MESSAGE 0031 mamas mo mam Mamas i ‘

on swacmg EXPOSURE

V Thesis for the Degree of M. A.-
‘ MICHIGAN STATE UNEVERSITY .-
Z ] JOHN J. :GALLOWAY '
x- 1973 '

ABSTRACT

MESSAGE COST FACTORS AND THEIR
INFLUENCE ON SELECTIVE EXPOSURE

BY

John J. Galloway

This study investigated the effects of what are
termed message cost factors on patterns of selective
exposure Sgainformation. An individual who selects a
message to bolster his prediSpositions, usually needs to
expend some personal resources, such as money, time, or
energy. Certain messages are perceived as more costly
than others to obtain and process. This study varied
cost attributes of an otherwise identical message. An
attempt was made to investigate the effects of such
attributes on selectivity.

A field experiment was conducted in which pre-
disposition measures were obtained toward a two-sided
issue. Letters were sent to all subjects making pamphlets
available on both sides of the issue. The content of the
letters was identical, except for the procedures outlined
by which pamphlets could be obtained: (1) letters with
postage-paid reply cards, (2) letters with reply cards

requiring stamps, and (3) letters providing addresses to

John J. Galloway

enable letters of request to be written and mailed.
Subjects were systematically assigned to the three
message cost conditions.

Two hypotheses were examined. The first predicted
lower overall willingness to expose oneself under higher
message costs. The second and main hypothesis predicted
an interaction effect, such that selective exposure would
be observed more under higher than low-levels of message
costs. The first hypothesis was confirmed. The second
was not confirmed. The frequencies were too small in the
experimental treatment cells of interest to adequately
assess the findings. Even so, selectivity under high
message costs was nearly twice that observed in the low
condition. Further research would appear warranted.

Only a low overall rate of reSponse to the letters
was Obtained. This was attributed to low saliency of the
issue. Little overall selectivity was observed. These
factors mitigated against the randomization procedures.
They also prevented analysis of the extent to which the
observed selectivity was in fact reinforcement seeking,

or de facto selectivity.

MESSAGE COST FACTORS AND THEIR

INFLUENCE ON SELECTIVE EXPOSURE

BY

John J. Galloway

A THESIS

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

MASTER OF ARTS
Department of Communication

1973

AC KNOWLED GME NT S

I wish to acknowledge the advice, assistance, and
friendship of Dr. Verling C. Troldahl, my thesis advisor.
He has contributed much to both the development of this
thesis and my own educational growth. Sincere thanks also
to Dr. Charles Atkin, who encouraged me to explore a topic
I was interested in and gave friendly advice on a variety
of problems.

Data collection was made possible through the pro-
vision of funds by the College of Agriculture and Natural
Resources Education Institute at Michigan State University.
I am particularly grateful to two members of the Institute.
Dr. Mason Miller helped both in early planning and in the
latter stages as a member of my guidance committee. The
special efforts of Dr. Carroll Wamhoff through the original
design and data collection stages are greatly appreciated.
Dr. Lawrence Sarbaugh provided my original contact with
Institute.

Finally, a very special "thank you" to my wife,

Wendy.

ii

TABLE OF CONTENTS

Page
ACKNOWLEDGMENTS . . . . . .

LIST OF TABLES . . . . .

Chapter
I. THE RESEARCH PROBLEM . . . . . . . .

1
Introduction . . . . . . . . . . 1
Cost Factors . . . . . . . . . . 4
Communication . . . . . . . . . 5
Other Behavioral Areas . . . . . . 5
7

Toward a Definition of Message Costs .
Theories and Costs . . . . . . . . 13
Learning Theory . . . . . . . . 13
Social Exchange Theory . . . . . . l4
Dissonance Theory . . . . . . . . 14
Related Research . . . . . . . . . l7
Reward Values and Costs . . . . . . 22
Conflicting Findings . . . . . . 23
Hypotheses for Present Study . . . . . 24
II. RESEARCH DESIGN . . . . . . .- . . . 26
Overview . . . . . . . . . . . 26
Description of Sample . . . . . . . 26
Data Collection . . . . . . . . . 27
The Issue . . . . . . . . . . 27
The Questionnaire . . . . . . . . 28
The Letters . . . . . . . . . 29
Operationalization of Variables . . . . 29
Criterion Variables . . . . . . . 31
Control Variables . . . . . . . . 32
Data Analysis . . . . . . . . . . 34

iii

Chapter

III. FINDINGS . . .

Description of Sample
Description of Respondents to the Letters.
Hypotheses . .

Past

History, Confidence, and Social
Utility . .

IV. DISCUSSION . .

Implications .

BIBLIOGRAPHY

APPENDICES .

Appendix
Appendix

Appendix
Appendix
Appendix

Appendix

A--The Questionnaire . . . .
Arguments in Favor

B--Pamph1et:

C--Pamphlet:

D--Letter (with cards): Low Cost
Condition.

E--Letter (with car

Cost Condition

F--Letter:

ds): Medium

High Cost Condition

iv

of
A Beef Marketing Tax
Arguments Against A
Beef Marketing Tax .

Page
35
35
39
44
46

54

59

66
67
74
83
91

94
97

10.

LIST OF TABLES

Effort Manipulations in Ten Studies . . .

Preference Rating Changes Under Different
Levels of Effort . . . . . . . . .

Exposure Patterns to Different Information
Sources . . . . . . . . . . . .

The Message Cost Manipulation . . . . .

Some Characteristics of the Sample, and
Comparisons of Respondents to the Letters
to Nonrespondents . . . . . . . . .

Effects of Message Costs on Overall Message
Selections . . . . . . . . . . .

Some Comparisons of Respondents to Letters
Offering Low-cost Messages with Respondents
to Letters Offering High-cost Messages . .

Effects of Message Costs on Type of Message
selected 0 O O O O O O O O O O 0

Effects of Message Costs on Type of Message
Selected: Post-Decisional Respondents
only I O O O I O O O O O O O 0

Effects of Message Costs on Overall Message
Selections: Post-Decisional Respondents
on 1y 0 O O O O O O O I O O O O

Page

l8

19

21
31

36

39

41

42

44

44

CHAPTER I

THE RESEARCH PROBLEM

Introduction

 

After reviewing the selective exposure literature,
Freedman and Sears in 1965 concluded that there is no
general psychological preference for supportive information
as against discrepant information, and that attempts by
researchers to clarify conflicting findings were inadequate.
Up to that time the proposition had not been seriously
challenged; for example, to Klapper (1960) it was of prime
importance in his analysis of mass communication effects.
The main selective exposure proposition, stated here in
terms of dissonance theory (Festinger, 1957), holds that
following a decision individuals are motivated to reduce
dissonance by seeking information which supports their
initial opinions and to avoid information which is opposed
to them.

It should be noted, however, that there are other
ways to reduce dissonance, and that at times there can be
other motivations to seek or avoid information; both points
having been argued prior to 1965 (e.g., Festinger, 1957,
and Sears and Freedman, 1963). Even so, the general
psychological preference interpretation had not before been

1

seriously challenged, no doubt because it "somehow seems

to deserve to be true . . ." (McGuire, 1968a), and because
of numerous field studies (e.g., Lazarsfeld, §E_§l., 1948;
Star and Hughes, 1950) which had shown that people do tend
to expose themselves disprOportionately to mass media
messages which reinforce their views. Freedman and Sears
conceded the occurrence of these voluntary exposure patterns,
and in a later review (Sears and Freedman, 1967) under-
scored the necessity to isolate factors which produce de
facto selectivity (Sears and Freedman, 1967).

From an instrumental (as opposed to incidental)
perspective, individuals seeking information appear to
expose to messages primarily because of gratifications and
utilities expected from them. Sears and Freedman and other
investigators suggest that reinforcement of predispositions
may be only one such need; utilitarian needs such as
general surveillance, decision guidance and task performance
may be others. In this thesis, only the supportive-
nonsupportive dimension which is the concern of selective
exposure investigators will be examined in detail, although
recognizing that the choice of a message may have reward
values on other dimensions. Also, message handling
behavior will be examined more from the viewpoint of
assumed instrumental (means-ends) seeking of information,
rather than its incidental acquisition or active avoidance.

This does not suggest that message cost factors are of no

influence in the latter cases; on the contrary it would
seem that they are. However, the analysis here will be
limited mainly to consideration of the instrumental types
of behavior.

Supportive information may be perceived as
potentially rewarding to an individual; however, he will
first need to obtain and process it. Thus, in the analysis
of the message acquisition process, from a temporal per-
spective the perceived costs of obtaining and processing
the message intervene between what are the two usual data
points in selective exposure studies: a measure of pre-
dispositions and the actual selection behavior. The two
data points are usually defined (dichotomously) as con-
sistent or inconsistent, i.e., along the reward dimension
of supportiveness. Therefore, perceived costs to the
individual in this process need to be considered as well
as, and in relation to, the rewards of the message. The
examination of these cost factors and their influence will
be the major objective of this thesis.

This is a preliminary attempt, however, and more
work is needed to evaluate the concepts and their relation-
ships. It will be suggested that message cost factors do
add some clarity to conflicting findings (e.g., those
reviewed in Freedman and Sears, 1965). Certain limited
empirical support from the study reported here will be

presented, and directions for future research examined.

Cost Factors

 

If a person goes to some trouble or effort to obtain
and process a particular message, it is often assumed he
must value the information that he thinks the message has
for him more than if he were not willing to expend that
amount of effort; his behavior indicates that the message
was "worth" it. Given several messages involving moderate
levels of effort to obtain and process, presumably he will
see some as "worth" it and some not. The unselected
messages may not be actively avoided, nor of no value;
rather the chosen messages are valued more (whether the
value derives from reinforcement, utility, or other motives),
and "worth" the effort expenditure. If the perceived cost
is high, but relative to the reward value not so high as
to prevent any message selection, the knowledge of expendi-
tures of resources introduces additional dissonant COgni-
tions, i.e., additional to whatever decisional dissonance
there may be were costs only minimal. To reduce the
increased magnitude of dissonance, the individual becomes
more likely to choose supportive than discrepant
dissonance-arousing messages. On the other hand, following
a choice between messages where little dissonance has been
aroused and there are only minimal expenditure in message
selection, an individual may not be motivated to choose
selectively; instead he may expose to either or both sides

of an issue.

Communication

 

In 1954 Schramm suggested that message selection
behavior could be conceptualized in terms of the ratio of
an individual's expectation of reward to the effort
required to obtain and process alternative messages. This
"fraction of selection" has been variously cited (e.g.,
Berlo, 1960), but neither Schramm nor other communication
scholars who have indicated its potential have examined
the concept in depth at theoretic or empirical levels.
Whether rewards and costs do bear this particular relation-
ship is unclear. Whereas rewards have been widely investi-
gated by communication researchers, including those con-
cerned with selective exposure, cost factors in general,
and message cost factors in particular, have been relatively
neglected. In inquiries outside the message selection or
exposure area, such as in counterattitudinal advocacy,
the effort variable has been operationalized. Some of
these studies will be examined under the heading Related

Research.

Other Behavioral Areas

 

Selective exposure researchers often measure pre-
dispositions toward an issue by paper and pencil tests and
then examine overt selection behaviors to observe the
consistency between pairs of responses. In other areas
of behavioral research the "consistency" question has been

argued in terms of an interaction with a hierarchy of costs.

For example, in psychology Campbell (1963) suggests apparent
inconsistencies can be explained in terms of "thresholds"
of response effort or difficulty. He examines a number

of research findings, including the classic La Piere (1934)
study with the Chinese couple who were refused accommoda-
tion in 0.4 percent of places visited as against 92.5
percent refusal rate in paper and pencil reSponses.
Campbell considers different normative thresholds were
involved: ". . . it is very hard to refuse a well-dressed
Chinese couple travelling with a European in a face-to-
face setting, and very easy to refuse the Chinese as a race
in a mailed questionnaire" (p. 160). Similarly with
Minard's (1952) study of prejudice among Pocahontas coal
miners. Campbell notes that it was easy (low threshold)
for miners to express prejudice in the town, but it was
more difficult for them to do so in the mines because of
social-distance factors.

Political participation can be conceptualized in a
similar way, in terms of effort expenditures, the range of
behaviors, and the consistency between them. For Milbrath
(1965) "levels of participation constitute a hierarchy of
costs. Time and energy costs are least for the activities
at the bottom of the hierarchy. Behaviors higher in the
hierarchy obviously require a greater expenditure of

energy . . . (p. 19). "There seems to be a kind of

threshold that must be crossed . . ." (p. 20). At levels

of minimal participation, measures of individual predispo-
sitions are liable to show less consistency with the
partisan vote, than at higher levels. Campbell suggests
that "If the items (in hierarchical effort responses) are
all symptomatizing the same diSposition, then the person
who shows the disposition in the situation with the highest
threshold should show it most frequently" (p. 158). Con-
flicting findings in selective exposure studies in
laboratory settings typically involve minimal effort in
message selection, i.e., it is easy for subjects to state
preferences for supportive or discrepant messages, or to
select both if there is an opportunity to do so.

Toward a Definition of
Message Costs

 

 

Consider an individual seeking information (for
reinforcement or some other reward value) and given a
choice between two identical messages, except that one is
perceived as more costly to obtain or process. Other
things being equal, he will undoubtedly select the low-cost
message. The difficult or costly alternative psychologically
implies nonaction or nonselection.l

Suppose now a similar situation where the messages

have high and low reward values, and the costs are held

 

1This argument is adapted from discussion on the
behavior of rats in T-maze experiments in Lawrence and
Festinger, 1962, pp. 38-39.

constant. Other things being equal the individual will
undoubtedly select the high reward message. In this
case, however, the low reward message does not psycho—
logically imply nonaction or nonselection. Rewards are
not deterrents to selection, but incentives; costs are
clearly both deterrents and dissonance arousing in
instrumental seeking situations, since they psychologically
imply the obverse of message selection. This is in accord with
Festinger's (1957) definition that ". . . two elements are
in a dissonant relation, if considering these two alone,
the obverse of one element would follow from the other"
(p. 13). The deterrents aspect has implications for
selective exposure investigations where message prefer-
ences are merely rank-ordered (e.g., Rosen, 1961; Mills,
§E_gl., 1959). The main selective exposure empirical
evidence, though not at all clear, leans more in favor
of the supportive part of the prOposition than the non-
supportive. Most studies have examined incentives to
selection with deterrents held constant at minimal or
near-zero levels. These procedures do not reflect most
individuals' day-by-day selection behaviors among messages
which have deterrent as well as incentive attributes.
Lawrence and Festinger (1962) distinguish between

deterrents and incentives as follows:

If one increases the magnitude of some variable and

observes hesitation of behavior, one may conclude

that the variable is a deterrent. If one decreases

the magnitude of the variable and then observes

hesitation of behavior, one may conclude that it

is an incentive (p. 162).
They term the distinction "crucial" for dissonance theory;
one would not want to say that any incentive, however
small, introduces dissonance. However, we need to be
cautious in generalizing to human behavior, and with rats
in T-maze experiments the distinction at the level of
operations may be clearer to make than in humans. Even
so, the distinction seems useful. In this analysis,
costs are conceptually and operationally defined as
deterrents, including message costs to be discussed
shortly.

In addition to being deterrents, however, and in
agreement with Steiner (1970), costs are viewed as affect-
ing an individual's probability about an outcome and of
being contingent upon available resources: "The costs an
individual incurs include all those expenditures of effort,
time, social capital, money, or other resources that are
believed by the individual to affect his probability of
obtaining an outcome" (1970, p. 191). Cost considerations
thus precede an outcome such as selection and acquisition
of information, and at an operational level in research
they follow the predisposition measures. In the message

selection process, therefore, cost factors are defined as

intervening.

lO

Confronted by a set of objects, a person dis-
criminates among them by evaluating various attributes of
the objects (Chaffee and Tipton, 1969); and attributes
are defined as the characteristics, traits or qualities of
the object as the person perceives them (Donohew and
Palmgreen, 1971). Message selection usually takes place
on the basis of summary statements (letters or notices,
verbal statements, message coverleaf, program guide, etc.)
about actual alternative messages. Outside the laboratory,
such summary statements include both reward and cost
attributes of the message to which they refer.

Reward attributes can includes supportiveness,
utility, etc. Cost attributes of the message seem likely
to include monetary expense, inaccessibility (spatial or
temporal), ambiguity, incomprehensibility, complexity,
etc. From an instrumental message-seeking frame of
reference, and from a functional analytic one, the cost
attributes of a message are mostly contingent attributes.

An individual considers the reward value of a
message in relation to the personal expenditures in money,
time, or physical or mental energy to obtain and process
it. Whether or not message cost attributes are perceived
to involve expenditures too great will depend partially on
the individual's resources of money, time, and energy.

Obvious correlates of these resources are the individual

11

difference variables of education, age, race, socio-
economic status, sex, etc. These have been widely
stressed by selective exposure reviewers as likely to
affect selection behavior (for example, Sears and
Freedman, 1967).

Which message attributes should be considered
message costs? Can expenditures of different types of
resources be directly observed, or must they be merely
assumed? In examining these questions, the acquisition
of messages will be distinguished from their processing
or decoding. Monetary and inaccessibility attributes
will be viewed as closely concerned with message acquisi-
tion. Ambiguity, incomprehensibility, and complexity,
will be subsumed under a higher-order concept. Further-
more, they will be viewed as primarily related to message
processing rather than message acquisition.

Monetary attributes of a message obviously involve
expenditures of money resources to obtain the message.
Similarly, messages perceived as relatively inaccessible
to obtain will involve greater expenditures of energy or
time. In a Spatial context, expenditures of energy may be
directly observed: for example, in terms of different
distances walked or driven, or even in the mere scanning
of voluminous material to obtain a desired message. In a
temporal context, time expenditures may involve different

durations or waiting or scheduling.

12

‘It has been suggested that expenditures of resources
may be incurred where the message content is perceived as
ambiguous, incomprehensible, or unduly complex. Expendi-
tures here are likely to be in terms of mental rather than
physical energy. They are thus less amenable to direct
observation. These expenditures may also be in terms of
whether an individual possesses sufficient time. Ambiguity,
incomprehensibility, and complexity, probably are suffi-
ciently unidimensional for them to be subsumed under a
higher-order concept, noise. In information theory, noise
is usually defined from a source viewpoint, and as a
property of the channel rather than the message. In
message selection situations, however, noise from a
receiver orientation may often be perceived as a message
prOperty. Research has shown adverse effects on the per-
formance of receivers in decoding noisy messages (for
example, in Smith, 1966, pp. 275-321). It seems likely
that such effects can be partially anticipated by an
individual in perceiving various attributes of a message.
Noise clearly involves difficulty in processing. It is
likely to be a deterrent to selection. And it affords a
more parsimonious conceptualization of a number of possible
cost attributes, such as ambiguity, incomprehensibility,
complexity, etc.

To summarize, message costs are defined as

attributes of a message which act as deterrents to selection

13

or exposure, are usually regarded as contingent in nature,
and appear to intervene in the process of instrumental
selection or exposure. Message attributes which may be
included are monetary expense, inaccessibility, and
decoding noise. An analysis of message costs also
necessitates taking into consideration the individual's
resources of money, time, and energy. The key concept
which relates the cost attributes of the message to the
resource properties of the individual, is expenditures.

An individual is seen as most likely to incur expenditures
only where perceived reward attributes in some sense

"outweigh" perceived cost attributes.

Theories and Costs

 

Learning Theory
Lawrence and Festinger (1962) summarize learning
theory viewpoints as follows:
Some theories of learning have proposed that effort
increases response inhibition. Other theories have
been inclined to regard effort as a negative incen—
tive. But, in one form or another, most learning
theories include the prOposition that increased
effort weakens the response strength (p. 6).
This suggests that the greater the cost, the fewer the
messages of any type that will be chosen. What has been
said so far is consistent with this view but goes beyond
the gross exposure or response rate. Those selection

responses within the higher cost categories are likely

to be made only if the reward value is high. And where

14

the dimension of reward is supportiveness--given that
there is dissonance and this is an apprOpriate strategy
to reduce it--more supportive messages will be chosen

at these levels than at lower levels.

Social Exchange Theory

 

Social exchange theorists (e.g., Homans, 1950;
Adams, 1965) have viewed social interactions as analogous
to economic exchanges, in which effects obtain from
relationships between outcomes and investments or rewards
and costs. However, greater weight is given to the
principle of equity in interacting with others with
respect to outcomes and investments, than to effects of
rewards and costs in instrumental seeking behavior by the
individual. To Homans, social interaction is governed by
desires for "distributive justice." Distributive justice
obtains when each person in the interaction achieves
rewards proportional to costs. That rewards and costs
are in some sense commensurate will be discussed under
the heading Reward Values and Costs. Homan's definition

of costs is discussed below.

Dissonance Theory

 

Festinger (1957) discusses two major determinants

of the magnitude of dissonance. The first, in part, is:

15

Other things being equal, the more important the

decision, the stronger will be the dissonance.

Thus, a decision to buy one automobile rather than

another will result in more dissonance than a

decision to buy one brand of soap rather than

another . . . (p. 37).
Here Festinger, at least implicitly, suggests the relevance
of reward values assuming that cars have higher reward
value for most peOple than soap. But, do cars not also
cost more than soap?

The second major determinant of post-decision
dissonance is the relative attractiveness of the unchosen
alternative (p. 37). This appears to be almost identical
with Homan's (1962) definition of cost: "Cost is value
foregone" (p. 288). Economists (see e.g., Samuelson, 1967)
usually define different types of cost, including what
they term "opportunity" cost which appears comparable
here to Festinger's second determination of dissonance
and Homan's definition. Thus, it is possible to infer
that costs are major determinants of dissonance, such that,
other things being equal, under high-cost conditions there
should be more post-decisional dissonance than under low-
cost conditions.

A number of strategies by which dissonance can be
reduced have been mentioned by Festinger (1957). Three
of these are:

1. Increasing the attractiveness of the chosen

alternative, decreasing the attractiveness
of the unchosen alternative, or both.

l6

2. Changing attitudes to be consistent with

overt behavior, particularly under forced
compliance and counterattitudinal conditions.

3. Seeking consonant (supportive) information.

Of course different situations may restrict free choice
among the strategies, and certain ones are more applicable
to particular study objectives: e.g., decision-making,
counterattitudinal advocacy, disconfirmation of expectancy,
forced compliance, listening to a persuasive communication,
etc. However, as well as being alternatives at the level
of theory, there is empirical support for:

a. Their functional equivalence, at least between
changing attitudes and seeking supportive information.
Clarke (1966) found that "individuals who changed their
beliefs were significantly more likely to seek supportive
information, than were subjects who remained unconvinced"
about the contents of an atLacking message. He concluded,
"Belief change and selective information seeking appear
to be functional equivalents . . . ."

b. Cost variables acting to intensify dissonance
in two of the three modes: first, attractiveness of the
chosen alternative (see e.g., Aronson and Mills, 1959;
Gerard and Mathewson, 1966) and second, changing attitudes
(see e.g., Zimbardo, 1965; Wickland §E_§l., 1967; Cohen,
1959). These studies will be examined below.

That cost factors are influential in two of three

functionally equivalent strategies for reducing dissonance

17

suggests that they could be in the third also, i.e., in

the seeking of supportive information.

Related Research

 

In previous studies experimental manipulations of
cost factors have mostly involved subjects' expenditures
of mental and physical effort, and time. Unfortunately,
however, none has used criterion measures of message
selection let alone selective exposure. Furthermore, in
only one study (Cohen, 1959) could it be said that message
costs were manipulated.

Cohen had all subjects read a counterattitudinal
communication. Half had been told simply that they would
expend a great deal of effort to understand the communica—
tion while the other half were told they would not have to
expend much effort to understand it. Subjects anticipating
greater effort expenditure expressed more attitude change
consonant with the counter-communication.

Effort manipulations in ten different studies are
summarized on the next page (Table 1). In each case it
would appear that greater dissonance was aroused in high-
cost conditions, and greater dissonance reduction took
place in high-cost conditions. In all studies, except
one, this meant a greater shift in preferences or attitudes

for high-cost subjects.

l8

 

owcmno
monufiuum Hmummno

Aummuv wsoouso owuommxw
CH mmflamn pmmmmuocH

mmsmno
moduwuum Hoummuu

ADHOMMm Hmnuom 0cm omumm
IHoHucm Hmossv mxmmu mo
mmmcm>fluomuuum Hmummum

wmcmno
mospflupm Hmumwuu

moouw mo
mmmcm>flpomuuum “mucosa

mmcmno
monuwuum nmumwuw

Aummuv
mEoouoo omuommxm
CH mwflamn ommmmuocH

mzoum mo
mmmcm>fluomnuum Hmummuw

mmocmuomwum
as wocmumflmcoo muoz

.cowuflmom Hmsflosufluumumucsou
mcflmfl>oumsw umaflss xomnomwm
anomfiosm ca Swamp mo sumcmn

ammo m mxmu on mcofiumummmum

coaumoficdasoo
Hmcflpsufluumumussoo

mcflummn on Howum mafia msfluwmz

xmmu Hmsuom Any
can .xmmu omummfloflucm Ame
How mcofluoahumcw m0 >uaxwamaou

mucwsmumum
Hmcflosufluumumucsoo mcflpoocm
cw cospmmw>oumafl mo mmummn

adoum m oucfi
coaumfluficfl mm mxoozm cauuomam

mGOADQmUme xmmu
mummuo on msoHuosuumcfl commuam

ummu m wxmu ou mcoflumummmum

msoum m ODCw coflumfluwcfl
mm mouoz osmomno mowomwm

AunOMMm soav umsmme gawk
mcfiuomamm Ho .AHHOmmm amaze
mumswmucoo Hmumfi you :mcflnmwm:

womoo>om
Hmcflmsufluumumucsou

mosmuowmxm
mo cOwDMEuflmcoomHa

coflumoflsssfioo
Hmsflosufluumnmucsoo
ou ocfismumflq

moccaamfioo pmouom

humoo>©m
Hmcflosufluumumuasoo

mommafimfioo Umouom
momoo>pm

HmcHUDUAuumumucsou

mosmuommxm
mo coflumEuHmcoomHQ

mommaameoo Umouom

mcflmeIGmeHomo

Amomav ownmnEHN

lemmas uwmcflummm
pcm smnumw

isomac “means can
smmoou .ocmflxoflz

repose xofluocmm
can Hmmmmnm

Aommav
meson mam mcwm

Abomav c0m3m£umz
ocm oumumo

immmav :mnoo
Aoomav mmom

mam poozouum

Ammmav
mHHHz paw concoum

AHmmHv COmGOH<

 

muasmmm

.coflumHsdflcmz puomum

>psum mo mama

“onusa

 

.mwwwsum awe cfl mcowumHsmwcmz uuommmnn.H mamas

19

This exception (Aronson, 1961) is worth noting,
since it was the closest to a free-choice message selection
study. Unlike the others it involved instrumental
decision-making and no counter-communications. Effort
was varied in relation to subjects obtaining rewarded and
unrewarded stimuli. One frequent interpretation of this
study (see e.g., Carlsmith and Freedman, 1968) is that,
on unrewarded trials, dissonance was reduced by increasing
the perceived attractiveness of empty metal containers
(which sometimes included money rewards). However, when
the high and low reward data are collapsed, as is shown
in Table 2, the change scores in preference ratings in
the two effort conditions show much greater shifts in the
Easy condition than in the Effortful condition. In other
words, the greater the effort expenditure in obtaining
rewards (high or low), the more stable or consistent were

the preferences with stated predispositions.

TABLE 2.--Preference Rating Changes Under Different
Levels of Effort.*

 

 

Easy Effortful
Mean + .97 + .02
Standard Deviation 1.18 .92
N 30 3O

 

*
From Aronson (1961).

20

A somewhat similar "consistency" effect was found
in experiments on rats by Lawrence and Festinger (1962).
As effort during behavior acquisition was increased,
resistance to extinction increased.

This was true whether one was dealing with 100
percent reward situations or with partial reward
situations. This uniformity of effect, however,
was not found in connection with an incentive
such as magnitude of reward. . . . Clearly,
incentives and deterrents do not produce the same
results and must be distinguished operationally
and conceptually (p. 163).
The well-known imprinting studies of Hess (1962) with
birds found greater behavioral consistency in terms of
resistance to extinction with increased effort expendi-
tures.

Correlational data in a field study of exposure
patterns to different message sources suggest that cost
factors were influential. Rogers and Atkin (1972) sur-
veyed female college students about their attitudes toward
hypothetical unwanted pregnancies and abortions, and
their rates of exposure to different information sources
on abortion. Their findings are presented in summarized
form in Table 3.

A cost-factors interaction interpretation necessi-
tates an assumption about attributes of messages from
different possible sources, specifically in this case about
monetary expense and inaccessibility attributes. Messages

from information sources such as electronic media and

close interpersonal ones are more accessible and lower

21

TABLE 3.--Exposure Patterns to Different Information
Sources.*

 

Attitudes Toward

Source of Hypothetical Abortion D1fference

Information Between

About Abortions In Favor Against' Groups

 

 

Friends

Boyfriend/husband

Roommate

Acquaintances 20% 21% 1%
Mother

Television

Radio

Magazines

Newspapers

Medical column in
newspapers

43% 33% 10%

Sex education books
Pamphlets 45% 27% 22%
Medical personnel

N=ll4 N=33

 

*
Adapted from Rogers and Atkin (1972).

in monetary expense (require fewer expenditures of effort
and money) than messages obtained and processed from the
printed media, in particular books, and from professional
interpersonal sources. A corollary to this assumption

from an information-avoidance rather than instrumental
seeking perspective (which is our main focus here) is

that greater resource expenditures are necessary to avoid
messages from some sources rather than others. The
interpretation outlined is clearly Open to question in
survey type data; message cost factors need to be

manipulated for a more adequate assessment.

22

Reward Values and Costs

 

In selective exposure studies the supportiveness
of a particular exposure alternative is often confounding
with other factors which may be rewarding for the indi-
vidual. The "most common" (Sears, 1968) confounding factor
is utility, defined as any information which is perceived
to help in the performance of a role or the successful
completion of a task (Katz, 1968). Where de facto
selectivity occurs, Sears (1968) suggests that the various
reward attributes of messages--supportiveness, utility,
interest, truth value, etc.--are often correlated. However,
in any one instance, whatever the reward attributes of a
particular message are, there will be some perceived costs
in obtaining and processing it. The higher the reward
value, the more likely it is that a person will be willing
to expend greater resources to obtain and possess it.
Hence, the two may covary to some extent and costs may
actually provide some sort of operational index of rewards.

Steiner (1970) writes:

Good ghings cost much; and things that cost much

are good. But it is not only the Puritan ethic

and the writings of Horatio Alger that maintain

this stance. Blau (1964), for example, contends

that love and approval that are easily gained

tend to be lowly valued (p. 201).
Steiner then reviews several studies (e.g., Gerard and
Mathewson, 1966; Aronson and Mills, 1959), and concludes

that "payoffs are commensurate with costs" (p. 202), and

suggests that

23

. . . peOple who are required to work inordinately
hard or endure uncomfortable circumstances in order
to obtain a goal tend to evaluate the payoff

(after it is obtained) more highly than do those
who are not required to incur such heavy costs

(p. 202, italics supplied).

 

Steiner is not concerned here with free-choice situations.
Presumably, if they were free-choice, heavier costs would
not be incurred, unless the payoff (reward) was highly
valued before it was obtained. Given free-choice situa-
tions therefore, such as in many selective exposure
studies, costs may in some sense provide an Operational

index of relative reward values.

Conflicting Findings

 

A basic question in laboratory selective exposure
studies has been: "If a person is given a choice between
supportive and nonsupportive information, will he prefer
exposure to the former, all other things being equal?"
(Sears and Freedman, 1967). However, the ceteris parabus
condition does not obtain in field situations where the
strongest evidence for selective exposure is usually found.
Messages in field situations do have different contingent
expenditures in effort and time to obtain and process them.

Freedman and Sears (1965) reviewed the findings
of seventeen laboratory studies on preferences. Typically,
laboratory procedures measure a subject's opinion on an
issue and then ask him which of several communications he

would like to read or hear. The dependent variable is

24

often really a measure.of interest in supportive, relative
to nonsupportive, information (Rhine, 1967)- It is
usually only a verbalized or check-marked expression of
preference (and, incidentally, often not altogether free-
choice because of "demand" characteristics--Orne, 1962).
In Campbell's (1963) terms the threshold of effort
or difficulty is easy or minimal; in Milbrath's (1965)
terms the response is at the bottom of the hierarchy of
costs; and, as noted earlier, in free—choice situations,
responses involving little expenditure of resources tend
to be ambivalent. Furthermore, Festinger (1964) in
discussing Brehm and Cohen's (1962) emphasis on commitment,2
distinguishes between merely stating preferences among
alternatives where no consequences follow, and making a
decision among them. Hence, had cost factors been
manipulated or held constant at higher than the near zero
levels, the supportive side of the main selective
exposure prOposition may have been evidenced more often

in these studies.

Hypotheses for Present Study

 

This study will focus on two criterion variables:

(a) frequency of selections, and (b) frequency of types of

 

2The commitment variable, however, has been found
wanting in empirical support (see e.g., Mills and Ross,
1964; Sears and Freedman, 1963).

25

selections--supportive only versus both sides of the
issue. The independent variable will be message costs.

H1: More information will be sought under
low-cost conditions than under higher-
cost conditions.

H2: Information on both sides.of an issue will
be sought under low-cost conditions, but
supportive information will be preferred
over both sides under higher-cost conditions.

The first hypothesis concerns gross levels of message
acquisition. It is, in effect, a manipulation check of

message costs. The second is an interaction hypothesis

and the main one of interest in this study.

CHAPTER II
RESEARCH DES IGN

Overview

In a field experiment in lower-central Michigan,
farmers were interviewed to obtain their opinions on a
prOposal to introduce a marketing tax on beef cattle.
Approximately two weeks later they were sent letters
offering pamphlets on both sides of the issue, but varying
certain cost attributes of the pamphlets. The requests
for information by respondents to the letters were noted
according to their consistency or otherwise with earlier
stated opinions in the interviews, and according to dif-

ferent message cost conditions.

Description of Sample

 

Strictly speaking, the study dealt with a popula-
tion of farmers who: (a) had at some time in the past
requested beef cattle publications from the Department of
Animal Husbandry, Michigan State University; (b) were listed
in telephone directories in the following locations
accessible via University-rented Watts lines: Lansing,
Grand Rapids, Ann Arbor, Detroit, Pontiac, Battle Creek,

and St. Johns; (c) self-identified as "farmer," including

26

27

those who were retired or only part-time; (d) were able

and willing to be interviewed within five or fewer attempts
at contact. The 303 farmers interviewed, however, could
also be viewed as comprising a judgementally selected non-
probability sample. This view was taken in preference to
considering them strictly as a population. This was
principally because the particular few hundred farmers

were selected mainly for reasons of convenience and
limitations of funds, rather than for purposes of possible
inferences about behavior being applicable only in this

highly limited population.

Data Collection

 

The Issue

 

The issue chosen centered on the desirability of
instituting a marketing tax for beef cattle, with the funds
collected to be used for promoting the beef industry. This
proposal was before the Michigan State Legislature, to
introduce a tax of 30¢ or 40¢ per head, depending on weight,
on all cattle sold in Michigan. Pretesting of the question-
naire and advice from agricultural experts suggested that
(a) this was a two-sided issue that the sample was likely
to be about equally divided on, and (b) it would be a
relevant tOpic in that a fair proportion of persons would

have some prior knowledge of it through farming cooperatives

28

and the farming press. Three further issues1 were

selected for inclusion in the questionnaire for the
following reasons: (a) to "mask" the main issue, i.e.,

to help avoid sensitizing individuals to inferring a link
between questions on the beef tax issue and later being
offered pamphlets about it, and (b) to provide agricultural
experts at Michigan State University with data they had

indicated would be useful in their work.

The Questionnaire

 

In mid-August, 1972, a questionnaire (see Appendix
A) was administered to members of the sample by telephone.
It had first been pretested on twelve randomly chosen
farmers with characteristics similar to those included in
the sample. Eight work-study student interviewers were
paid with funds provided by the College of Agriculture and
Natural Resources Education Institute. They attended
interview briefing sessions lasting one and a half hours
before their first interviews. Instructions were to
interview only the person whose name appeared on the
coversheet. However, if after a minimum of three call-

backs the designated person was still not available but

 

1These issues were also being considered by the
State Legislature. They were: unionization of farm workers,
collective bargaining in agriculture, and a Specific land
tax instead of the present property tax. The data in this
study were collected for personnel in the Cooperative
Extension Service and the Department of Agricultural
Economics at Michigan State University.

29

another family member was who was apparently quite
knowledgeable about the farm's Operations, they were to
be interviewed and the fact noted. The questionnaire's
introduction stressed that this was a study of farmers'
opinions, and if any person disagreed about the applica-
bility of the label "farmer," interviewers were instructed
to terminate the interview. Several questions related

to each of the four issues, and those on the beef tax
proposal were placed third in the questionnaire. Finally,
a number of demographic questions were asked: age,
education, number of cattle owned, and the respondent's

sex.

The Letters

 

On August 29, 1972, letters from the College of
Agricultural and Natural Resources Education Institute were
sent to all members of the sample offering to make avail-
able what were stated in the letters as "free informational

pamphlets,’ one of which was entitled "Arguments in Favor
of a Beef Marketing Tax" (see Appendix B), and the other
"Arguments Against a Beef Marketing Tax" (see Appendix C).

The letters were identical except for message cost factors.

Operationalization of Variables

 

Using a variant of Milgram's lost letter technique
of inquiry, Forbes and Gromoll (1971) dropped stamped and

unstamped letters in several cities and towns. Their

30

overall return rates were 79% for stamped letters and 45%
for unstamped letters. They suggested that the incon-
venience accounted for the differences.

In the present study a similar operationalization
was employed but with cost factors being message properties
rather than situationally bound. The sample was syste-
matically divided into three equal sized groups of 101.
Message cost was operationalized for the group as different
levels of effort and expense needed for persons to select
informational pamphlets offered through the mail by varying
the monetary expense and inaccessibility attributes of the
message. As mentioned, all the letters were identical
in content, but varied in the procedures by which pamphlets
could be obtained: letters with postage-paid reply cards
(see Appendix D), letters with reply cards requiring
stamps (see Appendix E), and letters providing addresses
to enable letters of request to be written and posted
(see Appendix F). These conditions are summarized in
Table 4. Separate addresses were used within each
condition so that greater expenditures of effort and
monetary expense would be required to select more than
one message. The two addresses, Fee Hall East and South
Kedzie Hall, both on MSU campus, had neutral connotations
compared with possible addresses such as Agricultural Hall

or Anthony Hall (location of Department of Animal Husbandry).

31

TABLE 4.--The Message Cost Manipulation.

 

 

 

 

 

Coidition Letter Procedure for Obtaining Pamphlets
Low lOl Offered "in favor" Letter enclosed separate postage—
and/or "against" paid reply cards to:
pamphlets (a) address A to obtain an "in
favor" pamphlet
(b) address B to obtain an
"against" pamphlet
Medium 101 Offered "in favor" Letter enclosed separate reply
and/or "against" cards requiring stamps to:
pamphlets (a) address A to obtain an "in
favor" pamphlet
(b) address B to obtain an
"against" pamphlet
High 101 Offered "in favor" Last paragraph of letter informed
and/or "against" that letters of request could be
pamphlets written and posted to:

 

(a) address A to obtain an "in
favor" pamphlet

(b) address B to obtain an
"against" pamphlet.

 

Criterion Variables

 

There were two criterion variables: frequency of
selections, and frequency of selections of message types.

Frequency of selections was defined simply as the
gross number of respondents who made any request for
pamphlets; this was the measure for the test of the first
hypothesis.

Message types were determined according to con-
sistency or inconsistency between predisposition measures
and the exposure or selection behavior; this was the

measure for the test of the second hypothesis. (Selecting

32

a pamphlet in favor of the beef tax was behavior consistent
with a favorable predisposition toward the issue, but
inconsistent with an unfavorable predisposition toward the
issue; and vice versa for selecting a pamphlet against the
beef tax.) Preceeding the predisposition measures in the
interview, individuals were asked whether or not they had
heard Of the beef tax proposal.2 If they had not heard of
it they were given a few key points about the prOposal and
then asked their opinions for or against.3 Those who had
heard Of it were asked if they had reached a decision
(Festinger, 1957, proposes dissonance as a post-decisional
state only), and, if so, were they for or against the
issue.4 The combined Opinion responses Of those who had
heard and those who had not heard Of the issue represented

the predisposition measures.

Control Variables

 

In addition to obtaining demographic data, measures
on three variables were Obtained to examine their possible
influence on the relationships hypothesized. Although the
evidence is by no means clear-cut, there have been some
studies which suggest that they influence selective

exposure rates:

 

2See Appendix A, question 10.

3See Appendix A, question 14.

4See Appendix A, question ll(a).

33

a. Past history on an issue (see e.g., Sears and

 

Freedman, 1967). In the present study this was operation—
alized simply as whether or not individuals had heard of
the issue.5

b. Confidence in decision (see e.g., Canon, 1964,

 

and Mills and Ross, 1964, who found that persons lower in
confidence seek supportive information more than those high
in confidence). This was operationally defined here by
asking individuals who had made up their minds how certain
6

they were that their decision was the correct one.

o. Social utility (see e.g., Clarke and James,

 

1967, who found that persons who expect to discuss an issue
with others are more likely to seek supportive information).
Individuals in this study were asked if they expected to
discuss the issue with any farmers in the next few months,
and if so, did they expect them to be mostly for or against
the issue.7

Finally, measures on the following demographic
variables were Obtained: years of formal education, number

of cattle owned, age, and sex.

 

5See Appendix A, question 10.
6See Appendix A, question ll(b).

7See Appendix A, questions 12, 13.

34

Data Analysis

 

The data were at the nominal level of measurement
only. Simple descriptive statistics were used for most of
the analysis. Chi-square tests of significance were
computed, where appropriate and where a sufficiency of
data permitted. The first hypothesis was tested by com-
parison with the Chi-square distribution. Relationships
relevant to the second hypothesis were described in
percentages. Characteristics of respondents to the letters
in relation to characteristics of nonrespondents were

analyzed through the use of contingency tables.

CHAPTER III

FINDINGS

Description of Sample

 

In Chapter II it was noted that all members of the
sample had previously requested beef cattle publications
from the Department of Animal Husbandry at Michigan State
University and resided within seven particular telephone
areas. The number of such theoretically possible persons
to include totalled 542, but this figure was reduced to a
sample of 303 studied. Attrition reason with numbers in
parentheses were as follows: unlisted telephone numbers or
numbers untraceable from address only (96), disconnected or
incorrect numbers (30), deaths (7), persons who had been
included in a questionnaire pretest (12), persons who did
not identify as "farmer" (36), refusals (32), and failure
to make Contact after four or more call-backs (26).

With the exception of Detroit and Pontiac areas,
the sample was fairly well distributed by area with the
greatest number (about one third) in any one area residing
in or near Ann Arbor.

Some personal characteristics of sample members

are shown in Table 5.

35

36

TABLE 5.--Some Characteristics of the Sample, and Compari-
sons of Respondents to the Letters to Non-
respondents.

 

 

 

 

 

 

 

 

 

 

Respondents Nonrespondents Total
to Letters to Letters Sample
Education:
Under 8 years 3% 16% 15%
9-12 years 47 56 55
13-16 years 41 22 24
17 years and more 8 5 6
100% 100% 100%
Cattle Ownership:
No cattle 11% 30% 28%
1-15 24 17 18
16-39 27 21 22
40-99 19 14 15
100-299 8 12 11
300 and more 11 6 6
100% 10 % 100%
Age:
Under 30 years 3% 6% 5%
30-39 22 14 15
40-49 19 22 21
50-59 30 25 26
60 years and older 27 34 33
100% 100% 100%
Sex:
Male 100% 95% 96%
Female 0 5 4
100% 100% 100%
Heard of Issue:
Had heard 14% 21% 20%
Had not heard 86 79 80
100% 100% 100%
N=37 N=266 =303

 

errors are involved.

Note: Entries do not total 100 percent where rounding

37

Nearly three-quarters owned beef or vealing cattle
and would thus be subject to the beef tax if it were to
come into effect. A third owned 40 or more head of cattle,
and an isolated few (3) owned as many as 1,000. The
largest percentage of cattle owners had between 15 and 40
head.

Over half the sample had received formal education
to within the last few years of high school. Between one
quarter and one-third had completed some college education.

More than half the sample was over 50 years of age,
and about one-third was over 60 years. Only one-fifth of
the sample was younger than 40.

Not surprisingly, the farmers were nearly all
men (95%).

Only one-fifth had heard of the beef tax proposal
before, even though there had been a fair coverage of the
issue in the farming press.

Description of Respondents
to the Letters

 

 

Only 12% (37) of the sample requested any pamphlets.
Such a low response rate suggests that respondents may have
differed systematically on certain attributes from the
non-respondents. Some comparisons in percentages are

shown in Table 5.

38

Respondents to the letters differed significantly
from non-respondents in the amount of formal education
they had received,1 and in whether or not they owned any
cattle.2

Nearly half the respondents to the letters had some
college education compared with about one-quarter of the
nonrespondents. About 90% of the respondents owned some
cattle as against 70% for the nonrespondents. However,
among those in the sample who were cattle owners, the two
groups did not differ beyond chance expectations with
regard to numbers of cattle owned. Education and cattle
ownership were considered Lo be indicators of socio-
economic status, and on only these two demographic
variables did the two groups differ significantly; no
differences were observed in age, nor in having heard about
the issue beforehand. It should be added that none of the
13 females in the sample responded to the letters.

Although this sex difference cannot be tested readily for
significance3 it appears appreciable--100% of respondents
to the letters were males compared with 95% of the non-

respondents.

 

1Chi-square = 6.95, d.f. = 1; p> .01.

2Chi-square, with correction for continuity = 5.27,
d.f. = l; p > .05.

3For a chi-square test the smallest expected fre—
quency should not be less than 5 (Cochran cited in Siegel,
1956, p. 110); in this case it is 1.6. And factorials
involved are too large for an Exact test.

39

Hypotheses

 

Two hypotheses were tested. The first was a
straight forward one sample test Of the effects of message
costs on overall selection behavior; this was confirmed.
The second and major hypothesis predicted an interaction
effect between choosing supportive messages and message
costs; insufficient data precluded its acceptance or
rejection, but the data direction was as hypothesized.

H1: More information on an issue will be

sought under low-cost conditions than
under higher-cost conditions.

Frequencies of selections by each respondent to

the letters within each message-cost condition are shown

in Table 6.

TABLE 6.--Effects of Message Costs on Overall Message
Selections.

 

Message Costs

 

 

Low Medium High Total
Percent 64.9 32.4 2.7 100%
F's 24 12 1 N=37

 

The differences are highly significant (p > .001),4 and

thus H1 is confirmed.

 

4Chi-square, with correction for continuity =
19.63, d.f. = 2.

40

In order to examine whether persons who responded
to the letters within the different cost conditions had
different demographic characteristics or not, the medium
and high-cost frequencies were collapsed and termed the
high-cost group (N=l3). This was compared with the low-
cost group (N=24). Proportions within each group on
several characteristics are shown in Table 7.

Compared with respondents to the letters at the
low-cost level, high-cost respondents had the following
characteristics: they were more likely to be younger,
slightly more likely to have had a college education,
and were less likely to be owners of cattle or to have
heard about the issue beforehand.

H2: Information on both sides of an issue

will be sought under low-cost conditions,
but supportive information will be
preferred over both sides under higher-
cost conditions.

The frequencies of message type selected by each
respondent within cost conditions are shown in Table 8.
Only 16% of all the respondents chose selectively as
against 78% choosing messages on both sides of the issue.
Slightly more selections for both pamphlets occurred
under low-cost than high-cost conditions, and supportive
pamphlets were preferred nearly twice as much by

respondents in the high-cost as in the low-cost condition.

The observed relationships cannot be readily tested for

41

TABLE 7.--Some Comparisons of Respondents to Letters
Offering Low-cost Messages with Respondents to
Letters Offering High-cost Messages.

 

Respondents Respondents

 

 

 

 

 

 

 

 

Who Sought Who Sought Total
Low-cost High-cost Respondents
Messages Messages
Education:
High school
or less 54% 46% 51%
Some college
or more 46 54 47
100% 100% 100%
Cattle Ownership:
No cattle 8% 15% 11%
1-40 54 46 51
41 or more 38 38 38
100% 100% 100%
Age:
Under 40 years 17% 38% 24%
40-60 42 62 49
60 years and Older 42 0 27
10 % 100% 100%
Heard of Issue:
Had heard 17% 8% 14%
Had not heard 83 92 86
100% 100% 100%
N=24 N=l3 N=37

Note: Entries do not total
errors are involved.

100 percent where rounding

42

TABLE 8.--Effects of Message Costs on Type of Message

 

 

 

 

Selected.
Low Cost High Cost Total
% f % f % f
Supportive pamphlet
only 12.5 3 23 3 16.2 6
Discrepant pamphlet
only 4.2 l - - 2.7 1
Both pamphlets 79.2 19 77 10 78.4 29
Neither supportive*
nor discrepant 4.2 l - - 2.7 1
TOTAL 100% N=24 100% N=13 100% N=37

 

*
S's predisposition was "not sure," and he chose an
"in favor" pamphlet.

Note: Data collection took place under three cost condi-
tions: low, medium, and high. The high-cost
condition in this table represents the original
medium and high categories combined, since there
was only one respondent in the original high-cost
category; that respondent selected supportively.

statistical significance.5 However, at the descriptive
level only, the main hypothesis receives partial support

bearing in mind the low frequencies.

 

5For a chi-square test the smallest expected
frequency should not be less than 5 (Cochran cited in
Siegel, 1956, p. 110); in this case examining the "sup-
portive" and "both" categories only in relation to cost
levels, it is 3.8. Factorials would be too large for
an Exact test.

43

Respondents who selectively exposed as against
those who exposed to both sides of the issue did not differ
greatly on demographic characteristics. However, the fre-
quencies are too small and unstable for any conclusive
statements about them.

To the extent that the observed reinforcement
information-seeking derived from dissonance rather than
other motivations, the hypothesized relationships should
hold for post-decisional persons, rather than those whose
predispositions were "not sure" or "not really decided
yet." Persons in the pre-decisional state totalled 81 in
the sample and 11 in the respondents. Ten of these 11
selected both pamphlets--6 in the low-cost, and 4 in the
high-cost condition--and the remaining person chose an
"against" pamphlet. Table 9 examines the main interaction
hypothesis in terms of post-decisional message selection
behavior only.

A comparison of Table 9 with the data in Table 8
shows much the same pattern of relationships.

Table 10 examines the first hypothesis in terms
of post-decisional overall selection behavior. A compari—
son of Table 10 with Table 6 shows a similar pattern with

the observed differences, again significant (p > .01).6

6Chi—square, with correction for continuity =
13.01, d.f. = 2.

44

TABLE 9.--Effects of Message Costs on Type of Message
Selected: Post-Decisional Respondents Only.

 

 

 

 

 

Low Cost High Cost Total
% f % f % f
Supportive pamphlet
only 17.6 3 33 3 23.1 6
Both pamphlets 76.5 13 66 6 73.1 19
Discrepant pamphlet
only 5.9 1 - - 3.8 1
TOTAL 100% N=l7 100% N=9 100% N=26

 

TABLE 10.--Effects of Message Costs on Overall Message
Selections: Post-Decisional Respondents Only.

 

Message Costs

 

 

Low Medium High Total
Percent 65.4 30.8 3.8 100%
f 17 8 1 N=26

 

Past History, Confidence,
and Social Utility

Insufficient data prevented analysis of the
hypothesized relationships in terms of these three
variables. Only five of the 37 respondents to the letters

had heard of the issue, and they were about evenly

45

distributed throughout the different treatment cells.

Only one of the 37 respondents to the letters was uncertain
about the decision he had reached on the issue. Finally,
only two respondents to the letters expected to discuss

the issue with any other farmers in the next few months.

CHAPTER IV

DISCUSSION

A field experiment was designed to assess the
effects of message costs on selective exposure to informa—
tion. Two hypotheses were examined: one predicted effects
on overall exposure rates, and the other--the main
hypothesis--predicted an interaction effect between message
costs and selectivity.

Findings came from a quite small sub-sample of
respondents (i.e., those who made any requests for informa-
tion), and an even smaller number of respondents who
requested selectively. Thus, the problem of the "dis-
appearing" sample so common in mailed surveys and the
possibility of individual difference variables being
unduly influential are questions raised about the results
obtained.

Indeed, respondents to the letters were higher in
education and socio-economic status (measured as cattle-
ownership) than nonrespondents, a finding not uncommon for
individuals who voluntarily expose to supportive communi-
cations (e.g., Star and Highes, 1950). Furthermore,

respondents who made relatively greater expenditures of

46

47

of resources to obtain and process information (high-

cost condition) differed from those who expended fewer
resources (low-cost condition): They were younger and
better educated, but were less often owners of cattle.
Even so, we cannot be sure about such characteristics

with so few persons in each treatment condition. But they
do point to a repeated theme in the literature, the
importance of demographic variables and the possibility

of observed selectivity being in fact de facto selectivity.
This and the small sub-sample size are points that should
be kept in mind in interpreting the results.

The experiment included provision to examine
separately and beyond the mainstream of the study, three
variables which have been evidenced, with some reservations,
to influence exposure patterns. They were past history on
the issue, level of confidence, and social utility. None
of the three could be adequately assessed in terms of
possibly facilitating selective exposure, again because of
low frequencies. However, as will be discussed shortly,
an absence of past history did appear to be associated
more with two-sided exposures than one-sided exposures.

Message-cost attributes of monetary expense and
inaccessibility were varied over three levels. The first
hypothesis, which also acted as a check of the manipulation,
was confirmed at high levels of significance for both

reSpondents and post-decisional respondents when considered

48

separately. The deterrent notion in the conceptualization
of message costs, and its correspondence at the level of
operations in terms of observed hesitancy, was well
supported. Clearly also the costs were attributes of the
message and not situationally located as in some of the
effort manipulations reviewed in attitude change studies.
However, further than that, H1 did not tell us a great
deal which could not be reasoned from common sense grounds.
The relationships of main interest were those in the
second hypothesis.

Because of low obtained frequencies the second
hypothesis was not tested for significance. However, the
results do point to an observed interaction in the right
direction. This was true in analyses both of all
respondents to the letters and post-decisional respondents
only. Thus, some tentative support is lent to the
hypothesis. However, on close examination of the original
wording, supportive information was actually not, as
stated, "preferred over both sides under higher-cost
conditions"; rather "information on both sides of an
issue" was preferred under both high and low costs.

Little overall selectivity occurred, only six out Of 37
cases (16%). However, as mentioned before, these few
cases were disproportionately located in the high-cost
condition which is clearly in the predicted direction.

A low level of precision, particularly in terms of the

49

small sub-sample size, can therefore be implicated for
failure to obtain more clear-cut results.

A follow-up telephone survey of a small number
(16) of randomly selected respondents to the letters and
nonrespondents was carried out for two purposes: to
examine reasons for the low response rate in the sample
as a whole (12%), and to assess what degree of sensitiza—
tion occurred from the "before" measures. With regard to
the response rate, it appears that it was not an unusually
busy time of the year for most farmers, although as had
been found out in administering the main questionnaire, it
was a period during which a number of farmers took their
vacations or were preparing for them. Presumably this
accounted for some nonresponses to the letters. The beef
tax issue was ranked only about fourth in interest relative
to five other agricultural issues mentioned. This is con-
sistent with few in the sample (20%) having heard of the
issue, and with the general finding that interest and
exposure are correlated (e.g., Rhine, 1967). Thus, either
the pretest sample which had been asked whether or not
they had heard of the issue was biased or too small and
the responses unstable, or the opinions of agricultural
experts on likely knowledge and saliency of the issue were
inaccurate, or both. The possible linking of administra-

tion of the questionnaire to the receipt of letters offering

50

information, thus sensitizing responses, did not appear
to have been a problem.

As mentioned before, the first hypothesis was in
effect a check on the message-cost manipulation. In the
three equal-sized treatment groups of low, medium, and
high message costs, the response rates to the letters
offering information were 24%, 12%, and 1% respectively.
Only the rate in the low-cost treatment group is approxi-
mately comparable with obtained rates in other studies.
But even at this level they are not fully comparable,
since rather than merely offering information the more
usual practice is to carry a persuasive appeal for pur-
poses of increasing response rates. And, of course,
acquisition of information (e.g., supportive) is only one
way to satisfy a need (e.g., dispositional reinforcement).

Dissonance theory has been criticized for its
imprecise definitions of fundamental concepts and vague
rules of correspondence with the empirical plane (e.g.,
Pepitone, 1966). The presence of dissonance must be
inferred; in most studies it is simply assumed. This study
was no exception. No attempt was made to gather data
either on the presence or magnitude of dissonance in
relation to the extent of support-seeking. One field
experiment has made such an attempt (see Troldahl, 1963).
However, on the basis of an overwhelming preference for

information on both sides of the issue (78%), much

51

selection behavior appears to have been motivated more
by utilitarian needs, such as surveillance or decision
guidance (see e.g., Blumler and McQuail, 1969), than
reinforcement needs.

This alternative explanation of utilitarian needs
overriding reinforcement as the main source of rewards is
given support from an examination of respondents’ past
histories on the issue and their education. First, on
past history Freedman and Sears (1965) cite data from
their jury studies that

. . . a two-sided presentation . . . is preferred
most by individuals who have not been previously
exposed to the partisan arguments of either side
. . . [on this basis they suggest that] any
information which will help the naive individual
become familiar with both sides of the issue will
be favored, regardless of whether or not it
supports his initial preference (p. 85).

Eighty-four percent of respondents to the letters
who had not heard of the issue beforehand (N=32) exposed
to both sides, whereas among those who had heard of the
issue beforehand (N=5), 40% exposed to both sides. With
suitable reservations about the paucity of data, some
support is given the Freedman and Sears argument. Second,
a greater amount of education was observed among respond-
ents than nonrespondents (49% of the former against 28%
oftjualatter had some college education). One effect of
more years of education may be a wider tolerance for

divergent ideas with accompanying behaviors of nonpartisan

selection and exposure.

52

Several aspects of the research design can be
examined in the light of the findings. Sears (1968, p. 780)
points out that very few selective exposure experiments have
been conducted "outside of college sophomores and other
staunchly middle-class groups." Numerous field studies have
been conducted in other populations, but, as Kerlinger
(1964) points out, ex post facto research lacks control of
independent variables which may account for the dependent
variable under observation. As well there is the
inability to randomize to treatment groups. Thus, plausible
explanations such as a general psychological tendency to
prefer supportive information, can be inferred if
respondents assign themselves into groups on the basis of
common characteristics.

Katz (1968) suggests the "ideal test" of selective
exposure should take account of time-order, and "involve a
situation in which communications are available on both
sides of an issue, and where there is a corresponding
division of Opinion" (p. 795). Several reviewers (Sears
and Freedman, 1967; Katz, 1968; McGuire, 1968b) have called
for the testing of interaction hypotheses rather than main
effects and direct observations of exposure rather than
reliance on self-reports.

The present field experiment has met these various
criteria, but yet a basic problem remains: was the
selectivity observed reinforcement-seeking or de facto

selectivity?

53

Because of the extent of sample shrinkage mitigating
against the randomization procedures, individuals could and
did group according to common characteristics: in the
case of respondents versus nonrespondents, on education and
socio-economic status; in the case of high—cost versus low-
cost, on age and education; and, unfortunately in the case
of the 2 x 2 design, cells contained too few respondents
for reliable demographic crossbreaks.

The low response rate appears to have resulted
primarily from the comparatively low saliency of the issue.
Thus, the present design seems to have had merit, but an
issue of greater interest would most likely have given more
clear-cut findings, not only as regards statistical sig-
nificance, but also data interpretation in terms of pure
reinforcement-seeking versus de facto selectivity.

As mentioned earlier, selective exposure experi-
ments typically involve measures at two points in time,

T and T and observe whether the two are consistent or

1 2'
inconsistent. At a higher level of abstraction this of
course accords with Festinger's definitions about the
state of two cognitions, one of which "follows from," or
is "the obverse of" the other, considering these two alone.
But in field conditions the ceteris parabus conditions

rarely obtain, and it has been argued that cost factors

interact with the consistency/inconsistency dimension such

54

that not only is there more frequently observed consistency
at higher levels of cost, but there is also a greater degree
of consistency. In most selective exposure experiments,
including the present one, the key variable of consistency
(and of course its T1 and T2 component measures) has been
treated only dichotomously rather than continuously. "The
extent to which two different estimates or measures are
correlated" is one definition of convergent validity
(Fishbein, 1966, p. 217). Where the "given variable" is
"disposition toward an issue" and the two measures are at

T and T as above, convergent validity appears quite

l 2

suited as a measurement approach to conceptualization of
consistency. For convergent validity, the T1 and T2
measures must each have construct and discriminant validity,
reliability, and preferably the highest level of measure-
ment possible. This and most other selective exposure
studies have not paid sufficient attention to conceptual-
izing the key variables from a measurement theory per-
spective and operationalizing them accordingly. Conse-
quently it is likely that much variation in the dependent

variable is not accounted for, information is lost,

precision is low, and findings tend to remain inconclusive.

Implications

 

It would appear that cost factors in general, and
message cost factors in particular, have been much ignored

relative to the research efforts on various reward dimensions

55

of individual behavior. First, this study reports that
they do intervene to affect absolute rates of exposure.
Although advertisers have long known it is important to
construct messages which are as easy to obtain (and
process!) as possible, researchers have found out "little"
about the determinants of absolute exposure rates (Sears,
1968, p. 782). Other variables are no doubt involved, but
the findings here suggest that message costs certainly
should be taken into account. Second, these factors may
interact along various reward dimensions. Limited support
for this view comes from the data in the case of reinforce-
ment rewards; and, if, as has been prOposed (e.g., Sears,
1968) supportiveness correlates with utility, attractive-
ness, truthfulness, and other rewards, then it is likely
that message costs are similarly influential along other
reward dimensions and in other contexts of instrumental
information seeking.

Much research is necessary before any implications
can be suggested on a practical level. However, specula—
tion is possible on likely practical implicatiOns should
future research confirm the relationships hypothesized in
this study. Persons concerned with the construction of
one-sided messages directed at mass audiences may find
that the more costly messages are, the fewer people they
will reach, and the fewer peOple with contrary views they
will reach. Thus, if one's intention were to persuade,

the cost attributes would most probably need to be minimal.

56

Conversely, and apart from arbitrary circulation
restrictions, if one's intention were to reach only as many
people as possible with similar views to the one being
advocated, the message might be best designed so as to
include higher message costs. However, if message costs
were too high in relation to message reward values, it
might be that few peOple at all would be reached. Among
other factors, it appears that the saliency of the issue
to the intended audience would need to be taken into
account. In this study low saliency was suggested as a
primary reason for few people having been reached.

Chapter I noted that several researchers have
Observed that costs and rewards are often broadly com-
mensurate, and it was suggested that under certain condi-
tions they may covary. One such condition seemed to be
free-choice between alternative messages in instrumental
seeking situations. If there are constraints to free-
choice--beyond an experimental manipulation of message
costs--individuals may expend resources because of the
constraints (such as social conformity, etc.) rather than
the message reward value expected. Therefore, given free-
choice between messages and the possibility of message
costs and rewards showing covariance, message costs may
be worthy of intensive research in order to pursue the

goal of scaling message reward values.

57

Message costs can be manipulated and, it will be
suggested, they can be measured on interval scales.
Laboratory research along these lines would need to be
carefully designed to avoid a possible ceiling effect on
message costs. Both Weick (1964) and Collins (cited in
Aronson, 1966, p. 126) report such effects in laboratory
studies where levels of effort were varied. Collins is
cited as suggesting that it is difficult to demonstrate
laboratory differences in effort, since ". . . in most
psychological experiments subjects are trying extremely
hard, even in the low reward condition." Relatively free-
choice message selection, under different cost conditions,
may thus be difficult to Obtain in laboratory settings.

On the other hand, field studies do not take account of
time order. Therefore field experiments may have the
most potential for this type of study.

This study was a preliminary attempt to examine
message cost factors and their influence on selective
exposure patterns. To extend the applicability of
indications from the present findings and to more clearly
delineate the relationships involved, more research is
necessary. Such research might determine which attributes
of messages are clearly cost ones, and whether the
relationships can be expected to operate in similar manner

for different types and valences of message costs.

58

It was suggested earlier that selective exposure
variables are continuous underlying and that they should
be operationalized at the highest level of measurement
possible. The same can be said for message costs. Pre—
dispositions could be scaled rather than treated
dichotomously; exposure could be measured as rates per
unit of time, or pages read, etc.; and message costs could
be measured as expenditures in terms of distance or time,
or monetary expenditures, to obtain and process different
messages. The exact nature of the rewards/costs relation-
ship in instrumental seeking situations is as yet little
understood. Does it operate as a ratio as Schramm
suggests, and if it does often do so what are the condi—
tions that obtain? In this investigation message costs
did affect overall rates of exposure. Beyond this gross
level it was suggested that message costs might interact
along the reinforcement-seeking dimension, and that one
of the conditions that could obtain is for there to be
free choice between "reasonably" salient messages. However,
more research is needed since an insufficiency of data in
this study did not enable conclusions about these

relationships.

. .~..._. ‘9...“4 an
- u .

div“ l“.
H._

La]

.,
Err. . 53.2 .

i||

. u .
111‘“ .31.... 51' 511%.!-

BIBLIOGRAPHY

59

Adams, J. S. _
in Experimental Social Psychology. Edited By
L. Berkowitz. VOIT 2. New York: Academic Press,

Aronson,

Aronson,

Aronson,

Aronson,

BIBLIOGRAPHY

"Inequity in Social Exchange." Advances

 

1965.
E. "The Effect of Effort on the Attractiveness
of Rewarded and Unrewarded Stimuli." Journal of

Abnormal and Social Psychology, 1961, 63:375:385.

 

E.

"The Psychology of Insufficient Justification:

An Analysis of Some Conflicting Data." Cognitive
Consistency. Edited by S. Feldman. New York:

 

 

Academic Press, 1966.

E.

"Dissonance Theory: Progress and Problems."

Theories of Cognitive Consistengy: A Sourcebook.

 

Edited by R. P. Abelson, et a1. Chicago:
Rand-McNally, 1968.

E.,

and Carlsmith, J. M. "Experimentation in

Social Psychology." Handbook of Social Psychology.
Edited by G. Lindzey and E. Aronson. Vol. 2.
Reading, Mass.: Addison-Wesley, 1968.

 

Aronson, E., and Mills, J. "The Effect of the Severity of
Initiation on Liking for a Group." Journal of
Abnormal Social Psychology, 1959, 1774181.

Arrowood,

 

 

John A., and Ross, Lee. "Anticipated Effort and

Subjective Probability." Journal of Pers. and
Social Psychology, 1966, 4:57-64.

 

 

Berlo, David K. The Process of Communication. New York:

Blau, P.

Holt,

 

Rhinehart and Winston, 1960.

Exchange and Power in Social Life. New York:

 

Wiley, 1964.

Blumler, J. G., and McQuail, D. Television in Politics:
Its Uses and Influences. Chicago: University of
Chicago Press, 1969.

 

 

60

61

Brehm, J. W., and Cohen, A. R. Explorations in Cognitive
Dissonance. New York: Wiley, 1962.

 

 

 

Brodbeck, May. "The Role of Small Groups in Mediating the
Effects of PrOpaganda." Journal of Abnormal Social
Psychology, 1956, 52:166-170.

 

Campbell, Donald T. "Social Attitudes and Other Acquired
Behavioral Dispositions." Psychology: A Study of
a Science. Edited by Sigmund Koch. Vol. 6.
New York: McGraw-Hill, 1963, pp. 94-172.

 

Canon, L. K. "Self-Confidence and Selective Exposure to
Information." Conflict, Decision and Dissonance.

Edited by L. Festinger. Stanford, Califi:
Stanford University Press, 1964.

'""”'."5'"’"”‘" T"?
1?

 

 

 

Carlsmith, J. Merrill, and Freedman, J. L. "Bad Decisions 31
and Dissonance: Nobody's Perfect." Theories of 99
Cognitive Consistency: A Sourcebook. Editediby
R. P. ABéISon, et aI} Chicago: Rand-McNally, 1968.

 

Chaffee, Steven H., and Tipton, Leonard P. "Conflict,
Information Seeking and the Discriminating
Attribute." Journalism Monographs, 1969, 14:55-74.

 

Clarke, Peter. "Some Correlates of Selectivity in Informa—
tion Seeking." Paper delivered at Association for
Education in Journalism meeting, Iowa City, 1966.

Clarke, Peter. "Selective Media Exposure: A Review of
Current Findings." Bulletin de l'Aerie.
Yugoslavia, 1967.

 

Clarke, Peter, and James, Jim. "The Effects of Situation,
Attitude Intensity and Personality on Information-
Seeking." Sociometry, 1967, 3:235-245.

 

Cohen, A. R. "Communication Discrepancy and Attitude
Change." Journal Of Personality, 1959, 27:386-396.

Donohew, Lewis, and Palmgreen, P. "A Reappraisal of
Dissonance and the Selective Exposure Hypothesis."
Journalism Quarterly, Autumn, 1971, 412-420.

Feather, N. T. "Cigarette Smoking and Lung Cancer: A
Study of Cognitive Dissonance." Australian Journal
of Psychology, 1962, 14:55-64.

 

Festinger, Leon. A Theory of COgnitive Dissonance.
Evanston, II1.: Row, Peterson, 1957.

 

62

Fishbein, Martin. "The Relationships Between Beliefs,
Attitudes, and Behavior." COgnitive Consistency.
Edited by S. Feldman. New York: Academic, 1966.

 

Freedman, J. L. and Sears, D. O. "Selective Exposure."
Advances in Experimental Social Psycholgfy. Edited
by L. Berkowitz. Vol. 2. New York: Academic,
1965.

Freedman, J. L. "Preference for Dissonant Information."
Journal Pers. Social Psychology, 1965, 2:287-289.

Forbes, G. B., and Gromoll, H. F. "The Lost Letter Techniqo»
as a Measure oerocial Variables: Some Exploratory
Findings." Social Forces, 1971, 50:113-115.

 

Gerard, H. B., and Mathewson, G. C. "The Effects of
Severity of Initiation on Liking for a Group: A
Replication." Journal of Exp. Soc. PsycholOgY: 1966,
2:278-287.

Hess, E. H. "Ethology: An Approach Toward the Complete
Analysis of Behavior." New Directions in Psychology.
Edited by T. M. Newcombe. New York: Holt,
Rinehart, and Winston, 1962.

 

Homans, George C. The Human Group. New York: Harcourt,
1950.

 

Homans, George C. Sentiments and Activities. New York:
Free Press, 1967.

 

Jecker, J. D. "Selective Exposure to New Information."
Conflict, Decision and Dissonance. Edited by L.
Festinger. Stanford, Calif.: Stanford University
Press, 1964.

Katz, Elihu. "On Reopening the Question of Selectivity in
Exposure to Mass Communication." Theories of
Cognitive Consistency: A Sourcebook. Edited by
R. P. Abelson, et al. Chicago: Rand-McNally, 1968.

 

Kerlinger, Frederick N. Foundations of Behavioral Research.
New York: Holt, Rinehart and Winston, 1964.

 

King, B. T., and Janis, I. L. "Comparison of the Effective—
ness of Improvised vs. Non-improvised Role-playing
on Producing Opinion Change." Human Relations, 1956,
9:177-186.

 

VP.

\
'1‘. .
- .u‘mq
.

“pa-r .— i1”... mi f“"f""~“"' w-m- '1“:J'
.

63

Klapper, Joseph T. The Effects of Mass Communication.
New York: Free Press, 1960.

La Piere, R. T. "Attitudes vs. Actions." Social Forces,
1934, 13:230-237.

Lawrence, D. H., and Festinger, L. Deterrents and Rein—
forcement. Stanford, Calif.: Stanford—University

Press, 1962.

 

Lazarsfeld, P. F.; Berelson, B.; and Gaudet, Hazel. $32
People's Choice. 2nd ed. New York: Columbia
University Press, 1948.

 

McGuire, William J. "Selective Exposure: A Summing Up."
Theories of Cognitive Consistency: A Sourcebook.
Edited By R. P. Abelson, et a1. Chicago: Rand-
McNally, 1968a. """"

McGuire, William J. "The Nature of Attitudes and Attitude
Change." Handbook of Social Psychology. Edited
by G. Lindzey and’E. Aronson. Vol. 3. Reading,
Mass.: Addison-Wesley, 1968b.

Milbrath, Lester W. Political Participation. Chicago:
Rand-McNally, 1965.

Mills, J.; Aronson, E.; and Robinson, J. "Selectivity in
Exposure to Information." Journal of Abnormal
Social Psychology, 1959, 59:250-253.

Mills, J., and Ross, A. "Effects of Commitment and
Certainty Upon Interest in Supporting Information."
Journal of Abnormal Social Psychology, 1964,
68:552-555.

Minnard, R. D. "Race Relationships in the Pocahontas Coal
Field." Journal of Social Issues, 1952, 8(1):29-44.

Orne, M. T. "On the Social Psychology of the Psychological
Experiment: With Particular Reference to Demand
Characteristics and Their Implications." American
Psychologist, 1962, 17:776-783.

Pepitone, Albert. "Some Conceptual and Empirical Problems
of Consistency Models." Cognitive Consistency.
Edited by S. Feldman. New York: Academic, 1966.

Rhine, R. J. "Some Problems in Dissonance Theory Research
on Information Selectivity." Psychological

64

Rogers, Everett M., and Atkin, Charles K. "Patterns of
Information-Seeking." Unpublished proposal sub—
mitted to Department of Health, Education and
Welfare, National Institute of Child Health and
Human Development, National Institute of Health,
Bethesda, Maryland, 1972.

Rosen, S. "Post-Decision Affinity for Incompatible
Information." Journal of Abnormal Social
Psychology, 1961, 63zl88-l90.

 

 

Samuelson, Paul. Economics: An Introductory_Analysis.
7th ed. New York: McGraw-Hill, 1967.

 

Schramm, Wilbur, ed. The Process and Effects of Mass
Communication. Urbana, 111.: University of
Illinois Press, 1954.

 

 

 

Sears, D. O. "Biased Indoctrination and Selectivity of
Exposure to New Information." Sociometry, 1965.
Sears, D. O. "The Paradox of De Facto Selective Exposure

Without Preferences for Supportive Information."
Theories of Cognitive Consistency: A Sourcebook.
Edited by R. P. Abelson, et a1. Chicago: Rand-
McNally, 1968.

 

Sears, D. 0., and Freedman, J. L. "Commitment, Information,
Utility, and Selective Exposure." August, 1963.
USN Tech. Rep. (ONR), Nonr -233 (54) NR 171-350,
No. 12.

Sears, D. 0., and Freedman, J. L. "The Effects of
Expected Familiarity with Arguments upon Opinion
Change and Selective Exposure." Journal of Pers.
Social Psychology, 1965, 2:420-426.

 

 

Sears, D. 0., and Freedman, J. L. "Selective Exposure to
Information: A Critical Review." Public
Opinion Quarterly, 1967, 31:194-213.

 

Shaffer, David R., and Hendrick, Clyde. "Effects of Actual
Effort and Anticipated Effort on Task Enhancement."
Journal of Experimental Social Psychology, 1971,
7:435-447.

 

Siegel, Sidney. Nonparametric Statistics for the Behavioral
Sciences. New York: McGraw-Hill, 1956.

Smith, Alfred G. Communication and Culture. New York:
Holt, Rinehart and Winston, 1966.

 

5' L via-7t!

W§:‘--QI_ “ ._M. ‘.. -.f NI':‘ '. .V. '-.':
«A - . '
'D
‘u, ._
.__ .~-<
‘ ' A

65

Star, Shirley A., and Hughes, Helen. "Report on an
Educational Campaign: The Cincinnati Plan for the
United Nations." American Journal of Sociology,
1950, 55:389-400.

Steiner, Ivan D. "Perceived Freedom." Advances in Experi-
mental Social Psychology. Edited’by L. Berkowitz.
Vol. 5. New York: Academic Press, 1970, pp. 187-
248.

 

 

Troldahl, V. C. "Mediated Communication and Personal
Influence." Unpublished Ph.D. dissertation,
University of Minnesota, 1963.

Weick, K. E. "Reduction of Cognitive Dissonance Through
Task Enhancement and Effort Expenditure." Journal
of Abnormal and Social Psyghology, 1964, 68:533-539.

Wicklund, R. A.; Cooper, J.; and Linder, D. E. "Effects of
Expected Effort on Attitude Change Prior to
Exposure." Journal of Experimental Social
Psychology, 1967, 3:416-428.

 

Yarren, R. B., and Festinger, L. "Preparatory Action and
Belief in the Probable Occurrance of Future Events."
Journal of Abnormal and Social Psychology, 1961,
603-606.

Zajonc, Robert B. "Cognitive Theories in Social Psychology."
Handbook of Social Psychology. Edited by G. Lindzey
and E. Aronson. 2nd 63), Vol. 1. Reading, Mass.:
Addison-Wesley, 1968.

Zimbardo, P. G. "The Effect of Effort and Improvisation on
Self-persuasion Produced by Role Playing."
Journal of Experimental Social Psychology, 1965,
1:103-120.

Aym‘- KN“?
' ..

.n'

, 1......» “_..uo=gll'v
’I 4
.p

,r‘

‘ u

\~_‘ g,.
H 0

".

 

‘ “I; .'.

‘Egr‘D‘FT '
V~ l

APPENDICES

66

APPENDIX A

THE QUESTIONNAIRE

67

‘~.-o..ocn_ “a...“ q
l

_..-.. -...-...~—
4;- _.

Wt]—

.11 .‘I‘ '.
A l

AGRICULTURAL ISSUES STUDY

 

 

Interviewer's Name Date completedi

lst Call 2nd Call 3rd Call
No answer
# busy

Respondent
not home

Refused
interview

Call back at

 

Hello, may I speak to (respondent's name) please?

This is and I'm calling from the Department
of CommunicatIOn at Michigan State University. We're doing
a study, and we're interested in finding out what farmers
think on a number of agricultural issues. I'd like to ask
a few questions, and what you say will help us to put
together an overall picture from talking to a number of
farmers, but your individual answers won't be made known

to anyone.

 

 

1. First, do you raise or feed any cattle, for meat
purposes . . . and if so, about how many do you
normally sell each year?

 

# of cattle No cattle

 

 

 

68

69

Now, your opinion on the possible "unionization of
farm workers.". Do you think that full-time farm
employees are likely to be unionized within the next
ten years?

 

NO J Yes

1

In about how many years do you think that they :3
might be unionized? g

 

 

 

 

 

 

73 74 75 76 77 78 79 80 81 82
‘* l 2 3 4 5 6 7 8 9 10

 

 

 

 

 

 

 

 

 

 

 

 

 

.
*1....:-.. ..

wu-M...~t '. c A _‘i" ". a...
i

If farm employees do unionize, would their unionization
hurt, help, or have no effect on Michigan agriculture?

 

 

 

Hurt Help No effect

 

 

 

 

 

 

 

 

 

Why do you think this is so? [IF NEC., PROBE: What
reasons . . . .?]

[_— _..-l
L_ ___1

Do you expect to discuss this issue with any farmers
in the next few months?

 

 

Yes ’No ——————> GO T0 0.7

L

Would most of them think that unionization will hurt,
help or have no effect on Michigan agriculture?

 

 

 

 

 

 

 

 

 

 

Hurt Help No effect

 

 

 

 

 

 

 

 

7.

70

As well as farm workers organizing, there is also talk
of farmers becoming more organized . . . sometimes the
term "collective bargaining in agriculture" is used.
Have you heard of this term?

 

Yes

i

 

 

 

 

 

NO

 

 

——------—-9 GO TO Q. 10

When you hear "collective bargaining in agriculture,"
what do you think of, what does the term mean to you?
[IF NEC., PROBE: What do you think its main influences

are?]

T 7

L J

What should farmers do, if anything, to become more
effective in collective bargaining?

[__

I _-

L J

10.

11.

12.

71

There has been talk of a possible marketing tax on
cattle, i.e. a tax on all cattle sold in Michigan.
The money would be used by a State-wide commission
for promoting the beef industry.
anything about this proposal?

 

 

Yes

 

 

 

 

No

 

 

 

I

; GO TO Q. 14

Have you heard

Have you made up your mind already for or against
this tax, or haven't you really decided yet?
. . are you

[IF "yes" ANSWER, PROBE: Which way .

for or against?]

 

Not really

 

decided yet

 

 

 

v

 

 

For Against

 

 

 

 

 

 

I 1

I'd like to know just how certain you
are that the position you've taken

is the right one on this issue

do you as somewhat sure, or very

sure that this position is the

right one?

 

Somewhat sure

 

 

 

 

 

Very sure

 

 

I

What are your reasons for being (in
favor of/against) a tax like this?

[_—
L__

—_l
___1

Do you expect to discuss this issue with any cattle
owners in the next few months?

 

 

[Yes

 

 

 

No

.L

 

 

 

I

GO TO Q. 15

13.

14.

15.

16.

72

Do you expect most of them to be for or against a
tax like this?

 

 

 

 

 

 

 

Most for Most against Don't know
fi\\‘ 1 .//7

GO TO Q. 15

 

 

 

 

 

The proposal is that you would have to pay about

30¢ or 40¢ per head (depending on weight) for all
cattle that you sell. The money would then go to

a State-wide commission to generally promote the beef
industry. Do you think that you would be for, or
against, a tax like this, or aren't you sure how you
would feel about it?

 

 

 

For Against Not sure

 

 

 

 

 

 

 

 

 

Farmers now have to pay a property tax. There has

been talk that instead of a property tax, farmers
should pay a specific land tax. This would be

based on dollars per acre for different classes of
land. The rates would be set at $9 per acre for the
better farm land, $7 for the next best, and $5 and

then $3 for other classes. Do you think you would be
for, or against, a tax like this instead of the present
property tax?

For Against‘ Not sure

GO TO Q. 17

 

 

 

 

 

 

 

 

 

 

 

Would you commit your land to a tax like this, and

in doing so agree not to change its use for 5 years?
[IF NEC., ADD: That is, would you agree not to develop
it to another class of land, say for housing,
business, etc.--you could of course put in whatever
crops you want.]

 

 

 

Yes No Not sure

 

 

 

 

 

 

 

 

 

73

17. Finally, can you tell me your age group please . . .
are you in your 60's, 50's, 40's, 30's or 20's?

 

60's and above

 

 

50's

 

 

 

 

20's and below

 

 

 

18. And, how many years of formal education did you
finish?

grade year in college

THANK RESPONDENT

Code respondent's sex:

 

 

 

 

 

Male Female
L____.

 

APPENDIX B

PAMPHLET: ARGUMENTS IN FAVOR OF

A BEEF MARKETING TAX

74

ARGUMENTS IN FAVOR
OF A

BEEF MARKETING TAX

75

ARGUMENTS IN FAVOR OF A BEEF MARKETING

TAX (CHECKOFF OR ASSESSMENT)*

Preface: In a recent telephone survey, over 300 farmers

in southern Michigan were asked their views on the proposed
marketing tax and beef industry commission. Most of the
farmers contacted owned cattle. Approximately half were in
favor of the issue, and approximately half were against it.
From respondents' answers,tW0 separate mimeographs have
been prepared--this one, and one entitled Arguments Against
a Beef Marketing Tax (Checkoff of Assessment). In both, the
words and phrases that respondents actually used have been
retained as much as possible, consistent with avoiding
redundancy among the main themes that respondents put for-
ward--i.e., the following arguments are direct, or almost

direct, quotations chosen from the interview responses.

--Advertising works for other things, why not meat?
Similar schemes, and in some cases almost identical ones,
operate in the cherry, potato, apple, bean, milk and pork
industries. Generally they have proved satisfactory, and
where there has been dissatisfaction by farmers, it has
often been through lack of understanding of behind-the—
scenes work which has gone on to consolidate or expand

markets. Advertising stimulates business. You've got to

76

77

advertise to sell anything these days, or at least to stay

in the race.

--Each type of farming needs its promotion, just as does
each type of manufactured product. The beef industry is
competing with other raw farm products and processed foods
for the consumer's attention. The industry needs to win
over that attention and capture increasing purchasing

power.

--Consumers now eat about 110 pounds of beef per person
each year. However, available money that the consumer

has for Spending on food products is increasing all the
time--at a rate relatively faster than per capita beef
consumption. We need to compete for these extra dollars
the consumer is spending, through promotion of beef and
beef products. Increasing affluence will lead to more
discernment in food selection, and hence to more beef
purchases, but we still need to work on the consumer's
awareness of the advantages of beef and keep reminding him

of them.

--People buy too much artificial and processed foods these
days--there is a real need for the public to be educated

to choose pure foods, nutritious ones, and especially beef.

78

--Beef prices are as high now as they were 20 years ago,
but the cyclical nature of farm prices suggests that they
may not retain their relative level, unless we act now!
The market could and should be expanded with a lot more
advertising. Sure, consumer demand is high now, but in

looking ahead, who will help us but ourselves?

--Processors and middlemen have their restricted advertis-
ing, but who is going to take the broader message of the
industry to the public to increase awareness unless we do?
The checkoff is a ghggp way to advertise beef. Successful
promotion programs have to be carried to the public, to be
planned for attack from different angles, and for these to
complement each other. This isn't an easy task--we need
the best talent and resources available, and the beef com-
mission will embody these--all for a few cents cost to

the farmer.

—-The long-run View needs to be taken by farmers. By out-
laying a few cents in checkoff we will get higher returns
later, through promotion and education, and the higher

returns will more than offset the checkoff costs involved.

--In the next 10 to 20 years, farmers are likely to be
asked to make payments for all sorts of things, e.g. pol-
lution control, etc. Rates are likely to be high on such

future payments, and in comparison, the marketing tax

79

payments seem almost negligible, but yet something we need
to keep up if we want the beef industry to remain viable.
And with the squeeze on the smaller farmer, we should do
what we can to help his returns and help him stay in

business.

--We seem assured of reasonable farmer representation on
the commission, rather than control of the funds by people
remote from our interests and lacking knowledge of the
industry. (Of 11 members, five will be breeders or feeders,
and one a dairyman. Of the other five, these will be

entitled to vote.)

--We need a better mouthpiece. Farmers in general, but
the beef industry in particular, need an effective lobby

to look after their interests.

--The present 6¢ checkoff for promotion has been voluntary
(refundable on request) and goes to the National Livestock
and Meat Board. The new scheme will, and should, be
mandatory, since where promotion benefits the whole

industry, either everybody, 9; nobody, should pay for it.

--The farm-to-table marketing system has its inefficiencies,
and the checkoff money should go toward ironing out some

of these. Through marketing research and other methods,

the commission should play an important role in developing

a more orderly system of marketing.

80

——We need to be informed as to changes in marketing situa—
tions. Information from research, in the same sort of way
that Michigan Milk Producer's Association works for pro-

ducers, will help in decision making, scheduling and farm

management.

--Michigan farmers can produce more cattle. There is plenty
of capacity, and yet we don't want to depress cattle

prices. But, through promotion programs and consumer edu-
cation, the commission can create or expand markets to
encourage greater numbers of cattle. With growing urbani-
zation and potential markets, we need to be active in doing

this.

--There is a need to educate the public to understand the
costs involved in raising and feeding cattle, i.e. land,
feed, etc. The city man has little awareness of farm
problems and of what the farmer receives for his pro-
ducts. There is a long list of middlemen, from shippers to
processors, to meat cutters, to merchandisers. Their
increased wages and Operating costs are reflected in the
price the consumer pays. There is a communication gap at
present between the city man and the farmer, which well-

directed information programs can help to overcome.

 

*Beef marketing tax, checkoff and assessment are here
used as interchangeable terms, although some respondents
stated a preference for one term over others.

This mimeograph was compiled from a recent survey by
MSU graduate assistants, but the views do not necessarily
reflect those of any staff member at Michigan State
University.

81

PROPOSED BEEF INDUSTRY COMMISSION--

BACKGROUND NOTES

Purposes: A bill which proposes the establishment of a
beef industry commission is at present being considered by
the Michigan State Legislature. If passed, the bill would
create an 11 member state beef industry commission within
the Department of Agriculture. The commission's primary
function would be to promote the sale of beef and beef
products. To do this it would engage in promotion pro-
grams, research, consumer education and marketing programs
in conjunction with the national livestock and meat board.
It would also function in a liason capacity among the beef

industry, food industry and consumers of the state.

Finances; The commission would have the power under the
bill to impose an assessment (checkoff, or marketing tax)
upon all sellers of cattle in the state. This amount would
be up to one-tenth of one percent of their gross receipts.
Sellers would be required to keep records of all purchases
and sales, and pay the assessment monthly. (The existing
scheme of 6¢ per head payment to the National Livestock

and Meat Board, through the Michigan Livestock Exchange
(refundable upon request) is not referred to in the bill;

the question of what would happen to it is apparently

82

left open at this stage.) It is proposed under the bill
that up to 50% of funds collected would be appropriated

to the National Livestock and Meat Board.

Avenues for changg: The bill allows for a referendum of

 

sellers, after three years of commission operation, to
determine if the commission should continue. Also, upon
written petition from sellers, a referendum could be held

to determine if the commission should continue.

Legislative moves: The bill passed the House of Represen—

 

tatives; it has not yet been decided upon in the Senate.

APPENDIX C

PAMPHLET: ARGUMENTS AGAINST

A BEEF MARKETING TAX

83

ARGUMENTS AGAINST
A

BEEF MARKETING TAX

84

ARGUMENTS AGAINST A BEEF MARKETING

TAX (CHECKOFF OR ASSESSMENT)*

Preface: In a recent telephone survey over 300 farmers in
southern Michigan were asked their views on the porposed
marketing tax and beef industry commission. Most of the
farmers contacted owned cattle. Approximately half were
in favor of the issue, and approximately half were against
it. From respondents' answers, two separate mimeographs
have been prepared—-this one, and one entitled Arguments
in Favor of a Beef Marketing Tax (Checkoff or Assessment).
In both, the words and phrases that reSpondents actually
used have been retained as much as possible, consistent
with avoiding redundancy among the main themes that
respondents put forward-—i.e., the following arguments

are direct, or almost direct, quotations chosen from the

interview responses.

--Beef prices are the highest they have been for 20 years--
the market is satisfactory-—there is no need to advertise
beef. The situation for lamb and pork is different, in
that to the consumer they are regarded as more or less
alternative meats to beef. Beef is constantly in front

of the public, and the consumer eats as much of it as he

can afford. The few extra pounds that would be sold

85

86

through promotion and consumer education wouldn't be worth

the added cost.

--At present, supplies of beef are short. What goes on to
the market sells pretty well, and with the increasing
spending-power of consumers, it is likely to stay that way.
Over time this might change, but the need isn't there now

and it doesn't warrant us paying a checkoff to promote beef.

--We don't need to create more white collar jobs. There
are already too many middlemen supported by the farmer,

in many cases receiving a higher net income for less work
than the farmer. Whose interests would be looked after--
those in administration, or ours? What would they know
about problems in farming, and how responsive would they
be to them? As well, there are a lot of questions left
open about what the commission would do, and how the money

would be spent.

—-It sounds as though net returns to farmers would be
increased, but instead they would be less. Either they
will be absorbed by costs of administration, or what's
left over will be so small as to prevent the commission
from being effective. Some limited good could come from
it, but farmers' experiences with bureaucracies are often

that the benefits to farmers seem to evaporate.

87

--Supply and demand operate best with little interference
from governments--although intervention of this type may
be a function for meat packers or producer organizations.
We are already paying a checkoff through the Michigan
Livestock Exchange, the funds going to the National Live-
stock and Meat Board. This payment is optional, and it

works well. We shouldn't have to pay any checkoff.

 

-—There are too many taxes now. The small farmer is being
squeezed, and can't afford to make more payments than those
that are necessary. If beef promotion is really the best
course to take, and if 51% of the producers agree on setting

it up, that's ok-—but if not, it shouldn't be imposed.

--There is a ceiling, in effect, on meat prices. With
reasonable consumer demand for beef, and some sort of
ceiling on prices present, there is no need to bother with

advertising. It won't do much good.

--Simi1ar schemes with other farm products don't appear to
have beenfited farmers much. The cherry tax doesn't seem

to have helped the cherry growers.

--What about others paying for promotion designed to
benefit the whole industry? Packers and stores have large
markups. We're not able to do this, yet we're expected to

provide the funds for the commission's activities.

88

--Prices of cattle are too high now for the feeder, and

this will raise prices even more.

--The commission might not be aggressive enough in its
promotion; also, it could tend to duplicate other adver-
tising of meat, e.g. by stores. Therefore, in a sense,
we would be "stuck" with the commission, and payments

made to it.

—-The diverse nature of the commission's programs suggest
that implementation will be difficult. The ideals for the
industry are fine, but how effective can they be? The

last thing that we need is another tax that won't provide

any benefits to the farmer or help increase his returns.

 

*Beef marketing tax, checkoff and assessment are here
used as interchangeable terms, although some respondents
stated a preference for one term over others.

This mimeograph was compiled from a recent survey by
MSU graduate assistants, but the views do not necessarily
reflect those of any staff member at Michigan State
University. ‘

89

PROPOSED BEEF INDUSTRY COMMISSION--

BACKGROUND NOTES

Purposes: A bill which proposes the establishment of a
beef industry commission is at present being considered by
the Michigan State Legislature. If passed, the bill would
create an 11 member state beef industry commission within
the Department of Agriculture. The commission's primary
function would be to promote the sale of beef and beef
products. To do this it would engage in promotion pro-
grams, research, consumer education and marketing programs
in conjunction with the national livestock and meat board.
It would also function in a liason capacity among the beef

industry, food industry and consumers of the state.

Finances: The commission would have the power under the
bill to impose an assessment (checkoff, or marketing tax)
upon all sellers of cattle in the state. This amount would
be up to one-tenth of one percent of their gross receipts.
Sellers would be required to keep records of all purchases
and sales, and pay the assessment monthly. (The existing
scheme of 6¢ per head payment to the National Livestock

and Meat Board, through the Michigan Livestock Exchange
(refundable upon request) is not referred to in the bill;

the question of what would happen to it is apparently

90

left Open at this stage.) It is proposed under the bill
that up to 50% of funds collected would be appropriated

to the National Livestock and Meat Board.

Avenues for changg: The bill allows for a referendum of

 

sellers, after three years of commission operation, to
determine if the commission should continue. Also, upon
written petition from sellers, a referendum could be held

to determine if the commission should continue.

Legislative moves: The bill passed the House of Represen-

tatives; it has not yet been decided upon in the Senate.

APPENDIX D

LETTER (WITH CARDS): LOW COST CONDITION

91

COOPERATIVE EXTENSION SERVICE

MICHIGAN STATE UNIVERSITY - EAST LANSING . MICHIGAN 48823

College ongricultuIe
andbhnndlhfimuom
Education Insdtutc
Agdaflnmchhfl

 

AND U.S. DEPARTMENT OF AGRICULTURE COOPERATING

August 29, 1972

Dear Farmer:

Currently we are preparing two brief publications, which will be distributed
free of cost if you wish to obtain either of them.

The Michigan State Legislature has before it a bill which preposes a tax
(checkoff or assessment) of approximately 20¢ to 40¢ per head on all cattle

sold in Michigan. This would establish a commission to administer a program of
beef promotion, consumer marketing, industry information, research and education.

There are varying opinions of\this proposal, and in the interests of providing
information on it we are preparing two mimeographs:

-ARGUMENTS IN FAVOR OF A BEEF MARKETING (CHECKOFF OR ASSESSMENT)
and
-ARCUMENTS AGAINST A BEEF MARKETING TAX (CHECKOFF OR ASSESSMENT).

If you would like either or both, complete the enclosed business-reply card/s
(postage paid) and your request will be forwarded shortly (free of cost).

Sincerely, /) ///
/E'rroll H. Wamhoff
. cting Director

College of Agriculture and Natural
Resources Education Institute

CHW:k1w

enclosure

92

93

Sample of Enclosure Cards:

 

 

 

 

 

 

 

 

 

Please send me
A
3 " ARGUMENTS AGAINST A
: BEEF MARKETING TAx
: ,5 (CHECKOFF. 0R ASSESSMENT)v
E .3; E
3 5 2
z 4:

Please send me
3 N ARGUMENTS IN FAVOR oF A
: BEEF MARKETING TAX ,
g (CHECKOFF. 0R ASSESSMENT)
'2. S E
V a 5:

 

 

APPENDIX E

LETTER (WITH CARDS): MEDIUM

COST CONDITION

94

COOPERATIVE EXTENSION SE RVICE

MICHIGAN STATE UNIVERSITY ° EAST LANSING ° -MICHIGAN 48823

College of Agriculture
andbhnndfumxuas
Education Institute
Agriculture Hall

 

AND U.S. DEPARTMENT OF AGRICULTURE COOPERATING

August 29, 1972

Dear Farmer:

Currently we are preparing two brief publications, which will be distributed
free of cost if you wish to obtain either of them.

The Michigan State Legislature has before it a bill which proposes a tax
(checkoff or assessment) of approximately 20¢ to 40¢ per head on all cattle

that are sold in Michigan. This would establish a commission to administer

a program of beef promotion, consumer marketing, industry information, research,
and education.

There are varying Opinions on this proposal, and in the interests of providing
information on it we are preparing two mimeographs:

-ARGUMENTS IN FAVOR OF A BEEF MARKETING TAX (CHECKOFF 0R ASSESSMENT)
and
-ARGUMENTS AGAINST A BEEF MARKETING TAX (CHECKOFF 0R ASSESSMENT)

If you would like to receive either or both, complete and stamp the enclosed
card/s, and your request will be forwarded shortly (free of cost).

 

Sincerely, .

’_ goal/fl’

arroll H. Wamhoff
Acting Director
College of Agriculture and Natural
Resources Education Institute

CHW:k1w

enclosure

95

96

Sample of Enclosure Cards:

 

 

 

 

 

Please send me
3 °" ARGUMENTS AGAINST A
: BEEF MARKETING TM 7
a (3 (CHECKOFF. 0R ASSESSMENT)
E. B ‘53
V a 2

Please send me
3 " ARGUMENTS IN FAVOR OF A ._
g BEEF MARKETING TAx
3,5 (CHECKOFF. 0R ASSESSMENT)
3- i=2 “2‘

Z <

 

 

 

 

 

APPENDIX F

LETTER: HIGH COST CONDITION

97

COOPERATIVE EXTENSION SERVICE

MICHIGAN STATE UNIVERSITY ° I'IAST LANSING ° MICHIGAN 48823

College of Agriculture
and Natural Resources

_.. .W—.l- - --.._——~...— .— . - _.. --M- . w---~-._ __.. -. _...-. .- . —- _. . -..

AND U.S. DEPARTMENT OF AGRICULTURE COOPERATING

Education Insritute
Agriculture Hall

August 29, 1972

Dear Farmer:

Currently we are preparing two brief publications, which will be distributed
free of cost if you wish to obtain either of them.

The Michigan State Legislature has before it a bill which proposes a tax
(checkoff or assessment) of approximately 20¢ to 40¢ per head on all cattle
that are sold in Michigan. This would establish a commission to administer

a program of beef promotion, consumer marketing, industry information, research
and education.

There are varying opinions on this proposal, and in the interests of providing
information on it we are preparing two mimeographs:

-ARGUMENTS IN FAVOR OF A BEEF MARKETING TAX (CHECKOFF OR ASSESSMENT)
and
-ARGUMENTS AGAINST A BEEF MARKETING TAX (CHECKOFF 0R ASSESSMENT)

If you would like to receive either or both, your request will be forwarded
shortly (free of cost)-dwrite to:

-(for Arguments in Favor of a Beef Marketing Tax)
M. Peterson
527 South Kedzie Hall
Michigan State University
East Lansing, Michigan 48823

—(for Arguments Against a Beef Marketing Tax)
L. Nash
A202 Fee Hall East
Michigan State University
East Lansing, Michigan 48823

     

roll H. Wamhoff

cting Director

College of Agriculture and Natural
Resources Education Institute

CHW:klw
98

"ll lll’llllll'llllllI