• TABLE OF CONTENTS
HIDE
 Copyright
 Title Page
 Table of Contents
 List of Tables
 Introduction
 Objectives
 Previous research
 Nature of data
 Theoretical and analytical...
 Empirical results
 Milk purchases and expenditures...
 Estimated demand and expenditure...
 Comparison and evaluation of regression...
 Other information on household...
 Summary
 Implications and interpretatio...
 Appendices






Group Title: Agricultural Economics report 19
Title: Household demand and purchasing behavior for fluid milk in Gainesville, Florida
CITATION PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00027478/00001
 Material Information
Title: Household demand and purchasing behavior for fluid milk in Gainesville, Florida
Physical Description: ii, 52 leaves : ; 28 cm.
Language: English
Creator: Prato, Anthony
Publisher: Dept. of Agricultural Economics, University of Florida
Place of Publication: Gainesville Fla
Publication Date: 1971
 Subjects
Subject: Milk consumption -- Florida -- Gainesville   ( lcsh )
Milk -- Prices -- Florida -- Gainesville   ( lcsh )
Milk trade -- Florida -- Gainesville   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
non-fiction   ( marcgt )
 Notes
Statement of Responsibility: Anthony Prato.
General Note: Cover title.
General Note: "March 1971."
Funding: Florida Historical Agriculture and Rural Life
Additional Physical Form: Agricultural economics report - University of Florida Dept. of Agricultural Economics ; no. 19
 Record Information
Bibliographic ID: UF00027478
Volume ID: VID00001
Source Institution: Marston Science Library, George A. Smathers Libraries, University of Florida
Holding Location: Florida Agricultural Experiment Station, Florida Cooperative Extension Service, Florida Department of Agriculture and Consumer Services, and the Engineering and Industrial Experiment Station; Institute for Food and Agricultural Services (IFAS), University of Florida
Rights Management: All rights reserved, Board of Trustees of the University of Florida
Resource Identifier: aleph - 001615904
oclc - 21028349
notis - AHP0349

Table of Contents
    Copyright
        Copyright
    Title Page
        Title
    Table of Contents
        Page i
    List of Tables
        Page ii
    Introduction
        Page 1
        Page 2
    Objectives
        Page 3
    Previous research
        Page 3
    Nature of data
        Page 4
        Page 5
    Theoretical and analytical framework
        Page 6
        Page 7
        Page 8
        Page 9
    Empirical results
        Page 10
        Page 11
        Page 12
    Milk purchases and expenditures for selected households
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
    Estimated demand and expenditure relationships
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
    Comparison and evaluation of regression results
        Page 29
        Page 30
        Page 31
        Page 32
    Other information on household milk purchasing behavior
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
    Summary
        Page 40
        Page 41
    Implications and interpretations
        Page 42
        Page 43
        Page 44
    Appendices
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
Full Text





HISTORIC NOTE


The publications in this collection do
not reflect current scientific knowledge
or recommendations. These texts
represent the historic publishing
record of the Institute for Food and
Agricultural Sciences and should be
used only to trace the historic work of
the Institute and its staff. Current IFAS
research may be found on the
Electronic Data Information Source
(EDIS)

site maintained by the Florida
Cooperative Extension Service.






Copyright 2005, Board of Trustees, University
of Florida





March 1971
'/. J, /c'>


Ag. Econ. Report 19


Household Demand and Purchasing
Behavior for Fluid Milk


in
Gainesville,


Florida


Department of Agricultural Economics
Florida Agricultural Experiment Stations
Institute of Food and Agricultural Sciences
University of Florida, Gainesville


Anthony Prato


-~-T- ~CI~ -- ~e I














TABLE OF CONTENTS


Page

LIST OF TABLES . . . ... . ii

Introduction . . . . .. .. 1

Objectives . . . . .. . . 3

Previous Research. . . . . .. 3

Nature of Data . . . . .. . 4

Theoretical and Analytical Framework . . 6

Empirical Results. . . . . ... .. 10

Socioeconomic Characteristics of Households . .. 10

Milk Purchases and Expenditures for Selected Households. 13

Estimated Demand and Expenditure Relationships . .. 23

Comparison and Evaluation of Regression Results. . 29

Other Information on Household Milk Purchasing Behavior. 33

Summary . . . . . . 40

Implications and Interpretations . . ... 42

APPENDICES . . . . ... ...... 45















LIST OF TABLES


Number Page

1 Selected socioeconomic characteristics: All 309 house-
holds. . . . . ... . 11

2 Average purchases and expenditures on fluid and other
milk products for whole milk users, fluid milk users
and all households: Gainesville, Florida, average
week in May-July, 1970 . . . ... 14

3 Average purchases and expenditures on fluid and other
milk products for six types of households: Gainesville,
Florida, average week in May-July, 1970. . ... 17

4 Changes in average milk purchases and expenditures with
respect to income for small and large households . 20

5 Changes in average milk purchases and expenditures with
respect to household size for low, medium and high
income households. . . . . ... 21

6 Regression coefficients and related statistical values:
Fluid milk demand equation . . .... 24

7 Regression coefficients and related statistical values:
Fluid milk expenditure equation. . . ... 26

8 Regression coefficients and related statistical values:
Whole milk demand equation . . . ... 28

9 Regression coefficients and related statistical values:
Whole milk expenditure equation. . . ... 30

10 Responses of fluid and whole milk users to selected
questions on milk purchasing behavior and filled milk
Gainesville, Florida, May-June, 1970 . ... 35

11 Responses to six types of households to questions on milk
purchasing behavior and filled milk: Gainesville,
Florida, May-June 1970 . . . . 37















HOUSEHOLD DEMAND AND PURCHASING BEHAVIOR
FOR FLUID MILK IN GAINESVILLE, FLORIDA

Anthony Prato*


Introduction

The dairy industry is an important source of revenue and employ-

ment in Florida. In 1969 some 527 commercial dairy farms produced 1559

million pounds of milk and sold 1543 million pounds of milk and cream.

Cash farm receipts from the sale of Florida dairy products totaled $121

million and ranked fourth among cash receipts from agricultural commod-

ities. Substantial revenue and employment were also generated by the

non-farm segments of the Florida dairy industry which, in 1969, included

some 12 bargaining cooperatives, 65 fluid and manufactured milk processing

plants and the distribution facilities of numerous food establishments.

Revenue and employment from Florida milk production, processing

and distribution depend in large part on fluid milk consumption and

expenditures. In 1969, 93 percent of all Florida produced milk sold to

federally regulated plants was utilized as fluid milk. Despite the impor-

tance of fluid milk sales, very little information is presently available

on consumer milk purchasing behavior and consumer demand for fluid milk

in Florida.




Anthony Prato is Assistant Professor of Agricultural Economics,
Institute of Food and Agricultural Sciences, University of Florida.



1Fluid milk includes whole milk, skim milk, low-fat (2 percent)
milk, buttermilk and chocolate milk and drinks.









At least two recent developments illustrate the importance of

consumer demand. One is the shipment to Florida of fluid milk (mainly

whole milk) produced and packaged in Minnesota and Wisconsin. For the

past two years this packaged milk, sold primarily in gas stations, has

been available to consumers at prices up to thirty-four cents per gallon

less than Florida whole milk sold in supermarkets and convenience stores
2
($0.99 vs. $1.33). Measurements of consumer demand are essential to

the understanding of how consumers reacted to Midwest milk and its impli-

cations for the Florida dairy industry. Another development is the

concern over the potential impact of filled milk on the Florida fluid
3
milk market. The nature of this impact would depend to a large extent

on consumers' receptivity to filled milk and the degree to which users

of filled milk would substitute it for whole milk.

These developments and other factors that influence consumer demand

for fluid milk are expected to alter milk consumption patterns and the

growth of the fluid milk market in Florida. Changes at the consumer level

will necessitate production and marketing adjustments. Without accurate

information on consumer demand for fluid milk and consumers' milk purchasing

behavior and attitudes towards filled milk in Florida, the dairy industry

will be hard pressed to make the kinds of adjustments that will maintain




2Whole milk prices varied over this interval during the observation
period of this study. This wide range of price variability provided a
unique opportunity to measure consumers' purchase response to milk prices.

3Filled milk is a product made of skim milk or skim milk solids
and water combined with a non-dairy fat such as coconut oil. Currently,
filled milk is not sold in Florida. Although its sale is not prohibited
by law, except in interstate commerce, there is a law that states that
the skim milk component of filled milk must be priced at the same level
as Class I milk used in regular fluid milk products.








and hopefully enhance industry and consumer welfare. It is towards an

improved understanding of consumer milk purchasing behavior and fluid

milk demand in Florida that the present study is directed.


Objectives

The objectives of this study are:

1) To determine purchases and expenditures on fluid milk

products for different types of households.

2) To measure changes in consumer purchases and expenditures

on whole milk and all fluid milk in response to changes in

milk prices and various socioeconomic factors.

3) To identify and assess consumers' attitudes towards filled

milk.


Previous Research

Several studies have been made of consumption, expenditures and

demand for fluid milk and manufactured dairy products. At the University

of Georgia, Atlanta consumer panel data (1958-1962) were used to estimate

annual consumption and expenditures on various fluid milk and dairy pro-

ducts for households of different socioeconomic composition and the
4, 5
quantity and expenditure response to selected socioeconomic factors.4

In addition, quantity and expenditure slopes and elasticities with respect




Robert Raunikar, J. C. Purcell and J. C. Elrod, Consumption and
Expenditure Analysis for Dairy Products, Fats, and Oils in Atlanta,
Georgia, University of Georgia, Agricultural Experiment Station, Technical
Bulletin N. S. 51, May 1966.


5Joseph Purcell, R. Raunikar and J. C. Elrod, Analysis of Demand
for Beverage Milk, Atlanta, Georgia Consumer Panel, University of Georgia,
Agricultural Experiment Station, Research Bulletin No. 43, October 1968.








and hopefully enhance industry and consumer welfare. It is towards an

improved understanding of consumer milk purchasing behavior and fluid

milk demand in Florida that the present study is directed.


Objectives

The objectives of this study are:

1) To determine purchases and expenditures on fluid milk

products for different types of households.

2) To measure changes in consumer purchases and expenditures

on whole milk and all fluid milk in response to changes in

milk prices and various socioeconomic factors.

3) To identify and assess consumers' attitudes towards filled

milk.


Previous Research

Several studies have been made of consumption, expenditures and

demand for fluid milk and manufactured dairy products. At the University

of Georgia, Atlanta consumer panel data (1958-1962) were used to estimate

annual consumption and expenditures on various fluid milk and dairy pro-

ducts for households of different socioeconomic composition and the
4, 5
quantity and expenditure response to selected socioeconomic factors.4

In addition, quantity and expenditure slopes and elasticities with respect




Robert Raunikar, J. C. Purcell and J. C. Elrod, Consumption and
Expenditure Analysis for Dairy Products, Fats, and Oils in Atlanta,
Georgia, University of Georgia, Agricultural Experiment Station, Technical
Bulletin N. S. 51, May 1966.


5Joseph Purcell, R. Raunikar and J. C. Elrod, Analysis of Demand
for Beverage Milk, Atlanta, Georgia Consumer Panel, University of Georgia,
Agricultural Experiment Station, Research Bulletin No. 43, October 1968.








to price and various socioeconomic variables were estimated.5' 6 Consump-

tion and demand for fluid milk were also analyzed using cross-section

data for 120 households in each of twelve Southern cities (none of which

were in Florida). Price and income elasticities of demand for fluid

milk products in Mississippi were estimated with time series data for

1963 to 1967 and again for 1962 to 1968.8 Consumption and utilization

of milk and consumer attitudes towards milk and milk products have also

been studied for Alabama families.9


Nature of Data

The lack of milk consumption data for Florida and the insufficient

variation in published retail prices of Florida fluid milk products,

prevent the estimation of consumer market demand for fluid milk in Florida.

After considering alternative methods of collecting primary data (interviews

over the telephone, at the points of purchase or in the home), the method

of personal interviews in the home was chosen. While this method is




J. C. Purcell, J. C. Elrod and R. Raunikar, Price-Quantity-Expendi-
ture Relations for Food, Atlanta Consumer Panel Quarterly Estimates,
University of Georgia, Agricultural Experiment Station, Technical Bulletin
N. S. 52, May 1966.

F. L. Corty and J. C. Purcell, Consumption and Demand, Fluid Milk
and Fluid Milk Substitutes in the Urban South, Southern Cooperative Series
Bulletin No. 58, October 1957.

8Verner Hurt and W. M. Gamble, An Analysis of the Demand for Fluid
Milk Products in Mississippi, 1963-67, Mississippi Agricultural Experiment
Station, A. Ec. M. R. No. 54, January 1969. C. M. Young, D. L. Trammell,
L. D. Welch and V. G. Hurt, The Mississippi Market for Fluid Milk, Mississ-
ippi State University Agricultural and Forestry Experiment Station, Bulletin
781, September 1970.

Ruth A. Hammett and J. H. Blackstone, Alabama Urban Homemakers and
Milk Products, Auburn University Agricultural Experiment Station, Bulletin
352, June 1964.








somewhat time consuming it provides the most favorable environment for

collecting accurate and complete data, aside from perhaps the method of

having the respondent keep a weekly diary of purchases, prices and re-

lated information. For convenience and to limit data collection costs

households were from the city of Gainesville, Florida (site of the

University of Florida).

A random sample of 334 households was drawn from the 1969 Gaines-

ville city directory. The sample represents approximately 1.6 percent

of the 20,000 households in the city limits of Gainesville, Florida.10

Household interviews were conducted using a questionnaire (see Appendix A)

containing questions on household purchases and prices paid for whole

milk, skim milk, fortified low-fat (2 percent) milk, chocolate milk and

drinks, buttermilk and canned and dried milk at various outlets during

the preceding seven-day period.1 Other questions pertained to socioeconomic

characteristics of the household, utilization of fluid milk in the household,

attitudes towards filled milk and other aspects of milk purchasing behavior.




1For a confidence level of 95 percent, a sample size of 397 from
a population of 20,000 would give a reliability factor of 15 percent
when 50 percent of the population possesses the attribute being measured
(see: H. Arkin and R. R. Colton, Tables for Statisticians, Barnes and
Noble, Inc., New York, 1950). Considering that more than 50 percent of
Gainesville households purchase and consume fluid milk (which is one of
the major variables being measured in this study) and that the required
sample size decreases with respect to the proportion of the population
possessing the characteristic being measured, a sample size of 300 would
provide a reliability factor of less than 15 percent. To insure that at
least 300 households were included in the analysis a slightly larger
sample was drawn. A sample of 334 households was obtained by taking
every 60th household entry in the 1969 Gainesville city directory (Polk's
Gainesville Street and Avenue Guide Including Telephone Numbers, Home
Owner Symbols and Zip Codes, City Directory, 1969).

11The four outlets used were: supermarkets, convenience stores, gas
stations and others. The last category includes home delivery, commissary
purchases and miscellaneous points of purchase.








Interviews were taken in May, June and July of 1970. Of the 334 question-

naires obtained, 309 contained all the information asked for and were

used in the analysis.

In view of the fact that Gainesville is a university community,

it can be expected to contain many households that possess socioeconomic

and cultural characteristics that are somewhat different from the charac-

teristics of households in other areas of Florida. A large proportion

of Gainesville residents earn medium to high incomes, have an above average

number of years of formal education, and are engaged in professional and

technical occupations. The presence of a large student population gives

the young Gainesville household a somewhat different combination of

characteristics than is typical of young households in non-University

communities.

Households with medium to high incomes and advanced education

are usually heavy users of fluid milk products. Hence, the average

Gainesville household is quite likely to consume larger quantities of

milk and to have different attitudes towards milk than the average Florida

household. This situation may seem to distract from the usefulness of

findings based on a sample of Gainesville households. However, the behavior

and attitudes of heavy users of milk have an important bearing on the

outcome (success or failure) of price and marketing programs in the

Florida dairy industry. Viewed in this manner, analyses based on a

sample of Gainesville households should provide information that is useful

to the Florida dairy industry.


Theoretical and Analytical Framework

Contemporary demand theory states that under certain conditions

an individual's preferences for all possible (non-negative) commodity








12
bundles can be represented by an ordinal utility function.2 For a given

set of prices and income the so-called "rational consumer" purchases

that commodity bundle which maximizes his utility function and exhausts

his income. The demand functions derived from this model show that

quantity demanded of an item depends on all commodity prices and income.

For justifiable reasons this demand model is an abstraction of actual

consumer purchase behavior. As a result several problems arise in applying

demand theory. Purchases of food items for home use are typically made

for an entire household whereas demand theory only considers the purchase

decisions of an individual. To treat a household as an individual does

not necessarily resolve the problem because household members may have

different preferences. In addition, preferences can vary over households.

A possible approach to this problem and the one used in this study is to

try to account for variability in preferences by including specific house-

hold variables in the demand function, e.g., the number of household members

in various age categories and the ethnic background of the household.

Another difficulty with applying the theory results from the presence

of all commodity prices in the demand functions. In applications, it is

impractical to include all prices in the equation used to measure demand.

Data for the present study were collected over a short period of time

during which actual prices of fluid milk items varied considerably whereas

prices of all other items remained relatively constant. Therefore, the

prices of nonfluid milk items were excluded from the demand functions for

fluid milk. Although prices of canned and dried milk and milk consumed

away from home remained fairly constant during the period of observation,




12Donald W. Katzner, Static Demand Theory, Macmillan, London, 1970,
pp. 15-27.









purchases of these items were expected to influence fluid milk purchases

and expenditures.

Based on these considerations household demand for fluid milk

was represented as follows:


(1) QF = G(PF, QCD, QA, E, N1, N2, N3, N4, N5, Y1, Y2, Y3)

where:

QF = weekly household purchases of fluid milk consumed at home
in pounds;

PF = price paid for fluid milk in cents per pounds (expenditure/
quantity purchased);

QCD = weekly household purchases of canned and dried milk in pounds;

QA = weekly household consumption of fluid milk away from home
in pounds;

E = ethnic background of household (0 for Caucasian, 1 for non-
Caucasian);

N1 = number of adults 18 years and older;

N2 = number of persons 11 to 17 years old;

N3 = number of persons 6 to 10 years old;

N = number of persons 2 to 5 years old;

N5 = number of persons less than 2 years old;

Y1 = low household income, less than $4,999;13

Y2 = medium household income, $5,000 to $13,999;

Y3 = high household income, $14,000 and above.

The household demand function for whole milk was represented as follows:



13
1Continuous data on household income were not collected. Instead
the respondent was asked to select the category of income that best
described total household income (see Table 1 for categories). For this
reason Y, Y and Y3 were handled as zero-one intercept shifter variables
in the demand and expenditure equations. For a complete discussion of
this technique see: D. B. Suits, "Use of Dummy Variables in Regression
Equations," Journal of the American Statistical Association, 52:548-551,
1957.








(2) QW = H(PW, QO, QA, E, N1, N2, N3, N4, N5, Y1, Y2, Y3)

where:

QW = weekly household purchases of whole milk consumed at home
in pounds;

PQ = price paid for whole milk in cents per pounds (expenditure/
quantity purchased);

QO = weekly household purchases of other milk in pounds; (fluid
milk minus whole milk plus canned and dried milk).


Other variables appearing in the whole milk demand function are as pre-

viously defined. All purchase and price variables were converted from

a quart, half-gallon or gallon base to a pound base in order to permit

aggregation over different package sizes. Canned and dried milk purchases

were converted from actual product weights to fluid milk equivalents.

Changes in fluid or whole milk prices may alter household expendi-

tures but have no effect on household purchases of these products. For

example, as the per unit price of milk decreases, households may simply

reduce their milk expenditures without varying total milk purchases.

This type of behavior could cause milk expenditures to vary in direct

proportion to price. On the other hand, if expenditures varied inversely

with price, purchases would vary inversely with price (8E/P = P(3Q/DP) +

Q < 0 implies DQ/8P < 0, where E (=QP), Q and P are expenditures, purchases

and price, respectively). The relationships between expenditures and

prices of fluid milk and whole milk were investigated by the following

expenditure relations:


(3) EF = H(PF, QCD, QA, E, N1, N2, N3, N4, N5, Y1' Y2 Y3)

(4) EW = H(PW, QO, QA, E, N N2, N3, N4, N5, 1, Y2, Y3)


where:









EF = weekly household expenditures on fluid milk consumed at
home, in cents;

EW = weekly household expenditures on whole milk consumed at
home, in cents.


All independent variables are as previously defined. Estimates of the

price coefficients of equations (3) and (4) would show how expenditures

and purchases on fluid milk and whole milk vary with respect to own

price.


Empirical Results


Socioeconomic Characteristics of Households

Table 1 gives the distribution of all 309 households by household

income, occupation and education of the head of the household, ethnic

background and the age-sex distribution of all persons in the sample.

The percentage distribution of households among low, medium and high

incomes is 26.53, 56.63 and 16.84 percent, respectively. Over half of

the households reported the occupation of the head of the household as

professional worker, technical worker or student. Education received

by the head of the household is quite advanced with 231 (75 percent)

having some college training or above. The relatively large proportion

of heads of household with professional, technical or student occupations

and advance educational backgrounds was expected for a University city.

Only 28 households or 9 percent of all households were non-Caucasian

which is considerably below their proportion in the population of Gaines-

ville. This unrepresentative proportion occurred because many of the

non-Caucasian households in the original sample had to be dropped since

both parents were away from home during the daytime hours when interviews

were being conducted. Adults accounted for the largest proportion of









Table 1.--Selected socioeconomic characteristics: All 309 households


Characteristic Number Percent


Household Income:a

Low-
less than $2,999 31 10.03
$3,000 $4,999 51 16.50

Medium-
$5,000 $6,999 51 16.50
$7,000 $8,999 68 22.01
$9,000 $13,999 56 18.12

High-
$14,000 $18,999 34 11.02
$19,000 $23,999 9 2.91
more than $24,000 9 2.91

Occupation:b

Clerical and sales workers 33 10.67

Professional, technical and
kindred workers 82 26.53

Managers, officials and
proprietors 19 6.14

Craftsman, foreman, farmers,
farm manager and kindred
workers 21 6.79

Operatives and kindred
workers 9 2.91

Service and private household
workers and laborers 32 10.35

Student 96 31.06

Retired, unemployed 17 5.55

Education:c

Grammar school 48 15.53
Some high school 13 4.20
High school graduate 17 5.50
Some college 92 29.77
College graduate--graduate school 139 45.00


Continued









Table 1.--Selected socioeconomic characteristics: All 309 households--
Continued



Characteristic Number Percent

Ethnic Background:

Caucasian 281 90.93
Non-caucasian 28 9.07

Age-Sex Category: Percentage of
Number of Persons Persons

Adults: 18 years and older 671 66.11
Males: 11-17 years old 91 8.97
Females: 11-17 years old 57 5.62
Males: 6-10 years old 44 4.33
Females: 6-10 years old 42 4.14
Males: 2-5 years old 40 3.94
Females: 2-5 years old 33 3.25
Children less than 2 years old 37 3.64



aReported on an annual basis.


For the head of the household.


CHighest level of education attained by the head of the household.


dCategory for members in all 309 households.










persons in the sample which reflects the large population of single

students living in Gainesville.


Milk Purchases and Expenditures for Selected Households

This section presents and analyzes the actual purchases and expendi-

tures on milk items made by various groups of households. For analytical

purposes all households were classified according to the kinds of milk

products they purchased; namely, users of "whole milk," "fluid milk,"

"other milk" and "non-users." Whole milk users are defined as households

that purchased some fresh whole milk. Fluid milk users are households

that purchased at least one of the following: fresh whole milk, skim

milk, low-fat (2 percent) milk, chocolate milk and drinks or buttermilk.

Other milk users are households that did not purchase any fluid milk but

did purchase canned or dried milk. Non-users are households that did

not purchase any form of milk.

Of the 309 households analyzed there were 271 whole milk users,

293 fluid milk users, fourteen other milk users and two non-users. Table 2

gives household and per capital purchases and expenditures on six milk

items for whole milk users, fluid milk users and all households combined.

Since fluid milk users include all whole milk users these two groups are

not mutually exclusive. The differences between the figures for whole

milk users and fluid milk users are accounted for by households that

purchased fluid milk products other than whole milk. Differences between

the figures for fluid milk users and all households reflect the inclusion

of other milk users and non-users in the all household category.

Skim milk and low-fat milk purchases are greater in households that

consume fluid milk products other than whole milk. Purchases of other

milk products except whole milk are about the same for whole milk users










Table 2.--Average purchases and expenditures on fluid and other milk pro-
ducts for whole milk users, fluid milk users and all households:
Gainesville, Florida, average week in May-July, 1970


Purchases Expendi-
Household group Purchases per Expenditures per tures per
and product household per household person person


pounds percent dollars percent pounds


dollars


Whole Milk Users:

Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk

Total fluid milk

Canned milka
Dried milkb

Total milk at home

Milk away from home


17.84
1.31
0.23
0.29


75.47
5.54
0.97
0.93


2.15
0.07
0.03
0.03


84.98
2.76
1.19
1.19


19.60 82.91 2.28 90.12

1.42 6.01 0.13 5.14
2.62 11.08 0.12 4.74

23.64 100.00 2.53 100.00

1.89 7.99d -


Fluid Milk Users:


Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk

Total fluid milk

Canned milk
Dried milk

Total milk at home

Milk away from home


16.50
1.65
0.21
0.29


73.30
7.33
0.93
1.29


1.99
0.13
0.03
0.04


18.65 82.85 2.19

1.37 6.09 0.13
2.49 11.06 0.12


22.51 100.00

1.79 7.95


81.56
5.33
1.23
1.64

89.76

5.32
4.92


2.44 100.00


All Households:


Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk


Total fluid milk


15.65
1.56
0.20
0.28


71.27
7.10
0.91
1.28


1.89
0.13
0.03
0.04


17.69 80.56 2.09


5.14
0.46
0.06
0.07

5.73

0.04
0.05

6.42


0.64
0.02
0.01
0.01

0.68

0.04
0.03

0.75


4.75
0.60
0.05
0.11

5.51

0.40
0.63

6.54

0.59


0.59
0.05
0.01
0.02

0.67

0.04
0.03

0.74


80.08
5.51
1.27
1.70

88.56


4.50
0.57
0.06
0.10

5.23


0.56
0.05
0.01
0.02

0.64


Continued










Table 2.--Average purchases and expenditures on fluid and other milk pro-
ducts for whole milk users, fluid milk users and all households:
Gainesville, Florida, average week in May-July, 1970--Continued


Purchases Expendi-
Household group Purchases per Expenditures per tures per
and product household per household person person

pounds percent dollars percent pounds dollars

Canned milk 1.41 6.42 0.13 5.51 0.41 0.02
Dried milk 2.86 13.02 0.14 5.93 0.76 0.04

Total milk at home 21.96 100.00 2.36 100.00 6.40 0.70

Milk away from home 1.74 7.92 0.57


aFluid equivalent at a conversion rate of 1:2.1.


Fluid equivalent at


a conversion rate of 1:10.6.


CDue to suspected poor recall on prices of milk consumed away from home,
these prices were not collected.

dilk away from home as a percent of total milk purchased for home use.










and fluid milk users. For all households combined whole milk accounts

for 71 percent of all milk purchased for home use and 80 percent of

total milk expenditures. Fluid milk accounts for 80 percent of total

milk purchases and 88 percent of total milk expenditures.

Purchases and expenditures on milk items were also computed for

each of six types of households. Household types were defined according

to size (number of members) and income category. The six types of house-

holds are:

1. Small size Low income

2. Small size Medium income

3. Small size High income

4. Large size Low income

5. Large size Medium income

6. Large size High income


A small household is defined to have up to three members and a large

household four or more members. Low, medium and high incomes are

indicated in Table 1.

Purchases and expenditures on milk products by the six types of

household are given in Table 3. These figures provide considerable in-

sight into the variation in milk purchases and expenditures with respect

to household size and income. This variation was analyzed by identifying

the changes in milk purchases and expenditures for various combinations

of the six types of households. Changes were first determined for a

given household size letting income vary (Table 4) and secondly, for a

given income level letting household size vary (Table 5). Table 4 shows

the successive changes in purchases and expenditures on whole milk, fluid

milk, all milk consumed at home and milk consumed away from home starting








Table 3.--Average purchases and expenditures on fluid and other milk pro-
ducts for six types of households: Gainesville, Florida, aver-
age week in May-July, 1970



Purchases Expendi-
Household type Purchases per Expenditures per tures per
and item household per household person person


pounds percent


dollars percent


Small Size-Low Income:
(61 households)

Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk

Total fluid milk

Canned milka
Dried milkb

Total milk at home

Milk away from home

Small Size-Medium Income:
(84 households)

Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk

Total fluid milk

Canned milk
Dried milk

Total milk at home

Milk away from home

Small Size-High Income:
(18 households)

Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk


Total fluid milk


8.56
0.28
0.00
0.00


78.89
2.58
0.00
0.00


1.15
0.04
0.00
0.00


8.84 81.47 1.19

0.64 5.90 0.05
1.37 12.63 0.06


88.46
3.08
0.00
0.00

91.54

3.85
4.61


10.85 100.00 1.30 100.00

1.56 14.37d -


9.67
2.28
0.51
0.51


64.38
15.18
3.40
3.40


1.30
0.15
0.01
0.08


12.97 86.36 1.54

0.92 6.12 0.08
1.13 7.52 0.06


15.02 100.00

0.87 5.79


10.03
1.31
0.12
0.48


64.34
8.40
0.77
3.08


77.38
8.93
0.60
4.76

91.67

4.76
3.57


1.68 100.00


1.45
0.19
0.02
0.07


11.94 76.59 1.73


76.31
10.00
1.05
3.68

91.05


4.07
0.13
0.00
0.00

4.20

0.36
0.68

5.24

0.98


3.95
1.13
0.02
0.24

5.34

0.37
0.50

6.21

0.40


3.92
0.64
0.04
0.18

4.78


0.56
0.02
0.00
0.00

0.58

0.03
0.03

0.64


0.53
0.07
0.01
0.04

0.65

0.03
0.03

0.71


0.57
0.09
0.01
0.03


0.70
Continued


pounds


dollars







Table 3.--Average purchases and expenditures on fluid and other milk pro-


ducts for six types of households: Gainesville,
age week in May-July, 1970--Continued


Florida, aver-


Purchase Expendi-
Household type Purchases per Expenditures per tures per
and item household per household person person


Canned milk
Dried milk

Total milk at home

Milk away from home

Large Size-Low Income:
(18 households)

Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk

Total fluid milk

Canned milk
Dried milk

Total milk at home

Milk away from home

Large Size-Medium Income:
(92 households)

Whole milk
Skim & low-fat milk
Chocolate milk & drinks
Buttermilk

Total fluid milk

Canned milk
Dried milk

Total milk at home

Milk away from home


pounds percent dollars percent

1.02 6.54 0.07 3.68
2.63 16.87 0.10 5.27

15.59 100.00 1.90 100.00

0.89 5.71 -


18.75
0.00
0.84
0.00


72.73
0.00
3.26
0.00


2.48
0.00
0.12
0.00


80.78
0.00
3.91
0.00


19.59 75.99 2.60 84.69

2.85 11.06 0.29 9.45
3.34 12.95 0.18 5.86

25.78 100.00 3.07 100.00

1.69 6.55 -


19.54
0.68
0.25
0.33


69.81
2.43
0.93
1.18


2.51
0.09
0.03
0.05


20.81 74.35 2.68

2.44 8.72 0.24
4.74 16.93 0.23 -

27.99 100.00 3.15 100.00

2.40 8.57 -


pounds


0.39
0.93

6.10

0.34


4.30
0.00
0.21
0.00

4.51

0.57
0.67

5.75

0.42


4.19
0.16
0.06
0.07

4.48

0.53
0.98

5.99

0.52


dollars


0.03
0.04

0.77


0.57
0.00
0.03
0.00

0.60

0.06
0.04

0.70


0.54
0.02
0.01
0.01

0.58

0.05
0.05

0.68


Continued








Table 3.--Average purchases and expenditures on fluid and other milk pro-
ducts for six types of households: Gainesville, Florida, aver-
age week in May-July, 1970--Continued



Purchase Expendi-
Household type Purchases per Expenditures per tures per
and item household per household person person

pounds percent dollars percent pounds dollars

Large Size-High Income:
(36 households)

Whole milk 32.91 74.81 2.85 79.61 7.73 0.64
Skim & low-fat milk 5.25 11.93 0.34 9.50 1.28 0.08
Chocolate milk & drinks 0.48 1.10 0.08 2.23 0.12 0.02
Buttermilk 0.12 0.27 0.02 0.56 0.03 0.01

Total fluid milk 38.76 88.11 3.29 91.90 9.16 0.75

Canned milk 0.74 1.68 0.06 1.68 0.16 0.01
Dried milk 4.49 10.21 0.23 6.42 0.95 0.05

Total milk at home 43.99 100.00 3.58 100.00 10.27 0.81

Milk away from home 2.80 6.36 0.60


See footnotes at end of Table 2.











Table 4.--Changes in average milk purchases and expenditures with
respect to income for small and large households



Starting Starting
Itema with S Y b with S Y1
11S2Y1


S1Y2


SIY3


S2


12 S2Y3


Household purchases of whole milk

Per capital purchases of whole milk

Household expenditure on whole milk

Per capital expenditure on whole milk

Household purchases of fluid milk

Per capital purchases of fluid milk

Household expenditure on fluid milk

Per capital expenditure on fluid milk

Household purchases of all milk

Per capital purchases on all milk

Household expenditure on all milk

Per capital expenditure on all milk

Household purchases of milk away
from home

Per capital purchases of milk away
from home


+1.11

-0.12

+0.15

-0.03

+4.13

+1.14

+0.35

+0.07

+4.17

+0.97

+0.38

+0.07


+0.36

-0.03

+0.15

+0.03

+0.57

-0.56

+0.19

+0.05

+0.57

-0.11

+0.22

+0.06


-0.69 +0.02


-0.58 -0.08


Except for the last two items, quantities and expenditures refer
to milk consumed at home. Quantities are measured in pounds and expen-
ditures in dollars.

bs and S stand for small and large households, respectively, and
Y' Y 2 and Y3 or low, medium and high incomes, respectively.


+0.79

-0.11

+0.03

-0.02

+1.22

-0.03

+0.08

-0.02

+2.21

+0.24

+0.08

-0.02


+0.71


+0.10


+13.37

+3.54

+0.34

+0.10

+17.95

+4.68

+0.61

+0.17

+16.00

+4.28

+0.43

+0.13


+0.40


+0.08










Table 5.--Changes in average milk purchases and expenditures with
respect to household size for low, medium and high income
households



From S1Y1 From SIY2 From SIY3
Itema
to S2 lb to S2Y2 to S2Y3


Household purchases of whole milk +10.19 +9.87 +14.16

Per capital purchases of whole milk +0.23 +0.24 +3.43

Household expenditure on whole milk +1.33 +1.21 +0.37

Per capital expenditure on whole milk +0.01 +0.11 +0.07

Household purchases of fluid milk +10.75 +7.84 +19.17

Per capital purchases of fluid milk +0.31 -0.86 +4.65

Household expenditure on fluid milk +1.41 +1.14 +0.69

Per capital expenditure on fluid milk +0.02 -0.07 +0.15

Household purchases of all milk +14.93 +12.97 +18.21

Per capital purchases on all milk +0.51 -0.22 +4.52

Household expenditure on all milk +1.77 +1.47 +0.51

Per capital expenditure on all milk +0.06 -0.03 +0.11

Household purchases of milk away
from home +0.12 +1.53 +1.11

Per capital purchases of milk away
from home -0.02 +0.12 +0.18



aExcept for the last two items, quantities and expenditures refer
to milk consumed at home. Quantities are measured in pounds and expen-
ditures in dollars.

bs and S stand for small and large households, respectively, and
Y1, Y2 and Y3 or low, medium and high incomes, respectively.








with small size low income households (S 1Y) in the first two columns

and with large size low income households (S2Y1) in the last two

columns. Plus indicates an increase and minus indicates a decrease.

A similar scheme is used in Table 5 to illustrate the changes in pur-

chases and expenditures for the two sizes of households at each income

level.

Increases in income are accompanied by increases in household

purchases and expenditures on whole milk, fluid milk and all milk consumed

at home for small and large households (c.f., Table 4). For small

households the increases in purchases and expenditures are greater be-

tween low and medium income households than between medium and high income

households. This relationship is reversed for large households. Household

consumption of fluid milk away from home increases with income for large

households whereas for small households consumption decreases from low to

medium income households and increases from medium to high income house-

holds. In large households per capital purchases and expenditures on

whole milk and fluid milk and per capital expenditures on all milk consumed

at home decrease from low to medium income households and increase from

medium to high income households. The pattern in per capital purchases

and expenditures is less uniform for small households.

At each income level, increases in household size are associated

with increases in household purchases and expenditures on whole milk,

fluid milk, all milk consumed at home and milk consumed away from home

(c.f., Table 5). The increases in household purchases of whole milk,

fluid milk and all milk consumed at home with respect to size are largest

in high income households and smallest in medium income households. In









low and high income households, increases in size are accompanied by

increases in per capital purchases and expenditures on all items except

milk consumed away from home. Per capital purchases and expenditures in

medium income households have a less uniform relationship with household

size.

In summary, this analysis indicates that household purchases and

expenditures on whole milk, fluid milk and all milk consumed at home

increase with income for each size of household and increase with house-

hold size at each income level. Per capital purchases and expenditures

on these items are not always positively related to household size and

income.


Estimated Demand and Expenditure Relationships

The demand and expenditure equations for fluid and whole milk,

equations (1) through (4), were fitted by ordinary least-squares.14

Table 6 contains the regression coefficients and related statistical

values for the fluid milk demand equation. Only three coefficients are

statistically significant above the five percent level, namely the coef-

ficients of price, the number of persons 6 to 10 years old and the high

income variable. The average fluid milk price elasticity shows that a

one percent decrease (increase) in fluid milk prices is associated with

a 5.7 percent increase (decrease) in household purchases of fluid milk.

The magnitude of this price elasticity may be somewhat distorted by the

fact that the price coefficient (slope), on which it is based, was esti-

mated using a weighted average price. The weighted average price depends



14
1To avoid perfect multicollinearity between Y1, Y2 and Y3, only Y
and Y3 were included in the regression equations. Regression coefficients
of Y and Y3 measure deviations from the income effect in low income house-
holds.








Table 6.--Regression coefficients and related statistical values:
Fluid milk demand equation



-Independent Regression Standard t-prob- Average
variable coefficient error ability elasticity


Fluid milk price (PF)

Canned & dried milk purchases
(QCD)

Milk consumed away from home
(QA)

Ethnic background (E)

Number of adults (NI)

Number of persons 11 to 17
years old (N2)

Number of persons 6 to 10
years old (N3)

Number of persons 2 to 5
years old (N4)

Number of persons less than
2 years old (N5)

Medium income (Y2)

High income (Y3)

Constant term = 121.034
R2= .331; 2 .284b F(,281)
R = .331; R = .284 ; F(11,281)


- 7.718***


- .386


.726

- 2.017

- .229


1.582


.829 .00000001


.255 .131


.570

7.029

2.659


.204

.774

.931


2.251 .482


7.375** 3.138 .019


- 1.584


2.226

.869

16. 195***


3.682 .667


5.291

4.501

5.991


.674

.847

.007


aComputed by evaluating coefficients at the means (see Appendix B
for actual means).

Adjusted coefficient of determination.

CF-value for testing whether R2 is significantly different from zero.

*Significant at the 5 percent level.
**Significant at the 2 percent level.
***Significant at the 1 percent level.


-5.765


- .080


.069

- .008

- .026


.039


.114


- .021


.015

.026

.151


= 11.526***c








on not only the prices paid but also the relative purchases of fluid

milk made at each outlet. Hence, the estimated price slope and the aver-

age price elasticity may confound the true effects of price changes with

the effects of other factors that influence household purchases of fluid

milk, such as the age-composition of the household. Since Y3 is a

zero-one variable its coefficient cannot be interpreted as a conventional

income slope. Nevertheless, the coefficient of Y3 shows that fluid

milk purchases are greater in high income households than in low income

households.

Table 7 contains the regression results for the fluid milk expendi-

ture equation. Eight of the coefficients are statistically significant

at or above the five percent level. The price coefficient is negative

and statistically significant and the average price elasticity shows that

a one percent decrease (increase) in fluid milk prices is associated with

a 0.55 percent increase (decrease) in household expenditures on fluid

milk. The negative price coefficient also implies that fluid milk pur-

chases vary inversely with price which is consistent with the negative

price coefficient of the demand equation (Table 6).15 The coefficient

of milk consumed away from home is positive and statistically significant,

thus suggesting that as households consume more milk away from home they

also spend more on fluid milk consumed in the home. This does not mean

that fluid milk consumed at home (QF) and milk consumed away from home

(QA) are complements.6 Coefficients of the age-composition variables,



If (DEF/3PF) = PF(DQF/DPF) + QF < 0, then (aQF/3PF) < 0, because
PF > 0 and QF > 0.

16QF and QA would be complements only if (3QF/aPA) + QF (aQF/aY)
< 0, where PA is the price of QA and Y is income. If QF is a normal
good, i.e., (3QF/DY) > 0, which the coefficients of Y2 and Y3 in














Table 7.--Regression coefficients and related statistical values:
Fluid milk expenditure equation1



Independent Regression Standard t-prob- Average
variable coefficient error ability elasticitya


Fluid milk price (PF)

Canned & dried milk purchases
(QCD)

Milk consumed away from home
(QA)

Ethnic background (E)

Number of adults (N1)

Number of persons 11 to 17
years old (N2)

Number of persons 6 to 10
years old (N3)

Number of persons 2 to 5
years old (N4)

Number of persons less than
2 years old (N5)

Medium income (Y2)

High income (Y3)

Constant term = 185.993
2 .470; 2 449; F(,281)
R = .470; R = .449 ; F(11,281)


- 8.792**


- 1.889


14.487***

-56.287 2

24.535* 1


48.076***


83.251*** 1


64.298*** 1


89.059*** 2

8.586 1

62.286** 2



= 22.655***C


3.566 .014


1.098 .086


2.452

30.230

.1.438


9.682


3.495


.5.834


'2.756

9.357

5.767


.00000001

.063

.032


.000001


.00000001


.00006


.0001

.657

.016


See footnotes at end of Table 6.


-.556


-.033


.117

-.020

.242


.100


.109


.072


.051

.022

.049








i.e., N1 through N5, are all positive and significant. Note also that

for the most part the age-composition elasticities decrease in direct

relation to age. Only the coefficient of the high income variable is

statistically significant and the positive sign implies that high in-

come households spend more on fluid milk than low income households.

The demand and expenditure equations specified for whole milk

are somewhat different than the ones specified for fluid milk. Besides

replacing fluid milk price by whole milk price, purchases of other milk

(fluid milk less whole milk plus canned and dried milk, QO) are combined

in a single variable which replaces purchases of canned and dried milk,

QCD. All other variables remain the same.

Table 8 contains the regression results for the whole milk demand

equation. Five of the coefficients are statistically significant at or

above the five percent level. The price coefficient is negative and the

whole milk price elasticity shows that a one percent decrease (increase)

in whole milk prices is associated with a 3.8 percent increase (decrease)

in household purchases of whole milk. The positive signs for the coef-

ficients of purchases of other milk and milk consumed away from home

suggest that heavy users of other milk and milk away from home are also

heavy users of whole milk. Once again, this does not imply that milk

consumed away from home and other milk are complements for whole milk.

As in the case of fluid milk the number of 6 to 10 year olds and high

income have a positive and statistically significant effect on whole milk

purchases.


Table 6 seem to suggest, then the condition for complementarity reduces
to: (3QF/8PA) < 0. But, (3EF/@QA) > 0 does not imply this condition.















Table 8.--Regression coefficients and related statistical values:
Whole milk demand equation1


Independent Regression Standard t-prob- Average
variable coefficient error ability elasticity


Whole milk price (PW)

Purchases of other milk (QO)

Milk consumed away from home
(QA)

Ethnic background (E)

Number of adults (N1)

Number of persons 11 to 17
years old (N2)

Number of persons 6 to 10
years old (N3)

Number of persons 2 to 5
years old (N4)

Number of persons less than
2 years old (N )

Medium income (Y2)

High income (Y3)

Constant term = 75.463
2 .454; 2 = b
R = .454; R = .431 ; F(11,2


-4.906***

.962***


.846*

-7.707

.773


-1.689


5.858**


-3.451


3.485

- .141

9.198*


.685

.117


.426

5.292

2.033


.00000001

.00000001


.048

.146

.704


1.615 .296


2.359 .013


2.680 .198


3.935

3.470

4.582


.376

.967

.045


259) = 19.604***


See footnotes at the end of Table 6.


-3.805

.312


.089

- .035

.094


- .047


.099


- .050


.026

- .004

.091








In the whole milk expenditure equation, Table 9, six coefficients

are statistically significant. The price elasticity shows that a one

percent decrease (increase) in whole milk prices is associated with a

0.524 increase (decrease) in whole milk expenditures. A negative price

coefficient also implies that whole milk purchases are inversely related

to whole milk prices as found in the whole milk demand equation (Table 8).

The positive and statistically significant coefficient of milk consumed

away from home implies that heavy consumers of milk away from home

spend more on whole milk. Except for the number of adults the age-compo-

sition coefficients are positive and statistically significant. The

two income coefficients are not statistically significant at the five

percent level.


Comparison and Evaluation of Regression Results

One would expect the price elasticity of demand for whole milk

to exceed, in absolute value, the price elasticity of demand for fluid

milk because of the availability of more substitutes for whole milk than

fluid milk. However, in the present analysis the variability, measured

in terms of standard deviations, in fluid milk purchases and prices

(3.587 and 2.271) is greater than the variability in whole milk purchases

and prices (2.985 and 2.199). This could explain part of the difference

between the whole milk and fluid milk price elasticities (-3.805 vs. -5.675).

Available cross-section studies do not provide estimates of the fluid

milk purchase-price elasticity or the whole milk expenditure-price elasticity.

However, the purchase-price elasticity for whole milk was estimated to be

-2.91 for Atlanta panel households; a value which is smaller than the

estimate -3.805 given here.7 However, the elasticities estimated in



17Purcell, Elrod and Raunikar, op. cit., p. 78.














Table 9.--Regression coefficients and related statistical values:
Whole milk expenditure equation1



Independent Regression Standard t-prob- Average
variable coefficient error ability elasticity


Whole milk price (PW)

Purchases of other milk (QO)

Milk consumed away from home
(QA)

Ethnic background (E)

Number of adults (N1)

Number of persons 11 to 17
years old (N2)

Number of persons 6 to 10
years old (N3)

Number of persons 2 to 5
years old (N4)

Number of persons less than
2 years old (N5)

Medium income (Y2)

High income (Y3)

Constant term = 179.537
2 =2 b
R = .443; R= .420 ; F(11,259)


- 8.162*

- 1.227


13.947***

-58.652

21.908


42.173***


82.896***


60.779***


88.154***

6.858

50.980


See footnotes at the end of Table 6.


4.023

.687


2.502

31.046

11.932


9.476


13.840


15.723


23.089

20.361

26.885


.043

.075


.00000007

.059

.067


.0001


.00000001


.0001


.0001

.736

.059


-.524

-.033


.122

-.022

.220


.097


.116


.074


.056

.017

.042


= 18.775***c








this study were derived from price data that exhibited more variation

than is typical of cross-section and time-series studies. Such price

variation may explain the comparatively large values of the price elas-

ticity estimates.

Perhaps the most salient feature of the regression results is

the difference between the statistical significance of the age-composition

variables in the demand and expenditure equations. In the fluid and

whole milk demand equations only the number of 6 to 10 year olds is

statistically significant whereas in the expenditure equation all five

age-composition variables (except adults in the whole milk expenditure

equation) are statistically significant. The question then arises: Do

the age-composition variables influence only expenditures on fluid and

whole milk and not the quantities purchased? In a Georgia study of pur-

chases and expenditures on whole milk, the numbers of persons over 18

years old, 6 to 10 years old, and 2 to 5 years old had a statistically
18
significant effect on annual quantities purchased of whole milk.8 By

comparison, only the numbers of 6 to 10 year olds and 2 to 5 year olds

had a statistically significant effect on expenditures for whole milk.

The Georgia study did not consider the number of persons under 2 years

of age but did show more age-composition variables to be significant in

the purchase equation than in the expenditure equation. Perhaps, it

would be appropriate to offer a possible explanation for the differences

in the significance of the age-composition variables.

One of the specification problems encountered in this analysis

related to the empirical representation of price in the demand and

expenditure equations. In theory own price is an exogenous variable.



18Raunikar, Purcell and Elrod, p. cit., pp. 25 and 27.
Raunikar, Purcell and Eirod, op. cit., pp. 25 and 27.








The weighted average price in the demand and expenditure equations fitted

in this study is not exogenous--it depends not only on the quantities

of milk purchased at various outlets but also on the size of package in

which milk is purchased. Since large households are likely to purchase

milk in bigger package sizes we would expect a large household to pay

a lower per unit price for milk than a small household even for milk

purchased at the same outlet. If this proposition is correct and if

larger households consume more milk, then as the number of persons in

an age category increases the weighted average price paid for milk would

decrease. The simple correlations between the age-composition variables

and the weighted average prices of whole milk and fluid milk were all

found to be negative thus supporting the proposition.

Under these circumstances age-composition variables could indirectly

influence purchases through their effect on weighted average price in

which case it would be difficult to separate and measure the net effects

of price and age-composition variables on quantities purchased. In

contrast, the regression coefficients of the age-composition variables in

the expenditure equations measure directly the net effects of these vari-
DE = P 3Q + Q 3P
ables on price and quantity because: N. N. 8N.



This argument could explain why more age-composition coefficients are

significant in the expenditure equations than in the demand equations.

Conceptually, this price specification problem could be avoided by esti-

mating a separate demand equation for milk in each package size and for

each outlet. Practically, however, this procedure would increase the

difficulty of estimating age-composition and income effects because of

the low level of aggregation.








All the equations fitted in this study have relatively low coef-

ficients of determination (R2) but all are statistically significant

at the one percent level. Low R 2's are not uncommon in cross-section
19
analyses of demand and consumer behavior.9 There are at least three

possible explanations for low R 's. First, the implicit assumption

that the parameters of the demand and expenditure equations are the

same for all households in the analysis is no doubt violated, even

though variables that account for some of the variation in preferences

over households were included in these equations. Secondly, the regression

analyses were carried out at a disaggregated level with the household

as the unit of observation. This alone increases the possibility of

considerable unexplained variability in the dependent variables. By

comparison, most time series regression analyses that employ average

data over a large market usually yield much higher R2's.

Finally low R2's may mean that the equations fitted do not ade-

quately represent the demand and expenditure relations for whole milk

and fluid milk. This possibility prompts the suggestion that household

demand and expenditure relations for whole milk and fluid milk should

be reconsidered and reestimated as further information and data on milk

purchasing behavior becomes available.


Other Information on Household Milk Purchasing Behavior

One of the purposes of this study was to obtain detailed information

about consumer milk purchasing behavior. The results discussed thus far




19For the Georgia study referenced in footnote 15, the whole milk
regression equations with quantities purchased and expenditures as the
dependent variables yielded R2's of 0.270 and 0.271, respectively. In the
Georgia study referenced in footnote 4, the R2 for the quantity purchased
regression for whole milk was 0.365.








have focused on milk purchases and expenditures in various types of house-

holds and household demand and expenditure relations for fluid milk and

whole milk. In this section household behavior with respect to other

aspects of milk purchasing and attitudes towards filled milk are reported.

Table 10 contains the responses of fluid milk and whole milk

users to several questions related to purchase behavior and attitudes.

The largest proportion of these households reported that the wife was

the only purchaser of milk. While the number of households in the other

members) category may seem high it includes all households made up of

students living together. Brand preferences for milk were weak. Less

than 24 percent of fluid and whole milk users indicated they purchased

only one brand. About 86 percent of the households reported that some

non-dairy beverages were consumed in place of fluid milk. This finding

suggests that non-dairy beverages compete with fluid milk in consumer

diets. The percent of milk consumed in beverage form is quite high with

over half of the households reporting beverage milk utilization to be

75 percent or more of total at home milk consumption.

Questions 5 and 6 try to identify consumer attitudes towards

filled milk. About 47 percent of the households surveyed said they

would be willing to try filled milk if it cost the same as whole milk.

This figure increased to 78 percent if the price of filled milk was 6

to 8 cents less per half gallon than whole milk. Of the households who

claimed they would try filled milk at the lower price, 32 percent said

they would continue to use filled milk along with whole milk provided

household members liked filled milk. The remaining 68 percent said they

would completely replace whole milk with filled milk.

Responses to the questions appearing in Table 10 for each of the

six types of households defined earlier are given in Table 11. This








Table 10.--Responses of fluid and whole milk users to selected ques-


tions on milk purchasing behavior filled
ville, Florida, May-June, 1970


milk: Gaines-


Question Fluid milk users Whole milk users

Number Percent Number Percent


1. Who buys milk for household:


Wife
Husband
Wife and husband
Other members)


43.34
7.84
26.62
22.20


41.70
8.48
27.30
22.52


2. What are brand preferences
for milk:

Purchase only one brand
Purchase one brand more
than any other
No attention paid to brands

3. Are other beverages consumed
in place of fluid milk:a

Yes
No

4. What percent of fluid milk
is consumed as a beverage:

Less than 25%
25 to 50%
51 to 75%
More than 75%

5. Would filled milk be tried
if it cost:

The same as whole milk
6 to 8 cents less per
half-gallon than whole
milk
Would not try filled milk

6. If family liked filled milk
would you:b


Use both filled and
whole milk


71 24.23


63 23.25


42.32
33.45


86.00
14.00


10.92
16.38
18.77
53.93


47.09


77.81
22.18


73 32.01


42.43
34.32


86.35
13.65


8.86
16.97
18.45
55.72


48.34


79.70
20.30


71 32.87


Continued









Table 10.--Responses of fluid and whole milk users to selected ques-
tions on milk purchasing behavior filled milk: Gaines-
ville, Florida, May-June, 1970--Continued


Question Fluid milk users Whole milk users

Number Percent Number Percent

Use only filled milk 155 67.99 145 67.13



An affirmative answer to this question does not mean that house-
hold members use other beverages to the exclusion of fluid milk, but
rather that other beverages are partially substituted for fluid milk.

bFigures are for households that stated they would try filled milk.








Table ll.--Responses to six types of households to questions on milk purchasing behavior and filled milk: Gaines-
ville, Florida, May-June 1970



Small size Large size
Low income Medium income High income Low income Medium income High income
Question 61 households 84 households 18 households 18 households 92 households 36 households


Pet. No.


Pet. No.


Pet. No.


Pet. No.


Pet. No.


1. Who buys milk for
household:
Wife
Husband
Wife and husband
Other members)
2. What are brand
preferences for milk:
Purchase only one
brand
Purchase one brand
more than any other
No attention paid
to brand
3. Are other beverages
consumed in place
of fluid milk:a


32.78
9.84
21.31
36.07


44.05
5.95
26.19
23.81


55.55
5.57
27.77
11.11


38.89
5.56
22.22
33.33


45.65
10.87
28.26
15.22


Pet.


58.33
2.79
22.22
16.66


13 21.31 20 23.81 9 50.00 3 16.66 23 25.27 14 38.88


21 34.43 41 48.81


6 33.33 8 44.44 38 41.76 12 33.33


27 44.26 23 27.38 3 16.67 7 38.90 30 32.97 10 27.79


90.16 71
9.94 13


84.52 18
15.48 0


100.00 16
0.00 2


88.89 78
11.11 14


84.78 29
15.22 7


4. What percent of fluid
milk is consumed as a
beverage:
Less than 25%
25 to 50%
51 to 75%
More than 75%


Yes
No


14.04
10.52
10.52
64.92


12.66
18.99
26.58
41.77


22.22
22.22
11.12
44.44


5.55
11.11
27.78
55.56


80.55
19.45


11.43
22.86
14.28
51.43


7.86
16.85
20.22
55.07









Table ll.--Responses to six types of households to questions on milk purchasing behavior and filled milk: Gaines-
ville, Florida, May-June 1970--Continued



Small size Large size
Low income Medium income High income Low income Medium income High income
Question 61 households 84 households 18 households 18 households 92 households 36 households

No. Pet. No. Pet. No. Pet. No. Pct. No. Pet. No. Pet.
5. Would filled milk be
tried if it cost:
The same as whole
milk 36 59.01 29 34.53 11 61.11 7 38.88 38 41.30 21 58.33
6 to 8 cents less
per half gallon
than whole milkb 46 75.41 55 65.47 14 77.77 16 88.88 74 80.43 31 86.11
Would not try filled
milk 15 24.59 29 34.53 4 22.23 2 11.12 18 19.57 5 13.89
6. If family liked filled
milk would you:c
Use both filled and
whole milk 15 32.61 20 36.36 2 14.28 5 31.25 21 28.38 14 45.16
Use only filled milk 31 67.39 35 63.64 12 85.72 11 68.75 53 71.62 17 54.84



aAn affirmative answer to this question does not mean that household members use other beverages to the exclusion
of fluid milk, but rather that other beverages are partially substituted for fluid milk.

The number of households in this category includes households that would try filled milk if it cost the same as
whole milk.

c
Figures are for households that stated they would try filled milk.









cross tabulation was done in order to determine whether the kinds of

answers given depend on the size and income level of the household.

The number of households responding to a question is in some cases

less than the total number of households of the given type due to the

non-responses of households not consuming milk.

For all six types of households the wife was the single most

important and the husband the least important purchaser of milk. In

medium and high income households the wife and husband purchased the

milk more often than other members and vice versa in low income households.

Loyalty to one milk brand increased with household income in each size

category. Also as household income increased the percent of households

paying no attention to brands decreased in both size categories. The

percent of households consuming other beverages in place of fluid milk

was lowest, 80.5 percent, in large size-high income households and

highest, 100 percent, in small size-high income households. Among small

size households, those having low incomes consumed proportionately more

milk in beverage form than medium and high income households. The percent

of large households consuming more than 75 percent of their milk in

beverage form was relatively constant with respect to income.

The percent of small households willing to try filled milk when

it cost the same as whole milk was about the same at the low and high

incomes, 59 and 61 percent, respectively, but dropped to 34 percent at

medium incomes. However, this percent increased with income for large

households. When filled milk cost 6 to 8 cents less than whole milk

the percent willing to buy filled milk was greater at low and high incomes

than at medium incomes for both sizes of households. As one might expect

the percent of households willing to try filled milk was greater for large








than small households. The most dramatic increase, 128 percent, in

the number of households willing to try filled milk when it cost 6 to

8 cents less instead of the same as whole milk occurred for large size-

low income households.

The number of households that would completely replace whole

milk with filled milk was greater than the number that would consume

both filled and whole milk for all six types of households. For small

households there was a direct relationship between the percent of house-

holds willing to try filled milk, at the lower price, and the percent

who would use only filled milk.


Summary

This report analyzes the milk purchasing behavior of 309 house-

holds that were randomly selected from all households in the city of

Gainesville, Florida. Data on the 309 households were collected by

means of a questionnaire that was completed during personal household

interviews. Various aspects of milk purchasing behavior were studied

including: (a) purchases and expenditures on whole milk and other fluid

milk products in various types of households, (b) changes in milk purchases

and expenditures with respect to household size and income, (c) household

demand and expenditure relations for fluid and whole milk, (d) purchaser

of milk for the household, brand preferences for milk and the utilization

of milk in beverage form and (e) consumer attitudes towards filled milk.

Cross tabular analyses show that household purchases and expenditures

on milk consumed at home increased with household income for fixed house-

hold size and increased with household size for fixed household income.

Per capital purchases and expenditures do not show a consistent positive

relationship with household size or income.








Demand and expenditure functions for fluid and whole milk were

estimated by ordinary least squares. Quantity purchased and expendi-

tures on whole milk and fluid milk were both inversely related to own

price. Estimated price elasticities of fluid milk demand and whole

milk demand were -5.765 and -3.805, respectively. Of the five house-

hold age-composition variables included in the regressions only the

number of 6 to 10 year olds had a positive and significant effect on

quantity purchased of whole milk and fluid milk. In contrast, all five

age-composition variables, with one exception, had a positive and signi-

ficent effect on fluid milk expenditures and whole milk expenditures.

Of the medium and high income (zero-one) variables used to explain pur-

chases and expenditures only the coefficient of high income was signifi-

cant, in particular, the high income coefficient was positive and signifi-

cant in the fluid milk demand and expenditure equations and in the

whole milk demand equation.

For the six types of households analyzed the wife was the single

most important purchaser of milk for the household. Households purchasing

only one brand of milk accounted for less than 25 percent of all whole

milk and fluid milk users. However, purchases of only one brand were

reported by 50 percent of the small size and 39 percent of the large

size households in the high income category. Fifty-six percent of all

whole milk users and 54 percent of all fluid milk users reported that

over 75 percent of all milk purchased for home use was consumed as a

beverage.

Consumer attitudes towards filled milk were generally favorable

in all types of households analyzed. Fourty-seven percent of all fluid

milk users and 48 percent of all whole milk users claimed they would try








filled milk even if it cost the same as whole milk. This figure increased

to 78 percent for fluid milk users and 80 percent for whole milk users

when filled milk cost six to eight cents less per half-gallon than

whole milk. On the average low and high income households were more

willing to try filled milk than medium income households. Also, large

households appeared more receptive to filled milk than small households.

Potential users of filled milk appeared more likely to completely substi-

tute filled milk for whole milk than to partially substitute it for

whole milk.

These findings by no means exhaust all the information one might

like to have on consumer milk purchasing behavior. Typically, the kind

of information which one considers to be most relevant depends on the

specific problems) one is attempting to solve. Nevertheless, the present

study does represent the first attempt to understand and analyze milk

purchasing behavior in Florida.


Implications and Interpretations

The implications and interpretations given below do not necessarily

apply to all areas of Florida since they are based on consumer purchase

behavior, attitudes and market events in Gainesville, Florida. Neverthe-

less, the fluid milk market in Gainesville does possess many of the distri-

bution and price characteristics of other Florida milk markets. In this

sense, other areas of Florida face milk marketing problems similar to

those found in Gainesville.

Household demands for fluid milk and whole milk were found to be

price elastic over the range of prices analyzed. Although this study

did not measure total milk sales before and after the introduction of

Midwestern milk, an elastic demand suggests that total sales of fluid








milk in Gainesville increased as households responded to lower milk

prices by increasing their total purchases and expenditures on fluid

milk. The greater part of any increases in total sales would have

gone to Midwest distributors because many households either partially

or completely substituted lower priced Midwestern milk for Florida pro-

duced milk. In an attempt to regain part of the Gainesville milk market

captured by Midwestern milk, Gainesville distributors found it necessary

to lower retail store prices of Florida produced milk. It remains to

be seen whether current store prices for fluid milk in Gainesville

($0.59 for half-gallon and $1.09 per gallon) are adequate to cover

processing and distribution costs.

The willingness of Gainesville households to try filled milk

takes on added significance in light of the fact that whole milk prices

were comparatively low during the time of the survey. Had whole milk

prices been at the levels existing prior to the introduction of Midwest

milk, consumers might have expressed an even greater willingness to try

filled milk. Evidence suggests that if a good quality filled milk becomes

available at prices below whole milk there is a strong possibility that

it would replace part or all of the whole milk in the diets of a significant

number of consumers. Current laws which regulate the cost of the skim

milk component of filled milk hold in check much of the potential impact

of filled milk. Moreover, the low milk prices that now prevail in Gaines-

ville may help to deter or delay the introduction of filled milk.

In summary, it appears that Gainesville milk prices, and most

likely prices in other areas of Florida, will continue to experience pre-

sures from changes in demand and supply conditions. On the demand side

there is the downward pressure on prices caused by actual or potential








shipments of packaged Midwest milk to Florida and its substitution for

Florida produced milk by consumers. This pressure builds up whenever

consumer prices of fluid milk in the two areas differ by more than the

inter-area transportation costs of packaged fluid milk. Price pressures

generated by the flow of Midwest packaged milk take on added significance

in view of the fact that minimum federal order prices for fluid milk in

Florida (and other non-surplus market areas)are based primarily on

Midwest prices of raw milk plus inter-area transportation costs of raw

milk. It appears, however, that price differentials and inter-area trans-

portation costs for packaged fluid milk are becoming as important a

determinant of consumer prices of milk in Florida as price differentials

and inter-area transportation costs for raw milk. On the supply side,

there is the upward pressure on milk prices caused by increasing costs

of production, processing and distribution.

This situation poses a serious challenge to the Florida dairy

industry, namely, to develop flexible pricing arrangements for fluid milk

that will assure both reasonable profits to industry participants and a

strong and growing demand for Florida fluid milk products. Hopefully,

information provided by studies of consumer demands and purchasing behavior

for fluid milk will contribute to the development of such pricing arrange-

ments.








































APPENDICES








APPENDIX A--Questionnaire


Questionnaire No. Date:

Address Interviewer:


GAINESVILLE MILK CONSUMPTION SURVEY
DEPARTMENT OF AGRICULTURAL ECONOMICS
UNIVERSITY OF FLORIDA


PART I


I would like to begin by asking you a few questions about purchases and
uses of fluid milk in your family (household)?

1. Who buys the milk for your family (household)?

I I Wife __I Husband | Wife and Husband i Other (specify: )

2. (Hand te.pondent ca.d 1)

Of the following three statements which one best describes your milk
purchasing behavior?


a. I buy only one brand of milk.


I b. I buy one brand of milk more often than any other.


I c. I do not pay attention to brands when buying milk.

3. (a) Do any members of your family (household) consume other beverages
in place of fluid milk?


I Yes: Can you name some of these beverages? (Go to pazt b)




I_ No: (Go to question 4)

(b) Why do they consume these beverages in place of fluid milk.








CARD 2
TABLE 1


Fluid Purchases and Prices Paid for Fluid Milk Products at Various Outlets, Last Week
Milk
Products Supermarkets Convenience Gas stationsther (specify:
Stores )

quarts at ea. quarts at ea. _quarts at ea. quarts at ea.
half- half- half- half-
Whole s at ea. gallons at ea. at C ea. at c ea.
Milk gallons gallons gallons --gallons
gallons at ea. gallons at ea. gallons at ea. gallons at ea.


quarts at ea. quarts at __ ea. quarts at ea. _quarts at ea.
half- half- half- half-
Skim at C ea. gllntat eat ea.
Milk gallons gallons gallons e gallons
gallons at ea. gallons at ea. gallons at ea. gallons at ea.


quarts at ea. quarts at ea. quarts at ea. quarts at ea.
Fortified half- half- half- half-
low-fat gallons at ea. gallons at ea. gallons gallons
(2%) milk gallons at ea. gallons at ea. gallons at ea. gallons at C ea.


Quarts at ea. quarts at ea. quarts at ea. quarts at ea.
Chocolate --
half- half- half- half-
Milk and allows at ea. allows at ea. allows at ea. al at ea.
Chocolate gallons gallons gallons gallons
Chocolate
Drink gallons at ea. gallons at ea. gallons at C ea. gallons at ea.

quarts at ea. quarts at ea. _quarts at ea. quarts at ea.
half- half- half- half-
Butter- gallons at ea. gallons at ea. gallons at ea. allows at ea.
gallons gallons gallons gallons
milk
gallons at ea. gallons at ea. gallons at ea. gallons at __ ea.







4. We would like some information on the kinds of milk products consumed at
home by your family (household), the prices you pay for milk and where
you buy milk. (Hand Table 1 to respondent--card 2.)

For each milk product shown in Table 1 would you please tell me the amounts
your family (household) purchased for home use and the prices paid last
week; that is Monday through Sunday.
Hand Table 1 (continued) to respondent--card 3.
For the following milk products please give me the amounts actually con-
sumed by your family (household) and the prices paid last week.


TABLE 1 (continued)

Consumption and Prices Paid for Other
Other Milk Milk Products at Various Outlets, Last Week
Other Milk
Products Convenience Other (specify:
Supermarket Stores )


cans cans cans
Canned Milk weight per can weight per can weight per can
per can per can __ per can



Dried or packages packages packages
Powdered weight per pkg. weight per pkg. weight per pkg.
Milk
__ per package __ per package __ C per package



5. Approximately what portion of your family's (household's) weekly purchases
of milk is consumed as a beverage? (Milk not consumed as a beverage
includes milk that is used with cereals, coffee, tea, in cooking, etc.)


6. How much milk did your family (household) consume away from home last week?








7. Suppose a new milk-type product having
value as whole milk and made with some
of whole milk was sold in Gainesville.


the same tast and nutritional
but not all of the ingredients


(The new product uses vegetable fat in place of butteLfat.)

(a) Would you try this new product if it cost the same as whole milk?



(b) Would you try this new product if it cost 6 to 8 cents less per
half-gallon than whole milk?



(c) If tepondent answered "Yes" oa "Maybe" to paitt a oa b, ask:
Suppose after trying this new product you found that your family
liked it. How would this influence your future purchases of the
new product and your future purchases of whole milk?




That completes the questions about your purchases and use of milk.


PART II

We are also trying to determine how milk purchases are influenced by such
things as size of family, occupation, income and other factors. So I would
like to turn now to a few questions about your family (household).
Record:


Type of Residence:

Ethnic Background:


I-I
I-
H


Apartment


I I House


Caucasian (White)


Negroid (Black)


Mongoloid (Yellow)









8. How long has your family lived in Gainesville?




9. (a) How many adults (including yourself), children and other members
now live at this address?




(b) (AWk only married people): What are your children' ages and sex?


10. What is the occupation of the head of this household?




11. (Hand ruepondent card 4)

Which of these seven categories best describes the highest level of
education achieved by the head of this household? (check only one)


I I Under 5th grade

I I 5th to 8th grades

I I 9th to llth grades

II High school graduate


5. I I Some college
r
6. __I College graduate

7. I I Graduate school


Male Female



Age






12. (Hand respondent card 5)
Which of the following income categories best describes your family's
(household's) total income before taxes and other deductions? (check
only one column.)
Annual Monthly Weekly

I- less than $2,999 I less than $249 i less than $58

I $3,000 $4,999 l_ $250 $417 $59 $96

1 $5,000 $6,999 i $418 $583 $97 $135

I $7,000 $8,999 0 $584 $750 I $136 $173

i_ $9,000 $13,999 $751 $1,167 $174 $269

I- $14,000 $18,999 I $1,168 $1,583 I $270 $365

I- $19,000 $23,999 I $1,584 $2,000 I $366 $461

II more than $24,000 -I more than $2,00 _I more than $462


Who

I I
I -- I
I I

I I


anvweAed thki quetionnaire (check one):


Wife

Husband

Wife and Husband

Other (specify:


















APPENDIX B


APPENDIX B--Mean values of the variables used
elasticities


to compute average


Tables 6 and 7

Units

pounds
pounds
pounds
cents
cents

number

number

number

number

number

0 or 1

0 or 1

0 or 1


4ean Value

18.660
3.866
1.787
13.940
220.164

2.170

0.460

0.290

0.249

0.126

0.569

0.174

0.078


Variable

QW
QO
QA
PW
EW


N1

N2

N3

N4

N5

Y2

Y3
E


Tables 8 and

Units

pounds
pounds
pounds
cents
cents

number

number

number

number

number

0 or 1

0 or 1

0 or 1


9

Mean Value

17.842
5.797
1.888
13.839
215.457

2.169

0.498

0.302

0.262

0.136

0.553

0.177

0.081


Variable

QF
QcD
QA
PF
EF


N1

N2

N3

N4

N5

Y2

Y3
E


~




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs