Informational Imperfection in Local Consumer Markets: Assessment and Implications

Assumptions for this Analysis

For the perfect information frontier framework to serve as a valid standard of assessment, several assumptions must be made. The first is that the identification of varieties of products and retail outlets is complete and accurate for some representative consumer in a particular market. A second is that the prices quoted are accurate. In this investigation the actual price represents the lowest price a seller was willing to quote when told that the information he provided would be widely distributed. For some products competent bargaining could secure a lower price. Third, fully informed consumers would accept Consumers Union's assessments of quality for different varieties of products.

Interpretation of Price Quality Relationships

The discovery of extensive price dispersion above the perfect information frontier need not signal an informationally imperfect market. A number of factors other than consumer ignorance might account for such dispersion. From an interpretive viewpoint the basic question is this: to what extent could the observed dispersion of prices above the frontier be reasonably attributed to such "other factors"? To the extent that the dispersion is not reasonably attributable to other factors, we will conclude that the market in question is informationally imperfect. We turn now to a discussion of the other factors that might produce such dispersion.

Intrinsically Subjective Characteristics

The quality data on which we will rely, based as it is on the work of Consumers Union, takes no account of the intrinsically subjective characteristics of products. An example might be the "style of an automobile." A quality estimate that took no account of style would tend to underestimate the characteristics of a variety of automobiles that rated high in style; the price, on the other hand, should reflect what the average consumer is willing to pay for style (as combined with other characteristics in a particular variety of car).

Non-Uniform "Product" and Market Sets

Consider cameras as an example. Both the delineation of the product set and the assessment of quality may depend on the individual's expected use of the product. The professional photographer, feeling the need for a full array of camera accessories, may include in his product set only cameras with a broad family of accessories. Not so, the enthusiastic amateur who may be content to get along with far fewer accessories. The same consideration affects the assessment of quality. Presumably a professional photographer would give heavy weight and high characteristic scores to a brand of camera providing a numerous and varied array of high quality of accessories. Again, the enthusiastic amateur may be expected to give lesser weight and lower ratings to such a consideration.

Analogous considerations pertain to markets that professional and amateur photographers would delineate. They might be expected to differ regarding appropriate search costs and hence might select different sellers for inclusion in "their" markets.

Together, non-uniform product and market sets might produce different frontiers or differences in the price-to-frontier ratios that are not attributable to informational imperfections.

Non-Uniform Quality Assessments

For products of variable quality there is no gainsaying the possibility that different quality assessments may be made by different consumers even for products whose characteristics are predominantly "objective." It is reassuring to learn that in 1970, among Consumer Reports purchasers-subscribers, 48 percent to 81 percent reported the purchase of models that were "top-rated" in Consumer Reports. [The product set included hair shampoo, color television, coffee makers, sewing machines, record changers, TV antennas, and AM-FM radios. Source: (Benson and Benson, 1970, Table T 21).] But it is not conclusive. Until convincing research has been undertaken on the matter of uniformity of quality assessments by fully informed consumers, each of us will have to estimate in each case how uniform or variable such quality assessments might be.

Characteristics of Retailers

In the charts, prices but not quality assessments, take account of such characteristics of retailers as their friendliness, convenience, etc. To the extent that consumers value such services, a finding of informational imperfection may be less justifiable.

A third type of information needed by consumers is general product information, i.e., knowledge of what characteristics of products are desirable. Salesmen/advertisers tend to provide such information only when it facilitates the selling of "my" brand/model. The result is that consumers probably receive relatively little general product information from salesmen/advertisers.

Obviously, consumers need price information and, obviously, salesmen/advertisers provide immense amounts of price information, especially in local and mail-order advertising.

Finally, we come to the category of information on which consumerists lay most emphasis--specific product information on the quality of particular brand/models. Almost invariably the quality information provided by salesmen/advertisers is partial, exaggerated, and presented in a way to forestall valid comparisons among varieties. We should pause to ask why.

The answer rests in the motivation and the competitive position of most sellers. The overriding objective of sellers/advertisers is to sell the set of goods produced or purchased by their organization. This motivation has several consequences. First, it leads to an emphasis on persuasion rather than information since experience and market research has shown that persuasion works better for them. Second, it leads to the employment of whatever devices work: exaggeration, omission, the use of psychological levers unrelated to information. For an apt discussion of such levers, see (Gardner, 1975). The last thing salesmen/advertisers want is a valid comparison of the quality of the brands/models. The reason lies in the competitive position in which most companies and hence most salesmen/advertisers find themselves. First, "winners" in a competition among brand/models are ordinarily few and seldom known in advance. Galbraith (1969, p. 30-31) cogently argues this point. Second, if your company's product is a non-winner in a fair comparison, you would rather thwart than facilitate such comparisons through your sales or advertising messages.

Almost viscerally, consumer reformers turn first to the regulation of advertising as a corrective policy. So shall we.

Regulation of Advertising as a Solution?

It is my thesis that the regulation of advertising will not work. There are several reason.

In the first place, Madison Avenue is too clever. Consider the following example of "corrective advertising," known to many of you. In 1971 the Federal Trade Commission had found the advertising for "Profile Bread" to be misleading. In the Commission's view the advertising for Profile Bread had suggested, misleadingly, that the consumption of Profile Bread would help people lose weight. That same year the makers of Profile Bread agreed to an FTC consent decree by which the company was not to advertise "Profile Bread" for one year unless 25% or more of its advertising budget was devoted to advertisements that would correct previously misleading copy. These ads had to be approved by the Federal Trade Commission.

As reported in the February, 1972 Consumer Reports (p. 64-65), the following advertisement, prepared by Madison Avenue and approved by the Commission, appeared on television:

I'm Julia Mead for Profile Bread. And like all mothers, I'm concerned about nutrition and balanced meals. So I'd like to clear up any misunderstandings you may have about Profile Bread from its advertising or even its name. Does Profile have fewer calories than other breads? No, Profile has about the same per ounce as other breads. To be exact Profile has seven fewer calories per slice. That's because it's sliced thinner. But eating Profile will not cause you to lose weight. A reduction of several calories is insignificant. It's total calories and balanced nutrition that counts. And Profile can help you achieve a balanced meal, because it provides protein and B vitamins as well as other nutrients.

How does my family feel about Profile? My children love profile sandwiches. My husband likes Profile toast. And I prefer Profile to any other bread. At our house delicious taste makes Profile a family affair.

Ask yourself which message is predominate--the retraction of a previously misleading advertisement, or the notion that Profile Bread is helpful to a person seeking to reduce his/her weight.

There are other problems with regulation as a corrective policy. The sheer volume of advertising messages is too great, posing an impossible monitoring problem. Further, it can be argued that the focus of regulation has been wrong (bur difficult to change). Since its beginning the objective of regulation has been on the elimination of deception. The trust of this paper is that the correct focus should be on the encouragement of advertising that informs over all informational categories, and not simply with respect to existence of products, brands and sellers.

A further obstacle to regulation is that the borderline between deception and non-deception is difficult to define. Consider the following example, culled from the advertisement of a savings and loan association:

ILLUSTRATION

For those with sufficient visual acuity, attention, and background knowledge this ad contains complete and accurate information. But for others? Now, in evaluating this ad from a regulatory viewpoint, consider the time and effort that would be required to apply David M. Gardner's proposal to ascertain whether an advertisement is behaviorally deceptive (Gardner, 1975).

Finally--and very important- the regulation of advertising does nothing to correct the omission-commission defects of information provided by salesmen.

Other corrective policies will be discussed toward the end of this article.

Price Discrimination Based on Objective Factors

Often sellers practice price discrimination by charging different prices to consumers of different objective characteristics, e.g., adults vs. children. Obviously, differences in prices arising from this source are not chargeable to informational imperfections.

3y contrast, some price discrimination is made possible by consumer ignorance. This would occur when a knowledgeable consumer successfully bargains for a lower price that is not generalized to other, less knowing consumers. Other examples abound. Since the data presented here assign but a single price to a particular retailer, neither kind of price discrimination affects these results. The net effect of this omission is to underestimate the extent of informational imperfection.

Comparing Markets: Other Questions That Might Be Asked

This paper confines itself to the problem of informational imperfections. Here the central question is: To what extent does there exist price dispersion above the perfect information frontier that is properly chargeable to informational imperfections in that market?

But there are two other questions that might be appropriately asked and answered by our price-quality diagrams. The first question is: what are the best achievable terms in Market I vs. Market II? The answer to this question calls for a comparison of the perfect information frontier in alternative markets--its height and its slope.

A second important question pertains to the richness of choice offered by particular markets. The answer to this question calls for a comparison of: (1) the number of varieties that are purchasable in Market I vs. Market IS; (2) the number of retail outlets from which a particular variety may be purchased in each market; and (3) the number of products that may be purchased "locally.'' [Depending upon "my" search costs, "my" local market may or may not include mail-order or telephone-order retailers.]

SAMPLE RESULTS AND THEIR INTERPRETATION

The author, in collaboration with Greg J. Duncan of the University of Michigan and Loren V. Geistfeld of Purdue University, is engaged in a pilot study designed to document the informational imperfections of local consumer markets over several dimensions: for different types of products, for markets of different sizes (Minneapolis vs. Ann Arbor vs. Ithaca vs. Corning), and over time. Unfortunately results from this study are not yet available. However, sample results from an earlier investigation are available and are presented here.

These sample results pertain to two products of approximately uniform quality--life insurance and Kodachrome 126 color film--and two products of variable quality--single-lens reflex cameras and men's ten-speed racing bicycles, for all four products the market is Ann Arbor, Michigan which in 1974 had a population of approximately 130,000.

Life insurance, our first sample product, should be of great interest to both consumers and students of markets: for those who purchase it, it takes three and one-half percent of their disposable income; what is more, consumers purchase it in large amounts and tend to stay with the same company for a long time. Term life insurance--the type .with which our example deals--is a product of uniform quality because it embodies a single characteristic: protection. Die, and the payment of policy's face value provides income replacement, or "protection," for your beneficiary.

Figure 1 tells the story for two variants of term life insurance, non-participating and participating. The results in both cases are emphatic. There is a high degree of price dispersion. And, what is equally undesirable from the viewpoint of a consumer in this market, only a minority of the low priced policies are "readily accessible" in Ann Arbor. The judgment that this is an informationally imperfect market cannot be avoided.

We turn now to our first product of variable quality, single lens reflex cameras. Figure 2 reveals an astounding degree of price dispersion. Compare $635.00, the highest asked and paid in Ann Arbor for reflex camera, with the corresponding frontier price of $170,00. Or note that, for variety 3, prices charged by Ann Arbor retailers run from $310.00 to $450.00, a range of $140,00.

Instead of precipitously reaching for a verdict of substantial "informational imperfection," let us check out alternative explanations. First, we ask whether the omitted characteristics of retailers could account for what we see. This is a tempting hypothesis. Could it be that the stores quoting the lowest price for a particular variety have poor reputations for service, reliability, and amenities? Alas, of the 11 retailers, none tends to be consistently low or high in price. So the facts give no comfort to this plausible speculation.

What of characteristics of cameras that do not enter the qualities of scores? The quality scores of Consumers Union are based primarily on "convenience, over-all optical quality of normal lenses, and freedom from blur in hand-held use at 1/30 and 1/60 seconds." Durability, though not measured directly, is unlikely to affect the results. [According to a letter from Monte Florman, Technical Director of Consumers Union, CU's engineers examined all of the tested cameras for evidence of poor durability. None was found.- Hence, it appears unlikely that the varieties vary significantly in durability. But without a valid test, one cannot be certain.] Concerning such matters as appearance, prestige of the maker, and weight, you will have to form your own judgment. It would be the author's judgment that differences attributable to these factors would not account for the difference between above-frontier and corresponding frontier prices.

A third consideration is the overlap of product and market sets. As for "product," Consumers Union has assessed cameras for the typical purchaser who uses but one lens and not for the serious hobbyist or professional who is concerned with the "family" of accessories compatible with a particular make. Clearly, this could make a difference: the professional/serious hobbyist group might properly be prepared to pay more for a larger family of high quality accessories. But how much more? Would it justify the difference between the frontier price of $170.00 and variety O's lowest price - $550.00 or even variety J's lowest price $359.00? If not, the tentative verdict of an informationally imperfect market would still stand.

FIGURE 1

TERM LIFE INSURANCE: PRICES AND ACCESS IN ANN ARBOR, FEBRUARY, 1975

FIGURE 2

PRICES AND QUALITY IN A LOCAL MARKET: SINGLE-LENS REFLEX CAMERAS IN ANN ARBOR, 1974

The "market" set in Chart 2 is incomplete, since would-be camera purchasers might buy from mail-order sellers whose prices are not chartered. If anything, the inclusion of mail-order prices would not affect the informationally imperfect interpretation since some mail-order prices might be lower but none are likely to be higher than the prices already depicted on Chart 2. As a fourth consideration, it certainly is possible that a different, fully informed assessor of quality might score the quality of varieties differently from Consumers Union. But how different? There is no way of knowing for sure whether this would alter the general configuration of the price-quality map. But a rereading of the Consumers Union report convinced the author that Consumers Union's assessment was persuasive. We do not know where you will arrive in this critical review of Chart 2. The author tentatively concludes that the near-zero correlation between price and quality is largely chargeable to consumer ignorance and that this market may be validly characterized as informationally imperfect.

CORRECTIVE POLICIES AND EDUCATION-INFORMATION-PERSUASION AGENTS

Dependence Upon Research Outcomes

If most markets are found to be informationally perfect or near-perfect--not my expectation--there is only one thing to do: proclaim it' No corrective policies would be necessary. If, however, markets exhibit widespread informational imperfections, then consideration of corrective policies are in order. A finding of widespread information imperfections, would imply that the education-information-persuasion agents in our economy have been less than optimally effective. It follows that corrective policies should be directed toward these agents.

Corrective policies might have three objectives:

1. To change the character of agents' activities;

2. To redistribute resources among various agents;

3. To create new agents or institutions.

Our first step is to identify the agents and to characterize their current activities and scope.

The Education-Information-Persuasion Agents

We identify five:

1. Salesmen/advertisers, supported by market research- ers;

2. The Educational Estate:

a. Economists;

b. Consumer educators in many guises: consumer economists, family economists, family management and home economics educators in universities, colleges, extension/adult education, high schools;

c. Consumer advocates, activists and their organizations;

3. Consumer product testing organizations;

4. The Media: the non-advertising components of television, radio, periodicals, and newspapers;

5. Government officials, elected and appointed communicating through the media and their own channels (publicly controlled media and publications).

For reasons of economy (and mercy!), we will describe the activities of the first three and prescribe policies affecting only salesmen/advertisers and consumer pro-duct-testing organizations.

Agents and the Information-Choice Process

In dealing with the information-choice process our agents tend to favor either of two models: (1) the rational model, or (2) the "realistic" model. The rational model--by which consumers deliberately seek to obtain and sift information and then to optimize--is embodied in the thinking, teaching, and publications of the educational estate, the consumer product-testing organizations, the non-advertising component of the media, and some of the activities of government officials.

Under the realistic model, by contrast, the decision-maker receives information, incidentally, either as the recipient of advertising/sales "messages," or from the shared experience of friends and acquaintances and then, perhaps, optimizes. This is the approach and model to which most advertising and selling is attuned.

Since the market research fraternity is split between a support role for salesmen/advertisers and membership in the educational estate, it should come as no surprise that market researchers pay attention, usually in different times and different places, to both the rational and the realistic model.

We turn now to a detailed characterization of the activities of the various agents and to possible corrective policies to which informationally imperfect markets seem to point.

Salesmen/Advertisers

By any standard the resources utilized by salesmen/ advertisers are substantial. In 1970 they accounted for 11% of aggregate consumption expenditures of $67 billion, distributed as follows:

Advertising        $21  B)

Public Relations     3  B)        $31 Billion

Direct Mail             7  B)

Personal Selling    36  B)

                          $67  B, or 10.9 percent of aggregate consumption expenditures in 1970.

[These estimates were supplied by Professor Ivan Ross, University of Minnesota. For an explanation of the estimates, see (Maynes, 1976, p. 323-324).]

With a broad brush we shall characterized each of our agents by the type, amount, and quality of information it provides. Our first category is existence information--communicating the existence of products, brands, varieties, retail outlets. It may be confidently stated that salesmen/advertisers provide this type of information almost always, copiously, and effectively. A second type of information that consumers need is an understanding of the functioning of markets, the contribution of salesmen/advertisers to an understanding of markets is probably negligible.

Consumer Product Testing Organizations

[The reader should know-and may wish to discount therefore-that the author has been a Board Member and Chairman of several Board Committees of Consumers Union since 1968.]

There are only two in the United States-Consumers Research, Inc. which was the first and Consumers Union, which is by far the largest and most influential. In view of the prominence they have achieved the resources they utilize are extraordinarily modest. In 1970 they totaled 13 million, or 0.002% of aggregate consumption expenditures (Maynes, 1976, p. 324).

What is the nature of the information they provide? For roughly one hundred and fifty product classes, mostly big-ticket and nationally branded, these two organizations provide existence information regarding the brand/models available, occasionally, calling consumers' attention to the existence of a new or particularly exotic product. Sometimes they try to improve the consumer's understanding of the operation of some markets, but this is not a major concern. The two types of information on which they concentrate are general product information and variety-specific quality information. The general product information they provide is copious in quantity, highly concentrated, and carefully communicated. Exactly the same comments apply also to their variety-specific quality information. But several additional comments are in order. Their quality information is usually restricted to characteristics that may be measured reproducibly. This implies that, for the most part, products whose chief characteristics are subjective are not usually tested. A second limitation is that these national organizations provide no information regarding the local sellers from whom the tested products may he purchased. Finally, both organizations take great pains to give the reader sufficient information so that he can make his or her own quality assessment should he want.

As in the case of sellers/advertisers, the information provided by these organizations is greatly influenced by the organizations' objectives. Their objectives enforced by their boards, by the market demand for their information, by their ideology, and by libel laws is to provide complete, reproducible comparative quality information. Combine these objectives and the attendant minuscule resources in the ands of these organizations and the outcome should not be surprising: high quality general and specific product information for a tiny fraction of the products that consumers buy.

Summing up, we find salesmen/advertisers dedicating tremendous resources to persuasion efforts that yield a modicum of reproducible information and consumer product testing organizations dedicating minuscule resources to the production and distribution of a small but high quality pool of general and specific product information. This summation point brings us to a discussion of two alternative corrective policies-proposals for: (1) a countervailing consumer information and (2) a local consumer information system. These will be discussed in turn.

Pro-Consumer Information As a Corrective Measure

The case for the allocation of resources to the cause of pro-consumer information is not hard to make. In 1970 the resources controlled by seller/advertisers exceeded those controlled by consumers by a ratio of 5,000 to 1. The analysis just completed suggests that seller-controlled information is long on persuasion and short on solid information. If these two conditions together are associated with widespread informationally imperfect markets--the point our research is designed to ascertain--then one corrective policy comes to mind almost automatically: provide consumer-control-led organizations with more resources. To put it another way the objective would be to establish consumer-controlled information-persuasion agents as effective countervailing agents vis-a-vis salesmen/advertisers.

The essence of this proposal is that a consumer controlled organization or organizations should have at their disposal substantial resources. The organization would be essentially free to allocate these resources as they will. One possibility would be that the consumer countervailing information organization would undertake product testing and information provision along the lines of Consumers Union, but on a much larger scale. But greater resources would also enable the consumer information organization to utilize the persuasive tools of advertising, to put it more personally, such an organization could put both Madison Avenue and a substantial fraction of the market research profession on the consumers' payroll!

How might such an organization be financed? Two alternatives seem plausible. First, a modest sales tax might be imposed on aggregate consumption for the benefit of "consumer education." In terms of 1970 numbers a tax as low as .0025% (1/4 of 1%) would have yielded $1.75 billion or roughly 135 times as much as the consumer product testing organizations had at hand that year. Alternatively, it might be argued that information-persuasion resources should be shifted from business to consumers. Such a logic might suggest the imposition of a tax on business expenditures on information and persuasion. Based again on estimates for 1970, a one percent tax on this base would have yielded $670,000,000. The base for the tax might be expenditures on advertising plus one-half of expenditures on the wage bill for sales personnel. The choice between these financing alternatives should perhaps be made in terms of political feasibility.

Proposal for a Local Consumer Information System

Suppose that many local markets are informationally defective and suppose further that the required information is largely local in character. Your "suppositions'' make the case for the development of a local consumer information system. What would a local consumer information system do? It would provide the following four types of information:

1. Local Price Information. Where is the cheapest place in my shopping area to buy products characterized by relatively little quality variation, such as term life insurance?

2. Local Accessibility to Products Quality-Rated by Consumers Union. What does the local price-quality map look like? What models lie on the perfect in- formation frontier? What retailers sell these models at the frontier prices? For a given model, what range of prices is available?

3. Experience Rating of Vendor Services. Where should I take my disabled television set (or car, high-fidelity system, child, or other consumer durable) to have it repaired cheaply and effectively?

4. Consumer Ratings of Retailers. What have been consumers' factions to their purchase experience with particular local retailers-their advertising, dealings with salespersons, their promptness, their post-purchase service, refund experience, and correction of consumer grievances?

The objectives of such a local consumer information system are four: (1) to provide individual pay-offs to the system's users; (2) to improve the working of the particular local market in which it operates, (that is, to make the market informationally-near-perfect or perfect); (3) to reproduce itself and thus to improve the functioning of markets other than the one in which it was spawned; and (4) to document the informational perfection or imperfections of local markets over many types of products and areas of varying population densities. For a detailed discussion of the proposal for a local consumer information system, see (Maynes, Morgan, Vivian and Duncan, 1976).

This paper will not discuss details of implementation. But if the case can be made for the establishment of local consumer information systems, many in our professions will be able to make a career in the perfecting of such systems.

REFERENCES

Benson and Benson, "Survey of Present and Former Subscribers to Consumer Reports," unpublished report (Mt. Vernon, N.Y.: Consumers Union, 1970).

J.K. Galbraith, The New Industrial State (New York: Mentor, 1969).

David M. Gardner, "Deception in Advertising: A Conceptual Approach," Journal of Marketing, 39 (January, 1975), 40-46.

Paul E. Green and Yoram Wind, Multi-Attribute Decisions in Marketing (New York: Holt, Rinehart and Winston, 1973).

E. Scott Maynes, Decision Making for Consumers, An Introduction to Consumer Economics (New York: Macmillan, 1976a).

E. Scott Maynes, "The Concept and Measurement of Product Quality," in Nester E. Terleckyj (ed.), Household Production and Consumption, Studies in Income and Wealthy Volume Forty (New York: National Bureau of Economic Research, 1976b), 529-560.

E. Scott Maynes, James N. Morgan, Weston Vivian, and Greg J. Duncan, "The Local Consumer Information System: An Institution to Be?" Proceedings of the American Council on Consumer Interest, April, 1976.

Nester E. Terleckyj (ed.), Household Production and Consumption, Studies in Income and Wealth, Volume Forty (New York: National Bureau of Economic Research, 1976).

William L. Wilkie, How Consumers Use Product Information (Washington: Government Printing Office, Stock No. 038-000-0002376, 1975).

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