MASS COMMUNICATION & SOCIETY, 2006, 9(3), 339–366

Late-Night Comedy and the Salience of the Candidates’ Caricatured Traits in the 2000 Election Dannagal Goldthwaite Young Department of Communication University of Delaware

This article examines the effects of exposure to late-night comedy shows on the salience of candidate traits prominent in the monologues of late-night comedians in the 2000 presidential campaign and assesses the moderating effects of political knowledge. Logistic regressions predict one’s likelihood of mentioning caricatured traits as a function of late-night comedy exposure, demographic, political, and media exposure variables. The results suggest that viewing late-night comedy was not directly associated with the salience of the candidates’ most caricatured traits, but among less politically knowledgeable individuals, the salience of certain caricatured traits did increase at higher levels of late-night comedy exposure. Across the models, the salience of caricatured candidate traits was associated with political knowledge and partisanship, suggesting that even if late-night comedy exposure does not increase the salience of candidate caricatures for the population overall, its content is hardly devoid of political meaning, as these caricatures are on the minds of political sophisticates.

In the 2000 presidential election, the breakdown of the news–entertainment distinction was exemplified by the case of late-night comedy programming. Journalists wrote articles on late-night’s potential effects on public opinion (Sella, 2000), late-night comedy hosts were interviewed on network and cable news programs to discuss their impact (Kurtz, 2000), and politicians appeared on NBC’s The Tonight Show with Jay Leno (Vickers, 2000a, 2000b, 2000d), CBS’s The Late Show with David Letterman (Burnett, Gaines, & Pope, 2000a, 2000b, 2000d), and Saturday Night Live (Michaels, 2000). Late-night comedy has also emerged as an area of reCorrespondence should be addressed to Dannagal Goldthwaite Young, Department of Communication, University of Delaware, 250 Pearson Hall, Newark, DE 19716. E-mail: [email protected]

340

YOUNG

search among political scientists and communications scholars (Chaffee, Zhao, & Leshner, 1994; Davis & Owen, 1998; Hollander, 1994; Moy, Xenos, & Hess, 2004; Pfau, Cho, & Chong, 2001; Young, 2004). This project extends this body of research by exploring one route through which late-night jokes might construct the political world: by priming the candidates’ most caricatured personality traits in viewers’ overall assessments of the candidates. These analyses also explore how viewers with different levels of political knowledge may be differentially affected through this process.

LATE-NIGHT COMEDY AS A SOURCE OF CAMPAIGN INFORMATION Data released by the Pew Research Center for the People and the Press (2000) indicated shifts in the sources from which young people reported receiving campaign information. The report noted that young people were increasing in their likelihood of citing comedy shows like Leno and Letterman as sources of campaign information and decreasing in their citing of network news programs. More recent reports released by the Pew Research Center (2002, 2004) indicated that the trend continued into 2002 and 2004, with young people reporting higher rates of exposure to late-night comedy programs than older people and record rates of campaign information being obtained, not only from Leno and Letterman but also from comedy shows like Saturday Night Live and The Daily Show with Jon Stewart on Comedy Central. A report by the National Annenberg Election Survey (NAES; 2004) in the 2004 election campaign concluded that viewers of The Daily Show with Jon Stewart were younger and more liberal than people who did not watch late-night but that they were also more knowledgeable about the issue positions of the major candidates. Taken together, these studies indicate that late-night comedy shows may be growing in political importance, particularly among young people who have historically been our least politically involved citizens (Delli Carpini & Keeter, 1996; Keeter, Zukin, Andolina, & Jenkins, 2002).

THE POLITICAL IMPACT OF ENTERTAINMENT PROGRAMMING “Widespread recognition of the new media’s presence on the American political scene occurred during the 1992 presidential election campaign,” wrote Davis and Owen (1998, p. 4). After then-Governor Bill Clinton became the first “talk show president,” political scientists and political communications scholars became interested in the impact these nontraditional sources of political information on out-

LATE-NIGHT COMEDY AND ELECTION 2000

341

comes like learning (Chaffee et al., 1994; Hollander, 1994; McLeod et al., 1996) and persuasion (Pfau, Cho, & Chong, 2001; Young, 2004). In addition to research on learning and persuasion, several recent studies have explored media priming theory in the context of nontraditional forms of political information such as The West Wing (Holbert et al., 2003), television crime dramas (Holbrook & Hill, 2005), and candidate interviews on late-night comedy programs (Moy et al., 2006). Here scholars have examined how exposure to such programs can change the basis on which viewers form their evaluations of public officials. For instance, Moy et al. found that the salience of how much George W. Bush “cares about people like me” increased during the time period that included his appearance on The Late Show with David Letterman in the 2000 election. Hence, viewing Letterman’s interview with Bush might have caused people to weigh this more personal characteristic of “caring” more heavily in their evaluations of Bush. Holbert et al. found that after viewing an episode of The West Wing, individuals judged traits such as compassion and warmth as more important to a successful presidency. Holbrook and Hill conducted experimental and survey research that indicated that exposure to crime dramas such as NYPD Blue was associated with a higher likelihood of mentioning crime as one of the country’s most important problems and with a priming effect in which presidents were more heavily judged on how well they dealt with the problem of crime. Their survey research also suggested that these effects were strongest among habitual viewers, indicating that the effect they witnessed was likely due to the chronic activation of crime in the minds of those viewers.

THE CONTENT OF LATE-NIGHT COMEDY PROGRAMMING Research on late-night joke content has indicated that over the past several years, the monologues of major late-night comedy programs like those hosted by Jay Leno and Dave Letterman have been surprisingly consistent with one another in their content, focusing predominantly on the personal traits of public officials (Niven, Lichter, & Amundson, 2003; Young, 2004). A content analysis of the late-night comedy jokes made by Jay Leno and Dave Letterman in the 2000 presidential campaign revealed that 521 jokes were made about George W. Bush and 383 about Al Gore (Young, 2004). Despite the large number of candidate jokes told, the jokes were homogeneous in topic. The study found that the most common character traits mentioned in late-night jokes were the same in both Leno and Letterman’s monologues, with Gore’s main caricatures being his stiff and dull personality and his tendency to exaggerate and Bush’s main caricature being his lack of intelligence. (Examples of late-night candidate jokes are included in Appendix B.)

342

YOUNG

MAKING SENSE OF JOKES: THE COGNITIVE IMPLICATIONS Generally, studies of the effects of late-night political jokes have treated exposure to late-night comedy as they have exposure to television news or newspaper reading, adding it as a predictor on one side of the equation, but without a theoretical explanation to account for how exposure to such content ought to influence outcomes like attitude change. But political humor is not equivalent to news or presidential debates. Although more young people are citing late-night comedy as a source of information (Pew Research Center, 2000, 2004), in general, viewers of talk shows are less likely than news viewers to be using the shows’ content for the intended purpose of obtaining political information. Instead, as Prior’s (2003) research illustrates, viewers of late-night comedy and other soft news programs are more likely to be in search of entertainment, and any political content encountered in the process is merely incidental. If the goal when viewing late-night comedy shows is to be entertained, then “getting the jokes” is central to the fulfillment of the viewing experience. Because appreciating humor requires subtle cognitive processes, viewers might experience outcomes other than being entertained, namely, the subtle shaping of their political judgments. Wyer and Collins’s (1992) comprehension–elaboration model of humor appreciation posits that the appreciation of humor begins as the introductory text of a joke is encoded based on the accessibility of constructs in memory. This encoding activates nearby constructs to “form general expectations concerning the range of concepts and knowledge that are applicable to an understanding of future events involving these elements” (p. 671). Cognitive elaboration is necessary to reinterpret the new information in light of the old information (similar to the reconciliation of incongruity in models of Koestler, 1964, and Suls, 1972) to appreciate the humor. The process of elaborating on a joke and filling in the missing information can also be thought of in terms of construct activation, in which information is brought from long-term into working memory. Schmidt (1994, 2001) argued that it is this cognitive elaboration that caused the enhanced recall of humorous images and statements he witnessed in experimental studies. There are numerous models which conceptualize knowledge as being stored in interconnected networks of information (Anderson, 1983; Collins & Quillian, 1969; Srull, 1981). For our purposes, it is useful to conceptualize knowledge as stored in nonhierarchical associative networks (Anderson, 1983) consisting of concept or belief nodes that are interconnected. Through the process of spreading activation (Collins & Loftus, 1975), the residual excitation experienced by recently or frequently activated nodes and networks renders certain constructs more accessible than others (Bargh, Lombardi, & Higgins, 1988; Higgins, 1989; see Price & Tewksbury, 1997).

LATE-NIGHT COMEDY AND ELECTION 2000

343

One direct result of having to engage in cognitive elaboration to comprehend a joke concerns the activation and subsequent salience of constructs central to the understanding of that joke. In the context of late-night comedy jokes about political candidates, the missing information can often be conceptualized as a proposition (e.g., Candidate A is not intelligent; Candidate B is boring). As construct salience is fostered by recent and frequent activation, it follows that listening to these jokes will render those constructs (propositions) increasingly salient over time as a result of frequent exposure. Because the content of late-night comedy jokes is indeed repetitive (Niven et al., 2003; Young, 2004), regular exposure to such simplified representations of the candidates should enhance the chronic accessibility of the candidate caricatures portrayed in these jokes. In Cohen’s (1963) terms, late-night jokes might not have told viewers to think that Gore was a dull exaggerator and Bush was unintelligent, but they may have caused viewers to think about these caricatures. This kind of effect does not assume that individuals are acquiring factual political information from viewing, but rather viewers’ judgments of political candidates are affected by the priming of these caricatured traits with which viewers are already familiar. Baum (2003) described such effects of exposure to soft news shows as a form of “learning,” even though they do not involve the acquisition of factual knowledge per se. He argued that “learning” includes not only “increasing the volume of factual political knowledge” but also cognitive benefits such as the construction and reinforcement of heuristic cues that people can later use to aid in political judgment formation (p. 174). As suggested by Roskos-Ewoldsen, Roskos-Ewoldsen, and Carpentier (2002), perhaps this process of frequent and repetitive exposure fostering chronic construct accessibility should be described as “political cultivation” rather than “political priming.” Because priming more accurately describes a temporary increase in construct salience, the authors argued, employing the term priming in the context of long-term effects such as those hypothesized here is inappropriate. This debate over the appropriate term to describe this effect is not solved here, but it should be noted that the proposed effects of regular exposure to late-night comedy shows should be considered in terms of chronic rather than temporary construct accessibility. Finally, this study does not examine the effects of exposure to late-night comedy caricatures on salience alone but rather the effects of such exposure on the salience of caricatured traits in the context of broader candidate evaluations. Hence, according to the definitions proposed by Iyengar and Kinder (1987) and Scheufele (2000), this study qualifies as a study of priming effects. H1: Late-night comedy viewing increases the salience of the candidates’ most caricatured traits, hence causing viewers of late-night comedy to be more likely to mention these traits in their evaluations of the candidates than people who do not watch late-night comedy.

344

YOUNG

POLITICAL KNOWLEDGE AS A MODERATOR Research on late-night comedy audiences has indicated that shows like those hosted by Jay Leno and Dave Letterman are viewed more by younger people (NAES, 2004; Pew Research Center, 2000, 2002, 2004). Because age and political knowledge have historically been negatively correlated (Delli Carpini & Keeter, 1996), it is appropriate to ask how political knowledge—or a lack of political knowledge—might be fostering or impeding the potential influence of late-night comedy. On an intuitive level, it would seem that apolitical viewers would be more likely to be affected by chance exposure to political content. Literature on the role of knowledge in message effects (McGuire, 1968, 1972) suggests that knowledge mitigates certain effects. Zaller (1992) illustrated how the politically aware are more likely to encounter messages in the political environment but are more resistant to their potentially persuasive effects. If late-night comedy viewers are younger and less politically knowledgeable than traditional news viewers (an assumption complicated by the NAES report [2004]), then late-night comedy viewers may be incidentally receiving political information (Baum, 2003) without the knowledge to counter its impact. This process would result in greater effects of exposure to late-night comedy among viewers with less political knowledge. Recent research in the areas of soft news and late-night comedy has indicated stronger effects among low-knowledge viewers. Baum (2005) found that people who watched entertainment talk shows perceived their opposing party’s candidate to be more likeable than viewers of traditional public affairs programming. This relationship was stronger among talk show viewers who were less politically aware. Young (2004) found that late-night comedy viewers with less political knowledge were more affected by late-night comedy exposure than those higher in knowledge in their ratings of candidate traits. Findings such as these do suggest that late-night comedy’s influence may be augmented by a lack of political knowledge. Political knowledge has been conceptualized and measured several different ways by political communication scholars (see Price, 1999). Some knowledge scales, such as those of Delli Carpini and Keeter (1996) and Zaller (1992), are measurements of civics knowledge and knowledge of national political parties and leaders. In contrast, Patterson and McClure (1976) measured political knowledge with items assessing participants’ knowledge of political campaign information such as presidential candidates’ stands on political issues. In general, these different measures are highly correlated (Price, 1999) but do capture distinct knowledge constructs, with one’s understanding of civics knowledge likely stored in long-term memory, and one’s knowledge of presidential candidates increasing throughout a campaign, hence representing newly acquired information. In the context of latenight comedy priming criteria on which to judge the candidates, knowledge of campaign information such as candidate biographies and issue positions, rather than civics knowledge, is expected to be more likely to serve as alternative judgment criteria.

LATE-NIGHT COMEDY AND ELECTION 2000

345

With relevant alternative criteria on which to base judgments of political candidates, individuals high in these forms of political knowledge (knowledge of candidate biographies and issue stands) should be less primed by late-night comedy show viewing. H2: The priming of candidate caricatures as a result of late-night comedy viewing will be stronger among viewers who are less knowledgeable about the candidates issue positions and biographies.

METHOD Data Data for this project was obtained for the Electronic Dialogue 2000 study, funded by grants from the Pew Charitable Trusts and the Annenberg Public Policy Center to Dr. Vincent Price and Dr. Joseph Cappella, principal investigators. The Electronic Dialogue project was designed to examine electronic discussion among citizens in the 2000 presidential election. In February 2000, a random sample of American citizens age 18 and older (N = 3,967) was drawn from a nationally representative panel of survey respondents maintained by Knowledge Networks, Inc., of Menlo Park, California. The Knowledge Networks panel included a national sample of households that had agreed to accept WebTV equipment in exchange for completing online surveys. Of those recruited through the February recruitment survey, 51% agreed to participate in the project. Participants were invited either to (a) discuss politics online and complete monthly surveys (discussants), (b) complete the monthly surveys but not discuss politics online (control group), or (c) complete the baseline and final surveys only (set aside). Analysis of group characteristics (demographics, age, race, gender, political interest, ideology, and party leanings) confirmed that no significant differences were found between the three groups, indicating that randomization was successful. The discussion group consisted of 906 individuals, the control consisted of 139, and the set aside (first and last surveys only) consisted of 969. The survey containing the late-night comedy exposure item and the open-ended questions to be coded with the content analysis was issued in September to discussants and control group members only. Means comparisons were run to verify that late-night comedy exposure and political knowledge (our two main variables of interest) were not confounded with group level characteristics (e.g., discussion group vs. control).1 1Late-night comedy exposure (discussion group M = .94, SD = 1.02, control M = .91, SD = 1.01), background knowledge (discussion group M = .72, SD = .22, control M = .71, SD = .22), issue knowledge (discussion group M = .52, SD = .20, control M = .51, SD = .22), civics knowledge (discussion group M = .72, SD = .27, control M = .71, SD = .27) total knowledge (discussion group M = .65, SD = .19, control M = .64, SD = .18).

346

YOUNG

Measures

Salience of candidate traits. The salience of candidate traits was obtained through a content analysis of four open-ended questions issued on the September survey (in the field from August 25 to September 4, 2000). Four open-ended questions were posed to respondents, two questions about each candidate: “At this point, what would you say you like most about [George W. Bush/Al Gore]?,” “At this point, what would you say you like least about [George W. Bush/Al Gore]?” (Total valid n = 685) The construction of the late-night comedy trait salience measures was roughly based on a measure of issue salience used by the American National Election Study to capture the extent to which an individual had “reasons for liking or disliking the two major party candidates for Senate” (Clarke & Fredin, 1978, p. 147). Later studies by Brians and Wattenberg (1996); Holbert, Benoit, Hansen, and Wen (2002); and Benoit and Hansen (2004) employed similar measures to capture “issue salience.” These studies measured issue salience by asking respondents what they liked and did not like about the candidates that might have influenced their votes. Individuals who mentioned policy issues frequently in response to this open-ended question were considered as having high issue salience. Rather than coding for the number of substantive issue-based reasons for liking or disliking a candidate, the coding scheme used in the following analysis coded those reasons that had been the focus of late-night comedy content, hence capturing “late-night comedy trait salience” rather than “issue salience.” It is important to note that for the content analysis that follows, only the “like least” responses were included. Because late-night joke content is designed to deride candidates’ weaknesses, the traits that appear in late-night jokes are almost exclusively candidate flaws, not characteristics that show up in response to “what do you like most” about a candidate. The coding scheme used to code the “like least” responses was based on Young’s (2004) late-night joke content analysis. The results of that content analysis indicated that the most frequent caricature category for each candidate was the same in both Leno and Letterman’s monologues, with jokes about Gore as stiff and dull comprising 20% of the Gore jokes made by Leno and 17% of the Gore jokes made by Letterman. The caricature of Bush as generally unintelligent was the most frequent Bush category in both programs, making up 24% of the Bush jokes made by Leno and 34% of the Bush jokes made by Letterman. The study also found a large portion of the jokes made by both hosts concerned Gore’s lack of integrity or honesty, such as jokes about his tendency to exaggerate (e.g., claiming he invented the Internet) and reinvent himself according to his audience, as well as his illegal fund-raising scandal. Hence, the most salient caricature attributes in late-night comedy overall in 2000 were Bush as unintelligent, Gore as an exaggerator and a liar, and Gore as stiff and dull. The content analysis of the open-ended responses

LATE-NIGHT COMEDY AND ELECTION 2000

347

captures mentions of candidate traits that are consistent with these late-night comedy caricatures, that is, mentions of Bush’s intelligence/competence, Gore’s honesty/trustworthiness, and Gore’s charisma. The reliability of the open-ended coding instrument was tested between the author and another graduate student coder. One hundred sixty units were selected for a reliability analysis, 4 individual responses from each of 40 participants. Krippendorff’s α was calculated for each of the variables, obtaining high reliability for each of the codes (α: mentions of intelligence/competence = .94, mentions of honesty or trustworthiness = .84, mentions of charisma = .88). To check for the correlative validity of the “late-night comedy trait salience” codes, correlations were run between each of the codes and closed-ended ratings of the appropriate candidate traits.2 As expected, mentions of Gore’s dishonesty were significantly negatively correlated with closed-ended ratings of Gore’s honesty (r =-.41, p < .001) and mentions of Bush’s lack of intelligence were significantly negatively correlated with closed-ended ratings of Bush on the knowledgeable item (r = –.20, p < .001). In both cases, the higher the rating on the positively valenced trait, the lower the likelihood of mentioning a lack of that characteristic as something the respondent liked least about the candidate. The correlation between closed-ended ratings of Gore on the inspiring item were significantly positively correlated with mentions of Gore’s lack of charisma (r = .11, p < .01). Although counterintuitive on its face, the data suggest that when rating Gore on the closed-ended trait items, respondents were likely projecting a sense of overall favorability toward Gore onto his inspiring rating. A look at the actual open-ended texts explains the pattern. Most individuals who mentioned Gore’s lack of charisma as something they liked least about him were Democrats (highly favorable toward Gore) who believed that his lack of charisma was preventing him from communicating his message and hence hurting his chances of winning the election. When responding freely, Gore fans were critical of his lack of charisma, but when rating Gore on a scale, they inflated their assessments. To verify this post hoc explanation, a partial correlation was run between the openended mentions of Gore as uncharismatic and the closed-ended Gore inspiring ratings, this time controlling for overall Gore favorability. Here the correlation was not statistically significant (p < .11), although it was negative (r = –.07). It seems that overall attitude toward Gore is responsible for much of the positive relationship between these two items. This post hoc examination would suggest that, if anything, the open-ended mentions of Gore’s lack of charisma are more useful 2Participants were asked how well various traits described the candidates. The traits inspiring, knowledgeable, and honest were issued in randomized order. Options were extremely well (3), quite well (2), not too well (1), and not well at all (0). Although this scale is not interval, it is used in a Pearson’s correlation for a general understanding of strength and direction of association with the open-ended 0,1 codes.

348

YOUNG

measures than the closed-ended ratings. Looking at comparative prevalence of these various mentions, respondents were more likely to mention dishonesty with Gore than Bush (23.5% vs. 9.2%), more likely to mention a lack of inspiration with Gore than Bush (16.9% vs. 2.5%), and more likely to mention a lack of intelligence with Bush than with Gore (8% vs .2%).3 Late-night comedy exposure was obtained on the September survey (in the field from August 25 to September 4, 2000) from 64% of the discussants (n = 580) and 63% of the control group (n = 88) due to panel attrition over the course of the project. Respondents were asked, “Do you happen to watch any of the following entertainment programs on television?” followed by a list of programs including Jay Leno and Dave Letterman. The options were “No” (coded as 0), “Yes, but not very often” (coded as 1), and “Yes, much of the time” (coded as 2). Although these are two different late-night comedy programs, the element of their content that is most important in the context of the theoretical mechanism presented here (candidate caricatures) was consistent across programs. That is, the monologue jokes about Bush and Gore made by both Leno and Letterman highlighted the same caricatured traits most often, with Bush as unintelligent and Gore as stiff and dishonest (Young, 2004). For this reason, Leno and Letterman exposure items were combined using a sum score. Almost half of the sample (48%) watched no late-night comedy, whereas one-third watched either program but not very often. Less than 1 in 10 respondents watched Leno or Letterman much of the time. (M = .93, SD = 1.03). Demographic variables serving as predictors in the models include age in years (M = 48.88, SD = 15.79), gender (1 = male, 0 = female; M = .53, SD = .50), and years of education (M = 13.66, SD = 1.71). Political knowledge was obtained using 24 available political knowledge items. Correct answers were coded 1 and incorrect answers, don’t know, and refused were coded 0. (For a complete list of items, see Appendix A.) The 24 items captured three different kinds of knowledge. The first battery consisted of 7 questions regarding the candidates’ biographies (four Democrat, three Republican; M = .75, SD = .21; Cronbach’s α = .72). The second consisted of 7 questions about the candidates’ issue positions (four Democrat, three Republican; M = .54, SD = .20; Cronbach’s α = .46). The third battery consisted of 10 civics knowledge questions, based on Delli Carpini and Keeter’s (1996) and Zaller’s (1992) political knowledge scales (M = .78, SD = .23; Cronbach’s α =.83). Finally, all 24 items were combined into one total knowledge battery (M = .69, SD = .17; Cronbach’s α = .82; see Appendix A for details of the items). 3In response to “Like least about Bush” (n = 512), 47 respondents (9.2%) mentioned Bush being dishonest, 13 (2.5%) mentioned his lack of charisma, and 41 (8%) mentioned his lack of intelligence. In response to “Like least about Gore” (n = 531), 125 (23.5%) mentioned dishonesty, 90 (16.9%) mentioned his lack of charisma, and 1 (.2%) mentioned his lack of intelligence.

LATE-NIGHT COMEDY AND ELECTION 2000

349

In addition to the main variables of interest, the logistic regression models also include various predictors likely to be correlated with late-night comedy viewing, political knowledge, or the salience of certain candidate traits. Partisanship, for example, is traditionally correlated with political knowledge and is likely a strong predictor of what comes to mind when asked what a respondent likes least about a given candidate. Political interest has been found to be a strong significant predictor of issue salience (Holbert et al., 2002) and is also related to political knowledge (Delli Carpini & Keeter, 1996). Because the caricatures in late-night comedy are driven by real-world events and are grounded in news content, it follows that other forms of media exposure should be included in the models to verify that any effects of late-night exposure have been separated from other forms of media use. Therefore the models include exposure to local, network, and cable news as well as listening to political talk radio and reading the newspaper. Finally, because one of the Gore caricatures dealt with his dishonesty and willingness to say anything to get elected, cynicism was included in the models. Because cynics view politics largely as a game in which candidates are more interested in winning then actually helping society (Cappella & Jamieson, 1997), these are the people for whom a willingness to say anything to get elected would be most salient. To be able to compare coefficients across models, cynicism was included in models predicting the salience of all three traits. Partisanship was obtained on the baseline survey in February. The scale ranges from 1 to 7, where 1 is strong Republican, 7 is strong Democrat, and 4 is moderate or independent (M = 3.90, SD = 2.21). Political interest was the average of two 4-point scale items. The first asked how much the respondent cared who won the election (coded 1 to 4 from 1 [not at all] to 4 [a great deal]) and the second asked how often the respondent followed public affairs (coded 1 to 4 from 1 [hardly at all] to 4 [most of the time]; M = 3.36, SD = .63). Cynicism was obtained from respondents on the February baseline with four force-choice questions about candidates. Cynical responses were coded 1 and trustful responses were coded 0. These were the four items presented: Which comes closest to your views: 1) Candidates are talking about issues I care about v. Candidates are ducking the important issues facing the country, 2) Candidates usually tell the public what it wants to hear v. Candidates usually tell the public what they actually believe is best for the country, 3) Candidates tell voters what they will really try to do if they are elected v. Candidates cannot be trusted to keep promises they make while campaigning, and 4) Candidates usually vote the way their big financial contributors want them to v. Candidates usually vote the way they think best serves the country. These four items were averaged to create a scale. Cronbach’s α = .62 (M = .74, SD = .28).

350

YOUNG

Media exposure variables (in addition to late-night comedy exposure) include the number of days in the past week the respondent watched national news (M = 3.50, SD = 2.52), cable news (M = 2.86, SD = 2.62), or local news (M = 4.54, SD = 2.44); listened to political talk radio (M = 1.70, SD = 2.31); and read the newspaper (M = 4.13, SD = 2.73). Analytical procedure Logistic regressions were run to predict the probability of mentioning the candidates’ most caricatured traits in late-night joke content in response to the “like least about” question. The dependent variables in the models were mentions of Gore’s lack of charisma, Gore’s lack of honesty, and Bush’s lack of intelligence. The predictor variables included in the models were late-night comedy viewing, age, gender, education, partisanship, political knowledge, political interest, cynicism, national news exposure, cable news exposure, local news exposure, political talk radio listening, and newspaper reading. To assess the extent to which late-night comedy viewing’s relationship to candidate trait salience was contingent on the respondent’s political knowledge, a Political Knowledge × Late-Night Comedy Exposure interaction term was added as a predictor in each model. The interaction term was constructed four ways to assess how different kinds of political knowledge interacted with late-night comedy viewing in predicting caricatured trait salience (total knowledge, knowledge of candidates’ backgrounds, knowledge of candidate issue positions, and civics knowledge).

RESULTS The results of the model predicting the probability of mentioning Bush’s lack of intelligence are shown in the first column in Table 1. Late-night comedy exposure was not significantly associated with the salience of this trait. The strongest predictor of whether someone mentioned Bush’s lack of intelligence in response to “like least about Bush” was total political knowledge (24-item scale), so that respondents higher in political knowledge were far more likely to mention this trait as a concern (p < .001). Interpreting the odds ratio, we see that respondents who received a perfect total knowledge score were almost 209 times more likely to mention this trait than respondents who scored zero on political knowledge. In addition, the more people watched cable news, the less likely they were to mention Bush’s lack of intelligence (p < .05), with each additional day of exposure reducing one’s odds of mentioning this trait by 15%. National news exposure was also marginally significant, such that greater exposure increased one’s likelihood of mentioning Bush’s lack of intelligence.

LATE-NIGHT COMEDY AND ELECTION 2000

351

TABLE 1 Logistic Regressions Predicting Mentions Bush and Gore’s Most Caricatured Traits Bush Lack of Intelligence B (SE B) Constant Demographics Age Gender Education Political variables Political knowledgea Party identification Interest Cynicism Media use National news Cable news Local news Talk radio Newspaper Late-night comedy R2 N

Exp(B)

–7.12 (1.76)

.00****

.00 (.01) –.13 (.37) .09 (.10)

1.00 .87 1.09

5.34 (1.60)

208.95****

Gore Dishonesty

Gore Lack of Charisma

B (SE B)

Exp(B)

B (SE B)

Exp(B)

–3.00 (1.18)

.05**

–5.88 (1.29)

.02 (.01) –.46 (.25) –.06 (.07)

1.02* .63* .94

.00 (.01) –.19 (.27) .10 (.08)

1.76 (.89)

5.80**

2.91 (1.02)

.00**** 1.00 .83 1.10 18.29***

.08 (.08) –.10 (.36) –.22 (.59)

1.08 .90 .80

–.27 (.05) .33 (.23) .98 (.43)

.76**** 1.39 2.68**

.34 (.06) .00 (.25) –.70 (.43)

1.40**** 1.00 .50

.17 (.09) –.16 (.07) –.13 (.08) –.03 (.08) .02 (.07) .06 (.17)

1.18* .85** .88 .97 1.02 1.06

–.03 (.06) –.01 (.05) –.02 (.06) .06 (.05) –.02 (.05) .11 (.11)

.97 .99 .98 1.06 .98 1.12

.09 (.07) .01 (.05) –.04 (.06) –.08 (.06) .01 (.05) –.01 (.13)

1.09 1.01 .96 .92 1.01 .99

.13 513

.16 524

.19 524

aPolitical

knowledge is measured as total 24-item knowledge scale. *p < .1. **p < .05. ***p < .01. ****p < .001

The probability of mentioning Gore’s dishonesty in response to “like least about Gore” was a function of several political and demographic variables (see column 2 of Table 1). Again, late-night comedy viewing was not associated with the salience of this trait. The strongest predictor in the model was partisanship, so that for every step closer one moved to the Democratic side of the 7-point scale, one’s odds of mentioning Gore’s dishonesty decreased 24%. As in the Bush unintelligent model, political knowledge was a significant positive predictor of mentioning Gore’s dishonesty. More cynical respondents were also more likely to mention Gore’s dishonesty. Respondents who answered all four items cynically were more than two and a half times more likely to mention Gore’s dishonesty than respondents who answered all four items in a trusting manner. The model predicting one’s likelihood of mentioning Gore’s lack of charisma in response to “like least about Gore” (column 3 of Table 1) suggests that Democrats were more likely than Republicans to mention Gore’s lack of charisma as what they liked least about him. For each step one moved closer to the Democratic side of the

352

YOUNG

7-point scale, one’s likelihood of mentioning Gore’s lack of charisma increased almost 1½ times. Another significant predictor was political knowledge (p < .01); such respondents higher in knowledge were more likely to cite Gore’s lack of charisma as something they liked least about him than were lower knowledge respondents. Respondents with perfect knowledge scores were 18 times more likely to mention Gore’s lack of charisma than respondents who scored zero items correct. Tables 3, 4, and 5 show the logistic regression results predicting probability of mentioning Bush’s lack of intelligence, Gore’s dishonesty, and Gore’s lack of charisma respectively. These models also include the interaction of late-night comedy viewing with total political knowledge, candidate background knowledge, issue knowledge, and civics knowledge. There were no significant interaction terms in either of the Gore models (Tables 4 and 5). This indicates that the relationship between late-night comedy viewing and the salience of Gore’s dishonesty or lack of charisma was not significantly different for individuals with different levels of political knowledge. Table 2, however, does illustrate significant interactions (p < .05) of several forms of political knowledge, with late-night comedy viewing in the model predicting the salience of Bush’s lack of intelligence. The strongest interactions were with total knowledge and civics knowledge (p < .05), and the interaction of late-night comedy viewing and candidate issue knowledge was only marginally significant (p < .1). Interpreting results of logistic regression coefficients requires a conversion from exponentiated form back into a probability. The predicted probability for each set of late-night comedy and political knowledge values was calculated by substituting the mean values for the control variables in the model and sample values of exposure to late-night comedy programming and total political knowledge. Political knowledge values used in the model to denote low, medium, and high knowledge respondents were the 20th, 50th, and 80th percentile knowledge values, respectively. 1/(1 + EXP(( – b0) – (Mx1, x2…*bx1, x2,…) – (x3*bx3) – (x4*bx4) – ((x3*x4)*bx3*x4)))4 Following these steps using the coefficients in column 1 of Table 2 produces the graph shown in Figure 1. According to the illustration, among respondents low in total political knowledge, increases in late-night comedy viewing were associated with increases in one’s probability of mentioning Bush’s lack of intelligence in response to “like least about Bush.” The pattern is a bit weaker among those with moderate levels of political knowledge. Among respondents high in political knowledge, the trend appears to flatten, and somewhat reverse, so that increases in late-night comedy exposure were associated with a slightly lower likelihood of 4In this model, x1, x2, x3 represent control variables included in the model. x3 is exposure to late-night comedy programming, x4 is political knowledge.

TABLE 2 Logistic Regressions Predicting Mentions of Bush’s Lack of Intelligence Basic Model With Total Knowledge B (SE B) Constant Demographics Age Gender Education Political variables Political knowledgea Party identification Interest Cynicism Media use National news Cable news Local news Talk radio Newspaper Late-night comedy Late-night comedy × knowledge Nagelkerke R2 N

Exp(B)

–9.21 (2.11)

.00****

With Knowledge of Biographies B (SE B)

Exp(B)

–8.58 (2.10)

.00****

With Knowledge of Issue Positions B (SE B)

Exp(B)

–6.97 (1.74)

.00****

With Civics Knowledge B (SE B)

Exp(B)

–8.32 (2.18)

.00****

.00 (.01) –.12 (.37) .09 (.10)

1.00 .89 1.09

.00 (.01) .07 (.36) .14 (.10)

1.00 1.08 1.15

.01 (.01) .11 (.36) .13 (.10)

1.01 1.12 1.13

.00 (.01) –.13 (.36) .09 (.10)

1.00 .87 1.09

8.19 (2.19) .06 (.08) –.14 (.36) –.16 (.60)

3612.5**** 1.06 .87 .85

4.67 (1.77) .06 (.08) .11 (.34) –.14 (.58)

106.60*** 1.07 1.12 .87

3.08 (1.26) .06 (.08) .21 (.34) –.06 (.58)

21.75** 1.06 1.23 .94

6.45 (1.96) .07 (.08) –.21 (.36) –.34 (.60)

630.46*** 1.07 .81 .71

.16 (.09) –.13 (.07) –.14 (.08) –.03 (.08) .02 (.07) 1.11 (.83) –1.31 (1.00)

1.17* .87* .87* .97 1.02 3.03 .27

.18 (.09) –.15 (.07) –.13 (.08) –.04 (.08) .01 (.08) 2.03 (.89) –2.61 (1.16)

1.19* .86** .88 .96 1.01 7.64** .07** .15 513

.11 512

.17 (.09) –.13 (.07) –.15 (.08) –.03 (.08) .02 (.07) .89 (.48) –1.44 (.78)

1.18* .88* .86* .97 1.02 2.44* .24* .10 512

.18 (.09) –.15 (.07) –.14 (.08) –.04 (.08) .02 (.07) 1.69 (.83) –1.88 (.94)

1.20* .86* .87 .96 1.02 5.44** .15** .14 511

353

aIn basic model, political knowledge is the combined 24-item scale of biography, issue, and civics knowledge. In second model, knowledge is candidate biography knowledge only, in third model it is knowledge of candidate issue positions only, and in fourth model, knowledge is civics knowledge only. *p < .1. **p < .05. ***p < .01. ****p < .001.

354

TABLE 3 Logistic Regressions Predicting Mentions of Gore’s Dishonesty Basic Model With Total Knowledge

Constant Demographics Age Gender Education Political variables Political knowledgea Party identification Interest Cynicism Media use National news Cable news Local news Talk radio Newspaper Late-night comedy Late-night comedy × knowledge Nagelkerke R2 N

With Knowledge of Biographies

With Knowledge of Issue Positions

With Civics Knowledge

B (SE B)

Exp(B)

B (SE B)

Exp(B)

B (SE B)

Exp(B)

B (SE B)

Exp(B)

–3.07 (1.27)

.05**

–3.24 (1.28)

.04**

–2.84 (1.20)

.06**

–2.89 (1.23)

.06**

.02 (.01) –.46 (.25) –.06 (.07)

1.02* .63* .94

.02 (.01) –.44 (.25) –.05 (.07)

1.02* .65* .95

.02 (.01) –.37 (.25) –.05 (.07)

1.02* .69 .96

.02 (.01) –.35 (.24) –.04 (.07)

1.02 .70 .96

1.86 (1.14) –.27 (.05) .33 (.23) .99 (.43)

6.43* .76**** 1.39 2.68**

1.54 (.91) –.27 (.06) .37 (.23) .99 (.43)

4.69* .76**** 1.45 2.70**

.80 (.77) –.27 (.05) .41 (.23) 1.04 (.43)

2.22 .76**** 1.51* 2.84**

.66 (.81) –.27 (.05) .37 (.24) 1.02 (.43)

1.94 .76**** 1.45 2.78**

–.03 (.06) –.01 (.05) –.02 (.06) .06 (.05) –.02 (.05) .19 (.54) –.11 (.73)

.97 .99 .98 1.06 .98 1.21 .90

–.04 (.06) .00 (.05) –.02 (.06) .05 (.05) –.02 (.05) .07 (.49) .05 (.60)

.97 1.00 .98 1.06 .98 1.07 1.05

–.03 (.06) .00 (.05) –.03 (.06) .06 (.05) –.01 (.05) .14 (.34) –.05 (.55)

.97 1.00 .97 1.06 .99 1.15 .95

–.01 (.06) .00 (.05) –.03 (.06) .05 (.05) –.01 (.05) .22 (.44) –.10 (.51)

.99 1.00 .97 1.06 .99 1.24 .90

.16 524

.17 522

.16 522

aIn basic model, political knowledge is the combined 24-item scale of biography, issue, and civics knowledge.

.15 522

In second model, knowledge is candidate biography knowledge only, in third model it is knowledge of candidate issue positions only, and in fourth model, knowledge is civics knowledge only. *p < .1. **p < .05. ***p < .01. ****p < .001.

TABLE 4 Logistic Regressions Predicting Mentions of Gore’s Lack of Charisma Basic Model With Total Knowledge B (SE B) Constant Demographics Age Gender Education Political variables Political knowledgea Party ID Interest Cynicism Media use National news Cable news Local news Talk Radio Newspaper Late-night comedy Late-night comedy × knowledge Nagelkerke R2 N

Exp(B)

–5.51 (1.37)

.00****

.00 (.01) –.19 (.27) .10 (.08)

1.00 .83 1.11

2.32 (1.26) .34 (.06) .00 (.26) –.70 (.43)

10.13* 1.40**** 1.00 .49

.09 (.07) .01 (.05) –.04 (.06) –.08 (.06) .01 (.05) –.51 (.67) .70 (.90)

1.09 1.01 .96 .92 1.01 .60 2.01 .19 524

With Knowledge of Biographies

With Knowledge of Issue Positions

With Civics Knowledge

B (SE B)

Exp(B)

B (SE B)

Exp(B)

B (SE B)

Exp(B)

–5.62 (1.35)

.00****

–5.46 (1.30)

.00****

–5.30 (1.34)

.00****

.00 (.01) –.12 (.27) .14 (.07)

1.00 .89 1.15*

–.00 (.01) –.10 (.27) .13 (.08)

1.00 .91 1.13*

.00 (.01) –.09 (.27) .11 (.08)

1.00 .92 1.12

.83 (.93) .34 (.06) .14 (.25) –.67 (.43)

2.30 1.40**** 1.15 .51

1.14 (.92) .32 (.06) .15 (.25) –.64 (.43)

3.14 1.38**** 1.17 .53

1.30 (.91) .33 (.06) .06 (.26) –.66 (.43)

3.65 1.39**** 1.06 .52

1.08 1.02 .96 .92 1.01 .35 3.60

.09 (.07) .01 (.05) –.05 (.06) –.07 (.06) .02 (.05) –.24 (.44) .39 (.71)

1.10 1.01 .95 .93 1.02 .79 1.48

.09 (.06) .02 (.05) –.05 (.06) –.09 (.06) .02 (.05) .00 (.50) .01 (.60)

1.10 1.02 .95 .92 1.02 1.00 1.01

.08 (.06) .02 (.05) –.04 (.06) –.09 (.06) .01 (.05) –1.04 (.67) 1.28 (81) .19 522

.17 522

.17 522

355

aIn basic model, political knowledge is the combined 24-item scale of biography, issue, and civics knowledge. In second model, knowledge is candidate biography knowledge only, in third model it is knowledge of candidate issue positions only, and in fourth model, knowledge is civics knowledge only. *p < .1. **p < .05. ***p < .01. ****p < .001.

356

YOUNG

TABLE 5 Likelihood of Issuing Valid Responses to “Like Least About Bush” Among Various Political Knowledge Groups

Low knowledge Medium knowledge High knowledge Total n Chi-square

Provided Substantive Response

Skipped the Question

Said Did Not Know

26% 31% 43% 100% 510

49% 32% 19% 100% 124

54% 26% 20% 100% 46

N 220 207 253 680

43.19***

FIGURE 1 Late-night exposure increases the probability of mentioning Bush’s lack of intelligence most among the least politically knowledgeable respondents.

mentioning Bush’s lack of intelligence in response to “What do you like least about Bush?”. Although only the interaction with total political knowledge is shown here, interpreting the interactions with issue knowledge and civics knowledge shows a weaker version of the same pattern. DISCUSSION In the 2000 election, journalists, candidates, and late-night comedy hosts often discussed the potential effects of late-night joke content but without much empirical

LATE-NIGHT COMEDY AND ELECTION 2000

357

evidence to support their claims. Without a theory-driven hypothesis or a rigorous analytical procedure, these discussions remain mere speculation. This project was one attempt to move from speculation to empirical analysis by taking into consideration the unique psychological mechanisms at work in understanding political humor and positing outcomes grounded in the content of late-night jokes. The results do not provide evidence of late-night’s ability to prime caricatured traits among the general population. However, consistent with past research (Baum, 2005; Young, 2004), they do point to the possible moderating role of political knowledge. The significant trends found predicting likelihood of mentioning Bush’s lack of intelligence as a function of late-night comedy and political knowledge (Figure 1) provide some support for H2. Late-night comedy exposure was associated with increases in the salience of Bush’s most caricatured trait among those with the least political knowledge but did not strongly affect those with higher levels of knowledge. Of interest, among nonviewers of late-night comedy programming, the three knowledge groups looked quite different in their salience of Bush’s lack of intelligence, so that more knowledgeable respondents were far more likely to mention it in response to “like least about Bush” than less knowledgeable respondents. As respondents with less political knowledge increased in their exposure to late-night comedy programming, the extent to which they invoked Bush’s lack of intelligence became comparable to that of their more knowledgeable counterparts. This finding is consistent with a model in which viewing late-night comedy helped shape the impressions of Bush among those who did not have other criteria to turn to. To better understand how these mentions of Bush’s lack of intelligence arose in response to “What do you like least about Bush,” we return to the open-ended responses to explore the language used by late-night comedy viewers and respondents with various levels of political knowledge. Take, for instance, the following respondent who reported watching Leno much of the time and whose total knowledge score fell below the sample mean. When asked, “What do you like least about Bush,” he responded, “His butchering of the English language.” A female respondent with below-average political knowledge who reported watching Leno much of the time and Letterman some of the time wrote, “He really is quite a scatterbrain and not very knowledgeable in all matters he should be.” However, as illustrated in Figure 1, Bush’s lack of intelligence was not a concern reserved for lower knowledge participants who watched late-night comedy. There were 12 respondents (of the total 680) who scored in the top 20th percentile of total political knowledge who reported never watching late-night comedy programming and who cited Bush’s lack of intelligence as a concern. Comments from these respondents ranged from the straightforward (“inexperienced and unintelligent” and “inexperience in foreign affairs”) to the descriptive (“He is just a clone of Dan Quayle” and “He looks like a deer in headlights if he is ever asked a question he hasn’t been prepped for”).

358

YOUNG

According to Figure 1, respondents low in political knowledge who did not watch late-night comedy had a very low probability of mentioning Bush’s lack of intelligence in response to what they liked least about Bush. What, then, did these respondents mention when asked to identify what they liked least about him? Looking at the original texts of the open-ended responses, it seems that those nonviewing respondents with the lowest political knowledge had very little to say in response to “like least about Bush.” Many either issued no response or reported “I don’t know,” as illustrated in Table 5. Of the 46 people who said, “Don’t know,” over half were respondents who scored in the bottom third on the total knowledge scale, compared to only 20% of high-knowledge respondents. Similarly, of the 124 respondents who skipped the “like least about Bush” question, half were from the lowest knowledge group and only 19% were in the highest knowledge group. Participants with the least political knowledge were the least likely to provide a substantive response (anything other than “don’t know”) to the “like least about Bush” question (26% compared to 43% of highest knowledge respondents). However, as shown in Table 6, when looking just at respondents low in political knowledge (scoring in the bottom third), if broken down into those participants who watched late-night comedy programming and those who did not, we find that those who reported watching late-night comedy appear more likely to issue a valid response and less likely to say “don’t know” to the “like least about Bush” question than those who did not watch late-night comedy. It must be noted, though, that the small sample size makes it difficult to verify the significance of this relationship (see Table 6). In the end, the strongest predictors of the salience of candidates’ most caricatured traits were political knowledge and partisanship. Political knowledge was positively correlated with the salience of all three caricatured candidate traits predicted in the models. Meanwhile, strong Republicans were more likely to mention Gore’s dishonesty as what they liked least about him and strong Democrats—not Republicans—were more likely to mention Gore’s lack of charisma as what they liked least about him. Although the latter finding may seem counterintuitive on its TABLE 6 The Relationship Between Viewing Late-Night Comedy and Issuing a Valid Response Among Only Those Respondents Low in Political Knowledge (Scoring in the Bottom Third)

No late-night Late-night Total n Chi-square

Valid Response

Skipped the Question

Did Not Know

n

45% 55% 100% 134

52% 48% 100% 60

60% 40% 100% 25

107 112

2.03 (ns)

219

LATE-NIGHT COMEDY AND ELECTION 2000

359

face, the open-ended responses of Democrats indicate a sense of frustration with the fact that Gore’s personality might hurt his chances of winning. Some even responded to the question as though speaking to Gore’s campaign strategists: “I don’t feel him!! I mean we’re just not vibing right now. No spirit to his tone and lacks ability to move a crowd . … He should just let it flow!!!” said a strong Democrat. Another strong Democrat replied, “He hasn’t projected his true personality, he never shook the image of being a ‘stiff.’ He has to show more of the common side of himself. Connect with the regular guy.” Turning to less consistent predictors, cynical respondents were more likely to mention Gore’s dishonesty as the aspect of Gore that they liked least. This is likely due to the fact that the caricature of Gore’s saying anything to get elected (e.g., inventing the Internet) played into a cynic’s view of politics as a game. Finally, participants who watched cable news were less likely to mention Bush’s lack of intelligence as what they liked least about him. Because party identification is included in the models, the explanation that cable news viewers are more likely to be Republican is less plausible here. Although one could speculate that the Fox News network’s more conservative coverage of politics might be responsible for this finding, without a content analysis of Fox News’s coverage of Bush, it is impossible to say. Turning to the proposed distinctions between the various forms of political knowledge, with background and issue knowledge thought to mitigate late-night comedy viewing’s effects more than civics knowledge, these findings suggest quite the opposite. It was civics knowledge that exerted the strongest interactive effect with late-night comedy, not background and issue knowledge as hypothesized. Background knowledge alone had no interactive effect in the models. Civics knowledge, measured here as knowledge about government structure and processes, can be thought of as political knowledge held in long-term memory. Some have used this measure as a control variable when assessing the effects of media use on knowledge acquisition (Mondak, 1995). Others have suggested that such general political knowledge represents the development and complexity of political schemata in memory (Fiske. Lau, & Smith, 1990; Hsu & Price, 1993). Price and Zaller (1993) concluded that such general political knowledge is a stronger predictor of news reception than actual news viewing itself. Why would civics knowledge have the strongest interactive effect with latenight comedy viewing when predicting the salience of late-night comedy caricatures? The original hypothesis posited that knowledge of candidate issue positions and backgrounds would serve as alternative judgment criteria and hence would mitigate the enhanced accessibility of candidate caricatures hypothesized to occur as a result of viewing late-night comedy. However, H1 was not borne out in the models. Viewing late-night comedy shows did not foster the salience of the candidates’ most caricatured traits. Instead, it was political knowledge and party identification that were most associated with the salience of Bush’s lack of intelligence,

360

YOUNG

Gore’s dishonesty, and Gore’s lack of charisma. Rather than illustrating a mitigating effect, perhaps the significant interaction of civics knowledge and late-night comedy viewing along with the patterns shown in Tables 6 and 7 illustrate a “catching up” effect in which late-night comedy viewing provided those with the least complex understanding of general politics something to turn to when asked what they liked least about George W. Bush. Responding to a comment by Bill Moyers in an interview on PBS, Jon Stewart of Comedy Central’s The Daily Show stated, “It’s interesting. I think we don’t make things up. We just distill it to, hopefully, its most humorous nugget” (Siceloff, 2003). The political jokes that made their way into Leno and Letterman’s monologues in the 2000 election were not purely fabricated. If they were, the humor reconciliation mechanism would never have worked because no one would have been able to contribute the necessary information to make sense of them. Instead, the jokes came directly from the news headlines, which were then distilled to their “most humorous nuggets.” Perhaps these nuggets brought something to the top of the minds of less politically sophisticated viewers that otherwise would not have been there. Although late-night comedy viewing may have helped some apolitical audiences shape their impressions of the candidates, there is one important aspect of political jokes that has not been addressed. Jokes made by late-night comedy hosts are unequivocally critical of the candidates. They mock not the most flattering features of public officials but rather their most prominent weaknesses. As Jamieson and Waldman (2003) observed, The stance of late-night humor is fundamentally cynical; each politician is defined only by his or her most glaring weaknesses, and the system produces only venal, corrupt candidates unfit for the office. The implication? Neither candidate is actually worthy of high public office. (p. 68)

These analyses predicted whether respondents mentioned late-night comedy candidate caricatures in response to “What do you like least about candidate x.” This method was chosen because the content of late-night jokes was overwhelmingly negative toward the candidates and rarely featured traits that one would like most about a candidate. Using this method, the findings suggest that late-night comedy may have the ability (albeit limited) to render some negative candidate traits more salient among less knowledgeable viewers so that they begin to look like more knowledgeable individuals. However, it leaves open the question of how late-night comedy content helps less informed viewers construct positive images of the candidates. Perhaps candidate interviews on late-night comedy programs offer candidates the opportunity to combat their “most glaring weaknesses” that dominate the hosts’ monologues (see Baum, 2005). Future studies ought to explore

LATE-NIGHT COMEDY AND ELECTION 2000

361

the content and tone of candidate interviews on late-night comedy programs as well as late-night comedy’s relationship with cynicism.

LIMITATIONS AND FUTURE RESEARCH As the late-night comedy exposure and salience measures were obtained at the same time, these data must be treated as cross-sectional. In addition, these data were obtained from one wave of a larger panel study that involved the discussion of politics and multiple politically oriented surveys. As attrition from the project was associated with political interest and expertise, the generalizability of these findings is not clear. As these analyses found a stronger association between late-night comedy viewing and trait salience among low knowledge viewers, perhaps these effects would be stronger with a less politically sophisticated sample. Future research ought to test the humor elaboration mechanism in a controlled experimental setting where salience is measured either through reaction time or thought listing exercises. Studies designed to assess the effects of late-night comedy content ought to consider constructing reception measures that account for attention paid to political content in the late-night comedy programs. Such studies would allow a comparison between Baum’s (2003) notion of incidental exposure to political content in entertainment programming and more intentional exposure as operationalized by Prior (2003). If late-night’s potential rests in its ability to communicate the themes of traditional political information to apolitical audiences in “distilled humorous nuggets,” then perhaps incidental exposure is most relevant.

ACKNOWLEDGMENTS An earlier version of this article was presented in a poster session at the annual meeting of the American Political Science Association, Philadelphia, PA (August 28–31, 2003). The Electronic Dialogue 2000 Project, from which individual-level data were obtained, was supported by grants to Dr. Joseph Cappella and Dr. Vincent Price from the Pew Charitable Trusts and the Annenberg Public Policy Center of the University of Pennsylvania. The views expressed here are those of the author alone and do not reflect the views of the funding agencies or of the organizations from which data were obtained. I express thanks to CMPA’s Research Director, Dan Amundson, as well as Kimberly Meltzer, Klaus Krippendorff, Joseph Cappella, and Vincent Price. Any errors or oversights are my responsibility alone.

362

YOUNG

REFERENCES Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press. Bargh, J. A., Lombardi, W. J., & Higgins, E. T. (1988). Automaticity of chronically accessible constructs in person X situation effects on person perception: It’s just a matter of time. Journal of Personality and Social Psychology, 55, 599–605. Baum, M. A. (2003). Soft news and political knowedge: Evidence of absence or absence of evidence? Political Communication, 20, 173–190. Baum, M. A. (2005). Talking the vote: What happens when presidential candidates hit the talk show circuit? American Journal of Political Science, 49, 213–234. Benoit, W. L., & Hansen, G. J. (2004). Presidential debate watching, issue knowledge, character evaluation, and vote choice. Human Communication Research, 30, 121–144. Brians, C. L., & Wattenberg, M. P. (1996). Campaign issue knowledge and salience: Comparing reception from TV commercials, TV news, and newspapers. American Journal of Political Science, 40, 172–193. Burnett, R., Gaines, B., & Pope, M. (Executive Producers). (2000a, March 1). The Late Show with David Letterman [Television broadcast with John McCain]. New York: Columbia Broadcasting System. Burnett, R., Gaines, B., & Pope, M. (Executive Producers). (2000b, September 14). The Late Show with David Letterman [Television broadcast with Al Gore]. New York: Columbia Broadcasting System. Burnett, R., Gaines, B., & Pope, M. (Executive Producers). (2000c, October 9). The Late Show with David Letterman. New York: Columbia Broadcasting System. Burnett, R., Gaines, B., & Pope, M. (Executive Producers). (2000d, October 19). The Late Show with David Letterman [Television broadcast with George W. Bush]. New York: Columbia Broadcasting System. Cappella, J. N., & Jamieson, K. H. (1997). Spiral of cynicism: The press and the public good. New York: Oxford University Press. Chaffee, S. H., Zhao, X., & Leshner, G. (1994). Political knowledge and the campaign media of 1992. Communication Research, 21, 305–324. Clarke, P., & Fredin, E. (1978). Newspapers, television, and political reasoning. Public Opinion Quarterly, 42, 143–160. Cohen, B. (1963). The press and foreign policy. Princeton, NJ: Princeton University Press. Collins, A. M., & Loftus, E. (1975). A spreading activation theory of semantic processing. Psychological Review, 82, 407–428. Collins, A. M., & Quillian, M. R. (1969). Retrieval time from semantic memory. Journal of Verbal Learning and Verbal Behavior, 8, 240–248. Davis, R., & Owen, D. (1998). New media and American politics. New York: Oxford University Press. Delli Carpini, M., & Keeter, S. (1996). What Americans don’t know about politics and why it matters. New Haven, CT: Yale University Press. Fiske, S. T., Lau, R. R., & Smith, R. A. (1990). On the varieties and utilities of political expertise. Social Cognition, 8, 31–48. Higgins, E. T. (1989). Knowledge accessibility and activation: Subjectivity and suffering from unconscious sources. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 75–123). New York: Guilford. Holbert, R. L., Benoit, W. L., Hansen, G. J., & Wen, W. (2002). The role of communication in the formation of an issue-based citizenry. Communication Monographs, 69, 296–310. Holbert, R. L., Pillion, W., Tschida, D. A., Armfield, G. G., Kinder, K., Cherry, D. L., et al. (2003). The West Wing as endorsement of the U. S. presidency: Expanding the bounds of priming in political communication. Journal of Communication, 53, 427–447. Holbrook, R. A., & Hill. T. G. (2005). Agenda-setting and priming in prime time television: Crime dramas as political cues. Political Communication, 22, 277–295.

LATE-NIGHT COMEDY AND ELECTION 2000

363

Hollander, B. (1994). The new news and the 1992 presidential campaign: Perceived v. actual political knowledge. Journalism and Mass Communication Quarterly, 72, 786–798. Hsu, M., & Price, V. (1993). Political expertise and affect. Communication Research, 20, 671–695. Iyengar, S., & Kinder, D. R. (1987). News that matters: Television and American opinion. Chicago: University of Chicago Press. Jamieson, K. H., & Waldman, P. (2003). The press effect: Politicians, journalists, and the stories that shape the political world. New York: Oxford. Keeter, S., Zukin, C., Andolina, M., & Jenkins, K. (2002, September 19). The civic and political health of the nation: A generational portrait. College Park, MD: Center for Information and Research on Civic Learning and Engagement. Retrieved February 1, 2006, from http://www.pewtrusts.com/ pdf/public_policy_youth_civic_political_health.pdf Koestler, A. (1964). The act of creation. London: Hutchinson. Kurtz, H. (Host). (2000, April 8). Jon Stewart zings the fourth estate. Reliable Sources [Television broadcast]. Atlanta, GA: CNN. McGuire, W. J. (1968). Personality and susceptibility to social influence. In E. F. Borgatta & W. W. Lambert (Eds.), Handbook of personality theory and research (pp. 1130–1187). Chicago: Rand McNally. McGuire, W. J. (1972). Attitude change: The information-processing paradigm. In C. G. McClintock (Ed.), Experimental social psychology (pp. 108–141). New York: Holt, Rinehart, and Winston. McLeod, J. M., Guo, Z., Daily, K., Steele, C. A., Huang, H., Horowitz, E., et al. (1996). The impact of traditional and nontraditional media forms in the 1992 presidential election. Journalism and Mass Communication Quarterly, 73, 401–416. Michaels, L. (Executive Producer). (2000, November 5). Saturday Night Live: Presidential Bash 2000 [Television broadcast]. New York: National Broadcasting Company. Mondak, J. J. (1995). Newspapers and political awareness. American Journal of Political Science, 39, 513–528. Moy, P., Xenos, M. A., & Hess, V. K. (2006). Priming effects of late-night comedy. International Journal of Public Opinion Research. Forthcoming. National Annenberg Election Survey. (2004, 21 September). Daily Show viewers knowledgeable about presidential campaign [Online report]. Retrieved February 1, 2006, from http://www.annenbergpublicpolicycenter. org/naes/2004_03_late-night-knowledge-29-21_pr.pdf Niven, D., Lichter, S. R., & Amundson, D. (2003). The political content of late night comedy. Press/Politics, 8, 118–133. Patterson, T. W., & McClure, R. D. (1976). The unseeing eye: The myth of television power in national politics. New York: Putnam. Pew Research Center for the People and the Press. (2000, February). Audiences fragmented and skeptical: The tough job of communicating with voters [Online report]. Retrieived February 1, 2006, from http://www. people-press.org/jan00rpt2.htm Pew Research Center for the People and the Press. (2002, June). Public’s news habits little changed since September 11 [Online report]. Retrieved February 1, 2006, from http://people-press.org/reports/display. php3?ReportID=156 Pew Research Center for the People and the Press. (2004, 11 January). Cable and Internet loom large in fragmented political news universe [Online report]. Retrieved February 1, 2006, from http://people-press.org/ reports/display.php3?ReportID=200 Pfau, M., Cho, J., & Chong, K. (2001). Communication forms in U.S. presidential campaigns: Influences on candidate perceptions and the democratic process. Press/Politics, 6, 88–105. Price, V. (1999). Political information. In J. P. Robinson, P. R. Shaver, and L. S. Wrightsman (Eds.). Measures of political attitudes (pp. 591–639). San Diego: Academic Press. Price, V., & Tewksbury, D. (1997). News values and public opinion: A theoretical account of media priming and framing. In G. A. Barnett & F. J. Boster (Eds.), Progress in the communication sciences (pp. 173–212). New York: Ablex.

364

YOUNG

Price, V., & Zaller, J. (1993). Who gets the news? Alternative measures of news reception and their implications for research. Public Opinion Quarterly, 57, 133–164. Prior, M. (2003). Any good news in soft news? The impact of soft news preference on political knowledge. Political Communication, 20, 149–171. Roskos-Ewoldsen, D. R., Roskos-Ewoldsen, B., & Carpentier, F. R. D. (2002). Media priming: A synthesis. In J. Bryant & D. Zillman (Eds.), Media effects: Advances in theory and research (2nd ed., pp. 97–120). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Scheufele, D. A. (2000). Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication and Society, 3, 297–316. Schmidt, S. R. (1994). Effects of humor on sentence memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 20, 953–967. Schmidt, S. R. (2001). Memory for humorous cartoons. Memory and Cognition, 29, 305–311. Sella, M. (2000, September 24). The stiff guy vs. the dumb guy. The New York Times Magazine, p. 72–80. Siceloff, J. L. (Executive Producer). (2003, July 11). NOW with Bill Moyers. New York: PBS. Srull, T. (1981). Person memory: Some tests of associative storage and retrieval models. Journal of Experimental Psychology: Human Learning and Memory, 7, 440–463. Suls, J. (1972). A two-stage model for the appreciation of jokes and cartoons: An information-processing analysis. In J. H. Goldstein & P. E. McGhee (Eds.), Psychology of humor (pp. 81–100). New York: Academic. Vickers, D. (Executive Producer). (2000a, March 1). The Tonight Show with Jay Leno [Television broadcast with George W. Bush]. New York: National Broadcasting Company. Vickers, D. (Executive Producer). (2000b, March 6). The Tonight Show with Jay Leno [Television broadcast with George W. Bush]. New York: National Broadcasting Company. Vickers, D. (Executive Producer). (2000c, August 21). The Tonight Show with Jay Leno. New York: National Broadcasting Company. Vickers, D. (Executive Producer). (2000d, September 19). The Tonight Show with Jay Leno [Television broadcast with Al Gore]. New York: National Broadcasting Company. Vickers, D. (Executive Producer). (2000e, October 3). The Tonight Show with Jay Leno. New York: National Broadcasting Company. Vickers, D. (Executive Producer). (2000f, October 16). The Tonight Show with Jay Leno. New York: National Broadcasting Company Wyer, R. S., & Collins, J. E. (1992). A theory of humour elicitation. Psychological Review, 99, 663–688. Young, D. G. (2004). Late-night comedy in election 2000: Its influence on candidate trait ratings and the moderating effects of political knowledge. Journal of Broadcasting and Electronic Media, 48, 1–22. Zaller, J. (1992). The nature and origins of mass opinion. New York: Cambridge University Press.

APPENDIX A POLITICAL KNOWLEDGE Candidate Biographies: 1) Thinking about the Democrats , to the best of your knowledge who a) Was a professional basketball player (Bradley); b) Is the son of a former United States Senator (Gore); c) Voted for tax cuts proposed by President Reagan in 1981 (Bradley); d) Served in the United States Senate (Both). 2) Thinking about the Republicans (John McCain and George W. Bush), to the best of

LATE-NIGHT COMEDY AND ELECTION 2000

365

your knowledge, who a) Is a state governor (Bush); b) Is a United States Senator (McCain); c) Was a prisoner of war in Vietnam (McCain). Candidate Issue Positions: 1) Thinking about Democrats, to the best of your knowledge, who a) Supports a universal health care program (Bradley); b) Favors increased government funding of political campaigns (Bradley); c) Favors giving patients the right to sue their HMO (Both); d) Favors tax-free savings accounts to help parents pay for college (Gore). 2) Thinking about Republicans, to the best of your knowledge, who a) Supports giving tax credits or vouchers to people who send their children to private schools (both); b) Has pledged to cut federal income taxes by over $1 trillion in ten years (Bush); c) Supports a ban on soft money campaign contributions (McCain). Civics Knowledge: 1) Which one of the parties is more conservative than the other at the national level? [Democrats, Republicans, DK] 2) Which one of the parties has the most members in the House of Representatives in Washington? [Democrats, Republicans, DK] 3) Which one of the parties has the most members in the U.S. Senate? [Democrats, Republicans, DK] 4) Who has the final responsibility to decide if a law is Constitutional or not? [President, Congress, Supreme Court, DK] 5) Which one of the following is the main duty of Congress? [Write legislation; Administer the President’s policies; Watch over the governments of each state; DK] 6) Whose responsibility is it to nominate judges to the Federal Courts? [President, Congress, Supreme Court, DK] 7) How much of a majority is needed for the U.S. Senate and House to override a presidential veto? [Bare majority (one more than half the votes), Two-thirds majority, Three-fourths majority, DK] 8) Do you happen to know what job or political office is currently held by Al Gore? [U.S. Senator, U.S. Vice President, Governor of Tennessee, DK] 9) What job or political office is currently held by Trent Lott? [U.S. Senator, U.S. Ambassador to the United Nations, Chief Justice of the U.S. Supreme Court, DK] 10) What job or political office is currently held by William Rehnquist? [U.S. Senator, U.S. Ambassador to the United Nations, Chief Justice of the U.S. Supreme Court, DK]

APPENDIX B SAMPLE JOKE TEXTS FROM LATE-NIGHT COMEDY PROGRAMS Jay Leno on Gore’s stiff appearance and dull personality (August 21, 2000): “Did you watch Gore’s [convention] speech? I thought he did a nice job. You know, I think we saw some changes. I think we saw a change from a stiff, wooden Al Gore into a softer, cardboard Al Gore” (Vickers, 2000c). Jay Leno on Bush’s lack of intelligence (October 3, 2000): “Bush said that he is going to use the same strategy [in the debates] that he used to get through Yale Uni-

366

YOUNG

versity. See, once the debate starts, he is going to push his podium real close to Gore, then cheat off of his answers” (Vickers, 2000e). David Letterman on Gore’s tendency to exaggerate or lie (October 9, 2000): “Al Gore is very excited for round number two in the debates. He’s already prepared five brand new made-up stories, so he’s ready to go. Got five big fibs he’s dying to tell.” (Burnett, Gaines, & Pope, 2000c). Jay Leno on Bush’s lack of knowledge and Gore’s tendency to exaggerate (October 16, 2000): “The final debate for Bush and Gore is on Tuesday. They both have a gentleman’s agreement for this one – If Bush forgets an important fact, Gore will make one up for him” (Vickers, 2000f).