A Study of Public Policy Impact on Consumer Decision-Making

ABSTRACT - This report describes how a consumer decision-making model was created and tested in research sponsored by the Federal Communications Commission. The FCC was considering a series of policy changes with regard to equipment and rules of operation. Discriminant and regression statistical results were combined into a simulation model which projected the impact of the proposed policy changes.

Citation:

Philip C. Burger and Alladi Venkatesh (1979) ,"A Study of Public Policy Impact on Consumer Decision-Making", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 508-513.

Advances in Consumer Research Volume 6, 1979      Pages 508-513

A STUDY OF PUBLIC POLICY IMPACT ON CONSUMER DECISION-MAKING

Philip C. Burger, State University of New York at Binghamton

Alladi Venkatesh, State University of New York at Binghamton

ABSTRACT -

This report describes how a consumer decision-making model was created and tested in research sponsored by the Federal Communications Commission. The FCC was considering a series of policy changes with regard to equipment and rules of operation. Discriminant and regression statistical results were combined into a simulation model which projected the impact of the proposed policy changes.

INTRODUCTION

Traditionally, marketers have focused their attention on the impact of corporate policies upon consumer decision-making and choice behavior. The policies of the government were a given condition external to the consumer decision-making framework. Political scientists and economists, however, have always been interested in the politics and economics of regulation at the aggregate rather than the individual level.

This study was commissioned by the Federal Communications Commission to explore the impact of federal policy making on consumer decisions. For our purpose, 'policy' can be defined as a series of rules affecting the availability of products and services with the express intention of achieving certain objectives. In the case of the government, such objectives have a broad social content and in the case of a corporation they are usually profit oriented. Both government and corporate policies can result in the creation and elimination of products which in turn have a direct impact upon consumers.

The Federal Communications Commission originally commissioned this particular study to determine which of several policies might be implemented to improve citizen's band radio service. Citizen's band radio had grown from under 40,000 users in 1959 to 20 million users in 1977. The frequencies available to CB users created certain limitations for personal radio communications. Being aware of this problem, the FCC was developing a number of alternative proposals. At the same time, it was getting concerned with possible consumer reactions to these policy proposals. There was, obviously, a growing desire within the FCC that any new policy should maximize the satisfaction of as many CB users as possible.

BACKGROUND

Typically, studies on public policy issues relating to marketing have been in the areas of consumerism, deceptive advertising, legal restraint, products labeling, and such other issues involving consumer protection (Aaker and Day, 1978). Commenting on the role of consumer research, Sheth (1974) noted "the bulk of research in consumer behavior for public policy is likely to be problem oriented, and most of the problems are likely to come from the negative side effects of mass marketing and mass consumption - for example, mass media effect on citizen's value." Recently, however, some authors have noted that substantive research is needed on consumer decision-making and choice behavior as inputs to public policy making or to evaluate its impact (Wilkie, 1976, Jacoby, 1977). There are very few systematic empirical or theoretical studies at the micro consumer behavior level in the marketing literature. What little work there is, has been done evaluating FTC policies on deceptive advertising, children's commercials, etc. [For a very interesting theoretical model designed to evaluate welfare effects of regulation at micro consumer level, using Lancaster's approach, see Colantoni, et al. (1976).]

Our study marks a serious departure from the existing public policy-marketing interface issues and examines consumer attitudes and usage patterns of a highly regulated product, CB radios. In evaluating the impact of policy making upon consumer attitudes and decisions, a segmentation design was developed based on CB user/nonuser attitudes, psychographics and demographics. In addition, the research analyzed the recent adoption behavior of CB radio users and extrapolated future adoption possibilities resulting from key alternative policy proposals. Thus the study, while indicating that some generalizations could be made about consumer behavior in response to policy making, also pointed the way toward picking a more optimal set of policy alternatives.

The Policy-Consumer Impact Model

Figure 1 shows a simplified model on which most marketing studies are based. The process underlying the model suggests that a given federal agency creates a set of rules concerning a relevant marketing issue such as advertising, product safety, etc., which are transmitted to industry (arrow 3 in Figure 1). Industry then uses such rules as inputs into its product and promotion policies which, in turn, impact customers (arrow 4). There is a feedback loop from the customer both back to the industry and the agency. An example of this is the use of seat belts. In spite of the auto safety implications of the usage of seat belts, car drivers have not been using them. Consequently, the government is trying to enforce this, not directly, but by insisting that manufacturers build an automatic seat belt mechanism in the cars.

Amore complicated set of relationships is suggested by the arrows 1, 2 and 3 in Figure 1. This system of effects may be described as the general paradigm for the present study. The sequence of steps are simply that the regulatory agency defines a policy which is transmitted to consumers through mass media or any other means (arrow 1). The consumer reaction in terms of letters, complaints and other types of communication is relayed back to the regulatory agency which then modifies its position, if necessary, and suggests a set of technical rules and regulations to industry (arrow 2). Based on such directives from the government, the industry makes its own product policies and the cycle continues (arrow 3 and 4). This paradigm clearly illustrates what happened in the case of the CB radio beginning in 1959. The early ruling concerning CB radios set aside the 27 MHz band for use of private citizens using 4 watt (low power) radios for their own purposes. These short range radios achieved very limited adoption. The regulation generated very little reaction among consumers initially. However, in 1973 when the "truckers strike" occurred as a result of the Arab Oil Embargo, the FCC was bombarded by license applications for the new radios. A great demand for the radios existed and there was a small supply. The industry responded by building many radios and importing many from Japan. Large numbers of applications were received by the FCC and a large number of complaints about channel crowding were also received (Figure 1, arrow 2). The FCC responded (arrow 3) by adding 17 additional channels to the existing radio configuration.

FIGURE 1

INITIAL POLICY - CONSUMER PARADIGM

Consumers responded again to the FCC with complaints about the telephonic quality of the radio and the FCC then began considering new alternatives which would change the configuration and the price of the radio service. In addition, it became clear that a certain culture was developing around the citizen's band radio. This culture manifested itself in usage behavior, special language, and other unforeseen uses for the radio itself which included locating police radars, providing and enhancing security of individuals, boating uses, and other legal and semi-legal uses.

It became clear in discussions with the FCC that a more refined approach to the policy impact on consumer decision-making was required. Therefore, the relationships shown in Figure 2 were introduced as the basis for the research. This new paradigm can be viewed as a circular flow beginning with the regulatory agency. The regulatory agency promulgates rules and technical specifications for products. These specifications impact on the cognitive structure of the consumers, the logic of which is created by the interfacing of the technical aspects of the product with the perceptual aspects of the products and service. In terms of CB radios, consumers perceive CB radios as a certain bundle of characteristics a subset of which is influenced by the policies of the FCC whose impact is also clearly perceived. One resulting hypothesis is that the policy impact alters consumer's beliefs-attitudes and which in turn affects purchase and usage behaviors.

FIGURE 2

EMERGENT POLICY CONSUMER PARADIGM

The segmentation aspect of the model (represented by arrow 4) posits that different consumer groups react in different ways which are unique to the segment. Arrow 5 in Figure 2 indicates that consumer behavior responses can be integrated into an indicator system which can be monitored by regulatory agencies. Thus, the model can be viewed as a stimulus-response model where, the regulatory agency creates a stimulus and the consumers respond over a period of time. The regulatory agency must then interpret the demand indicators to ascertain whether the policy contributes to consumer satisfaction or not.

Sampling Design

The sampling was done on a national basis. Two types of primary sampling units (psu's) were identified, metropolitan areas and the others. All the psu's were grouped into 65 strata. The secondary sampling units (ssu's) were identified as all areas with a population of approximately 2500 (e.g., a block or two in a major city; a census tract in some rural areas). Systematic random procedures were used to select 100 ssu's. Finally, households were selected from the ssu's. In each household, the interview was conducted with the CB radio user, and for nonuser households, the key male member of the household was questioned.

In the first stage, about 863 users and nonusers were personally interviewed. This was followed by 441 interviews with CB user households which were identified through telephone screening. The final sample included 1304 respondents of which 674 were nonusers and the remainder were users. Because of the disproportionate representation of the users in the sample, a weighting scheme was devised to achieve representative estimations in the analysis of sample data.

The data was collected as shown in Table 1. The questionnaire consisted of five major sections relating to CB ownership, usage, and attitudes. The first section included detailed information on the ownership of CB radios, as to when they were purchased, what price was paid, and the number of channels the radio had. The second major section of the data was the usage patterns of the CB radios. This data included time of day, number of hours per week, terrain of use, and purpose of use. A third major section of the questionnaire involved a 19 item semantic-differential scale measuring beliefs and attitudes toward various aspects of CB radios. The list of attributes was derived from detailed focused group interviews with both users and nonusers of CB radios.

TABLE 1

DATA COLLECTED

The fourth major section of the questionnaire involved 35 questions which were derived from the psychographic literature (Wells and Tigert, 1971). The questions devised by previous researchers were used with almost no modification. Of the fourteen life style factors proposed by Wells and Tigert, 12 were hypothesized to be related to CB radio usage.

The last major section included demographic information including age, income, occupation, residence, number of cars owned, education, and others. The questionnaire took about an hour to complete.

THE RESULTS

The data analysis used different techniques: (1) basic descriptive statistics, (2) cross-tabulation, (3) segmentation analysis, (4) discriminant analysis, and (5) regression. At the end, the statistical information was cast into a computer simulation program to test effects of various policy impacts on the dependent variables.

The dependent variables in the study, which were the basis for the demand indicators are: (1) projected numbers of users and nonusers by segment, (2) the pattern of usage, (3) the number of hours per year of usage by segment, and (4) attitudes in terms of satisfaction/ dissatisfaction with elements of the radio service.

TABLE 2

FACTOR ANALYSIS OF PSYCHOGRAPHICS

The segmentation analysis was undertaken using factor analysis as a prime technique. First, the psychographic variables were factored to see if the loadings and factor patterns were consistent with those found by Wells and Tigert. The results appeared to confirm their findings. The emergent factor patterns are shown in Table 2 with relevant segments identified by asterisks. Only those psychographic variables which discriminated best the users from nonusers in cross tabulation analysis and which also loaded highly on the factors were considered important in the segmentation analysis. Six meaningful segments emerged. The first one was the indoor sports fan. The indoor sports fan watches television, reads sports magazines, and values entertainment. This was a prime segment of the CB market. This segment accounted for 26.6% of the U.S. population and over 20% of this particular group owns CB radios. The second major segment was the car repair buff. These people constituted approximately 18.8% of the U.S. population and almost 31% of them owned CB radios.

The third segment was the person who belongs to the service organization. These people were the members of the CB radio clubs, ambulance services, volunteer fireman, and such. These people comprised approximately 6% of the U.S. population but over 31% of them own CB radios. A fourth major segment was the fashion conscious businessman. This person values his or her appearance and spends a great deal of time traveling. This segment accounted for 6.5% of the population and almost 18% of this group owned CB radios.

Segment 5 is called the residual under-30 group consisting of those individuals who were in the age range from 16 years to 30 years, but did not fall in the previous four segment categories. They accounted for approximately 11% of the population and 16% of them owned CB radios. The sixth and last segment includes the residual middle income person. These people are all those who did not fall in the previous five segments but who were classified as middle income ($]6,000-$25,000). They comprised about 11% of the U.S. population and almost 12% of them have CB radios. The null segment or the members of no segment accounted for approximately 18% of the U.S. population and 6% of them had CB radios. Additional segments were not considered important because of the low marginal information obtained by adding more segments.

After the segments were identified, all of the data analysis was performed on a within segment basis. The next major step was to see if attitudes and perceptions of radio service discriminated between users and nonusers in each segment. It was hypothesized, based on the results of the focus group interviews, that belief and attitude variables would tend to discriminate between users and nonusers.

A discriminant analysis was performed and the results are shown in Table 3. In the segment containing service organization members, a nonsignificant user/nonuser discrimination was discovered. Indoors sports fans valued being entertained. The car repair buffs valued entertainment, helping others and getting help. Also, the aggregate preference score which was a simple summation of all the attitudes values appeared is a discriminant for the car buffs.

The fashion conscious business segment was male, and valued making friends and being entertained. The residual middle income group valued entertainment and helping others. The necessity for boating communications was evaluated negatively. Thus, it was clear from the data analysis that attitudes were very powerful discriminators between users and nonusers. It was also clear from the data analysis that other variables such as demographics and car ownership, etc., were not as powerful discriminators.

TABLE 3

VARIABLE DISCRIMINATING USERS FROM NONUSERS

Regression analysis was also performed to predict the various kinds of CB usage behavior. As a first step, all of the behaviors were factor analyzed and found to group into three basic factors. One of these was, time of day, namely 2 P.M. to 10 P.M. daily. Another was use in the suburbs and small cities while the last was use in flatlands and low hills. These three factors became dependent variables of usage. For each segment, a series of regressions using attitudes as the independent variable were performed to see if attitude structure predicted usage pattern. The results are shown in Table 4. The R² range from 3% to 34% indicating that in some types of usage a very low predictive power but others have quite high predictive power. Overall, the prediction of usage pattern was much weaker than discriminating users/nonusers. It is important to note that, in general, the results show a tradeoff between telephonic needs and entertainment. The indoor sports fans wished to be entertained while other groups tended to value telephonic communication.

Using the Results to Test Policy Alternatives

A computer simulation program was created using the results from the regression and discriminant analyses. The program which was written in the FORTRAN language, contained all of the regression/discriminant b coefficients for each segment. The program used the raw data from each respondent as input, classified each respondent as a user or nonuser of the proposed CB service, determined the rate of usage by category and computed some aggregate sample statistics. The input data was weighted to reflect a sample representative of the U.S.

The key logic of the simulation required that changes in perception of the proposed radio service be related to the engineering aspects of the proposal. The program simulated changes in beliefs and attitudes of each individual respondent and computed adoption and usage figures based on the changes. Discussion with FCC personnel revealed each proposal primarily affected two aspects of radio perception: (1) the telephonic quality (the ability to transmit information over substantial distances with little or no interference) and (2) the entertainment value (being able to listen to other persons chat). The technical proposals served to enhance telephonic quality while reducing entertainment value. Price was also included in the simulation. The gross relation between price and adoption rate was investigated and found to be a decreasing function of price with an elasticity of b = -1.4.

TABLE 4

VARIABLES PREDICTING BEHAVIOR

CONCLUSIONS

The FCC was considering two primary proposals. The first of these proposals would be to add a new CB radio service at 220 MHz. This service would allow many more channels than are currently available. The net effect of this would be to decrease the entertainment value of the CB radio by diffusing users over a larger number of channels. However, on the positive side, the service would become much more telephonic in its ability to handle communications. A second major policy proposal was to open up a very high frequency band in the 900 MHz range which would increase the telephonic aspect of the CB radio quite substantially and would further reduce its entertainment value.

Given these logical specification, it can be seen that entertainment and phone communications appear in both the discrimination and the usage pattern predictions. Thus in the simulation, the question becomes whether the increase in telephonic communication would outweigh the decrease in entertainment value. The overall evaluations of the FCC Proposal showed that there was a small yet significant market for the new services.

REFERENCES

Aaker, David A. and George S. Day, "Consumerism: Search for the Consumer Interests" New York: The Free Press, 1978 (Third Edition).

Bloom, Paul N. and Nikhitesh Dholakia. "Marketing Behavior and Public Policy: Some Unexplored Territory,'' Journal of Marketing, 37, 4 (Oct. 1973), 67.

Colantoni, Claude S., Otto A. Davis and Malati Swaminathan. "Imperfect Consumers and Welfare Comparisons of Policies Concerning Information and Regulation," The Bell Journal of Economics, Autumn, 1976, 7, 2.

Dyer, Robert F. and Terence A. Shimp. "Enhancing the Role of Market Research in Public Decision- Making," Journal of Marketing, 41, 1, January 1977, 63-67.

Jacoby, Jacob. "Information Load and Decision Quality: Some Contested Issues," Journal of Marketing Research, 14, November 1977, 569-573.

Jones, Mary Gardiner. "The FTC's Need for Social Science Research." (Address before the Second Annual Conference, Association for Consumer Research, College Park Maryland, September 1971).

Richardson, Lee. Consumers in the Federal Decision- Making Process, California Management Review, 16, 2, Winter 1973, 79-84.

Ross, Ivan. "Applications of Consumer Information Public Policy Decisions," in Jagdish N. Sheth and Peter L. Wright eds., Marketing Analysis for Socielet Problems, Champaign, Ill, University of Illinois Press, 1974.

Sheth, Jagdish. "The Next Decade of Buyer Behavior Theory and Research" in J. Sheth (ed.) "Models of Buyer Behavior: Conceptual, Quantitative and Empirical,'' New York: Harper and Row, Publishers, 1974 (ch. 21).

Summers, John O. "Less Information is Better?", Journal of Marketing Research, 11, November 1974, 467-468.

Wells, William D. and Douglas J. Tigert. "Activities, Interests and Opinions," Journal of Advertising Research, 1971.

Wilkie, William L. "Consumer Information, Acquisition: Public Policy Perspectives," Association of Consumer Research Proceedings, 3, 1976, 334-340.

----------------------------------------