MSU

LIBRARIES
.—:—.

RETURNING MATERIALS:

 

 

 

Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

EVALUATING DECISION RULES AND PLANNING TOOLS IN FARM

DECISION-MAKING: A CONCEPTUAL FRAMEWORK

By
Ridgely Abdul Mu’min

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Agricultural Economics

1987

Copyright by
Ridgely Abdul Mu’min

1987

ABSTRACT

EVALUATING DECISION RULES AND PLANNING TOOLS IN FARM
DECISION-MAKING: A CONCEPTUAL FRAMEWORK

By

Ridgely Abdul Mu’min

The major purpose of this research was to develop and
test a model which would measure the potential value of
planning tools to farm management. The systems approach was
used to develop the Management Systems Research model as a
methodology for measuring such potential. In accomplishing
the primary goals other products were produced including a
validated managerial effectiveness measure, common be-
havioral attributes of effective farm managers, use of
economic tools by farm managers, information on the rela-
tionship between behavior, performance. and effectiveness,
and information on the criteria for decision aid selection.

The measurement system developed included five major
activities: 1. Selection of Farm System, 2. Behavioral
Analysis, 3. Measuring Effectiveness. 4. Testing Behavioral
Predictors of Effectivenss and 5. Quasi-Experiment.

Once the farm system was selected, an open-ended
questionnaire was developed and administered to a group of
successful cash grain farmers as identified by extension
personnel in Ingham and Tuscola counties. Information from

the interviews was used to develop a managerial index ques-

tionnaire, and helped in determining the appropriate effec-
tiveness measure for the farm manager.

The farmers selected profitability as the most impor-
tant criteria for evaluating a farm manager. Therefore,
management income per acre was chosen as a measure of effec-
tiveness, and this measure was tested for workability,
reliability, and validity. The reliability test consisted
of testing for consistency of the measure over time in
identifying the TelFarm cash grain farmers in terms of
managerial groups. Management income per acre was checked
for validity by testing the relationship between the average
management income per acre for each of the farmers against
locational and farm size variables.

With a validated managerial effectiveness measure and a
managerial index developed, the next step was to administer
the index to the TelFarm cash grain farmers and test the
relationship between behavior, performance and effective-
ness. The results of the analysis demonstrated that
managerial effectiveness could be predicted by a subset of
managerial questions focussing on cost realization and
financial decision-making.

Now that it is possible to make aprori predictions of
managerial success using an index, the stage is set for
using such an index for measuring the possible benefits of

improved decision aids and planning tools.

ACKNOWLEGEMENTS

First I would like to thank my wife. Dianne, and
children for sticking with me through this ordeal. A
special thanks is given to my brother, Wilbur Banks, for
purchasing my computer.

Dr. Ralph Hepp, my academic and major advisor, has
given me great guidance and support through the frustrations
of both graduate school and my dissertation. I. also,
appreciate the support and advise given by the other members
of my thesis committee. Dr. Sherril Nott. Dr. Gerald Schwab
and Dr. Richard Simonds. Separate thanks are given to Dr.
Glenn Johnson and Dr. Dan Ilgin who, although they were not
on my committee. took the time to give insights and valuable
background references.

I must also thank Dr. James Jay for helping me to get to

Michigan State and finish.

ii

TABLE OF CONTENTS

ACKNOWLEGEMENTS . . . . . . . .

LIST OF TABLES . . . . . . . . . . .

LIST OF FIGURES

LIST OF APPENDICES

PART

I.

Chap.

Chap.

II.

Chap.

Chap.

Chap.

Chap.

INTRODUCTION

1 The Research Problem: Background and

Objectives .

2 Methodology: The Systems Appproach
FEASIBILITY EVALUATION

3 Needs Analysis

3.1 Technology and The Structure of
Agriculture . . . .

3. 2 Adoption of Technology .

3. 3 Micro- -computer Technology and Present
Farm Applications

3.4 Statement of Needs

4 System Identification

.1 The Farm System .
.2 The Management Process

uhkuh

Decision Support Systems (DSS)
4.4 Computer Functions As It Relates to
Measuring Value .
4.5 System Definition

5 Problem Formulation: Purpose of Measurement

System . . .

6 Generating of System Alternatives

6.1 Measuring Behavior, Performance or
Effectiveness . .

6.2 Measuring Improvements in Managerial
Behavior . . . . . .

6. 3 Measuring Improvements in Managerial
Performance

iii

Page
ii

vi
vii

viii

14
15

15
18

20
22

24
24
26

.3 Management Informamtion Systems (MIS> vs.

36

39

41

45

51

51

52

54

III.

6.4 Measuring Improvements in Managerial

Effectiveness . . . . . . . . . .

Modelling and Implementation Design

Chap. 7 Concept Selection

7.

7.
7.

1

2
3

Concepts of Management and Information
that Guide the Present Study

Elements of The Decision Function
Differences in Farming Systems as They
Affect Management . .

Chap. 6 Modelling of Concept . .
6.1 Measuring the Management System
6.2 Model

Chap. 9 Model Specificaion or Implementation Design

IV.

Chap.

Chap.

Chap.

Chap.

(0000

9.

kwNI-b

5

Selection of Farm Type .
Behavioral Analysis . . . . .
Measuring Effectiveness

Testing Behavioral Predictors of
Effectiveness

Quasi-Experiment

Implementation and Results

10 Selection of Farm Type
10.1 Implementation
10.2 Results/Outputs

11 Behavioral Analaysis
11.1
11.2
11.3

11

12 Measuring Effectiveness<System Performance)

.4

Selecting Set of Sucessful Farmers .
Develop Model of Decision System

56

61

62

62
64

66

68
68
72

75
75
77
80

83
87

94

96
96
97

100
100
103

Design Instrument to Determine Behavioral

Incidents .
Evaluate Decision-Making Behavior of

105

Successful Farmers and Develop Instrument

to Measure Behavior and Performance

12.1 Implementation
12.2 Results

13 Testing Behavioral Predictors of
Performance
13.
13.
13.
.4
.5

13
13

1
2
3

Administering the Instrument to Farmers

Scoring the Test .

Testing the Hypotheses

Revising the Instrument . .
Implications for Quasi-Experiment

iv

107

127
127
131

138
139
139
144
160
162

Chap. 14 Evaluation of Secondary Outputs of Process
14.1 Decision-Making Behavior of Successful

14.

14.

14.

Farmers

164

164

Implications of Observed Decision-Making
Behavior to Farm Management/Agricultural

Economics

Performance (Effectiveness) Measures
Revisited .

Criteria for Development and Selection
Decision Aids

V. Conclusions and Recommendations

Chap. 15 Summary of Major Findings

Chap. 16 Evaluation .
16.1 Conceptual Framework Evaluation
16.2 Implementation Evaluation

Chap. 17 Recommendations

17.
17.
17.
17.
17.
17.

APPENDICES
BIBLIOGRAPHY

05019026.)»-

Extension

Research

Teaching

Farm Policy . . .
Recommendation for Model Application
Preliminary Assessment of The Role of
Microcomputers in Farm Decision-Making

 

169

170
of
172

174

175

182
182
185

190
191
192
194
196
196

198

200
221

Table

Table

Table

Table

Table

Table

Table

Table

Table

LIST OF TABLES

Anova-1 Management By Group and Year
Environmental Variables Affecting MIA
Testing Hypothesis I: MIA = C + b*TSCORE

Testing Hypothesis II: MIA = f<BEHAVIOR,
PERFORMANCE)

Testing Hypothesis III: MIA = f<PROD,MARK,
FIN. INFO)

Testing Hypothesis IV: MIA = f<ORGAN, ACOUIRE
COSTPR.FINBP)

Comparing Hypothesis IV without ACOUIRE or
without ORGAN . . . . . . . . . .

Testing Hypothesis IV without Six Top
Farmers . . . . . . . . . . . . . .

MIA = f(ORGAN. COSTPR, FINBP:CTY AND ACRE)

vi

Page
133

136

146

149

151

154

156

158

159

Figure

Figure
Figure
Figure
Figure
Figure
Figure

Figure

Figure

Figure
Figure
Figure

Figure

Figure

Figure

10:

11:

12:

13:

14:

15:

LIST OF FIGURES

The Systems Approach as a Problem Solving
Methodology . . . . . . . . . .

Feasibility Evaluation

Johnson's Problem Solving Model
System Identification

The Farm System

Analysis Phase of Management System
Measurement System .

Isolation and Measurement of Management
System

Page
8

10
12
25
27
37

42

47

Management Systems Research Model for Measuring

the Value of Decision Aids
Selection of Farm Type
Behavioral Analysis
Measuring Effectiveness

Testing Behavioral Predictors of
Effectiveness

Quasi-Experiment

Map of Michigan

vii

73

76

78

81

84

89

102

LIST OF APPENDICES

APPENDIX A: OUTPUT OF OPEN-ENDED QUESTIONNAIRE . . . . 200
APPENDIX B: OPEN-ENDED QUESTIONNAIRE . . . . . . . . 212
APPENDIX C: DECISION-MAKING ANALYSIS QUESTIONNAIRE . . 214

viii

PART I
INTRODUCTION
The use of micro-computers by farmers as a management
tool has become a real phenomenon in the last few years.
The cost of micro-computers, unlike the cost of other agri-
cultural production inputs, has been declining; thus placing
the computer within the budgetary constraints of smaller
operations. The question remains as to the possible impli-
cations of widescale adoption of this technology as it
affects farm viability.

In the past, technological innovations have increased
the concentration of production towards larger, more specia-
lized units (James and Pugh,1968; Holland.1960; Osborn.
1961). The characteristics of computers would indicate that
there may be significant economies of size associated with
their adoption and utilization. If such technology imparts
a significant competitive advantage to the farm user, then
farmers who adopt and use the new technology effectively can
be expected to have a greater probability of surviving and
expanding their operations. Therefore. small farmers without
access to computer power may be at a significant short and
long-term disadvantage. which may imperil the existence of
the smaller U.S. family farm and contribute to more concen-

tration in agriculture.

However, the rate of adoption has not been as rapid as
expected because of a number of factors. According to King
(1985), these include: 1) the extraordinarily rapid rate of
technological change in information technology has dis-
couraged farm investment due to the fear of obsolesence; 2)
widespread financial stress in the farm sector has reduced
levels of investment in general; and 3) farm information
systems and information services often do not meet the needs
of farm managers. To ascertain what these needs are
requires knowledge of the decision-making behavior of

farmers.

Chapter 1
The Research Problem: Background and Objectives
W

This research was motivated by a desire to develop com-
puterized information systems and farm management tools
adapted to the needs of small farmers. The level of adop-
tion of such tools would depend upon their perceived bene-
fits to farmers. It was hoped that by looking at how the
computer benefited larger farms, inferences could be made
for smaller operations. However, the studies available
focused on qualitative and not quantitative benefits.
Therefore, it was apparent that there was a need for
research on the quantitative benefits of computerized
decision aids.

Many of the computerized decision aids are merely com-
puterized versions of the same decision aids advocated in
the past by farm management professionals before the advent
of microcomputers. Therefore, another approach explored was
to examine research findings that had quantified the
benefits of non-computerized farm management tools.

However, this approach did not yield sufficient information
in this area.

Since the subject matter information on quantified
benefits of computerized or non-computerized farm management

tools was not available. the next step was to determine how

to develop such information. This led to a search for the
appropriate methodology for determining the value of such
tools. Focus now centered on the literature pertaining to
the value of information. Previous attempts at determining
such value can be categorized into the bayesian approach,
the production function approach, the residual analysis
approach. and the experimental approach. These approaches
along with their strengths and weaknesses are presented in
the main body of this text. However, none of these
approaches seemed sufficient to determine the value of
information or an information system in a multi-decision
environment.

In the process of investigating the different ap-
proaches to measuring the value of information, it became
apparent that there was a need to clarify the concept of
“management" which is at the center of the measurement
problem. Therefore, managerial behavior itself became a

primary subject for investigation.

91' !'
Therefore, this research focuses on developing the
methodological framework for answering such questions on the
potential value of computerized decision aids for farmer
managers. The specific objectives of this research are:

1. To develop a model of procedures to determine the

potential value of planning tools in farm decision-

making, and

2. To test the model by using cash grain farmers in

Michigan as subjects.

Chapter 2
Methodology: The Systems Approach
W

The problem is very ill-defined and nebulous, and
therefore calls for an approach that will help to give it a
form that can be modelled. The “systems approach" will be
used to identify and structure the problem and develop the
model of the measurment system.

This study can be considered disciplinary research. Ac-
cording to Johnson (1982), disciplinary research in eco-
nomics consists largely of research to develop and improve
economic theories, quantitative techniques and the measure-
ment of basic economic phenomena. This research is designed
to develop a measurement system to determine the value of
information or an improved information system as it supports
farm decision-making. The technique will be used to
evaluate micro-computers, however the method developed will
be appropriate to the generic problem of determining the
value of information in any non-static, multi-decision en-

vironment.

W

The "systems approach" is defined as a problem solving
methodology which begins with a tentatively identified set

of needs. It has as its result, an operating system for

efficiently satisfying a set of needs which are acceptable
or “good" in light of trade-offs among needs and the
resource limitations that are accepted as constraints in the
given setting (Manetsch and Park, 1982). There are two
prominent attributes of the approach: 1) it overtly seeks to
include all factors which are important in arriving at a
“good” solution to the given problem; and 2) it makes use of
quantitative models, and often computer simulation of those
models, to assist in making rational decisions. This
research will attempt to focus on the first attribute by
involving both managerial professionals and the managers
themselves in developing a useful model. However, instead
of producing a quantitative model, the technique will be
used to develop a conceptual framework and a working system
of procedures to use in evaluating planning tools in farm
management.

The major phases of the approach are: feasibility
evaluation, abstract modelling, implementation design,
implementation and system operation. As seen in Figure 1
reproduced from Manetsch and Park (1982), the movement of
the process is from a set of needs to be satisfied (at the
top) toward an operational system (at the bottom) capable of
satisfying the needs which exist. The solid lines and
arrows indicate the path of flow through the process as you
move through the sequence of functions which must be carried

out in order to achieve system objectives.

E .].].! E IHESIQD has as its goal the generation

Needs

1

 

Feasibility
Evaluation

 

 

Positive
———>

Information

‘ Implementation
Design
é——9[Implementation

Modelling

i

# Abstract

 

i
.L

 

4

 

System .
Operation

Normative
Information

Figure 1: The Systems Approach as a Problem Solving Methodo—

logy

of a set of feasible system "alternatives" capable of satis-
fying needs which have been identified (Manetsch and Park,
1962). The steps in feasibility evaluation are reproduced
in Figure 2. Feasibility evaluation begins with a careful
analysis of needs to determine whether the needs do, in
fact, exist and if they do, to state them in an opera-
tionally useful form. The "output" of Feasibility Evalua-
tion is a set of realizable alternatives which appear
capable of satisfying the identified needs.

uggglling receives as inputs, the feasible alternatives
of Feasibility Evaluation and has as its output, the broad
specifications for a system design and/or management
strategy to be implemented in the real world. In this case,
the model will consist of procedures or methods to follow in
evaluating the role of planning tools in farm management.
This phase begins by sifting through the different
conceptual alternatives and the selection of the concept to
be modelled.

Lnglgmgntgtign_flggign_has as its purpose to completely
specify the details of system and/or management strategy
designed in the abstract during the modelling phase.
"Completely specify“ means developing a complete set of
instructions that will lead to the operationalization of the
desired system. This would include specifying the necessary
preparations for implementation including data and budgetary
needs along with the procedures for accomplishing specific

funtions.

10 ;;

Primitive Need

3|:
Needs
Analysis ‘
System 4
Identification

Positive * Problem k Normative
Information Formulation Information

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Generation

of System 1
Alternatives

Physical
‘ Social and
E 3 Political

Realizability

 

 

 

 

 

 

 

 

 

 

Determination
of Economic and
Financial
Feasibility

 

 

 

To Abstract Modelling

Figure 2: Feasibility Evaluation

11

lgnlgmgntatign is to give physical existence to the

desired system. In this research, the procedures developed
will be tested by using a specific farm type as an example
of the measurement system in use. Since the systems
approach is iterative, information developed in this phase
on the workability of the system would act as feedback for
the modifications needed for implementation of a flg;king_,
Eman-

Similarities between the steps in the systems approach
and Johnson’s (1977) problem solving model (Figure 3 below)
indicate the generic nature of both processes. Also, the
systems approach can be observed to coincide with the basic
thesis development i.e., problem statement (feasibility
evaluation), investigating how others have brought light to
the problem in the form of a literature review (developing a
set of feasible solutions), developing of model to be tested
(abstract modeling), statistical methods and hypothesis
testing (implementation design), analysis and recommenda-
tions (implementation). However, the systems approach goes
further and actually involves the decision-maker or those
who will be affected by the system to participate in its
development. Plus, the approach goes toward development of
the working system to which much academic research falls
short.

The systems approach makes the assumption that it is
the end user and not a so-called ”expert" that will deter-

mine the adequacy of the system outputs. This is important

12

 

 

Hiroblem Definition

 

 

Observation

_¥
Positive ______J$%i Analysis Normative
Information e: 1 Information

Decision

H Execution H
Responsibility
Bearing

 

 

 

 

 

 

 

 

 

 

Figure 3: Johnson’s Problem Solving Model

13

when dealing with a normative discipline such as economics

where the role of the professional is to educate the

manager, but learn from other theorists. This research

assumes that there is such a thing as a "successful farm

manager“ and that we can learn from what they do and not

just expect them to accept what we teach. Although the

measurement system developed by the
used primarily by other economists,
not be so abstract as to render the
of correspondence to the real world
farmer faces.

Another benefit of the systems

systems approach will be
the methodology should
values measured devoid

conditions that the

approach to very un-

structured problems is the introduction of detailed diagrams

which guides the research and helps

sion of necessary parts to complete

to prevent the exclu-

the working system.

This does not mean that no parts can be omitted either

knowingly or unknowningly, but diagramming the process helps

to focus on the essentials.

PART II

FEASIBILITY EVALUATION

As stated previously, Feasibility Evaluation begins
with some expressed or unexpressed need, and ends with a set
of feasible alternatives. This phase parallels the Litera-
ture Review and Problem Statment phases of the conventional
thesis format. This phase will include: 1)Needs Analysis,
2)System Identification, 3)Problem Formulation and 4)Genera-

tion of System Alternatives.

14

Chapter 3

Needs Analysis

Feasibility Evaluation begins with the analysis of the
needs the system under consideration must satisfy. Needs
analysis attempts to penetrate the veil of ”apparent" needs
and identify underlying actual needs. Needs analysis must
carefully consider the needs of all persons and institutions
which will be involved with the proposed system.

3.1 Technology and The Structure of Agriculture

A review of literature suggests that although the
definition for a small farm is not settled (Fiske, 1983),
the facts show that by whatever definition their numbers are
declining. However, there are conflicting opinions both
inside government and academia on the implications of this
decline. Since Thomas Jefferson, the expressed intent of
the nation’s agriculture has been to support the model of
the self-sufficient farmer who owned his own land. The
emphasis, however, has shifted over time to the more commer-
cially oriented larger farms (Fiske, 1983). This pattern was
probably accelerated by the need for increased output per
worker neccessary to sustain a major war effort such as
World Wars I and II. Increased output and productivity even
after the war years has been equated by many as a national

symbol of progress (James and Pugh, 1968). Even the presence

15

16

of small farms was considered a "consternation" to commer-
cial farmers through their influence on resource prices and
agricultural policy (Brannen, 1968).

Within this question over national priorities, comes
the question of efficiency verses equitable distribution of
income. Murphree (1968) has pointed out that increased farm
size in areas of limited alternative employment opportuni-
ties results in stagnation of those rural communities under-
going this transformation. Since the result of increased
farm size has been the displacement of farm workers into the
unskilled labor market, if all factors are accounted for,
there may not be a net social benefit to increased labor
productivity. Murphree (1968) further points out that while
land values have been appreciating -- due to increased land
productivity -- with the increased trend of absentee land-
ownership, income transfers through land may be accruing to
a largely non-agricultural sector.

Different explanations have been given for the dispro-
portionate rate of small farm failures. A number of these
reasons center around increased capital requirements due to
the increased use of purchased inputs (James and Pugh, 1968;
Brannen, 1968; Babb, 1979). The increased use of purchased
inputs has been a result of the successful efforts of the
USDA and Land Grant Institutions to advocate the use of
improved technology. However, as some farmers have dis—
covered the rewards for the incorporation of innovations on

the farm accrue only to the early adopters, usually larger

17

farmers (Holland, 1980; Osborn,1980), then the rewards
accrue to the consumer (Steichen, 1968). In fact, there are
suggestions that the benefits of employing the new techno-
logies do not cover their high market cost when compared to
the replaced technologies (Steichen, 1968).

Reasons for small farm survival and demise have been
attributed to the characteristics of the farmers themselves
and their decision-making processes. Olson (1978), in his
investigation of the information needs of small farmers in
Michigan, discovered that small farmers were indeed
different in both physical and human resources available to
them, technology appropriate to them, and the formulation of
goals and identification of problems by them. White and
Boone (1976) in their investigation into the decision-making
processes of limited resource farmers in eastern North
Carolina found that small farmers did, in fact, rely on more
interpersonal sources for production and marketing informa-
tion, while larger farmers had more contact with what Rogers
(1971) would call “change agents". However, White and Boone
also found that within this limited information structure,
small farmers behaved rationally in their decision-making
processes. That is, they followed the five steps of
rational decision-making defined as: 1)0rientation,
2)Observation, 3)Analysis, 4)Implementation, and 5)Feedback
and Adjustment (White and Boone, 1976). They concluded that
small farmers could, in fact, be helped towards better

levels of production and standards of living if they are

18

steered to ”unbiased" sources of information.
3.2 Adoption of Technology

The problem of evaluating the effects of technology
centers around the question of adoption or nonadoption and
subsequent consequences. To evaluate the adoption potential
for a given technology requires knowledge of the character-
istics of the technology and of the potential adopter.
Leagans (1979) observes that to determine how farmers will
accept change one must understand why farmers do what they
do. The burden of blame for the non-adoption of technology
should be directed away from the farmer towards the short-
comings of the society or change agent to affect the
perceived attributes of the technology favorable to
adoption. Adoption depends on the cumulative valence of
incentives versus disincentives (Leagans, 1979). In other
words, people observe both limitations and incentives to
innovate simultaneously (Leagans, 1979). Drucker (1977)
also emphasizes the importance of perception in changing
behaviour. Communication begins with the receiver not the
sender. For effective communication of an idea, the sender
must put himself in the shoes of the receiver to understand
his point of view to ensure that the idea is expressed in
terms emotionally acceptable to the receiver (Drucker,
1977).

Rogers (1971) confirms the dichotomous nature of the
problem as he explains adoption in terms of both the attri-

butes of the technology and the potential adopter. The

19

external factors affecting adoptions include: perceived
attributes of the innovation, type of innovation-decision,
channels of communication, nature of the social system, and
extent of the change agents’ promotion efforts (Rogers,
1971). Internal factors which are characteristics of the
adopter are: socio-economic variables, personality, and
communication behaviour (Rogers, 1971). This format takes
both the attributes of the technology and the adopter as the
independent variables affecting adoption. However, micro-
computer technology as a planning tool is not set. Its
attributes change as its capabilities increase and cost
decreases. Both of which are continuing at accelerated
rates.

In the Organization for Economic Development’s (OECD)
study on the ”Social Assessment of Technology" (1978),
attention is drawn to the dynamic nature of new and yet
partially evaluated technology. The report emphasizes the
importance of an iterative process in determining the impact
of a new technological development (OECD, 1978). They
suggest that:

"...the most satisfying approach seems to be a conver-

gent investigation of needs and demands of the poten-

tial users on the one hand and of the opportunities and
promises of new technological discoveries on the other.

Critical comparison of the two trends leads to the

choice of particularly interesting technological deve-

lopments among which the most promising one can be
distinguished as the pivotal objective of the
analysis.“

(OECD, 1978:25)

This approach suggests that an investigation of the

impact of such a multifaceted and flexible technology like

20

the micro-computer must be iterative. The essential problem
is to determine which group of farmers could possibly
benefit most from the adoption of micro-computer technology
and then determine optimal hardware and software combina-
tions appropriate to that group or groups. There are as
many configurations that computer technology can be packaged
in as there are applications for their use. Furthermore,
the technology must be available in both the physical and
economic sense, i.e. within reasonble proximity and af-
fordable.

3.3 Micro-computer Technology and Present Farm
Applications

By definition, a computer is an electronic device capa-
ble of receiving and storing data, performing prescribed
numeric or logical operations on the data, and reporting the
results of those operations (Sonka,1983). There are two
parts of a computer system, hardware and software. Hardware
consists of an input device, a central processing unit,
video display, disk or other types of external memory,and a
printer. Computers can be categorized according to their
speed of performing calculations, capacity to store data and
instructions, portability, and cost (Sonka, 1983). The
characteristics that distinquish micro-computers from their
larger cousins, main-frames, are size, cost, and limited
internal memory. However, the internal memory and pro-
cessing ability of micros are quickly catching up with many

main-frame computers that have become standard on many

21

university campuses. Software can be considered the mind of
the computer and the hardware its brain. The hardware
provides potential, whereas the software provides applica-
tions. The key to usefulness of any system is appropriate
software to handle the desired task. Would-be computer
users are warned to first look at the software needed to
solve their computer needs before deciding on hardware
(McGrann,1982).

In a Purdue University survey of on-farm computer users
in Indiana, the most popular computer applications were in
the financial areas: general ledger, accounting, record-
keeping and financial statements. The most needed software
development was in the areas of crop production, access to
market data, market planning, and field records (Purdue
1982). The accounting software is probably more developed
due to its similarity with non-farm business needs. General
accounting packages can be adapted slightly to encompass the
categories most appropriate to farm situations. However,
the more farm oriented needs, like crop production software,
investment, and market planning have had to wait for subject
matter specialists to get involved with software develop-
ment.

The Purdue University study tried to measure the bene-
fits of micro-computers in a qualitative way when surveying
on-farm computer users. It found that the majority of the
farmers perceived the benefits to be improved control, time

saving, cost reductions, education and entertainment.

22

However, the time saving element depends on the context in
which time is measured and the quality of time considered.
Henshaw has pointed out that it takes just as much time to
input data into a computer as it takes to write it down in a
ledger. Plus, time must be spent in learning the skills
necessary to successfully use the technology. The farmer or
farm manager must learn typing, accounting, ration formula-
tion and other analytical skills to effectively use infor-
mation generated by computer programs (Doster, 1982;McGrann,
1982). However, the time that a computer can save is
realized in the analytical and computational aspects of data
processing. Whereas, the time needed to collect and input
data may be the same for a manual and computerized manage-
ment information system, the computer can quickly facilitate
the transformation of data to information. The immediacy of
results is a primary feature of computers which make them
most appealing in demonstrating the value of keeping records
(Henshaw, conversation). An important question is whether
people who are not good at or just dread mathematical

calculations can find in the computer a welcomed relief.

3.4 Statement of Needs
The need to specify a measurement system stems from the
observation that there is very little tangible evidence
quantifying the benefits of planning tools including on-farm
computers. To date studies have only exposed qualitative
benefits derived from testimonials from present computer

users. If we are to analyze the possible implications of

23

the technology on efficient utilization of inputs, distribu-
tion of income and structure of agriculture, we need to know
more about which groups might benefit most from its adoption.
The lack of quantitative analysis of the value of on-
farm computer use can be attributed to two causes. One is
that the technology is in its infancy with a very short
track record. The second is the nature of the technology
itself. A computer is basically a management aid. The
divergent opinions of those that study management as to
whether it is an "art" or a ”science" have made attempts at
defining it, isolating it, and measuring it difficult.
Together with the problem of measuring the value of
micro-computers as a planning tool to the farm manager are
the problems of determining the economic value of informa-
tion and measuring the management input to production. It
seems that to solve any one of these problems is to substan-
tially solve the other. Since one of the main functions of
management is decision-making and the major input to
decision-making is information - to determine the value of
information requires a measure of both the changes in
behavior and performance as a result of an improved informa-
tion system. The problem then becomes what to measure as

benefits and how to measure them.

Chapter 4

System Identification

System Identification forms a link between the state-
ment of needs and a specific statement of the problem that
the system is developed to solve. This phase leads to a
detailed specification of the variables involved in the
design and control phases and concludes with the determina-
tion of performance criteria which will assist in evaluating
system alternatives. This section begins by defining
management within the overall Farm System. The management
system is further identified in terms of how computerized
information systems are theorized to affect management.
This chapter concludes by identifying the proposed Measure-
ment System in terms of its relationship to the management

system and criteria for evaluating a working system.
4.1 The Farm System

A ”system” is a set of interconnected elements
organized toward a goal or set of goals (Manetsch and Park,
1982). A basic systems model along with definitions of
component parts are given in Figure 4. The systems approach
forces a conscious recognition by the researcher of what is
and is not under the manager’s control and who are the

relevant managers of that system.

24

\

 

 

l
CONTROLLABLE M SYSTEM 1}

INPUTS

 

 

25

  
  
 

SYSTEM ENVIRONMENT

Environmental
(exogenous) I I
Inputs

'DesiredI
Outputs|
l

 

 

 

Undesired
Outputs

 

 

 

System Design
Parameters

 

 

 

MANAGEMENT
SYSTEM

 

 

 

Figure 4: System Identification

‘W’NV

26

The “Farm System“ shown in Figure 5 includes the
farm/family with its different production enterprises and
consumption units which transform land, labor, and capital
into outputs such as production, income, cash balance, con-
sumption, education, net worth and debt. The transformation
process is determined by the design parameters of the system
which consists of land area, soil productivity, labor avai-
lability, input availability, input/output coefficients, and
per-capita consumption. The controllable inputs under the
direct influence of the manager(s) are levels and types of
production, marketing strategies, and investment policies.
The "System Environment" includes those inputs which are
exogenous to the system or outside of the direct influence
of the manager including: weather, the biological process
itself, output prices, input prices, government programs,
wage rates, interest rates, and credit availability.
Although the manager may not be able to directly influence
these factors, a knowledge of them and their probable direc-

tions is needed to make adjustments in his farm system.

4.2 The Management Process
According to Johnson (1977), management is best

regarded as a separate “enterprise" which:

"...uses resources (time, information, data processing

equipment. etc.) to p;g§g;g_gggigign§ about, among other

things, the amounts of inputs to use in producing speci-
fic products, levels of production and consumption,

27

Environment
Product prices
Input prices
Interest rates
Gov. Programs
Weather

Wages

GVQU'Iwat-e

 

 

 

 

 

 

 

_L

 

Credit availability
Biological Process

 

 

 

 

I
l
¥ I
Controllable 4+ System Desired l
Inputs Outputs |
I
1. Crops: mix, varieties 1. Crop production 1. Products I
2. Inputs: levels, types 2. Inventory 2. Income r
3. Herds: size, breeds 3. Livestock prod. 3. Cash I
4. Marketing 4. Consumption: 4. Net worth I
5. Credit policy investment and 5. Consumption l
6 Labor scheduling personal 6. Education
7. Hiring labor 5. Transportation 7. Savings I
8. Consumption levels 6. Processing 8. Status I
9. Increased I
T T productivity
Design I
Parameters Undesired I
Outputs |
1. Crop I/O ratios
2. Livestock IIO 1. Long work
ratios hours
Production cycle 2. Debt
Input availability 3. Pollution
Production capacities 4. Soil

\1 0301:50)

 

 

Physical capital
quantity 8 quality
Market limitations

 

 

 

 

Management System

depletion

 

 

' A

The Farm Family

 

 

 

 

Figure 5. The Farm System

28

decision rules to use, actions to take in carrying

out projects, projects, projects to execute in adminis-
tering programs, which programs fit under an overall
policy, and what policies to adopt.“

The management process produces decisions not products
like corn or milk. Management can be further divided into:
problem definition, observation, analysis, decision, execu-
tion and responsibility bearing (Johnson, 1977). Within the
management system, normative and positive information is
processed through decision rules to produce prescriptions.
The manager,s value system determines what is "good" or
"bad" while the prescriptions or decisions made by the
manager can be labeled “right" or "wrong“ depending on how

well these decisions are made or if they are appropriate for

accomplishing the manager’s intent.

MSW

There are two divergent schools of thought as to
whether management is a ”science“ or an "art“. By "science"
one means a verified, accumulated, and organized body of
knowledge that has been objectively obtained and systemat-
ically checked (Richards and Greenlaw, 1972). The concept
of “art” emphasizes the skills obtained through experience
by the manager in solving problems. One school argues that
management can never be a science due to the complexity of
problems that managers face. Although the manager may use
knowledge obtained from science, he must decide, even if

such knowledge does not exist (Richards and Greenlaw, 1972).

29

Therefore, one cannot categorically assume that the uti-
lizaton of scientific methodology will produce results
better than those achieved by experience-based decision
approaches.

In the past, economists by projecting the manager as the
”rational” (scientific) man, emphasize the allocative role
of management, but because of its Neoclassical underpinnings
can be criticized for its reductionist and deterministic
view of the managerial process. Johnson (1957) has pointed
out that "... economics turns out to be a necessary but
insufficient basis for management" and that "... static
production economics assumes away much of both the process
and problems of management" (1957: 444-445).

However, Teece and Winter (1984) have observed, an
economist’s view of management is disguised production
economics. They further distinguish between economics and
management in that economics concerns itself with problems
that are very different from management problems (1984).
The difference in focus can be summed up into three
categories:

“ 1. Economists have a certain brand of rationality

based on a reductionist approach instead of seeing the
problem as a whole.

2. Economists are not concerned with predicting the
behavior of individual decision units, but with
analyzing the pattern of behavior of whole populations
of economic agents.

3. Economics deals primarily with a world of
organized auction markets operating under high informa-
tion constraints.”(Teece and Winter, 1984)

30

The result has been the development of analytical tools
by economists that (a) underemphasize dynamics, (b) have no
embodied theory of innovation, (c) have an inadequate theory
of the firm, (d) and have no role for entrepreneurship
(Teece 8 Winter, 1984). Attempts of reforming the econo-
mist's static profit maximizing objective function approach
as an approximation of how the manager behaves include the
development of the expected utility hypothesis, dynamic
programming, multivalue maximization techniques, and game
theory.‘

Even these attempts at introducing dynamics into
economic models have suffered from the economist’s obsession
with developing rigorous mathematical models which are cum-
bersome in arriving at solutions and whose underlying as-
sumptions greatly restrict the generality of their conclu-
sions. Some problems include empirical evidence of pre-
ference reversals which damages the generality of the
expected utility hypothesis (Grethe 8 Plott, 1979). Even if
the mathematical, expertise and data requirements problem
necessary for arriving at solutions using these models is
solved, these models de-emphasize the time consuming and
expensive predecision stage that the manager must face
where: 1) the alternatives must be selected, 2) the proba-
bilities of occurence of states of nature determined, and 3)
the alternatives given utility weights based on estimating
the utility of multi-variate outcomes (Toda, 1976). These

models further give answers to only one problem, whereas any

31

particular decision will be nested with more grand problems

within a ”decision hierarchy“ (Toda, 1976).

Such a restricted view of the manager’s role by some
economists has led others such as Johnson to view managers
as essentially problem solvers. Now instead of the manager
fullfilling the role as the economist’s "rational man”, he
may use the tools of economics within certain of his
functions. In particular within the analysis phase the
manager can use the economic “...principle of matching added
costs against added returns under certain second-order
conditions, as a basis for defining an optimum“ (Johnson,
1957: 447). Such a principle is of use both "... 1) in
solving the breeding or engineering problem, which may be
either static or dynamic, and 2)in ascertaining within the
dynamic decision-making process, the Optimum amount of
information to acquire and the optimum amount of analyzing
worth doing“ (Opcit. 1957). The manager must set priorities
when it comes to determining which problems to solve first
and how much time to spend on each. The marginal returns
from the appropriate allocation of time to decision-making
could outweigh the benefits of a sophisticated method for
solving any particular problem.

Further, not all of the information needed for the
decision-maker is economic in nature and includes technical
as well as human and institutional information. Different

aggregates of information are needed for different sets of

32

problems. For instance, farmers use a different set of
information sources when establishing a new enterprise than
they do when managing an ongoing one (Johnson gt_al, 1961).

Mintzberg opposes the management sciences' classical
view which suggests that the manager organizes, coordinates,
plans and controls. Instead he emphasizes the “... inter-
personal, informational, and decisional roles" that the
manager plays (Mintzberg, 1975). The manager as resource
allocator falls under his decisional role which also
includes entrepreneurship, disturbance handler, and nego-
tiator. Within his interpersonal role, the manager acts as
a figurehead, leader, and liaison. His informational role
requires him to be a monitor, disseminator and spokesman
(Mintzberg, 1975).

While fullfilling these roles, the manager relies more
on verbal methods of information gathering rather than on
documents or outputs from sophisticated management informa-
tion systems. The manager is more likely to make decisions
based on "...programs remaining locked deep inside their
heads“ (Mintzberg, 1975 p53), instead of some systematic,
analytically determined procedure which is external and open
for evaluation. This internalization of the decision
process is what may limit attempts at improvement by the
management scientist, since only as these processes become
externalized can they be scrutinized and analyzed.
Mintzberg still accepts the usefulness of management

scientists, since he sees their role as understanding and

33

educating the manager as to his different roles which
includes, but is not limited to, systematic decision-making.

The issue as to whether management is an "art“ or
“science" raises questions regarding how the psychological
makeup of the manager may affect the way they make deci-
sions. The manager that sees himself as an "artist“ may
rely more on "seat of the pants” reasoning, while the
"scientific” type may prefer more systematic management
tools. Roberts and Lee used the Myers-Briggs Type Indicator
(MBTI), which seeks to measure personality types, at Texas
Tech University in 1977 to analyze personality types of both
professors and students in the Agricultural Economics
Department. Their result suggest that sensing and judging
types prefer to learn in a systematic manner, while the
intuitive and perceptive types prefer to learn in a
flexible, unstructured manner. Students exhibiting
contrasting personality types should not be taught or
treated in the same manner (Roberts and Lee, 1977).

Henderson and Nutt, also using the MBTI, report that
cognitive style seemed to influence executives’ choices when
choosing capital expansion projects under simulated condi-
tions. Therefore, if cognitive style affects both how we
learn, utilize information, and make decisions, then it
would influence the farm manager to choose the information
system most conducive to his decision-making process.

A related issue involves how much confidence a decision-

maker places in information to be used for decision-making.

34

Fischhoff and Macgregor assert that forecasts have
little value to decision makers unless it is known how much
confidence to place in them“ (Fischhoff and Macgregor,
1982). Therefore, a manager may choose not to use an

"improved" information system, because his cognitive style

may prejudice him against any ”scientific” management tool.

I! l' E! U" “1' IE]

The manager has to decide on what weights to give
different outcomes based on his value/goal system which may
change over time. The setting of goals and the evaluation
of values is another function that the manager is burdened
with, but is also prime areas for scientific analysis as an
aid to understanding and process improvement (Johnson,
1957).

Goals are decisions that act as guide posts for the
making of other nested decisions or represent constraints on
what alternatives are looked at when making decisions. As
such, no one or set of goals can be held to be constant over
any wide range of managers nor constant over time with the
same manager. Harman gt_11 (1972) out of a list of eight
goals discovered that four were picked first most often;
1)make more annual profits, 2)maintain or increase family
living, 3) avoid years of low profits or losses, 4)avoid
being forced out of business. The ranking of these goals
however varied over manager characteristics of age and
tenure of the operator, educational attainment, number of

dependents, assets, net worth, debt-asset ratio, off-farm

35

income, total land and cropland in the operation, total
acres owned, and the proportions of land and cropland owned
(Harman gt_gl, 1972). Such an extensive list of factors
associated with the selection of primary goals makes the
establishment of stable utility functions questionable.

Vogel (1981) argues that productivity and profitability
are not the small farmer’s top priorities. Woodworth gt_gl
(1978) and Orden (1977) independently found that small
farmers did not take full advantage of credit opportunities,
nor did they express strong desires to expand, indicating
strong risk-averse behaviour.

Unlike these findings, other researchers contend that
small farmers do recognize economic as well as socio-
cultural conciderations (Hansen gt_gl, 1981;Fiske,
1983;3chroeder, 1983). Hansen et al (1981), while using the
Farming Systems Research method of investigation in one
Florida county, discovered that to understand the small farm
allocation problem one must look at the whole farm as a
system. Goals of the family may include making a farm profi-
table, be self-sufficient and seek off-farm employment and
all of these without the exclusion of the others.
Efficiency, exorts Hansen et al (1981) must be measured
according to the overall goals of the farm household and not
just farm output efficiency. Schroeder (1983) also empha-
sized the blending and supportive nature of various goals
towards the total family well-being in her study of Illinois

small-scale farmers. Lifestyle goals and profit goals were

36

found not to conflict as the farm family attempts to
maximize utility (Shcroeder, 1983). Profit motives were
found to strengthen investment motives, and therfore
indirectly strengthen the lifestyle motives. Lifestyle
motives exerted a positive effect on investment and
therefore profit (Shcroeder,1983).

The profit maximizing motive which is hypothesized to
be the prime mover behind economic decision-making, may not
be the key factor for nonfarm as well as farm firms.
Drucker (1977) argues that to understand business behaviour
one must look at the business’ overall objectives which more
than not is survival, not short-term profit maximization.
Profits are essential for survival but not to the sacrifice
of other goals such as: perpetuity, harmonious relationship
within the economy and society in which it must function,
the ability to supply an economic good and service, and

being a change agent within that society (Drucker, 1977).

4.3 Management Information Systems (MIS)
vs Decision Support Systems (088)

Within the management process, it is the analysis phase
in which the introduction of computers into the decision
making process is theorized to have its greatest impact
(Capouch, 1981; Doster, 1982; McGrann, 1982; King, 1985).
Figure 6 represents this analysis phase with the emphasis on
the level of structure of problems to be solved. In

general, structured problems are those where there is very

37

mm

DATA PROCESSING

 

 

 

 

 

 

 

 

 

PROBLEM CLASSIFICATION~

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GJNSTRUCTUREED CSEM I -STRUCTURE£)) STRUCTUREID

INTUITION SIMULATION FORMAL OPTI-
- What if? MIZATION MODEL

(e.g. L.P.)

 

 

 

 

 

i ' L

HUNCHIEDUCA N91150: CTION RECOMMENDATION N)
( ‘ GUESS

DECISION

 

 

 

 

 

DEFINTIONS:

STRUCTURED: These problems are of the recurring type
which may be handled within some formal optimization model,
solved through using heuristics (rules of thumb), or be so
routine as to be almost an instuntaneous response to stimuli

- habit.

SEMI-STRUCTURED: Semi-structured problems while having
part of its solution adaptable to systematic modelling, will
usually not have a builtin optimization criteria. Results
of different scenarios must be weighed by the manager
against goals, values, and other decisions made or to be
made.

UNSTRUCTURED: Unstructured problems arise which are
completely novel to the manager, whose solution is not
amenable to mathematical or other systematized modelling,
and whose recurrence may be viewed as highly implausible.

Figure 6: Analysis Phase of Management System

38

little question as to the certainty of the information used
to make the decision or the expected results. Here the
manager is acting as though things are certain and can use
such “rational” economic tools as formal optimization -
techniques at arriving at prescriptions (King, 1985).
However, a large number of problems facing the farm manager
are of the semi-structured type, where there are probabili-
ties of outcomes other than certainty.

In the case of structured and some semi-structured
problems the concept of management information systems is
relevant. King, using Davis and Olson definition, describes
a management information system (M18) as:

”...an integrated, user-machine system for providing
information to support operations, mangement, and deci-
sion making functions in an organization. The system
utilizes computer hardware and software; manual proce-
dures; models for analysis, planning, control and
decision making; and a database." (King, 1985)

Criticism of M18 concept has centered around the fact it
only supports managerial work indirectly (Keen and Morton,
1978). M15 does not help managers integrate data from a
number of sources into the context of particular decisions
(King, 1985). Further, a MIS which may include optimization
models, rules of thumb and systematic models (simulation)
may not be useful in unstructured decisions which are unique
and outcomes uncertain.

The concept of a decision support system (088) is a

response to this problem. Sprague and Carlson (1982) define

086 as "interative computer-based systems that help decision

39

makers use data and models to solve unstructured problems."
The USS concept emphasizes help or support for decision-
makers and not a substitute for the decision-maker. Second,
088 are meant to be used directly by the decision-maker.
Third, they integrate both data and models. Finally, 085 are
designed for use in unstructured decision situations (King,
1985). However, the barriers dividing 088 and M18 seem to
be both imprecise and movable.

Further classification of decisions according to the
type of problems and information needs, while it may be
theoretically useful, may not be practically possible. For
instance, Johnson (1977) in reviewing lessons from the
Interstate Managerial Survey, indicated that although infor-
mation can be classified into categories of prices, produc-
tion, new technology, institutional, and human, the solving
of any specific problem requires information on all five
subjects for their solution. This state of affairs again
emphasizes the need for flexibility in a system designed to
aid the decision-maker so that all relevant information can
be processed properly. Adherence to set decision rules
based on limited data could lead to myopic decision patterns
adversely affecting the managerial process.

4.4 Computer Functions As They Relate to
Measuring Value

Computer can be useful in three major areas: 1)Elec-
tronic Data Processing (EDP), 2)Management Information
Systems (M18) and 3)Decision Support Systems (059).

Electronic Data Processing (EDP) consists of observa-

40

tion, classification, and storage of data usually producing
standardized reports, i.e. balance sheets, herd status, etc.
The criteria for evaluating such a system consist of
comparing the new system with the previous system in the
areas of: 1) average processing time, 2) labor loading and
3) equipment cost (Keen and Morton, 1978).

Management Information System is classified in the area
of Operations Research/Management Science, where the
emphasis on structured problems where the objective, data,
and constraints can be perceived (Keen and Morton,1978).
Measurement criteria include:

1) profitability

2) application to major problems

3) quality of decisions

4) user satisfaction

5) widespread use (Ein-Dor and Segev, 1981)

Type of software included in Management Information

Systems include:

1) Linear Programming

2) ration balancing

3) lease/buy decision

With Decision Support Systems, the impact is on semi-
structured problems where the computer can be used as an
analytical aid, but where management judgement is essential
(Keen and Morton, 1978)

Measurement criteria center on extending the range and
capability of managers’ decision processes to help improve
their effectiveness in terms of:

1) Decision outputs

2) Changes in the decision process
3) Changes in managers’ concepts of the decision

41

situation
4) Procedural changes
5) Classical cost/benefit analysis

6) Service measures
7) Managers’ assessment of the system’s value
8) Anecdotal evidence (Kleen and Morton, 1978)

Type of software commonly classified as DSS include:

1) simulation modelling
2) electronic spread sheets
3) market forecasting (charting)
4) statistical analysis
5) management games
6) teaching modules
As can be observed the measurement criteria for EDP and
M18 are more easily quantified than 088. These differences
will be explored more and are essential for isolating proper
measurement criteria and emphasize the need for behavioral
analysis especially in the case of 088.
4.5 System Definition
In this section, a more specific model of both the
management system and management evaluation system is
presented. The general systems model was presented earlier
where the “Management System“ was identified in terms of its
role as controller of the “Farm System” through the manipu-
lation of inputs. The function of a ”Measurement System“
would be to determine how the firm’s system may be affected
by improvements in the manager’s decision-making system.
The structure in which the Measurement System must
operate is described in Figure 7. The System that we must

develop is designed to measure the impact of improved infor-

mation and planning tools on the behavior and effectiveness

42

Envi ronment 4‘
. Cooperation

. Participation I
Macro policies

Learning curve I

 
 
 

 

 

 

 

Controllable
Inputs System Outputs

 

 
 

     
 
 

 

1.Decision aids Measuring impact of 1. Mangerial
2. Information improved information indeces

 

 

 

 

4. Incentives mangerial behavior
3. Criteria for

 

 

3. Training 1" 2. Knowledge of

 

decision aids

 

Parameters

1. Type of farm

2. Goals of farmer

3. Time

4. Funds for research

 

Management (System
Users)

Farm managers
Software developers

1

2.

3 . Pol icy makers

4 Management professionals

 

 

Figure 7: Measurement System

43
of the farm manager. The "managers" or primary users of the
results of the measurement system would include computer
users, software developers, policy makers, and managerial
professionals. These individuals would supply the decision
aids, information, training, and incentives necessary for
the system to function.

Beneficiaries of such a procedure will be farmers who
may make the decision to purchase a computer and who may be
interested in furthering their understanding of management.
Software developers would be interested in knowing the
extent of possible adoption of the technology and may obtain
ideas regarding software needs of potential adopters.
Management professionals would be interested in both
decision-making behavior and the relationship between
behavior and performance. The management teacher could
find the approach useful in focusing in on the most crucial
areas for managerial success. Policy makers would be
interested in the possible implications of the adoption or
non-adoption of the technology and the role of government
in such adoption.

However, the success of the system in producing tang-
ible results of measured changes in effectiveness measures,
information on how decisions are made, managerial indices,
and criteria for further decision aid development is limited
by outside factors such as cooperation between institutions
and personnel dealing directly with the farmers, participa—

tion by the farmers, macro- policies and the learning curve.

44

Other parameters that will cause modifications in the system
include: the type of farming system analyzed, the goals of
the farmers, time and funding sources. Therefore, these

factors must be concidered while developing a useful tool.

Chapter 5
Problem Formulation:

Purpose of Measurement System

Problem Formulation involves defining the specifica-
tions for models which would address the real problem. A
workable problem statement would indicate how the results of
the model are to be interpreted and implemented.

In an attempt to aid decision-makers we first need a
better understanding of the behavior of decision-makers
along with the constraints to their decision resources in
order to develop decision aides and educational programs.
Although research has shown that "good" decision makers are
“rational", the question still remains as to how many of
tools developed and advocated by agricultural economists and
other management professionals are being used by decision-
makers and if being used, how effective are these tools.
This research is an attempt to define the constraints to the
use of such tools with implications to their potential value
or needed modifications.

The purpose of developing a measurement system is for
purposes of comparisons. Measurement requires definition,
isolation and scaling. In defining the role of a business
or orgainzational entity Camphell, Dunnette, Lawler, and

Welck (1970) have distinguished among the concepts of

45

46

behavior, performance, and effectiveness. Behavior is
simply what people do in the course of work. Performance is
behavior that has been evaluated (i.e. measured) in terms of
its contribution to the goals of the organization. Effec-
tiveness refers to some summary index of organizational
outcomes for which an individual is at least partially
responsible such as profits, debt ratios and productivity.
Most of the performance measures used by finance or business
analyst are considered effectiveness measures by these
definitions, while the behavioral aspects of performance are
not generally concidered. In other words performance deals
with how a manager gets the firm to a position that can be
evaluated in terms of effectiveness measures.

For our purposes an even clearer distinction between
behavior, performance and effectiveness can be made. Figure
8 demonstrates how the isolation and measurement of the
management system is dependent on where one monitors in
relation to the throughput of information. The manager uses
positive information from the outputs of his farm system
along with his values and goals to formulate a prescription
for further action. Management behavior or decision-making
includes problem definition, observation, analysis and deci-
sion. To determine the effects of an improved decision-
making system one can measure changes in the behavior of the
manager which consists of measuring at point 1. One could
compare differences in decisions made or prescriptions

produced by the management system (performance), point 2.

47

 

 

MEASUREMENT ‘
SYSTEM

(2) (1) (3)

 

 

Management System
(Decision Making)

Problem Definition

 

 

\Desired Output

 

 

 

Prescription Observation

 

Analysis k Undesired

Output

 

 

Decision

 

(1) Decision Making (Managerial Behavior)
(2) Prescriptions (Managerial Performance)
(3) System Outputs (Managerial Effectiveness)

Figure 8: Isolation And Measurement Of Managment System

I
I
I

48

One could also measure changes in actual conditions of the
farm system (effectiveness), point 3.

The distinction between behavior, performance and ef-
fectiveness is important for our investigation since in the
short run a manager’s behavior may change without immediate
results being measured by performance or effectiveness of
the organization. That is, many factors outside of the
manager’s immediate decisions and actions may substantially
effect the outcomes of his farm system.

Therefore, to thoroughly evaluate the benefits of an
improved decision-making system we should attempt to measure
both how the decision-making behavior and performance of the
manager is changed as well as measure relevant system per-
formance criteria (effectiveness measures). While previous
studies have somewhat standardized the performance measures
used to evaluate the physical and financial status of a
farm, less attention has focused on the difference between
“right“ and "wrong” decision-making behavior. Furthermore,
the long time lag between action and result in agriculture
makes it difficult to determine improvements in effective-
ness measures unless considerable time has elapsed.
However, if a combination behavioral/performance oriented
measurement system could be developed that would give some
indication of possible future effectiveness, such a mea-
surement system or test could help in evaluating decision

support tools and educational programs designed to aid the

manager.

49

Any measure of changes in management due to information
must satisfy the criteria for scientific objectivity.
Therefore, any measurement concept must be evaluated with
respect to internal and external consistency, clarity and
workability (Johnson and Zerby, 1973). Internal consistency
(coherence) is a logical or analytical matter where a set of
concepts must bear logical relationship to each other. The
external test of consistency (correspondence) is the test of
experience. In order to apply the test of correspondence an
existing concept is compared with concepts based upon new
experience. A concept passes the test of clarity when it
can be communicated between people without ambiguity and
vagueness. The test of workability is a pragmatic test of
usefulness (Johnson and Zerby, 1973).

The major thrust of this research is to clarify and
measure the concept of management as it relates to farm
decision-making, to identify the areas of managerial
effectiveness that can be measured, and to identify areas of
managerial ability that can be tested.

While performing the primary tasks involved with the
development of the measurement system, other products will
be produced which may be equally important. Subject matter

information will be generated such as:

1. Validated managerial effectiveness measure(s),
2. Common behavioral attributes of “successful" (effec-

tive) farm managers,

50

Levels of use of economic tools by farm managers,
Information on the relationship between behavior,
formance, and effectiveness,

Criteria for decision aid selection for cash grain

farmers.

per-

Chapter 6

Generation of System Alternatives

This phase of the analysis outlines alternative ways of
solving the measurement problem focusing on how other
researchers have attempted to measure the value of informa-
tion. The chapter begins by developing a structure under
which these studies can be categorized and then reviews the
literature in the context of this classification scheme.

6.1 Measuring Behavior, Performance or
Effectiveness

The value of information is related directly to how
management is defined and measured, since information is an
input to the decision-making phase of the management
process. Information can be regarded as data in a form that
is of real or perceived value for making current or prospec-
tive decisions (Davis, 1974). According to Goldman (1953:
2), “...the amount of information in a message should be
measured by the extent of the change in probability produced
by the message." Fuller (1982) asserts that information is
valued only as it affects behaviour. Forster (1978) claims
that the value of information increases with the riskiness
of the decision to be made, and the value of the expected
gain or loss from that decision. These different schools of
thought stem from the issue as to whether management is a

direct input to production, a factor that must be measured

51

52

only as a residual entity, or measured as a separate
function.

Going back to Figure 8, we can classify past attempts at
measuring the value of information into categories of
1)Measuring Improvements in Managerial Behavior, 2)Measuring
Improvements in Managerial Performance and 3) Measuring
Improvements in Managerial Effectiveness. How one defines
the management function, and where one decides to measure
changes in that function, opens the concept to criticism

by one or more of the tests for objectivity.

6.2 Measuring Improvements in Managerial Behavior

Revisiting Figure 8, one may decide to measure manage-
ment within the "Management System ", where one would
compare managers’ traits, levels of managerial knowledge, or
management practices. Such an analysis was done by Vogel
(1981) where he compared management practices of small and
large farmers in southern Illinois. He found that on a
percentage basis large farmers used more professionally
accepted management practices than small farmers in respect
to cultural practices, financial management, and marketing
tools used. There are problems with using an index of a
farmer’s knowledge as a measure of management based on
threats to coherence in that: 1) one may not be able to
distinquish between knowledge and entrepreneural logic, 2)
it is difficult to distinguish between management ability

and management input, and 3) the inclination for the re-

53

searcher to incorporate subjective elements in the index
(Heady and Dillon, 1961). The underlying implications to
the value of information is that the "more information used
the better“ or the ”more certain practices used the better“.
The ”better“ being the subjective evaluation of the re-

searcher as to what is the "right” or "wrong“ way to behave.

“.l]. ! 2

Another method of determining the value of information
in decision-making would be to investigate farmers’ will-
ingness to pay for information services. One could keep
track of how much farmers pay for information and processing
equipment, along with assigning an opportunity cost for the
value of time spent in collecting and processing data for
decision-making purposes. Of course it could be argued that
farmers do not allocate the optimal ammount of resourses to
decision-making, and much of the information may be
redundant. However, redundant information may hold value in
that it may increase the confidence of the decisions made.
This improvement in confidence could be of great value to
decision-makers faced with a number of uncertainties.

Although the author has not found studies which

measured the amount paid for information, this method could
give some indication on the value of their present informa-
tion resources as compared to the value of other productive
inputs. However, the willingness to pay would be an
hindsight measure determined after adoption of the

technology. The value obtained in this manner can not be

54

extrapolated to represent how farmers would value an
improved decision-making system with substantially different
attributes.
6.3 Measuring Improvements in Managerial
Performance

One can also compare managers in terms of the prescrip-
tions or decisions which they produce. In practice it may
be very difficult to actually see a prescription since this
decision may only exist in the manager’s head. What is
observed is activity based on that decision and the final
result of such activity demonstrated in the farm system
outputs.

Within the teaching or training arena we use tests or
questions which we have already determined the "right“
answers to in order to test the student’s ability. However,
these questions must necessarily be abstractions from
reality. It is hoped that scores on these tests can be a
good indication of future managerial performance since per-
spective employers of these managerial students look at the
grades from such tests in the selection process. Unfortu-
nately not many real world problems are as structured as the
problems presented in such tests.

Rohrbach (1983) conducted an experiment comparing the
performance of graduate students in agricultural education
on a standardized test. The object was to evaluate the
effectiveness of computer aided instruction (CAI) in compa-

rison to conventional instruction. Rohrbach’s study indi-

55

cated ”...that the lecture-discussion method of teaching was
superior to the microcomputer-assisted technique in teaching
the types of principles and concepts under the conditions
described.“ The areas taught under the two methods were
principles and methods in cost recovery and investment
credit. Rohrbach admitted that his findings were not in
line with other evaluations of computer aided instruction,
however his findings would indicate that CAI may not be

appropriate in all cases.

5' I l' E .

Kleijnen suggests the simulation or gaming approach as
another method of measuring managerial ability. According
to Kleijnen (1980) ”...simulation can be characterized in
the terminology of information economics as: dynamic (in-
cluding feedbacks and feedforwards), non-linear, nonzero-sum
game,...complex management processes, and multicriteria
satisficing behavior based on imperfect knowledge.“

Debertin gt_gl. (1975; 1976) used the simulation gaming
approach to assess the value of information and feedback to
experimental managers of laboratory corn and soybean farms.
The studies used both college senior undergraduates and
farmers as subjects. Although the authors found a signifi-
cant return to information to both groups, the returns as
measured using students was higher than when using farmers
(Debertin gt_3L., 1976). This prompted the authors to point

out areas of difficulty with the simulation gaming approach

56

to valuing information including: 1) the higher response to
information by students might reflect marginal returns to
education, 2) the differences in student and farmer behavior
with the simulation model could mean that students were not
better managers but better "players“ of the game, 3) the
information used in the experiment and the simulation model
itself might not be representative of the real world, and 4)
students might be willing to accept the information given at
face value, whereas farmers may rely more on their own real
world experience.

E E 5 I E 'II'

The Smith and Kendall (1963) pproach to performance
scale development consists of getting a group of managers to
name and define the major components of performance for the
job in question. Using these definitions as guides, the
participants then are asked to describe specific behavioral
episodes that illustrate both effective and ineffective

performance (i.e. critical incidents) within each of the a

Q

priori factors.

The advantage of this approach is that the researcher
does not use his own normative judgement as to how the
manager should behave, but depends on the managers them—
selves to determine what is "right" or "wrong".

6.4 Measuring Improvements in Managerial
Effectiveness

This school of thought assumes that the value of infor-
mation should be evaluated based on performance (effective-

ness), where changes in system performance is the value

57

added by the managerial input once other production inputs
are accounted for.
W211
The Bayesian approach to measuring the value of infor-
mation assumes that managers are "Bayesian“ decision-makers
maximizing expected utility or expected returns. Perrin
(1976) paraphrases Schlaifer to say that
“...the value of a bit of information is the difference
between the decision maker’s current expectations of
(a) the payoff value that will result if he chooses his
act as well as he can without obtaining the informa-
tion, and (b) the payoff value that will result if he
were to obtain the information and then choose his act
as well as he can. In other words, the value of the
information is the increment in expected payoff that
can be realized by utilizing the information in
making the decision." (Perrin, 1976)
Perrin uses this method to obtain the value of soil test
information in corn response in Brazil. He concluded that
soil test information was quite valuable. However, he notes
that the ”... value measured is compared with the alterna-
tive of knowing nothing about the piece of land.“ (Perrin,
1976) Since farmers know a great deal about their land, one
would expect the net value of soil tests to be smaller than
measured.

Kleijnen (1980) goes further to critique the usefulness
of the Bayesian or information economics approach. He states
that much ”...information is needed as input to decision
calculus: one should know all the possible actions, all

possible states of nature, and the corresponding

payoffs...The decision maker is supposed to be purely

58

Latigngl, and maximize expected profit or, more generally,
maximize expected utility...Even in static models, the com-
putations may become tedious, and pose difficult mathemati-
cal problems...In dynamic models the mathematical problems
become tremendous...The models remain highly simplified...it
ignores behavioral attributes of information."(Kleijnen,
1980)

E I I. E I.

Kline (1963) argued that management is indeed a
separate input to production and can be measured like other
inputs to production. Some scientists like Welch and
Pudasaini have argued that increases in management ability
through education can be measured as changes in the marginal
value product of management in a Cobb-Douglas production
function (Pudasaini, 1983). Johnson (1963) argued that
management does not produce products other than decisions.
Heady (1963) argued that management and capital are almost
perfect complements, and therefore can not be separated

within the production function.

E 'l I E I .
Advocates of farm budget analysis and cost/benefit
analysis use the residual or net return approach to
measuring managerial ability. Nielsen and Crosswhite
(1959) used increases in net farm earnings as a measure of
project benefits when comparing Michigan farmers that were

and were not involved with an intensive township extension

59

program. However, there can be substantial threats to the
validity of such a measure. Other forces outside of the
manager’s control could effect net returns. Further. if a
program or project is designed to help one aspect of the
farm condition. other activities of the farm manager not
influened by the project could bias such an overall measure.
The constraints of the manager must also be taken into
account before he can be judged as to levels of success.
W

Rosenberg and Grey (1962) took the experimental approach
in assessing the value of market information on the varia-
tion in prices received by alfalfa hay producers.e Henderson
(1979) points out that although the controlled experimental
approach may be a reliable and useful technique. the politi—
cal consequences of depriving a set of decision-makers with
respect to information would generally preclude the use of
this approach. Further, even under experimental conditions
one has not solved the problem of differentiating between
decisions and their consequences. In other words. by
measuring consequences one measures the desired and
undesired ouputs of the system which is a function of both
the management system and constraints imposed by the
specific parameters of each farmer’s system. along with the
influences of exogenous environmental factors which may
alter the intended outcomes of decisions.

lllll'lfl"! l

A hybrid method of determining the potential value of an

60

information system was attempted by the authour in an
unpublished study in which he compared what a farmer would
have done over a five year period using his present decision
strategy and information system verses what he would have
done given an "improved” decision system. Although the
method used actual past data and attempted to simulate
actual response, this approach can be critisized based on
the assumption of static decision strategies. Shackle
(1961) argues that as"...soon as we permit time to elapse we
must permit knowledge to change, and knowledge cannot be
regarded as anything else“ (p. 41). He further asserts that
there "...is no assurance that anyone can in advance say
what set of hypotheses a decision maker will entertain
concerning any specific act available to him. Decision is

thought and not merely determinate response” (Shackle. 1961:

6).

PART III
MODELLING AND IMPLEMENTATION DESIGN

This part of the analysis will include the modelling
and implementation design phases of the systems approach
methodology. Modelling has as its input the feasible alter-
natives of Feasibility Evaluation and has as its output the
broad specifications for the "Measurement System" to be
implemented in the real world. The modelling process
includes the selection of the specific concept to be
modelled and the modelling itself. In this case the model
_will be procedures or steps to take in evaluating the value
of planning tools in farm decision-making.

Implementation design takes the general model developed
in the modelling phase and completely specifies the details
of the system. ”Completely specify" means developing a
complete set of instructions that will lead to operationali-
zation of the system. This would include specifying the
necessary preparations for implementation including data and
budgetary needs, along with the procedures for accomplishing

specific functions.

61

Chapter 7

Concept Selection

This chapter will define the theoretical concept
selected that will govern what will be measured in the
”Measurement System". The concept selected will be
tentative -- awaiting finalization upon the forthcoming

implementation and feedback.

7.1 Concepts of Management and Information that
Guide the Present Study

Management consists of planning the activities for a
system. monitoring and controlling those activities. and
finally evaluating system perfomance and learning from those
results. In this case, the system is the farm which con-
sists of production equipment and facilities. land, labor
and capital. The planning function includes deciding on
activities that affect both the short- and long-run perfor-
mance of the system. Monitoring and control are on-going
processes of decision-making as the manager tries to keep
his plans on track facing disequilibria created by the
environment and his own misconceptions of that environment.
How the system is evaluated depends on the goals and values
of the farm manager. his family and the surrounding com-

munity in which he must interact. The learning process

62

63

happens over time which forces adjustments in even the way
decisions are made.

Although monitoring and control are important activi-
ties that the manager must perform. it would take extensive
monitoring of the manager’s behavior over the course of the
planning horizon to determine how well he/she was doing. In
other words, the monitoring and control activities deal more
with the ability of the manager to carry through plans made
and may be considered more of an "art" where quick judgement
may be more appropriate than analytical processing.

However. the relative importance of planning in comparison
to control could vary according to farming systems. For
instance. the success of a dairy farm may be more dependent
on day-to-day recordkeeping and control than a cash grain
operation. These possible differences emphasize the need
for interaction with the farm managers themselves to get a
feel for what is important before developing programs to
improve management and instruments to evaluate management
success.

The ability to learn is an important attribute that a
manager must have. However, learning takes time and
requires feedback from the operating farm system. Since
farming is a biological process. the time required for
feedback on decisions made would make an evaluation of the
manager's increase in knowledge difficult to quantify except
in a structured learning environment. However, the

different psychological traits of managers should influence

64

how fast they learn, which techniques they choose to learn
and which methods may be most appropriate in introducing a
new technique such as using a computer to help make

decisions.

7.2 Elements of The Decision Function

Since the major goal of this study is to develop a
model of procedures to determine the potential value of
planning tools in farm management. the management system or
decision-making system of the farmer must be analyzed. The
outputs of this system are decisions. Decisions can be
grouped into categories of production. marketing and
finance. In turn, decisions made in these areas can be
expected to affect the performance of the farm business.
However, due to factors outside of the immediate control of
the manager, there is not a one-to-one relationship between
decisions made and results. Complicating the matter is the
fact that the objective function of the farmer may not be
known, making evaluation of performance by an outsider dif-
ficult. However. we may state that:

Effectiveness = s(production, marketing. and financial
decisions; ability to follow through;
quality of resources; environment;
and random factors)

Here ”s" represents the system’s function of trans-

forming these inputs into outputs which can be used to

evaluate the effectiveness of the farm manager. Basically.

65

system performance is dependent on the decisions the manager
makes. his ability to carry out those decisions and factors
that affect that system which are beyond the control of the
manager. The ability to follow through is related to both
motivation and skill. This skill factor can be considered a
component of the "art” of management discussed earlier.

As stated earlier. decisions are prescriptions which in
turn are products of the management system. A convenient
way of representing the relationship between decisions and
resources used by the decision-maker is:

Decisions = r(Information. Processing. Analysis, Time:

Constraints)

Information can be considered as categorized pieces of data
organized as factors or inputs to making decisions. Pro-
cessing is the storing and transforming of these factors in
a form appropriate to the particular decision. Analysis
involves applying different techniques to the information.
Each decision requires a certain amount of time between
realization of the need to make the decision and the final
derivation of an answer. The decision-maker is constrained
by the availability and quality of the different inputs to
the decision function. The decision-maker is governed by
his own decision rule. ”r“. which specifies the amounts of
each input to use toward the production of a prescription
for action. Of course. the adequacy of the decision rule
depends on the ability of the decision-maker to determine

the possible alternatives and their respective probabilities

66

and outcomes.

It is hoped that although each farmer may use a dif-
ferent methodology to solve his particular problems, there
might be common elements that may distinguish the more
successful farmers from the rest. Of course, the more
interesting question may not be what the decision rules are.
but how the farmer arrives at that rule. Therefore, the
decision rule itself is a function of even more primary
elements represented by:

Decision rules = g(Goals, Values, Personality, Philosophy:
subject to Training, Experience.
Motivations)
Here the decision rule chosen is a function of how the farm
manager views the world and his place in it.
7.3 Differences in Farming Systems as They Affect
Management

As stated earlier. the "Farm System” transforms inputs
into both desirable and undesirable outputs with the manager
evaluating the "goodness" or ”badness“ of those outputs and
formulating new prescriptions. However, farms are not homo-
genous in their outputs. inputs or means of accomplishing
the desired goals. The USDA has classified farms under
broad categories of cash grain, field crop, vegetable and
melon, fruit and tree nut. nursery and greenhouse, dairy,
poultry and egg. and cattle/hog/sheep. Each of these broad
farm types requires a different set of physical, human and
capital resources. The human resources would differ as to

skill levels and managerial responsibilities since each type

67
would be faced with a different set of problems.

Since a decision support system is designed to aid
decision- making. it would require substantially different
systems to satisfy the decision-making needs of farmers
faced with such a wide variation in farming systems. For
instance. a dairy farmer would need a day-to-day evaluation
of herd performance to ensure maximum returns. while a cash
grain farmer may have considerable stretches of time in
which no physical measure of crop performance can be
evaluated. Therefore. determining the potential value of
computers in farm decision-making requires a system-by-
system analysis if not a farm-by-farm one. How narrow or
broad such systems are chosen will determine the reliability
of measures developed while evaluating the decision-making

system.

Chapter 6

Modelling of Concept

This phase of the process will involve developing a
methodology which combines the essential parts of other
attempts to evaluate the value of information into a struc-
ture which satisfies the perceived demands on such a meas-
urement system. The general parts based on function will be
developed here leaving the specifics of how to implement

such a strategy given in the Implementation Design phase.

6.1 Measuring the Management System

In the previous chapter the System was defined in terms
of measuring the impact of improved information (Figure 7).
This Measurement System could include procedures to measure
changes in managerial behavior. managerial performance.
and/or managerial effectiveness (Figure 6). In terms of
elements of the decision function. comparing changes in
managerial behavior would include eliciting and comparing
decision rules or how decisions are made. Determining im-
provements in managerial performance would include comparing
the decisions made by managers. Using changes in effec-
tiveness as a measure of improved management would include

measuring the results of decisions on an operating farm

68

69
system.

The element of time plays an important part in deter-
mining the results of decisions and how the decisions are to
be classified. Decisions can be categorized as short.
medium and long run. or in the language of the systems
approach. decisions can affect the controllable inputs. or
decisions can determine and change the design parameters of
the system. The agricultural production process will
determine the time frame that is relevant in terms of what
can be manipulated in the short run. such as the amount of
fertilizer to apply. However. when one makes investment
decisions. the effect covers the long run and sets the
design parameters or basic input/output relationships for
that system. Therefore. in evaluating the results of a
decision. proper attention to the time frame over which that
decision is effective is important for relevant measurement
of system performance.

In an earlier section. previous attempts at measuring
the value of information were categorized in the three
areas: behavior. performance and effectiveness. Each
methodology had it advantages; however. it may be that no
one method is the best for all types of farming systems or
decision environments. One type of decision environment
could best be aided by changing behavior that might not be
immediately measurable in terms of effectiveness. while the
results of another type of decision may be easily traced by

effectiveness measures. Therefore. what is suggested here

70
is the development of measurement system that is flexible
across a broad spectrum of farming systems but in its use
allows the managers (researchers. software developers. etc.)
to develop concrete sets of evaluation criteria depending on
the uses for the measure(s) of value.

Measuring the value of planning tools and information
through comparative analysis can be accomplished by cross-
sectional analysis or comparisons over time. Most of the
behavioral and performance based methods of evaluation are
based on a cross-sectional comparisons while the element of
time is included when using the residual. experimental and
hypothetical decision methodologies. Cross-sectional
comparisons have the advantage of perhaps being quicker
while increasing the potential for specification errors
since some level of homogeneity is assumed across the
measurement frontier. On the other hand observing the
improvement of the same farmer over time eliminates the
question of non-comparability. but the element of time makes
the data collection and observation process more lengthy.
Also. over time. other elements besides the experimental
treatment have time to creep in reducing the ability to
isolate the effect of the treatment.

Therefore. an approach combining both cross-sectional
analysis along with observations over time could help
isolate the real effects of information. In particular. if
an instrument could be developed that measures behavior and

performance and is a good predictor of effectiveness. we

71

could then use the instrument for measuring probable changes
in effectiveness based on changes in behavior and then
follow these same managers whose decision-making behavior
have been altered over time to test their actual perfor-
mance. In essence. we would then have two separate measures
of the same entity which would allow us to make more conclu-
sive evaluations of the resulting measures.

If the values obtained using the management index are
similar to the values obtained in following the managers
over time. the value of decision aids obtained using the
managerial index approach could then be used as a quick
estimate. However. due to changes in the importance of
problems facing the farmer. no index can be expected to
retain its appropriateness over a long period of time.
Therefore. the adjusting of the index and its ability to
predict would have to be done periodically requiring the
time consuming process of observing another set of managers
over time.

Furthermore. no attempt will be made to measure the
value of an individual piece of information because
decision-makers use a package of information for making any
decision and no decision is made in a vacuum separate from
other decisions. Since it is an information system that a
farmer uses while managing his operations. it will be an

information system that is evaluated.

72

6.2 Model

Figure 9 is a representation of the general framework
of the model or procedures involved in determining the value
of an information system or set of decision aids. This
model is labelled Management Systems Research (MSR). The
Management Systems Research model is defined by the author
as a methodology for measuring improvements in the decision-
making system. where managerial behavior is the subject and
increased effectiveness the object. The decision-making
system includes the decision rules. decision aids. data
processing equipment. and both internal and external data
used to make decisions.

The first step involves the selection of the farm type
to be considered since decisions and therefore aids to such
decisions would be radically different across farming
systems. The process splits into two phases that can be
carried on simultaneously. One phase deals with the identi-
fication of common behavioral incidents of successful
farmers and from their responses. a behavioral instrument
designed to predict effectiveness is developed. On the
other branch. effectiveness measures are selected and tested
for scientific objectivity. i.e. coherence. workability.
internal and external validity and reliability.

Subsequently. the behavioral instrument is tested to
see if it represents a good index in terms of its predictive
ability of system performance (effectiveness). Once the

managerial index is validated it can then be used as the

73

 

Selection
of
Farm System

i . J 7

 

 

 

 

 

Behavioral Analysis Measuring Effectiveness
(Decision Making) (System Performance)

 

 

 

 

 

Testing Behavioral‘
4. Predictors of
Effectiveness

J

Quasi-Experiment

Value of Improved
Decision Support System

 

 

 

 

 

 

 

 

 

 

 

Figure 9 : General Framework For Measuring the
Value of an Improved Information System

74

basis for testing farmers in a quasi-experiment designed to
measure changes in managerial behavior. performance and ef-
fectiveness before and after the introduction of comput-
erized decision aids. Finally. the result of the actual

experiment will be a measure of the value of the set of

decision aids chosen for analysis.

Chapter 9

Model Specification or Implementation Design

The Management Systems Research framework serves as the
general model of the procedures. Now each component will be
broken down to reveal the specific steps to be followed in
measuring the value of decision aids. “Selection of Farm
System" is depicted in Figure 10. “Behavioral Analysis“ in
Figure 11. “Measuring Effectiveness“ in Figure 12. ”Testing
Behavioral Predictors of Effectiveness" in Figure 13. and

“Quasi-Experiment“ in Figure 14.
9.1 Selection of Farm Type

Selecting the specific type of farming system to be
analyzed (Figure 10) begins with identifying the universe
under consideration. How we categorize farm types may
depend on whether we are interested in a national or state
sample. The important enterprises and combination of enter-
prises may vary substantially across states. A combination
of tobacco and vegetable production may be common in North
Carolina. but rare in Michigan.

As mentioned earlier. the USDA categories of cash
grain. field crop. vegetable and melon. fruit and tree nut.
nursery and greenhouse. dairy. poultry and egg. and cattle-

Ihoglsheep may be a good starting point for classification.

75

76

Farmers

l

Categorize According
to Farm Type

1

CFarming Sys (939

 

 

Select Farm
Type

l

[/5 Farm Type‘\
—\\_(Cash Grainlj

 

 

 

 

 

 

 

v

to Behavioral to Measuring
Analysis Effectiveness

Figure 10 : Selection of Farm Type

77
However. these categories may need modifications depending
on the scope of the research. In general. the broader the
categories the less precise will be the measure of value due
to the non-homogeneity of the decision environment.
However. with very narrow categories the measure may have
limited usefulness. The tradeoffs involved would depend on
the purpose and use of the measure.

After the farming systems or types have been iden-
tified. then a type for research can be chosen based on
either the need. familiarity with that farming system. the
importance of that farm type to the economy. or just arbi-
trarily selected as the first to be investigated. Both
research and extension should be involved in determining
relevant categories and selecting the specific type of
farm for further analysis. Once the specific farm type is
selected that information is then used in two simultaneous

processes - Behavioral Analysis and Measuring Effectiveness.

9.2 Behavioral Analysis

The output of the behavioral analysis phase is an
instrument or test that can be used to predict managerial
effectiveness. By observing the decision-making behavior of
the successful farmers. critical incidents of common
behavior can be detected that can be used as indicators of

SUCCESS .

Once the farm type has been selected (Figure 11). means

78

Farm Type
(Cash Grain)

J,

J

Jr

A

1 .

 

Develop Model
of Decision

 

Select set of
Successful Farmers

 

 

 

System

Set of Major
Decisions

Design Instrument
to Determine
Behavioral Incidents

_ l

‘_....._._

 

 

 

 

Set of
Farmers

L__

Successful >

 

l

 

(Open-ended Questionnaire:>_______~

1

 

Evaluate D-M Behavior
of Successful Farmers

 

 

 

 

 

 

/1Decision-Maki

1\¥#Behavior

to Select

Decision Aids‘L

:19

 
     

Evaluation
Criteria

_‘_

 

 

Develop Instrument
to Measure Behavior
and Performance

 

Instrument
(Index)

 

 

i

To Testing Behavioral Predictors of

. Figure 11

to Test for
Validity
of Effec-
tiveness
Measure(s)

Effectiveness

: Behavioral Analysis

79

for evaluating the decision-making behavior of successful
farmers are prepared. This involves both developing an
instrument to question farmers and identifying "successful"
farmers.

The term "successful" farmer can be ambiguous depending
on the criteria selected and who is doing the selection.
There may be a set of farmers identified as ”successful”
based on more or less objective measures of effectiveness
such as net income. The Extension Service in each county
may have an available list of "successful" farmers. Farmers
Home Administration. ASCS. Farm Bureau and individual
commodity groups may each have separate lists of what they
consider the more "successful" farmers. Without having
access to extensive yearly records and an agreed upon
criteria. any selection process may be biased since the
sample is based on participation in programs sponsored by a
particular agency. If a cross sample of lists from dif-
ferent agencies within the same county was developed. the
names chosen in common could be an acceptable list of the
better farmers. However. since tendencies and not statisti-
cal accuracy is demanded in this phase. selection by the
Extension Service may be sufficient.

To develop an appropriate open-ended questionnaire for
soliciting behavior. a set of the major decisions made by
the farm type must be developed. Information for such a set
could come from previous studies on decision-making. exten-

sion personnel. or by observing a set of farmers. The

80

questionnaire is then developed around these key decisions
to obtain information on how the farm managers solve these
problems. The questionnaire is open-ended to allow for as
much farmer input as possible without leading them into
preconceived ideas about how they should make decisions.

This evaluation process can produce a clearer under-
standing of the decision-making behavior of farmers along
with getting a feel for what they use as criteria for
evaluating effective management. Information on decision-
making behavior can then be used to develop a more struc-
tured instrument to act as a test or managerial index
designed to differentiate between ”good" and "poor" manag-
erial behavior and performance. Since the function of
decision aids is to change managerial behavior. information
on decision-making behavior is then necessary for selection
of proper aids.

Knowing the evaluation criteria of farmers for defining
a successful manager is essential for testing the validity
of an effectiveness measure. Therefore. one output of this
system goes directly into the next phase designed to develop

valid and reliable measures of managerial effectiveness.

9.3 Measuring Effectiveness

This phase is designed to select and validate measures
of managerial effectiveness. The major input in this phase
is the farm type selected for analysis (Figure 12) and the

Output would be an effectiveness measure or set of measures

81

Farm Type
(Cash Grain)

1

Find Set of Farmers
With Historical Records

 

 

 

 

 

 

to Administer -1 ~<Farmers with Records)
Instrument

 

 

Select Effectiveness Measure(s)
From Records
(Theory and Workability)

 

 

 

 

(Effectiveness Measure(si)

 

 

from Evaluation . Test for Validity
Criteria (Internal & External)

L

(Validated Measure(si)

 

 

 

 

 

Test for Reliability
(Consistency)

Validated and Reliable
Measure(s)

to Testing Behavioral Predictors of Effectiveness

 

 

Figure 12 : Measuring Effectiveness

82

that had been tested for coherence. workability. validity
and reliability.

A major problem with selecting an effective measure is
identifying a set of farmers with historical records that
are both compatible and complete enough for further
analysis. The selection of an effectiveness measure is
dependent on the theory that identifies the measure as valid
and the practical problem of measurement. Of course. the
measure chosen will depend on the type of farm analyzed and
on whether the analysis is to be a cross-sectional compar-
ison or a comparison over time.

Once a measure is chosen it must be further evaluated in
terms of validity and reliability. Since one set of pos-
sible benefactors of the end product of measuring the value
of decision aids is the farmers themselves. it is important
that the measure be relevant to their value system.
Therefore. information obtained from the open-ended ques-
tionnaire to the successful farmers on how they would
evaluate the effectiveness of another manager would be
important in the screening process of measurement selection.

Other tests for validity could be statistical tests that
would check if the measure was actually measuring what it is
posited to measure. In other words. a measure of managerial
effectiveness should not be dependent on factors outside of
the manager's control such as soil type. closeness to

markets or some other advantage or disadvantage beyond his

control.

83

The measure should be reliable in that one should be
able to replicate the measure and arrive with basically the
same conclusions. Statistically. the test for reliability
could be accomplished by seeing if the measure can classify
managers into levels of effectiveness over an extended time.
since one would expect the relative position of management
abilities to be somewhat constant over time. unless severely
disturbed by concentrated programs to change the level of
management. Once a measure has been tested for scientific
objectivity. it can be used to test the relationship between
managerial behavior and effectiveness.

9.4 Testing Behavioral Predictors of
Effectiveness

This phase is designed to test the relationship between
behavior. performance and effectiveness. If a management
index can be developed that can be used as a predictor of
effectiveness. the stage is set for cross-sectional
evaluation of programs or systems designed to increase the
level of managerial effectiveness.

The first step in this phase (Figure 13) consists of
administering our prototype. or first attempt at an instru-
ment to measure managerial behavior. to the set of farmers
identified as having historical records. Once the instru-
ments have been completed then they must be scored according
to some standard. One method of scoring would be to use

management professionals as the experts and accepting their

84

Instrument from Farmers from
Behavioral Analysis Measuring Effectiveness

 

Administer Instrument to
Farmers with Historical Records

Completed Instruments
(Index)

V l

 

 

 

 

 

 

 

 

 

Score Test based on From Validated
Top Farmers and Reliable
(Effectiveness) Measure(s)
To Select Q iBehavioral Scores)
Dec181on A1ds JL
Test Hypotheses ,
(Effect.=f(behavior.performance))

 

     
      

Are
Predictors
Significant

no no

 
 

to Behavioral
Analysis '

To Measuring
Effectiveness

yes

 

Set of
Predictors

 

 

Revise Intrument]
(Index)

( Managerial Index x)

to Quasi-Experiment

 

 

Figure 13 : Testing Behavioral Predictors of Effectiveness

85

answers as the norm. Another method would be to go back to
the first set of successful farmers. give them the test. and
use their responses as the norm. However. a quicker method
would be to use the effectiveness measure chosen to divide
the farmers into management groups according to prior levels
of farm success. The answers of those farmers in the
highest group based on the effectiveness measure can then be
used as the norm for scoring the instrument.

The behavioral scores would be aggregated according to
production. marketing. and financial decision-making. To
reiterate. behavior refers to how the decision-maker makes
the decision or the process involved in making the decision.
This would include the type of information used. the place
from which the information is obtained. how the information
is obtained. stored. processed and analyzed.

There is a gray area between behavior and performance.
Performance evaluation would require comparing answers to
specific problems. For instance. the farmer could be given
a problem situation and supplied with relevant information
that could be used in its solution. This form of evaluation
is similar to tests given in an academic setting where one
assumes that there is a ”right" solution to the problem.
These performance scores can be aggregated in the same areas
of production. marketing and finance. These scores can be
another set of possible indicators of managerial effec-
tiveness.

As mentioned earlier. behavior and performance are

86

based on decision rules which are functions of the value
system. training. personality and experience. Part of the
evaluation process would include questions on these aspects
of the makeup of the individual managers. We can test
hypotheses that posit a relationship between the behavioral
and performance variables with the psychological. socio-
logical and demographic variables.

Statistical tests for significance that are performed
in testing these relationships will indicate the degree of
reliability of the managerial index as an indicator of
effectiveness. The parts of the index that passed the test
can then be used to establish a revised instrument based on
the set of predictors that are found significant. If no
relationship is found between the items on the managerial
instrument and the measure of effectiveness. one can
conclude that the behavioral and performance criteria were
not appropriate and go back to steps within the Behavioral
Analysis phase. One can also conclude that the effec-
tiveness measure may not be appropriate and therefore go
back to steps within the Measuring Effectiveness stage.

However. one could also conclude that for this type of
farm the decision environment is too diverse and the paths
to success may be so varied that a simple managerial evalua-
tion index is not feasible. Even if this is the case. much
information on the decision-making behavior of farmers would
have been compiled. Much of this information is directly

used in trying to produce a reliable index. however much

87

anecdotal information may not have been used since only one
or two farmers may have made these comments. However. these
comments can he extremely valuable in terms of getting a
feel for the general decision—making environment. In parti-
cular the attitude of the farmers toward certain institu-
tions and programs can help shape future directions for

those interested in aiding farm decision-makers.

9.5 Quasi-Experiment

The previous phases set the stage for actual measurment
of the value of decision aids. The mode of evaluation is
direct measurement via a quasi-experiment. As stated
earlier. measuring the value of information through compara-
tive analysis can be accomplished by cross-sectional
analysis or comparisons over time. An approach combining
both cross-sectional analysis between farmers along with
observing the same farmers over time could help isolate the
real effects of information.

In particular. if we could develop an instrument that
measures behavior and performance and is also a good pre-
dictor of effectiveness. then we could use the instrument
for measuring probable changes in effectiveness based on
changes in behavior and then follow these same managers
whose decision-making behavior have been altered over time
to test their actual performance. In essence. we would then
have two separate measures of the same entity which would

allow us to make more conclusive evaluations of the

88

resulting measures. If the values obtained using the
management index are similar to the values obtained in
following the managers over time. then the value of decision
aids obtained using the managerial index approach could be
used as a quick but reliable estimate.

The basic design for such a quasi-experiment is based
on the ”Nonequivalent Control Group Design" as described by
Campbell and Stanley (1963). In this design. both the
experimental group and the control group are both given a
pretest and post-test. however. the experimental group and
control group do not have pre-experimental sampling equi-
valents as required for a pure experimental design. As
described later the experimental group will be composed of
some type of previously assembled collective or self-selec-
tive grouping. The control group is then selected to best
match the experimental group on relevant characteristics.
Assuming that the control group is properly selected one can
regard this design as controlling for the main effects of
history. maturation. testing. and instrumentation (Campbell
and Stanley. 1963).

The first step in this phase (Figure 14) is the selec-
tion of participants. The type of farmer selected must meet
certain criteria which make random selection. especially of
the experimental group. almost impossible. The selection of
farmers and the appropriate decision aids and mode of
delivery must be iterative to ensure the proper fit. To a

great extent the farmers must be self-selective since parti-

89

 

 

Participation

Select Farmers for

 

 

 

 

(;Participants)

 

 

 

from Mangerial Test and Score
Index Using Index

 

< Predicted Effectiveness)

l

from Behavioral Select Decision“
Analysis - Aids

Observe and Keep Records

 

Qbecision Aids

 

Train a Set of
Farmers in use
of Aids

L
(Trainfd FarmerE>__J

Re-test Both
Groups

 

 

 

 

 

 

 

 

 

 

(ZSets of Scores)

 

 

 

Compare scores

1

 

 

of Both Groups
Over Time

Records for Comparison
Over Time

Evaluate Differences
in Effectiveness as
Value of Decision
Aids

 

 

 

 

 

 

Measured Value of
Decision Aids
(Field Value)

 

Compare Both

 

 

 

Measured Value of
Decision Aids
(Classroom Valued

’4' Measures

 

 

 

(Value of Decision Aids)

Figure 14 : Quasi-Experiment

 

90.

cipation is both optional and will require a lot of time and
effort.

The mode of delivery of the technology will also play a
role. Computers are expensive in both physical and human
capital terms. Therefore. one would expect that there are
few farmers of any given type that have access to computers.
One may find a few larger farms with computers. but very few
smaller sized operations would be expected to have them.
Therefore. to get a sample of larger farmers. a search
procedure would be necessary. To obtain a sample of smaller
farmers a delivery system would have to be installed. Such
a delivery system could be through current Extension
facilities and programs. For instance. North Carolina
Extension has placed computers in all of their offices.
Also. there is in place a program to provide management
assistance to smaller farmers through what is called the
Farm Opportunities Program in which paraprofessionals are
used to work intensively with a small group of farmers over
a three to four year period. If Extension and/or parapro-
fessionals are used then the selection process would have to
focus on the enthusiasm and the abilities of people within
these institutions to facilitate the incorporation of the
managers of smaller farms into the experiment.

Although the self-selection process removes the random-
ness. it does get around the moral issue of providing infor-
mation to one group while excluding others. Of course the

problem still remains of both attrition and the filtering of

91

information between the control group and experimental
groups. Also. the problem arises as to improvements in the
technology over time with the introduction of new decision
aids. For control purposes one would want to stick with one
set of aids; however. this must be tempered with the moral
obligation of giving to enthusiastic managers the best tools
possible.

After the experimental group is selected a control
group must be chosen to match the basic characteristics of
the experimental group as to farming systems. location and
demographic characteristics. Once the experimental and
control groups are selected the managerial index is given to
all participants. The Myer’s-Briggs' Type Indicator (MBTI)
is also administered to determine the effect of cognitive
types on the decision rules chosen. Then farmers from the
experimental group would be trained in the use of comput-
erized decision aids chosen for evaluation. Afterwards.
both groups are retested with the experimental group allowed
to use computers to help answer and solve the problems on
the index. A new set of scores are then available for
cross-sectional and longitudanal evaluation.

Since the managerial scores are used to predict effec-
tiveness. the difference in effectiveness measures can be
used as the expected value of the improved system. There
are several ways of evaluating these scores. One is to look
at the difference between groups at the second testing.

Another would be to look at the increase between testing of

92

the experimental group. Still another would be to take the
difference between means of the effectiveness measure before
and after training the experimental group and subtract from
that the observed difference of means of the control group
for the two tests. This method would control for the
learning process outside of computer training. This measure
would constitute the “Classroom" value of the decision aids.

These same farmers in both the control and experimental
group can then be observed over time and comparisons made
after a number of years. The selection of the time frame
must take certain things into consideration. Too short of a
period could make the measure biased due to the environmen-
tal effects of drastic weather or policy conditions that may
prevail in a given year. Too long of a period could result
in both high attrition rates and technology obsolescense.
Since the Farm Opportunities Program requires participation
from three to four years. four years may be a feasible. if
arbitrary. time frame.

The effectiveness measure chosen would then be compared
between the groups following the same basic procedure
outlined above so that the learning process that would have
occurred without the computer can be statistically elimi-
nated. The difference between means of the experimental and
control groups can then act as the "Field" value of the
decision aids. However. one would expect that the longer
one uses the system the more proficient one would become.

Therefore. it may be worth looking at the difference between

93

the first year and last year of participation as an
indicator of increased managerial ability over time.
However. due to environmental conditions the two years
chosen could be crucial. Although such a comparison could
be informative. there may be too many threats to the relia-
bility of such a measure since its ability to be replicated
would be nil.

The final step in the process would be to compare the
”Classroom" value with the "Field” value to see if they
support or contradict each other. Of course. how well they
support or contradict would be a subjective matter on which
the users of the measure must decide. However. if the
values do support each other then the "Classroom“ value and

approach could be used for further evaluation of decision

aids for these farm types.

PART IV

IMPLEMENTATION AND RESULTS

Implementation is to give physical existence to the
desired system. In the case of this research the system is
the set of procedures for establishing and using a measure-
ment system for evaluating computer benefits to farm
decision-making.

To determine the workability of the system. Michigan
will be chosen to test the methodology. Since the total
procedure is quite lengthy and cumbersome. the phases of
the system will be looked at separately. The phases include
Selection of Farm Type. Behavioral Analysis. Measuring Ef-
fectiveness. Testing Behavioral Predictors of Effectiveness.
and Quasi—Experiment. Each phase except Quasi-Experiment
will be designed and implemented in this analysis. Sugges-
tions on how the Quasi-Experiment can be accomplished will
be developed and discussed. but actual implementation will
not be a part of this analysis.

Therefore. each chapter will consist of implementing
the separate phases of the procedure and presenting their
results. The outputs resulting from implementation em-
phasized here will be those that are directly used in the
next phases. Secondary outputs or subject matter informa-

tion not specificly used in following phases will be

94

95

discussed in subsequent chapters.

Although there may be many questions on the validity.
usefulness and workability of this system of procedures. it
is hoped that the implementation of major parts of the
system will act as a preliminary test of the system. An
estimate of the time and effort necessary for carrying out
the procedures can be weighed against perceived usefulness
of the information obtained through system implementation.
The systems approach requires iteration. By reviewing the
results of this trial run one should be able to modify the

procedures used to accommodate other situations.

Chapter 10

Selection of Farm Type

As mentioned earlier the type of farm may greatly
alter the set of decision aids useful for farm managers. as
well as change the evaluation procedures and criteria for
such aids. In general. the broader the categories. the less
precise will be recommendations for decision aid selection.
and the less useful will be the measure of value for those
aids. due to the non-homogeneity of the decision environ-
ment. However. with very narrow categories the measure may

be generalizable to a limited number of farm situations.

10.1 Implementation

As mentioned earlier the USDA classifies farms into
categories of cash grain. field crop. vegetable and melon.
fruit and tree nut. nursery and greenhouse. dairy. poultry
and egg. and cattlelhog/sheep. The Michigan Department of
Agriculture classifies farm production into slightly dif-
ferent categories combining cash grain and field crops under
one heading of field crops. Included under field crops are
corn. soybean. wheat. oats. barley. rye. hay. dry beans.
potatoes. sugarbeets and spearmint. The other Michigan
agricultural categories include: fruit. vegetables. flori-

culture. livestock and poultry. and dairy. The Michigan

96

97

State University computerized record keeping system
(TelFarm) has labeled the Michigan Department of Agricul-
ture’s field crop category as cash grain or cash grain and
sugar beet. For our purposes field crop and cash grain will
be used synonymously.

Selection of a category for analysis should be based on
importance of that farming system to the state. the availa-
bility of information on historical farm performance and the
availability or accessibility to analysis of the important
decisions made by that farm type. The availability of
information on historical system performance is essential
for the successful accomplishment of the Measuring Effec-
tiveness phase of the overall methodology. For the
succesful completion of the Behavioral Analysis phase. one
must make an assessment of whether the major decisions made

within the farming system can be isolated and evaluated.

10.2 Results/Outputs

According to the 1965 Michigan Agricultural Statistics.
field crops accounted for 7.724 million harvested acres of
the 7.939 million total crop acreage in 1964. In dollar
terms field crops represented 81.6 billion out of the total
$1.79 billion in crop receipts for the same year.
Therefore. field crOps ranked first by far in importance of
the crops grown in Michigan. Cash receipts from cattle and
calves were $246 million in 1964 while receipts from hogs.
sheep and lamb. poultry. and milk were $203 million. 84.6

million. $29.4 million. $724 million respectively.

98

Therefore. as a category field crop production ranked first
in Michigan in terms of gross receipts.

Historical records on the farm system performance is
essential for selecting the proper effectiveness measure by
which a farm manager is evaluated. Some farmers keep
accurate records of their past performance; however. relying
on records kept by farmers can pose some problems in terms
of reliability and comparability. Fortunately. Michigan
State University has an on-going farm recordkeeping system
called TelFarm in which a number of cash grain or field crop
farmers have participated in over the years. The records
kept by TelFarm are based on standard recordkeeping pro-
cedures which allow for comparability across farms and over
time. There are 64 cash grain or field crop farms which
have continuously participated in the recordkeeping system
from 1961 to 1964. The information available on these farms
include: production performance such as yields. labor use
and efficiency. and financial information such as net
income. returns to assets and debt ratios. Therefore. the
availability of this information makes the selection and
analysis of effectiveness measures possible for farms
primarily engaged in field crop production.

The third criterion for selection is that the major
decisions for the specific farm type can be isolated and
analyzed. Fortunately. Hepp and Olson (1960) have invest-
igated what were the important decisions made by Michigan

field crop farmers. This information can act as a basis for

99

determining the decision- making system of field crOp
operators to be used in the Behavioral Analysis phase of the
methodology. However. the ability to isolate and analyze
these decisions can only be determined through the actual
implementation of that phase.

Therefore. field crop (cash grain) farms are selected
for implementation of the following phases of the analysis.
The farm type selected is now used in both the Behavioral

Analysis and the Measuring Effectiveness phases which

follow.

Chapter 11

Behavioral Analysis

Now that cash grain or field crop farms have been
selected for analysis the process of developing an instru-
ment to serve as a predictor of managerial success can be
initiated. From Figure 11 one can see that the major
functions involved in this process include designing and
administering an open-ended questionnaire to determine
behavioral incidents of successful field crop farmers. The
informational outputs of these functions allow us to develop
the test instrument. and add to information necessary for
the selection of decision aids for the quasi-experiment.
along with providing evaluation criteria for evaluating the
validity of effectiveness measures.

Implementation and results will be discussed in the
following sections. The following sections will include: 1)
select set of successful farmers. 2) develop model of the
decision system. 3) design instrument to determine behavi-
oral incidents. 4) evaluate decision-making behavior of
successful farmers. and 5) develop instrument to measure be-

havior and performance.

11.1 Selecting Set of Successful Farmers

W

Although the term "successful" farmer may be both ambi-

100

101

guous and subject to differing interpretations. one must
start somewhere. The first step in the selection process
begins with a tentative definition of what is a ”successful"
farm manager. In this case “successful" manager can be
defined as one whose farm seems to stand a better chance of
survival while accomplishing both individual and family
goals.

Michigan cash grain or field crop farmers are divided
into two main categories based on topography and crop mix.
Saginaw valley farmers have the opportunity to grow sugar
beets. which is a high returns per acre crop. while other
parts of Michigan due to less favorable conditions cannot
economically do so. The majority of Michigan’s cash grain
operators are found in the lower half of the state. For
these reasons it was decided to concentrate in two major
areas - south central Michigan and the Saginaw Valley (see
Figure 15). Ingham County in central Michigan and Tuscola
County representing the Saginaw Valley were selected based
on location. accessibility. and possible willingness of
Extension personnel and the subsequent farmers selected to
participate.

It was decided by Dr. Hepp and the author to use the
Extension service for help in locating willing participants
for this phase of the analysis. Farm management personnel.
familiar with the managerial behavior and a feel for the
willingness of their more successful farm managers to parti-

cipate. were asked to supply us with a list of candidates.

 

 

 

10,;

C. (“W r5 fit?“ ,.
53"" 'L._. . rmx ,IV‘II'W‘T ! T
K. L.__L7.:J- - -i ‘mm‘ééa-‘J. ! mi.
\-\! rm. ! r" “‘"L _.-I.°“""“ é
‘\.,,\ - I far-'1 l"“""‘;}-----~. w ...
.4 t i- 1-1 ”L {.mv. / \d-‘b-laz-L-ch‘)
"fl flJ‘l-é’] a .e
g y 05 m\1
O my .
C— 4] 9 (fi)i !n;;\.\
g/ ”3611...- it.
.o x; raw-mm
a) .. , . .
MICHIGAN mammgéfiémla. “.-.]..-

 

 

 

Figure 15: Map of Michigan

 

 

 

103

83351155.

There were seventeen farmers actually interviewed. All
but one were cash grain operators with the other being a
hybrid seed producer. Nine were chosen from central
Michigan and eight chosen from the Saginaw Valley. The size
of crop operations ranged from 400 to 4.600 acres for the

1965 crop season.

11.2 Develop Model of Decision System
11121331211133.1211
Research has indicated a number of important and often
made decisions that the farm manager must make. Important
decisions as determined by Hepp and Olson (1960) include:

Purchase or rental of machinery and equipment
Repairing or building new buildings
What to plant

Livestock mix

Whether to purchase land

Whether or not to rent out land
Sale land

Whether or not to expand

Whether or not to quit farming

10. Remodel or build new home

11. Insulating home

12. Prices and Marketing

(DQVODOIIBQNH

As was mentioned in an earlier section (Measuring the
Management System). the element of time is important in
classifying decisions and determining the effect of such
decisions. The above decisions can be classified according
to their short and long run effects and the frequency that
they are made. Determining what to plant. whether or not to

rent out land. determining prices and marketing can be

104

classified as short-run and requiring attention by the
manager. The sale of land. expansion. and staying in
farming are decisions which can definitely be classified as
having long term consequences. Other decisions such as the
livestock mix. remodeling or building new home. purchase or
rental of machinery and equipment. and insulating the home
would take an intermediate position on a short—long run
continuum.

Since the goal of this analysis is to measure possible
benefits of computer aided decisions. the question of com-
parability is important. It is more difficult to measure
the effects of long run decisions because of both the non-
recurring nature of such decisions. and the possibility of
environmental or exogenous effects distorting such compara-
bility. Therefore. the short and medium range decisions
will be the focus for developing the measurement system.

With this in mind the areas of decisions explored by this
study include:
Determining the portion of crops to produce
Adding a new crop or substantially changing the crop
m1x
Determining corn variety
Determining when to purchase inputs
Determining when to sell corn

3.
4.
5.
6. Determining the adequacy of on-farm storage capacity
7. Estimating the price you expect to receive for corn
6.

9.

0.

1.

NH

Determining whether to pacticipate in government corn
programs

Determining whether to purchase a piece of land
Determining how much to pay for rental land
Determining the need to replace equipment

105

11.3 Design Instrument to Determine Behavioral
Incidents

In the developmental stage of scale evaluation.
analysis of the decision function was done by asking a set
of questions for each of the decisions analyzed. This phase
of the research can be considered a "fishing expedition". in
that we are putting out a big net of questions to possibly
retrieve only a few areas that may distinguish the better
farmers from the rest of the pack. These questions included:

A. What factors would a "good" decision maker in
comparison to a "poor“ decision maker consider
in making this decision?

B. What calculations would be appropriate for
solving this problem?

C. Do you spend a lot of time collecting and
analyzing information for this decision? C1.How
much?

D. How constraining are the following factors on
making a good decision:

Small Large
Knowledge of how to make decision 1 2 3 4 5
Internal data from farm
External data (prices.quality etc.)
Government (policy. taxes etc.)
Data Processing capabilities
Time to make decision
Uncertainty from results of prior decions
. Uncertainty from outside sources
Other - specify

OOVQU‘éwNo-A

E. What are your major sources for external data?
F. How would you or what would you need to help you
improve the way you make this decision?
The above decisions can be categorized as predominantly
affecting production. marketing and financing; as stated.
these decisions can be considered short to medium run. Not

being sure that one can capture the essence of what makes

the more successful farmers by focusing on particular

106

decisions. effort was put forth to evaluate possible
philosophical and procedural differences. that may cut
across individual decisions within the areas of production.
marketing and financing. The set of questions designed to
get at possible behavioral incidents include:
A. Production
1. What do you consider the keys to getting good yields
in corn! wheat/ soybeans?
2 . How do you ensure that you will have enough time to
get your planting and harvesting done 7
B. Marketing - Crops
1. What is your general philosophy about crop marketing?
2. Would you consider either production or marketing as
more or less important for the overall success of
the farm business?
3. What marketing alternatives do you feel most

confident with. i.e. forward contracting. cash. hed-
ging etc.?

C. Financing and Capital Investment
1. What is your general investment strategy for land?
2. What is your general credit policy?

These questions may only scratch the surface as to how
farmers make decisions. However. if from this group we can
find some decisions which in the manner in which managers
behave in trying to solve them. significantly differentiate
between managerial success. then a more intensive evaluation
of those decisions are warranted. In fact the more inter-
esting questions may deal with how the manager decides on a
decision rule rather than comparing the decision rules them-
selves which this research analyzes.

Also. this research only emphasizes one of the facets

107

of management and planning. excluding the other aspects of
monitoring. control. evaluation and learning. Of course.
there are some managers that may be very good at planning
but not so good at follow through. The results of such
behavior can affect returns in similar manners.
11.4 Evaluate Decision-Making Behavior of
Successful Farmers and Develop Instrument
to Measure Behavior and Performance
This section presents the results of administering the
open-ended questionnaire to the successful farmers. These
responses are then utilized in developing the instrument to
measure managerial behavior and performance. The format of
this section is to first present the responses grouped
according to areas of general philosophy. production.
marketing. finance and information processing. then the

questions designed from these responses will be presented.

11.41 General Philosophy

KW

As a general observation most of these farmers were
reluctant to consider themselves successful and even more
reluctant to judge other farmers. However. the farmers did
give forty-two different keys to successful farming. These
statements can be grouped into eight categories and ranked
according to the number of mentions of items within that
category (see Appendix A). The categories include: finan-
cial management<19). production(10), timeliness(9). reacting

to changes (decision-making)(9). attitude(6). communica-

108

tions(7). marketing(4) and general(4). It seems that finan-
cial management has become the most important key to
successful farming and this is in accord with current
financial conditions in agriculture (Append. A. Table I).

Although marketing was viewed as important by farmers.
they felt that they could not do very much about it and
therefore to be successful they have to do a good job at the
other aspects of farming. Also. none of the farmers
mentioned the importance of doing their own repair work on
machinery. However. from subsequent questions it was feund
that all the farmers could and did do most of their repairs
outside of major overhauls on engines and transmissions with
two even doing those. Therefore. this item just may be
taken for granted by farmers. From this analysis. it seems
that a test for measuring managerial success should

emphasize financial management. production and timeliness.

:.!.[H 21!.

Related to the keys to success question was the idea of
setting criteria on which to evaluate a manager. The
responses to this question were categorized objective.
subjective and behavioral (Append. A. Table II). The
objective factors included profitability. production costs.
assets and liabilities. yields. and financial returns on
assets. Of these objective measures. profitability was
chosen most by these farmers as a criteria for evaluation.

Among the subjective factors. appearance seemed to be

109
an important evaluation criteria. which would indicate that
these farmers felt that looking prosperous was a good sign
of being successful. The fact that the more successful
farmers also felt that one could evaluate a farm manager by

observing his managerial behavior. encourages the development

of a managerial index.

SD ! E 1'] I D' 11 E] ! E .

From the farmers' responses it seems that independence.
the challenge or feeling of accomplishment that farming
offers. along with the love for the outdoors and closeness
with the family seem to be their major motivating factors
(Append. A. Table III). However. uncertainty and unpre-
dictability generated by the market and weather along with
the high asset requirements tend to reduce the level of
enjoyment of their profession.

The model presented earlier would present the hypothesis
that these motivating factors should effect the decision
rules chosen and the amount of time spent in collecting and
analyzing information for decision-making. The responses to
a set of questions on goals and perspectives could be tested
as to their discriminating ability in relation to decision
rules used or factors analyzed.

W

The questions developed from these responses are pre-
sented in Appendix C under ”Decision-Making Analysis
Questionnaire” and include questions 6.9.10.11.34.37 and 39.

The options presented to the farmers in questions 10. 37 and

110

39 were based on the categories established by grouping the
responses to similar questions asked in the open ended-

questionnaire.

 

6. How many farm organizations are you a member of?
9. How many community organizations do you participate in?__

10. How did you acquire the majority of your owned land?
(Check one)
___a) Inherited
___b) Bought from family member(s)
___c) Bought from non-relatives
___d) Other (specify)_

11. How many years have you been farming independently? _____

34. How much repair work do you do on your equipment on your
farm?

37. What are the most important keys to being successful in
farming? (Rank 3)

_a) Attitude: motivation. dealing with people etc.

_b) Timeliness: getting things done on time

_c) Managing money and finances: controling costs.
managing credit etc.

_d) Productivity: solid production practices. stewardship
of land. adequate buildings and machinery etc.

e) Decision-making: ability to react to change. planning.
pin pointing problems etc.

f) Communications and public relations: keep contact with
outside world. be able to get expert help etc.

39. Which of these statements best describes why you stay in
farming? (Rank 3)

_a) Family tradition

_b) Feeling of accomplishment

_c) Like to control assets

_d) Attachment to land

_a) I'm a “Workaholic”

___f) Locked in by assets

___g) Don’t have other skills

___h) Offers high standard of living
_i) Like working with family

111
11.42 Production

E I x. I! . C

The fourteen farmers that responded to this question
indicated that the most important keys to getting good
yields in corn were fertilization. variety selection. timing
of planting. improved soil productivity and pest control

(Append. A. Table V).

H . E l I. E El !i I H v !'
Timeliness was earlier depicted by these farmers as
being most essential for success. therefore they seem most
sensitive to unanticipated delays during planting and har—
vesting. To ensure against such delays they indicated that
having adequate and reliable equipment. along with preseason
preparations. diversity in crops and maturing dates. the
willingness to ’run until done’. experience. and two way
communications equipment were essential (Append. A. Table
VI). Of course there may be a thin line between having
adequate equipment and being over-equipped which ties this

strategty to the overall investment strategy.

W

The factors which are predominant for determining crop
mix are rotation. government programs. returns per dollar.
and in the case of the sugar beet producers. controlling the
ground for sugar beet production (Append. A. Table X). The
rotational considerations are most important because these

farmers felt that a good rotation could help in controlling

112
pests. increasing overall soil productivity and reducing
erosion. In other words. they felt that long term benefits
outweighed the benefits of trying to receive maximum profits
per year by changing crop mix in response to yearly price
changes. Government policy and programs along with their
prior participation in these programs set constraints on
their acreage allotments to corn and wheat. In fact it
seemed that farmers spend a lot of time trying to guess what
the government was going to do since they felt that
government programs set the market price. Government policy
was stated to be the biggest constraint to making the crop
mix decision. In particular. a delay in the releasing of
quota. loan price and set-aside information delayed the crop
mix decision which added uncertainty to other phases of the

planning procedure (Append. A. Table XXIII).

El! !'11 :1 . E H'

The farmers in general expressed great reluctance in
substantially changing their crop mix in response to short
term fluctuations in prices received for commodities. They
would consider such changes only if they could be guaranteed
an extremely high contract price. plus government programs
would have to be considered along with the effects that such
changes would have on available labor and equipment. the
riskiness of the changes and the erosion or weed problems it

might introduce (Append. A. Table XI).

12!": IE'!

113

In determining corn variety. maturity dates as well as
performance and quality were important factors to consider.
As sources of information. MSU yield trial results were the
most frequently mentioned followed by on-farm test plots.
experience and dealer recommendations (Append. A. Table

XII).

W15.

Cash discounts along with income tax considerations
seem to be the most predominant factors which effect the
timing Of input purhase. Other factors considered are the
interest rates for borrowed funds. the availability of the
input along with expected price changes. the effects on cash
flow and the desire to have the inputs far enough in advance
as not to pose problems at time of use (Append. A. Table

XIII).

12. If you had to grow corn in another part of the state
what would be your primary source of information on the best
varieties of corn to plant? (Rank 3)

___ a) Seed dealers

b) Farmers near your new farm

c) Extension
d) Past experience on old farm
f) Local elevators

___ 9) Farm journals
13. How many varieties would you plant? _______

14. If you found that you were consistently falling behind

on getting your planting done on time. what would be the

114
first thing you would try to eliminate the problem? (Select

one)

_ a) Start planting earlier

b) Get larger equipment

c) Reduce acreage

d) Diversify crops and/or varieties grown
e) Hire more help

f) Put in more hours

15. In which crop is downtime the least expensive during

harvest? (Rank 3 least)
_ a)soybeans ___ b)wheat ___ c)corn ___ d)navy beans
_ e)sugar beets ___ f)barley ___g)seed corn

16. Assuming that other crop prices remain the same. how
high would corn prices have to be projected for the coming
year before you would double your acreage of corn?

Give high price ______ .
17. How low would corn prices have to be projected as
falling in the next year to get you to cut your acreage in
half?

Give low price _____ .

16. If you were interested in adding a crop to your present
crop mix. what factors would most influence that decision?

(Rank 3)

_a) Projected average price for next 5 years
_b) Equipment requirements

___c) Production practices and costs

__d) Compatability with present rotation

_e) Labor requirements

___f) Lowest price over last 5 years

___g) Government programs

19. Assume that interest rates are 12%. and your fertilizer
dealer is offering a 10% discount in December for fertilizer

that you would use on corn in the spring. Would you? (Select

115
one)
___a) Borrow money to take advantage of discount
___b) Wait to purchase until spring
20. Soil tests calls for 100 lbs of nitrogen on corn for

your type of soil and variety of corn. Would you? (Choose

one)

_a) Use a bit more
___b) Use as recommended
___c) Use a bit less

21. Do you know your cost of production for each of your

major crap? _____

22. If yes. what was your cost of producing for corn in

1985?

Cash Expenses: per acre Fixed expenses or overhead:
Per acre

Labor (Hired) Land Rental Charge

Labor (Own & family) _____ Depreciation ____
Repairs _____ Interest on debt ____
Seed _____ Insurance ____
Fertilizer and lime _____ Taxes ____
Pesticides _____ Utilities _---
Drying costs _____ Other __--
Fuel .....
Other _____
Total ------------ Total _--_
Total cash and overhead ____
Yeild per acre __--
Cost per bushel ____

11.43 Marketing
W
The basic strategy for crop marketing can be summarized
as first and foremost knowing your cost of production for
each crop. evaluating the advantage of entering the gov-
ernment programs with this cost of production figure in

mind. then trying to stagger sales during the year so as not

116

to get stuck with corn for more than nine months and not
being forced to sell at harvest (Append. A. Table IX).
”Trigger selling" summarizes their strategies most
adequately in that they sell when prices reach that trigger

and do not wait in hope of getting a better price.

W

The farmers emphasized the necessity of knowing cost of
production before being able to make the sell decision
(Append. A. Table XIV). After knowing cost of production.
including storage and government interest charges.
strategies ranged from trigger selling to selling when
needed to pay off creditors or scheduled sales spread out
over the year. Again the emphasis was put on getting rid of

the commodity and not carrying it over into another season.

mm:

For most of the farmers on farm storage was looked at
as a method to ease harvest bottlenecks more than as a
marketing tool (Append. A. Table XVI). However. farmers did
mention that on-farm storage could provide some bonuses in
prices received by holding until elevators were short. Some
felt that it was cheaper to store on the farm. and on-farm
storage gave them the option of selling either on the market
or to the government. Another interesting strategy was to
have enough storage to cover only owned acreage production
while selling the excess from rented acreage at harvest.

This allows for matching fixed asset capacities to assured

117

volumes. allowing the rented acreages to fluctuate without

adversely affecting fixed asset utilization.

M

It seems to be the very strong opinion of these farmers
that it is a waste of time trying to guess corn prices. The
best one can do is to use the government loan rate as the
bottom floor for corn prices and use that figure when fore-

casting expected returns (Append. A. Table XVII).

E !' . !' . E ! E

The majority of these farmers always participated in the
government corn program and would not grow corn without such
programs (Append. A. Table XVIII). Some of them would at
least pencil in the anticipated income comparing "in" versus
“out" of the program. but they stated that they seldom came
up with figures that would indicate not to participate.
However. they indicated that they put their marginal land in
the conservation program in order to qualify. In deter-
mining how much they would pay in rent for a piece of land

the ASCS corn basis was often considered.

W25.

23. What are your reasons for entering the government loan
program? (Rank 3)

_a) Sets price floor

_b) Cheap source of operating capital

_c) Is an assured market

_d) Can utilize marginal land as set aside
_e) Never entered

24.
strategy for selling corn would be:

_a)
_b)
_c)
_d)
_e)

___f)

-_-9>
___h)

_i)

25. The

If the government got out of the corn market.

118

the best
(Rank 3)

Use market information to pick highest price

See what neighbors are doing

Sell at harvest

Sell some corn every month

Set trigger above cost of production and begain to
sell when price reaches that point

Sell when market is going up

Sell just after market starts going down

Store on farm until elevators offer premium

Sell as cash is needed to pay bills

advantage of on-farm storage is: (Rank 3)
Can participate in government loan program
Saves time during harvest

Can get better price by holding

Can store cheaper than at elevator
Utilizes labor in off season

11.44 Financing: Investment and Money Manage-

ment

Land.1nxsstmsnt.§tnatssx

The farmers indicated that the land investment decision
was more of a response than a plan in that they could only
purchase land as it became available with the further pres-
sure of expanding to allow family members to get into
farming (Append. A. Table VII). The expansion process
usually started with renting more land and then purchasing
that land when it came up for sale. Some farmers more than
others would evaluate the land price against the expected
returns for that land. while others would look at their

present income generating potential and asset position to

see if they could subsidize the purchase of that land.

9 l'! E l'

119

From the seven farmers that responded to this question
the only predominant incident was a policy to use one lender
to ensure against becoming over extended and loosing track
of one’s debt load (Append. A. Table VIII). The most
striking observation about their responses was the diversity
of opinions about the use of credit and rules of thumb used
to deal with the credit issue. Opinions ranged from
“...never use credit to purchase inputs" to "...borrow to
maintain cash flow." This area seems ripe for further
evaluation of both the positive and normative aspects behind

the selection of such diverse methods of handling debt.

W

Although most land purchase decisions were not con-
sidered until a particular piece of land became available.
the farmers indicated that they would match expected returns
against the monthly costs of that prOperty (Append. A. Table
XIX). They would test the quality of the land by getting

soil tests or from prior experience from renting the

property.

W

The most important factors for determining how much to
pay for rental land were rental terms. the quality of land
and expected yields (Append. A. Table XX). Reasons for
renting land included the desire to lease first in order to
test the productivity of the land before purchase. and

taking a longer lease in order to build up the productivity

120

of a piece of land.

W

The major considerations in determining whether to
replace a piece of equipment were the perceived productivity
differences between the old versus a new piece of equipment.
the ability to afford the replacement. and the dependability
of the machinery during prime utilization periods (Append.
A. Table XXII). Indications of productivity problems
included the size of repair bills and constantly being
behind on job completions. Other rules of thumb for equip-
ment management included running machinery until it dies.
buy only used equipment. buy back up equipment for parts.
take good care. purchase for expected growth in acreage.
size for the job and knowing your capabilities for doing
repair work.

Fifteen out of seventeen farmers interviewed could and
did do all of their repair work except major overhauls to
engines and transmissions. while the other two even did the
major overhauls. Therefore it seems that the ability to do

your own repair work is an essential ingredient in farming

SUCCESS .

WM

26. How often do you prepare a projected cash flow? (Select

one)

_a) Yearly
_b) Monthly
___c) Quarterly
___d) None

121

27. A good reason for renting more land would be: (Rank 3)

28.

_a) Spread fixed costs
_b) Increase labor utilizaton
_c) The rental land is close to own land

__d) To test productivity before deciding to purchase

_a) Act as buffer while making equipment systems change

In deciding to purchase a piece of land up for sale

which of these statements best describes how you would go

abou

t making the decision? (Select one)

___a) I have been renting it for a while and if I don't

29.

buy it someone else will and get the benefits.

b) I would compare the expected net returns over the
life of the mortgage compared to monthly payments
and purchase only if it pays for itself.

c) I would determine if the equity base in my present
operation could subsidize the purchase of the land.

d) If the expected net returns per acre are greater
than my average cost of land. I would purchase.

If at the end of tax year you had 320,000 more taxable

income than expected. would you? (Choose one)

30.

_a) Purchase machinery or equipment to get tax credit

_b) Pay taxes on the income and put the rest on the
mortgage

_c) Invest in more land

_d) Spend more on consumption

_e) Buy more inputs for next year to reduce this year’s
taxes

Indicate how many years you would finance the following

items: (If less than one year indicate with a fraction)

a)
b)
c)

d)
e)

31.

Production Equipment _____
Land _____
Production inputs

used up in one year _____
Machinery shed .....
Personal car _____

Assume you have the Option of renting land from someone

122

on fixed cash rent or shares. Choose the alternative you
would prefer under the following conditions. (Indicate "F“
for fixed cash rent and "S” for shares.)

Prices are volatile

Prices are stable and high

Prices are stable but low

The land is not well drained

The land does not have a ASCS corn basis

They want a long term lease on good land

Don’t know productivity of land

32. Assume that you have enough equipment and labor to work
an extra 40 acres of land and you currently grow a combina-
tion of corn. wheat and soybeans. Also assume that a piece
land comes up for sale that historically produces 100
bushels of corn. If the finance terms are at 10% fixed

interest for 15 years. how much would you be willing to pay

per acre?

 

33. In choosing the following types of equipment pick 3
attributes from the list which are most important when
making the decision to purchase. (Place letter of attribute

next to number)

Performance qualities
Dependability

Equipment Attributes
Planter __1 __2 __3 A. Price
B. Age
Tractor __1 __2 __3 C. Condition
D. Financing arrangements
Harvest E. Depreciation
Truck __1 __2 __3 F. Investment tax credit
G. Resale value
Combine __1 __2 __3 H. Size
I. Dealer service
J. Availability of spare
parts
K. Availability of back up
equipment
L.
M.

123

N."State of the art“
Flexibility of uses

Own ability to repair
Horsepower

Lease options
Compatability with other
equipment

mmomo

35. How many different lenders do you do business with? _____

11.45 Information Processing

Of the twelve specific decisions analyzed some more
than others seem to require more calculations than others
and some require more internally generated historical data
than others. Theoretically. computerized data storage and
processing equipment should be most useful where the deci-
sion requires either/or extensive and accurate calculations
and/or extensive records.

The decision on when to sell corn was mentioned most by
the farmers as one where calculations are very important
(Append. A. Table XXII). In particular. figuring the cost
of production for each crop is both crucial and difficult.
Only after costs of productions are known can the manager
set price thresholds for the sell decision. However. a
number of managers expressed their skepticism towards
sophisticated charting or forecasting models for corn and
felt that the combination of the government loan rate and
futures market prices were sufficient for estimating future
prices within the growing season.

Next on the list is the land investment decision where

124

forecasting future cash flows and the effects of that
purchase on other fixed asset capacities are so important.
When to purchase inputs along with the decision to change
crop mixes follow in line as to the neccessity of calcula-
tions. Calculations seemed to play the smallest role in the
storage decision. determining expected corn prices. deter-
mining corn variety. and whether to rent verses buy a piece
of land.

Determining the proportion of each crop to grow was a
decision that farmers felt would be most aided by extensive
internal historical records including acreages grown and
performances on each portion of land along with optimal
input requirements for each parcel. Historical records
from the farm could aid the corn variety determination
decision along with helping determine the level of partici-
pation in government programs.

The equipment replacement decision was most dependent
on the performance and reliability of the machinery.
Therefore. records on break downs. repair expenses and field
performance information such as time per job and fuel con-
sumption would aid greatly in determining whether a newer or

larger piece of equipment was warranted.

gg'! "Lilli"

Due to the length of the questionnaire all of the
Slaecific decisions were not analyzed to determine con-
SiLraints to making those decisions. However. of the ones

a.l'ialyzed. proportion of crops. changing crop mix. corn

125

variety. input purchase and when to sell corn seem to

be most affected by both the availability of external infor-
mation on prices or qualities along with what the government
programs for the next year was to be and tax policy consid-
erations. It seems that the timing of government programs

is a major determinant of management strategy in these key

decisions.

W
36. Rank the usefulness of doing extensive and accurate
numerical calculations when making decisions in the

following categories. (Circle 1.2.3.4 or 5 where 1 means

not useful and 5 means very useful)

Not Very
useful useful
a) Determining crop mix 1 2 3 4 5
b) Adding a new crop 1 2 3 4 5
c) Determining crop varieties 1 2 3 4 5
d) When to purchase inputs 1 2 3 4 5
e) Cost of production for individual
crops 1 2 3 4 5
f) Estimating the future price of corn 1 2 3 4 5
g) Determining the economic level of
fertilizer to use 1 2 3 4 5
h) Determinig whether to enter a
government program 1 2 3 4 5
i) How much to pay for land 1 2 3 4 5
j) Choosing between financial Options 1 2 3 4 5
k) Making an equipment purchase 1 2 3 4 S
1 2 3 4 5

1) Estate planning

36. How many years of historical records on the following
items do you have that are readily available?

No. of years Item

a) Crop performance by fields

b) Variety performance

c) Soil test data by fields

d) Fertilizer utilization for each crop

e)
f)

h)

i)

126

Chemical utilization for each crop

Land cost per acre

Fuel utilization per acre

Yearly equipment repair cost for each peice
of major machinery

Hours of planter use

Chapter 12

Measuring Effectiveness (System Performance)

Now that field crop farms have been selected for
analysis. an appropriate measure of effectiveness must be
selected and tested for validity and reliability. This
phase commences by locating a set of farmers with historical
records of farm performance. Then an effectiveness measure
is chosen based on theory and workability. The measure is
tested for internal and external validity. which incor-
porates the views of the farmers themselves on an appro-
priate measure. along with statistical tests. Then the
measure is tested for reliability by comparing the consis-
tency of the measure over time. Once a measure has been
tested for both validity and reliability it is then used in
the next phase where the behavioral predictors of effective-

ness are used to predict the effectiveness measure chosen.

12.1 Implementation
5 I !' E EC: !° N
As stated earlier there are 64 field crop farmers that
have been involved in the Michigan State TelFarm record-
keeping system for at least four years. Among the informa-
tion collected and stored on these farms are yields. net
income. acreage planted to specific crops. production costs.

assets and liabilities. and financial returns to assets.

127

128

Information from the Behavioral Analysis phase indi-
cated that if farmers interviewed were given the opportunity
to evaluate how effective another farm manager was. they
would choose profitability. production costs. assets and
liabilities. yields. financial returns on assets. and
compare these measures over time as criteria for evaluation.
Profitability was chosen most by these farmers as a physical
and comprehensive measure of performance.

However. there are different interpretations and conse-
quently measures Of profitability. There is the income tax
definition; there is the cash verses accrual method of
determining net income; and there is the economic definition
of profits based on the opportunity cost principal. The
definition chosen here will be based on economic theory and
is the definition used in evaluating profitability by the
TelFarm system.

In TelFarm net farm income is determined using the
accrual method of accounting where this year’s costs are
matched against this year's value Of production. The income
calculated may be different from actual cash expenses
subtracted from this years sales because of inventory
changes for both inputs and outputs. After net farm income
is calculated. assumed charges for unpaid family labor.
assumed returns to Operator labor. and assumed return on
operator’s average capital (equity) are subtracted to arrive
at a residual which is assumed to measure the value of the

management input. Since each farm would have different

129

total acres, dividing by crop acres brings us to a measure
comparable accross farms of similar types. This we will

call “Management Income per Acre".

Ill. Bl'l'l'!

As stated earlier. any measure must stand up to the test
of scientific Objectivity. In particular management income
per acre can be said to be a workable definition within the
limited confines of those farmers that keep good records.
The question of clarity is dependent on the user's under-
standing and the ability to relate the concept. An impor-
tant issue arises as to the validity and consistency of the
measure. or reliability. Does it measure what it portends
to measure and is that measure a consistent indicator of
that attribute over time?

One way of testing the consistency of management income
per acre is to first assume that managerial ability does not
change very much over a short period of time. since agricul-
tural production is faced with long lags. Therefore. even
though the level of management income may change on a yearly
basis due to influences outside of the control of the
manager. for a set of managers facing similar circumstances.
the relative position of these managers as measured by a
managerial variable should show signs of consistency. In
particular. if we measure a farmer’s relative position to
the average or mean. then his position should not change

drastically over a relatively short period of time - say

130

four years.

Analysis of variance will be the statistical technique
used to determine the consistency over time of management
income per acre as a measure of managerial performance.
Analysis Of variance (ANOVA) compares the variance of a
group from the mean to determine whether the diffences of
means of the groups can be attributed to differences between
the groups or just a factor of random occurrences (noise).

Another question arises as to whether we are measuring
managerial effectiveness. noise or some other phenomena. In
particular. going back to our farm system. there are factors
which contribute to the performance of the farm system that
are not under the immediate control Of the farm operator.
One can argue that the differences in management income per
acre is just a function of location. soil type or the
availability of specialized markets which allow a higher
return.

We know that soil types vary extensively in Michigan.
They vary across counties and even within counties. It
stands to reason that if a farmer were in a county that had
corn yields consistently higher than another county. his
returns per acre might be greater. Also. in Michigan some
counties. due to their soil type and climatic conditions.
can economically grow sugar beets. which is a high value
crop. Therefore. before one can say that the management
income variable is a residual attributable to the managerial

input. the effects of location should be investigated.

131

Earlier it was noted that Heady and others question the
validity of most measures of management because of the
alleged association between capital and management.
According to this argument one can not adequately distin-
guish between the two. Therefore. if a statistical rela-
tionship does exist between levels of capital on the farm
and our chosen measure of effectiveness. serious questions
Of validity do exist.

Along with the capital issue is the question of eco-
nomies of size. there are some that question whether larger
farms have an economic efficiency advantage due to size.
Therefore. it would be advised to see if our measure of
effectiveness is statistically associated with a measure of
farm size.

Linear multiple regression is the technique chosen to
test the spuriousness of management income per acre (MI/A).
The hypothesis tested will be whether average county yields
in corn. being in the Saginaw Valley. the amount of capital
per acre and farm size as measured by acreage are signifi-

cantly associated with MIIA.

12.2 Results
I l E E . ! Q I.

The first step in preparing the data for using analysis
of variance is to divide the farmers for whom we can
calculate their management income for the four years between
1961 and 1964 into three equal groups. The farmers are then

sorted in descending order in terms of their respective

132

differences from the mean of management income per acre
(DFMGT). The first third of these farmers are considered in
the high management group. the next third in the middle
management group and the bottom third in the low management
group. Now one must determine if the variance

in their respective effectiveness be best attributed to the
management groups or a random effect of time. In other
words. do members in the mangament groups stay within these
groups over time or just shift around aimlessly?

The results of ANOVA (Table 1) indicate that the
variance attributable to management group membership was
significant to at least the 0.001 level Of probability;
whereas. the variance attributable to time was not signifi-
cant. In other words. it is very unlikely that the variance
observed as being attributable to differences between the
three management groups happened by chance. Furthermore.
management grouping accounted for 36.7 percent (R2=.367) of
the variance in management income per acre. while grouping
by year accounted for only 0.16 percent of the variance.

Therefore. although we cannot say categorically that
management income per acre is a perfect indicator of
managerial performance. at least it seems to be able to
consistently group farmers over time. Of course dif-
ferences in the levels of this variable can be attributable
to other factors beside managerial ability such as dif-

ferences in primary resource qualities. in particular. land

133

Table 1: ANOVA-2 MANAGEMENT BY GROUP AND YEAR

A. One way ANOVA: DFMGT grouped over variable 1 (GROUP)
with values from 1 to 3

A N A L Y S I S 0 F V A R I A N C E T A B L E

Degrees of Sum of Error

Freedom Squares Mean Square F-value Prob.
Between 2 749512.6111 374756.41 71.10 .000
Within 245 1291435.4056 5271.16
Total 247 2040946.2167
R2 = Sum of Squares Between I Sum of Squares Total = .367

B. One way ANOVA: DFMGT grouped over variable 2 (YEAR)
with values from 1 to 4

A N A L Y S I S 0 F V A R I A N C E T A B L E

Degrees of Sum of Error

Freedom Squares Mean Square F-value Prob.
Between 3 3329.1226 1109.71 0.13
Within 244 2037619.0941 6350.90
Total 247 2040946.2167

R2 = Sum of Squares Between I Sum Of Squares Total = .0016

134
and climate. The extent of some of these other factors
which question the validity of this measure will be

addressed later.

E . ! . E 1!! I! .! I 1:. l l I I ! .
Since by grouping the farmers by their average manage-
ment income per acre figure over a four year period seemed
to account for a substantial part of the variance. the next
step is to test the relationship of average MIIA over the
period to related environmental factors and the capital
endowment of the farm. The environmental factors invest-
igated are the average county yields and whether the farm is
in the Saginaw Valley or not. Since Heady (1963)
hypothesised that management ability was associated with
capital structure. it might be interesting to see if the
amount of investment per acre is related to management
income.
Hypothesis: The null hypothesis is that there is no rela-
tionship between MIIA and Average county corn yeilds.
whether the farm is in the Saginaw Valley or not. the

dollar amount of investment per acre. or the number of
acres farmed.

Model:

Y=b1 + b2*X2 + b3*X3 + b4X4 + bsX5 + e

Where:

Y=MIIA : average management income per acre (1961-1964);
dollars.

b1 = constant.

135

ha = YCORN :Average county corn yields (1961-1964);
measured in bu/acre.

b3=CTY :Location (1=Saginaw Valley. 0=Not in Saginaw

Valley).
b4 = CAPAC :Average dollar value of total capital per
acre (1961-1964).
h5 = ACRE :Average acres farmed per manager (1961-1964).

Error term.

Of course. the price received by the farmer affects his
income. The individual farmer may be able to get higher
than the average prices due to his/her marketing strategy.
but this is part Of the managerial input and should be a
part of MIIA. If the farmer is getting a higher price due
to a locational advantage then this should be accounted for

by the locational variable (CTY).

8911111.:

The results of multiple regression (Table 2) reveal
that the combination of the four variables (YCORN. CTY.
CAPAC. ACRE) is a poor predictor of management income per
acre (MIIA). Using the 5 percent confidence level for
evaluating the F-statistic. for the equation. we can accept
the null hypothesis that there is no relationship between
the set of independent variables and MIIA. However. one of
the independent variables. ACRE. may show some influence on
MIIA. At the 10 percent level of significance on the T-
statistic we can reject the null hypothesis that the coeffi-

cient is zero. However. an R2 of 0.041 would indicate that

136

Table 2: ENVIROMENTAL VARIABLES AFFECTING MIA

Function: MULTIREG
Data case no. 1 to 71
Without selection

2-YCORN

3-CTY

4-CAPAC

S-ACRE

1-MIA

Var. Regression Standard Student

No. Coefficient Error T value Prob
2 1.7039E+00 1.3663E+00 1.247 .21
3 -1.6261E+01 2.2620E+01 -0.719 .47
4 ~6.0627E-03 1.3106E-02 -0.463 .64
5 6.6736E-02 2.9414E-02 2.337 .02

Intercept =-246.7215

Coefficient of Determination (R-Square)= 0.096
Adjusted R-Square = 0.041
Multiple R = 0.309
Standard Err of Est. = 66.024

A N A L Y S I S 0 F V A R I A N C E T A B L E

Sum Of Squares df Mean Square F Signif
Regression 32263.760000 4 6065.94000 1.74 .151
Residual 305399.400000 66 4627.26400
Total 337663.200000 70

137
these variables account for only 4.1 percent of the variance
in MIIA.

Therefore. it would seem that management income per
acre can not be adequately explained as the result of
location. county corn yields. size of farm or amount of
investment per acre. But size of the farm may have some
effect once the other variables are taken into considera-
tion. The positive sign on ACRE indicates that larger farms
may have an advantage at attaining high management income.
However. one cannot be sure of the direction of causality.
It could be that the ability to manage well may influence
the size of Operation attempted. or farming a larger
Operation could increase the manager’s abilities. or simply
that manamgement income per acre is affected by economies of
scale having nothing to do with ability.

Since the relationships between MIIA and the independent
variables are small and statistically insignificant except
for ACRE. one can argue that MI/A is measuring something
other than evironmental or structural variables outside the
immediate influence of management. Therefore. it would seem
that MIIA has passed this test for validity in that the
measure was not substantially influenced by this set of non-

managerial variables.

Chapter 13

Testing Behavioral Predictors of Effectiveness

The first step in this phase consists of administering
to a set of farmers with historical farm records the instru-
ment or test designed in a previous chapter to measure
managerial behavior and performance. Once the managerial
test has been completed by the farmers. they must be scored
based on some criteria.

Once the behavioral scores are compiled. statistical
tests for significance can be performed to indicate the
degree of reliability of the managerial index as an indi-
cator of effectiveness. A revised managerial test can then
be formulated based on the results of the statistical
analysis. where the best set of predictors can be used as
the basis for the new instrument. The new instrument can
then serve as the managerial index to be used in the quasi-
experiment. where comparisons can be made between the index
in measuring the value of decision aids and the value
Obtained by observing sets of farmers over time as they use
different decision systems.

The chapter will be divided into four main areas: 1)
Administering the Instrument to Farmers. 2) Scoring the

Test. 3) Testing the Hypothesis and 4) Revising the Instru-

ment.

138‘

139

13.1 Adminstering the Instrument to Farmers

The farmers chosen to participate in this phase of the
analysis are the same set used for validating management
income per acre as an effectiveness measure. These were 64
cash grain or field crop farmers that participated in the
TelFarm record keeping system from 1961-1964.

Out of these 64 farmers there were 47 usable question-
naires returned for a response rate of 56 percent. "Usable“

means completely filled out.
13.2 Scoring the Test

The method of scoring the test is based on the prin-
ciple adhered to throughout this analysis. and that is. that
the hghayigg_of the more successful farmers is the key to
their success as measured by effectiveness. The first step
in the scoring process is to select a set of farmers whose
answers are to serve as the criteria for scoring the
responses Of the other farmers. Since management income per
acre (MIA) was chosen as the effectiveness measure. the
criteria for selection of the top management group is that
they must have had a management income per acre value that
placed them in the top one third of all the managers for at
least three of the four years. and in the top one third as
determined by the four year average Of management income per

acre.

Based on the above criteria six farmers were selected

140

from the 47 farmers completing the questionnaires. Their
responses were used as the standard by which the other
questionnaires were scored.

The questionnaire itself - found in Appendix C - was
divided into five parts including: 1) Background Informa-
tion. 2) Production Decisions. 3) Marketing. 4) Investment
and Money Management and 5) General Management. The area
soliciting background information was not scored as such. in
that no score was given them other than a numerical code
representing the raw responses given.

However. the questions in the other areas were scored
according to the following criteria. Questions 12. 15. 16.
23-25. 27. 33.37 and 39. where the respondents were asked to
rank 3 of the alternatives given. were scored by giving
scores of 5.3 and 1 based on responses of 1.2 and 3 respec-
tively. Then the scores for each alternative were added
together; now this value was used as the score given to the
respondents for selecting the respective alternative.
However. although the respondents were asked to rank three
alternatives. they were scored based on the first two alter-
natives they selected regardless Of whether that alternative
was selected first or second. A demonstration is in order.

Let us take question 23 as an example of this type of
scoring.

23. What are your reasons for entering the goverment
loan program? (Rank 3)

a) Sets price floor
b) Cheap source of operating capital

141

c) Is an assured market

d) Can utilize marginal land as set aside

e) Never entered
Alternative "a" was chosen first three times. twice once.
and third once giving it a score of 19 ((3*5)+(1*3)+(1*1)).
Alternative "b" was chosen second four times. and third
once. giving it a score of 13 ((4*3)+(1*1)). The other
three alternatives were scored likewise with the result
being alternative a=19. b=13. c=10. d=3. and e=5. For the
next phase of the scoring. only the top three responses (a.b
and c) were left with their scores. while the other
responses were given a zero score. Now if a respondent
chose "a“. ”b” or ”c" for his first or second choice then he
was scored accordingly. For instance if the respondent
chose “a” first and "b" second. then he was given a total
score Of 19+13. or 32. However. if he chose ”a" second and
“d“ or ”e" first. then he received only 19 points.

There were other questions such as 19.20.26.26 and 29.
where the respondent was asked to select only one. In these
cases if the set of farmers chosen as the standard selected
a given alternative. that alternative was given a score of
5. When all six Of the respondents' selections had been
noted and tabulated. the sum of the values within each
alternative became the score given to subsequent respondents
that chose that alternative.

There were a set of questions — 13. 30a.b.c.d.e. 32.
34. 35. and 36 - where a specific number was asked for. In

these cases the correct answer was assumed to be one within

142

one standard deviation of the mean of the responses given by
the six more successful farmers. For instance. question 13
asked how many varieties would one plant. The mean was four
and the standard diviation was about one. Therefore. each
respondent was subsequently given a score of 30 if their
response fell between 3 and 5. Each one of these questions
was given a different value or score for answering cor-
rectly. The selection of the level was based on giving that
question a weight in proportion to the possible scores
obtainable on the other questions. and the relative
smallness of its standard deviation. If the standard devia-
tion was small. one could assume that the issue was quite
settled. Whereas. if the standard deviation was large for
these six respondents. then either the question did not
bring out a behavioral incident. it was too vague. or there
may be more than one right solution to the problem under
real world conditions. A smaller standard deviation would
warrant giving that question a proportionally higher poten—
tial value.

There were a few questions - 21 and 31 - where only a
yes/no or slf (share/fixed) response was asked for. In
these cases the majority response was selected as the
correct response and scored as to give questions 21 and 31
comparable weights with the other questions.

Question 36. which asked respondents to rank the use-
fulness of doing extensive and accurate numerical calcula-

tions when making decisions over certain categories of

143

problems. was scored by giving five points to each alter-
native level Of usefulness chosen by the six more successful
farmers. Subsequent respondents were then given the sum Of
each level of usefulness when they chose the appropriate
response. For instance. the problem of adding a new crop
was given a level of 3 on the usefullness scale by 3
respondents and 5 by the other 3. Therefore. if a subse-
quent respondent chose either 3 or 5 they would receive 15
points. but they would not receive any points for choosing
1.2 or 4.

Question 22. where respondents were asked to figure out
their cost of production for corn. yield per acre and cost
per bushel were used as the criterion variables. They were
scored by first determining the mean and standard deviation.
then for yield per acre the subsequent respondents were
given a score of 20 if their yield was no less than one
standard deviaton below the mean. and given 20 points if
their cost per bushel was at most one standard deviation
above the mean. Of course. if they did not respond. they
were given a score of zero. This question may cause
problems in subsequent interpretation since more than one
thing is being measured at once. Here both the ability to
do the computations (behavior and performance) and the
actual system response (yields or cost per bushel) which are
effectiveness measures are both being treated together.

Finally. questions which included multiple parts were

given a total score based on the sum of scores Obtained from

144

answering the parts.

13.3 Testing the Hypotheses
W

Now there are 47 completed and scored questionnaires.
or managerial tests that can now be used to test the rela-
tionship between effectiveness and behavior/performance.
The questionnaire is comprised of many questions which can
be categorized according to production. marketing. finance.
and information processing as well as further divided into
whether the questions measure behavior or performance.
Therefore. there are a number of hypotheses that can be
tested based on how one categorizes the questions. Each of
the hypotheses are then tested by using linear multiple

regression.

WW1
Hypothesis 1: There is no relationship between effectiveness

and the total score of answers to the questionnaire.

Hypothesis II: There is no relationship between effective-

ness and behavior or performace.

Hypothesis III: There is no relationship between effective-
ness and scores on production. marketing. finance. and in-

formation processing.

Hypothesis IV: There is no relationship between effective-

ness and any set of questions on the managerial test.

145

19.51.1113
Hypothesis 1: There is no relationship between effectiveness
and the total score of answers to the questionnaire.

This hypothesis is based on adding the individual
scores from questions 11 through 39 into'a composite score
(TSCORE) which reflects the overall behavior and performance
aggregated across the entire managerial test. The formal

model can be represented as: MIA = C + b * TSCORE + e

8mm
As shown in Table 3. TSCORE accounts for only 2 percent

Of the variance in MIA and is not significant at the .10
level. Therefore. one can argue that the null hypothesis of
no relationship between management income per acre. and the

total score received on the managerial test instrument.

holds.

Hypothesis II: There is no relationship between effective-
ness and behavior or performance.
The model to be tested based on this hypothesis becomes:
MIA = C + blsBEHAVIOR + bzfiPERFORMANCE + e
BEHAVIOR is the variable that represents a summation of all

the scores on the questionnaire that relate how a manager

behaves while making decisions. PERFORMANCE represents the

146

Table 3 : Testing Hypothesis: MIA = C + bfiTScore

Data case no. 1 to 47
Without selection

TSCORE
MIA
47 cases read 0 missing cases discarded
Regression Standard Std. Partial Student
Variable Coefficient Error Regr. Coeff. T value Prob
TSCORE 1.1022E-01 7.6505E-02 0.2049E+00 1.404 .16

Intercept =-133.763

Coefficient of Determination (R-Square)= 0.042
Adjusted R-Square = 0.021
Multiple R = 0.205

Standard Err of Est. 71.172

A N A L Y S I S O F V A R I A N C E T A B L E

Sum Of Squares df Mean Square F Signif
Regression 9965.090000 1 9965.09000 1.97 .167
Residual 227944.000000 45 5065.43000

Total 237929.000000 46

147

summation Of scores of each question where the farm manager
is asked to solve a problem. or state how he solved a
problem. An example may be helpful in making the dis-
tinction between a behavioral question and a performance
question. An example of a behavioral question is question
18:

16. If you were interested in adding a crop to your present
crop mix. what factors would most influence that decision?
The farmer is given alternative possible pieces of informa-
tion that he might use in making the decision. An example
of a performance question would be question 19:

19. Assume that interest rates are 12%. and your fertilizer
dealer is offering a 10% discount in December for fertilizer
that you would use on corn in the spring.

From here the farmer is given two possible courses of action
from which to choose. The distinction is that question 16
is trying to solicit how the manager makes the decision.

while question 19 asked the farmer to solve the problem and

give a prescription.

39.13.1115.

The results of linear multiple regression analysis
(Table 4) indicates that the null hypothesis is not
rejected at the .10 confidence interval as measured by the
probability of the F-statistic for the relationship MIA =

f(BEHAVIOR. PERFORMANCE). However. it is interesting to

148

note that the performance variable is much closer to being
significant than the behavioral variable. which may indicate
that subsequent questionnaires should attempt to include
more questions asking the farmers to solve problems or
asking them past prescriptions followed.

It is comforting to note at this point that there is a
high degree of correlation between BEHAVIOR and PERFORMANCE
as seen in the correlation table presented in Table 4.

This indicates that at least the questionnaire backs up the
hypothesis that there is a relationship between how farmers

make the decisions and the actual decisions they make.

Hypothesis III: There is no relationship between effective-
ness and scores on production. marketing. finance. and in-
formation processing.

The model to be tested here is:

MIA = C + b1*PROD + bzfiMARK + b3*FIN + b4*INFO +e
PROD is the summation of the scores on questions relating to
production decisions. MARK is the total score on the
marketing questions. FIN is the total score on the financing

and money management questions. and INFO is the total score

on questions relating to information.

149

Table 4: Testing Hypothesis II: MIA = f(BEHAVIOR. PERFORMANCE)

Function: MULTIREG
Data case no. 1 to 47
Without selection

BEHAVIOR
PERFORMANCE
MIA

47 cases read 0 missing cases discarded

Correlation Matrix
BEHAVIOR PERFORMANCE MIA

BEHAVIOR 1.000
PERFORMANCE 0.600 1.000
MIA ' 0.139 0.254 1.000

Regression Standard Std. Partial Student
Variable Coefficient Error Regr. Coeff. T value Prob
BEHAVIOR -1.6695E-02 1.4506E-01 -.2096E-01 -0.115 .90
PERFORMANCE 3.2930E-01 2.2477E-01 0.2669E+00 1.465 .14

Intercept =-123.714

Coefficient of Determination (R-Square)= 0.065
Adjusted R-Square = 0.022
Multiple R = 0.255
Standard Err of Est. = 71.10?

A N A L Y S I S O F V A R I A N C E T A B L E

Sum Of Squares df Mean Square F Signif
Regression 15454.300000 2 7727.17000 1.53 .226
Residual 222475.000000 44 5056.25000

Total 237929.000000 46

150

8151113.:

The results of multiple regression (Table 5) for this
configuration of variables were similar to the preceeding
one. in that the null hypothesis of no relationship between
MIA and now the questions aggregated according to their
reference to production. marketing. finance. and information
processing. still held. Non of the individual variables
were significant at the .10 probability of the t-statistic.
However. there were strong correlations between some of the
independent variables themselves which is to be expected if
the questions are measuring aspects of the same attribute.

The strong multi-colinearity between MARK. FIN and INFO
may point one in the direction of a different combination of
scores. that may lead to a set of variables that may indeed
be related to the dependent variable. This information is

used in establishing such a set in the next hypothesis to be

tested.

Hypothesis IV: There is no relationship between effective-

ness and any set of questions on the managerial test.

As stated above the previous regressions on MIA indi-
cate that there may be another combination of scores that
may indeed be considered indicators of effectiveness. The
previous analyses also indicate that a number of the ques-

tions may be discarded leaving a core of questions that may

151

Table 5: Testing Hypothesis III: MIA = f(PROD.MARK.FIN.INFO)

Function: MULTIREG

Data case no. 1 to 47
Without selection : 47 cases read. 0 missing cases discarded
PROD
MARK
FIN
INFO
MIA
Correlation Matrix
PROD MARK FIN INFO MIA
PROD 1.000
MARK 0.227 1.000
FIN 0.455 0.527 1.000
INFO 0.240 0.572 0.461 1.000
MIA 0.067 0.192 0.246 0.092 1.000

Determinant of matrix is .359372

Regression Standard Std. Partial Student

Variable Coefficient Error Regr. Coeff. T value Prob
PROD -6.2069E-02 4.6356E-01 -.2636E-01 -0.170 .66
MARK 1.9779E-01 3.1276E-01 0.1224E+00 0.632 .53
FIN 3.7116E-01 3.0707E-01 0.2372E+00 1.209 .23
INFO -1.0673E-01 2.3466E-01 -.6534E-01 -0.455 .65

Intercept =-139.924

Coefficient of Determination (R-Square)= 0.072
Adjusted R-Square = -.016
Multiple R = 0.266

Standard Err of Est. 72.506

A N A L Y S I S O F V A R I A N C E T A B L E

Sum of Squares df Mean Square F Signif
Regression 17127.900000 4 4261.97000 0.61 .523
Residual 220601.000000 42 5257.16000

Total 237929.000000 46

152

be the basis for a test that has a better chance at predic-
ting success.

The approach here is to fish through the variables one
at a time. and select the questions that show a strong
correlation with MIA. and then aggregate these scores into
variables according to decision type. i.e. production.
finance. etc. After the searching. testing and reaggrega-

tion. another model can be posited and statistically tested.

8.2511113.

After looking at individual questions and their statis-
tical association with the dependent variable. these ques-
tions stand out as possible raw material for aggregation and
testing: 6. 10. 16. 19. 22. 27. 29. 30 and 31. Questions 6
and 10 are background variables which stand alone and are
not appropriate for aggregation. Questions 16 and 19 are
related to both marketing and production. but much more
related to knowing and controlling one's cost of production.
Therefore. these two questions are aggregated; summing their
individual scores to form a new variable COSTPR. which is an
indicator of the farmer knowing his cost of production.

Questions 19. 27. 29. 30 and 31 are all related to
financial or money matters. All of these questions are
performance type questions except 27. where behavior is
solicited. Therefore. a new variable FINBP was calculated

by aggregating scores to questions 19. 27. 29. 30 and 31.

153

Now a new model can be posited:

MIA = C +b1*ORGAN + bzflACQUIRE + b3*COSTPR + b4fiFINBP + e
Where ORGAN is the organizational structure of the farm.
ACQUIRE is how the farm was acquired. COSTPR is the aggre-
gated variable measuring the performance Of the farmer in
measuring and using his cost of production information. and

FINBP is the aggregated variable measuring both behavior and

performance Of the manager when faced with financial and

money matter type decisions.

315111.13

Table 6 presents the results of multiple regression
analysis. Immediately. one can reject the null hypothesis
of no relationship between MIA and our independent var—
iables. since the F statistic was 6.46. which represented a
probability smaller than .001 for type I error. The
equation itself according to the coefficient of determina-
tion accounted for over 36 percent in the variance of man-
agement income per acre. The variable showing the greatest
association with MIA was FINBP or the varible representing
the farmers score on the questions categorized as pertaining
to finance and money management.

Therefore. this analysis would indicate that this sub-
set of variables from the original management test instru-
ment may indeed be predictors of managerial effectiveness as
measured by management income per acre. In particular.

higher scores on the questions pertaining to determining

154

Table 6 : Testing Hypothesis IV: MIA = f(ORGAN. ACQUIRE.
COSTPR. FINBP)

Function: MULTIREG

47 cases read: 0 missing cases discarded

ORGAN
ACQUIRE
COSTPR
FINBP
MIA

Minimum Maximum Mean
ORGAN 1.00 5.00 1.702
ACQUIRE 1.00 4.00 2.661
COSTPR 0.00 70.00 20.213
FINBP 30.00 157.00 112.651
MIA -305.50 61.00 -67.114
Correlation Matrix

ORGAN ACQUIRE COSTPR FINBP MIA

ORGAN 1.000
ACQUIRE 0.329 1.000
COSTPR 0.002 0.196 1.000
FINBP 0.194 0.321 0.425 1.000
MIA 0.266 0.360 0.364 0.543 1.000

Determinant of matrix is .637405

Regression Standard Std. Partial Student

Variable Coefficient Error Regr. Coeff. T value Prob
ORGAN 9.5262E+00 7.4613E+00 0.1656E+00 1.274 .20
ACQUIRE 1.5695E+01 1.4566E+01 0.1447E+00 1.076 .26
COSTPR 6.2669E-01 4.3996E-01 0.1927E+00 1.424 .16
FINBP 1.0623E+00 3.9695E-01 0.3626E+00 2.727 .00

Intercept =-260.212

Coefficient Of Determination (R-Square)= 0.362
Adjusted R-Square = 0.323
Multiple R = 0.616
Standard Err Of Est. = 59.166

A N A L Y S I S 0 F V A R I A N C E T A B L E

Sum of Squares df Mean Square F Signif
Regression 90793.400000 4 22696.30000 6.46 .000
Residual 147136.000000 42 3503.24000

Total 237929.000000 46

155

cost of production and on making financial. investment and
money management decisions are positive indicators of in-
creased management income per acre. Further. that the or-
ganizational structure of the farm. and how that farm was
acquired influence the ability of the behavioral and perfor-
mance variables to predict management income per acre.

Further analysis showed that one could drop either
ORGAN or ACQUIRE without either significantly reducing the
coefficient of determination. increasing the standard errors
Of COSTPR and FINBP. or substantially changing the magnitude
of the coefficients on COSTPR and FINBP. Table 7 presents
these results which indicate that the best estimate Of the
coefficients on COSTPR and FINBP are obtained by including
ORGAN and dropping ACQUIRE.

If we accept the hypothesis that MIA is a function of
ORGAN. COSTPR and FINBP. questions may still arise as to the
validity Of the scoring method and subsequent analysis. In
particular. since the six farmers selected for establishing
how the questions were to be scored were included in the 47
cases used for testing the model and developing the coeffi-
cients. one can then argue that the deck is stacked in favor
of accepting such an hypothesis of subsequent association.
Therefore. as a check. MIA was regressed against ORGAN.

COSTPR and FINBP without the six farmers used as the stan-

156

Table 7: Comparing Hypothesis IV without ACQUIRE or without
ORGAN

47 cases read 0 missing cases discarded

A. Without ACQUIRE

Regression Standard Std. Partial Student

Variable Coefficient Error Regr. Coeff. T value Prob
ORGAN 1.1907E+01 7.1602E+00 0.2069E+00 1.663 .10
COSTPR 6.7494E-01 4.3647E-01 0.2075E+00 1.539 .13
FINBP 1.1731E+00 3.6659E-01 0.4149E+00 3.019 .00

Intercept =-233.404

Coefficient of Determination (R-Square)= 0.365
Adjusted R-Square = 0.320
Multiple R = 0.604
Standard Err of Est. = 59.297

A N A L Y S I S 0 F V A R I A N C E T A B L E

Sum of Squares df Mean Square F Signif
Regression 66737.700000 3 26912.60000 6.22 .000
Residual 151192.000000 43 3516.09000
Total 237929.000000 46

*Mfliflfl-‘IHNMKMNNINN************M:*******I:I:INK“!******§*******§*******

B. Without ORGAN
Regression Standard Std. Partial Student

Variable Coefficient Error Regr. Coeff. T value Prob
ACQUIRE 2.1164E+01 1.4035E+01 0.1953E+00 1.509 .13
COSTPR 5.6192E-01 4.4016E-01 0.1726E+00 1.277 .20
FINBP 1.1509E+00 3.9611E-01 0.4070E+00 2.906 .00

Intercept =-265.146

Coefficient of Determination (R-Square)= 0.356
Adjusted R-Square = 0.313
Multiple R = 0.596
Standard Err Of Est. = 59.615

A N A L Y S I S 0 F V A R I A N C E T A B L E

Sum of Squares df Mean Square F Signif
Regression 65110.600000 3 26370.30000 7.96 .000
Residual 152619.000000 43 3553.92000

Total 237929.000000 46

157

dard for scoring the instrument. Table 6 shows the

results of that analysis and although the coefficients are
less significant. and the overall measure of significance of
the equation is reduced. the elimination of the six farmers
does not change the basic structure of the relationship.

The null hypothesis of no relationship is rejected at the
0.01 level of probability.

Earlier. when evaluating the validity of MIA. the
question arose as to whether the size of the farm influenced
management income per acre or did the ability to manage
affect the size Of the farm. Table 9 reveals that the
addition of the two environmental variables. CTY and ACRES.
does not add to the predictive ability Of the other
variables. The low t values also indicate that they are not
significant indicators of management income per acre. and
the sign of ACRE is even negative. Such results would
indicate that the addition of FINBP greatly reduces the
ability of the size of the farm to predict MIA. Therefore.
not only is MIA further validated. but financial managerial

ability seems to be essential for effectiveness.

158

Table 6: Tsting Hypothesis IV without Six Top Farmers

case no. 7 to 47
41 cases read: 0 missing cases discarded

Regression Standard Std. Partial Student

Variable Coefficient Error Regr. Coeff. T value Prob
ORGAN 1.1026E+01 6.0767E+00 0.1932E+00 1 365 17
COSTPR 4.7476E-01 5.0429E-01 0.1401E+00 0.941 .35
FINBP 1.1405E+00 4.2957E-01 0.4069E+00 2 655 01
Intercept =-229.277

Coefficient of Determination (R-Square)= 0.306

Adjusted R-Square = 0.250

Multiple R = 0.553

Standard Err of Est. = 62.006

A N A L Y S I S 0 F V A R I A N C E T A B L E

Sum of Squares df Mean Square F Signif
Regression 62722.100000 3 20907.40000 5.44 .003
Residual 142254.000000 37 3644.69000

Total 204976.000000 40

159

Table 9: MIA=f(ORGAN. COSTPR and FINBP:CTY and ACRE)
Function: MULTIREG
Data case no. 1 to 47
Variable Regression Standard Std. Partial Student
Number Coefficient Error Regr. Coeff. T value Prob
ORGAN 1.2130E+01 7.5646E+00 0.2106E+00 1 604 .11
COSTPR 6.1464E-01 4.5657E-01 0.1690E+00 1.340 .16
FINBP 1.2096E+00 4.2690E-01 0.4276E+00 2.634 .00
CTY 1.2667E+01 2.2396E+01 -.7413E-01 -0.575 .56
ACRE 1.1140E-02 5.1960E-02 -.3077E-01 -0.214 .63
Intercept =-230.007
Coefficient of Determination (R-Square)= 0.371
Adjusted R-Square = 0.294
Multiple R = 0.609
Standard Err Of Est. = 60.421

A N A L Y S I S O F V A R I A N C E T A B L E

Sum of Squares df Mean Square F Sign1f

Regression 66250.900000 5 17650.20000 4.63 001
Residual 149676.000000 41 3650.69000
Total 237929.000000 46

160

In conclusion the null hypotheses I. II and III were
accepted and hypothesis IV was rejected. In analyzing
hypothesis IV. it was discovered that sets of questions on
the managerial test were found to be significant indicators
of MIA. In particular. questions dealing with the farmer’s
ability to calculate and use his cost of production informa-
tion. along with his ability to solve financial. or money
decisions. were found to be significant predictors of
management effectiveness. A qualifying demographic char-
acteristic. organizational structure or how the farm was
acquired. added to the predictive ability of the two

managerial characteristics.

13.4 Revising the Instrument

The results of testing the relationship between manage-
ment income per acre and the managerial test indicates that
the attempt to establish a managerial index may be fruitful.
However. such a test should be slanted highly towards ques-
tions on financing. money management. investment and knowing
one's cost of production for individual commodities.

The production and marketing sections of the managerial
test did not serve as good predictors Of management income
per acre. Four separate hypotheses can be given for the
inability of these sections to serve as predictors. First.
these farmers may be very similar in their decision-making
behavior in production and marketing. therefore these areas

do not account for differences in effectiveness of the

161

managers. Second. the questions developed for these
sections may not have been adequate to find such dis-
tinctions. Third. the measure of effectiveness may be
slanted away from discovering differences that may in fact
occur between farmers in their decision-making behavior and
performance in production and marketing. Fourth. the
managerial test was slanted toward the planning phase of the
manager’s task. leaving the control and follow through areas
untouched. Therefore. farmers may be similar in their
planning decision-making in the area of production and
marketing. but their abilities to carry out such decisions
may differ.

Although the acceptance of any of the above hypotheses
would warrant further attempts to find better measures of
production. and marketing behavior/performance. there is
clear evidence that financial. investment and money matter
type decisions are indicators of management income per acre.
It would follow that one could use the questions from this
first attempt as the core for the establishment of future
managerial tests. Such tests should lean heavily towards
asking questions in the areas of finance. investment. money
matters and cost of production. that require the manager to
solve a problem or state how he has solved such a problem
before. It seems that focussing on the behavioral aspects
such as sources of information and decision rules used are
not as effective as comparing the actual decisions made when

trying to measure management decision-making. This brings

162

back in mind the role of simulation as a possible method of
putting the decision-maker in as close to a real problematic
environment as possible.

Therefore. at least two courses of action should be
followed. More investment. financial and money matter
problem situations should be developed for managers to solve
on a managerial test. Further work should be done to
develop simulated real world environments in which the
decision-maker must plan a course of action that requires
the ability to determine and use one’s cost of production

information. along with the ability to make financial and

money matter decisions.

13.5 Implications for Quasi-Experiment

The above analysis indicates that a revised question-
naire should be developed which emphasizes performance on
money and investment type decisions. However. since in-
vestment decisions take a longer time to measure their
effects. this could pose problems with conducting a compara-
tive study. Although. it was hoped that three to four years
may have been sufficient to analyze differences in effec-
tiveness. a long time frame such as five to seven years may
be warranted because Of the nature of the issues that make a
difference in farm success.

Another implication focuses on the problem of testing.
Problem-solving and particularly under realistic simulated

conditions takes time. Although. the questionnaire used in

163

this analysis probably took only one hour at the most. a
test involving many problems under simulated conditions
could take hours. Whether one could get a sufficiently
large enough set of farmers to go through with such an
extensive ordeal is problematic. Therefore. it becomes a
question of weighing the advantages of accuracy obtained by

a more rigorous test versus participation levels necessary

for statistical comparisons.

Chapter 14
Evaluation of Secondary Outputs

As stated earlier the open-ended questionnaire was
used in the development phase Of the managerial test to
establish behavioral incidents of the farmers selected by
extension personnel as being the more successful farmers in
their respective counties. The method of filtering such
information derived from administering the questionnaire was
a simple numbers game. in that responses were categorized
into similar groups and the groups which had the most
responses were selected as incidents of similar behavior.
However. there were many responses or even side comments by
individual or only a few farmers that hold interest when one

is trying to investigate the decision environment of farm

managers.

14.1 Decision-Making Behavior of Successful
Farmers

This section is dedicated to presenting information
that was not used directly into establishing the managerial
test. but hold interest in the light that it sheds on the
general structure of the problems facing farm managers. The
comments will be divided into four major categories: 1) time
management. 2) structure of decision-making. 3) discipline

and 4)institutional problems.

164

165

14.11 Time Management

The limitations that time presents is an important
element within the problem environment of the farmer. In
general economic theory. the object maximized is profit with
biases towards profits or returns to a fixed asset other
than the manager’s time. The managers interviewed in the
developmental stage of the managerial test point to the
importance of time management with such statements as:

1. ”Know how long to spend on a problem."

2. “Know which times are most critical and allow room for
the unexpected."

3. "Decisions are made quickly but with a lot of thought;
must have feel of situation."

4. “Gathering information takes time. not making deci-
sions."

5. "Computers could bring problems of information over-
load. Good managers will know what information is
important and how much time to spend on a particular
decision.“

6. "Use tape recorder to remember details of problems
while working."

7. "Labor is most constraining factor. Maximize labor
productivity!"

6. "Make preparations. Plan ahead.”

These statements emphasize the need for the manager to
conserve his precious time. However. a conflicting element
arises between saving time and keeping control as evidenced

by the following quotes:

1. ”If you run your own equipment. you can use older
equipment."

 

166

2. "If you do the work yourself you can spot problems more
easily."

3. ”It’s hard to get and watch help. therefore it is
better to do as much work yourself as possible.“

4. "Some farmers allow their wives to keep the books. but
if you do that you won’t know your day—to-day
financial situation.“

Instead of the manager allocating jobs to others. allowing
him to concentrate on decision-making. these statements
suggest that to make the right decisions requires the
manager to do most Of the work. Therefore. there is some
optimal trade off-region between doing the work yourself to
keep control. and allocating jobs to others to save time.

It is the farmers that have established the right combina-

tions. that are successful.

14.12 Structure of Decision-Making
Establishing the right combinations of human resources
and responsibilities requires an understanding of the deci-
sion environment facing the manager along with understanding

the human element.
1. ”Personality combinations are important."

2. "When making joint decisions in a partnership. try to
eliminate as many subjective factors as possible.
Education helps to remove emotionalism from decision-

making.“

3. "A lot of decisions are made for you. it’s a matter of
how you adjust.”

4. "Profits come from higher yields and reduced costs. not
expected higher prices."

5. "Invest in items that reduce the risk in farming."

167

6. "Could have learned about farming by working. but could
learn more being a finance or management major.”

7. ”Changes in environment shifts emphasis from production
and marketing to finance."

The implication here is that more emphasis needs to be
put on dealing with the human problems associated with joint
decision-making. and understanding more about finance and

money along with production agriculture in our teaching and

extension programs.

The importance of establishing and sticking to the right
rotation is an issue expressed by most of the farmers. This
brings into serious doubt the usefulness of marginality and
short term profit maximizing decision models.

1. "Long-run returns over short run expectations...‘

2. “Government programs limit ability to make ad-
justments."

3. ”A field is a system. it is hard to identify isolated
variables."

14.13 Discipline
Self discipline along with communicating the need for
sacrifice to family members are areas pointed out by the

following comments:

1. "... ability to follow through on plans. How often does
one miss a schedule?"

2. "Must be able to sacrifice personal wants for long term
stability."

3. ”Family stress management is needed."

4. "One expands to bring in a family member. but this can
cause a strain on the firm and the family.“

Here the rigours of being in a business which requires

participation by the whole family makes farming somewhat

168

different from other businesses. The farm is also the home.
The development of the farm depends on the life cycle and
aspirations of the family. The kind and direction of expan-
sion can also be separate problems as evidenced by these

statements:

1. “Looking at verticle instead of horizontal expansion.
Try to provide more services.”

2. “Do not expand to carry equipment."
The expansion problem may lead one to delve in uncharted
regions such as providing a further stage in processing. or
the requirements of discipline again forces one not to get

attached to the idea of expansion.

14.14 Institutional Problems
Along with the limitations on crop acreage adjustments
due to goVernment programs. other institutional issues are
of concern to these farmers such as:
1. "Tax policy dictates many investment decisions."

2. "Must pay taxes. Saving on taxes should not influence
equipment investment."

3. "Inexperienced loan officers have caused problem in
PCA's. Their bad advice may have caused over-capitali-
zation."

4. "Loan Officers are over reacting and causing land

values to decrease even more. causing another whole
rung Of foreclosures.”

5. "Loan Officers are jumping to fast from equity
financing to cashflow criteria."

As mentioned in an earlier chapter it was government

policy which seemed to add the most to the unpredictability

169
of returns in the minds of many of the farmers. Here tax
policies and the behavior of the banking industry are felt
to influence the financial health of farms. It would seem
that not only do farmers need to be educated in financial
matters. but more experienced and educated loan Officers
need to be employed in the federal banking system. The
ability of the loan officer to give advice on. and evaluate
financial matters should be in line with the tremendous
risks facing the farmers using the financing. In other
words. to put a "greenhorn" in a position to determine the
financial future Of a farm with substantial assets is
showing a lack of respect of the risk that such a farm
manager is taking. Further. long-term implications Of tax
policy to the over capitalization problem in agricultural
production should be investigated. along with how the
requirements of income disclosure may require information in

a form not compatible for decision-making.

14.2 Implications of Observed Decision-Making
Behavior to Farm Management/Agricultural

Economics
There are implications for farm management training
and agricultural economics research that stem from this
investigation on managerial behavior. The emphasis on mar-
ginality and understanding production function analysis may
be premature to teaching basic concepts of cost accounting
and understanding how to make lumpy investment decisions.

More emphasis should be placed on understanding interper-

170

sonal relationships and developing communication skills
instead of production agriculture. Further. emphasis should
be placed on expanding the managers’ concept of the world so
as to plan for potential problems that begin outside of his
immediate production environment. In other words. less
emphasis should be placed on micro-economic theory and more
on macro-economic and policy environments that affect the
long term survival of the farm business.
14.3 Performance (Effectiveness) Measures
Revisited

In our analysis on determining appropriate effective-
ness measures. other aspects such as workability had to be
considered as well as accuracy. validity. and reliability.
The extensive information needed and adherence to a set
definition is a hindrance to using management income per
acre as a criteria. An information system designed to
measure and record information necessary for the formulation
of the economist's definition of income may not be one
appropriate for determining taxable income. Since. many
farmers may only keep necessary information for filing
taxes. lack of conformity in the computing of profits could
cause problems in making comparisons between farmers. This
is a particularly accute problem when trying to develop a
quasi-experiment in which recording necessary data for com-
parisons on the control groups could pose such non—

comformity problems.

Another issue arises as to why the management test was

171

able to predict management income per acre without the use
of the production or marketing questions. Maybe if another
criteria besides management income per acre were chosen.
then evidence Of production and marketing decision-making
differences that affect that effectiveness measure could be
Observed. In particular. one may want to investigate
whether the respondents differed in yields per acre on their
major crops and if these differences were associated with
farm success. and if these differences could be predicted by
the production section of the managerial test. One could
also compare prices received on their major commodities and
do a similar two stage analysis.

It may be difficult to get farmers to give us enough
information to measure management income accurately and
consistently across a large enough sample for statistical
examination. Therefore. there may be a need to develop or
choose a proxi to this measure which is more workable.
However. such a proxy will have to include a measure of the
assets committed or some other factor which discribes the
asset risk that the manager is facing. Management income
per acre. by giving a charge to the use of capital and a
charge for the use of operator and family labor. does allow
one to compare farms not only between their respective
levels of income but also on the efficient use of assets and
labor committed.

However. management income per acre may not take into

account the level of exposure to the risk Of foreclosure due

172

to the level Of debt carried by the farm. A suggested
modification would be to discount MIA by a factor related to
the debt to equity ratio. A farm with more debt is exposed
to the possibility of forclosure and should be making
relatively more income than a similar farm with less

leverage. if we are to compare farm managers on the basis Of

ensuring long term viability.
14.4 Criteria for Development and Selection of
Decision Aids

There are certain implications of the above analysis
which point to areas of most probable usefulness of compu-
terized decision aids. Cost of production. lumpy in-
vestment analysis. determining how much land to rent and the
appropriate price. and long term implications of a certain
rotational scheme under different assumptions of the future
could be prime candidates for computer aided support. Also.
control of labor utilization could be a major concern when
developing recordkeeping systems. Farms may not need day-
to-day recordkeeping systems. but one providing maximum
utilization for decision-making with minimal updating.

The computer aid must solve the trade-off problem
between accuracy and ease Of use to show benefits. Farmers
are doers. not armchair managers. They do not want to be
bogged down in paperwork. but understand the need of putting
a pencil to paper on major decisions.

The computer system should be designed to aid decision-

making. not replace the tax accountant. The farmer will

 

173

still have to use the accountant for IRS purposes. but the
farmer needs more details on certain areas than required by
the government such as monitoring his and his family’s labor
along with the ability to allocate costs across different
enterprises. The computer system should provide the flexi-
bility for changes according to the user’s specific produc-
tion system. managerial structure and aptitude. In
particular. spreadsheet analysis is preferred over basic or
pascal languages because of the ease of learning and the
potential for the farmer to make adjustments to packaged
templates and to use the spreadsheet to formulated new
tools.

A specific suggestion is that a complete planning.
control and evaluation system for farms be done in an in-

tegrated electronic spreadsheet format. Such a system is

now being developed by the author.

 

PART V

CONCLUSIOINS AND RECOMMENDATIONS

The Management Systems Research model developed and
tested in the above analysis is an attempt to structure a
research process that starts with managerial behavior as the
subject and improving effectiveness as the Object. The

ability of this methodology to accomplish its overall

objectives is the purpose of the following discussions along

with determining how such a methodology can be profitably

used in the future.

174

Chapter 15

Summary of Major Findings

As stated in the Objectives the major purpose of this
research was to develop and test a model which would measure
the potential value of planning tools to farm decision-
making. The systems approach was used to develop the
Management Systems Research model as a methodology for
measuring such potential. In accomplishing the primary
goals other products were produced including a validated
managerial effectiveness measure. common behavioral attri-
butes of effective farm managers. use of economic tools by
farm managers. information on the relationship between
behavior. performance. and effectiveness. and information on
the criteria for decision aid selection.

The Management Systems Research model developed
included five major activities: 1. Selection Of Farm System.
2. Behavioral Analysis. 3. Measuring Effectiveness.

4. Testing Behavioral Predictors Of Effectiveness. and 5.
Quasi-Experiment. The first step of selection of a farm
system begins with categorizing farms into types. The USDA
classification scheme was chosen as a first cut which
consists Of cash grain. field crop. vegetable and melon.
fruit and tree nut. nursery and greenhouse. dairy. poultry

and egg. and cattle/hoglsheep. Since the Michigan State

175

176

TelFarm program combined the cash grain and field crop
categories. and this classification seemed to produce
adequate homogeneity across the farms considered. field crop
and cash grain were considered one category.

Once a farm system or type was selected. an Open-ended
questionnaire was developed and administered to a group of
successful cash grain farmers as identified by extension
personnel in Ingham and Tuscola counties. Information from
the interviews was used to develop a managerial index ques-
tionnaire. and helped in determining the appropriate effec-
tiveness measure for the farm manager.

The farmers selected profitability as the most impor-
tant criteria for evaluating a farm manager. Therefore.
management income per acre was chosen as a measure of effec-
tiveness. and this measure was tested for workability.
reliability. and validity. The reliability test consisted
of testing for consistency of the measure over time in
identifying the TelFarm cash grain farmers in terms of
managerial groups. The results of ANOVA showed that the
management grouping accounted for 36.7 percent Of the
variance in management income per acre over the four year
period between 1961 and 1964. while grouping according to
the year accounted for only 0.16 percent of the variance.

Management income per acre was checked for validity by
testing the relationship between the average management
income per acre for each Of the farmers against locational

and farm size variables. The results of multiple regression

177

demonstrated that the average county corn yield. location.
capital invested per acre. and average acres farmed per
manager were poor predictors of the average management
income per acre over the four years tested.

With a validated managerial effectiveness measure and a
managerial index developed. the next step was to administer
the test to the TelFarm cash grain farmers and test the
relationship between behavior. performance and effective-
ness. The results of the analysis demonstrated that
managerial effectiveness could be predicted by a subset of
the managerial questions focusing on cost realization and
financial decision-making. The specific financial
questions included in the financial management index (FINBP)
were 19. 27. 29. 30. and 31.

Question 19:

19. Assume that interest rates are 12%. and your fertilizer
dealer is offering a 10% discount in December for fertilizer
that you would use on corn in the spring. Would you? (Select

one)
_1_a) Borrow money to take advantage of discount

___b) Wait to purchase until spring

The more successful farmers chose a" which showed

their ability to use credit wisely to take advantage of

discounts. Therefore. they expressed an understanding of

the present value concept in evaluating short-term

purchasing decisions.

Question 27:

27. A good reason for renting more land would be: (Rank 3)
_2_a) Spread fixed costs

_1_b) Increase labor utilizaton
_c) The rental land is close to own land

178

___d) To test productivity before deciding to purchase
___e) Act as buffer while making equipment systems change

The more successful farmers chose "b" and a" in that
order. Therefore. these farmers expressed the need to
maximize labor utilization and to use rental land for a
flexible way to expand in order to better utilize equipment.
They also showed a preference for expansion through renting
instead of buying and minimized the use of rental land as a
way to control a piece of property for future purchasing.
The decision rule then filters down to using rental land as
a flexible way of expanding to better utilize intermediate
assets and labor.

Question 29:

29. If at the end of tax year you had 820.000 more taxable
income than expected. would you? (Choose one)

___a) Purchase machinery or equipment to get tax credit

_2_b) Pay taxes on the income and put the rest on the
mortgage

_c) Invest in more land

___d) Spend more on consumption

_1_e) Buy more inputs for next year to reduce this year’s

taxes
The more successful farmers chose "e" - to buy more
inputs for next year. and "b" - put the rest on the

mortgage. These farmers would not allow the desire to
reduce taxes to force them into increasing their long term
debt load. In fact. the more successful farmers would use
such unexpected profits to reduce their debt and increase
only short-term assets. Therefore. the decision rule seems
to be not to make long-term investment decisions based on

the desire to reduce short-term taxable income.

179

Question 30:

30. Indicate how many years you would finance the following
items: (If less than one year indicate with a fraction)

Years
a) Production Equipment __3.5____
b) Land --22__-
c) Production inputs
used up in one year __.5___
d) Machinery shed _-5-_-
e) Personal car _--5---

The conservative use of credit by the more successful
farmers is expressed through the short length of time they
would be willing to finance long-term assets. This conser-
vative nature conflicts with the economic or financial
management suggested policy of matching long-term financing
to long-term assets. The unwillingness Of the more
successful farmers to use such a principle reflects their
understanding of another finance principle of matching risk
with returns. These farmers understand the variable nature
of the returns to agriculture. and are not willing to over-
extend themselves facing such variability. The benefits of
financial leverage are discounted by the more successful
farmers with their knowlege of the riskiness of that
leverage. Investment discipline as a managerial quality is
highlighted by their response to this question. As a
decision rule. these farmers seem to be saying to finance a
long-term asset for a considerably shorter length than the
expected life of that asset.

Question 31:

31. Assume you have the option of renting land from someone
on fixed cash rent or shares. Choose the alternative you
would prefer under the following conditions. (Indicate ”F"

180

for fixed cash rent and "S" for shares.)

_S__ Prices are volatile

_F__ Prices are stable and high

_S__ Prices are stable but low

_S__ The land is not well drained

_S__ The land does not have a ASCS corn basis
_S__ Don’t know productivity of land

The more successful farmers chose fixed rental arrange-
ments only with a high degree of confidense in expected
favorable returns. Therefore as a decision rule. the lack
of knowlege of the productivity of the rental land and/or
volatility in prices is a cue to seek crop sharing type of
arrangements which would allow them to share that risk with
the landlord.

The farm organizational and historical development
variables also played significant roles in predicting MIA.
In particular. the farms that were purchased instead of
inherited and the incorporated or jointly owned seemed to
have higher mananagement incomes per acre. This could have
some serious implications toward the viability of sole
propietor farms in comparison to their more organized
counterparts. It may be that the inclusion of other people
into the decision-making process increases the managerial
resource pool which in turn increases the viability of the
farm. It also seems that the farm managers that puchased
their farms instead of inheriting them do a better job at
managing

Now that it seems possible to make apriori predictions
of managerial success using an index. the stage is set for

using such an index for measuring the possible benefits of

181

improved decision aids. The remaining test is to compare
the measure obtained from the management index against
Observed differences between a set of farmers over time: one
set using the computerized system and the other without such
aids. One such quasi-experiment is set to be conducted in

North Carolina with the farmers involved in the North

Carolina A&T Farmer Opportunities Program.

Chapter 16

Evaluation

The objective of the process started in this study is
to evaluate the role of planning tools in farm decision-
making. The present study presents the conceptual framework
and carries through such a methodology up to the quasi-
experimental phase of implementation. The total process can
be considered to have three phases: 1. Conceptual framework.
2. Implementation. and 3. Evaluation of the two methods for

measuring value of information.

16.1 Conceptual Framework Evaluation

The conceptual framework for Management Systems
Research consisted of five parts: 1) Selection of Farm
System. 2) Behavioral Analyis. 3) Measuring Effectiveness.
4) Testing Behavioral Predictors Of Effectiveness and 5)
Quasi-Experiment. At the end of the quasi-experiment the
measure Of the value of information or decision aids is
compared using the "classroom" or index measure against the
“field" or experimental method.

The systems framework seemed to work well in struc-
turing the problem and controlling the flow of events.
Therefore. in general. the general systems framework seems

to work as a methodology to develop and control this type of

182

183

model development. The Management Systems Research model
developed consisted of five parts of which the last four are
separate research topics in themselves. The present study
attempted to test the first four parts of the model while
laying the groundwork for the last - Quasi-Experiment.

In this iteration of the model the selection of the
farming system was helped by the pre-existence of a set of
farms with available historical effectiveness measures. The
selection of future farms for such an analysis must weigh
the problem of accessibility of information to the loss of
randomness and the ability to extrapolate findings to a
wider universe that such data requirements produce.

No distinctions were made between large and small
farms. In the case of cash grain farms in Michigan this may
not cause a problem in the homogeneity of the decision
environment. However. in other areas and with a different
set of major commodities the question of size or scale may
pose homogeneity problems.

The behavioral analysis section can prove to be the most
informative in producing subject matter information on how
the more successful farmers make decisions. However. this
portion of the model presents a different set of problems
for future replications. The development of the open-ended
questionnaire requires background information on the nature
of the important decisions made by that farm type. Such
background information on what are the major decisions may

not be readily available. In the present implementation

184

the results of a previous thesis were available that
provided the needed information. Therefore. the time
involved in establishing the open-ended questionnaire may
have to be expanded on future replications.

Although. only one measure was tested for validity and
reliability in the measuring effectiveness section of the
model. other measures could be evaluated. Since the work-
ability Of different measures may vary across farm types. it
is advisable to test more than one in case there is need for
a proxi to take the place of the more preferred measure.

Validity and reliability were tested statistically
using linear multiple regression analysis and analysis of
variance. However. other methods not presented here could
be tested against the present ones chosen for answering the
questions Of the appropriateness and objectivity of an
effectiveness measure.

Testing the behavioral predictors of effectiveness is
by far the most frustrating portion of the analysis since
the possible combinations of questions can be substantial.
A more efficient form of factor analysis could be employed
to simplify such a task. However. doing the factor
develOpment by hand does have some merit due to the ill-
defined boundaries between what is a production. marketing
and financial question along with the fine distinction
between behavior and performance. Going over the data more
closely allows one to get a feel for the decision

environment and the nuances of that environment facing the

185

farm manager.

Since the quasi-experimental part of the model was only
developed and not yet tested. it is difficult to evaluate it
based on the experience of the present study. However. some
possible problems with implementation are discussed in the

following section.

16.2 Implementation Evaluation

It was thought that the testing of the model on cash
grain farmers in Michigan would act as a model for replica-
tion Of the total methodology in North Carolina. However.
the results derived from testing the management index
against the effectiveness measure indicated that much of the
first steps in the process need not be duplicated for North
Carolina. In particular. administering the open-ended
questionnaire was done to develop an appropriate test for
that type of farm operation. The production and marketing
problems facing farmers may be more varied across locations
and farming systems than are financial type problems.
However. since only the financial managerial type questions
showed any relationship to success. and these questions are
general enough to apply across a wide diversity of farm
types and locations. it is felt that the managerial index
developed in this research can act as the nucleus for one to
be used in North Carolina.

Therefore. instead of repeating Selection Of Farm

System. Behavioral Analysis. Measuring Effectiveness. and

186

Testing Behavioral Predictors of Effectiveness. one can
start with the managerial index and proceed to the Quasi-
Experiment. Of course. the more appropriate way would be to
replicate these first steps and see how the index developed
from the replication compares with the present version.
However. this would take another full year if the necessary
information was available. This is not the case in North
Carolina. North Carolina does not have enough field crop or
cash grain farmers that have participated in a mail-in
recordkeeping system for a long enough period to carry out
the complete analysis.

Not only is the implementation of the complete model a
problem in North Carolina. but the question of workability
arises in other states since there are a limited number of
states that have programs such as Michigan’s TelFarm
program. Another procedure could be to ask the farmers for
financial information on the same managerial test sent to
them. To determine management income per acre requires
information from both the balance sheet and income state-
ment. along with estimates of family and operator labor
actually used on the farm during the analysis period.
However. experience has shown that farmers are either reluc-
tant to divulge such information in mailed survey or do not
have such information readily available.

The question of workability of the methodology also is
important for implementation of the quasi-experiment even in

North Carolina where at least a possible set of participants

187

may already be identified. A quasi-experiment is time con-
suming. One would need at least three to four years of
continuous Observation to be able to draw any real conclu-
sions. This of course takes money on the part of the
research institution and cooperation between research.
extension personnel and the farmers themselves. Due to the
financial crisis and high degree of attrition among farmers.
one can expect a significant number of the participants to
drop out before the end of the experimental period making
problems Of comparability between experimental and control
groups and statistical significance of the results very
important.

Finding and keeping willing participants is made more
difficult because Of the time involved in learning new
managerial techniques and keeping good records. The deci-
sion aids chosen must require the minimal amount of time and
effort for use until their benefits can be proven to parti-
cipants. Although it was hoped that one could administer
the managerial questionnaire a total of three times to the
experimental group. this may not be practical since this may
overburden the participants with tests. Also. it should
take at least one year before the user Of the decision aid
is both comfortable with using it and appreciative of its
value. Therefore. the managerial test should probably be
administered only twice to each group: once at the initia-
tion of the experiment and at the end.

Fortunately. the majority of the North Carolina farmers

188

in the Farm Opportunities program are field crop farms and
some with a few cattle or swine. Although. the replication
of the total methodology may not be necessary for carrying
out the quasi-experiment and subsequent determination of the
potential value of information for these type farms. the
procedures would need to be completely replicated for more
specialized farms such as dairy. feedlot. horticultural
crops. and orchards. In these systems production and
marketing could be either more essential for success and/or
more diverse within the categories.

The model can be evaluated between its potential
payoff and its recognized costs. The model was able to
identify incidences of behavioral incidents and develop an
instrument that showed promise in predicting managerial
effectiveness. The purpose of such a test was to eliminate
the necessity in measuring the value of a decision aid by
using a quasi-experiment. Hopefully. the results of the
quasi-experiment justify the substitution of such an expen-
sive and time-consuming method with such a managerial test.

From begining to end the methodology tested here on
Michigan cash grain farmers took one year to implement.
However. since the results of testing the relationship
between effectiveness and managerial index indicated that
specific production and marketing decisions. which would
limit applicability of the results to cash grain farmers.
did not play a major role with this set of farmers in

predicting effectiveness. a more generalized grouping of

189

farms across production-defined boundaries could limit the
need for the number of replications. It would seem that
many Of the financial and money management type questions
could be applicable to a wide variety of farm types. And
such questions that deal with cost of production could be
set up differently for each farm type. In other words.
instead of asking a feedlot Operator to figure the cost of
production for corn. a similar question could be posed on
cost of weight gain per animal.

In fulfilling the model’s primary objective in setting
up the structure for determining the value of computerized
farm management aids. the secondary outputs may prove to be
more important for long range planning for research. exten-
sion and teaching. The subject matter information developed
from such a methodolgy presents a current picture of the
managerial environment and means of coping with such en-
vironment by farm managers. The understanding of the
decision-making behavior of farmers can point to whether the
major problems facing them are of a production. marketing.

financial or some other as yet unclassified nature.

Chapter 17

Reccommendations

This research has produced information that may have
significant bearings on how the Land Grant system conducts
extension. research and teaching especially in the manage-
ment or economic areas. In general the focus of our efforts
should be on profitability rather than increasing output.
The farmers analyzed in this research seemed to be
somewhat homogeneous in the production and marketing
decision-making. Their effectiveness could only be
predicted by their measured ability to answer financial
management and cost of production type questions. However.
since all of these farmers were on a particular record-
keeping system. one might suspect that this group was biased
toward the more sophisticated set of farmers. In other
words. a more random group Of farmers may show more variance
in their production and marketing decision-making. On the
other hand. since these farmers xggg on a recordkeeping
system one could argue that the sample was biased in a
direction that would minimize the difference in financial
decision-making or any other type Of decision-making that
required financial data.

Therefore. although we can not be sure that this group

is representative Of all cash grain operators. the evidence

190

191

points to certain adjustments in the emphasis placed by the
Land Grant system on production agriculture. Replication of
this methodology over a broader area or in different parts

Of the country should add to the reliability and applicabi-

lity Of the findings and suggestions to a wider audience.

17.1 Extension

The findings of this research point to the extension
service providing more training to farmers on financial
decision- making. Particularly. emphasis needs to be placed
on investment analysis. credit policy. organizational
structure. and determining cost of production. This would
probably mean an increase in the number of trained extension
personnel in these areas.

The question remains as to the level of involvement of
extension personnel in the role of consulting. The accepted
role Of extension has been to educate. not advise. particu-
larly when it comes to financial matters. Extension can
walk that thin line between education and consulting by
providing the tools and helping in their application without
giving specific advice. In making decisions. the succesful
farmers exhibit similar rules of thumb or decision rules
which extension personnel could suggest as useful tools for
managers with similar problems. The computer could be a
useful tool in storing such decision rules. i.e. expert
systems. and applying such rules through appropriate

softwear. This is not to suggest that extension should not

192

continue to be involved in educating farmers on new produc-
tion technology. but such education should include how to

measure the profitability of such technology.

17.2 Research

Research on decision-making behavior takes time. money
and ”people skills". Although micro-economics deals with
relationships between firms. it does not go into the depths
necessary for predicting individual behavior. Therefore.
the tools developed using the micro-economic approach should
not be oversold to the individual decision-maker until they
have been evaluated for workability and usefulness. A major
component of workability for these tools is the availability
of relevant and reliable internal data from the firm. More
behavioral research should be undertaken to understand why
many farm managers do not keep adequate records.

When the emphasis is on increasing output on a national
basis. the knowledge of individual decision-making may be
minimized. However. if emphasis is placed on making agri-
culture more profitable. then research on behavior and
individual response to macro level policy and environment is
warranted as well.

In the past not much research has been done on the level
of adoption of management tools. nor on the realized
benefits of such tools. The present study points out that

studying the managerial tools as a technology towards

193

increasing profitability is needed and can be handled within
a manageable framework. Further. along with more research
on the managerial tools themselves. it would seem that the
Land Grant institutions should do more research on the
implications of their production-oriented research toward
profitability. For instance. economic analysis of the field
trial results of a new production technology should be
implemented and presented along with the physical results of
such trials. If that is not practical. then at least infor-
mation and analysis tools should be presented that would
allow the farmer to make such evaluation.

There are some other specific topics for research which
this present study suggests. Along with the replication of
the testing of behavior and performance against effective-
ness in other states. other questions arise within the scope
of the present data set. For instance. although the produc-
tion and marketing sections did not predict managerial
effectiveness as measured by management income per acre. it
would be interesting to know if this portion Of the index
could predict some other effectiveness measure. In
particular. the production questions may predict yields per
acre of corn on the individual farms. and the marketing
questions may predict variances in the price of corn
received. A corollary of this question is the issue of the
association of yield per acre and price received per bushel
with management income per acre. These further analyses

could help determine both the reliability of the total index

194

and help in understanding the importance of production effi-
ciency and marketing as related to financial success.

The present research has pointed out the general
conservative nature of successful farmers towards credit.
Agriculture in the past has exhibited both low returns to
assets and high levels of variability in returns. More
research is needed in the area of understanding the rela-
tionship between risk and returns in agricultural production
in order to develop tools that would give farmers assistance

in analyzing the riskiness of potential investments.

17.3 Teaching

The general findings of this research along with some
specific statements by some Of the farmers interviewed
indicate that farmers are realizing the importance of a
business oriented education instead Of one emphasizing how
to maximize production. In particular. if the schools of
agriculture do not provide financial management training as
a part of their programs. it is likely that farmers. serious
about farming as a money-making business. may gravitate to
the business schools.

Although micro-economics is a useful tool in evaluating
sector response. more courses on individual decision-making
behavior must be made available to help the real world
manager. Even within micro-economics. the tools and
theories should be given substance by relating them to

problems faced by the decision-maker and not just the

195

curiosity of the academician. Of course. the level of
relevant teaching and research will depend on the career
insentives faced by the economic and management profes-
sionals.

The economic tools which should be stressed more
include investment models - such as the present value concept.
money and banking. optimization models - such as linear
programming. Opportunity cost. cost/benefit analysis and
decision theory within a value of information framework.
Linear programming and the concept of opportunity cost are
most helpful for major transitional type questions such as
adding a new enterprise. expansion and determining the level
of off-farm employment to pursue.

With the present distress in agriculture and the
chronic low returns to assets facing the farm manager. the
importance of financial management training cannot be over-
emphasized since it seems that success is not just a
function of productivity but also the ability to control
assets. The possibility of high productivity with low
financial returns is a fact of life in agriculture and a
threat to the survival of the individual farm firm.
Therefore. a well rounded program for the farm manager
should include cost accounting. recordkeeping, and simple

systems analysis.

196

17.4 Farm Policy

One of the findings Of the present study indicates that
the farm managers who purchased their farms instead of
inheriting them do a better job at managing. This could
have policy implications to how the present financial crisis
in agriculture is handled. It may be an economically
beneficial strategy for the government to buy the mis-
managed farms and sell them to other farmers or the same
farmers once they have completed training in financial
management and can pass some type of managerial test.
Success in agricultural production may now be a factor Of
knowlege - managerial and scientific. and less a function of
hard physical work. The new manager must be sound in his
"scientific" ability to manage as well as skillful in the
"art" Of management.

The "art" of management as it relates to farming may be
difficult to teach in a classroom setting. This type Of
human capital may best be developed through the experience
of working on the farm. However. financial managerial
ability may be augmented through an intensive classroom
type program centered around computerized decision aids.

The value of such training and tools will be the output of

the intended quasi-experiment.
17.5 Recommendation for Model Application

Along with the use of the Management Systems Research

model for its primary intended use. i.e. establishing a

197

measurment system to determine the value Of improved
planning tools or decision aids. the model serves an added
purpose in focusing attention on the problem areas facing
managers and the areas where extension. research and
teaching should focus efforts. Since the environment of
the farmer changes. such a methodology used on a recurring
basis could help pinpoint future problem areas. For
instance. some conditions may warrant emphasis on production
agriculture. while other conditions such as the present
financial crisis warrant emphasis on cost reductions and
wise investment decisions. Such an iterative methodology
can serve as a barometer of these changes.

Another issue that is highlighted by this form of
analysis is that farm management advisors and teachers can
learn from the successful farmers what is needed by the less
successful farmers to improve the effectiveness of this set
of farmers. It may be more realistic to bring the less
successful farmers to the level of their more successful
peers than it would be to transform these farmers into the
”economic man" or any other type of ideal manager posited by
our given professions as the ideal. There must be a
compromise between bringing the manager up to the level of
using our sophisticated economic tools and develOping
managerial tools that are both useful and useable. The
model or methodology presented in this study could prove
useful in developing and refining such tools. The model at

present was able to produce a few general non-quantified

198

decision rules. but a more detailed iteration Of the model
could be used to help develop expert systems for farm

management.

17.6 Preliminary Assessment of The Role of
Microcomputers in Farm Decision-Making

The final output of the implementation of the total
measurement system or model is to measure the potential
value of computers to farm managers. Although the present
analysis was only the first step in the total process,
information obtained does reflect on the overall issue of
the role of computers in farm decision-making.

Computers are efficient at manipulating. storing and
presenting numerical data. Many aspects of production require
experience and a tactile awareness of the physical produc-
tion environment for efficient manipulation of the control-
lable inputs. Marketing requires a system and discipline
with a major part of that system being a knowlege Of one’s
cost of production. Financial management requires a firm
knowlege of the dollar and cents aspects Of the farm system.
From the present analysis it was discovered that the factors
that could predict effectiveness were those which required a
firm grasp of the numbers - dollars and cents - of the farm
system. Therefore. computers could play a major role in
helping the farm manager with the very areas that may
determine the farm’s economic survival. i.e. keeping up with

the cost of production and the financial aspects of the

199

firm. Although the quantified benefits of computers has not
yet been established. it is in the areas of financial
management and enterprise cost control that one can expect
to find the greatest monetary payoffs for the role of micro-
computers in farm decision-making for cash grain or field
crop farms.

The fact that all the farmers used in the present
analysis were on a monthly mail-in recordkeeping system -
TelFarm - presents other implications. Having good records
is not enough to ensure sound management. The ability to
use such information and develop the "right" prescriptions
is essential. Therefore. future computer applications
should not be confined to recordkeeping but should include
planning and monitoring. This would mean the development Of
planning tools such as simulation that allows for the
flexibility in use accompaned by training in financial

analysis. cost control and organizational planning.

APPEND I CES

APPENDIX A : OUTPUT OF OPEN-ENDED QUESTIONNAIRE
Table I
Keys to Success
N=17

Categories Dimensions Mentions

 

 

 

1. Financial management: 9 1
Managing money

Know and control costs

Good source of financing

Profitability

Deal with one lender

Cash flow

Don’t buy high priced land

Only own land that can be brought up to
standards

No debt load

HHHHwbmm

9 asawo°w>
g...

2. Production: 7 1
Stewardship of land

Keep up with new technology

Blend of labor. capiral and crops

Solid production practices

Adequate machinery

Necessary functional buildings

Adequate size of operations

Q'UFIU(TU3>
HHHHNNNO

3. Timeliness: 4

Do it when necessary
Good health

Be able to work

Plan ahead

U<1u1>
Hr40)0

0

4. Decision Making: 7
A. React to higher interest rates 8 lower
prices
Flexibility
Be able to handle unpredictability
Set up plan and track expenses
Analyze what was done
Do things yourself to know problem areas
Pin point problems

Q'UFIU()U!
Hrahswrdnsw

5. Attitude: 6

Managing people

Have to enjoy it

Attitude

Sort needs out from wants
Willing to do office work
Ambitious

Hrahawtvh30

'HFIU(1UI>

200

201

6. Communications: 4
A. Keep contact with outside
B. Promotiong of farming
C. Involvement in organizations
D. Listen to consumers
6. Good Marketing: 1
7. General: 4
A. Be good in prod.. mktg. and finance
B. Off-farm income
C. Asset valuation
D. Participation in government programs

bHHwaz

Hut-step

 

 

:iJ._

202

 

 

Table II
Criteria For Measuring Management
N=17

Type Category Dimension Mentions
A. Objective: 6 19
1. Profitability 1 5
2. Production costs 1 3
3. Assets and liabilities 1 3
4. Yields 1 3
5. Fin. returns on assets 1 3
6. Time 2 2
a. How got started 1

b. Comparison over years 1

B. Subjective: 4 27
1. Appearance 5 11
a. Physical view of crops 4

b. Appearance Of farmstead 3

c. Machinery appearance 2

d. Tile lines and drainage ditch appearance 1

e. Errosion 1

2. Production 4 9
a. Timing 4

b. Production Practices 3

c. Time efficiency 1

d. Machinery use efficiency 1

3. Financing 3 5
a. Lines of credit 2

b. Pays bills on time 2

c. Have handle on finances 1

4. Community status 1 2
C. Behavior: 2 19
1. Managerial ability 5 12
a. Extensive 8 good records 4

b. Marketing ability 3

c. Knowing financial status 2

d. Planning ability 2

e. Preventive measures 1

2. Attitude 5 7
a. Dedication 2

b. Accurate and honest 2

c. Willingness to change 1

d. Ability to listen 1

e. Good attitude 1

203

 

 

Table III
Like and Dislike About Farming
N=17
Like Menstions Dislike Mentions
1. Own boss 15 Unpredictable long hrs. 4

2. Feeling of accomplishment 10 Returns not commencerate

 

 

 

with risk 4
3. Being outdoors 6 Uncertainty 3
4. Closeness with family 4 Market risks 3
5. Trying new technology 2 Need too much money to

Operate bus. 2
6. Flexibility Of jobs 2 Financial pressures 2
7. Honest work 1 Lack of security 2
6. Feels knowledgable in this 1 Tied to physical process 1
9. Like people in agr. 1 Government control 1
10. Close to basic elements 1 Reduced no. Of suppliers 1
11. Habit 1 Cold weather 1
12. Attachment to land 1 Mental stress 1

Table IV
Information Processing
N=17
Processing Method
How Often Farm Computer Hand
done Computer Service Records Total

Yearly 6 6 5 17
Quaterly 0 3 1 4

Monthly 5 1 1 6

 

204

 

 

Table V
Good Yields in Corn
N=14
Factors Mentions
Fertilization 12
Variety 9
Timing of Planting 7
Improved Soil Productivity 6
Pest Control 5
Tiling/drainage 3
Other 6
Table VI
Time for Harvesting and Planting
N=15
Factors Mentions
Adequate and Reliable Equipment 10
Preseason Preparations 9
Diversity of crops 8 maturity dates 6
Run untile done 6
Experience 5
Two way communications 5
Make sure you have 1 month to do 15 days work 3
Switch between crops while planting depending
on weather 3
Reduce slack only if planning to change equip-
ment capacity 3

 

205

 

 

 

 

 

 

 

Table VII
Land Investment Strategy
N=6
Item Mentions
Stand alone returns vs. price F6: land 4
Only when available 3
Forced to bring others in 3
Size of equipment 2
Ground must be real cheap 2
Disinvest in high valued property 2
Low down payment. low risk 2
Security in ownership 2
Other 9
Table VIII
Credit Policy
N=7
Item Mentions
65331735373553? """"""""" " 4
Use accelerated payments plan 2
Don't borrow money for inputs 2
Borrow to maintain cash flow. operating capital 2
Other 12
Table IX
Crop Marketing Philosophy
Factors Mentions
ESEE'EIGE """"""""""""" “ 7
Corn/wheat in government program 4
3

Staggered Sells

206

Table X
Proportion Of Crops to Produce
N=16
Item Mentions

 

 

 

and for overall soil productivity 12
Government programs 11
Returns per dollar 9
Control ground for sugar beets 5
Spreading work load 5
Spreading price and weather risks 5
Equipment 2
Soil types and productivity 2
Other 5

Table XI
Substantially Changing Crop Mix
N=12
Item Mentions
Extremely high contract offer 9
Good government program insentives 6
Spread labor and equipment - time allocation 4
To help an errosion or weed problem 3
Riskiness of crop 3
Cost of production 2
Other 5
Table XII
Determining Corn Variety
N=14
Item Mentions

 

Use different varieties to stagger
maturing dates

Use MSU yield trials

On-farm test plot results

Own experience

Take advise from dealer

Use 3 year comparisons

Phase in new variety over 3 years

Fertility requirements Of each variety

Spread risk

Experience with chemical 8 prod. practices

Other

wNNNNthmmo

207

 

Mentions

 

10

Table XIII
When to Purchase Inputs
N=11
Item
"HE; """""""""""""""""""""

Cash discounts

Buy early to get lower prices and insure

availability
Interest rates
Cash flow effects
Trends in prices
Have well ahead of needs
Volume discounts
Adequatcy of on-farm storage

 

Other
Table XIV
When to Sell Corn
N=10
Item

10

Mentions

 

 

Cost of production

Trigger selling
Price must pay storage and interest or

let government have it
Sell to keep good line of credit
Scheduled or spread out sells
Don’t guess the market
Other

9
010

208

 

Table XVI
On-Farm Storage
N=11
Item Mentions
Ease harvesting bottleneck — 9

 

 

 

 

 

 

Keep on farm waiting for elevator shortage 4
Allows choice between government and market 4
Cheaper to store on farm 4
Prices lowest at havest 2
Always store at elevator 2
Other 2
Table XVII
Price of Corn
N=6
Item Mentions
ESJSFEEEEE’Ioan program sets price ---- _S—
Useless to guess price 3
Other 4
Table XVIII
Participation in Government Programs
N=10
Item Mentions

 

 

Always in corn program. if not then plant
something else

Anticipated income in vs. out of program

Put set aside on low yielding land

Own basis - past participation

Other

U'INNUIG)

 

 

209

 

 

 

Table XIX
Buying Land
N=13
Item Mentions
Cost per month vs. expected returns 11
Quality Of land 3
Compare monthly cost to total land cost/acre 2
Interest rates 2
Transportation - proximity 2
Terms 2
Other 5
Table XX
Pay for Rental Land
N=17 '
Item Mentions
Rental terms - fixed price vs share. risk.
management and landlord requirments 5
Lease first to test productivity before buy 4
Longer lease to build of productivity 4
Quality of land 3
People seek me out 3
Expected yields 3
Expected prices 3
Cost 3
Past ownership practices 2
Proximity 2
Spread fixed costs 2
10

Other

210

 

 

 

 

 

Table XXI
Equipment Replacement
N=17
Item Mentions
Productivity evaluation 12
Ability to afford 10
Dependability - esp. at prime time 7
Repair bills 5
Buy used equipment only 4
Constantly not getting job done on time 4
Backup equipment for spare parts 3
Run until it dies 3
Purchase for expected growth in acreage 3
Take good care 3
Size for job 3
Keep older equipment only if you have the
ability to repair them yourself 3
Must have better equipment when using
hired labor 2
Retire equipment before loosing all resale
value 2
New technologies force acquire new equipment 2
Depreciation used up 2
Compatability between equipment and land 2
Other 12
Table XXI
Information Processing
Extensive Extensive
Decision Calculations Historical Records
Prop. of crops 6 6
Add crop 6 1
Corn variety 3 6
Purcase inputs 6 4
Sell corn 14 0
Storage 0 0
Price of corn 0 0
Govern. programs 7 5
Land investment 11 0
Pay for rental land 7 1
Replace equipment 6 6

 

211

Table XXIII
Constraints on Making Decisions

 

Constraints
Internal External Government Time Wheather

Decision Data Data Policy

Prop. crops 2 5 6 4 2

Add crop 4 7 3 4 _
Corn variety 1 2 1 1 _
Input purchase 2 4 3 _ 2
Sell corn 3 2

 

APPENDIX B: OPEN-ENDED QUESTIONNAIRE

I. Description of Farm and organization
1.Type and size of enterprise. 3 people involved in D-M.
2. How long have you been farming?
3. How did you get started in farming?
4 Do you have Off-farm employment? How many hours/week?
A General Philosphy
5. What do you concider the major keys to successful
farming?
6. What do you like and dislike about farming?
7. What criteria would you use to evaluate whether a farm
manager is successful or not?

II. Management Information System
6. Reports; type.purpose. schedules
9.Processing: equipment. resonsibilities

III. Decision Making
A. Production - 10.What is your major crop?
1. General
11. What do you concider the keys to getting good yields
in corn! wheat! soybeans?
12. How do you ensure that you will have enough time to
get your planting and harvesting done ?

ii. Specific Problems or decisions
13. Determining proportion Of crops to produce.
14. Determining whether to substantially increase
or decrease the acreage Of your major crop?
15. Determining corn variety.
16. Determining when to purchase inputs.

B. Marketing - Crops
1. General

17. What is your general philosopy about crop marketing?

16. Would you concider either production or marketing as
more or less important for the overall success of

the farm business?

19. What marketing alternatives do you feel most
confident with. i.e. forward contracting. cash. hed-
ging etc.? End

ii. Specific Decisions

20. Determining when to sell your corn/soybeans.

21. Determining the adequatcy of on-farm storage
capacity?

22. Estimating the price you expect to receive for corn?

23. Determing whether to participate in the relevant
government corn program?

C. Financing and Capital Investment

212

 

 

i O

24
25

2.

26.
27.
28.

29.
30.

213

General
. What is your general investment strategy for land?

. What is your general credit policy?

Specific Probems

Determining whether to expand or reduce overall opera-
tions?

If you have decided to expand how do you decide on
either rental or purchase of land?

Determining how much to pay for rental land.
Determining the need to replace equipment.

Determining whether to do self repair on equipment and
buildings vs paying outsiders.

IV. Specific information or questions for each specific problem

1.

Planning - Identification, observation. analysis,
decision.

A. What factors would a "good" decision maker in
deference to a ”poor“ dec151on maker concider in making
this decision?

B. What calculations would be appropriate for solving

this problem?

C. Do you spend a lot of time collecting and analysing

information for this decision? C1.How much?

D. How constraining are the following factors on making
a good decision:

Small Large

Knowledge of how to make decision 1 2 3 4 5

Internal data from farm

External data (prices.quality etc.)

Government (policy. taxes etc.)

Data Processing capabilities

Time to make decision

Uncertainty from results of prior Dec.

Uncertainty from outside sources

Other - specify

(DOVOUIIBCONH

E. What are your major sources for external data?
F How would you or what would you need to help you
improve the way you make this decision?

Mon1toring-Control: implementation. observation. pro-
blem id..analysis. decision

G. What signals do you look for that might require
substantial alterations of your plans?

H. How do you monitor or keep up with these signals?

Evaluation-Learning:
I. What criteria do you use to evaluate whether
this was a good or bad decision? How often do you

evaluate?

 

.I

APPENDIX C: DECISION-MAKING ANALYSIS QUESTIONNAIRE

I.Background Information

N
I
O
t
B
m
D
V
3'
0
C
n
m
'0
0
fl
4
m
(I)
x
D.
O
K
0
C
t
O
.1
x
O
m
DI”
n
D'
m
is
m
’1
3
x)

3. DO you have two-way communication equipment to communicate
between the house and field?

6. Type of organization (circle): a) sole proprietorship.
b) partnership with brother(s). c) partnership with children.
d) corporation. e)other(specify)

7. Are you planning to expand your operations to bring a family
member into the farm business within the next 5 years? ________

8. How many farm organizations are you a member of? __________
9. How many community organizations do you participate in? ______

10. How did you acquire the majority of your owned land? (Check
one)
___a) Inherited
__b) Bought from family member(s)
___c) Bought from non-relatives
__d) Other (specify)

11. How many years have you been farming independently? _______

II.Production Decisions

(Note: Many of the questions in this and following sections will
ask you to rank the alternatives given. Please rank the top
three most appropriate by numbering them 1.2 or 3.)

12. If you had to grow corn in another part of the state what
would be your primary source Of information on the best varieties
Of corn to plant? (Rank 3)

___ a) Seed dealers

b) Farmers near your new farm
c) Extension
d) Past experience on old farm
___ f) Local elevators

___ 9) Farm journals
13. How many varieties would you plant?

214

215

14. If you found that you were consistently falling behind on
getting your planting done on time. what would be the first
thing you would try to eliminate the problem? (Select one)

a) Start planting earlier

b) Get larger equipment

c) reduce acreage

d) diversify crops and/or varieties grown
e) Hire more help

f) Put in more hours

15. In which crop is downtime the least expensive during harvest?

(Rank 3 least)
a)soybeans __ b)wheat __ c)corn ___ d)navy beans

e)sugar beets ___ f)barley ___g)seed corn

16. Assuming that other crop prices remain the same. how high
would corn prices have to be projected for the coming year before
you would double your acreage of corn?

Give high price ______ .

17. How low would corn prices have to be projected as falling in
the next year to get you to cut your acreage in half?

Give low price _____ .

16. If you were interested in adding a crop to your present crop
mix. what factors would most influence that decision? (Rank 3)

_a) Projected average price for next 5 years
_b) Equipment requirements

_c) Production practices and costs

_d) Compatability with present rotation

_e) Labor requirements

___f) Lowest price over last 5 years

___g) Government programs

19. Assume that interest rates are 12%. and your fertilizer dealer
Offering a 10% discount in December for fertilizer that you would
use on corn in the spring. Would you? (Select one)

___a) Borrow money to take advantage of discount
_b) Wait to purchase until spring

20. Soil tests calls for 100 lbs of nitrogen on corn for your
type Of soil and variety of corn. Would you? (Choose one)

_a) Use a bit more
___b) Use as recommended
___c) Use a bit less

21. Do you know your cost of production for each of your major
crop? _____

22.

Cash Expenses: per acre

Labor (Hired)
Labor (Own 8 family)

216

If yes. what was your cost of producing for corn in 1965?

Fixed expenses or overhead:
Per acre

Land Rental Charge
Depreciation

Repairs _____ Interest on debt ____
Seed _____ Insurance ____
Fertilizer and lime _____ Taxes _--_
Pesticides _____ Utilities ___-
Drying costs _____ Other _--_
Fuel _____
Other _____
Total ------------ Total -__-
Total cash and overhead ____
Yeild per acre ___-
Cost per bushel -___
III. Marketing
23. What are your reasons for entering the government loan

program? (Rank 3)

_a)
_b)

___c)

24.

_d)
_e)

Sets price floor

Cheap source of operating capital

Is an assured market

Can utilize marginal land as set aside
Never entered

If the government got out Of the corn market. the best
strategy for selling corn would be: (Rank 3)

___a)

25.

_b)
‘c)
_d)
_e)

_r>

-9)
__h)
_i)

The

_a)
_b)
_c)
_d)
_e)

Use market information to pick highest price

See what neighbors are doing

Sell at harvest

Sell some corn every month

Set trigger above cost of production and begain to sell
when price reaches that point

Sell when market is going up

Sell just after market starts going down

Store on farm until elevators offer premium

Sell as cash is needed to pay bills

advantage of on-farm storage is: (Rank 3)

Can participate in government loan program
Saves time during harvest

Can get better price by holding

Can store cheaper than at elevator
Utilizes labor in Off season

 

 

1d

IV.

26.

How Often do you prepare a projected cash flow?

_a)

___b)

27.

28.

_c)
_d)

A good reason for

_a)
_b)
_c)
_d)
_e)

217

Investment and Money Management

(Select one)

Yearly
Monthly
Quarterly
None

renting more land would be: (Rank 3)
Spread fixed costs

Increase labor utilizaton

The rental land is close to own land
To test productivity before deciding

Act as buffer while making equipment

to purchase
systems change

In deciding to purchase a piece of land up for sale which of

these statements best describes how you would go about making the

decision?

29.

than expected.

30.

a)
b)
c)

_a)
_b)
_c)
_d)
_a)

Production Equipment
Land

Production inputs
used up in one year

(Select one)

I have been renting it for a while and if I don’t buy it
someone else will and get the benefits.

I would compare the expected net returns over the life of
the mortgage compared to monthly payments and purchase
only if it pays for itself.

I would determine if the equity base in my present
operation could subsidize the purchase of the land.

If the expected net returns per acre are greater than my
average cost of land. I would purchase.

If at the end of tax year you had 520,000 more taxable income

would you? (Choose one)

Purchase machinery or equipment to get tax credit

Pay taxes on the income and put the rest on the mortgage
Invest in more land

Spend more on consumption

Buy more inputs for next year to reduce this year’s
taxes

Indicate how many years you would finance the following
items:

(If less than one year indicate with a fraction)
Years

218

d) Machinery shed
e) Personal car

31. Assume you have the option of renting land from someone on
fixed cash rent or shares. Chose the alternative you would
prefer under the following conditions. (Indicate "F" for fixed
cash rent and "S" for shares.)

Prices are volatile

Prices are stable and high

Prices are stable but low

The land is not well drained

The land does not have a ASCS corn basis
They want a long term lease on good land
Don’t know productivity of land

32. Assume that you have enough equipment and labor to work an
extra 40 acres of land and you currently grow a combination of
corn. wheat and soybeans. Also assume that a piece land comes up
for sale that historically produces 100 bushels of corn. If the
finance terms are at 10% fixed interest for 15 years. how much
would you be willing tO pay per acre? _________

33. In choosing the following types of equipment pick 3
attributes from the list which are most important when making the
decision to purchase. (Place letter of attribute next to number)

Equipment Attributes

Price

Age

Condition

Financing arrangements
Depreciation
Investment tax credit
Resale value

Size

Dealer service
Availability of spare
parts

Availability Of back up
equipment

Performance qualities
Dependability

."State of the art"
Flexibility of uses
Own ability to repair
Horsepower

Lease Options
Compatability with other
equipment

Planter 1 2 3

Tractor 1 2 3

Harvest
Truck 1 2 3

Combine 1 2 3

urdnio'flFIU(UUJ>

UJSUD'TJOZZIq N

34. How much repair work do you do on your equipment on your
farm? %

35.

36.

219

How many different lenders do you do business with? ______

V. General Management

Rank the usefullness Of doing extensive and accurate numeri-

cal calculations when making decisions in the following catego-
ries. (Circle 1.2.3.4 or 5 where 1 means not useful and 5 means
very useful)

a)
b)
c)
d)
e)

f)
g)

h)
i)
j)
k)
l)

37.

Not Very

useful useful
Determining crop mix 1 2 3 4 5
Adding a new crop 1 2 3 4 5
Determining crop varieties 1 2 3 4 5
When to purchase inputs 1 2 3 4 5
Cost Of production for individual
crops 1 2 3 4 5
Estimating the future price of corn 1 2 4 5

Determining the economic level of

fertilizer to use 1 2 3 4 5
Determinig whether to enter a

government program 1 2 3 4 5
How much to pay for land 1 2 3 4 5
Choosing between financial options 1 2 3 4 5
Making an equipment purchase 1 2 3 4 5
Estate planning 1 2 3 4 5

What are the most important keys to being successful in

farming? (Rank 3)

_a) Attitude: motivation. dealing with people etc.

_b) Timeliness: getting things done on time

_c) Managing money and finances: controling costs. managing
credit etc.

d) Productivity: solid production practices. stewarship of
land. adequate buildings and machinery etc.

e) Decision-making: ability to react to change. planning,
pin pointing problems etc.

f) Communications and public relations: keep contact with
outside world, be able to get expert help etc.

38. How many years of historical records on the following items
do you have that are readily available?
NO. of years Item

a) Crop performance by fields

b) Variety performance

c) Soil test data by fields

d) Fertilizer utilization for each crop
e) Chemical utilization for each crop
f) Land cost per acre

g) Fuel utilization per acre

220

h) Yearly equipment repair cost for each peice of
major machinery
1) Hours of planter use

39. Which of these statements best describes why you stay in
farming? (Rank 3)

___a) Family tradition
_b) Feeling of accomplishment
_c) Like to control assets
_d) Attachment to land
__e) I’m a "Workaholic"
___f) Locked in by assets
___g) Don’t have other skills
___h) Offers high standard of living
_i) Like working with family

BIBLIOGRAPHY

BIBLIOGRPHY

Brannen, Stephen J. "Structural Change of the Individual
Farm". P eedin 5° Struct of Southern Farms of
The Future. The Agricultural Policy Institute. N.C.
State and Alabama CES. A.P.I Series 30. Aug. 1966.

Campbell. John P.. Marvin D. Dunnette. Richard D. Arvey. and
Lowell V. Hellervik. "The Development and Evaluation
of Behaviorally Based Rating Scales." Jour al 0

Applied Psyghologx 1973. Vol 57. No. 1. 15-22.

Campbell. J.P.. Dunnette. M.D.. Lawler, E.E.. and Weick. K.E.
a a e 'a Behav O ' f man e and Ef ec ivene
New York: McGraw-Hill. 1970.

Capouch. Brian. "Field Records Systems”. On r Com.v lse:

Conferencg Proceedings. Nov 22-23. 1962. Purdue U.
School of Agriculture. W. Lafayette. Ind. P-257-263.

Davis. Gordon B. Management Information Systems: Conceptual
Eogndgtiong. Stguctnge, and DevegonmentsI New York:

McGrawHill Book Company. 1974.

 

Debertin. David L.. Gerald A. Harrison. Robert J. Rades. and

Lawrence P. Bohl. "Estimating the Returns to
Information: A Gaming Approach.” Amggicnn Journal
r' l m' 57(1975):316-321.
Debertin. David L.. Robert J. Rades. and Gerald A. Harrison.
"Returns to Information: An Addendum." Ameciggn
Joucnal Q; Agricultngal Economigs 56(1976):321-
323.

Doster, D. H..”A Computerized Management Information System?".

Qn Fagn Connnte; Use; anteggngg Pgoceggings. Nov 22-

23. 1962. Purdue U. School of Ag. W. Lafayette.
Ind.

Drucker. Peter F.. TechnolggyI Management and Sggiety. Harper

Colophon Books. 1977.

Ein-Dor. Phillip and Eli Segev. A Paradigm for Mana emen

 

Information Systems. New York. N.Y.: Praeger. 1961.
Fischhoff. Baruch and Don Macgregor. ”Subjective Confidence in

Forecasts". Jgnrnal Qfi Fogecasting. Vol. 1. 155

172 (1962).

221

222

Fiske. Emmett P..”On Developing A Comparative Framework For
Assessing The Needs Of Limited Resource Farmers:
The Western Washington Farmer Survey.“ Paper
Presented at The Annual Meetings of The Rural
Sociological Society. Aug 17-20. 1963. Lexington.

Ky.
Forster. D. Lynn. "Developments in the Economic Theory of
Information: Discussion“. Amegignn Jnnrnal gfl

Agrignltural Egongmigs. Dec 1976.

Fuller. Earl. "The Computer As A Tool To Improve Day-To-Day
Control of The Hog Enterprise". On F rm m ute

Use: gonzggence Eggceegings. Nov 22-23. 1962.
Purdue Univ. School Of Agriculture. W. Lafayette.
Ind.

Goldman. Stanford. Information Theory' New York: Dover
Publications. Inc. 1953.

Grethe. David M. and Charles R. Plott. "Economic Theory of
Choice An The Preference Reversal Phenomenon". AER.
Sept 1979 69(4): 623-638.

Hansen. Art et n1. "Farming Systems of Alchua County. Florida:
An Overview With Special Attention To Low Resource
Farmers". Center For Communitv and Rural Develgn-
nent. Institute Of Food and Agricultural Sciences.
U. of Florida. Gainesville. Fla. Jan 1981.

Harman. Wyatte L.. Vernon R. Eidman. Roy E. Hatch and P. L.
Claypool. "Relating Farm and Operator Characteris-
tics to Multiple Goals”. Sou er Journ l of Adri-

c t a ' . Vol 4(1): 215-220. July. 1972.

Heady. Earl 0. "Management In Relation to Agricultural
Adjustment and Economic Growth". The Management Innnt
in Agriculture. Agricultural Policy Institute.

April 1963.

Heady. Earl O. and John L Dillon. A ‘i u tu P oduc '
Enngtions. Ames. Iowa State U. Press. 1961.

Hepp. Ralph and Thomas Olson. ”Information Needs and Sources
for Michigan Small Farm Operators." Agricultural
Economics Report No.372. Department of Agricultural
Economics. Michigan State University. E. Lansing. Mich.
March. 1980.

Henderson. Dennis R. "Market Information: Some Research
Issues.” ma '0 n 'c e rt'

the Food and Agzignltncal Sectgz. Proceedings of

 

 

223

conference sponsored by North Central Committee
117. Madison, Wis. 1979.

Henderson. John C. and Paul C. Nutt. "The Influence of
Decision Style on Decision mking Behavior." Management
Sgiengg 26(1980): 371-386.

Holland. David. ”Production Efficiency and Economies of Size
In Agriculture.” a e V 1‘ N on Ru 1 Center

Small Earm Egoject National Rural Center.
Washington. D.C. 1980.

James. H. Brooks and Charles R. Pugh. "Forces Affecting the

Structure Of Southern Agriculture." Ins gtrngtuge nfi

§outnern anns 9; the Future. Proceedings of The
Agricultural Policy Institute. N.C. State and Alabama CES,

1968. A.P.I. 30.

Johnson. Glenn L. "Agricultural Economics. Production
Economics and The Field of Farm Management". Jnurnal of

Eagm Egongmigs. Vol. 39(2): 441-450. May 1957.

______ "Methodology for the Managerial Input.’ The Manage—
nent Input in Aggignltnge. Agricultural Policy
Institute. Southern Farm Management Research
Committee. Farm Foundation. 1963.

______ "Some Lessons From The IMS." Paper presented at the
Agricultural Development Council. Inc. Conference
on Risk and Uncertainty in Agricultural Development.
CIMMYT. Mexico. 1977.

_______ .Research Methodology for Econemists. E. Lansing.
Mich: Dept of Agri. Econ.. Michigan State U.. 1982.

Johnson. Glenn L.. Albert N. Halter. Harald R. Jensen and D.

Woods Thomas. A Study of Managegial Processes of
Midwestecn Fagnegs. Ames. Iowa: Iowa State University

Press. 1961.

 

Johnson. Glenn L. and Lewis K. Zerby. What Economists Do About,

Va u s-- ud h n wers t 0 es '0 s
They Don’t Dare Ask. East Lansing. Mi: Michigan

State University. 1973.

Keen. Peter G.W. and Michael S. Morton. Decisinn §unnort

§ystems: nganizgtinnal Pegsnegtivg. Reading.
Mass.: Addison-Wesley Publ., 1978.

King. Robert P. "Technical and Institutional Innovation in
North American Grain Production: The New Informa-
tion Technology“. Discussion Paper 316. August
1984. Strategic Management Research Center.
University of Minnesota.

224

______ ."Farm Information Systems: Needs. Methods. and
Responsibilities". Presented at the North Central
Regional Farm Management Extension Workshop.
Champaign. Illinois. May 7-9. 1985.

Kleijnen. Jack P. G. gomnntggg nng nggggg; Quansifxing
' anc' Ben ‘ s f n O mat' . Mass.: Addison-

Wesley Pub. CO. 1980.

Kline. R. G. "Effect of Personal Abilities on Coefficient
Levels“. e ana me u ' A 'cu t e.
Agricultural Policy Institute. Southern Farm
Management Research Committee. Farm Foundation.
1963.

Leagans. J. Paul. "Adoption of Modern Agricultural
Technology by Small Farm Operators: An
Interdisciplinary Model for Research and Strategy

 

Builder5"- W “69-
Ithaca, N.Y.: Cornell Univ. 1979.
Manetsh. Thomas J. and Gerald L. Park. Sygten Analysts and
'mu '0. w' ' ' ns Ec nom‘ and Soc'
Systems. E. Lansing. Mich.: Dept. of Elect. Eng.

and System Science. MSU. 1982.

Mcgrann, James M. "Introduction to Microcomputers : Farmer
and Rancher Use". Unpublished. 1982.

_______ "Texas A8M’s Role in Computer Software Development:
Where We Are and Where Is the Future". Unpublished.
also in SRDC Conference Proceedings. 1982.

Mintzberg. Henry. "The Manager’s Job: Folklore and Fact".
flarvnrg Busine§§ Egvzeg. 53(4): 49-61. 1975.

Mu’min. Ridgely A. "Design Of Decision Support System For
Hrabal Farm”. Unpublished. Summer 1984.

Murphree. Clyde E. "Discussion Of Forces Affecting The
Structure of Southern Agriculture". P 0 .din : .e
Structure ofi Sgutnern Facns of Ibé Fntncg. The Agricul-
tural Policy Institute. N.C. State and Alabama CES.
A.P.I Series 30. Aug. 1968.

Nielson. James and William Crosswhite. "The Michigan
Township Extension Experiment: Changes in Agricultural
Production. Efficiency and Earnings". East Lansing.
Mi.: Michigan State University Agricultural Experiment
Station Bulletin 274. 1959.

Olson. Thomas. "Nonformal Education Delivery Systems to
Reach Limited Resource Farmers in Michigan."

225

Unpublished Ph.D. thesis. M.S.U.. Department of Agri-
cultural Economics. East Lansing. Michigan, 1978.

 

Orden. David R. A An I of So conomic Characte
R sour ana enen P ac 'ces and Productivity_
efl Small Farm Onegetors in Floyd; and Brunswick County
Elr rglnle. M. S. Thesis in Agricultural Economics.
Blacksburg. Virginia: VPI 1977.

   

 

Organization for Economic Development. Soglal Aeeesssment of
Technology. OECD. 1978.

Osborn. Howard A. "Technology and The Small Farm: A Concep-
tual Framework In Production Efficiency and Technology
For Small Farms". Eener VI Of The Nati nal u al Center

Small Earm Pgojegt. Washington. D.C.: National Rural
Center 1980.

Perrin. Richard K. ”The Value of Information and The Value of
Theoretical Models in CrOp Response Research".
Amer. Jl Aqr. Econ; 56(1976): 54-61.

Pudasaini. Som P. "The Effects of Education in Agriculture:
Evidence From Napal". Amen. J. of Aqr. Econ;
65(1963): 509-515.

Purdue University Bulletin. Cooperative Extension Service.

Farmer Qe§l£§§i What Are The Priorit¥_ltems7 A
Survey of Computing Indian; Farnege. July-Oct. 1981.

 

Richards. Max D. and Paul S. Greenlaw. Manegement Decisions
and Behavior. Homewood. 111.: Richard D. Irwin,
Inc. 1972.

Roberts. Dayton Young and Hong Yong Lee. "Personlizing
Learning Processes in Agricultural Economics."
Amer. J. of Agri. Econ. 59(1977):1022-1026.

Rohrbach. Norman F. "Microcomputer Use in Teaching Graduate
Students in Agricultural Education." Ph.D. disserta-
tion, University of Missouri-Columbia. 1983.

Rosenberg. H. 2. and James R. Gray. "Reduction of Price

Variation in a Producer’s Market." Jou a 0 Farm
Egononlge. 44((1962):219-221.

Schroeder. Emily Harper. "The Lifestyle Dimension of Small-
Scale Farming: Farming As A Form of Consumption"
Presented at the annual meeting of The Rural Sociology
Society. Aug 17-20. 1983. Lexington. Ky.

Shackle. G.L.S. Dec' ' rde and Time In Human Afifelrs.
Cambridge; Cambridge University Press 1961.

 

 

226

Smith. P.. and L.M. Kendall. "Retranslation Of Expectations:
An Approach to the Construction of Unambiguous Anchors
for Rating Scales." ourn o - ied P 'cr lo ,
1963. 47. 149-155.

Sonka. Steven T. Comnutegs in Fapmlng. New York, N.Y.:
McGraw-Hill 1983.

Sprague. Ralph H. and Eric D. Carlson. Bu' d'n E ct ve

Deglslon Senngtt Systeme. Englecliff. N.J.: Prentice-

Hall 1982.

Steichen. Joe C. "Panel Discussion on What's Happening on
Southern Farms: A View From The ’Back 40’".
Proceedings; The Structure of Southern FELQS of The

Fu 1 e. The Agricultural Policy Institute. N.C. State

O

and Alabama CES. A.P.I Series 30. Aug. 1968.

Teece. David J. and Sidney G. Winter. "Economic Theory and
Management Education: The Limits of Neoclassical
Theory In Management Education". gee, 74(2): 116-
121. May 1984.

Toda. Masanao. "The Decision Process: A Perspective".
’ . 1976, Vol 3:

 

Vogel. John. "Big and Small Farmers Are Economies and

Attitudes Apart". Ezeigie Farmeg March 7. 1981. p 14.

Woodworth. Roger C. Sammy L. Comer and Richard D. Edwards.
“Comparative Study of Small. Part-Time. Retirement and
Large Farms: Three Counties in Central and West
Tennessee". School of Agriculture and Home Economics.
Tennessee State Univ.. Nashville. Tenn. Feb 1978 BUll.

no. 38.

""TilllfllfljllfilltfylflfmlflllS