Consumer Protection and the Value of Information

Citation:

Robert A. Mittelstaedt (1972) ,"Consumer Protection and the Value of Information", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 101-106.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 101-106

CONSUMER PROTECTION AND THE VALUE OF INFORMATION

Robert A. Mittelstaedt, University of Missouri

 

Th' enthusyasm iv this counthry, Hinnissy,

always makes me think iv a bonfire on an icefloe.

It burns bright so long as ye feet it, an' it

looks good, but it don't take hold, somehow, on

th' ice.              ---Mr. Dooley

INTRODUCTION

The central thesis of this paper is a simple one based on several assumptions: (1) If "consumerism" is to continue as a viable force in public affairs, programs to protect the consumer must enjoy a widespread level of public support (2) To insure public support consumer protection programs must provide noticeable and substantial benefits to the consumer. (3) The provision of useful information to the consumer is an important form of consumer Protection.

There may be debate about any of these three points. One could argue that public support is not necessary or that, in any event, it is independent of the results produced by various programs. Further, one could argue that information programs should rank considerably below product safety programs or anti-trust enforcement as vehicles for protecting the interests of consumers. However, this paper will take these three assertions as "given" and attempt to develop the idea that the concept of the "value of information" can be a useful guide to the construction of priorities for both those agencies charged with protecting the consumer and those researchers who wish to make their efforts relevant to consumer problems.

The value of information is a function of the expected opportunity loss (probability and cost) of suboptimal choices. If we accept the idea that "optimality" exists in the eye of the beholder, the ultimate judge of the value of information is the consumer. To measure this value, and thereby set our priorities, we must look at the consumer, his perceptions and his behavior. However, this paper is not simply a call for "more research" or an admonition that regulatory agencies should remain inactive until "more is known." There are pressing problems which demand attention now. Since resources for agencies and researchers are limited, some initial agreement on priorities will have to be based on a priori estimates of costs and probabilities. Because this organization would seem to be an ideal forum for "protectors" and "researchers" to explore common ground for establishing mutual priorities, this paper is an attempt to spark a dialogue. To this end the paper will look at several information programs, attempt to assess their probable value to the consumer and examine the implications of this estimated value for protection agencies and researchers.

BEEF GRADING

First, consider a very old information program--grade labeling of beef. Unlike the mandatory insPection program which concerns itself with the basic wholesomeness of the product, the grade labeling standards provide a means of describing product quality in uniform language--the familiar grades of "U.S. Prime," "U.S. Choice," "U.S. Good," etc. Because the grading program is voluntary not all beef is graded. However, the U.S.D.A. estimates that 80-85 percent of all fresh beef bound for the household market is graded [U.S.D.A. 1969, p. 102; Hutchinson, 1970, p. 1]. Of all beef graded in 1970, 80 percent was graded "U.S. Choice" [U.S.D.A. 1971, p. 147]. This figure, when considered with the observation that many retail outlets adhere to a "one-grade policy," suggests that the U.S. housewife is very likely to select "U S Choice" beef from her retail meat counter because she has no alternatives.

The concept of the value of information tells us, then, that as the probability of the typical housewife choosing "U.S. Choice" approaches 1.0 (because of lack of alternatives), the value of the information represented by the grade label to her approaches 0.0. Viewed in this light, it does not seem surprising that most housewives know very little about the grading system used, as a recent study documented [Hutchinson, 1970]. This does not mean that grading should be abandoned, for it has other useful purposes. However it does suggest that putting resources into consumer education programs about beef grading is not likely to achieve much until the grading scheme is revised to provide grade labels which are more "information-laden." Behavioral research could aid in the construction of a grading system which allowed discrimination on dimensions of significance to the housewife.

UNIT PRICING

Unit pricing is another example of a program designed to provide useful information to consumers. Studies have shown that consumers find it difficult in the absence of unit price information, if not impossible, to find the "least cost alternative," even when instructed to do so [Friedman, 1966; Houston, 1972]. Further, these studies have shown that, compared to the "best buys," the actual choices consumers make "cost" them. It has been concluded that the consumer could save money if unit price information was made available.

The operative word in the previous sentence is "could." By contrast, the value of information concept directs attention to the use people make of unit-price information and would measure the value on the basis of behavioral change. For example, looking at one table from the "Kroger study," there appears to be a shift of higher income consumers toward lower unit priced alternatives in cooking oil [McCullough & Padberg, 1971]. If the goal of unit pricing is to facilitate this shift, the value of the unit price information is measured by the magnitude of the shift in buying patterns. In these data, the average net savings is about .08 cents on every pint sold.

In their excellent review of the Kroger study and others, Monroe and LaPlaca have shown that there are serious methodological pitfalls in studies of this kind, which is to say that they may not all be "fair tests" of the program's value [Monroe & LaPlaca, 1972]. Furthermore, it is clear that consumers may use unit-price information in other ways--to reinforce previous choices or to justify shifting to higher unit priced alternatives. Elsewhere Monroe has suggested a model for dealing with this expanded concept of information [Monroe, 1971].

Initially, research effort should be directed toward the design of methodologically sound experiments, the development of adequate educational programs about unit pricing and the further development of a model to assess the full range of uses of the information. Ultimately, the value of unit price information rests, not on how much consumers could save, but on how, and to what extent, they use unit price information in their buying decisions. When finally computed, the value should be compared to the costs of providing the information. Admittedly, it will not be an easy task but it appears to be, in principle, a solvable problem. Finally, assuming the research is "sound," the advocates of unit pricing should be prepared to accept the results.

TABLE 1

CUMULATIVE DISTRIBUTIONS OF SALES OF COOKING OIL, DISTRIBUTION DEVIATIONS FOR TEST AND CONTROL

DECEPTIVE ADVERTISING

At first glance, the concept of the value of information might not appear to be relevant to the area of deceptive advertising. After all, no one can argue that false and misleading advertising is of value to the consumer. Furthermore, legally removing deceptive advertising claims from the market place does not assure that the remaining (presumably) "true" claims will provide any useful information to the consumer. However, if our initial assumption about continued public support was sound, we should direct our limited regulatory resources toward those deceptive ads which represent the greatest potential disutility to the consumer. The concept of the value of information, then, should turn us away from priorities based on the magnitude of the falsehood of particular claims and toward a set of priorities based on the probability that consumers may be misled and the costs, to them, of the deception.

This is not to say that the successful prosecution of a deceptive advertising case should rest on evidence that people were, in fact, misled, for the courts have already decided otherwise [FTC, 1944]. Nor does it imply that we spend endless hours debating what consumers' optimum and suboptimum choices might be. However, the concept of the value of information implies that the costs of deception are a function of the probability of being misled and the associated opportunity loss.

In constructing a priority scheme based on the likely costs of deception, first priority should go to insuring the truthfulness of claims about products involving physical risk. Even when the probability of being misled is small, the costs of following deceptive claims about the safety or healthfulness of some products is high.

When considering potential economic (as opposed to physical) disutility, the "costs" become less clear. The dollar amount of the purchase may be some guide but, without some probability estimate, it is not a perfect criterion. One might argue that the shoddiest selling practices tend to be associated with those relatively expensive goods and services that are purchased very infrequently--the "once-in-a-lifetime" purchases for most people--and attention might be directed to them first.

However, most advertising involves relatively low-prices, frequently purchased products. Calls for truth-in-advertising usually argue that the consumer is exposed to an overwhelming number of ads for products of this type and is unable to evaluate the claims made for them. This is undoubtedly true, but it does not follow that all of these claims, even if relied upon, represent equal potential losses to the consumer. With repetitively purchased products, the greatest disutility to the consumer comes when the consumer is unable to evaluate the claims made for the product, even after purchase, and continues to buy the product because he cannot refute the deceptive claims based on his own experience.

After all, there are some ad claims, the truth of which the consumer is able to judge for himself after purchase. To take several examples from ad claims recently "under fire," it would appear that the consumer is reasonably well-equipped to judge whether or not: (1) a crunchy candy bar crunches as loudly as it was made to appear to do in an ad, (2) a chain store has stocked the goods it has advertised, or (3) a soup contains as many vegetables as it was pictured to have.

On the other hand, the consumer is very poorly equipped to judge whether or not: (1) a disinfectant significantly reduces the incidence of common diseases, (2) a dentifrice reduces cavities, (3) a fruit drink contains a Particular vitamin, or (4) a gasoline reduces air pollution.

Again, this is not to argue that false and misleading advertising should be allowed to flourish, but only that the "cost of being misled" can be determined, in part, by looking at the question of whether or not the consumer is able to judge ad claims for himself, after the purchase, and alter his behavior accordingly. Researchers interested in making a contribution in the area of deceptive advertising might proceed by discovering the salient dimensions of products, for "false" claims may or may not be "relevant."

THE RELEVANCE OF RESEARCH

As a final point, consideration should be given to the development of a mutual understanding of the research process and what it can, and cannot, contribute. At the risk of muddying the water with a study which is at the core of a controversial issue, consider the "Johns Hopkins' aspirin study." Although a "clinical study," it is not fundamentally different from the research conducted by most of the people in this room.

De Kornfeld, Lasgna and Frazier [1962] administered 5 popular aspirin containing analgesics and a placebo to 198 post-partum patients who had not received a general anaesthetic nor had suffered obstetric complications. Upon patient request, a randomly selected treatment was given each patient and her responses to a question which asked her to categorize her pain as "absent," "slight," "moderate," "severe" or "very severe" before the administration of the dosage and at 15, 30, 45, 60, 120, 180 and 240 minute intervals thereafter were recorded. A decrease in pain from the preceding level to the next level was counted as a score of "+1," a drop of 2 levels was counted as a score of "+2," and so forth with zero's (no change) and negative scores possible. Mean "pain-relief scores" were calculated for each treatment. The major conclusions of the study, which rest on the differences in these scores as tested for significance by Harter's modification of Duncan's multiple range test, were that "all five preparations gave better relief from pain than did the placebo and there were no significant differences among them during the first 3 hours of observation" [De Kornfeld. et. al.. 1962. D . 751.

In assessing the results of this study and its ability to support these conclusions, the researchers among us may wonder if the random assignment of treatments overcomes the problems associated with the sample and the crudeness of the measurement instrument. The statistically-minded may question the conclusion that all 5 compounds are equally effective which is based on the acceptance of a null hypothesis and wonder, therefore, how one might prevent the demonstration of "significant differences in effectiveness" from being something other than a simple function of sample size. The public policy maker might wrestle with the question of whether the inter-individual variability (assuming that it is not just "noise") is sufficient justification for preserving wide consumer choice.

But if we all step back one pace and look at this study and its conclusions from the viewpoint of the consumer, we will probably question the value of information about "comparative effectiveness" which, when finally revealed, consists of the perceptions of effectiveness by a sample of consumers. In short, this "clinical study" provides data no different in kind from that which any consumer could generate for himself at modest cost with little risk. In fact, from the consumer's point of view, given inter-individual variability, the data from this study, based as it is on "averages," will fail to square with many consumers' experiential data. When that happens, it is not difficult to predict which data will be disbelieved.

SUMMARY

Again, it should be stated that it was not the purpose of this paper to make an apology for current marketing practices nor to dismiss consumer information programs as useless. There are urgent problems which demand our attention. What is needed is some shared sense of purpose among the public policy makers and enforces and those of us, who by training and disposition, are equipped to study how consumers behave. Ultimately, if consumer information programs are to avoid the cynical judgment of Mr. Dooley, the value of information to the consumer must be our guide.

REFERENCES

Charles of the Ritz Distributors Corp. v. Federal Trade Commission, 143 F. 2d. 676 (1944).

De Kornfeld, Thomas J., Louis Lasagna & Todd M. Frazier. A Comparative Study of Five Proprietary Analgesic Compounds. Journal of the American Medical Association, 1962, 182, 75-78.

Economic Research Service, Statistical Reporting Service, Consumers and Marketing Service. 1970 Supplement to Statistical Bulletin No. 333, USDA, 1971.

Friedman, Monroe Peter. Consumer Confusion in the Selection of Supermarket Products. Journal of Applied Psychology, 1966, 50; 529-534.

Houston, Michael J. The Effect of Unit-pricing on Choices of Brand and Size in Economic Shopping. Journal of Marketing. 1972, 36, 51-54.

Hutchinson, T. Q. Consumers' Knowledge and Use of Governmental Grades for Selected Food Items. USDA, Economic Research Service, Marketing Research Report No. 876, 1970.

McCullough, T. David & Daniel I. Padberg. Unit Pricing in Supermarkets: Alternatives, Costs and Consumer Reactions. Search, Agriculture, I (January 1971) 1-25.

Monroe, Kent B. The Information Content of Prices: A Preliminary Model for Estimating Buyer Response. Management Science, 1971, 17, B519-B532.

Monroe, Kent B. & Peter J. LaPlaca. What Are The Benefits of Unit Pricing? Journal of Marketing, 1972, 36, 16-22.

U.S. Department of Agriculture. Food For Us All: The Yearbook of Agriculture, 1969. USDA, 1969.

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