluation and increase participation in scheme-level assessment.

Strengthening the current system could increase the management and diagnostic
capability of botln policymakers and administrators, and thus improve the success of
resettlement in increasing smallholder agricultural production and redistributing incomes.
However, perhaps it is more important to create a geater demand for the information
generated by a monitoring and evaluation system. Without demand for evaluation
information, from both policymakers and scheme administrators, there is little benefit

gained from improving the current system.

53 Limitations of the Study
The principal limitation to this study is the absence of primary data on the

performance, impact and efficiency of resettlement in Zimbabwe. Thus, it is difficult to

155
draw more specific conclusions regarding the impacts of land resettlement policy on

different population groups in different regions. While the secondary data used provided
important insights into potential problems at the scheme-level, interviews with ministry
officials, scheme administrators and settlers on different schemes are required to conduct
a more extensive analysis of resettlement, as well as to propose more specific
recommendations for scheme and farm-level evaluation.

Furthermore, primary research would have permitted the author to assess the
Monitoring and Evaluation Unit’s current resources and constraints in geater detail, with
the goal of proposing specific options for streamlining and improving the coordination of

its data collection activities on resettlement schemes with those of other ministries.

5.4 hiture Research

Some issues for applied research on resettlement in Zimbabwe have been identified,
including; examining the viability of resettling landless and near-landless individuals,
investigating the legal security of women on resettlement schemes, develOping strategies to
prevent potential overgazing and soil erosion problems and analyzing the potential for land
tenure reform in the resettlement areas.

However, after more than a decade of resettling farmers, more extensive research
should be conducted to assess the regional impacts of resettlement, the long-term viability
of farming in certain areas and the impact of the resettlement area tenure system on
traditional intergenerational land transfer. Further research and analysis of these second
generation effects would enable policymakers to improve long-term implementation,
particularly in the provision of necessary financial, administrative and institutional support

to resettlement areas.

 

 

 

 

of

156

To extend the comparative research undertaken in this study, one could examine the
Kenyan land resettlement progam to assess the differential performance of the high- and
low-density schemes compared to other smallholders, and in particular, the long-term impact

of land registration on the distribution of benefits among settlers.

APPENDIX A

CASE STUDIES

APPENDIX A

CASE STUDIES

Peru: Land Redistribution Policy

l. Pre~Reform Agrarian Structure

Peru’s pre-reform agarian structure is typical of most countries in Latin America,
in that most of the land area is concentrated in the hands of a few large landowners who
also have considerable political power, while many smallholders cultivate small, marginal
parcels. 1961 census data reveal that 1.3% of the landholders possessed 84% of the
agicultural land while, 83% of the landholders held about 6%. The largest estate holding
were at least 500 hectares, while smallholders fag-med plots under 5 hectares; one-fifth of
which were under half 3 hectare (Alberts, 1981). 9

There are vast differences in the quality of agricultural land in Peru, ranging from
irrigated coastal land to remote jungle (selva) to densely populated and over-cultivated
highland areas. Therefore, figures on land concentration that do not consider variable land
quality tend to overestimate its true extent (Alberts, 1981). Standardizing the unit of land
area to account for this variability provides a clearer idea of land ownership in Peru. For
example, King (1977) estimates that family farms averaging about 8 hectares covered 4%
of the land area in the coastal region and 5% in the highlands. Of the irrigated land in the
coastal region, about 50% was in 180 holding of over 500 hectares, while only 6% was in
36,000 holding of less than 5 hectares.

Prior to the reform there were three principal farrrning systems in Peru; plantations
or estates, haciendas and peasant communities. Plantations, located on 80% of the coastal
farmland, mainly produced sugar, rice and cotton for export. Haciendas, located in the
highland areas, comprised both large livestock enterprises and smaller livestock-crop estates.
Peasant communities, collective goups with strong econonnic, social and political ties, were
found primarily in the highlands.

The differences in land quality and working conditions among these farming systems
led to skewed income distribution, both inter-regionally and intra-regionally. For example,
in 197 1/ 72 (illustrating that these differences existed even after some reform measures were
taken), the average household income in the Coast area was about two times that earned
in the highlands, but comparable to incomes earned in the Selva (Kay, 1983). In the Coast
region, 5.8% of the people earned less than 670 (3 15.40) soles per month and 9.3% earned
over 6000 soles ($138) per month (where the average monthly household income was 3300
soles or $76). In the highlands, however, 37% earned less than 670 soles ($15.40) monthly
while only 5.1% earned more than 6000 soles ($138) monthly. This inequity resulted in
widespread poverty and poor nutrition (Kay, 1983).

 

39 Peasants traditionally fragnented their plots as a method of risk avoidance by
locating them at different altitudes and ecological areas (Ghose, 1983). Traditional
fragnentation, however, has been severely compounded by inheritance and population
pressure.

162

163

2. Impetus for Reform

In the 1950s and early 1960s the government realized that the landholding pattern
could be blamed, in part, for Peru’s limited agicultural gowth. King (1977) cites that while
food production was increasing at 2.5% per year, the population gew by 3% and food
demand increased by 4.4% annually. With declining agicultural exports and increasing food
innports, reform in the agicultural sector slowly became a pressing issue. The impetus for
a land reform began in the early 1950s when peasant rebelled and invaded large estates.
This movement eventually gew to number 300,000 peasants who demanded greater access
to land resources.

3. Goals and Objectives of Land Redistribution

Peru’s land reform policy unfolded over portions of two political regimes, from 1964-
1968 and 1969-1975. The first government focused on improving the country’s social
structure, its productive capacity and the distribution of national income. When a new
government came into power through a military coup in 1968, the new priority became
income redistribution through structural reforms in the agricultural sector (Alberts, 1981).
Economic growth was also an important concern, but the government assumed that gowth
would be a natural result of the reform. The government added other objectives to the
development plan for 1971-1975, the most important of which included developing
cooperatives to increase rural participation in decision-making, accelerating the transfer of
property, increasing the health level of the population, providing social services to marginal
rural people, and increasirng, as much as possible, production and productivity, especially in
agriculture (Alberts, 1981).

The government also pursued several unstated objectives. For example, the
government increased supplies of marketable surplus production to the gowing urban
centers, and to foster industrial gowth. Furthermore, the government sought to legitimize
its power by involving the rural population in the political system, thereby creating a
stronger national state (Kay, 1983).

Policy Implementation: 1964.68 and 1969-1975

4. Land Expropriation

The first land reform bill passed in 1964, stated that initially public lands were to be
expropriated, followed by abandoned properties and those owned by absentee-landlords
(King, 1977). The potential for the reform’s almost exclusive regional impact quickly
became apparent. The government gave priority for land transfer to those regions with the
geatest degee of peasant upheaval (mostly in the highlands), while excluding coastal sugar
plantations, efficiently-farmed estates and corporately-owned land and ago-industrial
operations from expropriation (King, 1977). Due to limited funds and congressional
opposition, the level of expropriation was very low during this period (Kay, 1983). Those
families who did receive land at this time were awarded individual parcels.

In order to speed up implementation in 1969, all large estates were subject to
expropriation without exception. In fact, the government began this phase of the land
reform "by expropriating the most profitable and capital-intensive agicultural enterprises,
showing their determination and ability to liquidate the rural oligarchy' (Kay, 1983: 207).
Former farm workers took over most of the expropriated land and turned them into
collective farms.

 

 

C
a strong
oollectiv
Catholic
owner 0
farm int
1981). A
owners
eliminal

S. Pmn'

partly i
calculat
potenti:
1981).

AlthouE
budgeg
Expendi
annual}:

0n the t
‘0 half l
govemr
govemr
years tc

Iegardil

164

Collectives predominated during the Peruvian for several reasons. First, Peru had
a strong history of cooperative action, originating from Marxist ideology, early Indian
collective farming systems and the various economic organizations introduced by the
Catholic church. Second, government admirnistrators found it much easier to displace the
owner of an estate and hand the entire operation over to the workers, than to subdivide the
farm into individual parcels and redesign the production system for smallholders (Alberts,
1981). Additional provisions in the law included giving squatters the right to become legal
owners of 15-30 hectares of land if they could prove de facto tenure status, and the
elimination of semi-feudal labor arrangements (King, 1977).

5. Provisions for Compensation of Landholders

The government compensated the former landowners for the expropriated land,
partly in cash, partly in bonds. The amount of cash compensation was determined by
calculating the land value as the average of the declared tax value, the market value and the
potential income that would be realized from the estate, under good management (Alberts,
1981). The bonds could be used either as tax payments or for an industrial investment.
Although this compensation might seem to place a considerable burden on the government’s
budget, but Alberts shows that the ratio of reform expenditures to total government
expenditures never reached more than 1.1% in any one year, and averaged about 0.5%
annually.
In 1968, the government changed the compensation method, thereafter valuing land
on the basis of self-declared tax assessments for 1968. Actual cash compensation decreased
to half the former amount and the remaining portion was paid in low-interest, long-maturing
government bonds (King, 1977). The farmers who took over the land reimbursed the
government in installments over a 20-year period, and received a grace period of up to five
years to begin repayment.

When the government initially enacted the land reform law, it set no targets
regarding the amount of land to be transferred, the number of farms to be expropriated or
the number of farmers to be benefitted. Target figures given by the government between
1969 and 1975 estimated that between 336,000 to 500,000 families would eventually receive
land, on 10- 13 million hectares of expropriated land (Alberts, 1981). This was the equivalent
of redistributing about 41% of Peru’s agicultural land to 39% of the peasant families (Kay,
1983).

Assessment of Land Redistribution

6. Land Distribution

Table A.1 presents data on the achievements of Peru’s land redistribution. Even
though these figures are below the targets, the government expropriated a substantial
amount of land and redistributed it to many families. These figures hide the fact that
although only a small percentage of land was actually distributed to individual smallholders,
43% (mostly nn the highlands) was subsequently fa aede by individuals, not cooperatives,
which is contrary to the government’s initial intentions (Kay, 1983). The government had
difficulties establishing cooperatives in the highlands because hacienda tenant-labor relations
were more entrenched among the peasant farmers. Also, peasants were not familiar with
the type of centralized management system found on coastal estates and from which it was
easier to form cooperatives (Alberts, 1981).

 

 

l

Table .
Benefi

 

165

Table A.1: Land Area Expropriated, Distributed and Number of Families
, Benefitted During Peru’s Land Redistribution, 1964-1978.

 

Time Period Land Expropriated Land Distd.a No. Benet.b
(100,000s ha) (100,000s ha) (000s)

. 1964-68 2.7 3.0 10
NA 1.8 90
3.0 2.2
7.4 6.2 245
NA 6.8 292
8.6

 

 

 

 

 

 

 

 

 

 

a Cumulative amount of land adjudicated and distributed.
b Cumulative number of families benefiting from land redistribution.

Sources: TomAlberts, . u ._ '_ . .
140-195); Russell King, W (1977: 177-180).

 

Several problems arose in carrying out the land redistribution. First, the government
had difficulties accommodating the many squatter families who applied for titles to land.
Of a total of 160,000 families, 128,000 applied to be registered but only 55,000 received even
provisional certificates, and very few received permanent titles (Alberts, 1981). In addition,
most of these families received small 6 hectare plots on poor soils. Thus, instead of
resolving some of the minifundio problem, a geater number of new minifundistas emerged
from the land reform. Second, because the overall expropriation procedure was long and
complicated, the process of expropriating the assets, handing them over to the beneficiaries
and drawing up a contract required an average of 18 months to complete. This greatly
reduced the amount of land that could be redistributed (Alberts, 1981).

7. Post-Reform Agrarian Structure

Critics argue that the Peruvian government concentrated too much on the type of
land to be expropriated, instead of working toward the agrarian structure desired through
the reform process. The mixed successes and failures of the reform center around the
state’s desire to emphasize collective agiculture and therefore it is difficult to draw specific
conclusions about the smallholder sector because more is known about the reform’s effects
on the politically sanctioned collective sub-sector than about small farmers and the rural
poor (Alberts, 1981).

Using the total number of production units as a point of reference, it appears that
the distribution of ownership did not change drastically from 1961 to 1972. This is because
the government rarely subdivided the expropriated estates, but maintained them as

166

cooperatives. The predominant form of ownership did change, however, since the
government prohibited alternative forms of indirect exploitation, such as sharecropping and
renting land. Therefore, the legal status of many peasants did change through the receipt
of legal titles to land (Alberts, 1981).

In terms of the number of smallholding, between 1961 and 1972, the number of
farms with less than 5 hectares increased by approximately 384,000, while their average size
decreased from 1.5 to 1.4hectares(Alberts, 1981). Additionally, smaller farms were
subdivided even further.

In general, the beneficiaries of the reform were the rural upper-class (mostly former
permanent workers on large estates) who constituted almost 30% of rural population. After
the reform, they still had access to almost 50% of the nation’s cropland and obtained the
greatest increases in income (Kay, 1983). At the same time, the comuneros, comprising
29% of the population, held only 19% of the land.

The comuneros and the minifundistas were, for the most part, excluded from
redistribution. Political struggles surrounding implementation prevented improvements from
taking place in the minifundio areas. Peruvian officials were hesitant to grant peasants the
right political participation since they feared that extending the reform to the minifundistas
would adversely affect agicultural production (Alberts, 1981). The minifundistas did not
receive better access to technical assistance and credit since most of the reform effort went
towards establishing and maintaining the cooperatives (Alberts, 1981).

8. Agricultural Production

Although the land redistribution had different impacts on regional agiculture, there
was little change in the composition of agriculture and no subsequent decline in production
levels.

The reform had differential effects on agicultural production among the various
cooperative enterprises. Coastal area sugar, cotton and rice plantations were expropriated
and turned into cooperatives. The data indicate that sugar production was not affected by
the reform, and in fact increased until droughts caused output to fall (Alberts, 1981).

For the, cotton and rice cooperatives, the data are less conclusive. Studies generally show
that production increased slightly and real wages rose significantly following the reform

y, 1983).

(K1! In the highland areas, the pre-reform agarian structure permeated the newly
reformed enterprises, causing differential degrees of proletarianism and technological
development (Kay, 1983). Despite considerable variation across the highland livestock
cooperatives, it appears that real wages increased until about 1978 and private ownership
of livestock continued to grow, even though profits and investment fell (Kay, 1983).

The livestock-crop units were originally planned as collective enterprises, but at least
one third, if not one half, of the total units are held in individual plots and livestock herds
(Kay, 1983). It is difficult to draw specific conclusions since data on these units is sparse.
However, it appears that the wages derived from collective farming did not equal the income
that farmers earned from their own plots. One study showed that labor productivity
eventually fell on the collective units, leading to asset disinvestment on many units (Kay,
1983 .

) Official data do not indicate a decline in overall agricultural production associated
with the reform period. One does see, however, a stagnation and eventually a decline in per
mm production from the 19603 to the middle of the 19708. This is due primarily to a
declining resource base, the cultivation of increasingly less productive land (marginal), low

 

 

 

private .
against

product
primarl
inequal
urban n
and ex;

that gr
aperie
period,
1279’ ,

9. Inc:

situatig
distrib
reforn-
Cooper
reiom
period

incom
group
25%

lllCOfr
agriq

Urb

167

private investment in agiculture and price controls and import subsidies that discriminated
against agiculture (Alberts, 1981).

As Kay points out, the reform had little impact on the composition of agicultural
production because “the Higlnlands produce primarily for the rural market and the Coast
primarily for the urban market" (1983: 231). The reform accentuated these regional
inequalities. In addition, the same pre-reform production pattern was evident, where the
urban market absorbed 65% of total agicultural production, the rural market received 27%
and export markets 7.9% (Kay, 1983).

Looking at overall economic growth during the reform period (1964-75), one sees
that goss domestic product (GDP) gew at an annual rate of 5.4%. The agiculture sector
experienced an annual gowth of 3.4% per year from 1960 to 1975. During this same
period, however, agiculture’s percentage contribution to GDP decreased from 18.5% to
12.7%, while manufacturing’s contribution gew from 20% to 26.2% (Alberts, 1981).

9. Income Distribution

Although there are no time series data from which to analyze the post-reform
situation, the various studies reviewed indicate that the reform did not improve the
distribution of income among Peru’s rural population. In the cooperative sectors, land
reform altered income distribution to varying degrees. For example, in the cotton and rice
cooperatives, real incomes per capita for permanent workers increased by 78% after the
reform (1968 to 1972), but temporary workers’ incomes increased only 27% for the same
period (Kay, 1983).

Examining income distribution by quintile (excluding property income), personal
income declined 0.5% to 1.0% in all quintiles (1961 to 1972). Only in the highest income
group did personal income increase (+ 2.5%). In 1972, the lowest income goup earned
2.5% of all personal income while the highest goup retained 58%. Low average rural
incomes are due primarily to low levels of agicultural output and the slow gowth in
agriculture, relative to gowth in other sectors of the economy (Alberts, 1981).

Alberts (1981) concludes that at least the lowest 40% of the population saw their
incomes decrease after land redistribution and that there was most likely an increase in
income concentration. Although there are no estimates since 1972, it is thought that for
most of the labor force, real income has been decreasing since 1973.

It was mentioned in the previous section that, as a result of the land reform, the
minifundia sub-sector had gown larger, and their holding had become more fragnented.
Alberts (1981) found that these smaller farms had a more limited income-generating
potential. He gives two reasons for this. First, the minifundistas have limited means of
investing in their productive capacity because of a lack of access to credit and technical
assistance. Second, as the holding are increasingly subdivided below 5 hectares, production
costs per unit of product go up, resulting in higher transportation costs. This makes
economies of scale diffith to achieve and thus makes investment prohibitive for
minifundistas.

10. Employment Generation

As a result of the reform the rural population decreased, even though new
employment opportunities were created in agiculture. Prior to the reform, 67.2% (1961)
of the population lived in rural areas, whereas in 1972, 52.5% were rural dwellers (Kay,
1983). This decline in the rural population was due primarily to heavy out-migation to
urban areas. Kay shows that the agicultural population also decreased in proportion to the

 

44 . 4_4,_- —._ IIALM.

 

 

 

 

168

urban population. In 1961, 50.5% of the rural population was employed in agiculture
compared to 42.2% in 1972. In addition, the composition of the agicultural population
changed. The number of self-employed in the agricultural population rose from 50% in
1961 to 76% in 1972. The proportion of wage-laborers fell from 30% to a little more tlnan
20% over the same period (Kay, 1983). The increased number of new self-employed are
mostly smallholders who were settled on colonization schemes in the highlands.

The government planned to create over 300,000 jobs in agiculture through land
reform, however, ultimately only 171,900 jobs were created. Even so, employment gew by
0.9% - three times the gowth in male employment from 1961 to 1972 (Kay, 1983).
Therefore, the reform did have a positive impact on employment creation, but not enough
to stem much of the rural outmigation.

 

 

 

landRe

l. on;

i
(hacienl
system.
theu‘es
owners]
agrarian
Bolivia]
Bolivia]
penod.
oimel

2. Pre-

CXploit.
desolat
00mm
10 the .
less ax]

hacien

(haciel
nnnfiu
Privile
These
Comm

Comp,
feWet
6% o:
1961)
most
(2011cE

lg

Obiec
of In:
den/e;
(Car

the F

lo “1
111),,

169
Land Redistribution in Bolivia

1. Origins of Land Redistribution in Bolivia

In the late 1940s, widespread conflict arose in the rural areas between large estate
(hacienda) owners and their tenants over the repressive nature of the hacienda-tenant
system. A violent revolution ensued, during which peasants expelled hacienda owners from
their estates to take possession of their lands and break up their stronghold on land
ownership. During the aftermath of this revolution the government decided to draw up an
agarian reform law in an attempt to redress some of the severe inequalities found in the
Bolivian landholding system. It is very difficult to conduct a complete evaluation of the
Bolivian land reform because time series and census data are lacking for the post-reform
period. However, case studies and various reports provide some insights as to the impact
of the reform on the agarian structure and on the agicultural sector.

2. Pre-Reform Agrarian Structure

Before the reform, the highland areas of Bolivia contained one of the most
exploitative hacienda systems in Latin America. Part of this region, the altiplano, is dry and
desolate, and farms are concentrated on the margins of the plateau. The Indian
communities found on the altiplano provided labor and land for the haciendas’ development.
In the valleys, northeast of the altiplano, subtropical cash crops are gown under a different,
less exploitative hacienda system. South of this region lies a mountainous region containing
haciendas, a high concentration of minifundios, and Indian communities.

Prior to 1950, three principal farnning systems existed in Bolivia: the latifundia
(hacienda or large estate), Indian communities (indigenous farming communities) and the
mirnifundia (small, individually-owned family farrrns). The latifnnndia were held by a
privileged landowning class but farmed by sharecroppers, tenants and Indian servants.
These haciendas had gown in size and power by destroying the neighboring Indian
communities and incorporating them into their estates (Garcia, 1970).

As of 1950, 78% of the farms in Bolivia contained 20 has of arable land or less, but
comprised only 1.0% of the total farmland. Sixty percent of these were minifundia with
fewer than 5 hectares of land, or an average of 2.2 hectares per holding. On the other hand,
6% of the farms were over 1000 hectares, comprising 92% of the total farmland (Carroll,
1961). The land reform was intended to have its geatest impact in the highlands where
most of the haciendas were located, and where 93% of Bolivia’s population was
concentrated on 40% of the country’s land (King, 1977).

3. Goals and Objectives

The land reform decree issued by the government in 1953 cited the following
objectives: ”1) redistribution of land that does not perform a social function; 2) development
of Indian communities; 3) reform of agicultural labor relations; 4) promotion of agicultural
development; 5) conservation of natural resources; 6) stimulation of internal migation"
(Carroll, 1961: 176).

The government’s overriding goal, however, was to destroy the latifundia and reduce
the problem of severe land concentration. As Garcia points out, "agarian reform was born
to the political ideal of destroying, by one blow, the social structure of the hacienda, thus
liberating the labor and creative energy of the peasant masses” (Garcia, 1970: 309).

170

As it will be shown below, emphasis on this last goal, in addition to fiscal constraints,
prevented the government from developing the supporting services necessary to ensure the
achievement of its other objectives.

Policy Implementation

4. Land Expropriation

The 1953 reform included provisions callirng for the abolition of the latifundia, and
the exemption of large farms from expropriation where the owner had already made large
capital investments and improvements to the land (Carroll, 1961). The law identified
different types of farms to be inspected and categorized for expropriation, and limited the
size of property that could be maintained after the reform, depending on the ago-ecological
area (King, 1977).

The latifundia were to be expropriated without compensation if: 1) the property was
owned by an absentee landlord; 2) the production process was considered technologically
obsolete; or 3) the tenants or workers were exploited (King, 1977).

Medium-sized properties could be expropriated if they surpassed a predetermined
size, determined according to regional differences in land quality. The ceiling varied from
6 hectares for irrigated vineyards to 350 hectares for farmland in the altiplano.

Small properties could not be expropriated except if they exceeded a maximum of
3 hectares in vineyards or up to 35 hectares on the altiplano. Estates using more modern
agicultural practices were allowed to retain up to 80 hectares of good quality land (or up
800 hectares on the altiplano), with the remainder to be expropriated (King, 1977).

The distribution of land to peasants usually followed their occupation of the estate,
which tended to “reflect the political strength of local sindicato organizations [and] the
persistence of local landowners in fighting campesinos’ claims," rather than the legal
distinction cited above (Eckstein et al., 1978: 22). Redistribution allowed the hacienda
workers to take immediate possession of their existing plots, as well as some portion of the
hacienda’s land (King, 1977). The government compensated owners of expropriated
properties with 25-year bonds at 8% interest. The amount of compensation was based on
the land value determined from that year’s tax assessment (Carroll, 1961). Beneficiaries of
the reform were also expected to repay the government for the land allocated to them.

Assessment of Land Redistribution

5. Land Distribution

The reform was marked by upheaval. Armed peasants occupied haciendas, expelling
the landowners and destroying a considerable amount of estate assets, including livestock
herds (Carroll, 1961). The first five years, in fact, were subject to political struggles,
economic difficulties and high inflation (King, 1977). This led to much insecurity regarding
rights to land, which was compounded by the government’s inability to gant titles to
beneficiaries upon their possession of the property.

Although the government distributed much of the land by 1955, the land title
certification process was more lengthy (Table A2). As of 1960, about 59,000 families had
received land titles, although the number of elig'ble families was estimated at between
100,000 to 200,000 (Carroll, 1961). By 1970, only 37% of the redistributed farmland had

171

been titled, and less than 30% of the landholders had actually received a title to their land
(Eckstein et al, 1978).

Approximately 25 % of all land transferred was titled collectively, but the farm unions
(sindicatos) receiving the collective titles operated poorly. Thus, by the 19708 most of this
land had been redistributed to individual holders (Eckstein et al, 1978). The difficulties
experienced in distributing and titling the land is partly attributed to the judicial procedure
for determining which land was to be expropriated and establishing the respective rights of
landowners and beneficiaries. This process often extended over 6 or 7 years, and sometimes
lasted as long as 10 years (Eckstein et al., 1978).

Administrative, technical and financial resource constraints also linnited the
government’s ability to effectively carry out the land reform progam. Most of the country
was unmapped at the time of the reform and little data existed on soil types and land use.
By 1960, few landlords had received their compensation and none of the beneficiaries had
been required to pay for their holding (Carroll, 1961). In addition, any complementary
services introduced during the reform period generally by-passed the reform sector and were
applied to commercial development efforts in the eastern region (King, 1977).

i Table A2: Land Area Distributed and Number of
3 Families Benefitted from Bolivian Land
? Redistribution, 1955-1975.

 

Land Distributed Number of Families
(000s no)a Benefitted (000s)al

4.2 59
3.8 185
9.8 237
18.0 477

 

 

 

 

 

 

 

a Totals are cumulative as of the end of each year figured.

Sources: Schlomo Eckstein et al., W (1978: 11); Thomas F.
Carroll. WW... (1961: 176-77); Russell King. Iand
W (1977 113)

172

Colonization was to be a complementary policy to the land reform, designed to
redistribute part of the population and concentrate these human resources on agiculture
(Garcia, 1970). The colonization progams, heavily supported by the Bolivian government
and foreign donors, contributed to an increase in agicultural production (mostly in sugar,
rice, cotton and coffee production, or cash crops) but had little effect in stemming the
increasing population pressure found in some areas (Garcia, 1970). The attempts at
colonization in Bolivia are best summarized by Garcia. He points out that only settlements
formed around the important elements of the rural community, or formed near traditional
nnigatory routes of the rural population were successful.

6. Post-Reform Agrarian Structure

There have been marked changes in Bolivia’s rural areas since 1953, but these
changes cannot all be attributed to the reform. With respect to the concentration of land
ownership, from 1950 to 1970 neither the size nor the number of minifundia holding
changed. Farms of less than 5 hectares still comprised about 0.2% of the farmland, their
average size remaining at 1.4 hectares per holding. Due to the subdivision of large estates
into smaller holding and an increased number of small property owners, the number of
nninifundia decreased in proportion from 60% of all farmers before the reform to 14% after
the reform. Farms ranging in size from 5 to 100 hectares ("small properties“) increased
their total land holding from 1.4% to 31%, and the average holding size increased to about
30 hectares (Eckstein et al., 1978). Most of the new landholders in this size category were
reform beneficiaries.

The number of large farms decreased significantly after the reform, although the size
of the holding did not decrease. For example, the number of farms with holding of 100-
1,000 hectares fell from 7% to 1% of the total farm units, and their average size declined
from 350 to 280 hectares. Very large estates decreased in number by approximately the
same amount, but the average holding increased in size from 5,018 to 7,097 hectares
(Eckstein et al., 1978). Large estates increased in size due to the expansion of the eastern
frontier beyond the Andes mountains. Most of these new estates were formed into large-
scale ranches, some of which even exceeded the 50,000 hectare ceiling placed on large
holding (Eckstein et al., 1978).

The conflict which inspired the revolution continued for years following the reform.
After the reform was innitiated, some landlords in an efiort to regain their properties or to
receive some payment for their former land, continued to intimidate the new peasant
owners (J acoby, 1971). These conflicts indicated a trend toward the reassertion of the
landlords’ power. Some landlords regained all of their properties or were able to persuade
peasants to return to the estate as sharecroppers; especially those who had not yet received
a land gant, (King, 1977). Since 1971, many landowners had formed affiliations with the
military regime, thus gaining more political influence (King, 1977).

7. Distribution of Income and Benefits

Although the reform improved living conditions for some of Bolivia’s population
(particularly those who had worked on the expropriated haciendas), the reform had only a
regional impact. Since the reform, land use changes in other areas of the country resulted
only from expanding and intensifying farming in the tropical lowlands (King, 1977).

In the reform areas, however, available evidence indicates that geater security of
tenure encouraged farmers to invest in housing construction on their land. The
development of settlements, many along the roadsides, improved farmers’ access to

173

marketing and transport facilities. In the Cochabamba valley, for example, it was estimated
that the total value of production increased tenfold as a result of post-reform changes in
land use (King, 1977). King cites that farmers on one hacienda increased their marketable
production from 10% to 70% after the reform. Also, farmers on former haciendas began
stocking geater numbers of livestock than the more extensive estates had previously
retained. Overgazing in this area, however, could lead to erosion and deterioration of the
land (King, 1977).

The revolution and the ensuing reform destroyed the social, economic and political
vestiges of the hacienda, as well as other types of servitude (Garcia, 1970). However, the
reform had little effect on the Indian communities because it did not provide for the
necessary structural change in tenure, agicultural practices or expanded institutional
relationships with the rest of the economy (Garcia, 1970).

As for the minifundia, the reform failed to remodel or concentrate land parcels to
diminish the extreme fragnentation of the smallholding (Garcia, 1970). In addition, the
rapid manner in which the reform was implemented, and government-level resource
constraints ruled out any attempt at launching rural development strategies targeted at the
minifundia. In these areas, gowing population pressure on a limited resource base
eventually forced some farmers to migate out of the region in search of employment. In
areas too distant from potential employment sources, hidden unemployment gew because
these workers could not be absorbed by the existing farming practices carried out on small,
marginal plots (Garcia, 1970).

8. Agricultural Production

While little time-series data exist to evaluate trends in Bolivia’s agicultural output
before and after the reform, information from various reports and case studies outline the
major issues.

Initially, from 1952 to 1955, overall agicultural production dropped by 70-87% (there
are many different estimates), eventually rising to its pre-reform levels around 1960 (King,
1977). Most of this subsequent increase, however, was attributed to increased rice, sugar
and cotton production in the newly colonized eastern areas which were not affected much
by the reform.

In the early 1960s, however, annual gowth in the agiculture sector dropped from
2% to almost zero in 1970 (King, 1977). This decline was accompanied by severe economy-
wide problems which were exacerbated by high inflation, price controls, and sporadic
droughts. During this time wheat imports doubled and the marketing system, previously
maintained by large landowners, basically collapsed.

In the reform sector, the production of corrn, potatoes and wheat initially declined
but recovered by 1958-60, although incidents of drought tend to complicate this evaluation
(King, 1977). Recovery continued into the 19708, with annual gowth rates of 6.3% for
potatoes, and 4.8% for maize and rice, the main staple foods of the campesinos (Eckstein
et al., 1978).

Eckstein et al. (1978) found that some of the short-term decreases in food crop
production in the reform sector can be explained by the disincentives created by PL-480
wheat imports and government policies encouraging barley imports. Although they were not
able to obtain direct production estimates for the areas with the largest number of land
reform beneficiaries, they cite indirect evidence that basic food production increased
significantly as a result of the reform.

174

Using Bolivia’s 1950 agicultural census, they argued that the subsequent post-reform
increases in potato output, which occurred after the land was subdivided into smaller farms,
were due to the strong negative correlation between yield and farm size (Eckstein et al.,
1978). Potato yields increased by 242% from 1950 to about 1972, mostly due to increased
use of chemical fertilizer by campesinos. This is significant because potatoes are a
campesino crop for which labor intensity increases with fertilizer use, rather than decreases.
It is most likely that fertilizer was imported as a result of cultivator demand for the input.
If demand for fertilizer had come from large landowners, they postulate, it would have been
used on cash crops. This implies that some of the reform beneficiaries obtained better
access to inputs.

Finally, there is also evidence of increased vegetable production among campesinos
in the highland and valley areas. This appears to be the result of improved access to
markets among campesinos in areas with extensive land redistribution (Eckstein et al., 1978).

175
Land Redistribution in Indonesia: Transmigration Policy

l. Origins of Transmigration Policy

Transmigation originated in 1905 when the Dutch administration in Indonesia
attempted to relieve gowing population pressure on Java by moving nnigants to the outer
islands. Approximately 200,000 nnigants were settled by 1940, in addition to the large
number of spontaneous migants who left Java to work on Sumatran plantations. After
1942, fifteen years of war and political instability left the newly independent Indonesian
government with few resources to allocate to transmigation. Even though a more
comprehensive transmigation policy was planned in 1947 (3 1 million people over 15 years,
changed in 1951 to 49 million over 35 years), changes in government administration and
economic problems produced linnited results. Until the creation of the first five-year plan
(Repelita I), the government had no firm resettlement progam and settled varying numbers
of migants, from 25,000 a year to almost none (Oberai, 1988).

2. Land Tenure Policy

The 1960 Basic Agarian Law established a uniform tenure system. This law
supported individual ownership throughout the country, and placed rest4 ons on land
purchases and sales through the regulation of absentee land ownership. In addition,
the law regulated the transfer of use-rights through leasing, sharecropping and pawning
(Booth, 1988). The Basic Agarian Law protects customary land rights, except in cases
where they conflict with national interests (World Bank, 1988).

Traditional law, or gig], allows the community to allocate the right to cultivate, with
no guarantee of tenure security (World Bank, 1988). Incoming nnigants may be asked to
compensate the local community for the land they obtain, and sometimes migant holding
are subject to expropriation if the value of their land increases to the point where it
becomes valuable to the community. However, due to the absence of legal records or
precise boundaries, the establishment of a legal transfer of title from the local community
to nnigants becomes difficult (World Bank, 1988; Oberai, 1988). Therefore, government
sponsorship of the transmigation progam is essential in mitigating some of the conflicts
arising over land in the outer islands (World Bank, 1988).

3. Goals and Objectives of Transmigration

By 1966, Indonesia’s population had gown to more than 70 million, with average
population densities reaching 600 persons per square kilometer in some areas of Java. At
this time, it was estimated that two-thirds of the population lived below the poverty line, and
that per capita income was steadily decreasing (Oberai, 1988). Therefore, after 1966, the
emphasis of the transnnigation progam shifted from alleviating population pressure to
raising living standards for botln migants and the inhabitants of Java. It was also planned
that the progam would promote economic development for the outer islands (Oberai,
1988).

 

40 The amount of land that any one family could own was limited to 5 hectares of
wetland and 6 hectares of dryland in the most densely populated areas, and 15 hectares of
wetland and 20 hectares of dryland in the least densely populated areas. A minimum
holding size of 2 hectares was established for both wetland and dryland areas (Bootln, 1988).

176

The First Five-Year Plan (Repelita I; 1969-1974) emphasized increasing food
production and creating a more balanced population distribution throughout the country.
The Second Five-Year Plan (Repelita II; 1974-1979) placed more emphasis on improving
social welfare and stimulating regional development. Initially, a target was set for settling
250,000 families during this period. Later, in 1976, this target was revised to a total of
108,000 families.

The Third Five-Year Plan (Repelita 111; 1979-1984) emphasized settlements as
gowth points that would contribute to regional development and extend beyond agiculture
into processing and other industries. The plan was based on a target of 500,000 families or
2 million people. The oil price increases of 1973-74 and 1979-80 enabled the transnnigation
progam to gow from about 0.7% to almost 6% of the development budget by 1978
(Oberai, 1988). .

The Fourth Five-Year Plan (Repelita IV; 1984-1989) called for the settlement of
750,000 families. However, in January of 1986, the government of Indonesia was forced to
significantly reduce its development budget resulting in a 56% reduction in the
transmigation budget (World Bank, 1988).

Program Implementation Under the Five-Year Plans

4. Settler Selection

The selection of settlers was based on three elements: 1) families should nnigate
as units, 2) migation should be voluntary, and 3) people from highly populated areas should
be given priority to migate. The official criteria for selecting transmigants specified certain
personal characteristics, but farming skills - although desired - were not required. Several
studies showed that most had limited experience in farnning; about one-third had never
owned or managed a farm before and less than one-fifth had never farmed before (Oberai,
1988). As the progam expanded, the government stressed the need for transnnigants not
only with farming skills, but also other service skills. Some were homeless, landless and
unemployed persons transferred out of Java by the Social Welfare Department and by
village officials.

5. Site Selection and Management

Inadequate site selection was the primary cause of settlement failure in the 1950s
and 19603 because systematic surveys regarding water availability and soil types were not
carried out and management personnel had no incentive to be effective (Oberai, 1988).
Initially, transmigation sites had to be suitable for irrigated wet rice production, as well as
close to Java in order to reduce transportation costs (Oberai, 1988). As sites of this type
became scarce and more costly, attention turned to swamp reclamation for tidal irrigation.
However, although swamp soils have highly organic soils, tidal cultivation proved technically
diffith due to soil acidity and problems in dealing with tidal effects, annual flooding and
severe pest problems (Oberai, 1988; World Bank, 1988).

Suratman and Guiness (1977) report that the required surveys often involved only
cursory inspections of the potential site carried out over a period of one or two months, but
insufficient to provide information on which to build agicultural development progams and
understand inter-ethnic relations. Consequently, after several years of settlement,
transmigants experienced crop failure, declining soil fertility and land disputes with local
farmers that disrupted farming activities (Suratman and Guiness, 1977).

177

Although further improvements were made to the transmigation progam under
Repelita III, the results were mixed. Site location improved, but site plans and land clearing
were often inadequate, and infrastructure was poorly maintained (World Bank, 1988). In
addition, land and financial resources became geater constraints to progam
implementation (Oberai, 1988). Much of the available land was too mountainous, swampy
or endowed with poor, erodible soils and hence not suitable for agicultural cultivation
(Oberai, 1988).

The Government of Indonesia (GOI) has provided no cash compensation for land
ceded to the transmigation progam. It does, however, provide cash compensation for the
loss of any tree crops located in settlement areas (World Bank, 1988). G01 believes that
if nnigants paid for settlement land: 1) the price of land for local smallholders would
increase; and 2) it would make GOI development progams less replicable.

In the earlier stages of the transmigation progam, migants were expected to repay
the government for the transportation, housing and food provided to them (Hardjono, 1977).
However, by the 19608, the idea of repayment was discontinued because it added little to
the overall transnnigation budget and it was thought that migants could achieve desired
living standards more quickly if they were not overburdened by loan repayments (Hardjono,
1977).

The administration of each scheme remains under MOT (Ministry of
Transmigation) authority for at least five years or until self-sufficiency is achieved. Fiscal
and administrative responsibility is then transferred to the province which also assumes
responsibility for service and infrastructure maintenance (World Bank, 1988). The World
Bank cites road maintenance and staff transfer as the primary difficulties resulting from the
transfer to the provincial level.

6. Service Provision

Over the life of the transmigation progarrn, the provision of assistance to nnigants
has varied geatly. Suratman and Guiness (1977) state that in 1950s, it was assumed that
nnigants needed only to be settled on the land with a seed and food allowance to carry them
over until the first harvest.

Beginning in the early 1970s, government-sponsored nnigants received transportation
to the site, a small house and 0.75 to 1.0 hectares of cleared land (World Bank, 1988). They
also received subsistence supplies for one year, including materials for planting minor tree
crops and additional supplies in cases of crop failure. At the end of five years, settlers are
expected to be self-sufficient (World Bank, 1988). One survey found, however, that in the
first few months, 62% of the household heads interviewed had to find employment outside
of the scheme because they were unable to support their families with the food rations
supplied by the MOT (Suratman and Guiness, 1977).

Cash cr0p schemes were innitiated in 1978 and based on tree crops (rubber, coconut,
and oil palm). These schemes are open to both local smallholders and migants, the latter
receiving a small subsistence plot (World Bank, 1988). Oberai (1988) notes that one of the
most effective forms of assistance has been the cultivation of tree plantations for cash crops.
This progam provides a more secure livelihood for transmigants than the traditional
agicultural production schemes (Oberai, 1988).

There has been a tendency on the part of the indigenous people of the outer islands
to resent the nnigants who have access to better facilities and services (Oberai, 1988). For
example, a 1981 UNDP/ OPE Management and Monitoring survey found that the ratio of
agicultural extension workers to settlers was much geater in transnnigation areas than in

178

the local outer island farming communities (Babcock, 1986). The transmigation authorities
have attempted to mitigate this problem by settling some local people onto transmigation
sites, improving local land acquisition methods, and encouraging indigenous entrepreneurs
to establish relations with the new settlers (Oberai, 1988).

The enormity of the transnnigation progam places an adnninistrative strain on the
country, but the G01 has attempted to improve interagency administration and coordination.
Settlement personnel are charged with providing courses on topics such as animal
husbandry, agicultural skills, local government, cooperatives and family planning and health
(Suratman and Guiness, 1977). Most settlements, however, are plagued by a shortage of
staff, particularly agicultural officers, health workers, teachers and social workers (Suratman
and Guiness, 1977). Oberai (1988) reports that as staff numbers have increased in order
to meet the demands of the expanding progam, the quality of the management and service
personnel has decreased, especially among the on-site staff.

Assessment of Transmigration in Indonesia

7. Achievement of Transmigration Targets

As indicated in Table A3, the total number of migants settled has never equalled
planned government targets for the transmigation progam. The initial target set in 1951
was to move about 49 million people over a 35 year period. However, only about 4 million
people had been settled as of mid-1987. The reasons for the variable numbers of people
settled (Table A3) stem from changes in the amount of financial resources allocated to
transmigation, shortages of appropriate land for settlement and shortages field and central
government personnel. Oberai (1988) asserts that even though the transnnigation progam
represents a tremendous effort to reduce population pressures on Java and promote
regional development, at its current intensity, the progam will not correct the population
imbalance betwA.een Java and the outer islands. Java’s population is approaching 100
million and increasing at 2% annually. Even if the targets for Repelita III are attained, it
will only remove 20% of the annual W in Java’s population (Oberai, 1988).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

'lndigenousfamiliessettledin transmigration“

b Resettlement ofspoasored orspontaneous migrants within theprovince.
cData not available.

‘Numbermmedbynneendornmo.

°Asoriuneso,iss7.

Somme World MWOM m); {Mo Baritone.
(1977: 19-21); A. S. Oberai, . .o . 19'. .

 

 

180

8. Agicultural Production

It is difficult to assess the regional impact of transmigation on agicultural:
production because: 1) nnigants may need several years to reach full production levels1
2) many migants improve their operations over time through the addition of tree crops or
irrigation, but it is impossible to separate the transrrnigants’ contribution to aggegate
production from that of the local population; and 3) provincial data are very unreliable
(World Bank, 1988: 84).

Table A.4 provides information extracted from a World Bank analysis on the
regional impact of transmigation on rice production throughout Indonesia. This analysis
is based on the following assumptions that approximate the actual situation in each province:
1) 80% of all sponsored nnigants produce an average yield of 700kg/family of rice; and 2)
for each government-sponsored migrant family, one spontaneous migant family has settled
in the same province, with the exception of two areas (World Bank, 1988).

Although data for each province are not included in the above table, in certain
provinces transmigants have contributed from 45-96% of incremental rice production. It
is apparent from the above analysis that transmigants’ contribution to rice production is
higlnly variable at the provincial level and that this variability can be attributed to the
predominant farm models in the region and general soil fertility. According to this and
other World Bank studies, transmigants produce approximately 2 million tons of unmilled
rice, which is equivalent to 17% of outer island production and 5% of Indonesia’s total rice
production (World Bank, 1988).

This analysis should be interpreted with care. Oberai (1988) found that surveys
conducted on older upland settlements showed low and highly variable crop yields. In one
survey, about 30% of all farmers obtained 500 kg or less of unmilled rice per year. It was
found that these farmers had turned from rice to other subsistence crops, with their primary
source of cash coming from off-farm employment (Oberai, 1988). The World Bank has also
found that in some areas, farmers rely on off-farm employment to supplement their earning
from agiculture; often supplementing 40-80% of their household income.

When asked about their perceptions of their current rice production relative to that
of two years ago, 40% of all settlers interviewed reported that their production was lower
than two years ago, while approximately the same proportion cited increased production
(World Bank, 1988). On older tidal sites, 62% of the tidal farmers reported production
increases over two years ago, while 40% reported production decreases (World Bank, 1988).

 

41 For example, Hardjono (1977) found that nnigants entering tidal projects require
several seasons to adjust to their new environment. They must learn tidal cultivation and
often face a lack of fresh water in the dry season, malaria, cholera, and crop destruction by
wild animals. In addition, the long-duration variety used by these farmers produces only one
crop per year, meaning that they must find another source of income.

181

Table A4: Regional Contribution of Transmigration to incremental Rice Production by Region in Indonesia, 19m-
1934.

 

1984 Regional incremental Transmigrant contribution

production regional .
(000) pm. (000) (An' re)” (96) 1

7,15 1,627 4
22

 

 

 

3,030 2

 

24.4

 

 

 

 

 

 

 

 

 

 

'Assumes80%ofalifamfliupmduceng/familyofficemsmkg/familyavengeforautnnsmigann.
bAasumcsratioofonespontaneousfamily/sponsoredfamilywithriceproductionatr'atcsin(a)abovnre.

Source: Wodd Bunk. W (19% 8+5)-

The World Bank (1988) reports that in some tidal areas, reasonable production has
been maintained for 20 years or more without the use of fertilizers. However, it has
recently been found that new sites are experiencing serious agicultural problems related to
site location (deep peat or acid soils), planning agd pest infestation. For example, 56% of
the farmers interviewed on older upland sites4 reported decreases in rice production,
while only 3 1% showed production increases (World Bank, 1988). These inconclusive results
point to the need for establishing "long-term monitoring of yields and production on
representative upland and tidal sites“ (World Bank, 1988: 31). This issue will be addressed
later in section 2.3.6.

9. Distribution of Incomes

Table A5 presents the results of an income survey carried out by the Central Bureau
of Statistics (BPS) and the MOT in 1985, covering over 2,200 households in both successful
and critical sites (World Bank, 1988). Included in this table are comparative data from a
1984 National Socioeconomic Survey (Susenas) of 17,000 households in rural Java and 2,800
rural households in the receiving provinces (World Bank, 1988).

 

42 This survey information, compiled by the Central Bureau of Statistics (BPS) in 1985,
covers those migants who arrived at their sites before 1982.

182

1 Table A5: Comparison of Transnnigant, Javanese and Outer Island
Incomes in Indonesia, 1984-1985.

     
   

 

       
       
         
           
   

7 Areaa Survey Monthly household Annual household
, date income - Rp (USS) income - Rp (USS)

 

‘ Transmigation
sites 1985 58,300 (53) 699,600 (636)

! Rural sending
areas 1984 67,200 (62) 806,300 (733)

9 Rural receiving
‘ areas

 

 

 
  

   

   

 

 

 

1984 90,750 (83) 1,089,000 (990)

 

a Data on rural sending areas include individual smallholders, tenants and landowners.

    

Source: World Bank, 1 . on is °° ' - w as

'v (1988: xxv).

When compared to incomes in the rural areas from which they came (sending provinces)
and incomes in the rural areas to which they rnigated (receiving provinces), tine
transmigants’ incomes are lower. If the survey data are corrected for possible
underreporting of subsistence production, gigant incomes become roughly the equivalent
of Javanese incomes (World Bank, 1988) . In addition, transmigants who had been
settled in upland communities for at least five years had average monthly incomes of Rp
75,000 per household, which are slightly higher than those of the Javanese surveyed,
including landowners (World Bank, 1988).

Although many transmigants earn lower incomes than the average rural Javanese
farmer, the migants’ living conditions have been improved through transnnigation. All
nnigants own land, a house and some own livestock (World Bank, 1988). In a survey of
migants’ perceptions, two-thirds said their incomes had increased since transmigation, 17%
said their incomes were relatively unchanged, and 16% said their incomes had decreased
(World Bank, 1988).

It is evident that incomes vary by both farm model and by period of settlement. Data
show that farmers on older upland sites have low agicultural incomes, attributable to
declining soil fertility, which they supplemented with off-farm income. Recent settlers in
swamp reclamation areas have the lowest total incomes, while farmers with tree crops and
those on older tidal sites have the highest agicultural incomes. The proportion of total
household income derived from off-farm employment ranges from 40-80% (again depending

 

43 Only income from major commodities is reported in the BPS survey which excludes
income from home garden production. The World Bank (1988) has readjusted the BPS
expenditure data by Rp 1,000-2,000/capita/month (Rp 5,000-10,000/household/month) in
order to allow some comparability among populations outside the survey.

183

on farm model and period of settlement), indicating that some transnnigants find off-farm
employment more remunerative than work in agiculture (World Bank, 1988).

10.

World Bank (1988) surveys reveal additional information on the position of rrnigants,
relative to rural Javanese. Their data show that migants are able to meet their subsistence
needs from home production, and in addition, spend less on nonfood expenditures
(fuelwood, household items, durable goods, taxes, medical expenses and schooling) than
Javanese residents.

Table A6 presents rrnigants’ perceptions of their welfare, before and after
transmigation. When nnigants were asked whether their incomes had improved, decreased
or remained about the same, the majority of Repelita II and III nnigants replied that their
incomes had improved since transmigation. Compared to two years ago, 57% of the
Repelita II and 51% of the Repelita III nnigants felt that their incomes had improved.
With respect to transportation facilities, 63% of Repelita II and 69% of Repelita III
migants felt that transportation was worse in the transnnigation area than in Java. Few
nnigants felt that their health had declined in the transmigation area, but 44% of Repelita
II and 51% of Repelita III migants felt that their health status had not changed. For the
most part, an approximately equal proportion (except in the case of Repelita II tidal
farmers) felt that their health had improved since transmigation (World Bank, 1988).

184

1 TableA.6: MigantPerceptiomofWelfarebminsmigrafionSiteinIndonesia,l985
(PercentageofRespondents).

 

Currentincomevaprc— Currentincomevs. Transportatsitevs. Healtbatsitevain
migrationincome twoyearaago inJava Java

 

BetterWoaeSameBetterWorseSameBetterWoneSameBetterV/one Same

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

185

11. Employment Generation

During the Repelita III progam (1979-84) between 500,000-600,000 jobs were created
on both rainfed food crop schemes and on tree crop settlements (World Bank, 1988). This
translates to 1.3-1.6 jobs created per transnnigant household, or 12-15% of the incremental
labor force in Java and Bali (World Bank, 1988). There is, however, no data on the amount
of indirect employment created by the transmigation progam in services such as
transportation, marketing, social services and the input supplies to transnnigants.
Temporary employment, also immeasurable, has been created through labor-intensive rural
works projects (such as land clearing and the construction of roads and houses) and through
the development of tree crop schemes (World Bank, 1988).

The cost of employment creation for sponsored nnigants is much less than in the
industrial sector ($10,000-20,000 per job), yet geater than in service industries in Indonesia
(World Bank, 1988). For example, on upland schemes, the cost is approximately $3,300-
4,100 per permanent job. On tidal schemes, the cost increases to $4,500-5,500. On more
remote sites, the cost per job is believed to increase by 25-50%, according to the farm model
(World Bank, 1988). It should be noted that there is no information on the cost of
employment creation for spontaneous nnigants, who move at their own cost.

Babcock (1986) reports that with the exception of one or two cases, transmigation
progams have not generated important secondary industries. He notes, however, that ”in
the more prosperous project areas new market centers do gow up where commonly the
entrepreneurs, for example, the traders, the transport sector, are from non-transmigant
ethnic goups" (1986: 177). While there are some employment opportunities for low-level
professionals in the transmigation areas, their employment in this sector implies that
another rural area will be short-staffed (Babcock, 1986).

12. Regional Development

Transmigation has had a variable impact in the receiving areas on agicultural
production, infrastructure and service sector development and population gowth. World
Bank (1988) estimates indicate that transnnigants settled during Repelita III produce 2%
of the total outer island rice production and those settled since 1950 produce 33% of the
outer islands’ total rice production, or about 5% of Indonesia’s total production.

Oberai (1988) notes that transmigation has had a marginal impact in integating the
regions into the national economy - in terms of developing industry and trade, exploiting
and processing natural resources and improving transport and communications. He adds,
however, that the conservation and rational exploitation of Indonesia’s resources is becoming
a geat concern. The World Bank (1988) states that although transnnigation has reduced
demogaphic pressures in Java’s more critical areas and pernnitted the reforestation of
overcultivated lands, settlement in the outer islands has reduced the amount of forested land
and placed pressure on conservation areas and wildlife habitats. For example, the World
Bank found that during Repelita 111 between 30-50% of all land cleared for sponsored
nnigants had been previously forested. The relative proportion of forested land cleared for
transmigation sites differs depending upon the province, and ranges from 0.2% of the total
land area in Sulawesi to 4% in Sumatra (World Bank, 1988). The degee to which forested
land has been cleared by spontaneous nnigants is unknown (World Bank, 1988).

One of the clearest indicators of regional development in Indonesia’s transmigation
progam is the improvement of infrastructure, tlnrough the construction of both access and
main roads. In some areas infrastructure in the regions has doubled (World Bank, 1988).
A drawback to this development, however, is the burden it places on the districts when fiscal

186

responsibility for infrastructure maintenance is transferred to the provincial level. As a
result, many roads fall into disrepair once national government financing is terminated
(World Bank, 1988). The transnnigation progam has also led to the expansion of other
services in outer islands, especially extension, agicultural input supply, education and health
services. There is evidence that those towns, which depend primarily on central government
budgets to finance their development expenditures, have difficulties accommodating town
and urban gowth under the current planning arrangements (World Bank, 1988).

Babcock (1986) estimates that the Repelita I and II transnnigants made up 1.4% of the
outer island population in 1980, and accounted for 5.8% of the total provincial population
gowth in the settlement provinces. These figures, however, do not include spontaneous
nnigants whose contribution to population gowth was not recorded until 1983 (Babcock,
1986). During the Repelita III period, the proportion of sponsored transnnigants to
indigenous residents gew to as much as 40% of the 1980 population of some districts
(Babcock, 1986; World Bank, 1988).

13. Monitoring and Evaluation of Transmigration Projects

Under the Repelita III progam, the World Bank and the United Nations Development
Progamme supported a monitoring project which included an early warning system to allow
detection and rapid response to scheme-level problems (World Bank, 1988). Unfortunately,
GOI resource constraints forced the progam to be abandoned. Field-level problems also
hindered the success of the monitoring progam. For example, Ministry of Transnnigation
staff were charged with collecting too much information, some of which was not readily
available to them (World Bank, 1988). The monitoring and evaluation aspect of the
transnnigation progam is one of the most important components yet remains one of the
weakest ones.

14. Economic and Financial Analyses of Transmigation Projects

The World Bank carried out financial and economic analyses of nine transmigation farm
models (1988). They analyzed both upland and tidal sites over a thirty year period,
beginning in early 1980. The typical farm budget included benefits realized by the farmer
from all crop and livestock production (large and small animal) and from off-farm income
earned. The following costs were quantified in the analysis: the opportunity cost of labor
for settlers and the costs of site selection and preparation, settler selection, recruitment and
subsistence, input packages and project administration. The opportunity cost of capital used
for these analyses was 10%.

For all food-crop models, the rates of return were lower than prior appraisals had
indicated. For example, the rates of return from agiculture (excluding off-farm income)
were low, falling between negative values for the most prev ent low-input models to 4% for
upland sites with sufficient market access and input use . Upon adding the benefits
from off-farm income, the economic rates of return increased to an average of 3-6% for the
low-input/low-output upland and tidal sites. Only the more costly tree crop models (oil
palm, coconut and rubber plantations) had economic rates of return geater than the
opportunity cost of capital. The respective rates of returns to oil palm and rubber tree

 

44 About 75% of the Repelita settlements are located in upland areas, and few schemes
are irrigated.

187

schemes were 13% and the returns to coconut schemes were estimated at 11% (World

Bank, 1988).

188
Land Resettlement: The Case of Kenya

1. Agrarian Structure and the Origins of Land Resettlement in Kenya

Land resettlement in Kenya was a political response to population pressure in black
farming areas and to smalllnolders’ desire for land from which they had traditionally been
excluded (House and Killick, 1983). The colonial system of landownership had established
separate status and rights for Europeans, Africans and Asians through the allocation of land.
European settlers maintained exclusive freehold ownership in the White Highlands, the
highest potential land. A census executed prior to any land redistribution indicated that
over 90% of the European farms were holding of 400 hectares or geater and 50% of these
were holding of at least 2000 hectares (Hazlewood, 1985).

The reserves (or non-scheduled areas) were communally farmed land units allocated to
various ethnic goups, comprising 24% of Kenya’s total land area, or 74% of all arable land
(Leo, 1989). Six million Africans cultivated small plots of variable quality under a
communal tenure system. Squatters were allowed to reside on European farms in return
for their labor on large-scale farms in the highlands (Leo, 1989).

Consequently, Kenya’s dualistic land ownership system resulted in severe inequities in
resource distribution for millions of smallholders. Political instability developed, stifling
foreign investment in the country, and motivating the Kenyan government to develop several
land transfer progams.

TIA? colonial government irnitiated a land reform in 1954, according to the Swynnerton
Plan, that promoted freehold tenure for all Kenyans. It did not, however, address the
issue of land hunger among Africans, a major cause of political instability, which
subsequently impeded foreign investment in Kenya. The government thought that the
transfer of land from large European landholders to African farmers would create geater
political stability and confidence in the Kenyan government with, it was hoped, no negative
impact on agicultural productivity (von Haugwitz, 1972).

As Hazlewood notes, "the brief history of land transfer is complex, because schemes of
different types were introduced to meet the political and economic pressures of the time“
(1985: 445). The Million Acre progam, established in 1962, was the most important land
transfer initiative in terms of the amount of land redistributed and its impact on
smallholders. Several irrigation schemes were also created - the Perkerra scheme (1953),
the Mwea scheme (1954), the Ahero and Bunyala schemes (1968) and the Tana scheme
(1969) - but these schemes were smaller and much more management and capital intensive
than the Million Acre progam. In addition, study of the Million Acre progam offers the
most relevant insights regarding the implementation and evaluation of Zimbabwe’s rainfed
resettlement schemes.

The Million-Acre progam involved the purchase of over 400,000 hectares of European
farmland adjacent to African or non-scheduled areas (at a rate of 81,000 hectares per year
for five years). The government then redistributed this land under individual title to African
farmers in several different types of settlements; high and low-density settlements, and
cooperatively or individually owned large-scale farms (House and Killick, 1983).

 

45 The major components of the Swynnerton Plan were '1) a major progamme of land
reform, 2) increased availability of credit, and 3) a reorientation of research, extension and
marketing bodies” (House and Killick, 1983: 44-45).

189

Land transfer continued until 1971, but after 1965 the primary focus of the progam
shifted towards maintenance of large-scale farming units (Hazlewood, 1985). Leo notes that
throughout the 1970s and into the 1980s, “the policies that were followed reflected above all
a continued intense pressure, both from landless people and from the bourgeoisie, for the
multiplication of smallholding. Indeed they were less policies than a series of expedients,
each one designed under pressure to come to terms with a particular set of circumstances”
(1989: 179). Consequently, during this time the government developed two Harambee
settlements for carefully selected settlers to farm smaller sized low-density plots. Following
intense pressure from the landless, squatters were eventually given titles and settled on
abandoned European farms (known as Haraka schemes). By 1982, these settlements
together covered over 670,000 hectares and benefitted approximately 64,000 families (Leo,

1989). Although the number of beneficiaries appears large, it represented only 3% of the
population.

2. Goals and Objectives of Resettlement

The primary objective of resettlement was to induce political and social stability by
reducing inequality in land ownership between European and African landholders and
creating geater employment opportunities for Africans. In addition, the government
wanted to produce a structural change in agiculture by substituting smallholder farming for
large-scale farming, without causing a decrease in agicultural production (Hazlewood, 1985)
and a predetermined target income for settlers (von Haugwitz, 1972). Objectives
concerning improvements in social welfare were not explicitly stated and social welfare
concerns were addressed only through land redistribution which permitted increased access
to resources for smallholders.

Implementation of Kenya’s Resettlement Program

3. Land Selection

Commercial farmers sold land to the Land Development and Settlement Board on a
willing buyer, willing seller basis. The government valued land at 1959 prices (the last year
in which a significant number of land market transactions occurred) and paid sellers a
proportion in cash and the balance in annual installments backed by a 5% promissory note.
The purchase price allowed for a 12.5% return on the capital invested, as well as for the
value of permanent improvements on the land which contributed to a profitable farnning
activity (von Haugwitz, 1972).

Geogaphy, as well as politics, influenced land selection. Ethnic tensions necessitated
that certain tribes be separated from others by boundaries, and thus, settlements were
planned according to ethnic lines. Leo concludes that the settlement progam "not only
reaffirmed and hardened existing ethnic boundaries, but in fact reintroduced ethnic
uniformity in areas where mixing had already taken place spontaneously" (1989: 111).

Significant funding for land purchases came from the United Kingdom, the
Commonwealth Development Corporation and the World Bank. In addition, the Kenyan
Government received loans for a proportion of the land purchase and for settler credit.
Given the high level of external funding, the government felt compelled to make the
settlement progam pay for itself, at least irn part. This led the government to accelerate the
rate at which settlers repaid their land purchase and agicultural input loans, contributing
to low repayment rates (Hazlewood, 1985).

190

4. Settler Selection

The criteria for settler selection differed for the high and low-density schemes.
Committees interviewed applicants for high-density schemes, usually choosing those who
were unemployed and landless but had some agicultural knowledge. Priority was often
given to the previous employees of a farm whose land was to be redistributed. Both types
of settlers were required to deposit some working capital and pay a stamp duty and legal
fees at registration (von Haugwitz, 1972).

Smallholders on the settlement schemes received freehold tenure status contingent upon
repayment of their land and development loans, and compliance with recommended farnning
procedures and government regulations (Harbeson, 1984). Settlement regulations stipulated
that farmers could not subdivide, transfer, or improve their plots without the government’s
consent (Harbeson, 1984). Settlers were also not allowed to retain land in their former
areas but many farmers used the schemes as a way of increasing their total land holding
(Hazlewood, 1985).

5. Scheme Planning and Management

Resettlement planners developed farm budgets for each soil type found on a scheme
(usually two), although many schemes with a common income target had three different
model budgets for various farming systems, plot sizes and loan schedules. De Wilde (1967)
found that farm plans often paid insufficient attention to variations in soil fertility,
t0pogaphy and the scheme’s micro-climate. In addition, he found that farm plans were
sometimes developed for crop or livestock systems subsequently not viable for the settler.

During the initial planning period, the government encouraged farmers to form
marketing cooperatives since there were often no private traders to act as marketing
intermediaries (von Haugwitz, 1972). The ethnic and geogaphical origin of settlers
hampered scheme-level cooperation. Although settlers were often from the same tribe, they
were from different communities which made cooperative efi'orts difficult (Clayton, 1978).
Therefore, economic incentives played a significant role in motivating cooperative
participation. When settlers first joined a scheme, they were more likely to participate in
cooperative activities. As they gained more experience, cooperative participation decreased
as they marketed their production outside of the cooperative to obtain higher prices
(Clayton, 197 8).

The Million Acre program contained two types of individual smallholder schemes. High
and low-density schemes were develOped to meet the demand for land by both skilled and
unskilled agicultural workers. The government created 122 high-density settlements for the
landless and the unemployed, with average holding of 10 hectares. Each scheme held 3000-
4000 settlers on 4000 hectares of land who were targeted to achieve a subsistence standard
of living and L25-75 cash income, in addition to repaying any loan obtained for land
purchase and for other inputs.

Approximately 40 low-density schemes, located on higher quality land, were allocated to
farmers with more agicultural experience. About 100 settlers farmed each 2000 hectare
scheme, whose individual holding averaged 13 hectares. Each plotholder was to achieve
subsistence and a L100 cash income after repaying their loans. Although the government

Wanted to locate both types of schemes near the overpopulated African Areas in order to
beduce population pressure in these areas, this was not always possible. As Leo explains,
European farmers “were not prepared to sell high-potential, underdeveloped land at prices
that reflected its present worth rather than its ultimate potential, and that those who wanted

191

to do so were able to find buyers at better prices [than those offered by the Land
Development and Settlement Board]" (1989: 79).

The Department of Settlement theoretically transferred scheme-level management from
Department personnel to the scheme cooperative in a four-stage process. Initially, the
Department of Settlement would operate the scheme in consultation with the cooperative.
The scheme would reach the second stage of Settlement staff withdrawal from daily
operations after 5 years, with the stafi retaining responsibility for loans, extension services
and all legal aspects relating to the scheme. At the third stage, the cooperative gouped
individual schemes into complexes of 5-20 schemes to ensure geater control of cooperative
activities. The fourth stage, full scheme maturity, occurred after repayment of all settler
loans; approximately 30 years after initial settlement, thus integating the scheme into the
smallholder farm sector (Clayton, 1978). This phasing was not adopted in many cases
because the Department of Settlement preferred to retain more control over the schemes
to ensure that they marketed a desired amount of surplus production (Clayton, 1978).

6. Service Provision

The government planned to provide roads, soil conservation works, water supplies and
trading centers for each scheme (de Wilde, 1967). As de Wilde notes, however, there were
often inadequate water supplies for livestock and household use at the time of initial
settlement. Furthermore, although settlers could obtain seed, fertilizer, livestock, fencing
and other inputs on credit, Hazlewood (1985) attributes shortages of some inputs to many
settlers’ inability to achieve target incomes.

Each scheme was advised by virtually the same number of staff, regardless of scheme size
and in spite of the fact that high-density schemes were significantly more populated than
low-density schemes. This is significant for the provision of extension services because the
high-density schemes contained large numbers of landless settlers with limited agicultural
experience, who were most likely in need of more intensive extension services; at least
during the first years of settlement (Leo, 1989). '

Assessment of Resettlement in Kenya

7. Post-Reform Agarian Structure
Since resettlement, most of the organizational changes in agiculture have occurred in
the small farm sector, leaving the large-scale farming sector basically intact. Most of the
plantations and ranches remained under the control of non-Africans since these farms were
not easily subdivided and therefore had to be purchased as complete units, at higher prices
(Senga, 1976). The large farm sector, particularly in the middle and lower size gouping
(200-2000 hectares) decreased in size (based on data from 1960 to 1973), while the number
of farms smaller than 100 hectares increased (Senga, 1976). The increase in the number of
certain sizes of large farms is due to the repurchase or lease of smallholder plots by larger
farmers after redistribution (Cohen, 1980). There is no policy limiting or prohibiting land
concentration, even though the data point to greater efficiency in smallholder agiculture.
In sum, of a total of 50 million hectares of agicultural land, the government purchased
L .7 million hectares of land (roughly 3%). As of 1978, 800,000 hectares of this area had

192

been transferred to smallholders, 860,000 hectares transferred as larger holding,‘46 with
100,000 hectares remaining under non-Kenyan ownership (House and Killick, 1983).

, Therefore land ownership in Kenya remained skewed after the reform. For example,
one third of all agicultural land is still held in large farms, 60% of which are holding over
100 hectares (Cohen, 1980). Within the smallholder sector, 52% of the smallholders own
less than 15 % of the total agicultural land, in holding that average two hectares or less
(Cohen, 1980). In addition, as of 1980 there were still an estimated 300,000 landless, or
19% of Kenya’s 1.7 million rural households, who could not afford to purchase land since
the emergence of a land marliet and title security had caused prices to increase (Cohen,
1980; Collier and Lal, 1986).4

8. Agricultural Production and Income Distribution

Several studies reached similar conclusions regarding gowth in output and income for
settlement farmers in Kenya. Von Haugwitz (1972) found that for the period 1964 to 1967,
farms at high altitudes without permanent crops achieved higher levels of output and net
farm profits than low altitude farms; even though the low altitude farms often had more
available hired and family labor, as well as more credit and capital invested per acre. The
rate of increase in output and net profit was higher on the high-density farms than on the
low-density farms. When he compared high and low-density farms over the whole progann,
von Haugwitz found neither farm type was superior because at low altitudes the low-density
farms performed better but found the inverse at higln altitudes.

Leo (1989), however, cites data from the Government of Kenya’s Economic Appraisal
of the Settlement Schemes (1964-68) showing that output and farm profits per acre
increased more rapidly on high than on low-density schemes. Clayton (1978) points out that
much of this rapid increase in farm output can be attributed to an expansion in area
cultivated.

Data on returns to capital invested on the settlement farms also indicate that high-
density farms out-performed low-density farms. From this data it appears that farmers
achieved higher cash surpluses (or in the early years, lower cash deficits) on the high-density
schemes, although overall a geater percentage of low-density farms achieved positive cash
surpluses (von Haugwitz, 1972). This is significant because the government originally
designed the high-density farms to provide much lower target incomes than the lowdensity
farms. However, many farmers did not reach the targeted income level for their farm type
(Clayton, 1978; Leo, 1989). In fact, during the period 1964 to 1968, the geatest proportion
of farms attaining their target income was 69% for the lowest target income goup (Leo,
1989).

Hunt (1984) provides comparative data for settlement schemes (1967 / 68) and for large
farms (1970/71). Her study shows that in both sectors, the smallest size of farms achieved
significantly higher output per acre than larger sized farms, which supports the proposition

 

46 It appears that by 1984, half of these large farms would be subdivided into
smallholdings, leaving approximately 430,000 hectares still under individual ownership
(House and Killick, 1983).

47 In fact, a study by Migot-Adholla reveals that settlement officers encouraged
absentee urbancdwellers with sufficient income to purchase land from smallholders who
defaulted on their loan payments.

193

that la r productivity and output/ area unit are geater on smaller farms that on larger
farms. This was attributed to a higher cropping intensity and a more intensive use of
grazing land on the smaller farms. The data also revealed that as farm size increased labor
inputs per acre decreased.

Data from the traditional small farm sector indicated some of the same relationships
found in the settlement and large farm sectors (Hunt, 1984), including an inverse
relationship between farm size and output. While the average farm operating surplus per
hectare was geater for the smallest size farms (particularly those under 0.5 hectare, or a
little over 1 acre), income per household member increased with farm size, partly because
household and farm size are positively correlated. The fact that smaller farms achieved
higher returns may be explained by more intensive labor use and management (Hunt, 1984).
These data point to a geater efficiency in resource use on smaller farms (higher output and
sales per hectare, higher use of labor per unit of output and lower use of capital and foreign
exchange-intensive resources per unit of output) implying that both efficiency and equity can
be achieved through land redistribution (Hunt, 1984).

9. Employment Generation

Farming patterns on the settlement schemes changed very little from those practiced by
the large-scale farmers who formerly owned the land. The settlement farmers, however,
employed more labor intensive methods of production instead of the highly mechanized
methods used by the large-scale farmers. As a result, although wage employment decreased
as the large-scale farms were broken into smallholdings (Maitha, 1976), redistribution
created geater employment opportunities for smallholders and their families. Surveys of
settlement farms from 1964 to 1967 indicate that an average of 319 adults were employed
per 1,000 hectares on high and low-density sc emes, a number much larger than those
employed by large-scale farms (Clayton, 1971). When both family and hired labor were
included, Clayton found that about 568 workers were employed per 1,000 hectares on
rainfed settlement schemes.

In comparison to employment generated on other types of settlement schemes and farms,
rainfed smallholder settlement employed more workers than the large farms owned by
Africans, while the more capital-intensive irrigation schemes employed nine times as many
workers, or approximately 2,807 persons per 1,000 hectares (Clayton, 1971). However,
settlement alone was unable to generate sufficient employment during this period (1964-67)
since employment increased by 2.8% annually while the labor force gew by 3.5%.

10. Settler Welfare

 

48 Smallest size category refers to those farms under 20 acres (8 hectares) in the
settlement sector and under 250 acres (100 hectares) in the large farm sector. It should be
noted that many settlement plots are from 8 to 24 times larger than the typical Kenyan
smallholding of 0.5 to 1.5 hectares (Hunt, 1984).

49 Although her calculations for labor inputs are based on a worker-equivalent per
1,000 acres, Hunt’s (1984) data from 1967/ 68 show the same pattern of employment
eneration. For example, for settlement farms under 10 acres, a total of 808 workers are
employed (781 family laborers and 27 hired workers). Using the same worker-equivalent,
large-scale farms under 250 acres employed a total of 93 workers.

194

It is virtually impossible to quantify the effects of Kenya’s land redistribution on the
poverty and inequality under which a large percentage of the rural population lived.
However, of the 71,000 families affected, not all were given equal access to productive
opportunities in agiculture. House and Killick (1983) argue that by providing land to
numbers of landless and poor families, resettlement reduced poverty and land hunger,
encouraged increased smallholder production and generated higher incomes in the highland
areas.

House and Killick (1983) also found that racial inequality between European settlers and
Africans was most likely reduced by the transfer of resources, even though tlnis also
permitted a small landholding class of wealthy Africans to emerge, further differentiating
the African population. Furthermore, those who still farmed in the traditional smallholder
areas and the resettlement farmers felt that the resettled farmers benefitted from improved
access to services and resources, which created additional perceptions of inequality.

The government did not always accord the same opportunities to the landless as to
landed individuals. According to Leo, ”settlement involved an accommodation of landless
people, but the accommodation was not designed to offer them a serious opportunity for
prosperity, even at a peasant level” (1989: 120). Consequently, in order to stifle any
continuation of the Mau Mau rebellion, the landless were dispersed through resettlement
to assert control over them and partly assuage their land hunger. They did not, however,
receive the high quality land allocated to the more prosperous peasants on the low-density
schemes (Leo, 1989). Hazlewood (1985) also found that the need to "make settlement pay"
influenced the selection of settlers to the point that many landless did not benefit from
resettlement because they showed less ability to repay loans for their land.

Leo (1989) illustrates that throughout the planning and implementation phases of
Kenya’s resettlement progam, the high-density settlements were established to meet
political objectives, while the low-density settlements were intended to be more viable for
smallholder agiculture. Even so, there are data to support the thesis that many settlers on
high-density schemes achieved cash surpluses and increases in agicultural output in spite
of the biases against them.

11. Credit and Debt Repayment Capability of Settlers

Low repayment rates and high indebtedness among settlers on the resettlement schemes
can be attributed, for the most part, to oversights in planning. First, the budgets developed
for the schemes did not provide sufficient income to meet a family’s cash expenditures (von
Haugwitz, 1972). Second, it was planned that a farm would achieve full production in four
years. However, the first loan installments were due in full at the end of the first six months
and settlers often had difficulties generating sufficient income to begin payment on these
installments (von Haugwitz, 1972).

Third, there was little allowance in the farm budgets for crop failures due to drought,
payment delays, heavy plowing charges levied by private firms or the obligation of
cooperative members to purchase any capital equipment left on a farm (Harbeson, 1984).
Many farmers accumulated geater debt in drought years and could not make their loan
payments which then accrued an annual interest charge of 6.5% (von Haugwitz, 1972). As
a result, the high repayment installments and interest on subsequent defaults reduced
settlers’ ability to make improvements to their farming operations (von Haugwitz, 1972).

De Wilde (1967) found that repayment rates were lower among the high-density farmers
than low-density farmers. This is consistent with data on farms achieving positive cash
surpluses. In addition, settlers on the low-density schemes were generally better managers

195

and farmers who started with more working capital and were thus capable of producing a
higher yield per acre (von Haugwitz, 1972). A geater number of low-density farmers
achieved cash surpluses than high-density farmers, yet the proportion of low-density farms
was consistently less than 50% (Leo, 1989), roughly equivalent to the total percentage of
loans repaid (von Haugwitz, 1972).

Furthermore, there were different cultural perceptions among the tribes regarding their
obligation to repay debt on land that formerly belonged to European settlers. Many of the
new settlers still felt that the redistributed land belonged to them traditionally, leaving them
with little incentive to make the loan repayments (de Wilde, 1967; Leo, 1989).

Heyer (1976) cites that low repayment rates on settler loans was the primary reason why
the settlement schemes were not successfully integated into the national progam for
smallholder development. The large expenditures on resettlement were eventually criticized,
leading to staffing cutbacks on many schemes that often left them understaffed compared
to the other smallholder areas (Heyer, 1976).

12. Economic Growth

Land resettlement in Kenya was achieved without any decline in national agricultural
production (Collier and Lal, 1986). During the years when most resettlement was carried
out, 1964 to 1972, goss domestic product increased annually by 6.5 %. Marketed
agicultural output increased by 6.7 % annually, while agicultural subsistence output gew
by 3.7% (population gowth, however, reached almost 4% annually), and manufacturing
increased by 12% per annum (Hunt, 1984). However, after 1973, the Kenyan economy
began to show signs of instability and the average gowth rate began to fluctuate at lower
levels (Hunt, 1984).

APPENDIX B

AGRICULTURE SECTOR DATA FOR ZIMBABWE

r_

AGRICULTURE SECTOR DATA FOR ZIMBABWE

APPENDIX B

1‘ Table 8.1: Structural Change in Zimbabwe’s Large-Scale Commercial Farm Sector, 1979-

l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

- 1988.

‘

Farm 1979 1988 Net Change 1

, Size No. farms Area (ha) No. farms Area (ha) No. farms Area (ha) i

< 200 ha 1,324 104,674 947 75,136 -377 49,538

; (21.7)a (0.7) (20.3) (0.7) (8.0) (03)

" 200499 445 125,805 367 104,000 -78 -21,805

x (73) (0.8) (7.9) (0.9) (1.7) (02)
400-599 446 220,176 392 192,344 -54 47,832

1 (73) (15) (8.4) (1.7) (12) (02)

600-799 425 294,387 356 246,837 -69 47,550

‘1 (7-0) (20) (7e) (22) (15) (0-4)

: 800-999 42; 374,916 348 309,067 -75 -65,849
(6-9) (15) (75) (28) (1.6) (0-6)

f 1000-1999 1,372 1,931,189 1,063 1,491,043 .309 44,146

3 (214) (12-8) (223) (133) (6-6) (3-9)

. 2000-3999 905 2,503,591 673 1,853,162 432 650,429

(14.8) (16.6) (14.4) (16.5) (5.0) (58)

; 4000-5999 289 1,425,735 182 893,705 .107 530,030

. (4-7) (95) (3-9) (8-0) (23) (4-7)

; 6000-7999 142 974,229 99 687,039 .43 -287,190
(23) (6.5) (21) (6.1) (0.9) (2.6)

l > 8000 342 7,111,514 233 5,361,053 -109 -1,750,461 3

i, (5.6) (472) (5.0) (47.8) (23) (39.9)
Total 6,113 15,064,216 4,660 11,213,386 4,453 -3,850,830

L__ (100.0) (100.0) (100.0)”10) f f_ (.6)

a Figures in parentheses are percentages of column totals for numbers of farms and land area.

Source: adapted from M. Roth, “Analysis of Agarian Structure and Land Use Patterns in Zimbabwe,“
(1990: 29).

196

197

Table 3.2: Growth Rates of National Crop Production in Zimbabwe (%), 1973-1988.

 

Area Growth Rates Production Growth Rates Yield Growth Rates
1973—80 1979-88 1973—80 1979—88 1973-80 1979-88

 

0.8 0.2 7.1 -1.1 63 -1.3
-4.3 -5.5 -4.0 -5.3 0.3 0.3
-5.0 -3.7 4.2 -1.9 9.7 1.9
6.9 0.9 2.1 3.1 1.2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soyabcans
. Sunflowers
Groundnuts

 

 

 

 

 

 

 

 

 

 

 

‘ Dry Beans -15.1

Sources: M. Roth, “Analysis of Agarian Structure and Land Use Patterns in Zimbabwe,“ (1990: 131-33).

 

AA

 

 

 

 

A

vvv'

 

198

Table B3: Maize Production on Zimbabwe’s Large-Scale Commercial Communal Area and
‘ Resettlement Area Farms, 1978-1989.

 

Area Prodn
(000 ha) (000 1')

Large-Scale Commercial

Yield
(”r/ha)

Area
(000 ha)

Communal Area
Prodn Yield

(000 1) cr/ha)

Area

Resettlement Area
Prodn Yield

(000113) (000T) CT/ha)

 

201.8 1,114.6

5.5%

450.0

0.643

NA

 

721.9

3.786

4ND

0.700

NA

 

910.7

4.000

600.0

0.667

NA

 

1,736.0

5.889

IMHO

11!!)

NA

 

1,143.6

4.229

NA

NA
NA
NA
NA
NA

 

601.8

285.0

5.92

 

694.6

NA

 

1,112.5

 

1,197.6

 

506.6

 

682.4

 

 

 

 

 

 

 

 

 

 

 

Sources: M. Roth, “Analysis of Agrarian Structure and Land Use Patterns in Zimbabwe,“ (1990: 152);
Cusworth and Walker, 'Land Resettlement in Zimbabwe, A Preliminary Evaluation,“ (1988: 67-69); KH.
Wekwete, 'A Review of the Rural Land Resettlement Programme in Post-independent Zimbabwe,"
(forthcoming: 22); CSO Crop Forecasting Committee.

199
Table 3.4: Sorghum Production on Zimbabwe’s Large-Scale Commercial, Communal Area and
‘ Resettlement Area Farms, 1978-1989.

 

Large-Scale Commercial Communal Area Resettlement Area 1
Area Prodn Yield Area Prodn Yield Area Prodn Yield 3

(000113) (0001‘) (T/ha) (000113) (0001') (T/ha) (000118) (0001) (T/ha)
as 15.6 2.29 57 0.48

7.5 18.9 2.52 30 0.39
6.8 16.3 2.40 66
8.4 2.89
7.1 . 2.38
5.1
8.8
13.9
22.0
5.7

 

 

 

 

 

NA
NA
NA
NA

 

 

65
1.

 

 

 

 

 

 

 

 

 

 

 

 

 

E§§§§§§§§§§E

 

 

Source: M. Roth, “Analysis of Agrarian Structure and Land Use Patterns in Zimbabwe," (1990: 152);
KB. Wekwete, ”A Review of the Rural Land Resettlement Programme in Post-independent Zimbabwe,”

(forthcoming: 22); C80, WW (1989: 177); CSO Crop Forecasting Committee.

0|.

v._ r_ ...._ ...._ r_ r. I. 1h“; 1~UW= S

 

 

Table BS: Flue and Burley Tobacco Production on Zimbabwe’s Large-Scale Commercial, Communal
Area and Resettlement Area Farms, 1978-1988.

 

Large-Scale Commercial Communal Area Resettlement Area
Area Prodn Yield Area Prodn Yield Area Prodn Yield

(000 ha) (000 T) (T/hA) (000 ha) (000 1') (T/ha) (000 ha) (000 1) (TIM)
55.4 83.1 1.5 169 0.44 NA

381
60.1 107.3 1.79 360 197 0.55
63.7 11.9 0.19 365 251 0.63
69.1 1.76 367 195
88.2 1.94 1,080 W4
93.2 2.03 1,41) 645
2.33 774
1,7“)
NA
NA
NA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sources: CSO, W (1989: 183); Crop Forecasting Committee (CSO).

 

201

‘ Table 8.6: Cotton Production on Zimbabwe’s Large-Scale Commercial, Communal Area and
Resettlement Area Farms, 1978-1989.

    
 

 

  
 

   

Area

(000 hin) (000 1) (T/ha)

Large-Scale Commercial

Prodn Yield

Area

Communal Area
Prodn Yield

(000 ha) (000 '1') (TIM)

Area

 

Resettlement Area
Prodn Yield

(000 ha) (000 1) cr/ha)

 

  

1978

 

86.1

1.50

41

0.76

NA

 

   

 

1979

77.0

1.69

Z)

0.75

NA

 

 

 

1980

 

74.9

1.94

15

0.80

NA

  

 

 

1981

 

58.4

2.02

59

0.76

NA

  

 

 

 

1982

52.3

2.00

51

0.53

0.79

  

 

 

1983

 

59.9

1.85

65

0.51

NA

  

 

1984

2.01

100

0.70

5.36

  

 

 

1985

70.3

130

110

0.85

9.77

  

 

 

1986

114

0.87

8.82

  

 

§§§§§§§§§§

13.50

    

 

 

 

 

 

 

 

 

 

 

 

Source: CSO, W (1989: 181); C80 Crop Forecasting Committee.

 

4 Table 8.7: Groundnut Production on Zimbabwe’s Large-Scale Commercial, Communal Area and

1 Resettlement Area Farms, 1978-1989.

202

 

 

 

 

 

 

 

 

 

 

 

 

 

Large-Scale Commercial Communal Area Resettlement Area

Area Prodn Yield Area Prodn Yield Area Prodn Yield I

Year (000118) (0001') (TIM) (000118) (0001') (T/ha) (000113) (000T) (T/ha) -
1978 13.4 12.7 0.950 200.0 100.9 0.504 NA NA NA
1979 3.2 7.5 2.324 240.0 100.0 0.417 NA NA NA
, 1980 3.8 10.7 2.779 175.0 67.0 0.383 NA NA NA

I 1981 12.9 18.8 1.456 300.0 100.0 0.333 NA NA NA I

1982 11.9 16.4 1.374 240.0 95.0 0.396 NA 0.164 NA I

E 1983 10.7 92 0.855 180.0 225 0.125 NA 0979 NA]
1984 7.0 6.2 0.883 144.0 18.7 0.130 4.2 3.403 0.806
1985 6.9 6.9 MIX) 118.0 61.0 0.512 12.3 4.8“) 0.390
1986 NA NA NA 160.0 64.0 0.400 10.2 3.060 0.300
1987 5.1 13.0 2.549 177.0 55.0 0.310 10.8 4.320 0.400
1988 5.5 16.4 2.982 197.0 106.0 0.540 NA NA NA

1989 6.0 NA NA 160.0 72.0 0.450 NA NA NA '

Sources: C80, W (1989: 179); M. Roth, “Analysis of Agrarian Structure and
Land Use Patterns in Zimbabwe,“ (1990: 152, 159); C80 Crop Forecasting Committee.

 

 

 

 

 

 

 

 

 

 

, Table B.8: Livestock Production on Zimbabwe’s Large-Seals. Commercial and
: Communal Area Farms (0008 head), 1978-1989.

     
 
 
       
  

 

large-Scale Commercial

  

 

Year Cattlea Sheep and Pigs Cattlea Sheep and Pigs
: Goats

3,077 283 113 2,950 2,366 ~ 96

2,709 235 90 2,860 1,700 NA
2,410 220 93 2,869 1,149 39

2,391
2,400
2,358
2,23 1
1,963
NA
1,847
NA

 

    
 

 

    
 

 

  
   
   
   
   
   
   
   
   
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a Includes beef and dairy cattle.

Sources: CSO, W (1987: 158); M. Roth, “Analysis of Agrarian
Structure and Land Use Patterns in Zimbabwe," (1990: 44-48)

 

 

 

 

Table:
1983/

Natur _
chio

 

 

 

 

 

It”:

 

Source:
110).

Table B_
Pf0ducfi.

Natural
RCQOD

 

204

i , _, _, i , 2 WM“!

’ Table B.9: Average Area Planted, Yield and Gross Margins for Resettlement Area Maize Production,
Q 1983/84 - 1986/86.

 

 

 

 

 

Natural 1983/84 1984/85 1985/86 *
1 Region Area Yield Gross Area Yield Gross Area Yield Gross ‘
: (ha) (ks/ha) Margin/ha (ha) (ks/ha) Margin/ha (ha) (ks/ha) Margin/ha 9
II 1.8 1868 98 1.9 2871 315 1.8 2729 197 f
, m 1.6 1092 46 20 2465 287 20 2126 211 1
4 l
» IV 23 1377 150 20 2497 329 1.9 1764 216 '
\I 1;3 123 3 115 2:5

 

 

 

   

 

 

 

 

 

 

Source: Cusworth and Walker, ”Land Resettlement in Zimbabwe, A Preliminary Evaluation,“ (1988:

110).

    

E Table 8.10: Average Area Planted, Yield and Gross Margins for Resettlement Area Sorghum
I Production, 1983/84 - 1986/86.

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

Natural 1983/84 1984/85 1985/86
Region Area Yield Gross Area Yield Gross Area Yield Gross
(ha) (ks/ha) Margin/ha (ha) (ks/ha) Margin/ha (ha) (ks/ha) Margin/ha
II 0.0 0.0 0.0 0.0 l 0.0 0.0 0.4 706 75
III 0.7 1730 208 0.4 E 677 87 0.5 608 65
' f
w | 128
V 14 r

 

 

Source: Cusworth and Walker, “Land Resettlement in Zimbabwe, A Preliminary Evaluation,“ (1988:

111).

 

 

 

IA

205

Table B.11: Average Area Planted, Yield and Gross Margins for Resettlement Area Cotton Production. |
‘ 1983/84 - 1986/86. |

 

 

 

 

 

% Natural 1983/84 1984/85 1985/86

Z Region Area Yield Gross Area Yield Gross Area Yield Gross I
1 (ha) (ks/ha) Margin/ha (ha) (ks/ha) Margin/ha (ha) (ks/ha) Margin/ha ,
l

; n 13 1304 366 2.0 1272 424 1.4 1168 610 T
i

! III 1.0 724 213 15 1230 463 1.4 1038 545 '.
I IV 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 ;
I, v 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 !

 

 

 

 

 

 

 

 

 

_ __ _ W , , i, i ,7 __ _Z «J

Source: Cusworth and Walker, "Land Resettlement in Zimbabwe, A Preliminary Evaluation,“ (1988:
112).

BIBLIOGRAPHY

BIBLIOGRAPHY

Ahmed, Zahir. 1975. W New Delhi: Orient Longman Ltd.

Akwabi-Ameyaw,Kof1. 1988. - - - ,, '
mm Dissertati9h. University of Florida-

Alberts, Thomas. 1981. c. 2 2,, .2 ’ '
Research Policy Institute, University of Lund, Number 12.

Alexander, Robert J. 1974. W939. New York: Macmillan Publishing Co.,
Inc.

 

 

Babcock, Timothy. 1986. Transmigration: The Regional Impact of a Miracle Cure.” in 99mm!
, Edited by Colin MacAndrews. Oxford: Oxford

 

University Press Pte. Ltd, 157-189. ” 1 1

Barber, William J. 1970. “Land Reform and Economic Change among African Farmers in Kenya."

 

Barrows, Richard and Michael Roth. 1989. "Land Tenure and Investment in African Agriculture: Theory
and Evidence.“ Land Tenure Center Paper No. 136. Madison, Wisconsin: LTC.

Bell, Clive. 1990. "Reforming Property Rights in Land and Tenancy.“ WM
9152119159). 143-66-

Berry, Albert R. and William R. Cline. 1979. . 2 2,.2 1
@9339. Baltimore. The Johns Hopkins University Press.

 

Binswanger Hans P. and Miranda Elgin.1990.'Reflections on Land Reform and Farm Size. in
v 1 in Second edition, edited by Carl K. Eicher, John M.
Staatz. Baltimore: Johns Hopkins University Press, 342-354.

Bratton, Michael. 1990. "Ten Years After: Land Distribution in Zimbabwe, 1980-1990. " in m
‘ - ' 2.. Edited by Roy L. Prosterman, Mary N.
Temple and Timothy M. Hanstad. Boulder: Lynne Reiner Publishers, 265-291.

 

Bruce, John W. 1986. “Land Tenure Issues in Project Design and Strategies for Agricultural
Development in Sub-Saharan Africa." Land Tenure Center Paper No. 128. Madison: LTC.

H1988 'A Perspective on Indigenous Land Tenure Systems and Land Concentration. in mm

—W Edited by RE. Downs and SP. Reyna. Hanover, New Hampshire:
University Press of New England, 23-52.

206

N7

. 1990. "Legal Issues in Land Use and Resettlement." Zimbabwe Agriculture Sector Memorandum
for The World Bank. Madison, Wisconsin: The Land Tenure Center.

 

Campbell, DJ ., L.M. Zinyama and T. Matiza. undated. "Land Policy and Access to Land in Zimbabwe:
The Dewure Resettlement Scheme." mimeograph. Harare: The University of Zimbabwe.

Carroll, Thomas F. 1961. ‘The Land Reform Issue in Latin America." in W
W. Edited by Albert 0. Hirschman. New York: Twentieth Century Fund, 161-201.

Carruthers, ID. and ES. Clayton. 1977. "Ex-Post Evaluation of Agricultural Projects: Its Implication for
“mine" We: 28(3). 305-317

Casley. Dennis J. and Krishna Kumar- 1987. WWW
Baltimore: The Johns Hopkins University Press.

2 2. . Baltimore: The

 

Johns Hopkins University Press. ‘4

Central Statistical Office. 1987. WWW. Harare: CSO.

Central Statistical Office. 1989. W. Harare: CSO.

Chambers, Robert. 1969. WM. New York: Praeger Publishers.

Chimedza, Ruvimbo. 1988. "Women’s Access to and Control over Land: The Case of Zimbabwe." Dept.
of Agricultural Economics and Extension Working Paper. Harare: AEE.

Christodoulou, D. 1965. "Land Settlement: Some Oft-Neglected Basic Issues." W
W 14(10). 1-6

Clayton, Eric S. 1971. "Agrarian Reform, Agricultural Planning and Employment in Kenya." in Agrarian
WW- Geneva: International Labour Office, 119-141.

 

. _. ,1.» . t: . London: Wye College. Occasional

Paper No. 3.

Cohen, John M. 1980. "Land Tenure and Rural DeveIOpment 1n Africa." in Want
' .Edited by Robert H. Bates and Michael F. Lofchie. New York:
Praeger Publishers, 349-400.

Collier, Paul and Deepak Lal. 1986. :1! . Oxford: Clarendon Press.

 

Commercial Farmers" Union. 1991. "Proposals for Land Reform for Zimbabwe, 1991." Harare: CFU.

Cusworth, John and Judy Walker. 1988. "Land Resettlement in Zimbabwe, A Preliminary Evaluation."
Evaluation Report EV 434. London: Overseas Development Administration.

de J anvry, Alain. 1984. "The Role of Land Reform 1n Economic Development: Policies and Politics." in
D l ' ' .First edition, edited by Carl K. Eicher and John M.
Staatz. Baltimore: The Johns Hopkins University Press, 263-274.

deWilde,JohnC. 1967. -. . _.. -
M9. Baltimore: The Johns Hopkins Press.

Dorner, Peter. 1971. "Land Tenure Institutions." in W Edited by
Melvin G. Blase. Ames, Iowa: The Iowa State University Press, 1431.

 

Dovring, Folke, 1974. "Land Reform: A Key to Change in Agriculture." in W
W. Edited by Nurul Islan. New York: Halsted Press.

Duloy, John H. 1974. "Sectoral, Regional, and Project Analysis." in W. by Hollis
Chenery, Montek S. Ahluwalia, C.L.G. Bell, John H. Duloy, Richard Jolly. New York: Oxford
University Press, 195-m8.

Eckstein, Shlomo, Gordon Donald, Douglas Horton, and Thomas Carroll. 1978. W
Wh- World Bank Staff Working Paper 1:10.275.
Washington, D.C.: The World Bank.

Eicher, Carl K. and John M. Staatz, eds. 1984. WWW Baltimore:
The Johns Hopkins University Press.

Feder, Gershon and Raymond Noronha. 1987. "Land Rights Systems and Agricultural Development in
Sub-Saharan Africa" MW- 2(2). 143-69-

Freeman, Howard E., Peter H. Rossi and Sonia R. Wright. 1979. W
W. Paris: Development Centre of the Organisation for Economic Co-operation
and Development.

Gaidzanwa, RB. 1988. "Women’s Land Rights in Zimbabwe, An Overview." Department of Rural and
Urban Planning Occasional Paper No. 13. Harare: RUP.

Garcia, Antonio. 1970. "Agrarian Reform and Social Development in Bolivia." in

mm
W Edited by Rodolfo Stavonhasom Garden City. New York
Doubleday & Company, Inc., 301-346.

Gesellschaft fur Agrarprojekte in Ubersee (GFA). 1987. "Study on the Economic and Social
Determinants of Livestock Production in the Communal Areas - Zimbabwe." Final report. Hamburg:
GFA.

 

Ghose, Ajit Kumar. 1983. 1, u, 2,- 2. _' ' .. . Kent, Great Britain:
Croom Helm, Ltd.

Gittinger, J. Price. 1982. W. Second edition. Baltimore: The
Johns Hopkins University Press.

Government of Zimbabwe. 1981. , ' ° ‘. _
29291933931. Conference Documentation, March 23-27. Salisbury: GOZ.

 

. 1991. "A Framework for Economic Reform 0991-1995)." Harare: GOZ, January 18.

 

Green, Joy K. 1987. "Evaluating the Impact of Consolidation of Holdings, Individualization of Tenure,
and Registration of Title: Lessons from Kenya." Land Tenure Center Paper No. 129. Madison: LTC.

209

Harbeson, John W. 1981. "Land Tenure and Rural Development Policy in Zimbabwe: The Inheritance
and Prospects for Change." Mimeograph.

_. 1984. "International Influences on Land Reform in Africa." in 199999199] Dimming: 9“ and
m- Edited by John D. Montgomery. Boulder: Westview Press, 149-167.

Hardaker, J. Brian, Jock R. Anderson and John L. Dillon. 1984. "Perspectives on Assessing the Impacts

of Improvod Asriwltuml Tochnolosios In Dcvoloums Countries." WWW
Bauuiuia 28(2-3). 87408

Hardjono, J .M. 1977. W Jakarta: Oxford University Press.

Hazlewood, Arthur. 1985. "Kenyan Land-Transfer Programs and Their Relevance for Zimbabwe."
MW 3(2). 445461

Herbst, Jeffrey. 1987a." Conflict over Land: Communal Farmers versus Squatters." in [9119
I911_“1I_I1:____1011_11 1_1'1_1_I:I)1 1 :‘W 10 9 .1_1_01' g I1I11 :10 1 39919 I

W Phd. dissertation, Yale University, 129-156.

_. 1987b. "How the Weak Succeed: Tactics, Political Goods, and Institutions in the Struggle over
Land in Zimbabwe." Mimeograph.

Heyer, Judith. 1976. "Achievements, Problems and Prospects 1n the Agricultural Sector. in “91991
2 .Edited by Judith Heyer, J .K. Maitha, W.M.

 

Senga. Nairobi: OxfordUniversity Press, 1-31. 1

House, William J. and Tony Killick. 1983. "Social Justice and Development Policy in Kenya’s Rural

Economy" in W Edited by Dhmm Chi and Smir
Radwan. Geneva: International Labour Organisation, 31-69.

Hunt,Diana.1984. It _ ' _‘
Gower Publishing Company, Ltd.

 

1 in . Hampshir' C, walla:

Hunter, John and Carl Mabbs-Zeno. 1986. "African Land Tenure." WW9. 23(2),
109-122.

J acoby, Erich. 1971. mm London: Andre Deutsch Ltd.

Kay, Cristobal. 1983. The Agrarian Reform in Peru: an Assessment." in W“

W Edited by Aiit Kumar Ghose Kent. Great Britaim Croom
Helm, Ltd., 185-239.

King, Russell. 1977. MW. London: G. Bell & Sons Ltd.

Kinsey, B. H. 1982. "Forever Gained: Resettlement and Land Policy in the Context of National
Development in Zimbabwe." £19. 52(3), 92-113.

..1983 "Emerging Policy Issues in Zimbabwe’s Land Resettlement Programmes." 12999199991
—W 1(2) 163-196-

Leo, Christopher. 1989. W Harare: Nehanda Publishers Ltd.

 

210

Lewis, Arthur. 1964. "Thoughts on Land Settlement." in W Edited
by Carl Eicher and Lawrence Witt. New York: McGraw-Hill Inc, 299-310.

Lipton, Michael. 1974. "Towards a Theory of Land Reform." in '
3919:9199. Edited by David Lehmann. London: Faber and Faber Limited, 269-315.

Maitha, ”<- 1976 "The Kenyan 50000111)!" in MnflhualflmlamaLiuMnLAaBamis
W. Edited by Judith Heyer, J .K. Maitha, W.M. Senga. Nairobi: Oxford University Press,
33—67.

Mellor. John W. 1966. IhLBanuuimLAsiialtinlMamat Ithaca. New York: Cornell

University Press.

Meyer, Carrie A. 1989.
Publishers.

Ministry of Finance, Economic Planning and Development. 1986. WWW
1289:1285. Harare: GOZ.

Ministry of Lands, Agriculture and Rural Resettlement. 1989. 'The National Land Policy." Draft.
Harare: MLARR.

 

. New York: Praeger

Morna, Colleen Lowe. 1990. "Zimbabwe May Alter Policies On Land, One-Party State." W
W91 October 25 5

Mosley, Paul. 1985. "Achievements and Contradictions of the Peruvian Agrarian Reform: A Regional
Pempeetivef WM 21(3). 440-48

Moyana. H. V. 1984. WWW Harare: Mambo Press.

Moyo, Sam 1986. "The Land Question." in: .
Edited by Ibbo Mandaza. Senegal. Codesria, 165-201.

 

Nelson, Michael. 1973. In - , _
The Johns Hopkins University Press.

 

Ng, Ronald and Francis Lethem.1983. U ' . , .
1999921319999. Washington, D ..:C The World Bank.

Oberai,A.S.1988. .2 -, .'
New York. Praeger Publishers.

 

 

Okoth-Ogendo, H .W. O. 1976. "African Land Tenure Reform." in

MW
W91. Edited by Judith Heyer, J.K. Maitha, WM. Senga. Nairobi: Oxford
University Press, 152-185.

Palmer. Robin 1977- "Agricultural History of Rhodesia" in W
5919993319. Edited by Robin Palmer and Neil Parsons. London: Heinemann Educational Books,
Ltd., 221-754.

_. 1990. "Land Reform in Zimbabwe, 1980-1990." mm. 89(355), 163-181.

211

Parliament of the Commonwealth of Australia. 1980. 21311291299. Canberra: Australian Government
Publishing Service.

Pinstrup-Andersen, Per and Peter B ..R Hazell. 1987. "The Impact of the Green Revolution and
Prospects for the Future." in ,
by J. Price Gittinger, Joanne Leslie and Caroline Hoisington. Baltimore. Johns Hopkins University
Press, 106-118.

 

Raup, Philip M. 1967. "Land Reform and A9-icultural Development."in 111W
2999919599911. Edited by Herman M. Southworth and Bruce Johnston. Ithaca: Cornell
University Press, 267-307.

Riddell, Roger. 1978. W. Gwelo, Zimbabwe: Mambo Press.

Rohrbach, David D. 1988. "The Growth of Smallholder Maize Production in Zimbabwe: Causes and
Implications for Food Security." PhD. dissertation. East Lansing, Michigan: Michigan State
University.

Roth, Michael J. 1990. "Analysis of Agrarian Structure and Land Use Patterns in Zimbabwe."
Background paper for the Zimbabwe A9iculture Sector Memorandum. Washington D.C.: The
World Bank.

Ruttan, Vernon W. and Yujiro Hayami. 1990. "Induced Innovation Model of A9'icultural Development."
in MW Second edition. edited by Carl K Eicher, John M-
Staatz. Baltimore: Johns Hopkins University Press, 97-112.

St. Clair, Alan. 1990. "Land reforms are coming, but the waiting continues." W. October,
23-24.

Schmid, A. Allan. 1972. "Analytical Institutional Economics: Challenging Problems in the Economics of
Resources for a New Environment" Mnanlmaletslmia 54(5). 893-9012

. 1987. WW9. New York: Praeger Publishers.

Senga, WM. 1976. "Kenya’s A9'icultural Sector." in . -,'

W991. Edited by Judith Heyer, J .K. Maitha, WM. 2 senga. Nairobi: Oxford University Press,
69-110.

 

 

Sternberg, Marvin J. 1971. W. Geneva: International Labour Office.

Stevens, Robert and Cathy L. J abara. 1988. . - ' '
W9. Baltimore. The Johns Hopkins University Press.

 

Stoneman, Colin. 1982. "A9'iculture." in mm. Edited by Colin Stoneman. New York:
St. Martin’s Press, 127-150.

Stoneman, Colin and Lionel Clifi'e. 1989. W99. London: Pinter
Publishers.

Suratman and Patrick Guiness. 1977. "The Changing Focus of Transmigration." Winn
W2 13(2). 78-1012

 

212

United Nations. 1984. "Guiding Principles for the Design and Use of Monitoring and Evaluation in
Rural Development Projects and Programmes." Rome: The United Nations ACC Task Force on
Rural Development.

Von Haugwitz, Hans-Wilhelm. 1972.
Weltforum Verlag.

I 1.. l I? Mme!”

 

Weiner, Dan, Sam Moyo, Barry Munslow and Phil O’Keefe. 1985. "Land Use and Agricultural
Productivity in Zimbabwe.” WM 13(2). 251-2852

Wekwete, K. H. "A Review of the Rural Land Resettlement Pr09'amme in Post-Independent
Zimbabwe." University of Zimbabwe: Department of Regional Planning. Draft.

Whitlow, Richard. 1985. "Conflicts in land use in Zimbabwe." W9. 2, 309-322.

Whitlow, Richard and Bruce Campbell. 1989. "Factors Influencing Erosion in Zimbabwe: 11 Statistical
Analysis." 2, _ :4

World Bank. 1988. W. Washington, D.C.: The World Bank.

World Bank. 1988. W. New York: Oxford University Press.
World Resources Institute. 1990. W. New York: Oxford University Press.

 

Zimbabwe National Farmers" Union. 1987. "Settlement of Qualified Farmers in Resettlement Areas."
Harare: ZNFU, September.

"I111l1ljfllfllfllflfltflfl11lfll“