Political Communication, 25:401–422, 2008 Copyright © Taylor & Francis Group, LLC ISSN: 1058-4609 print / 1091-7675 online DOI: 10.1080/10584600802427013

Late-Night Comedy as a Gateway to Traditional News: An Analysis of Time Trends in News Attention Among Late-Night Comedy Viewers During the 2004 Presidential Primaries

1091-7675 1058-4609 UPCP Political Communication, Communication Vol. 25, No. 4, September 2008: pp. 1–48

LAUREN FELDMAN and DANNAGAL GOLDTHWAITE YOUNG

ComedyFeldman Lauren as a Gateway and Dannagal Goldthwaite Young

Recent reports published by the Pew Research Center for the People and the Press (2000, 2004) propose that young audiences are abandoning traditional news as a source of election information in favor of late-night comedy programs. However, additional evidence (Young & Tisinger, 2006) suggests that exposure to late-night comedy programming is positively correlated with traditional news exposure. This study extends this body of research by offering evidence that exposure to late-night comedy is associated with increases in attention paid to the presidential campaign in national network and cable news. The analysis uses data collected via the National Annenberg Election Survey during the 2004 presidential primary season, between October 30, 2003, and June 4, 2004. Cross-sectional results demonstrate that viewers of late-night comedy programs—specifically viewers of The Tonight Show with Jay Leno and The Late Show with David Letterman, as well as Comedy Central’s The Daily Show with Jon Stewart—pay more attention to the campaign in national network and cable news than nonviewers, controlling for a variety of factors. An analysis of time trends also reveals that the rate of increase in news attention over the course of the primary season is greater for viewers of Leno or Letterman than for those who do not watch any late-night comedy. Keywords late-night comedy, satire, news attraction, gateway hypothesis

Over the past decade, journalists and academics have begun to embrace the notion that entertainment-oriented programming and texts might play important roles in the political environment. Studies of shows such as the West Wing, NYPD Blue, and Oprah Winfrey reveal that just as traditional news programs can persuade, teach, prime, frame, and alter political efficacy and trust, so too can entertainment programs (Baum, 2005; Holbert et al., 2003; Holbrook & Hill, 2005). While producers of entertainment shows generally deny any intention of fostering political outcomes (Gay, 2004; “Ten Questions,” 2004; Colapinto, 2004), these subtle latent effects have been demonstrated repeatedly. One form of entertainment programming that has received particular attention from media effects scholars is late-night comedy. Political communication researchers have examined the impact of programs such as those hosted by Jay Leno and David Letterman Lauren Feldman is Assistant Professor in the School of Communication at American University. Dannagal Goldthwaite Young is Assistant Professor of Communication at the University of Delaware. Address correspondence to Lauren Feldman, School of Communication, American University, 4400 Massachusetts Avenue, NW, Washington, DC 20016, USA. E-mail: [email protected]

401

402

Lauren Feldman and Dannagal Goldthwaite Young

on political attitudes, knowledge, and behaviors. For example, several studies demonstrate that exposure to late-night comedy during an election campaign can shape viewers’ perceptions of the candidates (Moy, Xenos, & Hess, 2006; Pfau, Cho, & Chong, 2001; Young 2004b, 2006), while others indicate effects on broader constructs such as political trust (Baumgartner & Morris, 2006). The interplay between late-night comedy viewing and traditional television news consumption has also attracted scholarly interest (e.g., Holbert, Lambe, Dudo, & Carlton, 2007). Reports by the Pew Center for the People and the Press from 2000 and 2004 identified young people as the most likely to report high rates of learning from late-night comedy programs and the least likely to report learning from network news and newspapers. During the 4 years between the two surveys, the gap grew, with young people claiming to learn less from news and more from comedy programs over time. For many, this inverse relationship was interpreted as evidence of young people’s “desertion” of traditional news for late-night comedy shows. Further contributing to this narrative is Mindich’s recent research (2005), which documents young people’s growing disillusionment with traditional news, particularly television news. As argued by Young and Tisinger (2006), while the Pew data certainly point to a macro-level trend in which young people as a demographic are tuning out news and tuning into late-night comedy, they do not establish that the same individuals who are not watching news are watching late-night. In fact, the authors use two different data sources to illustrate precisely the opposite pattern. Those young people who reported the highest rates of exposure to and learning from late-night comedy programs were the same individuals who reported the highest rates of exposure to more traditional forms of news. In other words, the narrative of young people tuning into late-night comedy instead of news is more myth than reality. This observation should not come as much of a surprise; various forms of communication use have long been theorized to be complementary rather than competitive in nature (Chaffee & McLeod, 1972; Chaffee, 1986). More than three decades ago, Chaffee and McLeod (1972) argued that the relationship between interpersonal and mass communication is both cyclical and reciprocal, with “one principal effect of communication . . . further communication” (p. 237). More recently, Holbert (2005a) extended this logic to apply to different types of mass communication. While consistent with the notion that communication breeds more communication, the Young and Tisinger study still does not speak to causality. That is, are young people who tune into traditional news programs subsequently more likely to watch late-night comedy programs? Or is the correlation between late-night comedy and news consumption illustrative of Baum’s (2003b) “gateway” hypothesis, such that exposure to entertainment programming that contains political content sparks an interest in otherwise apolitical viewers to start paying attention to the news? The present research addresses these causal considerations. Using data from the 2004 National Annenberg Election Survey (NAES), the current study compares trends in television news attention over the course of the 2004 presidential primary campaign among three subgroups of respondents: those who do not watch latenight comedy, those who watch late-night comedy at least once per week and report Comedy Central’s The Daily Show with Jon Stewart as their most viewed program, and individuals who watch late-night comedy and report a program other than The Daily Show, such as The Tonight Show with Jay Leno or The Late Show with David Letterman, as their most viewed. By taking advantage of the NAES’s time series component, conclusions can be drawn about whether viewing various forms of late-night comedy increases the rate at which people attend to the campaign on traditional television news programs.

Comedy as a Gateway

403

Late-Night as a Gateway to Traditional News According to Baum (2003a, 2003b, 2005), because soft news programs such as talk shows, infotainment programs, and late-night comedy “piggyback” political information on top of entertainment programming, even the most apolitical viewers receive at least some exposure to political topics. For this reason, Baum hypothesizes that exposure to soft news will increase attentiveness to political issues, hence reducing the power of political engagement in models predicting attentiveness to high-profile political stories. The mechanism via which attentiveness will increase among apolitical viewers, posits Baum (2003b), is through soft news shows functioning as “gateways” to traditional news. Baum’s data supporting the gateway effect were obtained in the news context of foreign crises. He found that an increase in exposure to talk shows and other entertainment-oriented news programs—programs where viewers were likely to encounter news of foreign crises—was associated with a subsequent increase in consumption of traditional news. Here, soft news programs appear to have provided viewers with a cursory understanding of foreign crises, motivating them to attend to coverage of this topic in traditional television news broadcasts. This process is not unlike that outlined by early political socialization researchers, who conceptualized television news as a “bridge” for young people and other political novices to learn about the political world (Chaffee & Yang, 1988). However, in the contemporary media environment, it is talk shows and late-night comedy that may now serve as segues to television news. One important context in which various late-night comedy programs could foster attention to traditional news is during a presidential campaign—particularly during the primaries, when the potential for growth in learning, attention, and interest in the campaign is ostensibly the highest (Kennamer & Chaffee, 1982). Late-night comedy hosts such as Leno and Letterman cover the political campaign by joking about scandals, mocking campaign events, caricaturing political candidates, and interviewing the presidential hopefuls. Meanwhile, The Daily Show includes candidate interviews as well—but also satirizes the campaign process, discusses policies and issues that are at the core of the campaign, and parodies the journalistic practices that traditionally deliver campaign news to us. At a time in the campaign when it seems that only a few savvy New Hampshirites and Iowans are following the candidates (Dutwin, 2000), late-night comedians are already integrating stories of the campaign into their monologues and headlines. Late-night comedy might also be an effective gateway to traditional news attentiveness because some degree of knowledge about public affairs is necessary to appreciate the shows’ topical humor. As Jon Stewart has argued, “If [kids] came to our show without knowledge, it wouldn’t make any sense to them” (C-Span Newhouse School Forum, 2004). Although some Daily Show viewers may tune into the program already informed about the issues of the day (Young & Tisinger, 2006), for other viewers of late-night comedy, particularly viewers of Leno or Letterman, a desire to understand the political humor—to be in on the joke—may compel further information seeking from traditional news. Indeed, just as many young people report that they are motivated to watch the news because of its utility for conversations with peers, family, and coworkers (Drew & Reeves, 1980; Mindich, 2005), they may likewise attend to traditional news programs so that they can take part in the “conversation” of late-night comedy.

Conceptualizing Late-Night Comedy An understanding of the potential influence of late-night comedy on traditional news attentiveness requires a clear delineation of the political content found in various forms

404

Lauren Feldman and Dannagal Goldthwaite Young

of late-night programming. The Tonight Show, The Late Show, and The Daily Show all include political humor and political satire in the context of monologues, headlines, and other show segments. Likewise, all offer a platform for political figures and candidates to be interviewed by late-night hosts and, in so doing, gain exposure to a diverse viewing audience. In this way, late-night comedy’s content reflects two content categories represented in Holbert’s (2005b) entertainment and politics typology. Satirical segments from The Daily Show or from Leno or Letterman’s monologues, in which some aspect of policy or a politician’s personality or integrity is implicitly challenged, would constitute a form of traditional satire. Meanwhile, the politician and candidate interviews airing on these same shows offer a more explicit piece of political content, one in which the politicians’ words are less filtered and the meaning of the text is less enthymematic than in the context of satire (see Holbert, 2005b). According to Young (2004b) and Niven, Lichter, and Amundson (2003), late-night jokes in the monologues of Leno and Letterman focus on the executive branch of government and emphasize politicians’ personal failings. More specifically, in a study of latenight jokes from the 2004 presidential primaries, Young (2004a) documented that almost half of the jokes made about Howard Dean focused on his likely loss of the election, particularly following the post-Iowa speech in which he “screamed” a little too enthusiastically. In the case of John Kerry, the focus of jokes during the primaries was split between his rich wife, his appearance, and his pandering to voters. In spite of their emphasis on personal failings of candidates, the transcripts of these jokes reveal that their lead-ins often summarize news stories about political events. For example, during the 2004 presidential primaries, Leno and Letterman made numerous jokes mentioning the Iowa caucuses. These hosts typically used the caucuses as a backdrop for jokes highlighting personal and physical flaws of the candidates. For example, Leno portrayed Senator John Edwards as an ambulance-chasing trial lawyer: Democratic front-runner Howard Dean is being called a hero this week. It seems he was going to a campaign rally in Iowa and one of his campaign volunteers collapsed from a seizure and since he was a doctor he ran over and treated the guy until the ambulance came. He is not the only candidate that helped; since he was a trial lawyer, John Edwards chased the ambulance all the way to the hospital. (Leno, Nov. 14, 2003)1 Likewise, David Letterman joked about Governor Howard Dean’s short stature: The frontrunner of course is Howard Dean. People may be saying he might be too short to be president. Earlier today he was campaigning at a mall in Iowa and security kept coming up to him asking if he was lost. (Letterman, Jan. 16, 2004) By far the most popular category of jokes using Iowa as a backdrop included those mocking Howard Dean’s infamous speech the night that he lost the caucuses to John Kerry: Did you see Dean’s speech last night? Oh my God! Now I hear the cows in Iowa are afraid of getting mad Dean disease. (Leno, Jan. 20, 2004) The Iowa caucuses were also featured prominently on The Daily Show. However, while Leno and Letterman concentrated on readily accessible candidate caricatures,

Comedy as a Gateway

405

Jon Stewart extrapolated on the structure, process, and results of the caucuses while mocking the spectacle and the pivotal role of the press. In this short 100-word segment from The Daily Show’s headlines, for example, Stewart successfully explained the caucus process while pointing out that “candidate electability” is a media-generated phenomenon: Next Monday brings one of the most heavily scrutinized moments of the entire campaign, the Iowa caucuses. Now the caucuses are not a primary, but rather a series of town hall-style meetings, and, it’s interesting to note, the attendees discuss the candidates, then appoint delegates to vote for them, in district and state conventions where—I’m sorry, I’m being told it was NOT interesting to note that . . . . The big event is nearly a week away, but with no votes cast yet, Vermont governor Howard Dean is still the candidate to beat—because that’s what we’ve been told. (Daily Show, Comedy Central, Jan. 13, 2004) In addition to the political jokes and satire included in these programs, during the 2004 primaries and general election campaign, late-night comedy served as a venue for candidates to showcase a less formal side of their personalities. John Kerry rode a motorcycle onto the set of The Tonight Show with Jay Leno and poked fun at his often-mocked wealthy wife in Letterman’s Top Ten list. Kerry’s first television interview after the airing of the controversial anti-Kerry Swiftboat ads was with Jon Stewart on The Daily Show— the same program on which John Edwards announced his presidential candidacy. Such examples illustrate how exposure to various forms of late-night comedy might increase the salience of a political campaign among less politically engaged viewers. Based on Baum’s gateway hypothesis (2003b), by bringing political candidates or political events such as the Iowa caucuses to the attention of viewers, late-night programs could serve as an impetus for these audience members to pay more attention to traditional news coverage of candidates and events. Still, as might have been suggested by the opposing character of the Iowa-related jokes referenced above, late-night comedy is not a singular genre. Despite some common features, there are systematic differences in content and function between the overtly political The Daily Show and the eclectic fare offered on the shows hosted by Leno and Letterman. For example, data from the 2004 NAES content analysis of latenight jokes illustrated that Stewart was more likely than Leno or Letterman to mention issues or policies in his segments and was less likely than the other late-night hosts to mock personality traits of Bush or Kerry, suggesting that The Daily Show is a more substantive, issue-oriented program than the network late-night shows (NAES, 2004). According to such scholars as Baym (2005) and Young (2007), The Daily Show is better described as an alternative form of journalism than as “fake” news or pure entertainment. Recent content and textual analyses comparing broadcast news and The Daily Show during the 2004 election demonstrated that Stewart’s parody of traditional news was as substantive as the broadcast news programs he satirizes on the show (Fox, Koloen, & Sahin, 2007; Jones, 2007). Moreover, election stories on The Daily Show were, on average, longer than election stories on broadcast news (Fox et al., 2007). Perhaps reflective of these content differences, the various programs appear to draw distinct audiences. With about 1.5 million nightly viewers (Associated Press, 2006), The Daily Show attracts a smaller, more select group relative to the

406

Lauren Feldman and Dannagal Goldthwaite Young

approximately 6 million who tune into Leno and 4 million watching Letterman each evening (Stanley, 2006). Indeed, Young and Tisinger (2006), in their analysis of younger viewers of late-night programs, found that a preference for The Daily Show was positively associated with political knowledge and interest, as well as more liberal and democratic political orientations. Preferences for Leno and Letterman, on the other hand, were related only to local news viewing among a range of variables measuring political orientation and engagement. Late-night audiences were also found to differ on basic demographic characteristics, with Daily Show viewers more likely to be younger and male, and Letterman viewers being slightly more educated. These content and audience differences underscore the need to separate The Daily Show from the programs of Leno and Letterman in any consideration of late-night comedy’s effects.

Hypotheses and Research Questions Recent research that has informed our understanding of the gateway hypothesis in the context of late-night comedy viewing has dealt exclusively with exposure to traditional news as a correlate of late-night comedy viewing (Pew Center, 2000, 2004; Young & Tisinger, 2006). In Baum’s (2003b) original “gateway” research, however, he proposed that exposure to entertainment content that piggybacked political information on top of its softer fare would foster subsequent attention paid to political issues in more traditional news forms (p. 111). In keeping with this original conceptualization, we also examine the effects of exposure to specific late-night comedy programs on attention to the campaign in traditional news outlets. In so doing, we emphasize not the quantity of traditional news consumed, but rather a substantively important form of engagement with traditional news—one that emphasizes the quality of the viewing experience and the specific aspects of content attended to (e.g., campaign information). Thus, based on the gateway hypothesis (Baum, 2003b) and the capacity of latenight jokes to expose viewers to campaign stories and events, there ought to be a positive association between exposure to late-night comedy programming and attention to the campaign in traditional news. Moreover, since the gateway hypothesis implies that late-night comedy viewing is not only associated with attention to news, but actually fosters that attention, we expect that, over the course of the campaign, news attention among viewers of late-night comedy programs will accelerate at a faster rate than those who do not watch late-night. Although the momentum of the campaign should cause attention to campaign news to increase across the board, if the gateway effect is operating, acceleration should be greater for viewers of late-night comedy. However, as detailed above, previous research has identified key differences in both the audience and content of The Daily Show relative to the programs of Leno and Letterman. For this reason, our hypotheses distinguish between viewers of The Daily Show and viewers of Leno or Letterman rather than treating the late-night comedy audience as a singular entity. We thus offer two sets of hypotheses to test in the context of the 2004 presidential primary campaign. The first set of hypotheses speaks to the correlation been viewing particular late-night programs and paying attention to stories about the campaign in traditional news, while the second set addresses the rate of increase in attention to campaign news over time among viewers of specific late-night shows versus those who do not watch late-night programming.

Comedy as a Gateway

407

Hypothesis 1a: Viewers of The Tonight Show with Jay Leno or The Late Show with David Letterman will pay more attention to the campaign on television news than those who do not watch late-night comedy. Hypothesis 1b: Viewers of The Daily Show will pay more attention to the campaign on television news than those who do not watch late-night comedy. Hypothesis 2a: Over time, the rate of increase in attention to the campaign on television news will be greater for viewers of The Tonight Show with Jay Leno or The Late Show with David Letterman than those who do not watch late-night comedy. Hypothesis 2b: Over time, the rate of increase in attention to the campaign on television news will be greater for viewers of The Daily Show than those who do not watch late-night comedy. Although we hypothesize that the gateway mechanism will operate for audiences of both The Daily Show and Leno/Letterman, it is unlikely that the over-time patterns of news attention will be identical for both subgroups. However, because the nature of the differences in news attentiveness among these two groups is as yet unclear, we pose a research question: Research Question 1: Will the trend in traditional news attention over the course of the campaign differ between audiences of The Daily Show and audiences of The Tonight Show with Jay Leno or The Late Show with David Letterman?

Method Data used in this study come from the 2004 NAES, conducted by the Annenberg Public Policy Center at the University of Pennsylvania. The 2004 NAES employed a rolling cross-sectional design, a design that involves sampling a series of random cross-sections of the target population over a period of time (Johnston & Brady, 2002). With the NAES design, cross-sections were taken on a daily basis. A random sample of telephone numbers was first selected using random digit dialing (RDD). These numbers were then divided into random subsamples, or replicates, which were released for telephone interviewing each night. The logic of the NAES rolling cross-sectional design ensures that on a given night, the sample of numbers interviewed will be comparable to the sample interviewed on any other night (see Kenski, 2006a). The date of the interview becomes a random event independent of sample characteristics such as gender, age, education, and so forth. The 2004 NAES assessed the political attitudes, behavior, knowledge, and communication patterns of Americans over the age of 18 and was conducted between October 7, 2003, and November 16, 2004. It yielded an overall response rate (using the American Association for Public Opinion Research’s RR1) of 22%. The present analyses make use only of NAES survey data collected during the presidential primary season, between October 30, 2003, and June 4, 2004. During the primary campaign, an average of nearly 200 interviews were completed daily, yielding a total sample size during this period of 38,271. Measurement of Key Variables Late-Night Viewing. Two questions on the NAES were used to identify three audience subgroups. Respondents were asked “How many days in the past week did you watch latenight comedy programs like The Late Show with David Letterman, The Tonight Show with

408

Lauren Feldman and Dannagal Goldthwaite Young

Jay Leno, or The Daily Show with Jon Stewart?” A full 70.5% of respondents reported not having viewed any late-night comedy in the previous week. Those who reported watching on at least one day were then asked, “Which of the following late-night comedy programs do you watch most often?” Response options included The Late Show with David Letterman, The Tonight Show with Jay Leno, and The Daily Show with Jon Stewart, as well as an “other” category, for which respondents were invited to specify a different program. Of the total sample, 3.5% reported watching The Daily Show most frequently, 11% watched Leno, 12.5% watched Letterman, and 2.5% mentioned another program. Three indicator variables were created to designate non-late-night viewers, Daily Show dominant viewers, and Leno/Letterman dominant viewers, respectively; the latter category combined audiences of Leno, Letterman, and “other” programs.2 Attention to Television News Stories About the Campaign. A single item was used to measure television news attention. Respondents were asked “During the past week, how much attention did you pay to stories about the campaign for president on national network or cable news?” Response options included “no attention at all” (1) “not too much” (2) “some” (3) and “a great deal of attention” (4). The 16% of respondents who indicated in earlier survey questions that they watched neither national network news nor cable news were coded as paying no attention. Among the entire sample, 28% of respondents reported paying no attention to campaign news stories on television; 21% reported not paying much attention, and 46% reported having paid some attention. Nearly 16% of respondents reported paying a great deal of attention to news stories. Measures for control variables used in the analyses are described in Appendix A. Analytic Strategy Because of its large sample size, the NAES is particularly conducive to analyses involving small subgroups of the population, such as viewers of The Daily Show (Young, Tisinger, Kenski, & Romer, 2006). The present study takes advantage of this aspect of the NAES in a two-stage analysis. In the first stage, each daily NAES sample is combined to form one large cross-section. This maximizes sample size and permits an exploration of the news attention of various late-night audiences at the individual level using ordinary least squares (OLS) regression. In the second stage, time series methodology is employed to more directly address the causal nature of the relationship between late-night comedy viewing and attention to the campaign in traditional news media. By comparing the trends in news attention during the course of the primary campaign among The Daily Show dominant viewers, Leno/ Letterman dominant viewers, and nonviewers of comedy, conclusions can be drawn about whether viewing various forms of late-night comedy is associated with increases in the rate at which people attend to the campaign on traditional television news programs.

Results Stage 1: Cross-Sectional Findings Multiple regression was used to test Hypotheses 1a and 1b. This method permits us to rule out the possibility that antecedent variables such as political interest, mass media exposure, and demographics account for any observed association between late-night comedy viewing and campaign news attention. The regression results are presented in Table 1.

Comedy as a Gateway

409

Table 1 Ordinary least squares (OLS) regression model predicting attention to the presidential campaign in network or cable television news from October 30, 2003, to June 4, 2004 B (SE) Constant Demographics Age Gender (male) Education Race (White) Political involvement Republican Democrat Ideology (liberal) Follow politics Mass media exposure Cable TV news Network TV news Local TV news Newspaper Late-night viewership Daily Show dominant Leno/Letterman dominant R2 N

β

0.68 (0.03)*** −0.003 (0.00)*** 0.07 (0.01)*** 0.03 (0.004)*** 0.02 (0.01)

−.04 .04 .03 .01

−0.007 (0.01) 0.05 (0.01)*** −0.01 (0.01) 0.32 (0.01)***

−.003 .02 −.01 .27

0.11 (0.002)*** 0.09 (0.002)*** 0.03 (0.002)*** 0.004 (0.002)* 0.16 (0.03)*** 0.07 (0.01)***

.31 .23 .06 .01 .03 .03 .36 29,054

*p < .05; ***p < .001.

Perhaps unsurprisingly, exposure to national television news on cable is the strongest predictor of campaign news attention (β = .31, p < .001). A tendency to follow what is going on in government and public affairs (β = .27, p < .001) and exposure to national network news (β = .23, p < .001) also demonstrate strong associations with attention. However, even with these controls, both Daily Show dominant viewers and Leno/Letterman dominant viewers are significantly more attentive to television campaign news than individuals who do not watch late-night comedy. The relationship between late-night viewing and attention, relative to non-viewers, appears to be of equal strength for audiences of The Daily Show and Leno/Letterman (both βs = .03, p < .001). These results, then, provide support for Hypotheses 1a and 1b.3 Stage 2: Time Series Findings While the regression results both establish an individual-level association among latenight comedy viewing and attention to campaign news and help to rule out some thirdvariable explanations for this relationship, the cross-sectional nature of the data prohibits assertions about causality. We do not know whether individuals who watch late-night comedy are those who pay attention to television campaign news to begin with, or if late-night

410

Lauren Feldman and Dannagal Goldthwaite Young

comedy viewing promotes this attention. By analyzing time trends in the data, however, we can assess whether the rate of growth in campaign news attention varies among viewers and nonviewers of late-night comedy. To do this, the individual cases for each of the three subpopulations of interest (i.e., Daily Show dominant viewers, Leno/Letterman dominant viewers, non-late-night viewers) were aggregated to the daily level, and a visual representation of the data was obtained.4 Figure 1 displays mean levels of campaign news attention, tracked between October 30, 2003, and June 4, 2004, for audiences of The Daily Show, of Leno and Letterman, and for those who watch no late-night comedy. To remove the sampling variation produced by the relatively small daily samples for the three subpopulations, and particularly for Daily Show dominant viewers, the data have been smoothed using a 21-day centered moving average.5 A visual examination of the smoothed data in Figure 1 indicates that audiences of The Daily Show and of Leno/Letterman exhibit higher levels of traditional news attention than non-late-night viewers from the outset of the primary campaign. All three subgroups, however, demonstrate increasing trends in attention during the first two-thirds of the campaign, which then flatten and decrease somewhat toward the end of the primary season.6 This suggests that both groups of late-night comedy viewers might start out paying more attention to campaign news than nonviewers and merely maintain this advantage over time, as opposed to increasing their rate of attention as a result of watching late-night comedy. On the other hand, the initial gap in news attention between Leno/Letterman dominant viewers and non-late-night viewers appears to widen slightly throughout the

Figure 1. Attention to campaign news on television among Daily Show dominant viewers, Leno/ Letterman dominant viewers, and non-late-night viewers from October 30, 2003, to June 4, 2004 (21-day moving average).

Comedy as a Gateway

411

course of the primary campaign. This would suggest that Leno/Letterman dominant viewers might increase their attention to campaign news at a faster rate than non-viewers. Statistical tests are necessary to determine if these differences are significant. Another pattern observable in Figure 1 is the sharp discontinuity in attention levels that occurs among Leno/Letterman dominant and non-late-night viewers around the middle of January. The trends in attention for these subgroups appear to increase gradually from the beginning of the series until roughly the time of the Iowa caucuses (January 19, 2004), at which point the rate of change in attention becomes much sharper.7 In contrast, attention levels for Daily Show dominant viewers seem to climb more steadily during this period of the campaign, suggesting that these viewers may not be as responsive to external campaign events, either in general or as a function of the coverage given such events on late-night comedy programs. Again, statistical analysis is required in order to confirm these conclusions. The first step in a statistical analysis of these time series is to make them stationary by identifying and removing any trends in the data (Diebold, 2001). Each of the three series was thus subjected to a regression analysis that tested for linear and quadratic components. In these analyses, attention was predicted by the variable “day,” which ranged from 1 to 213 (representing each day in the time series), and its quadratic function, “day2.” To reduce problems with multicollinearity, the day variable was mean-centered prior to computing the polynomial term (Cronbach, 1987). To test for the effects of the Iowa caucuses, a dichotomous variable, “Iowa,” was created, coded as 0 for all days up to and including January 19, 2004, and 1 for all days after January 19. This was included as a third predictor in the regression models. The regression results are presented in Table 2. Model 1 includes only the linear and quadratic predictors; the Iowa variable is added in Model 2. Results for Model 1 indicate significant linear and quadratic components in the news attention of Leno/Letterman dominant and non-late-night viewers.8 The trend for Daily Show dominant viewers also demonstrates a significant linear component, though the quadratic component is only marginally significant. This suggests that, for viewers of The Daily Show, the trend in news attention during the primary campaign is dominated by a linear increase. As revealed in Model 2, the Iowa predictor explains additional variance in the attention of Leno/Letterman audiences and non-late-night viewers but not for Daily Show dominant viewers.9 Figure 2 plots the predicted scores from the best-fitting regression model for each audience subgroup (i.e., Model 2 for Leno/Letterman dominant and nonviewers and Model 1 for Daily Show dominant viewers). This graph depicts patterns similar to those observed with the smoothed data in Figure 1. An analysis of the residuals of the fitted regression models indicated no evidence of autoregressive (AR) or moving average (MA) components. This means that once removing the linear and quadratic trends and the effect of the Iowa caucuses, there is no correlation between news attention on a given day and news attention on previous days. As suggested by Young et al. (2006), it is thus possible to test the statistical significance of the trends observed in Figure 2 by comparing the parameters of the fitted equations from Table 2. To compare regression parameters across subsamples, independent-sample t tests are used (Kleinbaum, Kupper, Muller, & Nizam, 1998).10 A comparison of the intercepts across the three Model 2 equations reveals significant differences between the non-latenight group and both Daily Show dominant, t(418) = 4.53, p < .001, and Leno/Letterman dominant groups, t (418) = 9.82, p < .001. The intercepts for The Daily Show and Leno/ Letterman equations were not significantly different from one another, t(418) = 0.37, p = .71.11 This indicates that early in the primary season viewers of both types of late-night comedy

412

2.25 (0.02)*** −0.0001 (0.0003) −1.70E-05 (2.67E-06)*** 0.216 (0.03)***

2.27 (0.02)*** 0.001 (0.0001)*** −2.62E-05 (2.45E-06)***

−.037 −.340 .613 .58 213

.473 −.523 .50

β

Note. Day was mean-centered prior to computing the quadratic term. † p < .10. **p < .01; ***p < .001.

Model 2 Constant Day (linear) Day2 (quadratic) Iowa R2 N

Model 1 Constant Day (linear) Day2 (quadratic) R2

B (SE)

News attention of non-late-night comedy viewers

2.52 (0.03)*** 0.0003 (0.0004) −2.26E-05 (4.00E-06)*** 0.227 (0.05)***

2.54 (0.03)*** 0.002 (0.0002)*** −3.22E-05(3.52E-06)***

B (SE)

.085 −.329 .470 .50 213

.476 −.469 .45

β

News attention of Leno/Letterman dominant viewers

2.55 (0.07)*** 0.0003 (0.001) −8.56E-06 (1.11E-05) 0.201 (0.14)

2.57 (0.07)*** 0.002 (0.001)** −1.71E-05 (3.52E-06)†

B (SE)

β

.044 −.061 .204 .07 213

.214 −.122 .06

News attention of Daily Show dominant viewers

Table 2 Regression results predicting time trends in news attention among non-late-night comedy viewers, Leno/Letterman dominant viewers, and Daily Show dominant viewers

Comedy as a Gateway

413

Mean Attention to Television News about the Campaign

2.80

Daily Show Dominant Leno/Letterman Dominant No Late-Night Group

2.60

2.40

2.20

2.00 /04 6/2 04 3/ 5/2 4 3/0 5/1 4 /0 5/3 04 3/ 4/2 4 3/0 4/1 4 /0 4/2 04 3/ 3/2 04 3/ 3/1 4 /0 3/3 04 2/ 2/2 4 2/0 2/1 4 /0 2/2 4 3/0 1/2 04 3/ 1/1 04 3/ 1/0 /03 /20 12 /03 /10 12 /03 /30 3 11 9/0 /1 3 11 9/0 /0 3 11 0/0 /3 10

Figure 2. Predicted trends in news attention among Daily Show dominant viewers, Leno/Letterman dominant viewers, and non-late-night viewers from October 30, 2003, to June 4, 2004.

paid significantly more attention to traditional television campaign news than did non-late-night viewers. The initial advantage in attention observed in Figure 2 for Daily Show dominant viewers relative to Leno/Letterman dominant viewers was not, however, discernible at levels of statistical significance. To test Hypothesis 2a, which predicted that watching Leno or Letterman increases the rate at which people pay attention to campaign news on television, we focused on the differences between the Leno/Letterman dominant time series and the non-late-night viewer time series that were observed in Figure 2. In Figure 2, it appears that the slope of the curve for the Leno/Letterman dominant group—and, to a lesser extent, the size of the Iowa increase—is steeper than for the nonviewers. A comparison of the coefficients for the Iowa effect in Model 2 (0.227 vs. 0.216) failed to yield a significant difference, t(418) = 0.38, p = .71. Thus, the spike in news attention prompted by the Iowa caucuses was no greater for Leno/Letterman dominant viewers than it was for individuals who did not watch latenight comedy. To test the difference between the overall linear trends in the two series, we compared the coefficients for day from Model 1. Because it excludes the Iowa variable and is thus unaffected by problems with multicollinearity, Model 1 provides a more stable estimate of the linear trend than does Model 2. In this case, the difference in slopes (0.002 vs. 0.001) was significant, t(420) = 2.12, p < .05. This suggests that the growth in news attention occurred faster for Leno/Letterman dominant viewers than for non-late-night viewers.

414

Lauren Feldman and Dannagal Goldthwaite Young

Although the above result suggests that watching late-night comedy such as Leno or Letterman increases the rate at which people attend to the campaign on network and cable news, other factors correlated with one’s self-selecting into or out of the Leno and Letterman audience could also have influenced attention. If this were the case, Leno/Letterman dominant viewers may appear to have increased their news attention at a faster rate than nonviewers even if that increased attention were actually a result of a change in the composition of the audience rather than from the content of late-night programming. Following Young et al. (2006), we explored this possibility by examining differences in the demographic characteristics of Leno/Letterman dominant and non-late-night viewers. Table 3 displays mean differences in daily demographics (age, gender, education, political ideology, and party identification) between Leno/Letterman dominant and nonlate-night viewers during the period from October 30, 2003, to June 4, 2004. A t test was used to assess whether the mean differences were significant. These results, presented in Table 3, reveal that Leno/Letterman dominant viewers are significantly more likely to be male, younger, better educated, liberal, and Democrat than those who do not watch any late-night comedy. And, as was demonstrated above, Leno/Letterman dominant viewers pay more attention to the campaign on television news. To determine if these differences might have changed over the course of the primary campaign, correlations were estimated between the difference scores and time (i.e., “day”). The correlations reported in Table 3 indicate no evidence of a linear relationship between time and any of the demographic difference scores. Thus, it appears that the Leno/Letterman dominant audience did not, over the course of the primary season, differentially gain members who were more likely to follow traditional news about the campaign. The final column in Table 3, which presents correlations between differences in news attention and differences in demographics, reveals significant relationships between attention and both knowledge and age. These findings suggest that demographic differences between the two groups, though unrelated to time, may have had some influence on differences in attention. In other words, the mean differences in age and education between Leno/Letterman dominant viewers and non-late-night comedy viewers may be predictive of mean differences in attention; however, because the size of these demographic differences does not change with time (as evidenced by their lack of correlation with “day”), they would not be able to explain why, throughout the course of the campaign, the linear Table 3 Mean differences in demographics between Leno and Letterman dominant and non-late-night comedy viewers from October 30, 2003, to June 4, 2004, and correlations of those differences with linear time (day) and news attention Mean difference

t ratio

Correlation with day

Correlation with news attention

0.04 −2.31 0.15 0.08 0.02 −0.01 0.30

6.05*** 11.01*** 9.86*** 6.64*** 3.07** 1.81 20.94***

−.024 −.033 .050 .069 −.045 −.027 .145*

.028 .162* .323*** −.064 −.018 .122

Gender (male) Age Education Political ideology (liberal) Democrat Republican Attention to news *p < .05; **p < .01; ***p < .001.

Comedy as a Gateway

415

increase in news attention occurs at a faster rate for Leno/Letterman dominant viewers than for nonviewers. To reiterate, the greater increase in attention exhibited by Leno/ Letterman dominant viewers is not an artifact of the changing demographic composition of the audience over time. Consistent with our earlier results, we also see in Table 3 that there is a significant correlation between time (day) and differences in news attention (r = .15, p < .05). Again, this suggests that the difference in mean news attention between Leno/Letterman dominant viewers and nonviewers increases linearly over time. To determine the size of this increase and to ensure that this trend is not explained by demographic differences, we regressed the attention series on day and demographic variables.12 This regression was conducted hierarchically: Day was entered first, and then stepwise procedures were used to determine whether any of the demographic variables entered into the equation. As seen in Table 4, a small but significant linear time trend in the attention series emerges in the first step of the analysis (β = .15, p < .05).13 With each subsequent day of the campaign, the average difference in news attention between Leno/Letterman dominant and non-viewers grows by 0.0005 units. In the second regression step, age and education are the only demographic differences to enter into the model; however, their inclusion does not appreciably change the relation between time and attention.14 These results reinforce the finding that watching late-night comedy programs such as Leno and Letterman increased the rate at which people paid attention to the campaign in traditional news sources and suggest that this trend is not attributable to over-time shifts in the demographics of late-night audiences. This provides clear support for Hypothesis 2a. With regard to Hypothesis 2b, there were no significant differences between the linear trend in the time series for Daily Show dominant viewers and that for nonviewers, t(420) = .64, p = .52. This suggests that viewers of The Daily Show do not increase their attention to campaign news at a faster rate than those who do not watch late-night comedy. At the same time, Figure 2 illustrates that the time series for Daily Show dominant viewers is Table 4 Regression results predicting news attention differences between Leno and Letterman dominant and non-late-night comedy viewers from October 30, 2003, to June 4, 2004 News attention differences B (SE) Model 1 Constant Day R2 Model 2 Constant Day Education differences Age differences R2 N *p < .05; ***p < .001.

0.25 (0.03)*** 0.0005 (0.0002)* .02 0.23 (0.03)*** 0.0005 (0.0002)* 0.29 (0.06)*** 0.01 (0.004)* .14 213

β

t ratio

.15

8.67 2.13

.13 .30 .13

7.66 2.09 4.62 2.00

416

Lauren Feldman and Dannagal Goldthwaite Young

distinguished from that of the other two subgroups—which speaks to our research question. The pattern of increasing attention seen in the trend line for The Daily Show audience appears to be unaffected by campaign events such as the Iowa caucuses; it also does not subside toward the end of the campaign to the same extent as is observed for the other two groups. Indeed, the effect of the Iowa caucuses on The Daily Show time series was not statistically significant, and the quadratic trend was only marginally significant (see Table 2).15

Discussion The results of this study demonstrate that exposure to late-night comedy is associated with higher levels of attention to the presidential campaign in traditional television news. The cross-sectional regression results reveal a positive association between watching late-night comedy—whether it be Leno or Letterman, or The Daily Show—and attending to the presidential campaign on traditional television news programs that cannot be explained by news media exposure, a tendency to follow politics, or demographic differences. The time series analysis further suggests that viewers of The Daily Show and of Leno and Letterman are more inclined than non-late-night viewers to pay attention to traditional news at the outset of the primary campaign; most importantly, however, the attention of the Leno/ Letterman dominant group also increases at a faster rate than does the non-late-night group. This latter finding is consistent with a gateway effect among Leno/Letterman dominant viewers. It is interesting to note that the Leno/Letterman dominant viewers and non-late-night viewers respond similarly to campaign dynamics: For both groups, the Iowa caucuses accelerate attention to the campaign, and as campaign events subside, so does their attention. The effect of the Iowa caucuses on news attention is perhaps unsurprising; the caucuses are the first step in the nominating process and thus generate attention from candidates, the media, and the public. Indeed, late-night comedians included the Iowa caucuses in their joke material in 2004. In light of this, one may have expected the Iowa caucuses to prompt a sharper increase in traditional news attention among viewers of Leno or Letterman than among nonviewers; this, however, was not found to be the case. This is likely because late-night comedians such as Leno and Letterman were only one of many information sources that signaled people to the official beginning of the nominating season. In 2004 especially, the attention given to Howard Dean’s infamous scream both in the media and around the water cooler likely helped raise the salience of the campaign for Leno/Letterman viewers and nonviewers alike. Although statistical power issues preclude us from making an unequivocal claim about the distinctiveness of the time trend for Daily Show dominant viewers, the data do suggest that their increase in attention is more protracted and less subject to the influence of campaign events. Similarly, the news attention of Daily Show dominant viewers appears more resistant to decline toward the end of the primaries. As a result, the Leno/ Letterman dominant and non-late-night groups reach their attentional peak earlier in the campaign than do Daily Show dominant viewers but begin to taper more quickly, whereas the Daily Show audience reaches its peak later but does not experience as much decline. These trends speak to how The Daily Show may function differently from other late-night comedy programs. Unlike Leno and Letterman, the content of The Daily Show is consistently and reliably political—not just during campaign events and elections, but all the time. For this reason, its viewers may not be as dependent on the program’s coverage of discrete events such as the Iowa caucuses to motivate their attention to traditional news; nor will their attention to the news necessarily wane in the absence of campaign activity.

Comedy as a Gateway

417

Although the NAES afforded a comparison among late-night audiences that many other data sets would not, the small daily sample sizes for the two comedy subgroups— and particularly for The Daily Show—present some limitations. Not only, as already mentioned, did the small sample sizes impose restrictions on the inferences we can draw about differences between The Daily Show and the other viewing subgroups, they also prohibited us from stratifying the time series by age or political interest. Although we were able to demonstrate that news attention throughout the primaries increased at a faster rate for Leno/Letterman dominant viewers than for those who did not watch late-night, we could not assess whether these effects were concentrated among those with less interest in politics, as would be predicted by Baum’s gateway model (2003b). These and other potential moderating effects will need to be addressed in future research, as will the mechanisms via which Leno and Letterman motivate their viewers to pay increasing attention to television news. For example, is it that the political focus of their humor—or of their guests—raises the salience of the election campaign among the politically inattentive? Or is it something unique about comedy that stimulates news attention among viewers; that is, does a desire to be “in” on the jokes of Leno and Letterman propel further information seeking? And finally, while we have taken steps to tease out possible confounding characteristics of late-night comedy audiences that might be driving these increases in campaign attention, it is certainly possible that some other latent feature of late-night viewers is responsible for these trends. For example, in our OLS analyses, we lacked a control measure of total television exposure, which would have ruled out the possibility that the attentional advantage exhibited by late-night comedy viewers relative to nonviewers is an artifact of higher television use overall. A true crucial test of Baum’s gateway hypothesis in the context of late-night comedy programming would require an experimental design that followed participants over time. The results of this study should nonetheless comfort those journalists and scholars who have voiced alarm about young people turning to late-night comedy for their campaign and election information in the place of news. Late-night comedy audiences—of both The Daily Show and Leno/Letterman varieties—are not tuning into these programs instead of traditional news (see also Young & Tisinger, 2006). In fact, over the course of an election campaign, the audiences of Leno and Letterman increase their attention to campaign news at a faster rate than nonviewers—suggesting that the late-night comedy of Leno and Letterman fosters interest in conventional news and politics. The audience for The Daily Show, on the other hand, maintains high levels of news attention regardless of the program’s election coverage. Although this study was concerned with late-night comedy’s ability to serve as a gateway to traditional news use, we do not wish to imply that the latter is somehow a superior form of political communication than the former. As entertainment programs incorporate political themes and traditional news programs strive for ratings with stories of celebrities, oddities, and drama (Delli Carpini & Williams, 2001), it becomes increasingly difficult to argue that one genre is more serious or important than the other. Thus, we hope our results can be interpreted as evidence for how exposure to Leno and Letterman broadens engagement with political media content by encouraging attention to news from sources that have, at least historically, been deemed public affairs oriented. At the same time, our findings make clear that for contemporary media audiences, the choice between traditional news and entertainment is not a zero-sum game. Just as the boundaries between news and entertainment programming are becoming more fluid in terms of content and effects, so too are audiences’ patterns of consumption—making the continued study of entertainment media’s political impact ever more important.

418

Lauren Feldman and Dannagal Goldthwaite Young

Notes 1. This and subsequent joke content was compiled as part of a content analysis conducted by the second author. 2. Although our measurement and analyses treat Daily Show viewers, Leno/Letterman viewers, and those who do not watch any late-night comedy as three distinct subgroups, it is possible that people who report most often watching The Daily Show also watch Leno or Letterman on occasion, and vice versa (see Prior, 2007, for a similar discussion involving cable news audiences). To acknowledge the possibility of cross-viewership, we label the two late-night comedy viewing subgroups as “Daily Show dominant” and “Leno/Letterman dominant.” 3. We recognize that the large sample size of the NAES could render even small effects statistically significant. For this reason, we ran a second regression analysis that was confined to respondents between the ages of 18 and 29 (N = 5,879), which is the age group considered both to be the primary audience for late-night comedy and the least attentive to traditional news (Pew Center, 2000, 2004). In this analysis, the relationship between late-night viewing and news attention remains significant among Daily Show dominant viewers (b = .06, p < .001) and Leno/Letterman dominant viewers (b = .05, p < .001). Of note, with this younger subset, the association between late-night viewing and news attention is even stronger than it was with the full sample. In fact, interactions between age and both the Leno/Letterman and Daily Show variables in the full sample model were significant and negative (both βs = −.04, p < .01), indicating that the relationship between late-night comedy viewing and news attention is stronger for younger than older adults. 4. The average daily sample sizes for each subgroup were as follows: non-late-night (M = 126.6, SD = 40.5), Leno/Letterman dominant (M = 46.8, SD = 16.3), and Daily Show dominant (M = 6.3, SD = 3.2). 5. Moving averages, which pool data across days, reveal important patterns in the data that would otherwise be obscured by sampling error (Johnston, Blais, Brady, & Crete, 1992). The centered moving average for a particular day is that day’s value averaged with a specified range of values around it. For example, the 21-day centered moving average is the average value obtained from 10 days before through 10 days after a given day. This is why, in Figure 1, the trend lines begin and end 10 days after/before the true endpoints of the campaign period being analyzed (i.e., October 30, 2003, and June 4, 2004). 6. Although the trend in news attention among Daily Show dominant viewers appears to fluctuate toward the end of the campaign, a closer examination of the time series using autocorrelation techniques suggests that this movement is random error, likely owing to the variation in daily sample sizes for the Daily Show subpopulation during this time. After removing the linear and quadratic trends, the series is comprised only of white noise (i.e., it has a mean of 0, constant variance, and no serial correlation). This indicates that there is no systematic relation between time periods beyond the general linear and quadratic trend. 7. The discontinuity in the time trend appears in the graph on January 13 rather than on January 19. Although this slight discrepancy could be attributed to anticipation of the Iowa caucuses, it is more likely due to the fact that a centered moving average incorporates the values of data following a given day, and therefore changes tend to appear in the graph before they actually occurred (Kenski, 2006b). Thus, in this case, a change in attention levels on January 19 was plausibly picked up in the 21-day centered moving average on January 13. 8. A cubic component was also tested but was not significant. 9. Note that when the Iowa predictor is added in Model 2, the linear trend is reduced to nonsignificance. Although this might suggest that most of the linear trend in news attention for Leno/Letterman dominant viewers and non-late-night viewers is captured by the effect of the Iowa caucuses, this should be interpreted cautiously because of the collinearity between the day and Iowa predictors (estimates for day2 should be unaffected since centering was used to reduce its correlation with day). Multicollinearity also likely explains why the estimated effects of the linear and quadratic predictors drop to nonsignificance when the effect of the Iowa caucuses is added to the Daily Show model.

Comedy as a Gateway

419

10. The test statistic is obtained by calculating the difference between the estimates of two slopes (or intercepts) and dividing this by the standard error of the difference in slopes (or intercepts), as captured in the following formula:

t=

B1 − B2 SB1 − B2

.

The computation of the denominator, SB1 − B2 , involves pooling and summing the residual mean-square errors for each regression equation. The complete formulas used for comparing slopes and intercepts can be found in Kleinbaum et al. (1998, pp. 322–326). Necessary computational statistics are provided in Appendix B. The t statistic obtained is evaluated using (N1 + N2 – 2k) degrees of freedom, where N is the number of days in each series and k is the number of parameters in each model. 11. Note that nearly identical results were obtained when comparing the constants across the Model 1 equations. In both cases, the significant differences detected between the intercepts for Daily Show dominant and non-late-night viewers, and between Leno/Letterman dominant and non-late-night viewers, withstand adjustments for multiple comparisons. Using the Bonferroni correction, which divides the Type I error rate (0.05) by the number of comparisons (here, three), .017 becomes the criterion for significance. The p values obtained fall well below this cutoff. 12. Prior to conducting the regression, we examined the time series and determined that it did not contain any AR or MA components. 13. A quadratic time trend was also tested but was not significant. 14. Interactions between time and differences in both age and education were also tested; neither was significant. This further suggests that the audience differences in age and education did not, over time, become more or less important in predicting differences in news attention. 15. The nonsignificant effect of the Iowa caucuses among Daily Show dominant viewers does not necessarily indicate that this effect is significantly less than that for the Leno/Letterman dominant group. Thus, we used a t test to compare the coefficients for the Iowa effect found for Leno/ Letterman dominant viewers and Daily Show dominant viewers. This difference was not significant; however, given the relatively smaller daily sample sizes and wider sampling variability of the Daily Show audience, we likely did not have ample statistical power to detect a difference here. Similar conclusions can be drawn with regard to differences in the size of the quadratic effects across the Daily Show and Leno/Letterman equations, which were also not significant.

References Associated Press. (2006). Jon Stewart’s Oscar host role helps Comedy Central in the ratings. Associated Press, March 18. Baum, M. A. (2003a). Soft news and political knowledge: Evidence of absence or absence of evidence? Political Communication, 20, 173–190. Baum, M. A. (2003b). Soft news goes to war. Princeton, NJ: Princeton University Press. 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. Baumgartner, J., & Morris, J. (2006). The Daily Show effect: Candidate evaluations, efficacy, and the American youth. American Politics Research, 34, 341–367. Baym, G. (2005). Discursive integration and the reinvention of political journalism. Political Communication, 22, 259–276. Chaffee, S. H. (1986). Mass media and interpersonal channels: Competitive, convergent, or complementary? In G. Gumpert & R. Cathcart (Eds.), InterMedia: Interpersonal communication in a media world (3rd ed., pp. 62–80). New York: Oxford University Press.

420

Lauren Feldman and Dannagal Goldthwaite Young

Chaffee, S. H., & McLeod, J. M. (1972). Individual vs. social predictors of information seeking. Journalism Quarterly, 50, 237–245. Chaffee, S., & Yang, S. (1988). Communication and political socialization. In O. Ichilov (Ed.), Political socialization for democracy (pp. 137–157). New York: Teachers College Press. Colapinto, J. (2004, October 28). The most trusted name in news. Rolling Stone, pp. 58–64. Cronbach, L. J. (1987). Statistical tests for moderator variables. Psychological Bulletin, 87, 51–57. C-Span Newhouse School Forum. (2004, October 14). Interview with Jon Stewart by Ken Auletta from the New Yorker. Delli Carpini, M., & Williams, B. (2001). Let us entertain you: Politics in the new media environment. In L. Bennett & R. Entman (Eds.), Mediated politics: Communication in the future of democracy (pp. 160–191). New York: Cambridge University Press. Diebold, F. X. (2001). Elements of forecasting (2nd ed.). Cincinnati: South-Western Thompson Learning. Drew, D., & Reeves, B. (1980). Children and television news. Journalism Quarterly, 57, 45–54. Dutwin, D. (2000). Knowledge in the 2000 primary elections. Annals of the American Academy of Political and Social Science, 572, 17–25. Fox, J. A., Koloen, G., & Sahin, V. (2007). No joke: A comparison of substance in The Daily Show with Jon Stewart and broadcast network television coverage of the 2004 presidential election campaign. Journal of Broadcasting & Electronic Media, 51, 213–227. Gay, V. (2004, January 14). Not necessarily the news; meet the players who will influence coverage of the 2004 campaign, you might be surprised. Newsday, B6. Holbert, R. L. (2005a). Intramedia mediation: The cumulative and complementary effects of news media use. Political Communication, 22, 447–462. Holbert, R. L. (2005b). A typology for the study of entertainment television and politics. American Behavioral Scientist, 49, 436–453. Holbert, R. L., Lambe, J. L., Dudo, A. D., & Carlton, K. A. (2007). Primacy effects of The Daily Show and national TV news viewing: Young viewers, political gratifications, and internal political self-efficacy. Journal of Broadcasting & Electronic Media, 51, 20–38. 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. Johnston, R., Blais, A., Brady, H. E., & Crete, J. (1992). Letting the people decide: Dynamics of a Canadian election. Stanford, CA: Stanford University Press. Johnston, R., & Brady, H. E. (2002). The rolling cross-section design. Electoral Studies, 21, 283–295. Jones, J. (2007). “Fake” news versus “real” news as sources of political information: The Daily Show and postmodern political reality. In K. Reigert (Ed.), Politicotainment: Televison’s take on the real (pp. 129–150). New York: Peter Lang. Kennamer, J. D., & Chaffee, S. H. (1982). Communication of political information during early presidential primaries: Cognition, affect, and uncertainty. In M. Burgoon (Ed.), Communication yearbook 5 (pp. 627–650). New Brunswick, NJ: Transaction Books. Kenski, K. (2006a). The rolling cross-section design. In D. Romer, K. Kenski, C. Adasiewicz, & K. H. Jamieson (Eds.), Capturing campaign dynamics: The National Annenberg Election Survey 2000 and 2004 (pp. 68–78). New York: Oxford University Press. Kenski, K. (2006b). Visualizing data across the campaign. In D. Romer, K. Kenski, C. Adasiewicz, & K. H. Jamieson (Eds.), Capturing campaign dynamics: The National Annenberg Election Survey 2000 and 2004 (pp. 104–120). New York: Oxford University Press. Kleinbaum, D. G., Kupper, L. L., Muller, K. E., & Nizam, A. (1998). Applied regression analysis and other multivariable methods. Pacific Grove, CA: Duxbury Press. Mindich, D. T. Z. (2005). Tuned out: Why Americans under 40 don’t follow the news. New York: Oxford University Press.

Comedy as a Gateway

421

Moy, P., Xenos, M. A., & Hess, V. K. (2006). Priming effects of late-night comedy. International Journal of Public Opinion Research, 18, 198–210. National Annenberg Election Survey (2004). Daily Show viewers knowledgeable about presidential campaign. Retrieved February 7, 2008, from http://www.annenbergpublicpolicycenter.org/ Downloads/Political_Communication/naes/2004_03_late-night-knowledge-2_9-21_pr.pdf Niven, D., Lichter, S. R., & Amundson, D. (2003). The political content of late night comedy. Harvard International Journal of Press/Politics, 8(3), 118–133. Pew Research Center for the People and the Press. (2000). Audiences fragmented and skeptical: The tough job of communicating with voters. Retrieved from http://people-press.org/reports/display. php3?ReportID=46 Pew Research Center for the People and the Press. (2004). Cable and Internet loom large in fragmented political news universe. Retrieved from http://people-press.org/reports/display.php3?Report ID=200 Pfau, M., Cho, J., & Chong, K. (2001). Communication forms in U.S. presidential campaigns: Influences on candidate perceptions and the democratic process. Harvard International Journal of Press/Politics, 6(4), 88–105. Prior, M. (2007). Post-broadcast democracy: How media choice increases inequality in political involvement and polarizes elections. New York: Cambridge University Press. Stanley, A. (2006). Leno vs. Letterman: A battle of wits with no clear winner. New York Times. Retrieved from http://www.nytimes.com/2006/06/16/arts/television/16stan.html?ei=5070&en= 9cc024e928288b82&ex=1181102400&pagewanted=print Ten questions for Jon Stewart. (2004, September 27). Time Magazine, p. 8. Young, D. G. (2004a). The Chicken or the Egg? Tracking themes in late-night jokes and trends in viewers’ candidate perceptions during the primaries. Paper presentation at the annual meeting of the National Communication Association, Chicago, IL, 11–14 November. Young, D. G. (2004b). Late-night comedy in election 2000: Its influence on candidate trait ratings and the moderating effects of political knowledge. Journal of Broadcasting & Electronic Media, 48, 1–22. Young, D. G. (2006). Late-night comedy and the salience of the candidates’ caricatured traits in the 2000 election. Mass Communication and Society, 9, 339–366. Young, D. G. (2007). The Daily Show as new journalism. In J.S., Morris & J. C. Baumgartner (Eds.), Laughing matters: Humor and American politics in the media age (pp. 241–259). New York: Routledge. Young, D. G., & Tisinger, R. (2006). Dispelling late-night myths: News consumption among latenight comedy viewers and the predictors of exposure to various late-night shows. Harvard International Journal of Press/Politics, 11(3), 113–134. Young, D. G., Tisinger, R., Kenski, K., & Romer, D. (2006). The power of numbers: Examining subpopulations with NAES. In D. Romer, K. Kenski, C. Adasiewicz, & K. H. Jamieson (Eds.), Capturing campaign dynamics: The National Annenberg Election Survey 2000 and 2004 (pp. 244–268). New York: Oxford University Press.

Appendix A: Control Variables Included in the Models Mass media exposure: Respondents were asked how many days in the past week they had “watched the national network news on TV—by national network news, I mean Peter Jennings on ABC, Dan Rather on CBS, Tom Brokaw on NBC, and Jim Lehrer News Hour on PBS” (network TV news); “watched a 24-hour cable news channel, such as CNN, Fox News Channel, or MSNBC” (cable TV news); “watched local TV news—for example, Eyewitness News or Action News” (local TV news); and “read a daily newspaper” (newspaper). Follow politics: Respondents were asked: “Some people seem to follow what is going on in government and public affairs most of the time, whether there is an election or not. Others are not that interested, or are interested in other things. Would you say you follow

422

Lauren Feldman and Dannagal Goldthwaite Young

what is going on in government and public affairs most of the time [coded 4], some of the time [3], only now and then [2], or hardly at all [1]?” Political ideology: Respondents were asked: “Generally speaking, would you describe your political views as very conservative [coded 1], conservative [2], moderate [3], liberal [4], or very liberal [5]?” Party identification: Respondents were asked: “Generally speaking, do you usually think of yourself as a Republican, a Democrat, an Independent, or something else?” Indicator variables were created to represent Republicans, Democrats, and Independents/ other, respectively. Education: Respondents were asked: “What is the last grade or class you completed in school?”—some high school or less [coded 1], high school graduate [2], some college or associate’s degree [3], 4-year college degree [4], and graduate or professional schooling after college [5]. Additional demographics: Respondents were asked to report their age, gender, and race/ethnicity.

Appendix B: Computational Statistics for Comparing Regression Parameters Across Equations Table B1 Mean-square errors for regression equations Equation Non-late-night viewers Leno/Letterman Daily Show

Model 1 MSE

Model 2 MSE

0.015

0.012

0.030 0.213

0.028 0.212

Table B2 Means and variances of regression predictors Predictor Day (linear) Day2 (quadratic) Iowa

M

Variance

107.00 3,780.67 0.64

3,798.50 11,487,930.17 0.23