The Effects of Alcohol Advertising

Charles K. Atkin, Michigan State University
Martin Block, Michigan State University
ABSTRACT - This report summarizes the findings from a series of investigations examining the content and impact of advertising for beer, wine, and liquor. Several surveys and experiments tested the relationship between advertising exposure and brand awareness, alcohol knowledge, images of drinkers, attitudes toward drinking, consumption behavior, and heavy and hazardous drinking.
[ to cite ]:
Charles K. Atkin and Martin Block (1984) ,"The Effects of Alcohol Advertising", in NA - Advances in Consumer Research Volume 11, eds. Thomas C. Kinnear, Provo, UT : Association for Consumer Research, Pages: 688-693.

Advances in Consumer Research Volume 11, 1984      Pages 688-693


Charles K. Atkin, Michigan State University

Martin Block, Michigan State University


This report summarizes the findings from a series of investigations examining the content and impact of advertising for beer, wine, and liquor. Several surveys and experiments tested the relationship between advertising exposure and brand awareness, alcohol knowledge, images of drinkers, attitudes toward drinking, consumption behavior, and heavy and hazardous drinking.


The alcohol advertising project involved seven separate phases of research studying the content and effects of advertisements. Two studies focused on the quantitative and qualitative attributes of ads appearing in magazines, newspapers, and television. Five studies measured audience exposure and response to alcohol ads, particularly among teenagers and young adults. The research was sponsored by four federal agencies: The Bureau of Alcohol, Tobacco, and Firearms, The Federal Trade Commission, the National Institute on Alcohol Abuse and Alcoholism, and the Department of transportation. Based on the needs expressed by these sponsors, the following purposes guided the investigation:

a. Advertising statistics. Before collecting new data on advertising content effects, national statistics from industry sources were compiled. This preliminary phase of the project described alcohol advertising expenditures in the mass media, and reported the quantity of advertisements for magazines and network television.

b. Content analysis. The characteristics of alcohol advertisements presented in the three major mass media were measured along numerous content dimensions. The study focused on factors shown by social scientists to have important implications for influencing the audience; in addition, selected advertising practices relevant to policy-makers are assessed. This investigation sought to systematically describe the types of advertising content appearing in magazine vs. newspaper vs. television media and for liquor vs. beer vs. wine products.

c. Field survey. The major survey study assessed the mount of exposure to alcohol advertising and the impact f ads on the audience. The research analyzed the relationships between reported exposed and various cognitions, values, and behaviors in the field setting.

d. Self-report study. This survey was conducted to obtain individuals' own assessment of how advertising influences :heir orientations toward alcohol. The questionnaire study examined the impact of alcohol advertising on the acquisition of product information, formation of brand images, reinforcement of brand preferences, inducement to rial purchases, stimulation of consumption behavior, and initial experimentation with alcohol.

e. Advertising response study. This phase of the research studied responses of individuals to selected types of message appeals featured in specimen advertisements. The investigation measured the personal perceptions, interpretations and evaluations of advertising content, and assessed some short-term effects of the ads.

f. Experimental study. A series of experiments examined the relative impact of six types of advertising message content. Four experiments focused on attributes of characters appearing in ads such as race and age, and two other experiments examined drinking moderation disclaimers and amount of product information.

g. Diary report survey. This study measured recent concrete incidents of response with self-reports of behavior during the previous day when advertisements were encountered. The research focused on exposure to advertising messages, along with selected immediate effects.


In the content analysis, a sample of advertisements was drawn from the July 1978 through February 1979 time period in three media: 41 magazines (the 21 largest circulation magazines, and 20 major periodicals aimed at youth or minority audiences) and 12 metropolitan and campus newspapers were monitored, and periodic segments of TV prime-time and sports programming were videotaped. This yielded 309 magazine ads, 135 newspaper ads and 91 TV ads, for a total of 535 different advertisements.

The analyses dealt with three basic kinds of content: portrayal of characters (e.g., age, physical attractiveness), content presentation techniques (e.g., space occupied by bottle, use of slogan), and themes/appeals (e.g., romance theme, informational appeal). Trained coders examined both visual and verbal aspects of ads, and assigned each message into precisely defined categories along several dozen content dimensions. Content profiles were computed according to medium (TV, magazines, and newspapers) and product (liquor, wine, TV beer, and print beer).

The basic survey and experimental research involved questionnaire administration, interviewing, and testing with the same master sample of 1,227 respondents with diverse backgrounds from different regions of the country. Because the investigation was primarily concerned with responses of young people to alcohol advertising, most respondents were between the ages of 12-22.

First, all respondents filled out the questionnaire for the field survey. Then, they were shown specimen advertisements for the response study phase of the investigation. For each respondent, one of these ads involved an experimental manipulation of the message content. Finally, the drinkers aged 18 or older were given a questionnaire for the self-report study.

In the field survey, a comprehensive questionnaire was constructed to measure amount of advertising exposure and a number of orientations toward alcohol; separate versions of the instrument were devised to assess magazine-oriented liquor advertising vs. TV-oriented beer-wine advertising.

Exposure to advertising was measured in terms of stated quantity of reading magazines or watching television programs that frequently carry alcohol ads, frequency of exposure to certain brands and to specimen ads pictured in the questionnaire, and degree of attention to product categories and to specimen ads.

Orientations toward alcohol were measured in terms of cognitions (brand awareness, alcohol knowledge), values (image of drinkers, brand preferences, attitude toward drinking), and reported behavior (alcohol consumption patterns, heavy and problem drinking, drinking/driving). Demographic variables and exposure to interpersonal influences were also measured.

In the analysis, cross-tabulations were computed between reported advertising exposure and the orientation variables, to determine the extent to which higher levels of exposure are associated with stronger cognitions, positive values, and frequent drinking. Separate computations were made for younger vs. older respondents, and for beer-wine vs. liquor advertising.

For the self-report study, a questionnaire was designed to ask respondents a series of straight-forward items dealing with self-perceived effects of ads on cognitions (information gain, image formation), brand preferences, and behavior (trial purchasing, consumption stimulation); one section focused on the influence of ads during the period of initial experimentation with alcohol. Separate versions of the questionnaire dealt with beer, wine, and liquor advertising. The sample included 535 adults, primarily in the 18-92 age range; they were subset of the master sample. Analyses compared self-reported impact on males vs. females, older vs. younger respondents, high vs. low social status subgroups, and those reporting above-average vs. below-average exposure to advertising.

The advertising response study used a total of 30 magazine and TV advertisements selected to represent key types of advertising message content: social appeals, excessive consumption appeals, dubious claims, escape appeals, comparative claims, hazardous portrayals, and indecency appeals. Six subsets of ads were grouped for testing; the procedure involved showing an advertisement and then asking a specially designed series of open-end and close-end questions about the message. Personal interviewing was used to collect the data. The overall sample was divided into six subgroups of approximately 200 each.

In each of six experiments, advertising content was manipulated to produce different treatments shown to random subgroups of respondents. After seeing the experimental message, respondents filled out a questionnaire containing rating scales for both the advertisement and the product. The manipulations involved white vs. black characters, older vs. younger characters, celebrity vs. ordinary endorsers, seductively posed vs. non-seductive characters, high vs. medium vs. low information content, and strong vs. mild vs. no moderation disclaimer. Multiple versions of each manipulation were constructed to minimize idiosyncratic responses to particular brands or models appearing in ads. Approximately 900 persons served in each experiment. The analyses compared rating scores from subgroups receiving the different experimental treatments.

Finally, the diary report survey involved questionnaire administration to a separate sample of 178 college students, who were asked to describe their responses to TV beer commercials and magazine liquor ads the day before the survey. The questions dealt with media exposure, contact with ads, attention to ads in terms of numbers and brands, and reported effects experienced immediately after seeing ads. The analyses simply tabulated the proportions of the sample reporting each behavior, for both beer and liquor.


The reporting of key results will be organized according to the type of variable studied regardless of methodology. The presentation begins with a description of the advertising messages distributed in the mass media, along with an assessment of the extent to which respondents report being exposed to the advertisements. The next sections describe the findings for various levels of response to message exposure, moving in sequence through the hierarchy effects from cognitive/informational variables (awareness and knowledge), to affective/evaluative variables (images, attitudes and preferences), to reported overt behavioral variables (drinking). The final section assembles certain findings into substantive categories reflecting policy-relevant issues.

Alcohol Advertising Distribution and Exposure

There is a substantial quantity of television beer commercials and magazine liquor ads readily accessible to mass media audiences, including youthful segments who frequently watch and read those vehicles carrying the most advertising. Surveys show that individuals notice large numbers of these ads, and pay close attention to between one-fourth and three-fifths of the messages encountered. Ads on television attract greater attention than magazine ads, possibly because of the more intrusive nature and higher complexity of the messages. Exposure and attention is higher for adolescents than older adults. Thus, there is ample opportunity for alcohol advertising to influence the cognitions, values and behavior of the public, especially young people.

Brand Awareness and Alcohol Knowledge

Since brand advertising seeks to convey a distinctive brand identity through use of unique themes and memorable symbols, these messages may exert a strong influence on brand awareness. Aside from brand distinctions, the generic product information in ads may contribute to basic learning about the subject of alcohol and drinking, although the impact may be limited because many ads lack substantive factual material.

The content analysis indicates a strong emphasis on brand imagery. A graphic or brand-name logo appears in almost half of the ads, and slogans are used in over half the ads. The name of the brand is mentioned about five times per ad. Furthermore, one of the most widespread persuasive appeals centers on brand uniqueness; this is emphasized in more than one-fifth of the ads. By contrast, basic product information is seldom featured. Aside from description of the product's geographical origin in magazine liquor ads, very limited mention is made of inherent qualities such as processing length, calorie level, ingredients, or price. Most liquor ads do cite the alcohol proof level, typically in small print.

In the field survey, there is a positive association between advertising exposure and brand awareness. Respondents highly exposed to alcohol ads are much more able to recall brand names, identify slogans and symbols, and know key claims featured in advertisements.

Regarding more basic generic learning about alcohol and drinking, the field survey shows a moderately positive association with exposure. For example, respondents reporting a high exposure to advertising are more likely to know about liquor proof, beer ingredients, and the appropriate drinks to have with a good meal.

There are also relevant findings from the self-report study. More than one-half of the sample state that they acquire substantive knowledge about product attributes from ads, in terms of learning about ingredients, production, beer calories, liquor proof, and new mixed drinks; about one-fourth report learning about taste and cost factors.

In addition, the self-report study shows that more than half gain brand awareness from ads, in terms of forming images about which brands are most popular, which are high class, and which will impress people. Another set of questions retrospectively probed learning from ads during the initial period of adolescent experimentation with alcohol: about two-fifths say ads helped them find out which brands would impress others, and which brands famous people drink; about one-third report they discovered how to make mixed drinks, and learned which brands are best; one-fourth recall that they used ads to ascertain cheap prices; one-sixth remember learning information about the taste of different drinks.

Images of Drinkers

Because advertisements project distinctive portrayals of the demographic, social, and psychological characteristics of drinkers, the audience may develop corresponding mental images of the type of person who drinks alcohol. In particular, people may perceive that drinkers possess positive attributes after seeing flattering depictions in ads.

The content analysis research shows that advertisements present a number of favorable portrayals of drinkers. Characters in ads tend to be above-average in physical attractiveness and social status (except in the case of TV beer commercials, which feature a broader range of characters). Most characters fall in the relatively youthful 25-35 age category, although few are at or below the 91-year old level. In addition, substantial numbers of ads depict drinkers as masculine, sociable, romantic, elegant, feminine, adventurous, and relaxed.

Drinker images were measured along several dimensions in the field survey. Respondents who report high exposure are somewhat more likely to have favorable perceptions of

Whiskey and beer drinkers. Those reporting heavy exposure to liquor ads tend to perceive whiskey drinkers as more friendly, relaxed, fun-loving, happy, manly, successful, sophisticated, and good-looking; those reporting heavy exposure to beer advertising tend to perceive beer drinkers as more adult, fun-loving, young, friendly, and happy. There is also a slight tendency for reported advertising exposure to be associated with perceived pervasiveness of drinking in society; those in the heavy exposure category estimate that the typical person consumes about two more drinks per week than do the lightly exposed respondents.

In the advertising response study, ads projecting romance, affiliation, and psychosexual appeals were tested. After seeing an ad depicting a romantic couple, about one-sixth say that the typical drinker of the advertised product is romantic; after seeing an ad portraying affiliation among friends, few consider the typical drinker of that brand as being sociable. When exposed to an ad showing a sexy female or lustful lovers, about one-tenth get the impression that the users of the product are sexy/swinging persons.

Finally, one of the experiments manipulated the race of characters in parallel pairs of ads. White respondents exposed to black models are significantly more likely to consider the typical drinker of the advertised product to be black; those exposed to white models tend to perceive product drinkers as white, to a slight degree. In addition, black characters generate somewhat more positive ratings of both the advertising and the product.

Attitudes toward Drinking

By depicting alcohol consumption as attractive, acceptable, and rewarding, advertising may create and reinforce favorable attitudes toward drinking. Specifically, exposed individuals may form positive values regarding the amount, situations, and benefits of alcohol and drinking.

The content analysis demonstrates that many facets of advertisements present favorable portrayals of alcohol. Among promised or implied benefits promoted in ads are social camaraderie, escape, refreshment, relaxation, social approval, romance, and elegance.

The field survey contained 50 items that represented overall attitude toward drinking, including measures of agreement with statements about alcohol (e.g., "It's OK for a teenager to get drunk every once in awhile", "alcohol helps people relax and unwind"), the range of situations considered appropriate for drinking (e.g., "during lunch", "after work"), and propriety of drinking in various amounts, at various ages. This attitudinal variable is positively associated with advertising exposure to a moderate degree.

The response study examined two major appeals relevant to alcohol attitudes: psychological escape, and social interaction. After seeing an escape ad, an average of three-fifths of the respondents say they derived the feeling that the product will "help you get away from your ordinary situation". Those seeing an ad with a social theme involving affiliation among friends are slightly more likely than a control group to say that drinking will make a social gathering more enjoyable; there is also a weak tendency for respondents seeing a social romance theme to say that alcohol helps make an evening more romantic.

Finally, one question in the self-report study indicates a potentially important impact on attitude. Adult respondents were directed to think back to the time they began experimenting with alcohol; when asked whether ads helped make them feel that drinking was normative, three-fifths say "yes" or "maybe."

Brand Preferences

One major purpose of advertising is to create and reinforce preferences for the advertised brand relative to competing brands in the product category. More positive evaluations may result after repeated exposure to the brand name, attractive symbols and rewarding benefits associated with the brand, and claims of brand quality. Preference is considered primarily as an attitudinal feeling of favorability; actual consumption of various brands will be described in the section on drinking behavior.

In the field survey, brand preference was measured with evaluation rating scales for youth and listing of favorite brand names for adults. Examining data for beer, wine, and liquor brand preferences, there is a positive relationship with advertising exposure.

Three questions in the self-report study dealt with the strengthening of prior dispositions toward favored brands. More than one-third say they occasionally develop stronger liking, become more favorable, or feel more certain regarding their favorite brands from seeing ads.

The response study measured self-perceived changes in respondent preferences for 15 different specimen ads. Most indicate that their brand liking was unchanged after seeing the ad. An average of 16% say they became more favorable toward the brand, and :0% say they became more negative. Ads using psychological escape appeals or social appeals generate the most positive change, while psychosexual ads produce unfavorable change.

The response study also examined the impact of comparative advertising messages for Eagle Rare (compared to Wild Turkey) and Cutty 12 vs. Chivas Regal. Among non-exposed control respondents, an average of just 18% say they prefer Eagle Rare or Cutty 12 over the more popular comparison brands. Those exposed to advertising are much more favorable; an average of 44% indicate a preference for the advertised brands.

Finally, there are brand preference data from the six experiments. In each experiment, respondents gave evaluations of the product along an 11-step scale. Averaging the three evaluation scales with the likelihood scale, the results show that certain techniques are more effective. Scores Are about a half-point higher for ads using a moderation disclaimer rather than no disclaimer, for ads depicting younger vs. older characters, and for ads featuring celebrities vs. ordinary endorsers. One-fifth-point differences occur for ads containing high rather than medium or low information, ads portraying black vs. white characters, and ads using sexy vs. non-sexy characters.

Youth Drinking

There are several reasons why alcohol advertising might bs expected to produce drinking by youth: ads may reduce inhibitions that restrict the consumption of alcohol, by showing that this activity is socially acceptable and normative in society; ads may persuade non-drinkers or occasional drinkers to consume more alcohol, by portraying rewarding consequences such as romance/sociability, masculinity/femininity, and escape; famous or attractive characters in ads may influence impressionable young people to model their behavior; and advertising may stimulate regular consumers to acquire and drink more alcohol, by a simple reminder to act. On the other hand, advertising impact may be outweighed by peer influences, or minimized by countervailing social persuasion or personal resistance to advertising appeals.

The major data base is the field survey, which measured both reported advertising exposure and drinking among 12-18 year-old junior and senior high school students. A liquor consumption index combining specific brand consumption and overall weekly drinking rates is closely related to liquor advertising exposure. For 11 brands of liquor listed in the questionnaire, an average of 31% of the high exposure group vs. 15% of the low exposure group say they have tried each one. On the item measuring weekly quantity, 9% of the heavily exposed vs. 3% of the lightly exposed youth report drinking five or more liquor drink in a typical week; at least one drink per week is consumed by 45% vs. 2770.

Beer drinking is also associated with beer-wine advertising exposure, while the relationship for wine drinking is not significant. For six brands of beer listed on the questionnaire, an average of 52% of the high exposure youth vs. 37% of the low exposure youth indicate they have tried each one; the difference for the three wines on the list is much smaller. Regarding consumption in a typical week, 16% of the heavily exposed vs. 10% of the lightly exposed say they drink five or more beers per week; 46% vs. 29% have at least one beer. For wine, at least one glass per week is consumed by 21% of the heavily exposed vs. 17% of those less exposed to advertising.

For those adolescents who had not yet begun drinking, a question asked about the probability of starting in the future. For all three types of alcohol, the heavily exposed non-drinkers are more likely to expect future consumption by a substantial margin over lightly exposed nondrinkers. For example, 20% vs. 10% say they "definitely" or "probably" would drink liquor, and an additional 3970 vs. 76% indicate "maybe "

Another approach assessed the role of advertising in youthful alcohol experimentation by asking adult respondents for retrospective recall. Very few respondents identify advertising as a major influence on their initial decision to start drinking; about one-third think it was a minor influence. Most claim that peers had the strongest impact on their experimentation.

Adult Drinking

Evidence on advertising influence was gathered in all three survey studies, with the most extensive data coming from the field survey. This investigation shows that the index of liquor drinking is related to liquor advertising exposure. Reports of both recent drinking and typical patterns of drinking were measured. For each of 10 categories of liquor (e.g., scotch, rum, vodka) the high exposure group reports more drinking than the low exposure group; during the previous month, the high group consumed a total of 16 drinks while the low group had 8 drinks. In addition, 78% of the highly exposed say they typically consume at least one mixed drink per week and 50% report having at least one straight drink per week. This compares to 51% reporting mixed drinks and 18% reporting straight drinks in the low exposure group. The typical weekly total shows a difference of 4.0 vs. 2.1 straight and mixed drinks.

Exposure to beer-wine advertising is moderately related to beer consumption and weakly related to wine consumption. Adding together all brands of beer, the high exposure group says they drank an average of 30 beers over the previous month, compared to 15 beers for the low exposure group. The question about the number of beers per typical week shows that 32% of the highly exposed respondents consume five or more; this compares to 19% for the less exposed respondents. In terms of total drinks per typical week, there is a difference of 5.7 vs. 3.2 beers.

The self-report study dealt with subjectively perceived impact of advertising on drinking. About one-seventh of the sample report at least monthly advertising stimulation to go to the refrigerator for a drink, to go out and buy alcohol, to buy alcohol advertised at discount prices, and to get a drink of preferred alcohol after seeing an ad for another brand. Another one-fifth say these effects happen at least a couple of times per year. Respondents estimate that they would consume about two fewer drinks per month if there was no advertising for alcohol.

Another set of self-report questions focused on trial purchases. When asked whether advertising influences decisions to try specific new brands of alcohol, one-fifth reply affirmatively. In addition, one-fourth frequently find out about new brands and one-tenth are stimulated to try new brands on a frequent basis; more than half say these two effects occur occasionally. For almost one-third of the respondents, advertising stimulates trials at least a couple of times per year. Another indication of this effect is the finding that consumers estimate they would have tried three or four fewer brands of alcohol if there were no ads promoting these products.

In the diary report survey, college students were asked if ads seen the previous day influenced their drinking behavior. Among those exposed to an ad, 5% say the ad stimulated a purchase of beer and 3% a purchase of liquor. Furthermore, 11% indicate that they decided to have a drink of beer and 1% a drink of liquor after seeing an ad that day.

There is little doubt that alcohol advertising exerts an influence on the frequency and quantity of adult alcohol consumption. A quantitative estimate of the contribution of alcohol advertising is difficult to calculate, but it is likely that ads account for a 10% to 30' increase in the total amount of alcohol that would be consumed without advertising.

Heavy or Excessive Drinking

Concern has been expressed that alcohol advertising encourages drinkers to consume beyond normal levels of moderation. Some ads overtly or subtly advocate excessive consumption of the advertised product, and the cumulative impact conveyed by ordinary ads may be that heavy drinking is appropriate and rewarding. Thus, ads may prompt excessive drinking and contribute to alcohol problems.

In the field survey, the alcohol advertising exposure index is positively related to reported heavy drinking. When asked how many drinks they consume in an evening at a bar or party, the highly exposed respondents report an average of 4. 5, compared to 2.9 for lightly exposed respondents. In response to a question concerning the frequency of having at least five or six drinks in a single day or night, 33% of the high exposure group vs. 16% of the light exposure group say this happens at least once a week.

Two other survey items dealt with problem drinking. When asked if they are concerned about their drinking, 18% vs. 11% indicate that they are worried about drinking too much; in addition, 8% vs. 3% report having gotten in trouble at school or on the job because of their drinking. A pair of the alcohol attitude items (above) pertains to heavy drinking; 45% of the highly exposed vs. 2975 of the lightly exposed agree that it's OK for an adult to get drunk every once in awhile, and 32% vs. 20% approve of teenagers getting drunk.

The response study tested five advertisements that appeared to promote heavy drinking. Respondents were asked what message the advertiser is trying to get across; reference to drinking/serving a lot of alcohol are made by one-fourth of the sample. When asked if the sponsoring company wanted a person to drink a large, medium, or small amount of alcohol, three-fourths say a large amount; however, unexposed control respondents give similar high estimates. Respondents also gave their own opinion about excessive drinking. After exposure to an ad, one-third say it is acceptable for a person to occasionally drink heavily rather than moderately, and one-fourth feel it is appropriate for a person to consume six or more drinks in an evening; again, these responses do not differ from the reactions in unexposed control group.

Drinking and Hazardous Activities

A number of activities requiring physical coordination and clear-headed judgement, such as driving a vehicle, are more dangerous to perform during or after alcohol consumption. Alcohol advertising may contribute to the conduct of hazardous activities in two ways: ads that specifically depict potentially dangerous behaviors may influence the audience attitudes, such that the activities are seen as safer and more acceptable to perform while drinking; and advertising in general may influence the basic frequency and quantity of alcohol consumption, such that there is a greater likelihood that the audience will be drinking before or during any activities, including those that are hazardous.

The content analysis shows that a small proportion of ads portray alcohol in the context of hazardous activity. The proportions are 5% for TV ads and 67: for magazine ads; the most frequent cases are pictorial portrayals of water sports. Driving is almost never associated with alcohol in advertising.

One portion of the field survey focused on the drinking and driving topic. About half of all post-high school age respondents report that they drive within an hour after drinking at least once a month; one-third of the 11-19th graders also report monthly driving after drinking. Slightly higher proportions drink while riding in another person's car. In addition, two-fifths of the high school youth say they drink in a parked car at least once per month. When asked about driving while drunk, about one-tenth of the driving-age respondents report a recent incident. Most individuals feel that they can consume two-to-three drinks without affecting their ability to drive safely. The vast majority say that driving after a few drinks is an unsafe activity.

These drinking/driving items are positively related to alcohol advertising exposure; those highly exposed to ads are more likely to drive after drinking, drink while riding, drink while sitting in a parked car, drive while intoxicated, and to feel that they can consume more alcohol without affecting their ability to drive. There is no difference on the question of whether drinking while driving is unsafe.

Aside from these cognitive and behavioral items, one question in the attitudinal portion of the survey dealt with the maximum number of drinks that are proper to consume right before driving; the average number is 3.0 for the highly exposed respondents vs. 2.4 for those lightly exposed.

In the field survey, one set of items measured opinions about the safety of performing six other activities after having a few drinks (swimming, playing golf, raft riding, mountain climbing, skiing, and boat driving). On the average, 68% of the lightly exposed respondents feel these behaviors are not safe, compared to 65% of those highly exposed to advertising. This indicates that advertising has a negligible impact in making people feel that it is safe to pursue possible hazardous activities after drinking.

Responses to Selected Advertising Appeals

This final section examines the reactions of respondents to several controversial and policy-relevant appeals used in alcohol advertising: endorsements by celebrities, psychosexual themes, and youthful character portrayals.

Regarding celebrity endorsements, the content analysis shows 10% of TV ads and 3% of magazine ads use a famous individual, typically a former athlete. This appeal was tested in the experimental study, where three nearly identical pairs of ads portrayed either a celebrity (actor Telly Savalas, model Cheryl Teigs, and basketball player Happy Hariston) or a non-celebrity as an endorser. Although there is no difference in believability, ads featuring a celebrity are rated as significantly more "effective," "important," "strong," and "interesting;" ratings are slightly higher on most other scales as well. In the celebrity condition, the product itself is rated higher on all scales; respondents say they are somewhat more likely to get the product, too. Furthermore, the celebrity character is rated as significantly more trustworthy and competent. These effects are very strong among younger respondents under the age of 18; the celebrity endorsement has little impact on older persons.

One finding from the retrospective portion of the self-report study is also pertinent. Two-fifths of the respondents say that they used ads to find out which brands famous people drink, during the period when they initially experimented with alcohol. Finally, field survey data show that half of the highly exposed respondents can identify which whiskey Telly Savalas likes and which wine Orson Welles likes.

Psychosexual themes appear in various forms in many alcohol advertisements. Here are some relevant findings from the content analysis: 45% of IV ads and 10% of magazine ads project a masculine image while 30% and 9% project a feminine image; romance is a theme in 19% and 16% of the ads, and hedonistic pleasure or sexual disinhibition are featured in 5% and 10% of the ads. In addition, 5% of TV ads and 8% of magazine ads contain indecent language or pictures.

The main set of findings on this topic come from the response study, which focused on four specimen ads with psychosexual themes of a potentially indecent nature: two portray provocatively posed females in swimming attire, and two depict young lovers and slogans with sexual overtones ("sip into something comfortable" and "Giocobazzi someone tonight"). When asked "what went through your mind while looking at the ad," an average of one-fourth mention sexual thoughts. For the two ads showing the female form, about two-thirds reply affirmatively when asked whether they personally consider the ad to be sexy. Almost nine-tenths say they perceive sexual connotations for each of the suggestive slogans. Less than one-tenth report being offended by the slogans, while about one-fifth feel the pictures of females are offensive. Finally, less than half say they like these ads, which is much lower than for other types of appeal.

One of the experimental studies dealt with the sexual suggestiveness issue. Sexy and non-sexy versions were created for three ads, with the same characters (a male, a female, and a male-female couple) posed either suggestively or more normally. The findings show that respondents give consistently higher ratings to ads featuring suggestive poses. The product advertised by the sexy characters also attains higher rating, although there is little difference in intention to get it. The characters are evaluated less positively in the suggestive condition Typical drinkers of the sexily advertised product are seen as somewhat more sexy people. There is a sharp difference in response according to age level: those under 18 years old react more favorably to the sexual theme, while older respondents give slightly more positive ratings in the non-sexy condition.

Findings from the field survey indicate that sexy advertisements affect audience images of typical drinkers of alcohol. Among those heavily exposed to liquor ads, 15% perceive whiskey drinkers as "sexy," compared to 10% among those lightly exposed. Similarly, respondents heavily exposed to beer advertising are more likely to hold the perception that beer drinkers are sexy, by a 16% to 11%o margin. Finally there is a slight tendency for those highly exposed to alcohol ads to hold the attitude that "drinking alcohol improves one's chances of sexual success;" 1470 agree with that statement, compared with 10% of those less frequently exposed to ads.

Regarding youthful character portrayals, the content analysis indicates that only 1% of TV and magazine ads are targeted to the under-21 age group. While no ad characters appear to be in the 10-19 age group, 31% of TV characters and 49% of magazine characters fall in the 20-29 age segment.

In the response study, respondents were asked to judge the age of 12 characters presented in the specimen ads. An average of 4% estimate the age level to be under 21; young respondents below the age of 18 are twice as likely as older respondents to perceive the characters as under 21

One of the experiments tested the impact of ads featuring either young characters (approximately age 20 or 21) or older characters (ranging from 38 to approximately 50). The young advertising portrayals have a stronger impact than do the same messages depicting older models; this effect primarily occurs for respondents under 18 years old. These young peopLe give higher ratings to the ads with young characters, and feel more favorable toward the product as well; in particular, they are significantly more likely to say they will get the product. Character age makes little difference to adult respondents.

In the field survey, persons heavily exposed to alcohol advertising are much more likely to hold the image that the typical drinker is "young"; 42% in the high exposure group vs. 28% in the low exposure group feel that beer drinkers are young, and the image of whiskey drinkers shows a 21% vs. 15% difference.

The evidence on this diverse set of issues indicates that there is some cause for concern about responses to certain types of advertising appeals. Although relatively few ads are judged to contain celebrities, underage characters, or explicit psychosexual themes, these appeals may still produce occasional cases of socially significant influence. Compared to adults, adolescents seem to be especially vulnerable to ads that use celebrity endorsers, sexy models, and youthful looking characters.