This research was supported by grants from the John and Mary R. Markle Foundation and the University of California, Los Angeles.

Commercial advertising has always been a central feature of American culture. As encountered in the mass media, it is pervasive and inescapable. Most Americans take for granted the "rules" of commercial advertising, even though they may not be aware that any formal guidelines exist and may have little or no idea what the legal effect of such guidelines might be. Commercial advertisements are widely accepted as fair and legitimate marketing.

Contrast the world of political advertising. In recent years, political advertising has become essential to campaign strategy (at least in major campaigns), and many regard it as far more intrusive than routine commercial advertising. But the world of political advertising is very different from the world of commercial advertising. There really are no "rules" when it comes to the content and form of political advertising. Political advertisers are not accountable to any regulatory body, voluntary or otherwise, for the accuracy of their claims. They readily engage in so-called "comparative" advertising. They blatantly criticize their competitors. They complain incessantly about the fairness of the comments made about them, while their opponents are doing the same. There is no acknowledged forum for the review of these claims and counter-claims. The press attempts to provide some sporadic checks on political advertisers by running "ad-watch" reports, but these reports by their very nature tend to fuel public cynicism. Considerable evidence suggests that the negativity associated with contemporary political campaigns has created an "avoidance" mentality which is serving to shrink the electorate and the level of political participation generally (see Ansolabehere and Iyengar, 1995).

The current state of political advertising has aroused considerable concern within the world of commercial advertising. Major advertising firms and professional associations have widely deplored the lack of accountability of political advertisers and their unwillingness to adhere to a code of ethics (see Advertising Age, April 29, 1996; New York Times, April 29, 1996; Washington Post, July 30, 1996). What exactly is Madison Avenue concerned about? Perhaps commercial advertisers fear that the apathy -- and all too frequently, aversion -- induced by political advertising campaigns may damage the credibility, and ultimately the persuasiveness, of more traditional forms of advertising. As Alex Kroll, former chairman of the American Association of Advertising Agencies, put it: "We must stop politicians from ruining our reputation." (Advertising Age, April 29, 1991) Kroll’s was not a solitary voice. In 1984, then AAAA chair John O’Toole claimed that political ads were "giving advertising a bad name." (Advertising Age, June 24, 1996) and in 1996, Burt Manning went so far as to assert that the "smear and scare" tactics of political advertisers meant that "today, the issue is survival of brand advertising" (Advertising Age, June 24, 1996). Our goal in this paper is to provide some evidence on the issue of whether political advertising, does, in fact, "contaminate" commercial advertising.

We set out to consider two rival possibilities, both of which rest on the assumption that commercial advertising is evaluated more favorably than political advertising. The assimilation hypothesis, derived from social judgment theory, suggests that exposure to political advertising campaigns encourages people to "assimilate" or equate their feelings about related attitude targets (for a discussion of social judgment and other theories of attitude change, see Petty and Cacioppo, 1986). The essence of this concept is that negative reactions to political advertising will color attitudes toward other forms of advertising. The competing possibility, which we have termed the "contrast" hypothesis, suggests that the negative response to political campaigns actually makes commercial advertising appear more appealing than it would have been in the absence of political advertising. By accentuating the negative attributes of political advertisements, political campaigns strengthen the standing of commercial advertisers.

In the sections that follow, we will first provide some background on the scope and extent of commercial and political advertising and the regulatory environment in which advertisers operate. Next, we describe recent scholarly research into the content and effects of political advertising. We then describe our experimental methodology for assessing the impact of political advertising on receptiveness to commercial advertising and summarize the findings. Finally, we consider the implications of our evidence for the current debate.

Even though the use of political advertising has spread exponentially, both in terms of the sheer frequency of exposure and the increased length of political campaigns, political advertising is still miniscule compared with commercial advertising. The total cost of the 1996 election (all races combined) amounted to approximately $2.5 billion (Center for Responsive Politics, 1999). This figure is less than the annual advertising budget for major U.S. corporations. During the height of the 1996 campaign, the research firm CMR found that fewer than one percent of all televised advertisements (750,000 out of 93,000,000) in the top 75 media markets were sponsored by political candidates or organizations (Goldstein, 1998). Clearly, the public’s distaste for these advertisements is based on factors other than sheer frequency.

The most distinctive feature of contemporary political campaign advertisements is the negativity of their content and tone. Political advertisers frequently engage in so-called "comparative" advertising in which the opposing candidate’s program and performance are criticized and even ridiculed. Highlighting the opponent’s liabilities and weaknesses usually takes precedence over identifying the sponsor’s program and strengths. In the most comprehensive tracking of campaign advertising to date, scholars at the Annenberg School of Communication have found that such "negative" advertising makes up approximately one-third of all campaign ads used in presidential campaigns (Jamieson et al., 1998). The level of negativity is actually significantly greater when one considers frequency-weighted indicators of content (Prior, 1999). In 1996, for instance, while fewer than one-half of the ads produced by the major candidates featured negative appeals, these appeals accounted for some seventy percent of the candidates’ ad buys (Goldstein, 1998). While we do not have comparable data for any commercial advertising campaign, the "comparative" element is unlikely to be so prominent; when compared with commercial ads, political ads are much more negative in content.

Unlike commercial advertisers, political advertises do not adhere to any codes or procedures intended to protect the public from inaccurate and unsubstantiated claims. All commercial advertisers voluntarily subscribe to a "code of advertising ethics" administered by the Advertising Division of the Better Business Bureau. This code includes provisions for dealing with complaints of false or misleading claims. Complaints directed at specific ads are reviewed and arbitrated by a panel appointed by the National Advertising Review Board. After reviewing the evidence from both sides, the panel may find the complaint to be valid and require that the ad in question be modified or discontinued. The panel may also refer the complaint to the appropriate governmental agency. If the advertiser fails to comply with a request for modification or termination, the panel may issue a "notice of noncompliance" identifying the advertiser.

Political advertisers are not subject to comparable voluntary guidelines. First Amendment protections make it virtually impossible to impose involuntary restraints on the content of political advertising. The American Association of Political Consultants has shown no inclination to encourage any form of self-restraint. The result is a free-for-all environment in which candidates repeatedly attack and counter-attack the claims of their competitors. The only accountability is provided by the press, in the form of sporadic "ad-watch" news reports that scrutinize specific ads for their accuracy (for a review of research into the effects of these reports, see Pew Commission, 1998). The very nature of ad-watch journalism, however, is bound to exacerbate public cynicism over the fairness and credibility of political advertising.

The harsh tone of political advertising, the often controversial techniques employed by political advertisers, and the fact that the competing claims made in campaign ads are beyond review, have raised questions about the goals of political advertisers. Many critics have suggested that political advertisers seek votes at any cost, even including a degraded sense of public regard for the candidates and the electoral process. Perhaps the amount of negativity featured in political campaigns is designed to shrink the "market" rather than increase the sponsor’s relative share. Discouraging people from voting is much more feasible than persuading supporters of one candidate to vote for the opponent. It is well known that most Americans hold fast to their partisan attachments and that the act of voting generally serves expressive (as opposed to instrumental) needs (for a review of research on political participation, see Rosenstone and Hansen, 1992). Since people acquire their affiliation with the Democratic or Republican parties early in life, the probability that they will cross party lines in response to an advertising campaign is slight. And since the motivation to vote is typically symbolic or psychological (in the sense that one’s vote is unlikely to be pivotal in determining the outcome of the election), increasing the level of controversy and conflict in ad campaigns is bound to discourage voters from making a choice and casting a vote. In effect, negative campaigns create an "avoidance" set within the electorate (see Houston et al., 1998, 1999).

Although the scholarly evidence is mixed, experimental studies substantiate these claims. Carefully controlled manipulations of advertising tone demonstrate that exposure to negative (rather than positive) campaign advertising heightens political cynicism and diminishes voter turnout (see Ansolabehere and Iyengar, 1995; Houston et al., 1998, 1999; Ansolabehere et al., 1999). It is hardly coincidental that the public’s views of elections and the importance of voting have soured as political advertising campaigns have become increasingly reliant on negative appeals. In 1960, for example, only one in four Americans endorsed the statement that "public officials don’t care much about what people like me think." By 1990, the cynical response was given by six of ten Americans (see Rosenstone and Hansen, 1992).

Exposure to political campaigns has extracted a similar toll on the public’s views of political advertising. There is ample survey data showing that the public dislikes media-based political campaigns. According to the most recent surveys by the Pew Center, a majority of the electorate (some 60 percent) felt that campaign commercials were not useful in helping them choose a candidate during the 1998 elections and more than two-thirds (68%) judged the campaign as "nasty" (Pew Center, 1998). And in a recent survey of voters in Virginia, some three-fourths of the sample indicated that negative campaigns were likely to discourage people from voting (Freedman, 1999).

Does the fallout from exposure to political advertising spread to commercial advertising in general? We attempt to answer this question experimentally, by manipulating exposure to political advertising and then measuring participants’ attitudes towards political and commercial ad campaigns. We also manipulate the tone of political advertising in order to assess the impact of negative political campaigns on the audience’s confidence in political and product advertisers. Our results indicate that exposure to political advertising in general -- and negative political advertising in particular -- strengthens viewers’ relative confidence in commercial advertising. People do not assimilate their generally unfavorable ratings of political ads to the commercial advertising arena. Nor do they express more favorable attitudes toward commercial advertising in the aftermath of exposure to political advertising. However, because campaigns heighten distaste for political advertising, the net effect is to boost the relative appeal of commercial advertising. Thus, exposure to political campaigns enlarges the contrast between commercial and political advertising.

We rely on experimental methods. The advantages and disadvantages of experimentation are well known. Unlike surveys, experiments provide accurate measures of exposure and hence yield precise causal inferences about the effects of campaign advertising. Of course, experiments have their own liabilities. Most are administered upon "captive" populations -- college students who must serve as guinea pigs in order to gain course credit. A further weakness of the typical experiment is the somewhat sterile, laboratory-like environment which bears little resemblance to the cacophony and confusion of election campaigns.

We enhanced the realism and generalizability of our experiments in several ways. Each was administered during a political campaign (1996) and featured real candidates (presidential candidates Bob Dole and Bill Clinton) or statewide ballot propositions (Proposition 209 or Proposition 211) as the sponsors.

The experiments were administered at three separate sites in the Greater Los Angeles area. One was located in predominantly Democratic West Los Angeles, in a popular shopping mall. The second was based in a small shopping area in Moorpark, a conservative northern suburb of Los Angeles. The third site was located in Manhattan Beach, a coastal city southwest of Los Angeles. This variety of locations helped to ensure a large and diverse subject pool. Subjects, who were recruited by the use of flyers, announcements in newsletters, and by personal contact in shopping malls offering payment of $15 for participation in "media research," were reasonably representative of the Southern California voting-age population.

The experimental "laboratory" consisted of a two-room office suite located in or near a retail shopping area. One of the rooms was used as a viewing room and the other for filling out questionnaires. The viewing rooms were configured to resemble, as closely as possible, the normal conditions in which a person views political advertisements. Comfortable couches and chairs were arranged in front of a television set, with houseplants and wall hangings placed around the room. Respondents were offered coffee, cookies, and soft drinks to enjoy during the viewing sessions. In most cases, family members or friends took part in the experiment at the same time, so that respondents did not find themselves sitting next to a stranger while viewing the political advertisements.

The ads corresponded to those being aired in southern California by Bill Clinton and Bob Dole. They covered a variety of issues including illegal immigration, drug abuse, federal spending on social programs, and cutting taxes. In other conditions, we substituted ads for and against Proposition 209 (the "California Civil Rights Initiative") and Proposition 211 (the "Securities Fraud - Lawsuits Initiative" ).

On their arrival, subjects read the instruction sheet which described the research (in general terms) as a study of consumer reactions to advertising. They then completed a brief pretest questionnaire which covered information concerning their personal background, media habits, and recent purchasing behavior. They were then shown into the viewing room where they watched a videotape containing nine advertisements. In the control conditions all nine ads were sponsored by non-political organizations. In the "treatment" conditions, subjects watched either one or two political ads. When the tape included one political spot, it appeared in the fifth position. In the case of the two-ad conditions, they were placed in the third and fifth slots. Following the presentation, subjects completed an extensive posttest questionnaire which included a series of questions about product and political advertising. Finally, they were debriefed and paid.

Our first comparison set was non-specific and consisted of attitudes towards political ads and Miller Beer ads. Subjects were asked to rate "political ads in general, not ads from a particular candidate or organization." Next, they were asked to rate "the ads you’ve seen for Miller Beer, not any one Miller ad in particular." (The videotape did not contain any beer ad.) They rated both types of ads in terms of a series of attribute terms: unfair, informative, nasty, misleading, funny, colorful, and persuasive. For each attribute, subjects indicated the degree (using a four-point scale that ranged from "very well" to "not well at all") to which the attribute described Miller Beer ads and political ads. We combined the unfair and misleading items into an index of veracity, the colorful, nasty, and funny items into an index of attractiveness and, finally, the informative and persuasive items were averaged to form an index of information value. Each index score was standardized to range from 0 to 1 indicating the degree to which subjects rated political and Miller Beer ads positively.

Our second indicator was more focused and targeted specific ads included on the videotape. Subjects were asked to think back to the "American Airlines," "Advil," and the "Dole/Clinton/Proposition" ads presented on the tape and indicate whether each of these ads elicited the following six feelings: angry, hopeful, sad, happy, proud, disgust. We computed the average number of negative feelings and subtracted it from the average number of positive feelings to obtain a measure of net affect.

We begin by comparing general evaluations of beer and political advertisements. Figure 1 shows the average ratings for each type of ad campaign in the control and treatment conditions. Political ads were rated much more harshly on the attributes of veracity and style. While Miller Beer ads were placed on the positive side of the scale for both veracity and style, political ads were rated toward the bottom of the scale. Although neither type of ad campaign was seen as a source of useful information, surprisingly, the beer ads were considered more informative. However, this difference was significant only among the subjects who watched a political ad.

(Figure 1 here)

The substantial differentials in the ratings of beer and political ads were unaffected by exposure to political advertising. None of the six between group differences was statistically significant. Our initial evidence thus suggests that exposure to political ads is inconsequential, at least with regard to attitudes towards a well-known genre of product advertising. Miller Beer ads were evaluated more favorably than political ads, with or without exposure to the latter.

A similar pattern of results was obtained for the manipulation of advertising tone. Figure 2 provides the mean ratings of beer and political ads in the positive and negative advertising conditions. The stark differences between ratings of beer and political ads were unchanged by the tone of the treatment ad. However, we begin to see faint traces of a contrast effect in the negative ad conditions; the ratings advantage enjoyed by beer ads became slightly more pronounced among subjects who saw one candidate or campaign attack the other.

(Figure 2 here)

Overall, people find little to like in political advertising: political ads are dismissed as unappealing, untruthful, and uninformative. Beer ads, on the other hand, are seen as appealing and truthful, albeit uninformative. Neither exposure to political advertising in general nor negative political advertising in particular influenced these attitudes.

The second phase of the analysis was directed at specific product and political ads. We asked subjects to indicate their feelings about the American Airlines and Advil ads included in the presentation, as well as the political ad. Statistical analysis of this measure is restricted to subjects in the treatment conditions since the control group was not exposed to the political ad. We focus on differences in subjects’ affective responses in relation to the number of political ads seen (1 or 2), the type of political ad (from a presidential campaign or initiative campaign), the candidate sponsoring the presidential ad (Clinton or Dole), and the tone of the ad (positive or negative).

Prior to examining the effects of political advertising on affective reactions to specific product ads, we present the mean net affect score for the product ads (averaged across the American Airlines and Advil ads), the positive political ads, and the negative political ads (see Figure 3). Clearly, the product ads elicited more positive reactions than political ads with the advantage being especially pronounced in the case of negative political ads. The product ads elicited a preponderance of positive reactions (a mean value of .14 on a –1 to +1 scale). At the other extreme, negative political ads elicited generally negative reactions (a mean of -.15). Positive ads were rated in-between, the mean score of -.04 indicating a virtual tie between positive and negative reactions.

(Figure 3 here)

Given subjects’ greater positivity toward product ads, is this advantage threatened by exposure to political advertising? We first examine the effects of the number and type of political ads on the mean affect scores for the product ads, the political ad, and the difference between the product and political ad scores. Positive differences indicate an advantage for product ads and vice versa. As shown in Figure 4, product ads were evaluated more positively than political ads regardless of experimental condition. The effects of the manipulation were limited to affect for the political ads. The mean affect score was most favorable among subjects exposed to a single presidential ad while exposure to a presidential and proposition ad produced the most negative response. Because of this significant movement in the ratings of the political ads, we also found significant experimental effects on the differences between the ratings of political and product ads.

(Figure 4 here)

As shown in the bottom panel of Figure 4, the number of political ads in the presentation significantly increased the contrast effect. Exposure to a presidential and proposition ad enlarged the difference between evaluations of product and political ads to a greater degree than exposure to any one political ad. Among subjects who saw only one political ad, the proposition ad widened the gap between product and political ads to a greater extent than did the presidential ad.

By making people more disenchanted with political ads, exposure to political campaigns enhances the public standing of product ads. As negative political ads are generally disliked more then positive political ads (e.g., Basil, Schooler & Reeves 1991; Thorson, Christ & Caywood 1991b), it is to be expected that the reputational advantage for product ads is furthered when people are exposed to negative political ads. Figure 5, which shows the difference in the net affect ratings of product and political ads broken down by tone and type of political ad, provides support for this expectation. A two-way ANOVA of the difference revealed a significant main effect for tone (F[1,1027]= 46.4, p<.001). No matter what type of political ad people saw, the evaluative difference between political and product ads was larger when the political ad was negative. Exposure to negative political campaigns makes political ads look especially bad. Figure 5 also elaborates on the finding that exposure to proposition ads produces larger contrast effects than exposure to presidential ads (F[1,1027]= 6.7, p=.01). It is now clear that this finding is equally true for positive and negative ads.

(Figure 5 here)

Finally, we examined the role of the presidential ad’s sponsor. So far, we have found that product ads benefit from proximity to any presidential ad. Yet to the extent that ads from different candidates elicit different feelings, we may find variability in the size of the contrast effect. Given the multiple differences between the Dole and Clinton candidacies, this analysis is merely exploratory. Figure 6 presents the results from an ANOVA of sponsor and tone. The main effect for sponsor is clearly visible, showing that Dole ads compared more unfavorably with the product ads than the Clinton ads (F[1,1072]= 6.6, p=.01). This effect does not depend on the tone of the ad. Figure 6 also confirms our results concerning the tone of the ad. No matter who the sponsor, negative presidential ads lead to larger contrast effects (F[1,1072]= 22.3, p<.001).

(Figure 6 here)

In sum, exposure to political ads does affect the standing of product ads vis-à-vis political ads. However, the effect is diametrically counter to the claims made by leaders of the advertising industry; exposure to a single political ad is sufficient to strengthen the distinction between product and political advertising thus making the former more attractive. The size of this contrast effect increases when people see two political ads instead of only one.

The contrast effect varies with type of political campaign (presidential or local), the tone of the political ad, and, in the case of the 1996 presidential campaign, the candidate sponsoring the ad. Product ads are set apart from political ads more clearly when people view negative political ads, ads from lesser known races such as referenda campaigns, and ads from less popular candidates.

Our most striking result is that product and political advertising are distinguished so clearly. Despite the inherent bias of all forms of advertising, people perceive product ads as generally truthful and interesting. In contrast, political ads are dismissed as dishonest, unappealing, and uninformative. When judged against political advertising, product advertising enjoys considerable public support. The evidence shows that political advertising hardly constitutes a threat to the advertising industry.

What accounts for the significant reputation gap between the two genres of advertising? We suggest two explanations. First, public distaste for political advertisements may stem from the belief that electoral choice and consumer choice are not equivalent activities. Purchasing a particular brand of soap or cereal is one thing, selecting the next president or senator quite another. The fact that voting is a "serious" task tends to undermine the legitimacy of "non-serious" forms of political communication. People may be averse to political advertising simply because it clashes with widespread norms concerning the nature of citizenship and campaigns for public office.

We are more certain of our second explanation, namely, that the negativity of the message breeds distrust of the medium, and, more broadly, of the political process. Our results show clearly that voters were especially turned off by political ads for both candidates and initiatives when they featured negative appeals. Commercial advertisers seek to depict their clients in the most attractive terms possible, often entertaining and amusing the audience in the process. Positivity is the currency of product advertising. In seeking to depict their clients’ opponents in the most unattractive terms possible, political advertisers typically anger, threaten, and repulse the audience. The overwhelming negativity of political advertising, while intended to weaken public support for the opposition candidate, also rubs off on the message itself.

The advertising industry takes for granted that the public’s dislike for political ads will spillover to product ads. In this study, however, we find a somewhat different result. Exposure to political ads drives down evaluations of political ads while leaving product ads untouched. In comparison with political ads, product ads appear even more attractive and credible. In other words, our results indicate a contrast effect. Watching political ads that they perceive as less truthful and appealing, people realize that product ads, relatively speaking, are not so bad. Rather than calling for a greater curbs on political advertisers, commercial advertisers should instead encourage the greater use of political advertising in general and negative political advertising in particular. If form holds and the next campaign features a sufficient number of nasty negative ads, many voters, instead of making the trip to the voting booth on election day, may well make the trek to the mall –or stay home to enjoy another Miller Light.

Figure 1: Average Ratings for Miller Beer and Political Ads: Information

Value, Veracity and Attractiveness

Figure 2: Average Ratings for Miller Beer and Political Ads: Information

Value, Veracity and Attractiveness by Tone

Figure 3: Mean Affective Index for Product Ads, Positive Political Ads and Negative Political Ads

Figure 4: Mean Affective Index and Difference Score by Type and

Number of Political Ads Seen

Figure 5: Difference Score by Type of Ad and Tone (One-Ad Design Only)

Figure 6: Difference Score by Tone and Sponsor (One-Ad or Two-Ad

Design, Presidential Ads)