I
I

 

DERIVATION OF ESTIMATES OF MARGINAL FACTOR COST
FUNCTIONS FOR CREDIT FOR CASH~CROP, DAIRY,
AND BEEF CATTLE FARMERS IN SELECTED
AREAS OF MICHIGAN 1960

Tstis for the Douro. of Ph. D.
MICHIGAN STATE UNIVERSITY
Sidney C. Bell
1961

 

 

‘ CCT 1, 51993

- .‘vfl:

 

IIIIIIIIIIIIII IIIIIIIIII

This is to certify that the

thesis entitled

DERIVATION OF ESTIMATES OF MARGINAL FACTOR COST
FUNCTIONS FOR CREDIT FOR CASH-CROP, DA 1? \Y, AND
BEEF CATTLE FARMERS IN SELECTED AREAS OF MICHIGAN

presented by

Sidney C. Bell

has been accepted towards fulfillment
of the requirements for

Ph.D. degreein Ag, Econ.

Jaw M

Date My 31. 1961

0-169

DERIVATION'GF ESTIMATES oF MARGINALIEACTGR cosr FUNCTIONS
'Foa.annIr:Foa case-c309,:nAIRY; AND BEEF CAITEE
'FARMERS IN SELECTED AREAS OF MICHIGAN
1960

by
Sidney c. Bell

AN ABSTRACT

Submitted to the School for Advanced Graduate Studies
of Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

mm (IF PHILOSOPHY
Department of Agricultural Economics

1961

Approved [41M L

 

Th
describn
for cas}
areas of
personal
tore, th
supply 0;
The team
affect tt
”tinted

The .
“PM vet-I
for dairy I
Stratified
obtain . a

Seve
the 1Ildiv
These equ .
dependent
equation I
credit obt

A be
fitted for

“dared 1T:

ABSIRAOT

The primary objective of this study was to derive and
describe estimates of the marginal cost functions for credit
for cashncrop, dairy and beef cattle farmers in selected
areas of Michigan. The necessary data were collected by
personal interview from.these three types of fauna. There-
fore, these functions represent the farmers' estimate of the
supply of credit available to then.at various interest rates.
The secondary objective was to determine the factors which
affect the quantity of credit the different types of farmers
estimated they could borrow.

The areas selected.to represent the different types of
farms were Saginaw County for cashpcrop, St. Clair County
for dairy, and Lenawee County for beef cattle farms. .A
stratified sample was randomly selected within each area to
obtain a sample to represent the farlers of that area.

Several regression equations were fitted.to the data of
the individual type farms and to the data for all fauna.
These equations were fitted.with quantity of credit as the
dependent variable. This was done to derive a best fitting
equation with the variables that affect the quantity of
credit obtainable at a series of interest rates.

A.best fitting equation was selected from.the equations
fitted for each type of farm and for all farms. Items con-

sidered in selecting the "best fit" were, 11) the adjusted

N

 

mltiple
standard .
coefficieI
(4) value
(5) distri
regression
independex
The 1
factors at
they can I)
I3) avails
ramp, (5!
(7) ‘89 an
The vI
were used I
to derive I

I
the Cash. C 1'

multiple correlation and determination coefficients, (2)
standard error of estimate, (3) number of variables whose
coefficients were significant and level of significance,
(4) value and sign of estimated regression coefficients,
(5) distribution of the residuals when plotted about the
regression line, and (6) the intercorrelation among the
independent variables. ,

The selected equations indicate that the important
factors affecting the quantity of credit farmers estimate
they can borrow were, (1) net worth, (2) interest rate,

(3) availability of land contracts, (4) credit rating of
far-er, (5) gross farn.ineome, (6) not farm income, and
(7) age and education. I

I The variables selected.in fitting the above equations
were used in fitting regression equations with the interest
rate as the dependent variable. These equations were used
to derive the marginal factor cost functions for credit for
the cashpcrop, dairy and beef cattle farmers. The function
most representative of the data for cash-crop farmers was
curvilinear, while the ones for dairy and beef cattle
farmers were straight line functions.

The marginal factor cost functions were also derived for
the individual types of farms from the data of all farms
using the types of farms as independent variables. The

results of fitting these functions indicated they were not
iii

useful as a function to represent a particular type farm.
The two most probable uses of the results of this study

are in the areas of: (1) research work -- (a) in farm

budgeting and programming and (b) supply response work, and

(2) extension work with farmers in the field of credit.

Some secondary uses of these functions might be, (1) in the

teaching field and (2) as an aid to various lending agencies.

DERIVATION Q ESTIMATES fl? MARGINAL PAM 0081‘ FUNCTIONS
FE. “IT Fm CASH-CEO, DAIRY, AND BEEF CATTLE
FARMS IN SEIECTED AREAS fl? MICHIGAN .
1960 . .

. by
Sidney C. Bell

A THESIS

Submitted to the School for Advanced Graduate Studies
of Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of
NOTE! a? PHILOSGHIY
Department of Agricultural Economics

1961

To
My Wife, Doris
And my three children
Betty, Bonnie and Bonnie

 

T

their I
The ad
the out
with hi
Am

the Depa
in the f
The
received
this the:
R- 1:. GUS
tistical 3
Appre

the rough
°f Agl'icul

 

 

ACKNOWLEDGMENTS

The author is grateful to the many persons who gave
their time and assistance in the development of this study.
The guidance of Dr. Glenn D. Johnson was very helpful and
the author is especially grateful for having been associated
with him for the period of his graduate study.

Appreciation is expressed to Dr. L. L. Boger, Head of
the Department of Agricultural Economics, for financial aid
in the form of a research assistantship.

The author would like to acknowledge the assistance
received from Dr. John Brake who read the rough drafts of
this thesis and made many helpful suggestions. Also to Dr.
R. L. Gustafson who aided in formulating part of the sta-
tistical procedure used in this study.

Appreciation is expressed to the secretaries who typed
the rough draft and to the statistical pool of the Department
of Agricultural Economics for doing the computational work.

CHAPTER

TABLE OF CONTENTS

I C IN TRODUCTI ON 0 O O O O O O O O O O O O O O O 0

Organization of Thesis . . . . . . . . . . .
Objectives of the Study . . . . . . . . . .
Description and Location of Sample Areas . .

Relevant characteristics of Saginaw County

Relevant characteristics of St. Clair County

Relevant characteristics of Lenawee County
Survey areas . . . . . . . . . . . . . .
Stratification of sample . . . . . . . .

Field Techniques . . . . . . . . . . . . . .
Design of the Questionnaire . . . . . . . .
Reliability of the Data . . . . . . . . . .

Processing the Data . . . . . . . . . . . .

II. ESTIMATING MARGINAL FACTOR cos'r FUNCTIONS FOR

CREDIT FOR CASH-CROP, DAIRY,AAND BEEF CATTLE
General Description of the Cash-Crop Farms .
General Description of the Dairy Farms . . .
General Description of Beef Cattle Farms . .
Simple Correlations Among the Independent
Variables . . . . . . . . . . . . . . . .

'Regression Equations Fitted . . . . . . . .

PAGE

COQQOSUIH

11
12
12
13
15
17
20

SI?

33

37
41

First fit . . . . . .
Cash-crop farms . .
Dairy farms . . . .
Beef-cattle farms .

Second fit . . . . . .
Oath-crop farms . .
‘Dairy farms . . . .
Beef-cattle farms .

Third fit . . . . . .
Cash-crop farms . .
Dairy farms . . . .
Beef-cattle rm. .

Selection and.Acceptability of

Equation 8 o s o o o e
Cash-crop farms . . .
Dairy runs a o o o 0

Beef cattle farms . .

Comparison of the best fits

Best Fitting

The Marginal Factor Cost Functions

Cash-crop farms . . . . . . . .

Dairy farms . . . . .
.Beef cattle farms . .

III.

CREDIT FOR CASH-CBGP, DAIRY AND BEEF CATTLE

ESTIMATING A MARGINAL FACTOR COST FUNCTION FOR

FARMERS FRCMDATAFCEAILFARMS . . . . . .

ix

42
43

43

45

47

49
49
5O

51
51
53
57
59
66
66
70
72

75

 

IV. St)“:

BIBLIOGIWI

Ge

SI

Re

Se

Re

PI

 

 

 

General Comparison of the CaShPCrop, Dairy,and
Beef Cattle Farms . . . . . . . . . . . . . . 75
Simple CorrilationS»Amang»the‘Indepondent
Variables . . . . . . . . . . . . . . . . . . 79
Regression Equations Fitted . . . . . . . . . . 81
‘First fit . . . . . . . . . . . . . . . . . . 81
Second fit . . . . . . . . . . . . . . . . . . 82
Selection and Acceptability.of Best Fitting
Equation . . . . . . .'. . . . . . . . . . . . 84
The Marginal Factor Cost Functions . . . . . . . 87
IV. SUMMARY.ANDICONCBUSIONS . . . . . . . . . . . . . 91
I 'Results and Conclusions, . . . . . . . . . . . . 92
Selecting the best fitting equations . . . . . 94
Summary of factors affecting quantity.of
credit . . . . . . . . . . . . . . . . . . . 96
Summary of factors affecting interest rate . . 101
Marginal factor cost functions . . . . . . . . 102
f Possible Implications of Results . . . . . . . . 109
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . 115
APPENDIX........................118

 

l.

2.

 

LIST OF TABBES

TABLE PAGE

-

1. .Average Tota1.Acreage, Years of Farming, Age,
and.Education by Size of Operation, 29
Cash-crop Farms, Saginaw County, Michigan,
1960 O O O O O O O O O O O O O O O O O O O, O O O 25

2. Average Off-farm Income, Gross Farm Income,
Net Farm Income, and Total Income, by Size of
Operations, 29 Cash-crop Farms, Saginaw
County, “Chigan,'19601soweosoeeooo 26

3. Average Credit Rating, Not Worth, and Quantity
of Credit, by Size of Operation, 29 Cash-crop
Farms, Saginaw County, Michigan, 1960» . . . . . 27

4. Average Tota1.Acreage,‘Years of Farming, Age
and Education, by Size of Operation, 30 Dairy
Farms, St. Clair County, MiChigan, 1960 s o c . 29

5. Average Off-farm Income, Gross Farm Income, Net
Farm Income, and Total Income by Size of
Operations, 30 Dairy Farms, St. Clair»County,
M1ch1gan,1960.................30

6. Average Credit Rating, Net Worth, Interest Rate,
and Quantity of Credit, by Size of Operations,
30 Dairy Farms, St. Clair County, Michigan,

196000ooeooooooooosodoosee32

7. Average Total Acreage, Number of Feeders, Years
of Farming, Age, and.Education, by Size of
Operation, 29 Beef Cattle Farms, Lenawee County,
Michigan,1960.................33

8. Average Off-farm Income, Gross Farm Income, Net
Farm Income, and Total Income, by Size of
Operation, 29 Beef Cattle Farms, Lenawee
County, MIChigan,1960....-... cases 35

9. Average Credit Rating, Not Worth, Interest Rate,
and Quantity of Credit, by Size of Operation,
29 Beef Cfittle Farms, Lenawee County, MiChigan,

1960 O O '3 O O O O O O O O O O O O O O O O O 0 O 36

 

10.

11.

12.

13.

14.

16.

 

10.

11.

12.

13.

14.

15.

16.

Simple Correlations Among the Independent
Variables, 29 Cash-crop Farms, Saginaw
County, Michigan, 1960 . . . . . . . . . .

Simple Correlations Among the Independent
Variables, 30 Dairy Farms, St. Clair
County, Michigan, 1960 o o o o o o o s o 0

Simple Correlations Among the Independent
Variables, 29 Beef Cattle Farms, Lenawee
County, Midhigan, 1960 . . . . . . . . . .

Average Size, Years of Farming, Age, and
Education, by Type of Farm, 88 Farms,
Selected.Areas of Michigan, 1960 . . . . .

Average Off-farm Income, Gross Farm Income,
Net Farm Income, and Total Income, by Type
of Farm, 88 Farms, Selected Areas of
MIChigan,1960..............

Average Credit Rating, Net Worth, Interest
Rate, and Quantity of Credit, by Type of
Farm, 88 Farms, Selected Areas of Michigan,

1960 O O O O O O O O O O O O O O O O O O 0

Simple Correlations Among the Independent
Variables, 88 Farms, Selected Areas of
Michigan,19600000000ooeoooo

xii

38

39

4O

76

77

79

80

..... .\ ‘
1 K
..... C ‘ ‘
K K
0000000 ‘ K
A K \
K 1
oooooo
K t ‘
x t ’
V. K
' K
K
‘
1 \
I oooooooooo
K A

 

FIGURE

1.

2.

Outll
OI

Mar .2

Mar;

 

 

FIGURE

1.
2.

3.

6.

LIST 0F FIGURES

Outline Map of Michigan Showing Location
ofSampleAreaS...........

{Marginal Factor Cost Functions for Cash-
Crop Farmers . . . . . . . . . . . . .

Marginal Factor Cost Functions for Dairy
FarmerS...............

Marginal Factor Cost Functions for Beef
CattleFarmerS............

Marginal Factor Cost Functions for Cash-
Crop, Dairy, and Beef Cattle Farmers,
Derived from Data for All Farms . . .

Marginal Factor Cost Functions for Cash-
Crop, Dairy, and Beef Cattle Farmers,

Derived from Data for Individual Type Farms

Marginal Factor Cost Functions for Cash-
Crop, Dairy, and Beef Cattle Farmers,
Derived from.Data for Individual Type
Farms Using Grand Mean Values . . . .

PAGE

10

69

71

74

89

106

107

CHAPTER I
INTRODUCTION

Credit plays a very important part in the general field
of agriculture and is indispensable to most individual
farmers. The adoption of many new and improved scientific
and technical developments in recent years in MiChigan's
agriculture has required larger and larger capital investments
in establishing and maintaining an economical size farming
unit. Capital, in many cases, must be obtained through regu-
larly ostabliShed lending institutions in the form.of credit.

The relative importance of credit and the need; in-
resoarch for empirically derived marginal factor cost func-
tions of credit were the factors that brought this problem
to the author‘s attention. There has been much speculation
about the shape and.type of marginal factor cost function for
credit that farmers are facing in the capitAl market.

In analyses employing the assumptions of static theory
of a perfectly competitive firm, the supply of credit is
often assumed to be unlimited at the going interest rate.
‘Under these assumptions, the quantity of credit used does
not affect the interest rate. Hence, the individual farmer

 

1Hildebrand., Peter E., Farm Organization and Resource
IFixitz: Modifications of the Linear Programming Model,

'Unpublished Ph.D. TheSIE, Michigan State University, 1959 and
ZMeKee, Dean E., Economic A. raisal of Ad ustments in Dair

in.Mich;gan to Meet Chang ng Conditions, ect current
progress at chigan State U varsity.

I.

 

-2-

can acquire unlimited funds by paying the market rate of
interest. These assumptions have been used in models of
agricultural organizations. ‘For example, Clark.Edwards!
thesisz, makes limited.use of this assumption, stating,—

I'perfectly elastic, continuous supply functions

are used for variable services in model 1.

Numerous imperfections in the factor and money

market exist in the farm economy and perfectly

elastic functions for inputs are far less real-

istic than perfectly elastic demand functions

for products. The perfectly elastic functions

are frequently used in analysis of the farm

economy and they are included here for compari-

son with the results of the other models.”
Edwards used two other types of supply functions in his
models. One was an upward sloping continuous supply func-
tion. The other function had a point of discontinuity. The
supply function was formed by the acquisition function for
quantities of services greater than the initial quantity on
the farm and.by the salvage function for quantities less
than initial, the point of discontinuity being dotormined.by
the initial use of the service. This problem.has been
treated by M; KaleCki, where he discusses the effect of risk
on cost of credit.3

In most linear programming models used for farms, the

supply of credit is assumed.fixed or a quantity is assumed

 

zEdwards, Clark, Resource Fixitzt Credit Availability
and.igricultura1 Organizat on, Unpubl s ed . es s,
c an State Un vers ty, 1958. w .

, .3Kalecki,tM;, "The Principle of Increasing Risk,"
Economics, XI (new sor., 1944), pp. 55-62. I

 

III Ill

-3-

available to carry out certain programs for the farm. Smith4
assumes in one of his models that the farmer was willing and
able to borrow 37,500, plus whatever extra funds may be
required for carrying one feeding program to completion simul-
taneously with starting a new let (should systems taking more
than a year prove profitable). I

In recent linear programming work, there has been an
improvement over the methods mentioned above. For example,
Hildebrand5 in his study of modifications of the linear pro-
gramming model assumed a supply function for credit based on
credit for land mortgage, credit for purchasing additional
land (there were 2 land contracts, one based on 6 per cent
interest, the other on 7 per cent both requiring 10 per cent
down payment): a chattel mortgage and credit from.machinery
and silo dealers. The sources of credit had varied interest
rates with interest paid annually.

In another linear programming study, Dvorak used an
assumed supply function for credit.6 He compiled a supply

curve for credit based on values from previous studies

 

4Snith, Victor_E., FPerrectve Discontinuous Input
Markets,“ Journal of Farm Economics,.V01. 27, (August 1955)
p. 538. A ,

5Hildebrand, Peter E., 9p. eit.

6Dvorak, F. E., Progra% ’ g‘The‘Organization and Capital
Use For a Cash Crop Farm n t aginaw Valle: and Thumb Area

of Michigan, Unpubl Shed MAS. Thesis, c gan State Un ver-
8 toy, 1959o _

 

 

.\

-4—
(which used assumed values), or suggested values by bankers,
machinery dealers, and professors of Michigan State University
acquainted with the area. Dvorak had two sources of credit,
a general source where credit could be obtained without pur-
chases and a specialized source where credit could be
obtained only if assets were purchased. His supply function,
like the others mentioned, did not take into consideration
such characteristics of the farmer as his ability to earn,
his age, education, etc. Although net worth.was considered,
‘Dworak assumed.that every farmer could borrow the same per-
centage of his net worth in terms of credit and at the same
interest rate.

Trant7, in his study of institutional credit for dairy
farmers considered the supply function of credit from.the
lender‘s viewpoint only. He considered only institutional
lenders and did.not take into consideration other sources of
credit such as land contracts with individuals, feed, machin-
ery, and livestock dealers, friends and families.

These and other studies have used somewhat inadequate
supply functions.for credit. The models used were unrealistic
in that the amount of money a farmer could borrow was not
related to his characteristics. This study is designed to

consider the personal characteristics, age, education, credit

 

7Trant, G. 1., Institutional Credit‘and The Efficiency of
Selected Dair Farms, Unpu lished Ph. . Thes s, gan State
University, 1960.

-5-
rating, etc., of the farmer and will attempt to estimate their
effects upon the quantity of credit.

In summary, there is a growing need for credit because
of increasing capital requirements in farming in Michigan.
‘With.this increase in use of credit in financing farming
operations comes a need to know more about the credit or
capital market Michigan farmers are facing. Further, the
increased number of research studies, linear programming and
otherwise, that are using supply functions for credit as a
variable input is the primary reason why an attempt should be
made to derive supply functions for credit for the principal
types of Michigan farmers.

Organization of Thesis

In the first part of Chapter I, the description and
location of the sample area will be discussed with details
on selection of survey areas, stratification of sample, and
how'the data were collected. In the last part of Chapter I,
the design of the questionnaire, the reliability of the data,
and details on processing the data to derive the marginal
factor cost functions will be presented.

Chapter II will present the various regression equations
used to derive marginal factor cost functions for the cash»
crop, dairy and beef cattle farms. The best fitting equations,
the basis for selecting them, and a detailed discussion of
their 'fit' and acceptability will be presented. In the last
section of Chapter II the derived marginal factor cost

-6-

functions of the three types of farms will be presented

mathematically and graphically.

The general organization of Chapter III will be similar

to Chapter II except that the marginal factor cost functions

for cashpcrep, dairy and beef cattle farms will be derived

frembthe combined data of all the farms.

Chapter IV includes the summary and conclusions derived

from the data of the thesis. The final section of the chap-

ter presents some possible implications of the results.

Objectives of the Study

The primary objectives of this study are:

1.

2.

3.

To derive and describe estimates of the marginal
factor cost functions for credit for cashpcrep,
dairy and beef cattle farmers in selected areas on
the basis of data collected frmm individual types
of farms in these areas.

To derive and describe estimates of the marginal
factor cost functions for credit for cashpcrep,
dairy and beef cattle farmers using the data from
all farms. This will be accomplished by using the
type of farm as an independent variable, thus the
general function can be used for cash-crop, dairy,
or beef cattle farms.

To compare (1) and (2).

The secondary objectives of this study are:

1.

To determine the factors which affect (a) the

-7-
quantity of credit the different types of farmers
estimate they can borrow and (b) the interest rates
they have to pay. I

2. To derive estimates of the quantitative effects of
each of these factors.

3. To determine if there are any differences in factors
and in their quantitative effects among the different
types of farmers.

Description and.Lecation of Sample.Aroas

The sample areas were Saginaw County for cashpcrop farm-

ers, St. Clair County for dairy farmers, and.Lenawee County

8 .These counties were selected

for beef cattle farmers.
because they are the leading counties in dollar volume of
sales within the state in their respective type of farming,
with the exception of St. Clair in dairying. Sanilac is the
number one dairy county in terms of dollar volume of dairy
products sold, but a survey was carried out with dairy
farmers there in 1959. Therefore, to avoid the possibility
of contacting the same farmers again this year, it was not
chosen. St. Clair County, the number two county in teams of

dairy products, was selected for dairy farmers in lieu of

Sanilac County.

 

8Beef cattle farmers are defined.in this study as farmers
that buy feeders (calves, heifers, or steers) and feed.them
out before selling them, and who have greater than 40 per cent
of their gross farm.income from sale of beef cattle.

-3-
Relevant Characteristics of Sgginaw County

Saginaw County is the fourth largest county in Michigan
in terms of tillable acresg; it is located in the mid-western
portion of the 'thumb' (see map figure 1). The soils of
Saginaw County were developed under poor natural drainage
conditions from loam, clay loam or silty clay loam.parent
material. The soils are relatively high in organic matter,
nitrogen and lime. They are moisture retentive, have good
natural fertility, and are durable under cultivation. The
principal soil series are Sims, Parkhill, and Kawkawlin.

About 65 per cent of the area of the county is occupied
by excellent agricultural soils. They were developed on
nearly level clayey plains where natural drainage was suffi-
ciently slow to permit a relatively rich accumulation of
organic matter and to prevent severe loss of nutrients by
leaching. Most of these soils require artificial drainage.
When tile drainage with adequate outlets is provided the
soils are very productive.

Saginaw County had about 60 per cent of its total farm '
income from.the sale of field crops in 1954-.10 The major
factors accounting for the type of farming in this area are
the level, genenally highly productive soils (when drained);

 

9Hill, Elton B., and Mawby, Russell 6., T es of.Farmp
ing in Michigan, Special Bulletin 206 (second egition)
September, 1954. .

10Michigan Statistical Abstract - Bureau of Business and
Economic Research MSU, Second edition 1958, p. 88.

,V
m
7'
,—
,.
»
u
" O < - ‘

 

-9-
the nearby good markets; the sugar beet processing plants;
the length of growing season which ranges from 130 to 160
days; and the moderated temperatures which favor dry field
bean production. Crop yields are well above the state
average.
Relevant Characteristics of St. Clair County

St. Clair’County is the eighth largest county in Michigan
in terms of tillable acres. It is located in the lower eastern
portion of the I'thumb" (see map figure 1). The soils of St.
Clair County were developed under very poor natural drainage
conditions from.loam, clay loam, or silty clay loam parent
materials. The soils are relatively high in organic matter,
nitrogen, and.lime, are moisture retentive, have good natural
fertility and are durable under cultivation. The principal
soil series are Brookston, Blount and Hoytville. The
Roscommon, AuGres, and Pests soil series occur in the north-
west corner of the county.

The topography is nearly level with some low depressions
and narrow sandy ridges. The principal problems in crop pro-
duction are poor drainage and maintenance of good soil struc-
ture. When tile drainage with adequate outlets is provided,
the soils are very productive because the surface is deep,
fine-textured and well supplied with humus.

Dairying is the most important enterprise for St. Clair
County with.greater than 40 per cent of its total farm.income

frfiJ!‘ 1

-10-

 

 

 

MICHIGAN

    
 

EEZE] Sample Areas

Saginaw - Cash—crop farmers
St. Clair - Dairy farmers
Lenawee - Beef cattle farmers

 

 

 

 

 

 

Figure 1. Outline map of Michigan showing location of sample
areas.

-11..

11 It is close

from the sale of dairy products in 1954.
enough to the large nearby markets to favor dairy production
and general farming yet not close enough to have a large
percentage of part-time farmers.

Relevant Characteristics of Lenawee Count!

Lenawee County is the third largest county in Michigan
in terms of tillable acres. It is located in the southeast-
ern part of the state (see map figure 1). The soils of
Lenawee County were formed mainly from clay loam, silty clay
loam, silty clay or clay parent materials. The drainage of
these soils ranges from.moderately well to imperfectly-
drained with the latter conditions generally associated with
the more level soils. The principal soil series are St.
Clair, Nappanee, Morley and Blount with some Fox, Oshtemo,
Waueson, and Berrien.

The topography is level to rolling and is generally
favorable for farming operations. The soils are deep, high
in fertility, and durable under cultivation except on the
steeper slopes. The tightness of the clay which reduces the
rate of water movement through the soil and.maintenance of
good soil structure on the surface are problems in the use
of this land for cropping purposes.

The most important source of farm income is from the

sale of livestock, mostly cattle, hogs, and.sheep. The major

 

llIbid.

~12-

factors influencing the selection of farm enterprises in this
area are the generally productive soils, the relatively long
growing season (150 to 170 days), and the good local and
nearby markets.
Surve: areas

Survey areas were selected within each sample county.
Five townships were randomly selected (using a table of ran-
dom numbers) within each sample county. Two sections were
randomly selected from each.township, the first section
selected became a survey area and the second section an
alternate area to be used if the desired number of qualified
farmers was not located in the first section. This made a
total of five survey areas and five alternate survey areas
for each county selected for a particular type of farming
area.
Stratification of sample

.A stratified sample was randomly selected within each
survey area. The sample for the cashpcrop type of farming
was stratified by size of farm measured in acres, with three
divisions, 0-80 acres, 81-160 acres and 161 acres and over.
in attempt was made to sub-stratify by size of gross income,
under $7,500 and $7,500 and over, to minimize the inter-
correlation of these two factors.

There were six records collected.from each survey area
or its alternate survey area, two for each size of farm

division, with an attempt to get one each of these in the two

,0

 

 

-13-
divisions of gross incmme. The latter was not possible in
every instance but was accomplished in most survey areas.
‘Uith.five survey areas this made a total of 30 records for
the cashpcrep farmers.

The survey of the other types of farming areas was
accomplished in much.the same procedure. The dairy farm
sample was stratified by size of herd, with.breakdewns of
under 29 cows, 20-39 cows, and over 30 cows. Sub-stratifi-
cation by gross income was carried.eut as possible. The
beef-cattle sample was stratified by number of feeders the
farmer purchased and fed out each year as well as by gross
income. There was a total of 30 records collected from.both
dairy farmers and beef cattle farmers. Thus a grand.total
of 99 schedules was completed. Two schedules, one cashpcrep
farm and one beef cattle farm, were later discarded because
of incomplete data.

Field.Techniques

The confidential nature of some of the information
required in the study created certain interviewing problems.
The time of the year the survey was taken, a very busy
season, created additional problems. To overcome these
difficulties each qualified farmer contacted was given the
opportunity to designate the most convenient time for the
interview. It was explained and ro-emphasized that this
information would be used in a strictly confidential manner.

Approximately 90 per cent of the qualified farmers who were

-14-
contacted gave interviews.

Schedules were obtained in each survey area by starting
in a randomly selected corner of the section and working in
a counter clock-wise direction. .A schedule was taken from
each qualified farmer in succession until the required.nump
her in each group was obtained. ‘Unqualified farmers were
omitted and the next farmer contacted, “Farmers were con-
tacted until six schedules were obtained from each sample
area or until all farmers in that area had been contacted.
This same procedure was used in working the alternate areas
to obtain records for strata not filled in the regular survey
area.

A.farm qualified as (l) a cashpcrop type of farm if
greater than 40 per cent of total farm.income was derived
from.sale of crops as cash crops rather than through live-
stock, (2) a dairy farm.if greater than 40 per cent of total
farm income was derived from the dairy enterprise, and (3) a
beef cattle type farm if greater than 40 per cent of total
farm.income was derived from the sale of cattle. The census
definition of a farm was used.to determine the smallest farm
that could qualify for the sample.

If a farm qualified by census definition and by type,

a schedule was completed provided the farmer would cooperate.
Every effort was made to complete a schedule if the farm
qualified.

A farm was considered within the sample area if the farm

 

 

 

 

 

-‘ f .II..||

-l5-
house was within the sample area regardless of where the
majority of the farm land was located. The farm was not
considered within the sample area, regardless of the per
cent of the farm land in the sample area if the farm dwell-
ing was not in the sample area.
Design of the Questionnaire12

The questionnaire was designed with a two-fold purpose
in mind. The first purpose was to collect data on independent
variables considered to affect the dependent variable, quan-
tity of credit. It was decided, based on knowledge gained
from 19 years on the farm and six years of intensive study
of agricultural economics along with consulting with members
of the Department of Agricultural Economics, that the follow-
ing independent variables should be considered in this study.

1. Interest rate

2. Not worth

3. Farmer‘s credit rating

4. Gross farm income

5. Net farm income

6. Size of farm

7. Age

8. Education

9. lFarming experience

10. Off-farm income

 

13A complete copy of the schedule is included in the
Appendix.

-16-
11. Availability of land contracts
The first part of the schedule was designed to produce direct
and indirect information.about these variables.

The second part of the questionnaire was designed to
produce estimates of the amount of credit individual farmers
could borrow under existing conditions. .A general question,
asking the farmer to estimate the maximum amount of money he
could borrow from all possible sources, was asked first.
Most of the credit agencies were listed to aid the farmer in
his recall of these agencies. These were individual loans
and were not cumulative. Each one was based on the present
equity position of the farmer and was used primarily to get
the respondent thinking in terms of the sources of credit
available to him. The next question was, 'consider for a
few minutei that you are going to borrow‘gll the money that
you can possibly get. Now tell me the details of these
loans." The amount of the loan and interest rate, starting
with.the source of lowest interest rate first, were recorded
until the farmer indicated.that he could not borrow any more
money regardless of the interest rate.

The questionnaire, when completed, had data for all the
independent variables, except the farmer's credit rating.
The credit bureaus of the respective counties were contacted
to get these ratings. They had agreed in advance to supply
this information. The credit ratings were grouped into four

general categories, poor, fair, good and excellent and.these

-17-
later were quantified to l, 2, 3 and 4 respectively to
facilitate function fitting. There was one question on the
questionnaire relating to this variable. Respondents were
asked to name three credit or business references to be used
in case the credit bureau did not have a record or enough
information on any of the farmers to give a general credit
rating.
Reliability of the Data

The estimates on quantity of credit secured.from the
farmers were considered by the author to be reliable estimates
for several reasons.

1. Some of the farmers, especially the beef cattle
farmers, had established "credit limits” at their
banks. The banker had taken not worth statements
from these farmers and had told them.how much their
credit limit would be in advance so the farmer would
not have to fill out forms, etc., to find out how
much credit he could.get when he wanted to make a
loan. These credit limits were not verified. The
quantity was recorded and used as reported by the
farmer.

2. Data were collected from.the individual farmers on
outstanding real estate and chattel loans as of
December 31, 1959. This information, as supplied
by the farmer, was checked against the information

the various credit bureaus had and generally the

 

3.

4.

5.

6.

-18-
quantities agreed. This indicated.that the farmers
were in a cooperative mood and were attempting to
give true answers to factual questions.
The land contracts included in the estimates were
contracts the farmers had'been offered or had been
discussed by the farmers with the potential seller
of the additional land.
Most of the farmers had a fair estimate on how much
the Federal Land Bank would loan them because of
loans made by farmers within the neighborhood or
because they had loans outstanding with the Federal
Land Bank at present. They were not sure in some
instances what the present interest rate was on
Federal Land Bank loans.
The additional quantities of credit above loans for
1and.such as for purchase of machinery, equipment,
livestock, etc., were fairly good estimates because
a very high percentage of all farmers had at one
time or another used.this type of credit. They.
seemed to know the source of this type of credit
and to have a good idea of how much of these types

of purchases they could procure on credit.

One of the weakest estimates was the amount of open

account credit or installment buying they could do
considering all the other loans. Usually this

quantity was small.

-19-
7. The data used to derive gross farm income and net
farm income were checked, in almost all interviews,
against the farmer's income tax returns for 1959.

The quantities of credit used to derive the marginal
factor cost functions were based on accumulative combina-
tions of the above types of loans. This is the weakest
point of the estimates because it is difficult for anyone
to estimate the quantity of credit he can get after two or
three loans have been made. Also, in some instances the
farmers were not aware of credit sources that will extend
credit at high interest rates after other loans have been
made. The interest rates payable for certain types of loans
were not well known by some farmers. The smaller farmers
with less credit experience than the larger farmers did.not
appear as well informed about the various credit opportuni-
ties as tho larger farmers.

This study was carried.out with farmers rather than
lenders of farm.credit for various reasons. The author
wanted to get an estimate of what the farmers thought they
could get in quantities of credit and the interest rates
they would have to pay. The relationship between the charac-
teristics of the farmer and the quantity of credit that he
can get on the capital market can be determined.better from
farmers than from.lenders' information. The typical response
of lenders when given a hypothetical farm and asked how much

they will lend is that they do not know without knowing

-20-
something about the man. If this study had been carried out
with lenders of farm credit, the many individuals who make
up the largest source of agricultural credit outside the
Federal lending agencies would have been omitted. Thus
biases are likely to arise from the lender‘s viewpoint as
well as in a study with.the present orientation.
Processing the Data

As pointed out above, the questionnaire was designed to
get direct information on most of the independent variables
considered in this study. The data on these variables, not
worth, size of farm, age, farming experience and off-farm
income were used in the form collected from.the farmers. The
value of two variables, gross farm income and net farm.income,
had to be computed from basic data on cash farm receipts and
cash expenses as supplied.by the farmers. (Almost all of
these figures were checked against the income tax form for
1959 as reported by the particular farmer.) The farm.was
charged with depreciation on all machinery and equipment,
inventory changes (plus or minus) on equipment and livestock,
and family labor to derive net farm income for the farm.13

The data for quantities of credit were collected in the
form of a series of loans that the farmer estimated would
give him.the maximum amount of credit. For each separate
loan of the series the interest rate was estimated by the

 

13See schedule in Appendix.

-21-

farmer. In all instances after a farmer had related the
details of these loans, he was asked, "Is there another
combination of loans that would give you more than this
amount of money if you were willing to pay higher interest
rates?“ The combination of loans that gave the farmer the
maximum.ameunt of credit was used in deriving the marginal
factor cost function of credit.

The loans, as pointed out repeatedly to the farmer
during the interview, were on a cumulative basis. Each addi-
tional loan of each combination was made with the provisiontmet
all prior loans had been made. One separate series of loans
or one combination was used in deriving the marginal factor
cost function. This was the combination that gave the
farmer (as estimated by him) the maximum amount of credit.
Cross combinations, that is using two or three sources from
one combination of loans and then shifting to another com-
bination,were not used. If cross combinations had been used
the estimates of interest rate would probably have been some-
what smaller. This is because the farmer could possibly have
obtained a smaller quantity of credit, if this was all he
desired to obtain, at a lower interest rate.

To make these data comparable for all combination loans
secured, the weighted average interest rate was computed for
the combination of loans for each farmer. This combination
loan for most farmers consisted of five separate loans with

applicable interest rates. This gave all farms approxhaately

-22—
five observations or values for interest rate and quantity of
credit. In processing the data or using the data to fit a
regression equation, each one of these values was used as a
separate observation, with the fixed values of the other
variables, not worth, credit rating, gross farm income, etc.,
repeated for each value of interest rate and quantity of
credit for that particular farm. Therefore, the 30 farms of
each group have approximately 150 observations or an N of
150 when fitting the regression equations.

All of these observations were not considered as inde-
pendent observations when testing for significance. Stu-
dent‘s 't' test was used for testing the level of signifi-
cance. “This is the ratio of the estimated regression
coefficients and their respective standard errors. When
using this 't' test the number of farms was considered as N
which gave approximately 25 degrees of freedom for each test
of significance. Therefore the statistical significance
tests are biased somewhat toward the low side.

Computing the weighted average interest rate converted
the cost of credit into average factor cost units. Therefore,
after the best fitting regression equation had been selected
with quantity of credit as the dependent variable and then
recalculated with interest rate as the dependent variable,
this interest rate was in terms of AFC (average factor cost).
To derive the marginal factor cost function of credit the

following procedure was used.

~23-

The equation for interest rate was: *

d
Interest rate (Y) - a+b1Q(quantity of credit)+:£_bIXI=AFC
. . =2

To convert to total cost multiply both sides by Q (quantity of
credit) because interest rate (Y) multiplied by quantity of
credit (0) - total cost M

d
2
or - aQ + blq + 521.11%

to convert to MPG (marginal factor cost) take derivative with

respect to Q

we as 9 Y a 00 (‘0 + biaz + D1110)
j; QL—— . ‘90 j-r-

 

e d
. i=2 1 1

By varying the quantity of credit (0) and using the
values of the his as computed from the regression equation
with interest rate as the dependent variable, and using the
mean value of the other variable in the equation, the supply

schedules were computed.

 

*dsnumber of independent variables.

CHAPTER II

ESTIMATING MARGINAL FACTOR COST FUNCTIONS FOR CREDIT
F6R.CASH-CRGP, DAIRY, AND BEEF CATTLE FARMERS

In this chapter, a marginal factor cost function of
credit will be derived for cash-crop, dairy and beef cattle
farmers based.on the data collected from.these respective
types of Michigan farmers. In the first section of this
chapter, following this introduction, the three types of
farms will be described and compared by size of farm.for
most of the variables considered in this study. The second
section of this chapter will present the regression equations
that were fitted. The selection of the best fitting equa-
tions for each type of farm will also be discussed in that
section. The third section.will present the derived mathe-
matical marginal factor cost functions with.graphs of these
functions.

General Description of the CashpCroplFarms

The records obtained from the sample of cash-crop farm
operators were sorted into three groups according to the size
of farm, in acres, and comparisons were made between small,
medium, and large farms, Table l.

The small farms (average size -- 74.4 acres), were
operated by older men than the medium size farms (average size
-- 122.6 acres), or the large farms (average size -- 200.6

acres). The average age for the operators of the small farms

 

 

~25-
was 46.9 years compared to 43.3 years for the operators of
medium farms and 41.2 years for large farm.eperators. This
was not caused by the operators of small farms‘beginning to
farmblate in life. The average years of farming experience
was 18.4 for the operators of small farms compared to 19.1
and 17.0 years for*the operators of medium and large farms,

respectively.

TABLE 1
AVERAGE rout ACREAGE, mans 0F FARMING , AGE AND EDUCATION,

BY SIZE OF cranium, 29 CASE CROP FARMS,
SAGINAw COUNTY, MICHIGAN, 1960

_v__ 'JA‘ ' 1;

Size —Nfimber ' ‘ Avera e H
(Total of SIze Years of :3. Education

 

 

 

_A_cres) Farms (Acre s) Farming_ (Years) (GradesL
0-80 9 7e.4 18.4 46.9 ' " 8.3 ’
81-160 10 122.6 19.1 43.3 9.6
160 and over ___1o 200.6 41.0 41,2 10.1
Ayorage _‘___ _-. _-_ 134.5 --.18.2-. -_, 43.7 , _ 9.4

 

There was a negative correlation between age and educa-
tion of the farm.eperators, the older farmers having less
formal education than the younger farmers. Thus, the
younger, more educated men were farming the larger farms.

. The operators of the small farms had a higher average eff-
farmvincomo than the operators of the medium.or large size
farms. In fact, six out of nine of the small farm.eperators

had a regular 8-hour a day job in town which they held down

lilll Ill. '1.

 

 

-26-
in addition to their farming operations. Their farm work was
done in the afternoons, after work, at night and weekends.
There was a high negative correlation between off-farm

income and size of farm for the cash-crop farmers, Table 2.

TABLEZ

AVERAGE min-FARM INCOME, Gnoss FARM mom, NET FARM INCOME
AND TOTAL INCOME, BY SIZE as CPERATIONS,
29 CASE CRcr FARMS, SAGINAw COUNTY,
IMICHIGAN, 1960

 

 

 

 

 

 

Size Gross Net
(Total , . arrests--‘- -Farn- , . Farm . Tota1*
Acres) Income In come Income Income
0-80 2,762 4,330 814 3,627
81-160 f 1,549 7,541 2,440 3,989
161 and over '965 ‘ 'g_;5,ooe " ' 4,5;;_ 5,482
Average _ . _ ,l,724<- . 9,119 ,. . 2,649 A4,392

 

*Tetal income includes off-farm income and net farm
incomeplus other income which included wife's salary,
dividends on stock, interest on bonds, etc. -

The gross and net farm incomes increased as the size of
farms increased, with the small farms having an average of
$4,330 gross and $814 not compared to $15,006 gross and
$4,511 not for the larger farms. When total income was
considered, difference in size of farm was not measly so
great because the smaller fans had enough off-farm income

to eff-set their low farm income. Thus, they compared
favorably with the larger farms in total income. The

’1‘

-~. - o.— _ -
a 0- 0
~ - - o.- c

-27-
smallest (0-80 acres) had an average of $3,627 compared to
$3,989 for the medium (80-160 acres) and $5,482 for the larger
farms (160 acres and over).

The average credit rating was lowest for the group of
small farmers and highest for the large farmers. The credit
rating was established by contacting the Saginaw Credit
Bureau, which gave general credit ratings in four general
classifications, poor, fair, good and excellent, based on
their records of the farmer's credit history. These general
credit ratings were converted into numerical values by assign-
ing the following valueslz poor = 1, fair - 2, good a 3 and
excellent - 4.

TABLE 3
AVERAGE CREDIT RATING, NET WORTH AND QUANTITY 0F CREDIT,

BY SIZE OF OPERATION, 29 CASH CROP FARMS,
SAGINAw COUNTY, MICHIGAN, 1960

 

 

 

 

 

Average
Size Quantity
(Total Credit Net Interest of
Acres) Rating Worth Rgte Credit
.Dol. .Pct. 1.
0-80 2.9 30,952 6.5 33,377
81-160 3.5 47,529 ' 6.4 55,400
161 and over 3.7 58,400 6.2 71,750
Average 3.4 46,133 6.4 54,203

 

 

1The residuals forecredit rating, when plotted about the
straight regression line substantiated this assumption of
linearity with the 'units' in which this variable is measured.

-27-
smallest (0-80 acres) had an average of $3,627 compared to
$3,989 for the medium.(80-160 acres) and $5,482 for the larger
farms (160 acres and over).

The average credit rating was lowest for the group of
small farmers and highest for the large farmers. The credit
rating was established by contacting the Saginaw Credit
Bureau, which gave general credit ratings in four general
classifications, poor, fair, good and excellent, based on
their records of the farmer's credit history. These general
credit ratings were converted into numerical values by assign-
ing the following valuesl: poor = 1, fair - 2, good = 3 and
excellent = 4.

TABLE 3
AvERAGE CREDIT RATING, NET WORTH AND QUANTITY CF CREDIT,

BY SIZE OF OPERATION, 29 CASH CRCP FARMS,
SAGINAw COUNTY, MICHIGAN, 1960

 

 

 

 

 

.Average the

Size Quantity
(Total Credit Net Interest of
Acres) Rating worth Rgte Credit
D01 o P01} o 551 e
0-80 2.9 30,952 6.5 33,377
81-160 3.5 47,529 ‘ 6.4 55,400
161 and over 3.7 58,400 6.2 71,750
Average 3.4 46,133 6.4 54,203

 

 

1The residuals forecredit rating, when plotted about the
straight regression line substantiated this assumption of
linearity with the 'units' in which this variable is measured.

a

'\

-23-

The average credit rating was lowest for the small
farmers and highest for the large farmers. This indicated
that the operators of the larger farms, probably through.more
extensive use of credit, had built up a better credit rating
than the operators of smaller farms. Credit rating was also
correlated with the not worth of farmers. The group of large
farmers had an average net worth of $58,400 and an average
credit rating of 3.7, compared with the smaller farmers with
an average net worth of $30,952 and an average credit rating
of 2.9.

Not worth and credit rating were positively correlated
with quantity of credit and.negatively correlated with interest
rate. The group of small farmers estimated they would have
to pay the highest average interest rate for the smallest
quantity of credit, compared to the estimates of the medium
and large size group of farmers.

General Description of the Dairy Farms

The records obtained from the sample of dairy farm.epera-
tors were sorted into three groups according to the size of
farms, based on average number of cows during the year and
comparisons were made between small, medium and large dairy
farms. These classifications were made to determine by
inspection if there were serious correlations among the inde-
pendent variables under consideration.

The small size dairy farm (0-20 cows) had an average of
15.4 cows, the medium.size (21-30 cows) had an average of 25.6

-29-
cows and the large size dairy farms (31 cows and ever) had an
average number of 36.6 cows, Table 4. Size of farm, in terms
of acres operated, was closely related to size of farm in
terms of cows milked, the small size group (0-20 cows) had an
average size' of 127 acres compared to an average of 308 acres

for the large size group (31 cows anll over).

TABLE4

AVERAGE TOTAL ACBEAGE, YEABS OF FARMING, AGE AND EDUCATION,
.BY SIZE OF OPERATION, 30 DAIRY FABIB,
ST. CLAIR COUNTY, MCHIGAN,1960

 

 

Tize Number Size

(Number of to a k Years Ago Education
of cows Farms acres cows F Years Grades
0-20 10 127 ' 15.4" 24.2 52.5 '6 8.8 '
21-30 10 206 25 .6 1'7 .5 41 .1 10 .4
31 and oval-1.0 308 " 36.6 ‘ 18.4 ' 45.8 9.0
Average 30 214 . 25 .9 - 20 .0 t 46 .5 9 .4

 

There was not much correlation between farming experience
(measured in terms of years of farming) and size of farm. The
farmers en the smallest farms had the most experience, an
average of 24.2 years. These on the medium size farms had
the least experience, an average of 17.5 years. The age of
the dairy farm operators was closely correlated with their
experience, the oldest farmers having the most experience and
the youngest farmers the least.

Education of farm operators was not closely related to

-30-
size of farming operations or t. farming experience. It was
inversely correlated with the age of the farm operators.

The oldest group of farmers had an average education of 8.8
grades compared to an average of 9.0 grades for medium age
group and an average of 10.4 grades for the youngest group.

There was a wide difference in the amount of off-farm
income by size of dairy farms, with the small size group
(0-20 cows) having an average of $1,272, while no operators
in the medium group and only one, in the large size group had
any off-farm income, Table 5.} This seemed to indicate that
with 21 cows or more, the farming operation was a full time
job which required all the time of the operator. Another
factor was that dairying, unlike cash-crop farming or other
type farming, is a 12 months job. '

TABEE 5
AVERAGE (FF-FARM INCOME, GROSS FARM INCOME, NET FARM INCOME
AND TOTAL INCOME, BY SIZE OF OPERATIONS, 30‘DAIRI’FARMS,
. . . 0 ST. CLAIR CTINTY,‘MICHIGAN, 1960, .

_—L— ‘ 4

11m Avera e orTirm
(Number tiff-farm - Gross farm Net farm TotaI

 

 

 

 

 

 

of cows) Income Income Income Income
0-20 1,272 6,862 1,938 3,333
21-30 0 10,716 2,811 2,838
31 and over __1.90 ' h ' .17 .' 666 A A \ (5,670 6, 373

 

Average. . 488 11,748 - A 3,473 4,181

 

 

 

0--

.,--

_ K.

o-

ur-o -0

40‘

o a

-31-
Gross farm.income and net farm income were positively
' correlated with size of farming operations (in terms of cows
milked), The small number group had $6,862 gross and $1,938
not compared to the largest number group having $17,666 gross
and $5,670 net farm income.

Total income, which included off-farm income, not farm
income and other income, suCh as interest payments, dividends,
wife's salary, etc., was somewhat correlated with size of
farming operations or number of cows milked. But this corre-
lation was not high because the small size group had enough
off-farm income to bring their average total income above the
medium size group, whiCh did not have any operators with any
outside income. The net farm.income of the large size group
was high enough to more than off-set the outside income of
the small dairy farmers. Their total income averaged $6,373
compared to only $2,838 for the medium size group and $3,333
for the small size group.

The average credit rating for the operators of dairy
farms was not highly correlated with size of operations or
not worth (see Table 6). There was an increase in the average
credit rating from.3.4 to 3.8 from the small to the medium
size group. But there was a decrease from.3.8 to 3.4 from
the medium size group to the large size group. The average
net worth increased from $57,236 to $86,950 from.the medium
to the large size group.

Not worth was correlated with the size of operation as

l.

«E

 

-32-
it increased about 75'per cent from the small size group to
medium size group and again from.medium size to large size
group. Interest rate was not closely related to the net
worth of the operators or to the quantity of credit. The
average interest rate decreased from the small size group to
the medium.size group and then increased from.the medium to
the large size group. Although the average net worth and
average quantity of credit increased from the small to the
medium.te the large size group. There was a negative corre-
lation between the average credit rating and average interest
rate, with the farmers having the highest credit rating (the
medium size group) having to pay the lowest average interest

 

 

 

 

rate.
TABLE 6
AVERAGE CREDIT RATING, NET WORTH, INTEREST RATE
ammo QUANTITT.0E CREDIT, BY_SIZE 0E 0PERATI0Ns,
30 HAIR! DARMS, ST. CLAIR COUNTY,
MICHIGAN, 1960
‘315. A
(Number Credit Net Interest Quantity
of cows) [Rating Worth Rate of Credit
D0]. e Pat e 01 e
0 - 20 3.4 31,511 6.2 30,390
21 - 30 3.8 57,230 6.0 49,245
31 and over 3.4 86,950 6.1 80,585
Average 3.5 58,566 6.1 53,374

 

The quantity of credit farm operators estimated they could

-33-
borrow on the capital market, from friends, and from other
sources was very highly correlated with their net worth.
General Description of the Beef Cattle Farms
The records obtained from the sample of beef cattle
farmers were sorted into three groups according to the size
of operation, based on number of feeder cattle. Comparisons
were made among small, medium and large farms. The three
groups were, 0-50 feeders, 51-150 feeders and 151 feeders and

over, Table 7.

TABLE 7

AvERAGE T0TAL.ACREAGE, NUMBER OF‘FEEDERS, YEARS OF
FARMING, AGE AND EDUCATION, BY SIZE 0E
OPERATION, 29 BEEF CATTLE FARMS,

LENAwEE COUNTY, MICHIGAN, 1960

 

 

 

 

 

gize
(Number Number Size _ Average
of of Totai N0. of Years Ago Education

Feeders) Farms .Acres ‘Feeders iFarming (Years) jGrade)_
0-50 10 175.9 34.2 19.6 42.7 11.0 ’
51-150 9 244.2 100.8 17.6 42.0 11.0
151 & over 10 223.7 196.8 24.4 53.1 9.7
Average 213.5 110.9 20.6 46.1 10.6

 

Size of farm (in total acres) is normally closely related
to size of operations based on number of feeders. In this
sample, however, the average size of farm for the medium size
group (51-150 feeders) was larger than the average size of

farm for the large size group (151 feeders and over). There

-34-
were two unusually large farms in terms of acreage in the
medium.oize group. Thus, the data do not show much correla-
tion between these factors. Size of farm operations in
terms of number of feeders was only slightly correlated with
farming experience. The operators of the largest farms had
the most experience but the operators of the medium.size
farms had.the least.

Age, as usual, was closely related to farming experience.
It seems that most farm operators start farming at about the
same age regardless of what size farm they operate. Age
and education were negatively correlated, with the youngest
group of farmers (average age -- 42.0) having completed 11.0
grades of school as compared to the oldest group of farmers
(average age -- 53.1 years) having completed only 9.7 grades
of school.

Average off-farm.income was positively correlated with
size of farm measured in terms of number of feeder cattle,
Table 8, which was quite unusual, because normally these two
factors are negatively correlated. .As to number of farmers
having off-farm income, there were four out of ten for the
small size group, compared to only one out of nine for the
medium size group and two out of ten for the large size group.
Thus, it was the relative size of non-farm income earned by
the medium and large size groups that caused.this unusual
correlation, rather than the percentage of the farmers work-

ing off the farm for each group.

-35-
Gross farm income, as in most instances, was highly

correlated with size of farm, but not farm.income was not
highly correlated with this factor. Net farm income actually
declined from.an average of $3,769 for the small size group
to an average of $1,836 for the medium size group, with the
large size group having an average of $10,050. Total income
was similar to net farm income as the smaller farms did.not
have enough off-farm income to offset the higher net farm

income of the larger farms.

TABEE 8

AVERAGE OER-FARM INCOME, GROSS FARM INCOME, NET FARM INCOME
AND TOTALLINCOME, BY SIZE OF OPERATIoN, 29 BEEF CATTLE FARMS,
_ LENAWEE COONTT,.MICHIGAN, 1960

 

 

Average Per Farm

 

 

 

 

 

 

-§ize A

(Number of TIT—ram Afiross Farm _ . .Net Farm “Total“
Feeders) Inggme Income Income Income

.. D01 e , DOi e 50]. e‘ E]. e

0 - 50 493 13,835 3,769 4,032
51 - 150 533 31,577 1,836 2,880
151 and over 800" " 266,536; 10,050 11,295
Average . 611 ....... 37,513-. A 5,335 6,179

 

The average credit rating for the operators of different
size groups was positively correlated with size of farm, net
worth and quantity of credit, Table 9. The average credit
rating increased from 3.5 for the small size group, to a

perfect 4.0 for the largest size group. Every farmer in the

-36-
size group, 151 feeders and over, had an I'excellent" credit
rating or the very highest rating assigned by the various
credit bureau exchanges. Not worth was positively correlated
with size of farm. The smaller farms had an average net
worth of only $34,170, compared to $85,204 for the medium
size and $177,416 for the large size group. Although
farmers in the large size group had not been farming many
more years than those in the smaller size group, their opera-

tors had greater net worths.

TABLE 9

AVERAGE CREDIT RATING, NET WORTH, INTEREST RATE.AND
QUANTITY OF CREDIT,.BT SIZE OF OPERATION,
29 BEEF CATTLE'FARMS, LENAWEE COUNTY,
MICHIGAN, 1960

 

 

 

 

Average

Size _iuant1$y
(Number of Credit . Net Interest of
Feeders) Rating Worth. Eggs Credit

0 - 50 3.5 34,170 6.2 53,910
51 - 150 3.9 85,204 6.1 126,978
151 and over 4.0 117,416 6.0 156,520
Average 3.8 78,713 6.1 111,980

 

Not worth and credit rating were positively correlated
with quantity of credit and negatively correlated with interest
rate. The group of small size farmers estimated they would
have to pay the highest average interest rate of 6.1 per cent,

‘ 1
~ K
.. - -. -.. . -
' v.
.. . ‘
A..- ’7)-.- - ‘ . A
' K

 

 

 

-37-
while being able to borrow only $53,910, compared to 6.1
per cent for $126,978 and 6.0 per cent for $156,520, for the
medium and large size groups, respectively. The farmers with
the lowest credit rating estimated they would have to pay the
highest interest rates for credit and that a smaller quantity
of credit would be available to them on the capital market.

Simple Correlations Among the Independent Variables

To get a more accurate measure of the simple correla-
tions which exist among the various independent variables than
presented in the above description, it was necessary to compute
the simple correlations of these variables. Standard errors
of the regression coefficients are positive functions of the
intercorrelations of the independent variables. Inspection
of the simple correlations among the various independent varia-
bles aids in selecting the variables to be included in the
revised equations. These were computed between each pair of
variables used in the various equations for the different
types of farms and appear in Tables 10, 11 and 12.

As it can be readily seen, the simple correlations among
some of the independent variables were high enough to affect
the estimated coefficients for these variables tending to
cause them to have compensating errors. An effort was made
not to use both of the variables where high intercorrelation
existed. In a few instances, pairs of variables were used
which had fairly high simple correlations. This does not
bias the regression coefficient estimates but it does tend

.ooosaaoo be: ones neansfiae> cacao sou nsofiaoaoaaoo oaaafia one I 02*

 

 

 

 
  

co.H vooaasoo dean
oz oc.H sofiasoscm

oz oz co.” ewe

02 mm. oz co.H Show mo oswm

oz 4...”. 02 an. co. H 0200:.“ Ed.“ aoz

wow on . aw. . mm . I or. am . co. H ofioofi flash on cue
ea. on. .62 on. we. on. oe.H mesons assess
«in. $6. AH. I mm. mm. «b. no. cc. H £6.33 aoz
om.| ca.| Ha. HH.| ba.1 oa.l mm.l n«.| cc.H ease oaeaeasH
ooeaasoo sofioso 1mmd mush oaoosa oaoosfil1mmmosa mono: cosh eapsfias>
Honda” snow no anew Show afiueao.. .voz. vaeaoasH soozoaeozu

osfim aoz macaw oHDsHh
azouzoaoumwl

Ocad gin-HMO; «ML—.7560 awn—54m dam gan. undo «N
mHAMdHM§> HZNEZQHHQZH Hfl9.02924 MZGHBdAHmmaD mHhZHn

0H mflm<a

 

Oo.H

wosaazoo 6min

 

  

 

 

mu. co.z sozooozofl
eo.u mo.u oo.z coo
oz zm. oz oo.z atom on ouzm
oz we. oz no. oo.z .aoozz some zoz
w. oz. on. oz ze. 5.. 8. z 385- 56 Bozo
. oz. oz. oo.a oo.: oo. oz. oo.z oezzes ozoeeo
oz. oo. on. «a. no. zs. oo.u oo.z zone: ooz
zo.u oz.u no. z~.u ee.n oz.u mo.u ,oo.n oo.z moat oaoeezez
aosaozoo aofiaoo Imwd Bush oaoosfi oaoosz mmfivsa gazes ouch sansfias>

easel uses o. ateo auoo ozooeo wooz ouoeoozz

euzm ooz aaoao

 

 

3oz .zqozzozz .zazooozzozo .3 £2qu EH4: on
noumozzzz azzozomzezz zoo ozozo.ozozoozomzoo ommzzo

AH mundfi

lll|l“|’ O

 

‘-I.IID ‘III’ I'll! 1|-‘1‘6 'Vll‘l'llfl.‘

NH “NAM-<9

 

 

cl. el. e .10 all e] eta el. 0
a o o .e e s all e
a o .11 ol. o e s
e e o e e
e a e
e e
.11 eul e
e e
e
‘xl‘ll II- I6. I‘IO‘.I t UIIAD-‘ t |I00ltUIiO 90.00.04 Vi... t l I .-

1|

-40-

 

  

  

oo.z aooaazoo coon
oz oo.z :ofiaoosom
oz om.| oo.z owd
oz oz oz oo.z sash mo oszm
oo. oz oz an. oo.z oaoosd Baum aoz
oz bo.| Hm. oz oz oo.z oaoosfi Bash macaw
oz. oz oz zz. we. ow. oo.z muzzee ozooeo
an. o«.1 ow. oz mm. Ho. no. oo.H Sane? eoz
wn.t No. no. MH.I mm.| HH.I Ho.l no. oo.d coda amohoan
oosaazoo :ozaeo «M4 sumo oaoosfi oaoosz Amadeus noses owes canozzo>
osdfl loom oo Shoo Show aHooao ooz auoaoozH soosozooaH
ouzm aoz naoao oznszas

oooz .zoozoozz .zozooo zmzazzz .mzmez-mzozdo zoom om

mmflfldHMN> EZHQZHNHQZH fins $2924 mchH<HHmmco HHmZHm

NH mnmdfi

s ' n o--

. hr—‘m . -
4
.

A“---

to ca
be Is
variz
regre

bles

varia

in th

 

 

 

-41-
to cause the estimated coefficients of these variables to
be less reliable. High correlations between two independent
variables increase the standard error of the estimated
regression coefficients and thus tend to cause these varia-
bles to be less statistically significant.
Regression Equations Fitted
As indicated in the introduction, the independent
variables to be used in deriving a supply function of credit
in this study were:
1. interest rate
2. net worth
3. farmer's credit rating
4. gross farm income
5. net farm income
6. size of farm
7. age
8. education
9. availability of land contracts
10. off-farm income
11. farming experience
First‘git

 

After a preliminary examination of the data and plotting
of scatter diagrams for each of the independent variables
against quantity of credit, two variables, farming experience
and off-farm income, were eliminated from consideration in

the equations for the first fit. The first set of equations

-42-

consisted of three equations to be used for all three types

of farms and for all farms as a group. These equations were

simple linear regressions and were fitted by the least squares

method. The three equations were:

(1) quantity of credit (Y) - flinterest rate (X1), net worth (X2),
credit rating (X3), gross farm income (14)] '
with the following results:
cash-crop farms

Y'= -13.357o16 + 1,743.44x1 + .11x2 + 7,945.81X + 2.05x
(1,162.20) (.08) (1,859.00) as (.29) 4*

if = .76, 112 == .58, standard error of estimate $14,249

dairy farms

Y a -10,386.92 + 4,793.3611 +'.72X2 - 3,497.23!3 - .03314
(2,430.97)* (.04)*** (1,699.08)* (.30)

fi‘a .89, fi? = .79, standard error of estimate $14,475

beef farms

Y - -77,251.22 + 9,372.8511 + .86x + 12,463.54X3 + .20x4

2
(3,906.65)* (.06)*** (5,186.07)* (.11)

fifa .87,‘R‘2 - .76, standard error of estimate $27,903

(2) quantity of credit (Y) - flinterest rate (X1), credit
rating (X3), net farm income (X5), size of farm (X6147

 

* - Significantly different from zero at the five per
cent probability level.

** - Significant at the one per cent probability level.
*** - Significant at the .1 per cent probability level.

,0

(3)

-43-
with the following results:

cashzcrop farms

6
(1,183.70) (1,868.22)*** (.66)*** (26.02)***

Y - -17,431 +1865.79§X1 + 8,509.78X3 +'3.78X5‘+ 114.25X

fi'a .75, fig = .56, standard error of estimate $14,591

dairy farms
Y - -56,268.79 + 12,624.34!1 - 2,969.8513 + .3515
(4,187.44)** (2,581.43) (.52)
+186.40X6
(18.55)***

§'= .65, E? a .43, standard error of estimate $23,968

beef cattle farms

Y - -201,038.52 + 16,063.30x1 + 49,241.05x + 1.93x5

3
(7,006.27)* (8,391.67)*** (.76)*
+ 55.62x '
6
(39.12)

fifa .50, fig = .25, standard error of estimate $49,233

quantity of credit (Y) = f interest rate (XI) +~net worth
(X2) + gross farm income (14) + age (X7) + education (X8)
with the following results:

cash-crop farms
Y - 18,714 + 494.531:1 + .20x + 2.14x + 272.50x +2,437.80:x8
2 4 7
(1,153.98) (.08)* (.33)*** (162.24) (767.26)**

-44-
fi'= .75, fig = .56, standard error of estimate $14,653

dairy farms

(2,022.61)* (.04)*** (.28) (1.05.67)*

.

8
(518.28)***

- 4,228.33“

'1? .92, 8'2 == 85, standard error of estimte $12,320

beef cattle farms

Y'= 48,882.83 + 9,188.77X1 + .8912

(3,985.22)* (.06)***(.ll)* (226.21)

+ .241 + 167.34X
4 7

+ 809.09X8
(1,061.54)

ET: .87, fig = .76, standard error of estimate $28,400

The results of fitting the first set of equations indi-

cated that none of the equations fitted were good fits:
(l) the multiple correlation coefficient was very low in
some of the equations, (2) in all of the equations there
were only one or two regression coefficients significant at
the one per cent probability level, (3) the standard error
of estimate was fairly high in all of the equations and (4)
there were several high intercorrelations among the inde-
pendent variables in the equations.

Second lj‘_:_i._t_

The first set of regression equations furniShed the

-45-
basis for determining which variables should be included in
the second fit. The most important criterion in selecting
the independent variables was the degree of significance of
their estimated coefficients as indicated by the 't' test
used. The simple correlatién between a particular variable
and quantity of credit along with.the simple correlations
among the independent variables were also considered in
selecting variables to use in the second fit.

The equations used in the second fit were not the same
for all types of farms. The first equation for cash-crop
and dairy farms was fitted with seven variables. .All the
variables did not have significant estimated coefficients.
This fit permitted determination of simple correlations among
the seven independent variables and how much the multiple
correlation coefficient was increased when thle- many varia-
bles were considered. The second revised equation fitted
to the data of the cash-crop and dairy farms was fitted
using only the variables whose estimated coefficients were
significant in at least one of the three equations used in
the first set. This criterion was applied to the variables
for both the first and second equation of the second fit for
beef cattle farms.

The equations used in the second fit were simple linear
regression equations fitted by the least squares method. The
equations were:

for cash-crop farms

(1)

(2)

for

(1)

-46-
quantity of credit (Y) a f [Interest rate (X1),
net worth (X2), credit rating (X3), gross farm income
(X4), net farm income (15), size of farm (X6),
education (X8)_7
with the following results:

Y = ~23,837.60 + 1,920.43X1 + .25X + 6,466.26X

2 3

(1,146.47) (.10)* (1,968.54)**
(.54)* (1.21) (36.53)* (715.19)

'P: = .77, fi2 = .59, standard error of estimate, $14,019

quantity of credit (Y) = f [Interest rate (X1), net

worth (X2), credit rating (X3), net farm income (X5),

education (X8)_7

with the following results:

3
(1,236.28) (.10) (2,119.21)**

Y = -24,527.54 + 2,176.02]:1 + .18x2 + 7,326.73x

+ 3.9115 + 1,691.32X8
(.95)*** (738.22)*
'P: = .72, 112 = .52, standard error of estimate $15,192

dairy farms

quantity of credit (Y) = f [Interest rate (X1), not
worth (X2), credit rating (X3), gross farm inCome
(X 4) , net farm income (X5), size of farm (X6),
education (18)]

with the following results:

(C

K

I.

(2)

for

(1)

(2)

-47...
Y = 15,322.37 + 5,097.40}:1 + .69X2 - 2,025.71x3
(2,214.34)* (.04)*** (1,453.86)

+~.30X4 + .65x +1.st6 - 3,972.67X8

5
(.31) (.30)* (15.30) (511.79)*#*

ff - .92, .112 == .85, standard error of estimate $12,329

quantity of credit (Y) = f [interest rate (X1), net

worth (X2), credit rating (X3), size of farm (X6),

866 (X7)_7

with the following results:

Y a 3,940.34 + 3,931.26 + .76X - 4,212.15X3 - 18.62X

2 6

(2,562.39) (.05)***(1,646.29)* (17.96)
- 12600“
7
(124.83) -
fi = .89, fi2 = .79, standard error of estimate $14,454

beef cattle farms
quantity of credit (Y) = f [Interest rate (X1), net

worth (X2), credit rating (X3), net farm income (X517
with the following results: A Y
Y = -82,688.59 + 10,130.01x + .90x

1 2
(3,970.97)* (.05)***(5,130.88)*

+ 12,873.36x3

+ 8.12X
5
(4.20)
H = .87, fi2 = .76, standard error of estimate $27,865
quantity of credit (Y) = f [Interest rate (X1), net

‘ worth (x2), credit rating (1:3)]

’1

-48-
with the following results:

+ .92X

Y - -74,944.14 + 8,299.121 + 14,364.19x3

1’ 2
(3,888.30)* (.05)*** (5,114,43)¢.

fi'= .87, fig = .76, standard error of estimate $28,095

The results of fitting the second set of equations
showed a slight improvement over the first set of equations.
There were a few more variables that were significant in these
equations, especially in the equation fitted to the data of
the beef cattle farms. These were the result of the elimina-
tion of highly correlated variables. The standard error of
estimates in most instances remained about the same,not Show-
ing any important improvements. Therefore, as a group the
second fit did not Show any improvement over the first fit.
Only one equation showed enough improvement to be considered
better than the equations of the first set. This was equa-
tion one for beef cattle farms.

The results of this fit indicated, (1) that at least one
important variable was not included in the various equations
used, or (2) that the right list of variables had not been
used, or (3) there were too many highly intercorrelated
variables in the equations used.

41.2-29.4.

The procedure for selecting and fitting the first equa-

tion for the third fit was the same for cash-crop, dadry, and

beef cattle farms. The best fitting equation from.the first

-49-
two fits was selected and the residuals of this equation
were computed and plotted.to determine if there were any
revisions, additions or deletions that might possibly
improve the fit of this equation. From plotting the resi-
duals, it was evident that at least one fairly important
variable had been omitted. The availability of land con-
tracts, when plotted on a scatter diagram with quantity of
credit, seemed worth including in the equation to get a
better fit.

Cash-crop farms
The equation selected as the best fitting of the first

 

and second fits was equation one of the first set. A.land

contract variable was added to this equation to form the

first equation for the third fit, which was quantity of

credit (Y) = f [Interest rate (XI), net worth.(X2), credit

rating (X3), gross farm income (X4), land contract (X9);7.
The results of fitting this equation were:

Y = -38,796.51 + 4,716.54X1 + .23X + 9,319.391

2 3
(1,136.97)*** (.07)** (1,669.56)***
+ 1.22X4 + .31X9
(,29)**# (.05)*es

§'= .82,‘fi2 = .67, standard error of estimate $12,685

JDairygfarms
The best fitting equation of the first and second fits,

equation three of the first fit, was selected and land con-

'tract was added to this to form the first equation of the

-50-
third fit. This equation was quantity of credit (Y) = f
[Interest rate (X1), net worth.(Xz), gross farm income (X4),
age (X7), education (X8), land contract (19147. The results
of fitting this equation were:

- .14X

Y - 15,288.64 + 6,836.42X1 + .79X - 265.18X7

2 4
(2,033.89)** (.04)*** (.28) (101.87)**
(502.96)*** (.10)***

fi7= .93, fig 2 .86, standard error of estimate $11,876

Beef cattle farms

The best fitting equation from the first and second fits
was equation one of the second fit. With the addition of
land contract this equation'became quantity of credit (Y) =
f‘[Interest rate (X1), net worth.(xz), credit rating (X3),
net farm income (X5), land contract (X9)_7K The results of
fitting this equation were: A

Y = -121,550.88 + 16,747.3211 + .84X2 + 13,713.64X3

(3,997.29)*** (.05)**# (4,835.85)**

+ .007X5 + .29X9
(.43) (.06)***

§:= .89,’§2 = .79, standard error of estimate $26,245

The equations used in the third fit were better fitting
equations for all types of farms, in almost every respect.
The multiple correlation coefficient improved in each equa-
tion for all types of farms, compared to the prior best

-51-
fitting equation. There was: a greater number of signifi-
cant variables, with a higher degree of significance, in
each of these equations for the different types of farms.
Also, there was a slight improvement in the reliability of
the estimating quality of these equations as the standard
error of estimates were reduced for each equation, compared
to all prior fits.

Selection and Acceptability of Best Fitting Equations

The regression equations computed above were fitted to
the data for the particular type farm for the purpose of
selecting the best fitting equation to be used in computing
the marginal factor cost function of credit for that panti-
cular type of farm.

The basis for selecting the best fitting equation was
primarily the "goodness of fit" of the equation. Things con-
sidered in the goodness of fits were, (1) the adjusted
multiple correlation coefficient and coefficient of multiple
determination, (2) standard error of estimate, (3) number of
variables whose coefficients were significant and level of
significance, and (4) distribution of the residuals when
plotted about the regression line. Other items considered
in selecting the best fitting equation were, (1) the signs
and values of the estimated regression coefficients, and (2)
the intercorrelation among the independent variables.

Cash-crop_Farms

The best fit of the above equations for cashzceop farms

-52-
was the equation of the third.fit, which was quantity of
credit (Y) - f [interest rate (X1), net worth.(X2), credit
rating (13), gross farm income (X4), land contract (X9)_7K
The results of fitting this equation were Y - -38,796.5l +

4,716.54X1 + .23x2 + 9,319.39}:3 + 1.22x + .31x9

4
(l,l36.97)***(.07)** (1,669.56)***(.29)*** (.05)***

The equation was acceptable because: (1) The adjusted
multiple correlation coefficient was .82 with.sn.§2 of .67.
(This was the highest of any equation fitted to cash-crop
farms data.) (2) The standard error of estimate was $12,685,
which was fairly low considering the range of data was from
$1,000 to $115,000 with a mean of $48,053.25. (3) All of
the variables in this equation had estimated coefficients
which were significant at less than one per cent probability
level. (4) The signs of the estimated coefficients agreed
with the expected sign in each case. All of the estimated
coefficients had positive signs, which would indicate as
net worth, credit rating, gross farm income, and size of
land contract available increases in quantity the estimate
of the quantity of credit a cash-Crop farmer could borrow
increases.

One of the questionable features of this equation was
the value of some of its coefficients. 'For example: (1)
the coefficient fem net worth was only .23, which seems to
be on the low side, because one would expect, as the net

worth of a cash-crop farmer increases he should be able in

-53—
fact to borrow a greater portion of it than 23 per cent as
indicated by this equation. (2) The coefficient of 9,319.39
for credit rating seems to be reasonable. .A change from
a I'fair" to a 'good' credit rating probably would enable a
cash-crop farmer to borrow this additional money. (3) The
coefficient of 1.22 for gross farm income seems to be some-
what on the high side. This could be in compensation for the
low coefficient of net worth, because these two factors had
a simple correlation of .73, although when making real estate
loans, farmers can easily borrow more than their estimated
gross income. (4) The coefficient of .31 for land contracts
seems to be low because a farmer probably would.be able to
get more than 31 per cent credit when buying land on a land
contract. Two of the estimated coefficients seemed to be
high and two low compared to expected values. The high
coefficients tend to compensate for the low coefficients
when considering the complete equation.
Dairy Farms

The best fit of the above equations for dairy farms
was the equation of the third.fit which was quantity of
credit (Y) = f [Interest rate (X1), net worth.(Xz), gross
farm income (X4), age (X7), education (X8), land contracts
(Xg)[7. The results of fitting this equation were Y =
15,288.64 + 6,836.4211 + .79X2

(2,033.89)** (.04)*** (.28) (101.87)*** (502.96)***

- .14x4 - 265.18x7 - 4,436.02X8

+ .37119
(,1o)*sa

-54-

This equation was acceptable because: (1) The adjusted
multiple correlation coefficient was .93 with an i? of .86
(this was the highest of any equation fitted to the dairy
farms data). (2) The standard error of estimate was $11,876,
which was fairly low considering the range of data was from
$5,000 to $156,500 with a mean of $46,248.42. (3) All of
the variables except gross farm income had estimated coef-
ficients which were significant at less than the one per
cent probability level. (4) The signs of the estimated
coefficients agreed with.the expected signs for almost all
variables. One exception to this was the negative coeffi-
cient for gross farm income. This coefficient was positive
in equation three of the first fit. This was the equation
picked as best fitting prior to adding available land con-
tracts as a variable. The simple correlation between gross
farm income and land contract was .19, gpble 11. Therefore,
although this simple correlation was low, the addition of
land contracts changed the sign of the regression coeffi-
cient for gross farm income. Gross farm income and net worth
had a simple correlation of .71. Therefore, some of the
effect of gross farm income could be reflected in the coeffi-
cient of net worth. Another negative coefficient that was
not expected was that of education. This negative coeffi-
cient is easier to rationalize than for gross farm income.

In the first place, there was not too muCh variation in the

quantity of education among the farmers. Secondly, the

-55-
younger farmers had the most education, (-.32 intercorrela-
tion between age and education), also the intercorrelation
between quantity of credit and education was -.16. There-
fore, the negative coefficient for education was not too
illogical.

The values of the estimated regression coefficients
seemed to be fairly close to the expected values, for
example: (1) The coefficient of net worth was .79, which
indicates that as the net worth of a dairy farmer increases,
he could borrow 79 per cent of the increase. For some forms
of net worth this would be too high, while for other forms,
too low. Therefore the .79 seems to be reasonable as an
average value for this coefficient. (2) The coefficient of
-.14 for gross farm income is too low. It indicates that as
gross farm income increases, a dairy farmer can borrow less
credit than before; this is illogical and in disagreement
with reality. This negative coefficient compensates for one
of the other coefficients that is too high, probably not
worth, which is highly intercorrelated with gross farm
income. (3) The coefficient of -265.18 for age seems
reasonable as an average expected value, although this would
vary for different age groups. With an increase in age from
59 to 60 the quantity of credit would probably drOp as much
as $265.18. On the other hand, however, an increase in age
from 25 to 26 would probably increase the quantity of credit
3 dairy farmer could borrow. This indicates that a

-56-
curvilinear equation with first a positive slope and then a
negative slope might fit this better. The straight line
form equation does not reflect adequately all the effects
of age. (4) The coefficient of -4,436.02 for education seems
to have the wrong sign in view of economic and logical rea-
soning. This coefficient indicates that as a dairy farmer
increases his education by one year, the quantity of credit
available to him is reduced by $4,436.02. This regression
coefficient is probably reflecting the effects of some other
variable or variables. The simple correlation between age
and education was -.32. This negative correlation indicates
confounding effects with respect to education, though there
were no clear indications of compensating errors for other
variables in the coefficient of education. Another possible
reason for this negative coefficient for education could be
bias in the reapondentst answers, 1.6., the lower educated
farmers could have over estimated their quantities of credit
as compared to the estimates of the higher educated farmers.
(5) The coefficient of 6,836.42 for interest rate seems to
be reasonable as an average expected value. This indicates
that if the farmer is willing to pay an average of one per
cent more he can acquire $6,836.42 more credit. By shifting
to different combinations of loans, some of which require
higher interest rates, a farmer is usually able to obtain

more credit with resulting higher interest rate.

-57-
Beef Cattle Farms
The best fit of all the equations computed for beef
cattle farms was the equation of the third fit, which was
quantity of credit (Y) - f‘[Interest rate (XI), net worth
(12), credit rating (23), net farm income (X5), land con-
tract (x9)_7. The results of fitting this were Y = -121,-
550.88 + 16,747.32X1 + .84X2 + 13,713.64x3 + .0071:5 + .29x9.
(3,997.29)*** (.05)***(4,835.85)** (.43) (.06)***
This equation was acceptable because: (1) the adjusted
multiple correlation coefficient was .89 with an R? of .79;
(this was the highest for any equation fitted to beef cattle
farms data), (2) the standard error of estimate was $29,245,
which was fairly low considering the range from.$8,000 to
$239,500 with a mean value of $99,904.12, (3) all of the
variables, except net farm income, had coefficients which
were significant at less than the one per cent probability
level, (4) the sign of the estimated coefficients agreed
with the expected sign in each case. The coefficients of all
variables had positive signs, which indicates as a beef
cattle farmer‘s net worth, credit rating, not farm.income and
size of land Contract available increases in quantity, the
estimate of the quantity of credit he could borrow increases.
The values of the estimated regression coefficients
seemed.te be fairly close to the expected values in most
instances. For example: (1) The coefficient of not worth
was .84 which indicates that/as net worth of a beef cattle

-58-
farmer increases, he could borrow 84 per cent of its increase.
For some forms of net worth as pointed out in the discussion
of this factor for dairy farms, this would be too high while
for other forms too low. Therefore this coefficient of .84
seems reasonable as an expected value of this coefficient.
(2) The coefficient of 13,713.64 for credit rating seems to
be reasonable as an average expected value. This indicates
a beef cattle farmer could borrow an additional $13,713.64
without any change in his net worth or other factors as his
credit rating changes from say "fair” to "good". (3) The
coefficient of .007 for net farm income seems to be very low.
Further, the coefficient of this variable was not significant
at even the 10 per cent probability level. This coefficient
indicates that as a beef cattle farmer's net income or his
expected net farm income increases by $1,000, he can borrow
only an additional $7, which as pointed out above seems
very low. (4) The coefficient of .29 for land contract seems
to be low as a farmer should be able to borrow more than 29
per cent of a land contract with usual down payments. (5) The
coefficient of 16,787.32 for interest rate seems to be high.
This indicates if a farmer is willing to pay one per cent
more in interest rate he can acquire $16,747.32 more credit.
By shifting to different combinations of loans which require
higher interest rate in some cases, the farmer is usually
able to obtain additional credit with resulting higher
interest rate. The quantity $16,787.32 seems to be a high

 

 

-59-
estimate of this additional quantity.
Comparison of the Best Fits

0ne of the first differences evident among the best
fitting equations was that the variables included in these
equations were different for each type of farm. Only three
variables, interest rate, net worth, and land contract were
included in all the equations, while two variables, credit
rating and gross farm income, were included in two of the
three types of farm equations. Net farm income, age and
education were included in only one equation. All other
variables were eliminated before the final equation. There-
fore, based on the number of best fitting equations the
variables were used in, interest rate, not worth and land
contract were the more important variables which affected
the quantity of credit. Second in importance, based on
number of equations, was credit rating and gross farm income,
followed by not farm income, age and education.

One of the causes of different variables in the final
equations for cash-crop, dairy and beef cattle farmers was
the process of eliminating highly correlated variables. In
the instances of two highly correlated variables, for example
gross farm income and net farm income, only one variable
was selected for the revised equation. Thus some variables
that were fairly important to the different types of farms
may have been eliminated before the final equation. Omission

of a factor from the final equation for a particular type of

-60-
farm does not mean this factor is insignificant or unimpor-
tant. The effects of the variables that were eliminated
were reflected in the values of the remaining variables
which were correlated with the omitted variable.

Another difference was in the values of the different
regression coefficients for net worth, which was only .23
for cash-crop farmers as compared to .79 for dairy farms
and .84 for beef cattle farms. This seems to indicate that
net worth is not as important to a cash-crop farmer as it
is to a dairy or beef cattle farmer in determining the quan-
tity of credit he can borrow on the capital market. These
coerticients indicate as the cash-crop farmer's net worth
increases he can borrow only 23 per cent of the increase
compared to 79 per cent for dairy and 84 per cent for beef
cattle farmers. The low coefficient for cash crop farmers
indicates that there was a possibility of some other factor
reflecting part of the effects of net worth. The high coeffi-
cients for dairy and beef cattle farmers indicates that there
was a possibility their values reflected some positive effects
of another factor or factors.

Gross farm income and net worth in the cash-crop equa-
tion had a simple correlation of .73. The estimated coeffi-
cient of gross farm income was 1.22 which seems high based
on economic and logical reasoning. When these two factors
are considered together they seemed to indicate that part of
the effect of net worth was reflected in the coefficient of

-61-
gross farm income. Another possible reason for the low
coefficient of net worth was that the simple correlation
between not worth and size of farm was .35, while the simple
correlation between gross farm income and size of farm was
.76. Therefore gross farm income rather than not worth is
more likely to reflect the effect of size of farm.

On the other hand, the coefficient of -.14 for gross
farm income for dairy farmers indicated that all the effects
of this factor were reflected in values of other coefficients.
The simple correlation between gross farm income and net
worth was .71. This and a fairly high .79 coefficient for
net worth seemed to indicate in terms of economics that part
of the effects of gross farm income were reflected in the
coefficient of net worth. For dairy farmers, net worth
probably also reflected most of the effects of size of farm.
Little of the effect of size of farm could have been
reflected in the gross farm income coefficient which was -.14.

The coefficient of .007 for net farm income for beef
cattle farmers indicated most of the effect of net farm income
was reflected in another variable or variables. This coeffi-
cient was reduced from 8.12 to .007 as a result of adding
land contract as a variable. This and the fact that the sim-
ple correlation between land contract and net farm income was
.49 indicates that part of the effects of net farm income was
probably reflected in the coefficient for land contracts.

Another factor that could be reflecting part of the

-62-
effects of net farm income is net worth. The simple correla-
tion between net farm income and net worth is only .25 but
that between gross farm income and.net worth is .61. Gross
farm income was not included in the final equation but its
effects should be reflected in net farm income (their simple
correlation was not computed) based on a prior knowledge.
Therefore, the coefficient of .84 for net worth which is
relatively high in terms of economic and logical reasoning
probably reflects part of the effects of gross and net farm
income.

Even though there are indications that the regression
coefficients for net worth, gross farm income, and net farm
income have compensating errors, the results indicate that
net worth is not as important to cash-crop farmers as to
dairy and beef cattle farmers in determining the quantity
of credit they can borrow. This conclusion is also supported
by the types of net worth these farmers own. For example:

A much higher percentage of a cash-crop farmer‘s net worth
is machinery and equipment than either dairy or beef cattle
farmerst net worth. A relative high percentage of the dairy
and beef cattle farmers! net worth is cattle, either dairy
cows or feeder cattle. Agricultural lenders will loan a
much higher percentage of value for cattle than for maChinery
and equipment which depreciates fairly rapidly.

There was quite a degree of difference in the coeffi-

cients of interest rate, with 4,716.54 for cash-crop farmers

-63-
compared to 6,836.42 for dairy and 16,747.32 for beef cattle
farmers. These coefficients indicate that interest rate had
more effect upon the quantity of credit a beef cattle farmer
could borrow than for either a cash-crop farmer or a dairy
farmer; by paying a one per cent increase in interest rate,
the beef cattle farmer estimated he had available almost
three times more additional credit than either of the other
type farmers.

The coefficient of credit rating was 9,319.29 for cash-
crop farmers and 13,712.64 for beef cattle farmers. Inter-
correlations do not appear important in this instance. This
difference seems to be caused primarily by the differences
in the type of individual managing these reSpective types of
farms. For example, practically all the beef cattle farmers
were fairly well established farmers with a wide range of
credit experience as compared to quite a number of the cash-
crop farmers who were part time and/or ”in and out" farmers.
Generally, cash-crop farmers surveyed appeared to be poorer
managers than beef cattle farmers. This difference is based
primarily on personal observations by the author during the
study.

One of the surprising results was that net farm income
was not significant in any of the equations and that it was
included in only one best fitting equation. This could be
explained by one of several possibilities: (1) there were a

few dairy and beef cattle farmers in the sample who had

-64-/.S’
negative net farm incomes; (2) gross farm income rather than
not farm income is often used as a_criterion for the farmer's
repayment ability by the lending agencies; (3) net farm income
is subject to wider percentage fluctuations than gross farm
income and (4) high intercorrelations existed between this
factor and other factors, thus the coefficients for the
other factors probably reflected the influence of net farm
income.

The only variable used in all best fitting equations that
did not indicate any large degree of difference for the dif-
ferent type of farms was land contract. The various coeffi-
cients for land contract were .31 for cash-crop farms, .37
for dairy farms, and .29 for beef cattle farms. Therefore,
the availability of land contracts affected the quantity of
credit available for all types of farms about the same.

The relatively low regression coefficient for land con-
tracts may be explained by the method used in measuring
available land contracts. For example, if a farmer estimated
he could purchase $60,000 worth of land on land contract with
$20,000 (one-third value) as down payment and if the farmer
had the $20,000 to finance the land contract, this was
recorded as a $60,000 available land contract which is the
independent variable in the study. However, the credit
acquired was treated as $40,000. Therefore, the regression
coefficient for available land contracts should not be one

as land contracts are not 100 per cent financed.

-66-

Another possible reason for the low coefficients for
land contracts are the simple correlations between land
contracts and interest rate. These are -.36, -.31 and -.38
for cash-crop, dairy and beef cattle farmers respectively.
Though these simple negative correlations are not so large,
the coefficient for interest rate was greatly affected by
introducing land contracts. The respective changes were from
2,176.02 to 4,716.54 for cash-crop farmers; from 3,931.26 to
6,836.42 for dairy farmers; and from 8,299.12 to 16,747.32
.for beef cattle farmers.

The Marginal Factor Cost Functions

A marginal factor cost function for credit for farmers
treats interest rate as a function of the quantity of credit
with other variables, not worth, credit rating, etc., held
constant. The list of variables used in fitting the average
cost functions (from which the marginal cost functions of
this study were derived) was based on the regression equa-
tions selected as ”best” when the quantity of credit was
treated as the dependent variable.

gash-crop_Farms

When the best fitting regression equation was recalcu-
lated with interest rate as the dependent variable, the
equation fitted to the cash-crop farms data was interest
rate (Y) = f [Huantity of credit (X1), net worth (X2),
credit rating (13), gross farm income (X4), land contract
(X9)J7, with the following results:

‘4‘

 

 

(.0000053)***(.0000047)***(.11)***(.000021)

- .0000231:9
(.0000033)#**

B’- .58 'fi2 = .34, standard error of estimate .87.

. . MFG = 7.74 + (2).00002211 — .000017X2 - .6213 + .000031X4
- .000023X9

(X1 = Q.- quantity of credit)

By varying the quantity of credit (XI) and using the
mean value1 of the other variables the marginal factor cost
schedule was computed and is illustrated in Figure 2.

This marginal factor cost function when plotted, with
interest rate on.X axis and quantity of credit on Y axis,
was a straight line because it was derived from a straight
line function. The residuals for this equation were plotted
about the regression line and their pattern indicated that
a curvilinear function might fit the data better than a
straight line.

.A curvilinear function was fitted to these data by
squaring the quantity of credit term (0) and then plotting
this against the scale of the unsquared term. This permitted
the function to curve upward when plotted with interest rate
on the'2 axis and quantity of credit on the X axis.

 

1For mean values see Appendix B, Table 1.

 

C c
. .
p
e
o
. a
o .
- -
V.- .
o
_ e
u
I
e e . _ / \ .
e -
-
. . . .
. A
I - ,
u
e z
t \ g
/\ e -
\ -
_
e
I\ a
.
\ .
f o
.
n
l
-
. —
e
K . I /
r t
{K / . ,\
I
e /\ . n
.. . v1.
\ e
./
s e
o - o
. e ex \
. e
e
v ‘ _
\ -d

1‘

-68..

The equation fitted was interest rate (Y) - f [Huantity
of credit (11)z, net worth.(xz), credit rating (13), gross
farm income (X4), land contract (19)].

with the following results:

Y - 8.05 + .00000000023112 - .00001812 - .5413

(.000000000052)***(.0000048)***(.ll)***
+ .000023x4 - .00002519
(.000021) (.0000034)***

R‘ - .59, 32 - .34 standard error of estimate .86

. . MEC - 8.05 + (3) .00000000023x12 - .000018x2 - .54:3

By varying the quantity of credit C11) and using the
mean value of the other variables, this curvilinear function
was computed and is illustrated in Figure 2. The residuals
when plotted about the regression line indicated that the
curvilinear function fitted the data better than the straight
line function. This seems logical because as the cashzcrop
farmer uses large quantities of credit over $60,000, the
interest rate startmzto increase quite rapidly. This agrees
with the general observation that after a certain limit is
reached, farmers are not able to borrow more money regardless
of the interest rate they are willing to pay. Thus, even-
tually the MFC function will be a straight, vertical line
with perfect inelasticity.

.maoaaoo aosouzmso zoo ozozoosso omoo aoooso Headwaez .m oz=MHz

AM u Hz AmosomsoSav azoozo oo hazusoza
OAN com Om.” omH ObH cm...” cm." Owe." 06H ONH 6.2.. OOH am am Ob Ow on 0% cm. GN OH O

1 d a 1 a 1 s 1

 

. mJV

. o.m

. m.m

6.0

-69...

¢

Aoz+ z+oz+mz+zzoouz

zoz+ez+mz+~z+mzxoouz o o
. o.o

. OX:

 

 

 

one: Mwouoasz

-70-
‘Dairy Farms

When the best fitting regression equation was recal-
culated with interest rate as the dependent variable, the
equation fitted to the dairy farms data was interest rate
(Y) = f [quantity of credit (X1), net worth.(X2), gross
farm income (X4), age (X7), education (X8), land contract
(X9 )J

with the following results:

Y = 5.41 + .000010X1 - .0000086X2 + .00000012X

4
(.0000030)***(.0000028)*** (.000011)

+.0015X7 + .036X8 - .000017X9

(.0040) (.023) (.0000039)***

§'= .37, B2 = .14, standard error of estimate .45.

+ .0015X7 + .036X8 - .000017X9 (X1 = quantity of
credit)

+ .00000012X4

By varying the quantity of credit (X1) and using the

mean values2

of the other variables, the marginal factor cost
schedule was computed and is illustrated in Figure 3.

This function when plotted, with interest rate on the
Y axis and quantity of credit on the X axis, was a straight
line because it was derived from a straight line function.

The residuals for this equation were computed and plotted

about the regression line and their pattern indicated that a

 

ZFor mean values see Appendix B, Table 1.

.mhofihsm zzfioo Loo :ozeosso umoo aoaooh Hosfimzmz .m oazwzz
zK u zx zoos-oases: ozooao oo zozossso
ozm com oozomz ooz ooz ooz oez ooz omz ozz ooz oo om or oo on 3. on on oz

.I‘ 1‘ fl ‘ q

C

 

1.- J.‘ G ‘ J 1‘1 1|“ ‘ I J

1 d d i l

-71-

o.b
m.b
o.m

m.w

 

. c.0H

 

 

 

mead anchoaaH

-72-
curvilinear function might fit the data better than a
straight line function. The results of fitting a curvi-
linear function were not as good as the straight line
function; therefore, the curvilinear function was discarded
and the straight line function was selected as the best
fitting MFC function for dairy farms.

This straight line function is reliable only over the
range of data for the dairy farmers, which was, as indicated
above, up to $156,500. Probably after this, the function
will be steeper than the slope of this straight line. This
conclusion is based on the conviction that after the average
dairy farmer surveyed borrows over $150,000, he would not be
able to get much more regardless of the interest rate he
would be willing to pay. Thus eventually the slope of func-
tion should be expected to change from .000020 to infinity.

Beef Cattle Farms

When the best fitting regression equation was recalcu-
lated with interest rate as the dependent variable, the
equation fitted for the beef cattle farms was interest rate
(Y) = f quantity of credit (X1), net worth (X2), credit
rating (X3), net farm income (X5), land contract (X9);7

with the following results: A

Y = 5.86 + .0000058X1 - .0000031X2 - .081'X3 - .0000063X5

(.0000014)*** (.0000015)* (.092) (.0000080)
- .0000065X9
(.0000011)***

 

 

 

\

N N i . iii-ll
liltll
'I?‘

 

 

-73..

2

fi’= .48, fi' = .23, standard error of estimate .487.

. . MFC = 5.86 + (2) .0000058(X1) - .0000031x2 - .081X3

- .0000063X5 - .000006519

By varying the quantity of credit (X1) and using the
mean values3 of the other variables, the marginal factor
cost schedule was computed and is illustrated in.Figure 4.

This NFC function, when plotted with interest rate on
the Y axis and quantity of credit on the X axis, was a
straight line because it was derived from a straight line
AFC function. The residuals for this equation were computed
and plotted about the regression line. Their pattern indi-
cated that a curvilinear function might fit the data better
than a straight line function. However, when a second
degree term was fitted for quantity of credit, no important
curvilinearity was revealed.

The linear function is reliable only over the range of
data for the beef cattle farmers which was, as indicated
above, up to $239,500. Beyond this quantity the function

will probably turn up.

 

3For mean values see Appendix B, Table 1.

.mzoaaoo ozeooo hoop Loo newnesso amoo zoaooo Hssfimzoz .v oaswfiz

«Man «N amossmzozov pzooao mo zpzosoaa

 

$ A d a d a a 1 a u a a A d a

can cow. OGH omH 05H cm: and” cm." and“ ONH OHH OOH om ow Ob cm on 9v on om o o

. 03v

mlm

0.6

-74.-

m.o
of“.
m..~.

 

o.w
l mOw
. c.m
.. m.m

.. o4:

 

 

 

oeez anemoasm

CHAPTER III
ESTIMATING A MARGINAL‘FACTOR.COST FUNCTION FOR CREDIT
FOR CASH-CROP, DAIRY AND BEEF CATTLE FARMERS
FROM DATA FOR ALL FARMS

In this chapter a marginal factor cost function will be
estimated on the basis of the data collected from all the
farmers included in this study, 29 cash-crop, 30 dairy and
29 beef cattle farmers. This function will treat type of
farm as an independent variable.

In this chapter the first section following this intro-
duction will present a general comparisOn of the three types
of farms, cash-crap, dairy and beef cattle. The second
section will present the results of the first set of equa-
tions used for the individual farm types, when fitted to the
data from all farms. The third section will present the
results and a discussion of the revised equations. In the
fourth section the equation selected as best fit will be
presented. Also, in that section this best fitting equation
will be used to derive the marginal factor cost function for
credit.

General Comparison of the Cash-Crop, Dairy
and Beef Cattle Farms

The cash-crop farms were smaller in total acres opera-
ted than either the dairy or beef cattle farms, averaging
only 134.5 acres compared to 214.0 for dairy and 213.5 for
beef cattle, Table 13. This was caused in part by the

stratified sample of cash-crop farms, which required one-

-76-
third of the farms to be 80 acres or under. The dairy and
beef cattle farm samples required only a farm with a few
milk cows (20 or under) or a few feeders (50 or under) and

this did not necessarily require a smaller farm in total

 

 

 

acreage.
TABLE 13
AVERAGE SIZE, YEARS OF FARMING, AGE, AND
EDUCATION, BY TYPE OF FARM, 88 FARMS,
SEIECTED AREAS as MICHIGAN, 1960
Size Farming

Type of Number (total Experience Age Education
Farm of farms acres) Lyears) (years) (gradeslg
Cash-crop 29 134.5 18.2 43.7 ' 9.4
Dairy 30 214.0 20.0 46.5 9.4
Beef-cattle 29 213.5 20.6 46.1 10.6

 

Within the individual type of farmbreakdowns, the
younger farmers had more education and in some instances
operated larger farms. When the three types of farms were
compared, the cash-crop farm operators were the youngest with
the least education while the beef-cattle farm Operators were
the oldest with the most education. Thus, there was a high
correlation between education and type of farm as well as
between education and age of farmer.

The cash-crop farmers had the highest average off-farm
income of the three types of farms, about 250 per cent

greater than either the dairy or beef cattle farmers. This

-77-
was because of the large percentage of operators of small
cash-crop farms had either a full-time or a part-time job
in addition to their farm operation.

For gross farm income, net farm.income and total income,
averages per acre were calculated in addition to the average
per farm to give a better comparison among the types of
farm. The average cash-crOp farm was much smaller than the
average dairy or beef cattle farm. The beef cattle farmers
had a very much higher average gross farm income per farm
and per acre, with the dairy farmers having the second high-
est per farm and the cash-crop farmers having the second
highest per acre, Table 14.

TABLE 14
AVERAGE OFF-FARM INCOME, GROSS FARM:INCOME, NET FARM INCOME

AND TOTAL INCOME, BY'TTFE 0F FARM, sstARMs,
SELECTED AREAS OF MICHIGAN, 1960

 

 

 

Type of 0f?:farm fiross Farm Net Farm ‘Total
Farm Income Income Income Income
Per Per . Per Per Fer Per
farm acre farm acre farm acre
Tdolij' (dol.7”(dol.5 Idol.) Idol.) Idol.;(dol.;
Cash-crop 1,724 9,119 67.80 2,649 19.69 4,392 32.60
Dairy 488 11,748 54.90 3,473 16.23 4,181 19.56

Beef-Cattle 611 37,513 175.70 5,335 25.00 6,179 28.92

 

The beef cattle farmers also had the highest net farm
income,as measured by both average per farm and per acre,

although the difference was not nearly as great as in gross

 

 

 

-78-
farm income. The average net farm income per acre was
$25.00 for beef cattle farms compared to an average of
$19.69 for cash-crop farms and $16.23 for dairy farms.
When the total income was compared by type of farm, the cash-
crop farmers had enough off-farm income to more than off-set
the higher net farm income of the beef cattle farmers on an
average per acre basis but not on an average per farm'basis.

The average credit rating by type Of farm was positively
correlated with net worth, with the cash crop farmers having
the lowest average net worth of $46,133 and lowest credit
rating of 3.4, compared to an average net worth of $58,566
and credit rating of 3.5 for dairy farmers and a net wOrth
of $78,713 and credit rating of 3.8 for beef cattle farmers,
Table 15, The average interest rate farmers estimated they
would have to pay for credit was negatively correlated with
credit rating and net worth. The interest rate was also
negatively correlated with quantity of credit, with the
quantity of credit being positively correlated with credit
rating and net worth.

The most significant difference among the types of
farms was the percentage of net worth the different types of
farmers estimated they could borrow. The beef cattle farmers
estimated they could borrow 142.3 per cent of their net
worth, compared to 117.5 per cent for cash-crop farmers and
only 91.3 per cent for dairy farmers. Thus, the type of

farm should be a significant variable when estimating the

-79-
marginal factor cost function for credit.
TABLE 15
AVERAGE CREDIT RATING, NET WORTH, INTEREST RATE AND QUANTITY

OF CREDIT, BY TYPE OF FARM, 88 FARMS,
SELECTED AREAS OF MICHIGAN, 1960

 

 

Type of Credit Net Ifiterest Quantity Of
Farm Rating, Worth Rate Credit
(dol.) (pct.)f (dol.)
Cash-crop 3.4 46,133 6.4 54,203
Dairy 3.5 58,566 6.1 53,374
Beef-cattle 3.8 78,713 6.1 111,980

 

Simple Correlations Among the Independent variables

There were some high simple correlations among the vari-
ous independent variables when these were computed for the
variables used in the equations fitted to the data of the
individual type farms. Therefore to determine if these or
other variables have high simple correlation when used in
equations fitted to all farms data, the simple correlations
among the variables used in the various equations were com-
puted. These appear in Table 16.

The simple correlations among the independent variables
were much lower for all farms than those for the particular
types of farms. There was only one simple correlation
above .40. This was between net worth and gross farm income

with a simple correlation of .62.

 

 

 

O
e K
’. K K
K
\ O / \ ° / \ . /
K ‘ K ' ..
a ‘ " °
1 ‘ ‘ . M
) O
Q
. O
O l

 

 

oo.H somaucoo damn

 

   
  

oz co.H sewaeo=dm

oz mo.u oo.H moo

oz oz oz oo.a seem co exam

oz oz oz ow. oo.H «goons seam omz

m on. ma. oz. oz oz oo.H oaoosfi spec mooso
a ma. oz oz me. an. on. oo.H . masons osooso
mm. mo.u om. oz oz no. ow. oo.H nope: omz

oo.a oo.u oo. ma.u sz.u mo.n em.u so.u oo.z moan amoeooaa
posaasoo neapmosuw ewd Bash cacosfi oaoesfl mafiama Sages ouch eancfias>
damn no Eaa% Bash vfidoau eoz amoaoazH GodzomousH

oufim 9oz mmcaa mansfiha
#:odsomodzH

oooz .zdonon as nomad ameomqmm .mzmom mm
mmnm<am¢> szmozmmmozH was ozez< ooneeammmoo mgmsz

0H HAMdB

 

 

 

 

-81-
Regression Equations Fitted

The data from all the farms were combined and used to
derive a marginal factor cost function with the type of farm
as one of the independent variables. The equations fitted
for all farms were, as for the individual type farm, simple
linear regression equations fitted by the least squares
method.

The First Fit

The equations used in the first fit for all farms data
were the same as the equations used in the first fit for the
individual type farms. 'The equations fitted and results of
these were:

(1) quantity of credit (Y) = f [interest rate (X1), net
worth (X2), credit rating (X3), gross farm income (X4147
with the following results:

Y = 35,497.95 + 5,346.84X1 + .71x2 + 3,150.62X3 + .74x4
(1,487.5l)*** (.03)***(1,704.84) (.07)***

R'= .87, R2 = .76, standard error of estimate $23,478

(2) quantity of credit (Y) = f [interest rate (X1), credit
rating (X3), net farmincome (X5), size of farm (X6147
with the following results:
Y = 78,988.34 + 8,633.13X1 + 16,850.11X3 + 2.37x5 + 133.33x6
(2,634.42)** (2,874.87)*** (.49)*** (20.38)***

._ ._2 .
R = .50, R = .25, standard error of estimate $41,177

 

\

\

.10! I4

 

-82-
(3) quantity of credit (Y) = f [interest rate (X1), net
worth (X2), gross farm income (X4), age (Xfi), education
(2:8)] '
with the following results:
Y = -17,464.02 +»47,404.07x1 + .72X + 7.68x

2 4
(14,591.53)** (.04)***(.69)***(1,144.15)

- 525.77):7

- 1,962.64X8
(5,563.10)

R'= .87, R2 = .75, standard error of estimate $23,580

One of the apparently surprising results of this fit was
that the estimated coefficients of all the variables in equa-
tion two were highly significant, although the'R'2 was very
low, .25, and the standard error of estimate high, $41,177.
The first and third equations indicated a fairly good fit
witthz of.75, and standard error of estimates of approxi-
mately $23,000, although both of these equations included
some variables whose estimated coefficients were not signi-
ficant even at the 10 per cent probability level.

Improving the standard error of estimates and increasing
the number of significant variables, for the same equation,
were two factors considered in selecting new equations to be
fitted to all farms data.

Second‘Fit

There were not any equations fitted to all farms data

 

 

J
_!

 

 

~83-
comparahle to the second fit of the individual type farms.
The procedure for the second fit involving the data for all
farms was similar to the procedure used for the third fit
for the individual type farms. The best fitting equation
from prior fits, in this instance from the three equations
used in the first fit, was selected and modified to include
available land contracts and type of farm as independent
variables. The equation fitted was quantity of credit (Y) =-
f [Interest rate (XI), net worth (12), credit rating (13),
gross farm. income- (X4), land contract (19), cash-crop farm
(x10), dairy farm (11]), beef cattle farm (112)]
A with the following results:
Y - -50,191.14 + 8,310.0111 + .75X2 + 3,437.75!3 + .2414
(1,341.05)*** (.03)*** (1,467.60)* (.97)*
+ .291:9 - 3,757.72!” - 10,999.35111 + 13,821.12112
(.03)*** NC ‘ NC NC
ii .. .91, 52 - .82, standard error of estimate $20,139
The dummy variable technique was used for the type of
farm,fer example, for cash-crop farms data 110 - l; 211 - 0;
112 - 0, while for dairy farms data 110 - 0; 111 - l; and
112 - 0. The computer would not handle three duny variables;
therefore, 112 was deleted from computer computation and the

regression coefficient for 112 was computed later}1 When the

 

180a Appendix A for complete details of the method used.

~84-
coefficient for;xlz was computed, the coefficients of x10
and X11 and the “A” value had to be adjusted. Since the
coefficients of 1110 and x11 had to be adjusted, the standard
errors for these variables were not accurate, therefore,
neither the standard error nor the ”t" test values are pre-
sented for X10, X11 and.X12. The standard errors computed
for X10 and.X11, even though not applicable, indicated that
the type of farm variables were highly significant.

This equation indicated improvement in three important
aspects over the prior best fitting equation. These improve-
ments were: (1) The coefficient of multiple determination
improved from .76 to .82. (2) More variables were significant
and at a higher level of significance. (3) The standard error
of estimate decreased from $23,478 to $20,139, which improved
the estimating quality of the equation.

Selection and Acceptability of Best Fitting Equation

The regression equations computed from the data of all
farms were fitted to these data for the purpose of selecting
the best fitting equation to be used in computing the marginal
factor cost function for credit. The best fit of all the
equations computed from the data of all farms was the equa-
tion of the second fit which was quantity of credit (Y) = f
[interest rate (X1), net worth (X2), credit rating (X3),
gross farm income (X4), land contract (X9), cash-crop farm
(X10), dairy farm (X11), beef cattle farm (X12147. The
results of fitting this equation were Y = -50,191.14 +

 

M-" ""r' I

’.\

{It‘llull‘l

I
I \
Ix
O O
l
O
\ v1
. I l
'\
O \
, Q.
‘ v 4
O
.
O.
Q I
E
O
1
O 1
.I
Q. I
\ u s I A
0
v s.
\
‘ 6
vi
.
\. \
9\

 

~85-

8,310.01X1 +-.75x2 + 3,437.45x + .24x4 + .29x9 - 3,757.72x

3 10

- 10,999.35X11 + 13,821.12X12.

The bases for selecting the best fitting equation were
the same as those for selecting the best fitting equation for
the individual type farm. These were: (1) the adjusted
multiple correlation coefficient and multiple determination
coefficient, (2) standard error of estimate, (3) number of
variables whose coefficients were significant and level of
significance, (4) distribution of the residuals when plotted
about the regression line, (5) the signs and values of the
estimated regression coefficients and (6) the simple corre-
lations among the independent variables.

This equation was acceptable because: (1) The adjusted
multiple correlation coefficient was .91 withanR2 of .82.
(2) The standard error of estimate was $20,139 which was
not too high considering the range of data was from $1,000
to $239,500. (3) The variables all had coefficients which
were significant. (4) The signs of the estimated coeffi-
cients agreed with the signs expected on the basis of theory
and a prior knowledge. The variables, interest rate, net
worth, credit rating and land contract had positive coeffi-
cients. This is in agreement with what farmers can in fact
do. The type of farm variable had cash-crop and dairy farms
with negative coefficients and beef cattle farms with a
positive coefficient. This agrees with the expected because

of the higher gross farm income and type of net worth the

-86-

boef cattle farmer, with other things equal, can borrow
more money. This was substantiated by the marginal factor
cost functions for credit when computed by type of farm.

The values of the type of farm coefficients seemed to
be fairly close to expected values. For example: (1) The
coefficient of cash-crop farms was ~ -3,757.72, which indi-
cates the cash-crop farmer‘s credit or available credit on
the money market would be S3,757 .72 less than the average
farmer's. The coefficient for dairy farm was ~10,999.35
which indicates his available credit is $10,999.35 less than
the average farmer. The coefficient for beef cattle farm.aas
13,821.12, which indicates the beef cattle farmer could get
$13,821.12 more credit, with other things equal, than the
average farmer. These values seem to be very reasonable
estimates of the actual differences that exist among these
types of farmers and quantities of credit they can borrow.
(2) The coefficient of not worth was .75, which.indicates
that as a farmer's net worth increases he can borrow 75 per
cent of the increase. This compares favorably with .75 for
dairy and .84 for beef cattle farmers from the regression equa-
tions for the individual type farms. This value is muCh
higher than .23 for cashpcrop farmers on an individual farm
basis. It tends to be an average for cashpcrop, dairy and
beef cattle farmers. The differences in this value for the
individual type farms are too great to combine into one value

to represent all farmers. (3) The coefficient for credit

-37-
rating was 3,437.75 which indicates as a farmer's credit
rating changes one classification he can borrow $3,437.75
additional money or purchase this much more on credit. This
value seems to be reasonable because a farmer's credit rating
greatly affects his ability to borrow and in Eost instances
>the amount of credit extended to hime (4) The coefficient
of 8,310.01 for interest rate seems to been the high side
of expected value because of policies of lending institu-
tions. They do not increase the size of a farmer's loan
just because he is willing to pay a higher interest rate.
They loan money at a fixed interest rate and the quantity of
the loan is determined by the farmer‘s assets and other
qualities. Farmers can usually get Eore money by changing
the combinations of loans with higher interest payments.
However, $8,310.01 seems a high estimate of the average
increase due to one per cent Change in the interest rate.
The Marginal Factor Cost Functions

The regression equation selected as the best fitting
equation was fitted with.the quantity of credit as the
dependent variable. To facilitate the computing of the mar-
ginal factor cost function this equation, using the same
variables, was refitted with interest rate as the dependent
variable and quantity of credit one of the independent
variables. Thus, the equation fitted was interest rate (Y) =
f fiuantity of credit (XI), net worth (22), credit rating

(Kg), gross farm income (X4), land contrast (X9), cashpcrop

 

e
r

\ e

e

. .—

I

I
K
i
K

I
’ e
\ - e
e '—
\
q s
. l\ 1
. —
l
K 9.
.t’
K
e

 

 

 

.maasm Has new sass Seam eo>fiaod .naoaach
sauces moon was .mafisd .aeaelnmse Lem macfiaossm pace aeeosm Hesfiwasz .m oasmam

“M.” HN Amuasmsonvv afiuoao we maavssza

 

4 a

com Om." aid as...” owa Omdctva and ONH OHH OOH cm om Ob cm on 0* on ON OH 0

m.¢
o.m

m.m

-89..

0.0

 

maoaasm oaoaso «mom m.©

opossum neaelnmso .o.b

.m.b

maeaash hafisn

  

l

O.m
. m.w
i 6.9

Aha

 

 

 

ovum anoaowsH

-90-

were indicated and discussed in the chapter on the individual
type farm. .After the quantity of credit passes beyond this
range the function will likely curve upward until it becomes
a vertical (straight) line. This is because of the fact, that
after a farmer borrows his limit at the market rates of all
lending agencies, open accounts, land contracts, etc., he will
reach a point where he cannot borrow any more money regardless
of the interest rate he is willing to pay.

There wasenot much difference in the marginal factor
cost functions derived by this method, as can be readily seen
in Figure 5, for the different types of farm. The cashpcrop
function (based on the one derived for only cash-crop farmers)
was decreased and the beef cattle function increased. Due to,
the large degree of difference in the marginal factor cost
functions fler the type of farm between types of farms, this

function does not seem to offer very mush.

 

CHAPTER IV
SUMMARY AND CONCLUSIONS

The primary objective of this study was to estimate and
describe the marginal factor cost functions for credit for
selected cash-crop, dairy and beef cattle farmers of Midhigan.
The data necessary to derive these marginal factor cost
functions were collected by personal interview from these
three types of farms. Therefore, these functions represent
the farmers' estimate of the marginal factor cost of credit
available to them at various interest rates. The secondary
Objective was to determine the factors which significantly
affect the quantity of credit the different types of farmers
estimated they could borrow.

Three counties were selected, one for each type of farm,
i.e., Saginaw for cash-crop, St. Clair for dairy and Lenawee
for beef cattle. These three counties constituted the sample
area. See Chapter I for details.

The questionnaire used to collect these data was designed
to collect data on certain factors considered to significantly
affect the quantity of available credit. These factors were:
(1) interest rate, (2) net worth, (3) farmers‘ credit rating,
(4) gross farm income, (5) net farm income, (6) size of farm,
(7) age, (8) education, (9) land contracts, (10) off-farm

income and (11) farming experience.

-92-

Results and Conclusions

Some of the more important characteristics of the farms

will be given and compared, by type of farm to get a clearer

picture of the basic characteristics of the farms from which

data were collected. These were:

1.

2.

The cash-crap farms were smaller in total acres
operated than either the dairy or beef cattle farms,
averaging only 134.5 acres compared to 214.0 for
dairy and 213.5 for'beef cattle farms.

The cash-crop farmers were slightly younger, with
an average age of 43.7 years as compared to 46.5
for dairy farmers and 46.1 years for beef cattle
farmers.

The cash-crop farmers, although younger, did not
have as many years of farming experience, averaging
18.2 years compared to 20.0 for dairy and 20.6 for
beef cattle farmers.

Within the individual type farms, the younger
farmers had more education than Older ones.
Although the cash-crop farmers were younger they
had only an average of 9.4 years of education com-
pared to 9.4 for dairy and 10.6 for beef cattle
farmers.

The dairy farmers had the lowest average off-farm
income with $488 compared to $1,724 for cash—crop

and $611 for beef cattle farmers.

 

us—
J‘s

10.

-93-
The beef cattle farmers had the largest average
gross farm income per farm and per acre, by a large
margin, with $37,513 per farm and $175.70 per acre
compared to only $9,119 and $67.80 for cash-crop
and $11,748 and $54.90 per farm and per acre
reapectively for dairy farmers.

The beef cattle farmers also had the largest

‘ average net farm income with $5,325 per farm and

$25.00 per acre compared to only $2,649 and $19.69
for cash-crop farmers and $3,473 and $16.23 per
farm and per acre reSpectively for dairy farmers.
There was not as much difference in total income

as in other incomes because the cash-crop farmer
had more off-farm income and other income, such as
wives' salaries, interest payments, dividends, etc.,
than the dairy or beef cattle farmers. Thus, the
cash—crop farmers‘ average total income was $4,392
per farm and $32.60 per acre compared to $4,181

and $19.56 for dairy and $6,179 and $28.92 per farm
and per acre, reSpectively, for beef cattle farmers.
The beef cattle farmers had the highest credit
rating with an average of 3.8 (out of a possible
4.0) compared to a 3.4 for cash-crop and 3.5 for
dairy farmers.

Cash-crop farmers had the lowest net worth, with an

average of $46,133 compared to $58,566 for dairy

-94-
and $78,713 for beef cattle farmers.

11. Based on their estimates, beef cattle farmers could
borrow an average of $111,980 with an average
interest rate of 6.1 compared to cash-crop farmers
who could borrow an average of $54,203 with an
average interest rate of 6.4 or dairy farmers who
could borrow an average of $53,374 with an interest
rate of 6.1.

12. One of the most significant differences among the
types of farms was in the percentage Of net worth
they estimated they could borrow. These percentages
were: beef cattle farmers--l42.3 per cent, cash-
crop farmers--ll7.5 per cent and dairy farmers--
91.3 per cent.

Selecting the BeSt Fitting Equations

Several equations were fitted to the data for the indi-
vidual type farms and to the data of all farms. *First,
equations were fitted using the quantity of credit as the
dependent variable and the other factors, interest rate, net
worth, credit rating, gross farm income, net farm income,
size of farm, age, and education, in different combinations
as the independent variables. Land contracts were not con-
sidered in the first and second fits but were added as one
of the independent variables in the equations for the third
fit.

The ”best” equation was selected from the equations

-95-
fitted for each type of farm and for all farms. Things con-
sidered in selecting the ”best“ fit were, (1) the adjusted
multiple correlation coefficient and multiple determination
coefficient, (2) standard error of estimate, (3) number of
variables whose coefficients were significant and level of
significance, (4) sign and magnitude of coefficients in
relation to theoretical consideration and related facts,
(5) simple intercorrelations and (6) the distribution of
unexplained residuals.

The best fitting equation for cash-crop farms was
quantity of credit (Y) = f [Interest rate (X1), net worth
(X2), credit rating (X3), gross farm income (X4), land
contract (X9);7

with the following results:

Y = -38,796.51 + 4,716.54X1 + .23X2 + 9,319.39X3 +

(l,136.97)*** (.07)** (1,669.56)***
1.22X4 + .31X9
(.29)*** (,05)***

R’= .82,R'2 = .67, standard error of estimate $12,685

The best fitting equation for dairy farms was quantity
of credit (Y) = r [interest rate (x1), net worth (x2), gross
farm income (X4), age (X7), education (X8), land contract
(X9 L]

with the following results:

Y = 15,288.64 + 6,836.42X1 + .79X2 - .14X4 - 265.18X7

((2,033.89)** (.04)*** (.28) (101.87)**

-96-
+ 4,436.02x8 + .37x9
(502.96)*** (.10)***
R’= .93,}?2 = .86, standard error of-estimate $11,876
The best fitting equation for beef cattle farms was
quantity of credit (Y) = f [Interest rate (X1), net worth
(X2), credit rating (X3), net farm income (X5), land con-
tract (x91J7
with the following results:
Y = -121,550.88 + 16,747.32X1 + .84:X2 + 13,713.64X3 +
(3,997.29)*** (.05)*** (4,835.85)**
.007X5 + .29X9
(.43) (.06)***
R’= .89,'R2 = .79, standard error of estimate $26,245
The best fitting equation to the data of all farms was
quantity of credit (Y) = f [interest rate (X1), net worth (X2),
credit rating (X3), gross farm income (X4), land contract
(X9), cash-crop farm (X10), dairy farm.(X11), beef cattle
farm (X12);7
with the following results:
Y = -50,191.14 + 8,310.01X1 + .75x2 + 3,437.75}:3 +.24x4
(1,341.05)*** (.03)***(1,467.60)* (.07)*
+ .29x9 - 3,757.72x10 - 10,999.35):11 + 13,821.12X12
(.03)*** NC NC NC

2 = .82, standard error of estimate $20,139.

fi= e91, E
Summary of Factors Affecting Quantity of Credit

As previously pointed out the various regression

-97-
equations were fitted with quantity of credit as the dependent
variable. This was done to derive a best fitting equation
for variables affecting quantity of credit.

The basis for specifying the order of importance of these
factors were: (1) the number of final best fitting equations
which included this factor, (2) the value of the estimated
regression coefficient for this factor, (3) the significance
level of this factor as determined by the 't' test used and
(4) simple correlations with other factors.

The regression equations selected as the best fitting for
individual farm types,(i.e., cash-crop, dairy and beef cattle)
indicated the following.

1. Net worth was the most important variable for all

types of farmers. The respective coefficients of
.23, .79 and .84 for cash-crop, dairy and beef cattle
farmers indicated that net worth was more important
to dairy and beef cattle farmers than to cash-crop
farmers. The ”t” test of 3.30, 17.87 and 18.80
indicated that net worth significantly affected

the quantity of credit for all three types of
farmers.

One of the probable reasons for the low regres-
sion coefficient for net worth for cash-crop farmers
was the simple correlation of .73 between net worth
and gross farm income for these farmers. Part of

the effect of net worth was probably reflected in

2.

-98-
the coefficient of gross farm income, which was
1.22. On the other hand, part of the effect of
gross farm income (coefficient of -.l4) was
probably reflected in the coefficient of net worth
for dairy farmers. The simple correlation between
these two factors for dairy farmers was .71. By
the same reasoning, part of the effect of net farm
income (coefficient of .007) was probably reflected
in the coefficient of net worth for beef cattle
farmers although the simple correlation between net
worth and net farm income was only .25 for beef
cattle farms.

Interest rate was second in order of importance as
rated on the above basis. The respective 't' test
of 4.15, 3.36 and 4.19 for the coefficients of this
factor for cash-crop, dairy and beef cattle farmers
indicated it was highly significant for all three
types of farmers. The coefficients 4,716.54,
6,836.42 and 16,747.32, respectively, for cash-
crop, dairy and beef cattle farmers indicated that
interest rate has a greater affect on the availa-
bility of credit for beef cattle farmers than for
either dairy or cash-crop farmers.

There were not any other independent variables
highly correlated with interest rate. Land con-

tracts had the highest correlation with -.36, -.31

 

‘dh'

 

 

 

In

3.

-99-
and -.38 for cash-crop, dairy and beef cattle farmers,
respectively.

Land contracts or the availability of land contracts
rated a very close third in importance. This factor
was highly significant as indicated by ”t' tests of
6.32, 3.57 and 4.69 for cash-crop, dairy and beef
cattle farmers. There was not much difference in
the coefficients for the different type farms which
were .31 for cash-crop, .37 for dairy and .29 for
beef cattle farmers.

The simple correlation of .49 between net farm
income and land contracts for beef cattle farmers
indicated a possibibity that the regression coeffi-
cient for land contract did reflect some of the
effects of net farm income, (whose coefficient was
.007).

Credit rating was included in the best fitting equa-
tions for cash-crop and beef cattle farmers. It

was significant in both of these equations with ”t”
test of 5.58 for cash-crop and 2.84 for beef cattle
farmers. The respective values of the coefficients
of 9,319.39 and 13,713.64 indicated that credit
rating had greater influence on the estimated quan-
tity of credit for beef cattle farmers than for
cash-crop farmers.

There was not much indication that the values of

-100-
these coefficients were affected by other factors.
There were not any very high simple correlations
between credit rating and other factors. The
highest were .40 and .39 with gross farm income
for beef cattle and cash-crop farmers reapectively.

After these four factors determining the degree of
importance becomes more difficult; therefore, the other
factors will not be assigned an order of importance.

Gross farm income was included in two Of the final equa-
tions, cash-crop and dairy cattle. Its regression coefficient
was highly significant for cash-crop farmers ('t' test of 4.19)
but was not significant for dairy farmers ("t” test .51). As
pointed out above this factor was highly correlated with net
worth. Probably the regression coefficient of 1.22 for cash-
crop farmers and -.14 for dairy farmers reflected some
effects of net worth.

Net farm income, age, and education were the other varia-
bles included in at least one final equation. Each was
included in only one final equation. Net farm income was in
the final equation for beef cattle with a coefficient of
.007 and a 't' test of .02. It is readily apparent that it
was not statistically significant. The effects of this factor
were probably reflected in the coefficients of land contracts
(intercorrelation .49) or net worth (intercorrelation of .25).

Age and education were included in the final equation

for dairy farmers. Both of these factors were significant as

 

 

 

 

-101-
indicated by “t” test of 2.60 for age and 8.82 for education.
The value of their coefficients indicated education had a
greater effect on quantity of credit than age. These values
were ~265.18 for age and -4,436.02 for education. As pointed
out in the discussion of the acceptability of this equation,
the negative sign for education seems to be contrary to
logic. Though the simple correlations between either of
these factors and other factors were not large (the highest
was a -.32 between age and education), the regression coeffi-
cient for education may reflect some of the negative effects
of age.
Summary of Factors Affectipg Interest Rate

This summary of factors affecting the interest rate is
based on equations which treat interest rate as a function of
a list of independent variables. The list of variables was
secured from the best fitting regression equations treating
quantity of credit as the dependent variable. Therefore, some
variables which might have had statistically significant
effects on interest rate could have been eliminated during the
process. of revising the equations. The variables considered
were: quantity of credit, net worth, credit rating, land
contracts, gross farm income, net farm income, age and educa-
tion.

When the regression equations were fitted with interest
rate as the dependent variable, the results indicated that

all the factors did not significantly affect interest rate.

~102-
For example: The regression coefficients for age, education,
net farm income and gross farm income indicated these factors
had very little effect on interest rate. Quantity of credit,
net worth and availability of land contracts were the factors
that had significant effects on interest rate for all farms.
Credit rating had a statistically significant effect on
interest rate for cash-crop farmers but not for beef cattle
farmers.

Quantity of credit seemed to have the greatest effect of
all the factors considered. This factor had a greater effect
on the interest rate for cash-crop farmers than for either
dairy or for beef cattle farmers. It had a lesser effect
upon beef cattle farmers than for dairy farmers. These
effects are indicated by their respective marginal factor
cost functions.

Marginal FaCtor Cost FunctiOns

A marginal factor cost schedule of credit was derived
for each type of farm in Chapter II by varying the quantity
of credit and using mean values1 (for the individual type of
farm) of the other variables in the equation. The equations
for marginal factor cost were used in computing these

d

schedules were MFG = a + 2le + b7X. or a + 3b Q? +
d i=2 i 1 1

bin1 depending upon type of function fitted.
i=2

 

1For mean values see Appendix B, Table l.

—4-"'-

 

iflbwnj

 

-lO3-
The marginal factor cost function for cash-crop farmers
was MFC = 8.05 + (3) .00000000023xf - .000018X2 - .54x3 +
.000023X4 - .000025X9 (X1 = quantity of credit).
The marginal factor cost function for dairy farmers was

MFG = 5.41 + (2) .OOOOIOX1 - .0000086X + .OOOOOOIZX4 +

2

.0015X - .000017X9 (X1 = Q - quantity of credit).

7
The marginal factor cost function for beef cattle farmers
was MFG = 5.86 + (2) .000058X1 - .OOOOO3l‘X2 - .081'X3 -

.0000063X - .0000065X9 (X1 = Q,= quantity of credit).

5
These marginal Cost functions were plotted on one chart,

Figure 6, to get a better comparison of the functions for the

different type farms. It is readily apparent that the average

dairy farmer studied thought he had to pay higher interest

rates for quantities of credit up to $35,000 than either the

average cash-crop or beef cattle farmer. On the other hand,

the average beef cattle farmer studied estimated he could get

credit at a lower rate of interest than the other two types

of farmers after a small quantity has been secured. This may

be due to several reasons. (1) The average cattle farmer

as pointed out before had a much larger gross farm income

than the average dairy or cash-crop farmer studied. (2)

Feeder cattle offer a fairly risk free enterprise from the

lender's viewpoint. (3) Beef cattle farmers were better

credit risk men. A high percentage of the beef cattle

farmers had an "excellent" (the highest) credit rating.

Cash-crop farmers estimated they had to pay the same

-104-
interest rate as beef cattle farmers up to a quantity of
about $20,000. Thereafter the cash-crop farmer's interest
rate increased faster than either dairy or beef cattle
farmers. This may be due to several reasons. (1) The
curvilinear function used permitted the interest rate to
increase much faster than a straight line function. (2)
Type of assets owned by the cash-crop farmers were different
from those owned by other type farmers. A fairly high per-
centage of the net worth of cash-crop farmers is in the form
of machinery and equipment which is not considered as good
for security as dairy cows or beef cattle. (3) A greater
fluctuation in gross and net farm income for cash-crOp
farmers as compared to dairy or beef cattle farmers.

These marginal cost functions are reliable only for the
range of data for the reSpective types of farms. The range
of data for quantity of credit was up to $115,000 for cash-
crop farmers, $156,500 for dairy farmers and $239,500 for
beef cattle farmers. Although the range of data for cash-
crop was up to $115,000, the curvilinear function is probably
unreliable for this amount. The interest rate begins to
increase very rapidly after about $50,000 -- probably too
fast, even though this function fitted the data better than
a linear function over most of the relevant range.

Quantity of credit to be borrowed had the greatest
effect upon the interest rate of all the variables considered

for all the individual type farms and for all farms combined.

~105-
Quantity of credit had a greater effect on the marginal
factor cost of credit for cash-crop farmers than either dairy
or beef cattle farmers. Not worth and availability of land
contracts are the only other variables affecting the interest
rate to any significant degree. The regression coefficients
for these factors indicated they also affected the cash-crop
farmers more than either dairy or beef cattle farmers.

After the marginal factor cost functions for credit were
obtained as discussed above another set was derived by using
the grand means2 (mean of data for all farms) of the other
variables instead of the means for each type of farm. The
functions eliminate differences among types of farms due to
different values of the independent variables. These func-
tions were computed and are illustrated in Figure 7. As it
can be readily seen there were not any significant differences
in the marginal factor cost functions at the higher quantities
of credit when computed using the type of farm.msans for the
variables, not worth, credit rating, not farm.income, etc.,
than when computed using the grand mean of these variables.

it the lower quantities (actually the more meaningful
quantities) of credit there were some differences. ‘For
example, from Figure 7 where marginal factor cost functions
were computed with.the grand mean values, a loan of $25,000

would cost all types of farmers an interest rate of

 

3For grand means see Appendix B, Table l.

-106-

.eaasm came Homew>aesfi now some seam seaweed .maosasm

oaaoco use: use madam .aoaennaso new usofiaessm once seesaw Homewas:

as u a

z Anessasesoo caeoao as assessed

.o seamen

com cad owH era and and OVA and ONH 6.: OOH Om ow Ob cc om ov a»... on OH O

 

 

I Q ‘I I. ‘

maosash oaaoce moon

opossum mason

maesash measlnnco

m.¢

: o.m

0 l0
0 O
0 l0

.m.©
. o.b
, m.b
. c.m
. m.w
. c.m
. m.a
. o.OH

. n.0H

 

 

comm auoaeasH

.mosas>
some endow wean: mason came Hcsow>fldsz new case Beam eo>Haoe .naoaacm
oaaoso moon was wages .aoaousmse new nacwoessa once aooeom Homemasz .b oasmfim

as u as Ana-senses: :8vo as $3525

OON OOH OOH On...” OOH OOH OOH OOH ONH OHH OOH OO Ow Ob OO Om Ouv Om ON OH O

‘1 d d W C d I ‘ 1 J i d

 

m.¢
O.m
m.m
.0.0

.m.O
naeaasm oavvso poem

-107-

.O.b
.m.b
I maesasm madam AO.w
.m.w
.O.m
m.m
jOOH

maoaaom mesonsmeo .m.OH

 

 

 

ovum emeaevsH

mu."

 

-108-

approximately 5.5 per cent. Whereas, the marginal factor
cost functions of Figure 6 indicate a loan of $25,000 would
cost cash-crop farmers 5.5 per cent, dairy farmers 5.8 per
cent and beef cattle farmers 5.4 per cent. By using the
grand means, the beef cattle farmers‘ interest rates were
decreased and dairy farmers' interest rates were increased.
The marginal factor cost curves that would best represent the
average cash-crep, dairy or beef cattle farm studied are in
Figure 6 where the means of each particular farm type were
used. The mathematical marginal factor cost functions can be
used with individual quantities for the variables, net worth,
gross farm income, etc., to fit a particular farm, to compute
either the quantity of credit forthcoming at a certain price
or the interest rate payable for a certain quantity of credit.

One other marginal factor cost function was computed.
This was from the data for all farms to obtain a general
function with the types of farm included as independent
variables. The marginal factor cost function based on the
data for all farms was MFC = 6.64 + (2) .0000090X .-

l
.0000063X2 - .27.X3 - .00000014X4 - .0000098X9 + .07OX10 +
.0079X11 - .071X12. (X1 = Q = quantity of credit). The
marginal factor cost schedules were computed by varying
the quantity of credit (X1), using the grand means for X2
through X9 while using X10 = l; X11 = O; X12 = O for cash-
crop farmers, X10 = 0;.X11 a 1;.X12 = 0 for dairy farmers

and X10 = 0; X11 = O; X12 = l for beef cattle farmers. These

~109-
were computed and are illustrated in Figure 5, Chapter III.
Based on these results as indicated by Figure 5, there was
only a small degree of difference in the marginal factor cost
function for the three different types of farmers. For
example, for a $15,000 loan the cash-crop farmer's cost would
be 5.5 per cent, dairy farmer's cost 5.4 per cent and 5.3
per cent for beef cattle farmers. Of course, the biggest
objection to the use of this function to represent a parti-
cular type of farm is that this difference of approximately
.1 per cent remains constant throughout the function. This
is very different to what was indicated by the marginal
factor cost functions for the particular types of farms,
Figure 6 and Figure 7. Therefore, this function would not
be useful to represent the individual types of farms.

Based on the various marginal factor cost functions
computed and schedules plotted the marginal factor cost
functions of credit computed by using individual type farm
data and mean values for the particular type farm represent
'the cash-crop farmer's, dairy farmer's and beef cattle
farmer's marginal factor cost functions of credit better
than any other function computed. These marginal factor
cost schedules are illustrated in Figure 6.

Possible Implications of Results

As pointed out in the introduction, this study was not

designed as a problem solving type study but as a descriptive

study. The primary objective was to derive and describe the

/O

 

 

 

~110-
marginal factor cost function for credit for Michigan‘s cash-
crop, dairy and beef cattle farmers. Though the results do
not solve any problem there are several areas in which.they
may be used. The two primary uses of the results are in the
areas of:

l. (a) Research work, especially in budgeting and
programming where the quantity of credit available
to a farmer to carry out certain projects or adjust-
ments on the farm is required.

(b) Supply response work where there is a need to
knew the quantity of credit farmers think they
have available to use in estimating their produc-
tion of farm commodities.

2. Extension work with farmers in the field of
credit.

There have been numerous studies made in the past, a
few of these were mentioned briefly in the introduction, which
have required.some kind of estimate as to the quantity of
credit available to a farm or a farmer under a Specified set
of circumstances. In the past, almost all of the researchers
carrying out these types of studies had to assume a set of
rules or guides which would determine the quantity of credit
available to the farm or farmer. The mathematical marginal
factor cost functions derived in this study can be used to
estimate the quantity of credit available, with the researcher
having to estimate or know only the interest rate the farmer

~111-
can pay or is willing to pay for credit. The value of the
other variables such as net worth, gross farm income, not
farm income, etc., will probably be known by the researcher.
If not, then the marginal factor cost function with the mean
of the particular type farm that he is working with can be
used to estimate the quantity of credit such a farm could
regard as available on the capital market.

These results can be used by extension economists work-
ing with farmers in the field of credit. For example, it
could be pointed out that farmers think credit is available
at a cheaper rate for use in dairying than for use in cash-
crop farming or at an even cheaper rate for use in buying
feeder cattle and feeding them out to market weights. Also
that it might be easier for the cash-crop farmer to expand.by
adding livestock to his program rather than expanding through
more acres of cash-crops based on the cost of credit available
for these additional enterprises. The variables used in making
estimates have to be adjusted to fit a particular farmer;
even then the functions Should not be expected to estimate
the exact amount of credit a farmer could borrow on the capi-
tal market. In addition to substantial standard errors of
estimate, it must be stressed that the estimate is of what
farmers think they can borrow not of what they can actually
borrow. These marginal factor cost functions should serve,
however, as broad outlines as to the quantity of credit a

particular type farmer or a particular farmer can borrow.

~112-

The results of this study should benefit extension
economists by giving them.a clearer picture of the credit
market the different types of farmers think they face. The
knowledge of the type of marginal factor cost function the
different types of farmers think they face should be of use
to extension personnel working in the field of farm credit.
These functions can be applied to the situation of a begin-
ning farmer or a person wanting to start farming (with
modifications) to get an estimate of the quantity of credit
he could borrow to commence farming based on his present
thinking, not worth, credit rating, estimated gross farm
income and other factors.

Another important way these results can be used by
extension economists is to give farmers an indication of the
importance of their credit rating. The results of this
study indicated that a farmer‘s credit rating is very impor-
tant not only in aiding the farmer in getting a loan but
also in determining the size-of that loan. Based on the
estimated coefficients of $9,319.39 for cash-crop farmers
and $13,713.64 for beef cattle farmers, the value of the
farmer‘s credit rating is very high. These coefficients
indicate that as a farmer's credit rating changes from one
classification to another, i.e., from 'fair' to 'good' or
from 'good' to "excellent”, it increases the quantity of
credit available by approximately $10,000. This informa-
tion should be very useful in stressing to farmers the

-ll3—
importance of keeping an unblemished credit rating.

Some secondary uses of these marginal factor cost
functions might be:

1. In the teaching field

2. As an aid to the various lending agencies.

These marginal factor cost functions and.the method
used to derive them could be used in the classroom to illus-
trate how marginal factor cost functions can be derived from
farm data and.also to make realistic comparisons among the
three different types of farms.

Various lending agencies probably would be interested
in what farmers think concerning the quantity of credit
available to them.ander the present policies of these lending
agencies. If these marginal factor cost functions based on
farmerst estimates of the quantity of credit available to them
varies to a significant degree from what the lenders are will-
ing to loan, this would indicate the farmers are not well
informed as to the policies of the lending agencies. Of
course these functions include farmers‘ estimates of the
value of land contracts available to them.on a credit basis
over which established lending agencies have little control;
therefore, these would not be expected to coincide too
closely with marginal factor cost functions derived from
the lender's viewpoint.

Another study of marginal factor cost functions for
credit, with the data collected from lending agencies might

“fl”;.-—w’

 

 

 

 

-ll4-
be worthwhile. .As mentioned above some allowances would
have to be made for land contracts and other sources that
are not controlled by the regular established lending agen-
cies. .A study of this kind should be valuable for two
reasons:

1. The marginal factor cost function derived from the
lender's viewpoint would serve as a check on the
functiOns derived in this study.

2. If there was much difference in the two methods,
this would indicate that farmers are not well
informed concerning their credit possibilities and
that more information should be extended to
farmers, through the various organizations, on
the policies of the different lending agencies.

Other studies should give more attention to different

kinds of collateral (closely related to net worth). At
least three categories of collateral are important, (1) land
and buildings, (2) livestock (dairy cows, feeder cattle,
etc.), and (3) maChinery and equipment. It is likely that
these could be used in a general marginal factor cost func-
tion for a combination of different types of farms, probably
without a type of farm.variable, to produce improved results
for individual type farms.

‘w-- -

 

 

 

BIBLIOGRAPHY

BIBLIOGRAPHY

Bell, Sidney 0., Financing‘Dairy'Production Adjustments in

the Tennessee Va 9: a o . a ama, unpub s ed . .
es s, urn n vers ty, 1 5 .

Bolger, Ross, A.Partial Evaluation of the Michigan Township
Extension ram. n enmar owns or he. or od
I953 to I958, fising Cobb-FEEglas sIEalysis, finpubIIshed
E. S. TEesIs, Depar ment of r c ural conomics,

Michigan State University, 1959.

Baadford, Lawrence A. and Johnson, Glenn In, Farm Management
Analysis, John Wiley and Sons, New York,.19

Brake, John Rm, Financial Seasonality of DairyiFarming and
its Relation to sk and ncerta t , unpubl shed . .
TEesIs, 56 artment of IErIcultural Economics, Michigan

State University, 1956.

Brake, John 11., Prediction'of Fertilizer Cong? tion in Two

Regions of the n ted ates, . . es s, ort .
arol a ta e o ege, 959. .

Dvorak, Frank E., neg ramming the. Organization and Capital

‘Use for a Gas -crop arm 6 ag aw a1 oi an
um ea 0 c ' an, unpu sped . . Thes s,

EIcfiIgan SIaIe UnIversity, 1959. .

Edwards, Clark, ResOurce Fixity, Credit AvailabilitE and
agriculture r an za on, unpu s e . . es s,

c [ HI I 6 I If _ .

gan a e n vers y, 1958.

 

 

Hildebrand, Peter E.,‘Farm.0rganization and.nesource'Fixity:
Medification of the ear gra g Mbdel, unpu shed
. . es s, c gan ta 9 Un vers ty, 959.
Hill, Elton B. and Mawby,Russell 6., Types of Figgigg in
Michigan, Special Bulletin 206, secon e on ,
optember 1954. . ,

Johnson, D. Gale, "The Nature of the Supply Function for
Agricultural‘Products,‘ American Economic Review, Vbl.
40, September 1950. -

Johnson, D. Gale, Ferward.Prices for agriculture, Chicago,
The University of cage ss, 4 .

-
o
O
-.4
--
-

 

 

-117-

McKee, Dean E. , Economic Aggraisal of Adjustments in 139.15;-
inch ano e ann n ons,wo
current progress at ch gan tate Un vers ty.
Mic Statistical Abstrac -- Bureau of Business and

cono c Researc , gan State University, Second
Edition, 1958.

IPickler, Eugene B., Use andAvailabilit; of Credit in
Northern Michi an arms, unpu l shed . . es s,
['g'an State UH I .

FREE vers ty, 1959.

 

Smith, Victor E., ”Perfect vs Discontinuous Input Markets;
A Linear Programing Analysis ," Journal of Farm
Economics, Vol. 37, August, 1955.

Trant, Gerald 1., Institutional credit and.The Efficiency
. of Selected Da rg Farms, unpu e . . es s,
0 an a e

vers ty, 1959.

Walker, Helen M. and Lev, Joseph, Statistical Inference
Henry Holt and Company, Inc. , 1555.

soil-F“. a

 

 

 

 

MD]!

APPENDIX A
Procedure1 for securing regression coefficients for
the three dummy variables in the regression equation fitted
to data from all farms. (See pp. 83 to 84). Estimated
values for b1 : i - 1, . . . , 9 were secured by the method
of least squares; 1’10 and bll were also estimated by least
squares with 112 omitted. The problem is to obtain 1’12 and

to modify the estimtes of b 0 and bll accordingly. The

l
dummy variables in this instance are 110, 111, and 112, which

are identified as 11 , 12 and 13 in the following:

Type 1 farms (cash-crop farms)

m

+231“;

Y
91 1‘4 11 1

1""B

Type 2 farms (dairy farms)
_ , d .
Y2 - 302 + 1E4 3111 + “2
Type 3 fans (beef cattle farms)
d _,
Y3 -= 1303 + 134 31x1 + :13
For estimating, combine into one equation

(1

 

1This procedure was formulated by Professor R. 1..
Gustafson, at the suggestions of John Brake and Glenn L.
Johnson, Department of Agricultural Economics, Michigan State
University.

-120-

wherexl-liffarmistypel
0 otherwise

x -liffarmistype2
2 0 otherwise

Estimate 0f B03 - b0 Estimate 0f B03 3 b0
. . (B02 - B03) - b2 . . B02 - b0 + b2

‘5 c‘ ‘

Define a new parameter, namely

NB +NB +NB
a 1 01‘ '2 02‘ “3'03‘
Boa N {N1+N2+N3-N)

 

N1(Est. 1301) + N§(Est. B an N3(Est. B03)

Estimate of B" - A;

0

N (b + b1)‘+ N2(b + b ) + Nabo
-L—Q-—-r, ‘LL—N

, (N14- N2‘+’N3)bo'+‘N1bJ' '+'N2bg
’ N

N N -
b0 + 1 b1 + N2 b2
Also, define new coefficients of 1.1, X2 and 13 in the over-

all equation
d

2B1 +u
1-414

* *
Y-Bo+31x1+13§x2+3313+
so that the resulting equation is equivalent to the original
set of three.
i.e., we want

3 *
BO + B1 1- B01

 

O
. e
H 4 K 1
I
w . n
a .o 7
. y .
\j I C
.
u
. o
. ..
— ~ . _ . ./.
. . .. . i .
. . . . e .
I ../
* o o .
-\- . - $
/\ , \ s. v u o .
a i o -
fl .
. o a .
. . _ .
i . . ,
, c
.. n . I . L .
. . . t
. \
. o .
. e . . I
. .\ 1
. . o
. . . v
o _
I
/
Ox
. o .
. 1
-. l o .. .
. v c . .
, . o . .1
In '. a
-H . .
I \. A
. .
. . . _ p
u \ o o o
. . .
. . r 1
/ . . -
v . ..
j
.
.
. o
n I
.
.. p
. .
r \
n
c
:\

 

Erom

-121-

a *_
BO + B2 B02

3 *

this, we have

all
0

N1

Estimate of B; n est. 301 - est. B

be +b1

N
.. 2 1:
_(_b0 4’ +fi- 2?
g ' N‘ N
(1 - filhl .. fighz
Estimate of B; - est. B02 - est. B;
N1 N
abo-i-bz '(bo+rb1 +fi2b2)
N1
9 - N" b1 + {1 $1132
Estimte of B; :- est. B03 - est. 3;

“1
”N"

N N
-'fi!"b1-fi'§b2

bo - computed a in original equation

N
=bo- (bo-l- b1+ޤb2)

b1 - computed coefficient of 11 in original equation
b2 .. computed coefficient of 12 in original equation
3; - derived a value

B - modified coefficient of 11

B; - modified coefficient of 12

B3 a derived coefficient of 13

APPENDIX TABLE 1.

APPENDIXB

The mean values of the Factors Used
in the Marginal Factor Cost Schedules

for Cash-crop, Dairy and Beef Cattle
Farmers

 

 

 

Land contracts

 

 

 

 

Independent shpcropfifieiry so a e ran

variable farms farms farms mean

Net worth 46,759.18 58,592.01 79,661.76 62,220.06

Credit rating 3.42 3.50 3.81 3.59

Gross farm 9,278.48 11,959.46 37,922.36 20,208.58
income

Net farm 2,713.92 4,161.74 5,380.61 2,844.02
income

Age 43.99 46.87 46.12 45.69

Education 9.37 9.48 10.56 9.82

18,746.75 4,795.03 26,088.24 16,688.66

 

 

---..-_

no-‘-

r-voa-
.

-'. a

o---—

j-o‘.l'l.

til--|

.
o
n o
q
o
v 4
.
. .
.
n
.
.
Ole--‘..r! ‘0 O ...‘..I O‘»‘I'I.I.'llt“‘-‘
.
_
0
Us Os es
.
a .
O
a
C O
A
e e e _
.
o o
_
.1900-.ot.00u,. tll-n\|.-‘-.!|90f
.
.
0\ Ca ‘
o
. l
..
.
O O
O C O
o
u o
o

D\ I\ O\
O
C C
a
. O O
.
a
‘
‘
| I | | ' - i U ’ - 0 IV - 'l - o. - 'I - | O O ’ 11
m
.
Ix O\ Q.
, .
C
c
o e .
.
O C .
_
a

mu C

 

 

CONFIDENTIAL Schedule No.
, Telephone No.

Name

Address

 

1. Farm Size

1 '

 

‘ fined en ed ota

Tillable

 

. - .......

Non-tillable open

 

woodland and other

 

Total

 

 

 

 

 

Tillable acreage leased out ‘-° Net tillable acres _

2. How many years have you operated a farm? ‘ .1 ‘ ‘ ' years.

a. Familyicharag'tgg‘ Egg; , q .. A- ,
bay member ,Ige ,Sex. chfion Month! grindd arm or
.' ggade men s vs no months ,a u

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Value of farm labor other than operators

 

4. Did you and your family have any income last year from
non—farm source such as: Sale of products from land rented
out , cash rent, boarders, old age assistance, pensions,
veteran's allowances, unemployment compensation, interest,
dividends, or help from non-resident members of the
family?

Ye s No Total annual amount

.04

—.
o

i . .
. . o
a .
- .

O

"O“-‘I..

c u ' 0 .0 I all
. _
. .
p
. _ .
o o n
. . .
p _ .
. .

o e
. . .

. o
. . .

. .

c .

. . .

. ., o
m

.

o a
. .
. a
.
. . a
. .
006l!rltu
. 1 .
.
. _ .
. . ,
. a
. .
.
4'0! 1.
. .
.
c .
. .
. o .

5.

(To enumerater:

-124-

To determine your estimated farm income
1958

CDSHZFIRMCRECEIPTS
Cattle : Dairy ' (culls,
selves, breeding stock,
.120 e)

4 Beef
Swine
Poultry

Sheep, wool and wool
payments

Other livestock

‘Dairy products

Eggs

Other produce raised:
Grain, hay, other crops
Fruits and vegetables
Forest products
Soy beans, beans, etc.

Machine work off farm

Agricultural payments

Machinery Sales

Other cash farm.receipts

Total cash farm receipts

IPrice

l o ‘ .
) u

I c 1 .
u o

O
l ' I I I
O
I

I O t D I
y I

I t 4 v n
v

I . O .

.....

- g - - -

value

P .
O 0
I e
O

O
I
D
O

s c v. Q a

1959

Price

0 I I I
0 I n n
l ' I
0 e y
f
0 g A p .
l . , .
I I 0

Value

‘ I-

got totals only if you cannot get individual
items for each year.)

Iv—ri

‘

. . . . . _
. . . .
. . . .
. . . . _

. _ . o

.

. . . . .
. . . . . a
. . . . c
. c c . .
a . . . .
a a . - . .
. . o .
. . — .
. . . . . a
. . . . . .
c o o . . . .
. . . . .
. . o o

. . . . .

 

 

b.

-125-

CASH EXPENSES:
Hired labor

‘Feed.purchased

Seeds, plants, spray materials

Machine hire
Supplies purchased
Repair and Maintenance:
Machinery (overhauls,
tires, etc.)
Improvements r
‘Livestock expense except
- poultry
Poultry purchased
Fertilizer including lime
Gasoline, fuel oil, grease
Taxes on farm property
Insurance on famm property
Electricity and phone
(farm share)
Other cash.farm.expenses
Capital investments:
Dairy cattle purchased
15...: cattle purchased
Hogs purchased
Sheep purchased
Other livestock purchased
Machinery purchased
Farm improvement pur.
Total Cash Farm Expenses
Interest and debt payments
Rent
Totals

1958

Price

 

s I o s
v o e . .
‘ ' I I J ‘ 0 ‘
J I p »
v - . . . A '

D
f
r
I

Q a. - Q.

.....

- a u - .

‘fiev.r-

.....

value

 

\ v .
| a . . I I '
O o
l . - .
s D

.
9 r

e s p
o ' o

' c

I I I , e .
. u 0 . t '
v , t ‘
O ‘ e
s . v s

- g a. m ~

00000

I

1959

Price

 

 

- o
. a .

.....

o a o - c

ooooo

value

 

 

 

- -
a
- u
.- ‘
o o
. a
e n
or-
o .
o .
O ‘
g a
o -
. n
a o
o -
o .
o -
. -
c.
. 4

g o
a o

- u
. a
v o
- o
o

- .
o .
- .
e o
- 1
- -

- -

 

—. -
o .-
o -o
... -
o -

o o o

.- -
. - .
o-o .
I s o

o--

0 o 4
l . -
o -- o

a _.

- o o
o u o
e a -
- o —

v--

ac-
D—Q-

. .
. - .
. -
. o -
n - -
-
. - .
o . -
r— o e
- - m
.- . .
. ~o
— o a
- m e
A _
_ . a

‘0‘

-o -
. V —
o -

 

 

u

w

~126-

0. FIRNHE‘RNINGS SUHHIHY
(to be completed in office) 1958 1959

Cash farm receipts

 

‘Less cash farm expenses

I
¢
‘ I
P

Net cash.farm.income
Net inventory change (plus or minus)
Net farm family income

‘Less family labor (other than
operators)

NETiFIRMCINCGHE

6. Inventory information

Increase Decrease

Ldyesteck (all livestock, dairy 1958
and‘beef cattle, swine, etc.$

 

 

page 5 6 6) 1959 ______D
Machinery equipment (page 8) 1958

" 1959 . __
Feed, seed and fertilizer 1958‘
' 1959 __
Totals _______
' Net 1958 __

1959

-127-

7. Livestock Inventory 1959

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Inventory Increase

 

or Decrease

 

.A" Sufitrapt 9..
Beg.Inventory No. K . . No. No.butchered End. Inv.
d N N V u
Lin 0. Value born bought Sold or died e. a1 e
DAIRY
e Cows
Heifers ggg
- J ...... .4 ..... +--
calves . »
Bulls
BEEF - ..... - -r jro ...... T .- - - - -- ..... - -
COWS __ ' 4*,
Heifers ‘ 1 ~ ‘ L
Feeders + . 1 f
. ...... r _ ....... q . ,
Calves A on ‘
- - -L-- -- -- --J_ - h a,4
Bulls
secs . -- .J ._ _- my - 41, _-
Sows
. ..... i ----..-..r---.-. ....... - ----}---- -- ..-- --
Bears
Pigs ‘ __ T
SHEEP .;_ - - - .............. _
. Ewes .
Rams Ti
lambs g» h T
POULTRY , --, ..... - - -_ _ -
Hens
Roosters
-. ...... --r .......... --- N - - - - -
Broilers
........ ¥- - - ....-..-.---.-.--.-.-.-.
Totals

 

“out

J-

. .

-Il‘a“-:-'|

Itau--0|.o

‘1'-
a

o

o

_
..-|
o

a

. .
o
. .
0'- 0|.
.
c A
_
a
.
u
_ .
a
..-.v
. .
_
.
‘--l l
h
a
o
c
I.-. .1
.
. .
o n

. .ls

-128-

8. livestock Inventory 1958

 

A5571

 

FSEbtract

 

Beg.Inventory

Kind

iDAIRY
Cows

NOo

Value

No...
born

N0. .‘,N0o.

 

bought So 91

or died
i____p_.

No.butchered

NOo

End. Inv.

Value

 

 

Heifers

 

Calves

 

Bulls

. . . e

 

BEEF
Cows

~ . - a . s a. -. o s o —

--------

 

Heifers

. .a ———————

 

- - - Q ‘

 

Feeders

calves

- c.. o Q

Q - o — - - 0 ~ - — § ‘

 

Bulls

p § we . s o w 1 .. ‘ a. - -

 

H068

a o a e .

 

_3ows

Bears

.............

g n. o - - .

 

Pigs

......

. - o a a

s a ‘ < Q

a , a -

 

SHEEP
Ewes

.....

 

Hams

 

ILambs

 

iPOULmHY
Hens

. - a 9 a - \

 

Roostersiih-

 

Broilers“

a a o -. -

a a. o - w

.. .....

 

Totals

 

 

 

. . .. c .

 

 

 

 

 

Inventory Increase

- a g o 0.

~ ‘ t g. a

 

..........

or Decrease

 

 

.
o
e

.

C
o

. v- :1 t
.
‘
a
.
- ...'.O

n
.
_

o
.
o
.

.
O
.
a

It... It-
.
I

o
_

o o a
. . . a
o o
. u .
. . .
-..-!to-t -v-i.a -ur
. o .
o .
. . .
. . o
_
. g c
o . .
. o .
O o .
l0-|0!|ll40‘ttl It!
. . g
_. i .
. a .
h _
L- .‘Y - I? .s'lt’s
. e
o . .
. . o
.—
.
w a
. . .
-D’-"“ "" -o. “
. .
o
. o .
o . .
"r‘ltll‘- » n .
. _
_ c
. . .
. .
.
. a .
. . .
. . .
. o _
a .
o a
. . . .
L. 740' 0". ..!I..
u
.
. g o .
c
s [5-4.1- 1"}1'0 ..
. .
. o u
4 . .
. .
. . o o
o .

. r l

a
'tulv
c
l
.
.
0
tr-
.
v- '

---».-

---!

o o a
..

o.—
moo----~

.
.

o
a .l.
.

o
o
a

.
a .
. .
o

a
POI.
-
. .
. .
.

C

-‘ o--
C
0

-.

 

 

~129-

9. Machinery 6 Equipment Investment

Item

Tractors:

 

,y—r —

.
u
o
.
o
.
.
n
.
I
.
-
.
.
_
a
a
.
.
—
q
.
.
Olt
.
’x
. .

. .
. H

. t 0 0
. . .
H

. . e

- - 't'l.

-"‘I

Q-

.
----.
.

n u - e
v
a
a _
o
.
. . .
. a A
9
. . .
.
< e
h
.
u o
. . .
. . .

a . a
. a .
o .

c . .
. . .
. .
o o
a . .
lla‘tl' o
. . o
o o . c
o a o
. . .
.1"-r.|o
o . .
e . a
a a . .
a g
o
.
a . o
o .
”too”-
. . .
.
. .

a a . .
. a
I! | 9 0o; O
o o .
o o —
o .
a o
. . .
c. | .. El 0.
.
o . .
a . a
a A .

o .
n

- m
a .
. . .

a

.

.
_ .
. n

I V

. c
o v
. . .
.

a o

c .

|\

. a

_ .

. .
. .
o o
a o
. . .
u
a Q
o .
v ' e - e.
. . .
. ¢ .
v
_ . .

O . m

.

. .

a

D. 0+] .|

r a

_ - n

. o a
a

. c

. u

_ . .
. a a
. ~ .
a . .
. .

1:, -ol

I
a .
. .
e
. w
. .
m
.
c. .
. .
g
a .
. a
.
C v
o
‘ .
. g
-
"l?
.
. .
. .
U' 0
_
. .
a r
_ _.
.
. o
o v u
n
.
. .
. .
. .
a o
. .
_
.
a a
. a
.1 ll.
. a
. o
A
.
. .
.
I 9
. a
u e

c o
.
. .
. .
a
a
o. .
o v
on. IA:
.
_ .
a
Q
.
04-0
.
a
O o
a
lelt.n
i .
m
. a
.
.
-Ylt
. .
r
u

-130-

 

Totals -- Ending 1959 inventory
Minus items bought in 1959
plus item sold in 1959
total .. \ -..
Plus 10% of this total - """Efialng 1958

 

 

inventory.

Ending 1958 inventory
Minus item bought in 1958
Plus item sold.in 1958

 

otal
Plus 10% of this total ginning

a be
1958 Inventory
Inventory Increase Decrease

1959 ' 1959

1958 ' ' " " 1958 ‘

 

 

w—r~'

It-

~- 0..

v a u
_ .
a .
. L .
. _ .
. u o .
. o .
. .

. .
v _
p .

/

n .
.

p

o
. n
- . u
. . .
. .
. o

\

. 1

p
| I.s I.
c

. _

.

. c _
. . .
u
o . .

m
_
-1! e I.
.
. . .
. _
. . .
. .
o . .
.
. . _
. o t '
, o
o . .
.. .
c
. . .
J -1.
a -
a
a
o
. 0 e u
. v
A . v
_ .
. . .
. .
a I .. I4.
. o .
. .
.
o u .

-‘o

~131-
10. Major changes in farm organization for 1960 relative

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..... ..

-- 4 ‘ * .-
........... - ‘ -. - s . a - - . .. o - i
..................... - .. ~ » , .

.................
~ - -
vi
...... .. .s o . s s .- -. a s o s e a o a . m -o s
. - ‘4 ~ 0 ‘ - V - h - ‘ . - ~ 0 ‘

 

 

 

 

 

Wei-"

 

 

o . ‘ - e . — e - c — o a - a -
. - . - - - - .. - a o n. . o -
- _ . . . v n . . -
- . . a. 9 ~ - - v a — - ,. o o -
o o c « - o n - - n , - - - a
r . - . . o a - a - e _ - r a
n o . a - o - - c m a . . a
- o -e - - . o . c . w -- -.v-.‘-O

- o - . .- - .v o o - . - ‘9
. - - - c - q o o m .
a- - - -1 o a - - . . .

- - - c. . .. . a - . g o . g -
a -- - a o o - e . - - m a a .A -

 

 

 

..-o
.- o
‘ .-
a v
Q-
-7—
.— -
v -

- .,
c .
- --
a a
. -—
o .-

 

.. .- .-. v-
, .
-0.‘------‘-.o
,‘
.~.-.--~‘-. *- - I
O v--*'..-\-r'-‘-

 

 

 

03” ”snow

 

use consummaH

 

one momma
"nommb

 

 

 

 

 

 

 

 

 

 

 

 

"move: couscous:

 

-132-

L

 

L

 

"move: dohsoom

 

 

1|.

 

1H.

 

"newsmaaoa Ho¢Wn m

 

 

 

Aegean

 

GEO

 

 

 

one
oousasm

 

. final
Hang

"no a apes sesame as _

 

6239:

 

t .hdhom.

 

some
09:.

 

annoas

 

3.5.5
[he coasmm

 

odd:
asown

 

omha

 

Quad ha QZH H<,MHHBHAHdeH-hu nuaoflm .HH

 

I .ll Q

a o -‘

I . ‘ O I
- . I -

0 ‘

I I

I I O

I O t
I ‘

I I I

' I n
p t -
0 ' I
I I

r ..

-‘.

.0! D I .

-.. | - .0 D
C A 0 C -
a O .I C - I

I lo 0 I...

I t O .I' - I
C I U - I |
' -|-- - 01..
-I‘I‘ - - ".
‘l. ‘l. - I.
t t In ‘ II t I
- n u - -I’ ‘
0 I. - e N .
s I | I I 1
t - I ‘ - u I
- O I t e ..

—--‘

f“’

A-
I

L

4--
0
t

‘I. .I-I
*‘.l l-‘
.

4'. c 'ill

.

-

.

c

Q 0|--l. .0

o

'0! §I-OII

.

'.-ll. 0 ‘

.

_

o

.

~

.

'III . . .

‘.- I O I O

-

.

v

”‘0. "I
I... e ‘ U

o
‘ '4 -
I 0 0--
as ~I .
II' II-
.
0 ‘l O
.
O el-‘u
.
.
a
I I I-
.
u I. ll
—
I I. I II
.
.
uloll|o

a

.
I.».
I...

.

_

,~.
.
a.

up.
.
I...
p
.
at

t

o

o
u u.
I

a

,
.xol'

—

e
to.

.
also-

-

.

o

- O
D
U
A
0 n 0
| I
I O
.I a 0
0 I v
I .l I
I O I
I
I I O
l .0
| t.
I I I
t |
C | I
‘ | -
I I ‘i
O
O
t
O
I I O
I I Al
I I II

I. -l

I t .0
I s ’.
1. 0 I
I I .|
' |
0 I .
I. U
| O
. O t
I u
t I I
r. u I
0 I
I l I
e . t
I I
II.- I

I - I
I, .- 0
.n
t I. I.
. )I
‘ I
a
o
t . v
1 I V.
I

'11.
s I
o I O
l | O
o t ‘
. ll ‘
l -‘
o I s
a t
a I
I 0
l s C
Itl .I
I I

o

I!
' I -
n I:

I.
| I

-133-

12. Under your present situation, considering your equity
in your land and other assets, what is the maxim- amunt
of money that you think you could borrow from the following
sources

Interest Terms Conditions
Alt. 7 rate (length or (security,
lean, etc.) other than
a. Federal Land . etc.)
Bank ‘ -

b . Far-er ‘ s Hone

 

“line -
c. Comercial Banks

 

d. Insurance Com-

 

panics
e. Individuals (Land

 

contracts, etc.)

r. Production Credit
Assoc.

g. Machinery dealers

 

h. Live stock dealers

H l I II II

 

i. Fertilizer dealers

 

1. Open accounts (tile
pay-ants, etc.)

 

k e “1101'. '

 

 

“'5

 

.
— o
a .
. -

a -o c o o - _

I o - o o o - c
. o - o
- .7 ,.

o - o - - .—
- . . -
- - - o 4 . a

o o - - e .g
. . . . .
_ _ ~ 9 » >9

-.II—.-a

1-'---. C'-

o . - - .
o . o - o
0 ~ 0 .
0 — -

'0 o -
- - .. -
C o o e .

 

-134-

13. Consider for a few minutes that you are going to borrow

all the money that you can possibly get. Now tell no
$53 details of these loans.

14. Does'the price of land purchased on land contract depend L
on interest rates?

. u - .

yes no

 

 

15. Is there another combination of loans that would give
you more than this amount of money if you were willing
to pay higher interest rates? Details?

....-‘O-.I

 

w—

-135-

16. NET“WORTH STAIEMENT
(As of‘Decelber 31, 1957)

 

Assets

_ Liabilities

 

Land, per acre

Buildings
Machinery (page 8)

Livestock (page 5)
Feed, seed, supplies
Household equipment
Stocks, bonds

9 . n g - ‘ -

Cash on hand

 

Cash in bank
Accounts receivable .-.“L~
TOTAL 3‘”

........

Fara mortgage (page 10) 3

 

Other mortgages (p. 10)
Bank notes (page 10) A
Personal notes 1
Other notes
.Acceunts payable

Taxes, rent, ins. due

 

 

 

 

0ther debts

mm. + 3
Net worth '
TOTAL I ‘

 

17. THREE CREDIT REFERENCES
(Business reference)

- _ a v. - -v. - u _ . ~ I - - -

 

. - . - . - - - ‘ f . .. - a - I. - .1 . - . a .......

 

 

~ .. s Q ~ o. o a a - a - o ‘ o - ~ -

2.

‘-o.-.‘ . . ‘1‘--~ .....

 

 

3.

 

 

 

Is~ao+

 

. . 0 ¢
. 4.- . .-

q - - -

o o o o

-\ - o o

- o . .

.. - - a

- - — . a
o - a o

. - . -
. -- . 1 .

I
o -. - . .

 

. n g L

. v o a
I

.o o . vo

- 4 u

o -
, -
“.-
-
c

,—
..
.—-
I
I
.
«a
--
--
u
-.-

o

. _ .
- - .
-.-< -
o g .
.4 .-A.V
——- -

 

 

o
. c
4 o
— .

 

c
I -
c v
V V n
u-
v .I
. a
u
u
. o
- c
. o
-

-..~..-.

- o
- o
o .
a -
o o
r o
.- -

 

 

ROOM USE OE‘E-LY.

 

"mwwfiflfiflrvauzmmnw