The Effects of Alcoholic Beverage Advertising and Marketing Practices: the Current State of Affairs

ABSTRACT - This paper addresses the current controversy over the effects of alcoholic beverage advertising focusing on the need for valid research in the area. Based on a literature review that highlights the current lack of relevant research the paper discusses the methodological problems confronting researchers in this area and presents a series of research opportunities.

Citation:

Edward T. Popper (1986) ,"The Effects of Alcoholic Beverage Advertising and Marketing Practices: the Current State of Affairs", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 442-445.

Advances in Consumer Research Volume 13, 1986      Pages 442-445

THE EFFECTS OF ALCOHOLIC BEVERAGE ADVERTISING AND MARKETING PRACTICES: THE CURRENT STATE OF AFFAIRS

Edward T. Popper, Northeastern University

ABSTRACT -

This paper addresses the current controversy over the effects of alcoholic beverage advertising focusing on the need for valid research in the area. Based on a literature review that highlights the current lack of relevant research the paper discusses the methodological problems confronting researchers in this area and presents a series of research opportunities.

INTRODUCTION

On April 16, 1985 the Federal Trade Commission (FTC) rejected a petition from Center for Science in the Public Interest (CSPI) which had requested increased regulation of alcoholic beverage advertising. In rejecting the petition the FTC stated that they found "no reliable basis on which to conclude that advertising significantly affects alcohol abuse. Absent such evidence, there is no basis for concluding that rules banning or otherwise limiting alcohol advertising would offer significant protection to the public (Rock 1985)."

Obviously the FTC rejection of the CSPI petition is not the end of the alcoholic beverage advertising controversy. However, the basis for the rejection highlights a central issue facing all regulators (and legislators) in this area - the near total lack of valid and reliable research on the effects of alcoholic beverage advertising. Until this research vacuum is filled the controversy is likely to be characterized by adversarial posturing rather than reasoned analysis.

The purpose of this paper is to stimulate research in an area that provides researchers with an opportunity to contribute to an issue of pressing public concern as well as to confront significant methodological challenges. In order to provide the context for the research issues, the paper will, first, briefly describe the current public policy debate (highlighting the role of research in the various proceedings). Next, the paper will review the current state of research in the area (including both consumer and economic research). Following the literature review will be a discussion of the methodological issues confronting researchers in this area. Finally, the paper will address some of the research needs and opportunities in the area of the effects of alcoholic beverage advertising.

THE CURRENT CONTROVERSY

The sales and advertising of alcoholic beverages have been controversial since the repeal of prohibition. However, the controversy had traditionally focused on the sale of the product rather than advertising and other marketing practices with the principal advertising control being the voluntary NAB ban on advertising for distilled spirits. In recent years this has changed as some states have limited or banned broadcast advertising of any alcohol beverages (i.e., Mississippi and Oklahoma) and other states limited certain promotional practices (for example, in 1984 Massachusetts banned happy hours).

On November 21, 1983 this controversy was brought to the FTC in the form of a petition from CSPI ( and cosigned by approximately thirty other organizations and individuals) requesting that the FTC act against alcoholic beverage advertising. Among the requested actions were a ban of all alcoholic beverage advertising, promotional, and marketing practices that: were aimed at or reach large numbers of youth or heavy drinkers; use athletic, show business or music celebrities; depict alcohol use in connection with risky activities; use subliminal techniques; link consumption of alcoholic beverage consumption with social, sexual, or business success or athletic prowess; encourage excessive consumption; and/or discourage the treatment of alcohol abuse. The petition also requested that any remaining print advertising for alcoholic beverages be required to include one of a series of rotational warnings (Jacobson et al 1983, pp. 21-24).

While the CSPI petition provided a limited amount of empirical support for its request it contended: "While the available literature does not demonstrate a causal connection between (alcoholic beverage) advertising and harm, the evidence is sufficient to justify a Commission action in light of the disastrous health and social consequences of alcoholic beverage consumption (Jacobson et all 1983, p. 20)." The petition went on to describe the allegedly unfair and deceptive techniques in great detail without providing empirical data on the effects (or even interpretation) of those techniques. The petitioners supplemented their argument by presenting a monograph, The Booze Merchants, prepared by CSPI which purported to analyze the advertising and marketing practices of the alcoholic beverage industry (Jacobson, Atkins and Hacker 1983). While the monograph provided a number of examples of alcoholic beverage advertising the interpretation of its effects appeared to be based solely on the authors' subjective judgements rather than any empirical research.

The FTC's initial response to the CSPI petition was to instruct its staff to undertake a careful review of the petition and its charges and to make a recommendation to the Commission. The report resulting from this investigation recommended that the petition be denied (Levine et al 1985). The Commission Staff took the position that given that alcoholic beverage consumption was legal and might even have some beneficial effects when consumed in moderation (c.f. Gordon and Gordon 1984) simply demonstrating a linkage between alcoholic advertising and consumption of alcoholic beverages would be insufficient. Rather, it would be necessary to demonstrate that the advertising resulted in alcohol abuse.

The FTC staff report (Levine et al 1985) included a detailed review of the literature (including that research cited in the CSPI petition) which concluded that there was insufficient reliable evidence linking alcoholic beverage advertising causally with consumption, much less abuse. Indeed the report cites a number of aggregate studies that actually fail to show any relationship between advertising and consumption (Levine et al, Appendix, A, p. 10).

While denying the industry wide bans and rules requested by CSPI, the Commission indicated that it would "continue its ongoing review of alcohol advertisement and marketing practices to identify any that warrant challenge as deceptive or unfair under Section 5 of the FTC Ace (Rock 1985). It is clear that the Commission's need for research evidence linking the advertising practice to alcoholic beverage consumption and abuse will be as critical in the case of an individual advertiser as it was in considering industry-wide intervention. Thus it might be inferred that without such research future FTC action is unlikely.

At the same time as the FTC was considering CSPI petition, the anti-alcoholic beverage advertising forces were opening a second front. An organization called SMART (an acronym for Stop Marketing Alcohol on Radio and Television) was organizing a petition drive on Congress. Organized by many of the same groups who were co-signers of the CSPI petition to the FTC, SMART was attempting a major grass roots lobbying effort aimed at convincing Congress that their constituents wanted an end to broadcast alcoholic beverage advertising. As of this writing congressional hearings on the anti-alcohol advertising issues had not been scheduled but were clearly in the works.

Another Federal agency that is currently addressing issues of alcoholic beverage advertising is the Bureau of Alcohol Tobacco and Firearms (BATF). An arm of the US Department of the Treasury, the principal alcohol related task of BATF is the collection of alcohol taxes. The BATF also is the Federal agency with statutory authority to regulate alcoholic beverage advertising. In this role they issue Trade Rules and Regulations (for example, the long standing ban on active athletes in alcoholic beverage advertising is a BATF rule) and preclear all voluntarily submitted alcoholic beverage advertising (by way of contrast, the FTC will not preclear any advertising both as a matter of policy and in an effort to avoid prior restraint of speech).

The BATF has had an ongoing rulemaking proceeding, since 1978, examining essentially the same issues as those raised in the CSPI petition to the FTC. Indeed the BATF was the recipient of a nearly identical CSPI petition. In the seven years that the BATF has been engaged in their current rulemaking there has been little indication of progress towards arriving at a rule (with the exception of a 1984 ruling reaffirming the BATF's support of nonmisleading advertising and a proposed rule banning subliminal advertising), However, the agency has indicated that it intends to actively pursue a rulemaking incorporating industry self-regulatory codes into their rules (George 1984).

Although neither the US Congress nor the BATF faces the same evidentiary burdens as a law enforcement agency like the FTC it is unlikely that either legislative or regulatory action will be based on pure emotion. While both the Congress and the BATF will undoubtedly receive impassioned appeals from parties to all sides of the issue their conclusions will certainly be influenced by the quality of support furnished. Unfortunately, as the next section of this paper will summarize, neither side will have good support to offer.

THE PRESENT STATE OF RESEARCH

Mirroring the policy questions this brief review will address the following issues: first, does alcoholic beverage advertising result in increased alcoholic beverage consumption (or attract new users)? Next, what is the relationship between alcoholic beverage consumption and abuse and what is the role of advertising in this process? Finally, the review will examine the relevance of other research into advertising effects.

In 1976 a Rand Corporation study concluded that there was a total lack of empirical evidence on the relationship between media advertising and alcoholic beverage consumption much less abuse (Comstock 1976). Unfortunately, as of 1985, there has been little change in this situation.

Since the methodological problems confronting researchers looking at the alcoholic beverage advertising - consumption link are less severe than those relating to the advertising - abuse link (see discussion in the next section) most of the existing research has addressed this question. In one of the few internal industry studies of the effects of alcoholic beverage advertising on consumption to be made public, Ackoff and Emshoff found that while a 50% increase in advertising for Anheuser-Busch resulted in a 7% sales increase a 25% decrease in advertising resulted in a 14% sales increase (1975).

Given the applicability of statistical analysis of aggregate industry data to the advertising-consumption question a number of researchers have pursued this approach. Most of these researchers have failed to find a positive relationship (Grabowski 1976: Bourgeois and Barnes 1979; Ashley et al 1980; McGuiness 1980; Duffy 1982; Hagen and Waterson 1983) between alcoholic beverage advertising and consumption. Unfortunately, many of these studies have limitations that, while in no way undermining their validity, reduce their applicability to the issues facing US policy-makers. Many of the studies (e.g. Bourgeois and Barnes; Duffy: Hagen and McGuiness) draw their data from countries other than the US. Given the cultural variations in both the consumption of alcoholic beverages and the exposure to advertising/media it is not possible to reject the hypothesis that the advertising/consumption relationships in another country are different than in the US without supporting data. These studies provide no such data. Further, all of the statistical studies examine the effects of incremental advertising expenditures. Given that policy makers are examining a total ban, incremental data have only marginal relevance. Finally, Ashley, et al, not only fail to find a causal relationship between advertising and consumption, they find evidence of a causal relationship between consumption and advertising (a not too surprising given the tendency of some advertisers to set advertising budgets as a percentage of sales).

Another research approach has been the small scale experiment in which a small number of subjects is exposed, under controlled conditions, to advertising for an alcoholic beverage and then their subsequent consumption is monitored (McCarty and Eqing 1983; Kohn and Smart, 1984; Kohn et al 1984). Only Kohn and Smart found even a temporary effect on consumption and none found an enduring effect. Once again these findings can have only limited relevance to the policy issues (in spite of their internal validity) in that the brief exposure to advertising and the unnatural research environment cannot be readily generalized to the environment in which most advertising exposure and alcoholic beverage consumption occurs. Indeed, the McCarty and Ewing study uses a method so contrived that serious questions can be raised as to the validity of the study. Finally, as will be discussed in greater detail below, one must question the ethicality of experiments in which subjects are induced to consume alcoholic beverages.

A final stream of research into the effects of alcoholic beverage advertising on consumption, content analysis, will be discussed only in passing as it isn't directly relevant to any of the public policy issues. Some of the researchers in this stream appropriately use content analysis to objectively report what themes, illustrations, or situations appear in alcoholic beverage advertising (e.g. Strickland et al 1982). Others use this methodology far more cavalierly, inferring advertiser intent and advertising effect in the complete absence of data supporting such inferences (e.g. Defoe and Breed 1979; Marsteller and Karnchnapee 1980). An unfortunate consequence of this latter type of research is that it appears both superficial and adversarial, thus subjecting all research considered in public policy settings to higher levels of scrutiny (a result which, in the final analysis, night benefit both researchers and policy-makers).

It is obvious that there is a relationship between alcoholic beverage consumption and alcohol abuse. Without consumption there can be no abuse. This is not to say, however, that there is a causal relationship between consumption and abuse. A majority of the people who drink alcoholic beverages in the United States could be termed moderate (or social) drinkers not alcohol abusers (US Dept. of the Treasury and US Dept. of Health and Human Services 1980). Thus consumption leading to abuse is the exception --not the rule.

If alcoholic beverage consumption does not inexorably lead to alcohol abuse, than it is clearly insufficient (given the current public policy debate) for researchers to look only at the advertising-consumption question. Rather, researchers interested in supplying public policy makers with relevant data must examine the relationship between alcoholic beverage advertising and alcohol abuse. Unfortunately only a very few researchers have attempted to confront the methodological obstacles and complexities to work in this area.

One of the largest studies of the effects of alcoholic beverage advertising attempts to address some of the issues surrounding effects on alcohol abuse (Atkin and Block 1980) also is one of the most controversial. While the study indicates relationships between alcoholic beverage advertising and both consumption and abuse it has been criticized by the FTC staff report (Levine et al 1985), the researcher who was the government's technical officer on the study (Mizerski 1980), and other researchers (Strickland et all 1982, pp. 659-660; Strickland 1984; Kohn et al 1984, p. 36). The criticisms of this report focus on sampling problems, the weaknesses of self-report data (particularly in the area of alcoholic beverage consumption), inadequate measures, inappropriate causal inferences, and statistical weaknesses. The FTC staff concluded: "it is impossible to make any meaningful statements about the effects of alcohol advertising on alcohol consumption based on Atkin and Block's results" (Levine et al 1985, Appendix A, p. 39). Strickland et al (1982) and Strickland (1984) are troubled by atkin and Block's use of content analysis, particularly when that use goes beyond content and attempts to infer advertiser intent.

It is particularly troublesome that in spite of Atkin and Block's clear statements of the limitations of their study (i.e. Vol. 1, p. 1) they frequently contradict these stated limitations and speak of effects in causal terms (i.e. Vol. 4, p. 2). Obviously, no study can be perfect. However, when authors go beyond the stated limitations of their research the objectivity of the work has to be questioned. Unfortunately, these questions undermine the otherwise valuable contributions of an important exploratory study.

A final study to be discussed (Strickland, 1981) found "advertising...to have meager effects on the level of consumption and these effects rarely translated into effects on alcohol problems" (p. 32). Unfortunately this study's total reliance on self report data undermines its results (particularly since its author not only neglects to mention any of the problems associated with self-report data but, instead, says that the strength of his findings are supported by the study's "methodological adequacy").

This brief review has demonstrated that researchers in the area of the effects of alcoholic beverage advertising are plagued by the methodological complexity of the area. The next section will focus on those methodological issues.

METHODOLOGICAL ISSUES

To provide data that is useful in a public policy forum presents methodological challenges more severe than those typically facing an academic researcher. Key among these is the need for maximizing external validity without destroying internal validity. Obviously there is a tradeoff between these two forms of validity. However, the policy maker require external validity to a greater extend than the theorist. At the same time the resulting reductions in internal validity can provide meaningless data. Clearly, researchers in this area need to make hard choices (as do all researchers). However, unlike other research settings, the policy arena requires that the researcher explicitly state the limitations of the study (both internal and external). Otherwise the less sophisticated policy-maker may misapply the findings and the research may be labeled adversarial.

While there are a wide range of additional methodological issues that researchers in this area must address this paper will briefly touch on three of them: causal designs, ethical issues, and self report measures. The need for causal designs is a direct result of the policy issues. Obviously, it is inappropriate to apply correlational data to these issues since reverse causality is often a valid and viable explanation. Unfortunately, this eliminated many useful research methods including much of econometric analysis. This area also raises serious ethical questions for the researcher. To what extent can researchers causally examine the advertising abuse (or even consumption) link without actually attempting to induce consumption? This problem becomes particularly acute for researchers attempting to investigate advertising effects on alcohol abusers. Clearly, in this area researchers must be particularly attuned to the effects of their research on their subjects.

A final methodological issue to be addressed is the use of self report measures. Many of the studies described above have been widely criticized for their use of self-report measures of advertising exposure, alcoholic beverage consumption, and abuse. Given the well documented tendency for that method to systematically over and/or understate responses (e.g. Clancy et.al. 1979) it is clearly inappropriate when studying such sensitive issues as alcoholic beverage consumption and abuse. Unfortunately, few alternatives exist. Thus, an early issue for researchers to address is the development of more reliable measures.

RESEARCH OPPORTUNITIES

As the review of the literature in the area of effects of alcoholic beverage advertising has demonstrated, there exists little credible research evidence regarding these effects. This is not to say that there are no effects. Rather, the current state of research in this area is so poor, it is not possible to ascertain what effect this advertising has.

As various government agencies consider a range of rules that would govern alcoholic beverage advertising it is possible that consumer research could provide valuable inputs to the proceedings. This potential research falls into two categories: general and specific.

Underlying all regulatory considerations in the area of alcoholic beverage advertising is the fact that alcoholic beverages can be (and frequently are) abused. Thus, requests for the banning of certain advertising and/or promotional practices are often based on the allegation that the practice leads to alcohol abuse. As the literature review demonstrated, there currently exists no research support for this allegation. Therefore, the general research question of whether alcoholic beverage advertising has an effect on alcohol abuse is an appropriate question for consumer researchers to address.

The question of what effects alcoholic beverage advertising has on alcohol abuse is remarkable complex and will require considerable research effort. It is not even clear that existing research methodologies are appropriate for addressing this question. Thus, an early stage in the research process would be to have researchers address the methodological question itself. One avenue for this might be an interdisciplinary research conference bringing researchers together (including researchers who had never worked in this area), updating them on the state of current research, and exposing them to the methodological problems and concerns in the area. The output of such a conference could be twofold: first, new methodological options and approaches could be considered; second, the conference could be expected to stimulate researchers to begin working in the area.

Before the question of the potential abuse related effects of alcoholic beverage advertising can be considered, a preliminary question must be addressed: whether alcoholic beverage advertising leads to alcoholic beverage consumption. Obviously, if alcoholic beverage advertising does not lead to consumption (or increase existing levels of consumption) it cannot lead to abuse. This could be an important distinction, since research on the question of consumption effects may be less formidable, methodologically, then research on abuse effects. While a finding that alcoholic beverage advertising resulted in (or could result in) increased alcoholic beverage consumption would be insufficient to conclude that such advertising led to abuse, a finding that no consumption effect existed would imply the lack of an abuse effect. Thus, the general research question that might be addressed first is what effect alcoholic beverage advertising has on consumption of alcoholic beverages. Only if that research indicated the existence of a consumption effect would research into the abuse related effects of alcoholic beverage advertising be necessary.

The general research projects described above can be expected to be long term efforts, spanning a number of years, and, because of the complexity of the research, it is likely that any results will be controversial and require replication by a number of researchers. Because of the currency of regulatory interest on issues relating to alcoholic beverage advertising, it would be useful if research were conducted that might shed light on specific issues even if some of the general questions were left unanswered. Carefully conducted copy research might provide a useful source of more readily accessible information on those narrower issues.

Copy research cannot address the broad issues of the effects of advertising on alcoholic beverage abuse and/or consumption. These issues are behavioral and copy research (with the exception of quite complex research systems) is not an appropriate source of behavioral data. However, copy research is appropriate for examining the the communications effects of an advertisement (or an advertising campaign). Further, copy research can provide a surrogate measure of behavior by measuring whether one of the communications effects of an advertisement is to change behavioral intent. Finally, copy research can provide causal data. That is, it can demonstrate that the communication effect that was observed was actually caused bs the ad.

The copy research that is suggested here can also aid in the pursuit of the general research questions described above. Without a measure of what is communicated by an advertisement it is impossible to measure the behavioral effects (i.e., consumption and/or abuse of alcoholic beverages) that might result from that advertising. Finally, the copy research may provide researchers with the opportunity to test the methodological approaches they are developing.

REFERENCES

Ackoff, R.A. and Emshoff, J.R. (1975) "Advertising Research at Anheuser-Busch, Inc. (1963-1968)," Sloan Management Review, 16, 1-15.

Ashley, R., Granger, C.W.J. and Schmalansee, R. (1980) "Advertising and Aggregate Consumption: Aa Analysis of Causality," Econometrica, 48, 1149-1167.

Atkin, C. and Block, M. (1980) Content and Effects of Alcohol Advertising, Final Report submitted to BATF, FTC, NIAAA, and U.S. DOT.

Bourgeois, J.C. and Barnes, J.G. (1979) "Does Advertising Increase Alcohol Consumption?" Journal of Advertising Re- search, 19, 19-29.

Clancy, K., Ostlund, L. and Wyner, G (1979) "False Reporting of Magazine Readership," Journal of Advertising Re- search, 223-230

Comstock, G. (1976) Television and Alcohol Consumption and Abuse, The Rand Corporation.

Defoe, J.R. and Breed, W. (1979) "The Problem of Alcohol Advertisements in College Newspapers,'t Journal of the American College Health Association, 27, 195-199.

Duffy, M. (1982) "The Effect of Advertising on the Total Consumption of Alcoholic Drinks," Journal of Advertising, 11, 105-107.

George, T. (1984) Unpublished Speech to the National Association of Beer Wholesalers, BATF.

Gordon, T. and Gordon, W.B. (1984) "Drinking and Mortality: The Framingham Study," The American Journal of Epidemiology, 120, 97-107.

Grabowski, H. (1976) "The Effects of Advertising on the Interindustry Distribution of Demand," Explorations in Economic Research, 3, 21-75.

Hagen, L.W. and Waterson, M.J. (1983) The Impact of Advertising on the United Kingdom Alcoholic Drink Market, The Advertising Association.

Jacobson, M.F., Atkins, R. and Hacker, G.A. (1983) The Booze Merchants, CSPI.

Jacobson, M.F., Hacker, G.A. and Silverglade, B.A. (1983), Petition for Regulation of Unfair and Deceptive Alcoholic Beverage Marketing and Advertising Practices, CSPI.

Kohn, P.M. and Smart, R.G. (1984) "The Impact of Television Advertising on Alcohol Consumption: An Experiment," Journal of Studies on Alcohol, 45, 295-301.

Kohn, P.M., Smart, R.G. and Ogborne, A.C. (1984) "Effects of Two Kinds of Alcohol Advertising on Subsequent Consumption," Journal of Advertising, 13, 34-48.

Levine, B.C., D'Amato, D.G., Popper, E.T., and Keenan, D (1985) Recommendations of the Staff of the Federal Trade Commission Re: Omnibue Petition for Regulation of Unfair and Deceptive Alcoholic Beverage Advertising and Marketing Practices, FTC.

Marsteller, P. and Karnchanapee, K (1980) "The Use of Women in the Advertising of Distilled Spirits, 1956-1979," Journal of Psychedelic Drugs, 12, 1-12.

McCarty, D. and Ewing, J.A. (1983) "Alcohol Consumption While Viewing Alcoholic Beverage Advertising," International Journal of Addictions, 18, 1011-1018.

McGuiness, T. (1980) "An Econometric Analysis of Total Demand for Alcoholic Beverages in the U.K., 1956-1975," Journal of Industrial Economics, 29, 85-109.

Mizerski, R. (1980) "Unpublished 'Methodological Critique of Atkins and Block (1980)" Submitted to FTC.

Rock, E.R. (1985) "Letter to M.F. Jacobson, CSPI," Official FTC Correspondence.

Strickland, D.E. (1981) "The Advertising Regulation Issue: Some Empirical Evidence Concerning Advertising Exposure and Teenage Consumption Patterns," Paper presented to Conference on Control Issues in Alcohol Abuse Prevention.

Strickland, D.E. (1984) "Contents and Effects of Alcohol Advertising: Comment on NTIS Pub. No. PB82-123142," Journal of Studies on Alcohol, 45, 87-100.

Strickland, D.E., Finn, T.A. and Lambert, M.D. (1982) "A Content Analysis of Beverage Alcohol Advertising," Journal of Studies on Alcohol, 43, 677-682.

U.S. Department of the Treasury and U.S. Department of Health and Human Services (1980) Report to the President and the Congress on Health Hazards Associated with Alcohol and Methods to Inform the General Public of these Hazards.

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