The Costs of Prohibiting Deceptive Advertising--Are They As Substantial As Economic Analysis Implies?

ABSTRACT - Research on consumer perceptions of advertising is used to question the implication from economic analysis that the prohibition of deceptive claims may create a net cost to consumers by rendering advertising unable to convey the truth accurately and/or by causing advertising to omit additional truthful information. An experiment involving original vs. rewritten versions of deceptive ads suggests that ads may easily be rewritten to convey the truth in place of the falsity and also to avoid elimination of other useful information.

Citation:

Ivan L. Preston and Jef I. Richards (1989) ,"The Costs of Prohibiting Deceptive Advertising--Are They As Substantial As Economic Analysis Implies?", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 209-214.

Advances in Consumer Research Volume 16, 1989      Pages 209-214

THE COSTS OF PROHIBITING DECEPTIVE ADVERTISING--ARE THEY AS SUBSTANTIAL AS ECONOMIC ANALYSIS IMPLIES?

Ivan L. Preston, University of Wisconsin-Madison

Jef I. Richards, University of Texas-Austin

ABSTRACT -

Research on consumer perceptions of advertising is used to question the implication from economic analysis that the prohibition of deceptive claims may create a net cost to consumers by rendering advertising unable to convey the truth accurately and/or by causing advertising to omit additional truthful information. An experiment involving original vs. rewritten versions of deceptive ads suggests that ads may easily be rewritten to convey the truth in place of the falsity and also to avoid elimination of other useful information.

INTRODUCTION

Although a traditional guiding proposition of advertising regulation has been that the prohibition of deceptive claims is entirely beneficial, the economics analysis that has become popular in the 1980s has offered the countering proposition that such prohibition may have costs as well as benefits. It could, therefore, in some cases produce a net loss that renders the prohibition socially unjustifiable.

A recent expression of that position, however proposed that net losses are far more likely to occur than the present authors believe is likely. This paper, therefore, presents consumer perception research designed to test whether the benefit of eliminating a deceptive claim is likely to be accompanied by undesirable costs resulting from additional unwanted changes in the perception of advertising messages.

The present authors have no disagreement with the argument of Craswell (1985) that the Federal Trade Commission (FTC) and other regulators are wrong in assuming that prohibiting deceptive advertising creates only benefits and no risks or costs. "The cost of preventing deception is rarely zero," he says (p. 679), and he discusses extensively the variety of potential costs. A major point is that when ad claims convey truthful and useful information to some consumers although false beliefs to others, removing such claims will remove the good effects as well as the bad, and possibly create additional bad effects. When the result is a net social loss or at best a benefit not free and clear of accompanying costs, the best social decision may be to allow the advertising to continue.

We wholly agree that a system geared to balancing the losses in regulatory actions against the gains is superior to the FTC's rigid declaration that "Deception. . . never offers increased efficiency or other countervailing benefits" (International Harvester 1984, p. 1056). He who never looks is certain never to see.

To cope with this problem, Craswell proposes the standard that "An advertisement is legally deceptive if and only if it leaves some consumers holding a false belief about a product, and the ad could be cost-effectively changed to reduce the resulting injury" (p. 678). The first portion of the statement essentially reflects the FTC's traditional standard, although that standard, by specifying deceptive capacity rather than actual deception, actually involves the existence only of the potential for creating false belief rather than actual false belief. The second portion of the statement reflects Crassell's proposed addition to the standard.

The problem we see with Craswell's position lies not with such standard but rather with the suggestion he conveys that ads often cannot be cost-effectively changed, and that substantial costs, therefore, will often occur, and that, therefore, it will often be socially desirable to avoid prohibiting messages even though they meet the traditional criterion of having the potential to create false beliefs. We feel that substantial costs often will not occur, and thus that Craswell's proposed chance, although fair and appropriately reflective of the emerging economics-based analyses of the Reagan years, would upon implementation make very little difference in the extent to which advertising claims are prohibited.

Our basis for this belief lies in existing research and theory on consumer perception of advertising claims. Seminik (1980) found that corrective ad claims could be written without adversely affecting brand image. Russo, Metcalf & Stephens (1981) concluded that "One of the striking findings of our experiment is that the main thrust of the ads was not attenuated by removal of the misleading claim" (p. 130). Oliver, Walbridge & Rheinstein (1984) found that rewritten versions of deceptive ads reduced consumer perception of the deceptive claims. Preston and Richards (1985) argued that many FTC cases prosecute advertising claims that create "eradicable miscomprehension," which means their potential for deceptiveness can be eliminated easily and without the substantial concurrent costs that Craswell fears.

A contrary finding was shown by Jacoby, Nelson & Hoyer (1982). However, they examined additions to ad copy dictated by the FTC rather than the more typical elimination of prohibited claims that would be handled by the ad industry's professional copywriters. We may presume reasonably that any changes made under Craswell's standard would be handled by professionals.

The analysis that follows adopts a method similar to that of Oliver et al along with the framework of Preston & Richards' (1985) discussion of miscomprehension, which was based on Jacoby, Hoyer & Sheluga's (1980) explication of that term. Miscomprehension occurs when the message the consumer sees conveyed by an advertisement is inaccurate in terms of the ad's literal content. When the consumer examines an ad and then reports an exact restatement or trivially different rephrasing of the literal content, or an implication that can be logically derived from the literal content, he/she has comprehended the ad's content. When any other message is reported to be conveyed, he/she has miscomprehended.

Although a certain degree of miscomprehension is a normal part of human behavior, some messages may produce more of it than others. In fact, certain advertisements may represent the phenomenon of "induced miscomprehension," which means they have been deliberately written to produce more miscomprehension than they would if written more simply and clearly. If so, these highly miscomprehended messages presumably could be rewritten in a simpler style that will produce far less miscomprehension. Miscomprehension that can be so eliminated or reduced amounts to "eradicable miscomprehension," while that which cannot be eliminated because the literal content already takes the least ambiguous form constitutes "ineradicable miscomprehension."

An examination of the case record suggests that the instances of deceptiveness prosecuted by the FTC typically appear to involve high degrees of eradicable miscomprehension. Under such circumstances, it seems reasonable to contend that the benefits to be gained by eliminating deceptiveness can be obtained by simple, easy rewriting that has virtually no risks or costs.

METHODOLOGY

To demonstrate this we conducted an experiment in which students at our university saw advertising claims prohibited by ten recent FTC orders. Each of ten pieces of advertising copy was seen by each student in one of two versions. The first version had the wording actually used in the advertising, which the FTC had alleged could create in consumers' minds a false implication. In these and numerous other FTC cases the false implication was neither stated in nor logically derivable from the literal content, and, therefore, consumer's tendencies to report it as conveyed would represent miscomprehension.

The second version of each ad presented the same content rewritten to prevent consumers from seeing the false implication conveyed. The observed degree of miscomprehension, we of course reasoned should be significantly less for this rewritten ad content than for the original content.

For example, in Adria Laboratories (1984) the complaint charged that ads for Efficin pain reliever represented falsely that "Efficin is not associated with most of the side effects and contraindications with which aspirin is associated." This was implied by the complaint to occur because the advertising stated literally that the product "contains no aspirin." The ad did not state literally the alleged misrepresentation nor say anything else that appeared to pertain to it. The alleged misrepresentation, therefore, was not a logical implication from the ad content, and its conveyance to consumers would constitute miscomprehension by those consumers. The alleged misrepresentation would constitute, in Craswell's terms, a pragmatic implication (Craswell 1985, p. 716- 19).

The FTC complaint had asserted the truth to be that "Efficin is similar to aspirin, and that the ingredient in Efficin has been associated with most of the same side effects and contraindications as aspirin." An additional complaint charge was that the failure to disclose this truth was a separate violation. The order, to which the advertiser agreed through a consent settlement, therefore forbade the use of "contains no aspirin," or of any other representation comparing the product's safety to any product containing aspirin, unless the following statement or suitable equivalent was clearly and prominently disclosed: "Efficin . . . has side effects similar to aspirin." This disclosure presumably would eliminate the miscomprehension producible by "contains no aspirin" and 50 eliminate the alleged misrepresentation and the alleged failure to disclose.

In our experiment the student subjects were shown slides of simulated ads consisting of verbal content drawn from the actual ads. Substitute product names were used to eliminate any effect of prior familiarity. For the Efficin ad, the simulation reflecting the actual verbal content of the ad stated: "Introducing Extra Strength Nalgecin tablets. Fast pain relief that contains no aspirin or acetaminophen. Recommended and prescribed by physicians for many years. Fewer tablets needed because of the extra strength."

The rewritten version said, "Introducing Extra Strength Nalgecin tablets. Fast pain relief. Recommended and prescribed by physicians for many years. Fewer tablets needed because of extra strength. While it contains no aspirin or acetaminophen, users who experience side effects from aspirin may have similar reactions to Nalgecin."

Each subject saw ten such ads, five representing actual content, five representing our rewritten content. Each person was shown only one version of each ad, and so could not know that two versions existed. Sixty subjects were used, with each version of each ad seen by 30. After seeing each ad, subjects were asked to respond on 7-point agree-disagree scales (1 = maximum agreement; 7 = maximum disagreement) to seven statements, with the instructions: "Place an X on the scale in the position which best represents your degree of agreement or disagreement that the advertisement made you feel that it was trying to convey this claim to you."

Within each set of seven statements associated with each ad, one statement reflected the truth with respect to the misrepresentation alleged by the FTC; for Nalgecin this statement was "May cause side effects associated with aspirin." A second statement was worded to be false with respect to the same topic; for Nalgecin it was "Avoids side effects of aspirin." A third statement reflected truthfully a different claim made in the ad, irrelevant to the charged misrepresentation and presumably conveying useful information; for Nalgecin this statement was "Works fast." The remaining four additional statements were based in various ways on the content of the ad, and were included only to pad the task so that undue attention would not be paid to the three critical statements.

Concerning those three, we hypothesized that the "true" and the "false" statements would receive different response for the actual and rewritten versions, reflecting the expectation that subjects would see the alleged misrepresentation less and the truth more in the rewritten version. Thus we expected the "true" statement would be agreed to more in the rewritten version, meaning the average number score given it would decrease, and the "false" statement would be disagreed with more (its average number would increase). We also hoped that the "irrelevant" statement, based on content not varied from actual to rewritten version, would not be treated differently from one to the other. If it did change, it would confirm Craswell's concern that prohibiting a challenged claim might create costs by eliminating good effects along with the bad.

RESULTS

For Nalgecin. the "false" statement for the actual version of the ad received a mean response of 3.13 (1 = maximum agreement), showing that a majority of subjects agreed, thereby seeing the false representation just as the FTC alleged they would. (See Table 1 for Adria Laboratories). Seventeen of them were on the "agree" side of the scale, 8 were on the disagree side, and 5 were on the midpoint (neither agreeing nor disagreeing). The FTC need not assert that the capacity to deceive exists for all consumers; rather, it requires that the phenomenon potentially occur only for a "substantial portion," with 20% usually being sufficient. Therefore, the degree of deceptiveness found for Nalgecin, although not the maximum possible, was well within the amount the FTC should find conclusive.

For the rewritten version of the ad, the mean for the "false" statement was 6.70, showing almost maximum disagreement that the false representation was conveyed. This time only one subject was on the "agree" side, with 28 on the "disagree" side and one on the midpoint. Thus in the case of the Nalgecin ad the rewriting produced the effect we expected of reducing the tendency for the challenged deceptiveness to occur. The difference of 3.57 was highly significant by one-way analysis of variance, between-subjects .

The "true" statement showed consistent results. For the actual version of the ad, the score of 5.70 showed subjects disagreeing about conveyance, thus agreeing with the FTC charge of failure to disclose. Twenty-three subjects disagreed, two agreed, and five took the midpoint. But for the rewritten version the score was 1.37 (29 agreeing, one disagreeing), strong evidence that the failure to disclose was corrected. Again, the difference was highly significant (Table 1).

There was also the "irrelevant" statement, irrelevant to the deceptiveness charge and not changed between versions. For Nalgecin it received a mean score of 2.33 for the actual version and 2.47 for the rewritten version, a difference so small as to constitute no evidence that the two versions prompted different responses. In other words although the differences in response between actual and rewritten version were statistically significant for "true" and "false" statements, they were not for the irrelevant statements. With respect to the questions raised by Craswell about benefits and costs, this translates to a finding that the advertisement was successfully rewritten so as to eliminate the misrepresentation, yet without eliminating other information of value to consumers.

The same results occurred for each of the ten sets of three statements, representing the ten ads studied. The other nine ads involved generally similar situations to that for Efficin; for example, the alleged false implication about Daily Greens, a food supplement made of dehydrated vegetables, was that its use was associated with a reduction in incidence of some hum.an cancers (Pharmtech Research 1984). Our version of the actual ad included the statement, "According to the National Academy of Sciences, a regular diet of cruciferous vegetables... is associated with a reduction in the incidence of certain cancers." The FTC did not call that statement literally false, but called it deceptive in conveying the alleged false implication. Our rewritten version, therefore, added to the above statement (with name masked as Vegetables) the additional wording the "While !2.e National Academy of Sciences has not said the Vegetables will help to reduce cancers, like the vegetables from which they are made, aren't they a better gamble than no cruciferous vegetables at all?"

The differences between actual and rewritten ads for the true and false statements were not as great for the other nine ads as they were for the Efficin (Nalgecin) case. Still, in every instance (Table l) they changed in the direction of reducing subjects' tendency to see conveyance of the alleged misrepresentation, although in some cases not to a statistically significant degree. Averaged over the entire set of ten ads, however, the differences were statistically significant for both the true and false statements.

For the irrelevant statements, the difference averaged across the ten ads was not significant. Three of the ten individual cases showed a significant change, although only one of them involved a reduced tendency to see the statement conveyed. In the other two there was actually an enhanced conveyance of the irrelevant statement.

We are disappointed that the differences in scores for the true and false statements between actual and rewritten versions were not greater than they were. We had hoped to virtually eliminate, not merely reduce, the tendency for ads to convey the deceptive claims. Perhaps the rewrites needed greater emphasis and firmer focus on the critical points. Still, although we are only amateur copywriters we reduced the miscomprehension considerably. For both true and false statements the mean shifted across the midpoint of 4.00 to the opposite side of the scale, meaning that the false implication switched from a net conveyance to a nonconveyance. Meanwhile, the net treatment of the irrelevant claim, which conveyed true information that was useful to consumers, did not shift significantly from its p( s, ions of net conveyance. (It was only a slight net conveyance, due probably to the fact that the allegedly deceptive claims typically were given stronger graphic treatment than were these ancillary truthful claims.)

TABLE 1

RATINGS OF "FALSE" STATEMENTS

RATINGS OF "TRUE" STATEMENTS

RATINGS OF "IRRELEVANT" STATEMENTS

DISCUSSION

Our principal conclusions, then, harking back to the ideas of Craswell which prompted the experiment, is that the reduction of deceptiveness appeared to be accomplishable easily, bringing benefits without significant costs. Advertisers inclined to do so could achieve these effects not in rewriting but in their original writing They are professionals in communication. When their copy was obscure, we feel it was often because they deliberately strove for a writing style that encouraged deceptive implications to occur. Indeed, there may be greater costs involved in achieving such masterful obscurity than in merely writing plainly.

Which of the following, for example, does the reader think took longer to concoct? (1) the phrase "Anacin's ingredient is aspirin," which states the simple unambiguous truth, or (2) the phrase "Anacin contains the pain reliever that doctors recommend most," which the FTC thought would convey, in conjunction with such references as "special fortified formula," the belief that Anacin's ingredient was something more special than ordinary aspirin? (American Home Products 1581). 'The identity of Anacin's ingredient is in every single instance obscured," said the FTC opinion. "The strained syntax of many of the advertisements ... fosters the impression that Anacin contains something other than aspirin." In another case the opinion stated that "Like much advertising we find deceptive, the ads are drafted with a artful choice of words" (Thompson Medical 1984).

The same point might be made for an example used by Craswell (1985, p. 674), that of the Chrysler advertising which described a six-cylinder car's superiority over the six-cylinder Chevrolet Nova in such a way as allegedly to imply that all of its models had the same superiority over all Nova models. There is no reason to doubt that any copywriter who wished to restrict the comparison could easily have done so in an unambiguous manner.

Ineradicable miscomprehension poses a serious problem for successful communication in any walk of life, but in advertising most of the miscomprehension that occurs is eradicable. That most advertisers know this full well is shown by the recurrence of certain types of implications again and again in FTC cases. Preston (1977) identified a dozen types of implications attacked by the FTC, most of them frequently enough that advertisers could certainly be said to be on notice as to their existence and their standing at law.

Research such as ours might also place advertisers on notice. Craswell observed that "One reason it often appears difficult to tell whether changes in an advertisement would have been beneficial is that evidentiary and measurement techniques have never been fully developed--in part, because this has never been recognized as a legally relevant issue" (p. 727). That is a worthy point, but there is no reason why that situation should stay the sa.ne in the future. The evidence used in litigated cases to confirm the existence of false implications consists typically of the advertisers' own copytesting, thus they clearly know the effects on message conveyance of what we called in our experiment the "Actual" version of the ad. All the FTC need do is use the same copytesting methods on a "rewritten" version and the evidentiary problem will be solved. It seem unlikely that either the ability to eradicate miscomprehension or the minimal cost of doing so can long remain phenomena that advertisers will deny.

REFERENCES

Adria Laboratories (1984), FTC Decisions, 103, 512-27.

American Home Products (1981), FTC Decisions, 98, 136 -427.

Commodore Business Machines (1985), FTC -- Decisions, 105, 230-45.

Craswell, Richard (1985), "Interpreting Deceptive Advertising," Boston University Law Review, 65 (4), 660-732.

Cynex Manufacturing (1984), FTC Decisions, 104, 464-77.

Dancer-Fitzgerald-Sample (1980), FTC Decisions, 96, 1 -17.

D'Arcy-MacManus & Masius, (1982), FTC Decisions, 99, 324-44.

Descent Control (1985), FTC Decisions, 105, 280-90.

Estee Corp. (1983), FTC Decisions, 102, 1804-12.

International Harvester (1984), FTC Decisions, 104, 949- 1088.

Jacoby, Jacob, Wayne D. Hoyer, and David A. Sheluga (1980), Miscomprehension of Televised Communications, New York: American Association of Advertising Agencies.

Jacoby, Jacob, Margaret C. Nelson, and Wayne D. Hoyer (1982), "Corrective Advertising and Affirmative Disclosure Statements: Their Potentia! for Confusing and Misleading the Consumer," Journal of Marketing, 46 (Winter), 61-72.

Litton Industries (1981), FTC Decisions, 97, 1-83.

Oliver, Richard L., R. Hoyt Walbridge, and Peter H. Rheinstein (1984), "A Study of Physicians' Perception of Advertising Judged Deceptive by the FDA," Advances in Consumer Research, 11, 224-28.

Pharmtech Research (1984), FTC Decisions, 103, 44860.

Preston, Ivan L. (1977), "The FTC's Handling of Puffery and Other Selling Claims Made By Implication,"' Journal of Business Research, 5 (June), 155-81.

Preston, Ivan L. and Jef. I. Richards (1985), "The Relationship of Miscomprehension to Deceptiveness in FTC Cases," Advances in Consumer Research, 13, 138-42.

Russo, J. Edward, Barbara L. Metcalf, and Debra Stephens (1981), "Identifying Misleading Advertising," Journal of Consumer Research., 8 (Sept.). 119-31.

Seminik, Richard J. (1980), Corrective Advertising: An Experimental Evaluation of Alternative Television Messages," Journal of Advertising, 9 (No. 3), 21-30.

Thomas L. Baker (1982), FTC Decisions, 100, 46187.

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