Trump’s 2016 Campaign as a Particular Test Case

Although the 2016 presidential campaign provides a clear opportunity to explore what happens when the legacy media cover—and the electorate considers—an entertainment-driven presidential campaign, the Trump campaign does not provide a perfect analytical window on these questions because the 2016 election was, historically speaking, unusual.

A confluence of factors contributed to Trump’s unexpected success, not least the state of the economy (Erikson and Wlezien Reference Erikson and Wlezien.2014; Vavreck Reference Vavreck2009). Other important factors in 2016 included a Republican party that “simply did not do the things a party would do to guide voters’ preferences or limit their choices” (Masket Reference Masket2016); a two-party system that has arguably alienated many voters and laid the ground work for populist unrest (Miller and Listhaug Reference Miller and Listhaug1990; Strömbäck and Dimitrova Reference Strömbäck and Dimitrova2006); and a steady decline in public esteem for political institutions (Gallup Reference Gallup2015; Harrington Reference Harrington2017; see also Pew Research Center2014)—all this before we even get to the weaknesses of the Democratic nominee and her campaign strategy, not to mention the role of gender politics in a campaign that pivoted on traditional gender attitudes and hostile sexism (Bock, Byrd-Craven, and Burkley Reference Bock, Byrd-Craven and Burkley2017; Maxwell and Shields n.d). To this catalogue of complicating factors must also be added the role of Russian cyberwarfare (Jamieson Reference Jamieson2018), and the rising appeal of populism and authoritarianism channeling simmering racial resentments (Hochschild Reference Hochschild2016; Inglehart and Norris Reference Inglehart and Norris.2016; Wade Reference Wade2016; Walsh Reference Walsh2016).

Moreover, the role played by alternative news sources and social media networks in the surprising success of the Trump campaign (Cornfield Reference Cornfield2017; Enli, Reference Enli2017; Wells et al. Reference Wells, Shah, Pevehouse, Yang, Pelled, Boehm, Lukito, Ghosh and Schmidt2016) demonstrate how digital and social media have to some extent supplanted traditional media as a source of campaign information for voters (see Boczkowski Reference Boczkowski2016). According to one recent analysis, key segments of the Trump electorate were not tuning in much to mainstream media, attending instead to Breitbart and other popular conservative sites (Benkler et al. Reference Benkler, Faris, Roberts and Zuckerman2017), although the effect of alternative and “fake news” sources on the election should not be overstated, according to another recent analysis (Allcott and Gentzkow Reference Allcott and Gentzkow2017).

Yet this historical confluence of confounding factors notwithstanding, research strongly suggests that heavy mainstream media coverage of Trump throughout the campaign—particularly in the critical early months of the primary season—was a key factor in his ultimate success (Patterson Reference Patterson2016a, Reference Patterson2016b, Reference Patterson2016c). Several political communication scholars have argued that “the news media helped to nominate Trump” (Azari Reference Azari2016) and that Trump “drove coverage to the nomination” (Wells et al. Reference Wells, Shah, Pevehouse, Yang, Pelled, Boehm, Lukito, Ghosh and Schmidt2016).

The case of Trump is unique in another important sense that is critical to understanding not just the amount, but the type of press coverage Trump received. As we will discuss further, the genesis of Trump’s fame in a distinctive genre of reality television (along with, to a lesser degree, talk radio, cable television, and World Wide Wrestling, as well as his self-applauding books and regular stints on talk radio) made him a particular kind of celebrity politician—one with a low-brow and divisive political style, and one particularly likely to be mocked by the establishment press. Generalizing from these particulars to how the press may cover celebrity candidacies of the future may therefore be challenging. Nevertheless, how the mainstream press covered the Trump campaign offers scholars the first opportunity at the U.S. presidential level to study how legacy media norms and routines may collide with celebrity politics—even as the media’s economic incentives ensured that the media couldn’t look away from a candidate like Trump.

Crashing the Gates: Donald Trump and Celebrity Politics in the Hybrid Media Era

The Trump campaign and the challenge it posed to traditional journalism and presidential politics reflects the growing celebritization of politics, evident decades ago in politicians like John F. Kennedy becoming celebrities, and in candidates like Ronald Reagan who moved from entertainment stardom into politics (West and Orman Reference West and Orman2003). In his influential work on celebrity politicians, John Street (Reference Street2004) cites the examples of two actors-turned-governor of California—Ronald Reagan and Arnold Schwarzenegger—and former pro wrestler-turned-governor of Minnesota Jesse Ventura (to which we might add more recent examples like comedian-senator Al Franken). Street describes these as candidates “whose background is in entertainment, show business or sport, and who trade on this background (by virtue of the skills acquired, the popularity achieved or the images associated) in the attempt to get elected” (2004, 437). Barack Obama’s “global supercelebrity” standing, complete with a European tour and rock star status before he was even elected, evinced the further entwinement of celebrity and politics (Kellner Reference Kellner, David Marshall and Redmond2016; see also Marsh, 't Hart, and Tindall Reference Marsh, Hart and Tindall2010; Wheeler Reference Wheeler2013). As these various examples illustrate, “celebrity politicians” can be either career politicians who become political celebrities, borrowing the techniques of entertainment world celebrity to further their political careers, or they can be entertainment realm celebrities who trade on their fame and stage skills to enter the world of politics.

Today’s political environment is more friendly to celebrity politicians of both sorts, given that, as Kellner puts it, contemporary U.S. politics is “informed by the logic of media spectacle and thus requires that candidates become masters of the spectacle” (Kellner Reference Kellner, David Marshall and Redmond2016, 116)—a dynamic some scholars have called the “tabloidization” of politics (Kellner Reference Kellner, David Marshall and Redmond2016; Turner Reference Turner2004). But from a broader analytical perspective, a growing literature traces the rise of a style of politics in which politicians (whatever their background) “[incorporate] matters of performance, personalization, branding and public relations into the heart of their political representation” (Wheeler Reference Wheeler2013, 87). According to Marshall, celebrity politics are just one expression of the larger phenomenon of “celebrity power,” in which “the disciplinary boundaries between the domains of popular culture and political culture have been eroded” (Marshall Reference Marshall1997, xiii). The phenomenon of celebrity politics is also linked to what Turner calls “democratainment,” reflecting the mass public’s declining identification with formal political parties and institutions, and the fact that “the consumption of celebrity has become a part of everyday life in the twenty-first century” (Wood, Corbett, and Flinders Reference Wood, Corbett and Flinders2016, 594). Indeed, Street argues, “celebrity politics, and the cult of the personality that it embodies, can be seen as a product of the transformation of political communication” (Street Reference Street2004, 441). This transformation, however, has not been consistently recognized in the mainstream of political communication research.

For its part, the political communication literature has identified the evolution of a “hybrid media system,” in which new media platforms and logics compete but also interact with traditional media. In this evolving media ecosystem, older forms of media exist alongside newer invasive species of media, the two sometimes struggling for dominance but both contributing to the news and information flows that shape politics (Chadwick Reference Chadwick2013). A defining feature of this evolving system is what Baym refers to as “discursive integration,” in which the markers dividing news from entertainment are becoming increasingly hard to discern (Baym Reference Baym2005). This transformation has created unprecedented opportunities for nontraditional figures to enter politics, particularly those who understand entertainment-derived values and aesthetics. A variety of actors now traverse the increasingly porous boundaries between traditional and entertainment-driven politics, across an increasing variety of media platforms (Lawrence and Boydstun Reference Lawrence, Boydstun, Walgrave and Van Aelst2017). In today’s mash-up media environment, “politically relevant” media include everything from comic strips and YouTube videos to Hollywood movies and, yes, reality TV shows (Baym Reference Baym2005; Williams and Delli Carpini Reference Williams and Delli Carpini2011). And ultimately, some scholars argue, the performative aesthetics and affective attachments between celebrities and “fans” that are more familiar to the world of entertainment are increasingly the basis for political representation as well (Street Reference Street2004; van Zoonen Reference van Zoonen2005).

While both literatures shed light on the changes in media and politics that made his campaign possible, Donald Trump presented the legacy press with a stark new version of celebrity politics, both in terms of his pre-established fame and (alleged) fortune when he entered the presidential arena, and in terms of his particular pathway to and style of celebrity. When he entered the 2016 presidential race Trump was already an entertainment household name (think: 14 seasons as host of The Apprentice). A Gallup survey of registered Republicans in July 2015, shortly after Trump announced his campaign, found that 92% recognized his name, compared with 81% who recognized the name Jeb Bush (son and brother to two former U.S. presidents), not to mention Ted Cruz (at 66%), and Marco Rubio (at 64%) (Dugan Reference Dugan2015). Trump’s name recognition was also likely due in part to his frequent coverage on FOX News prior to the election and to the increasingly tight bonds between that network and the Republican voter base. (As one journalist who has covered FOX closely recently observed, “Trump is the logical extension of the takeover of Republican party by television and entertainment” [National Public Radio 2018].)

Trump’s experience with the genre of reality TV was particularly important to this performative style. Not only did The Apprentice increase Trump’s name recognition, but the show’s casting calls in cities around the country—particularly in the South and Midwest—gave Trump, who assiduously attended these events that were described as record-breaking by the show’s producers, practice in the skills of campaigning (Frontline Reference Frontline2016). Although the show was for much of its run not as highly rated as Trump has often suggested (Bradley Reference Bradley2017), it helped him hone his unique style, skills bolstered by his numerous appearances on talk television and radio, and pro wrestling events.

Reality TV mattered for the unique case of Donald Trump in another way as well. Research suggests that for heavy viewers of the genre, watching reality TV involves a negotiation of what is “real” and “not real” about such programs. The pleasure in viewing reality TV hinges in part on the tension between the apparent spontaneity and authenticity of happenings within a show and the viewers’ sense that the show is still scripted—what Rose and Wood describe as “contrived authenticity” (Baruh Reference Baruh2010; Rose & Wood Reference Rose and Wood2005). This reality TV aesthetic seems particularly conducive to Trump’s political style—not just the daily installments of new drama but also the bombastic, winking showmanship and frequent crossings into hyperbole and fakery.

As Street observes in a recent essay, “celebrity performances are shaped by the conventions of the genre from which they emerge” (Street Reference Street2017, 16). For Trump, reality TV sharpened a performative style that is particularly aggressive, brash, iconoclastic and, some would say, mean-spirited, encapsulated in the pithy one-liner he inevitably uttered in each episode of The Apprentice: “You’re fired!” For his fans, Trump’s performative style may serve as a proxy for a style of political representation and leadership for which they yearn (Street Reference Street2017, 18).

And that style is key to understanding the challenge Trump presented for journalists. Even more than his lack of previous political experience, Trump’s transgressive political performance sets him apart from predecessors—particularly, his refusal to adopt and adapt to prevailing norms of presidential politics. Previous entertainers-turned-politicians like Reagan and Schwarzenegger shifted from entertainment to politics by rebranding themselves as more or less conventional politicians. Trump deviates from this pattern by continuing—even as president—to behave as an entertainer, “reformulat[ing]…politics as reality TV entertainment” (Schäfer-Wünsche and Kloeckner Reference Schafer-Wunsche and Kloeckner2016, 2). For a swath of American voters, Trump’s roots as an entertainer and, as Forbes once put it, a “business illusionist” (Lappin Reference Lappin2011) proved to be a powerful source of his appeal. Indeed, being an “outsider” was one of the most defining aspects of Mr. Trump’s support among voters (Huang and Yourish Reference Huang and Yourish2016). Although politicians have long succeeded by claiming to be Washington outsiders, Trump is “the first major U.S. presidential candidate to pursue politics as entertainment and thus to collapse the distinction between entertainment, news, and politics, greatly expanding the domain of spectacle” (Kellner Reference Kellner, Armano and Briziarelli2017, 6, emphasis added).

In his examination of the Trump phenomenon, Street observes that Trump “is not a ‘politician’ . . . because he fails to meet the standards expected of a democratic representative [and] expresses no desire to be such a figure” (Street Reference Street2017, 7–9). Indeed, Street argues that Trump must be understood primarily in terms of entertainment, not politics: That Trump is a celebrity first and foremost.

Guarding the Gates: The Legacy Press and Traditional Presidential Politics

Trump’s candidacy, unlike any previous U.S. presidential campaign in the degree to which it was saturated with and driven by entertainment aesthetics and the extent to which it flouted norms of traditional presidential politics, offers an extreme test of how the press reacts when celebrity and conventional politics collide.

One way we might expect that clash to manifest is by examining whether and how the press attempted to “winnow” Trump from the candidate field. As a long line of research has demonstrated, an important structuring feature of presidential elections are the press’s decisions about which candidates are “unserious” or “unelectable” and therefore deserving of less coverage, less favorable coverage, or both. This media handicapping of candidates during the all-important primary phase of a presidential election, when news outlets have to decide which contenders among many to devote their limited resources to covering, can influence candidates’ visibility and, thus, electability (Belt, Just, and Crigler Reference Belt and Crigler2012; Haynes et al. Reference Haynes, Gurian, Crespin and Zorn2004; Matthews Reference Matthews and Barber1978; Patterson Reference Patterson1994; Reference Patterson2016a). In the absence of objective criteria of presidential readiness, the establishment press act as gatekeepers of “electability.”

The literature on journalistic winnowing suggests two main winnowing triggers. First, candidates who lack name recognition, struggle to attract donors, or fail to gain ground in public opinion polls and primary contests are generally subject to winnowing. The press sees these candidates as “likely losers,” and coverage of these candidates is usually less positive, less frequent, or both. By the same token, candidates who are polling well deserve more coverage, journalists reason, because voters are clearly more interested in those candidates and thus they are more likely to win anyway (Hopmann, Van Aelst, and Legnante Reference Hopmann, Van Aelst and Legnante2012; Patterson Reference Patterson2005).

Second, candidates whose background, political views or personal demeanor seem outside the norm, or “unpresidential”, are also subject to winnowing. In particular, the press may deem some candidates not ready for the klieg lights of the presidential stage because of high profile campaign trail gaffes—from Edmund Muskie’s 1972 press conference in which he appeared to cry while defending his wife against politically-motivated attacks, to Howard Dean’s “scream” at an Iowa post-caucus event, which, although not the only factor in Dean’s demise, helped cement journalists’ impression that he was not ready for the presidential stage (Hindman Reference Hindman2005; Patterson Reference Patterson1994). Post-gaffe, candidates can find their media coverage and their campaign chances shift as a narrative of unelectability takes hold.

Most often, historically speaking, candidates who seem unpresidential and thus unelectable are the same candidates who struggle to gain ground at the polls. Think, for example, of Herman Cain’s 2012 campaign, which featured both a struggle to gain electoral steam and a candidate given to unconventional and sometimes unintelligible statements. In other words, the two dimensions of “unelectability”—and the press’s two main winnowing triggers—tend to co-occur. Campaigns featuring gravitas- and poll-challenged candidates, such as the 1992 presidential bid of Democrat Larry Agran, whose highest office held before seeking the presidency was mayor of Irvine, CA, can find themselves figuratively or even literally written out of campaign coverage (Meyrowitz Reference Meyrowitz and Kendall1995).

But public perceptions of presidentialness and electability may be shifting, perhaps shifting sharply, as politics becomes more influenced by celebrity and entertainment, as reviewed earlier, leaving the press in a more uncertain position vis-à-vis their winnowing power. The 2016 election thus presented a sharp and unique test of the press’s two main winnowing criteria: What to do when a profoundly unpresidential candidate (by traditional measures) enters the presidential stage, bringing with him an established brand name, high entertainment value, and an audience from beyond the political arena—and thus a real chance of being elected?

The literature reviewed earlier would likely have predicted that the press would write off a brash political novice and self-aggrandizing celebrity like Trump—much like the examples found in coverage of Trump’s campaign announcement—particularly since, according to some accounts, Trump’s campaign staff were signaling to journalists that the candidate himself did not expect, nor even set out, to win (Wolff Reference Wolff2018). But Trump’s celebrity status and his entertainment-driven campaign, unseemly or laughable as it may have been to journalists accustomed to more traditional candidacies, posed a deep dilemma: If one key threshold for defining a “serious” presidential candidacy is the candidate’s name recognition, not to mention their ability to self-fund their campaign, then Trump—long a household name and a self-reported millionaire—perhaps had to be taken seriously as an electoral contender. A show business candidate, fueled by his own celebrity and dismissive of the traditional rules of politics, posed a new challenge for legacy media.

As the data we present will suggest, the press’s urge to write off Trump persisted alongside Trump’s high entertainment value, which prompted media outlets to give him high amounts of coverage. News routines that took shape in an earlier era in the evolution of the American political and media systems did not adequately equip journalists to cover the unconventional entertainer-turned-candidate Trump. Traditional journalistic practices proved insufficient to the task of alerting the public to what seemed like a reality show but what would turn out to be an effective populist political campaign.

Evidence from Interviews

We began exploring this conundrum by considering the insights of people with a view behind the press’s curtain, obtained through interviews the lead author conducted, along with political scientist Peter Van Aelst, with twenty-four journalists, editors, and political strategists who were active in the 2016 campaign. To conduct these interviews, we contacted a range of political journalists and strategists active in the media sector, including national-level journalists (e.g., New York Times, US News & World Report, NPR), journalists with leading regional newspapers (e.g., LA Times, Sacramento Bee), foreign correspondents (e.g., The Guardian), consultants, and ghostwriters. During each interview, we asked for the contact information of relevant colleagues. We then contacted those people in turn, resulting in a non-random but diverse snowball sample of interviews. All interviews were conducted between April and September of 2017 (within ten months of the 2016 election).^{Footnote 2}

For our purposes here, we examined the interview transcripts for any discussion of the press’s treatment of Trump vis-à-vis his celebrity status. The selected quotations below are representative of the relevant comments interviewees made. Our interview data illustrate, first of all, that even though many journalists did not take the Trump campaign seriously, they felt compelled to cover it. As Mike McPhate of The New York Times described it:

The U.S. press was not alone in thinking of Trump as a joke. An editor and reporter for The Guardian, Martin Pengelly, explained the amount of news coverage Trump received this way:

The interviews also revealed how Trump’s refusal to play by the standard political playbook left journalists flummoxed. One veteran national newspaper reporter described the slow process of realizing Trump could “get away” with saying things that other politicians could not:

Political consultant Kevin Eckery described this collision between press standards and entertainment from the perspective of journalists and press outlets doing everything they were supposed to do as individual agents, yet suffering a “system failure” in the aggregate:

As some interviewees observed, the end result was that Trump was treated to different coverage than his general election opponent. One White House correspondent told us:

These themes in our interviews were echoed by other reporters who covered election 2016. The performance staged by the Trump campaign was so unconventional, so seemingly risky and beyond the pale that, as Susan Page, Washington bureau chief for USA Today, put it: “You watch Trump, it’s like a high-wire act, you want to see what he’s going to say next, and will he fall off.”^{Footnote 4} And for at least some reporters on the scene, it was difficult to convey through traditional journalistic means just how unconventional the campaign was. As Alec MacGillis, politics reporter for ProPublica put it, “What the press wasn’t able to do, and what it was almost not set up to do, was to get across the sheer ridiculousness or surreality of Donald Trump running for president.”^{Footnote 5}

Content Analysis Methods and Data

For a more explicit and quantitative test of the idea that the legacy press treated Trump as much as a “clown” as a serious politician throughout the campaign, we examined the press coverage given to Trump in two leading newspapers: the New York Times and the Washington Post. We examined these legacy newspapers, rather than television news for example, because we suspect that while the collision between traditional standards of journalism and Trump’s entertainment trappings also played out on TV—especially cable television, with its more urgent entertainment imperatives—the Times and the Post offer a hardest test case: If any news outlets would be able to navigate that collision skillfully, resisting the urge to dismiss Trump through “clown” coverage, especially as Election Day neared, it would be newspapers such as these that are well-resourced and whose reputation relies on serious journalism.

We examined coverage of Trump’s campaign in these two newspapers at eleven key moments in the campaign, selected after consulting multiple retrospective accounts of the 2016 election (e.g., Guardian Reference Guardian2016; Reuters Reference Reuters2016). These moments capture a mixture of events that reflected either traditional key moments in any election (e.g., passing the delegate threshold) or moments involving highly controversial Trump statements, aiming for a rough distribution of moments across the course of the campaign. Some of the moments reflected relatively well on Trump, others reflected more poorly, but all were major events in some way. We cannot claim these were necessarily the largest or even most important events of the campaign, but they are certainly among them. Specifically, we examined all New York Times and Washington Post stories, both straight news articles and editorial and op-ed pieces, about Trump during the 48 hours following each of the following key moments:

• • Trump’s announcement of his campaign (June 16, 2015)

• • The first GOP debate, notable not only as the first candidate forum but also for the altercation between Trump and Fox News host Megyn Kelly (August 6, 2015)

• • Trump’s call for a travel ban on Muslims (December 7, 2015)

• • The Iowa caucuses, where Trump came in second to Ted Cruz (February 1, 2016)

• • Super Tuesday II, when Trump won Florida, prompting Marco Rubio to quit the race (March 15, 2016)

• • The subsequent exit of Trump’s other close rivals, John Kasich and Ted Cruz (May 3, 2016)

• • Trump’s attainment of the requisite delegate threshold (May 26, 2016)

• • Public reaction to Trump’s slurs against the parents of slain U.S. soldier Humayun Khan (August 1, 2016)

• • The first debate between Trump and Hillary Clinton (September 26, 2016)

• • The release of the Access Hollywood tape (October 7, 2016)

• • FBI Director James Comey’s announcement that the bureau would re-open its investigation into Hillary Clinton’s emails (October 28, 2016)

Our data comprised articles that appeared in print (both hard news stories and editorials/op-eds) as well as online-only items (blog posts), producing a total of 1,517 in-print and online-only stories. Although blog posts comprised 72% of the articles in our study, the findings are strikingly similar for print and non-print stories, with no major substantive differences, and so we analyze print and online articles together.

We trained undergraduate students to code each story for four variables that offer direct insight into the question at hand.^{Footnote 6}

• • The overall tone of the story in portraying Trump (ordinal: positive, neutral, negative)

• • The overall tone of the story in portraying Trump’s electoral chances for the race the story was discussing, be it primary or general (ordinal: positive, neutral, negative)

• • Whether the story treated Trump as a “serious” political candidate (binary: 1, 0)

• • Whether the story treated Trump as a “clown” (binary: 1, 0)

The “overall tone toward Trump” variable offers face validity of our data, as it has been used in previous studies of the 2016 election. The “overall tone of Trump’s electoral chances” code allows us to test the argument that the press’s serious versus clown treatment was not necessarily in line with their estimations of his electoral chances. The final two codes (“serious” and “clown”) allow the most direct test of our argument that the press reacted to the conundrum of how to cover the Trump presidency through schismatic coverage. The “serious” code was used for articles giving substantive consideration to Trump or his campaign with discussion of the possible benefits or drawbacks of a Trump administration. Treating Trump seriously could take several forms, ranging from descriptions of the potential positive foreign policy implications of a Trump administration to discussion of Trump’s campaign as a potentially serious threat to democratic norms. The “clown” code was used for stories that explicitly used dismissive, sarcastic, and/or demeaning descriptors to describe Trump. Treating Trump as a clown included disparaging his persona, as in stories that described him as “Donald the Doofus,” a “ meandering fluff-head,” a “doughy showman,” a “celebrity bomb-thrower and provocateur,” or a “TV caricature.”

We intentionally treated the “serious” and “clown” codes as mutually-exclusive, with the aim of tracking the predominant signal audiences would have received from each article. According to the coders and our own examination of the data, there were almost no instances of a story that could have conceivably been coded as both “serious” and “clown” (had our coding allowed for such double coding). Importantly, note that the “electoral chances” variable was distinct from the “serious” and “clown” variables. Stories coded as “serious” or as “clown” could portray Trump’s chances positively or negatively (or not at all).

For a richer view of how these dueling treatments of Trump played out, we offer some examples from our data—sometimes in the words of journalists themselves or of the sources they quoted in hard news stories, sometimes in the words of editorials and op-ed pieces. Examples of “serious” coverage, meaning coverage that might reflect positively or negatively on Trump but that engaged with him substantively as a presidential candidate, include the following, presented chronologically:^{Footnote 7}

• • “It’s disturbing that someone with so little interest in the truth, who is happy to stoke xenophobic fires to advance his cause, and who seems to have little cause other than the glorification of himself, has somehow persuaded so many on the right that his presidency would make America great instead of be a disaster . . . . I don’t doubt the good faith of Trump supporters . . . . The answer isn’t to elect a ‘strong man’ who will find unnamed ways of forcing his views upon others. That’s a dangerous path, and it’s not one that the Constitution envisions or experience recommends.” (December 7, 2015)

• • “The plain truth is that a Trump presidency would not only fracture American society along ethnic, racial and, we now know, religious lines. It would also demolish American prestige on the world stage and alienate our most important allies.” (December 8, 2015)

• • “Trump is filling a vacuum left by years of inattention to voters who have been patronized and taken for granted. The fissures he exposed in the GOP will not go away.” (May 2, 2016)

• • “His evident lack of any kind of self-control has ominous implications for how he would respond as president and commander in chief to real crises and emergencies.” (August 2, 2016)

• • “For so many of us, this is the first time in our lives that we are forced to question whether our country will remain free and democratic if a major-party candidate wins.” (August 2, 2016)

• • “From this point of view, when Trump opened his first debate against Clinton with a critique of Chinese and Mexican trade and then continued into a detailed discussion of currency manipulation, taxes on exported goods and other technical issues related to international commerce, it was not the specific issues he was discussing that mattered. Rather, what Trump communicated to his supporters was his worldview and his sense of American identity.” (September 27, 2016)

As these quotes illustrate, “serious” coverage of Trump took several forms, including critiques of Trump’s policy positions, concerns about his ability to govern, and the implications of his candidacy for democracy, red flags about his divisive statements, and also recognitions of the appeal of his candidacy for voters.

Examples of “clown” coverage that engaged with Trump as the butt of a joke include the following:

• • “Donald Trump entered Iowa as a big-talking strip of New York sirloin. He exited as a second side order of high corn. Still, he remains undiminished as satiric red meat. The steely knives of America’s political cartoonists are close by, primed for serrated illustration.” (February 3, 2016)

• • “Before he left the race, Senator Ted Cruz shared a parting thought: ‘If anyone has seen the movie Back to the Future, Part II, the screenwriter says that he based the character Biff Tannen on Donald Trump—a caricature of a braggadocious, arrogant buffoon who builds giant casinos with giant pictures of him wherever he looks.’ (It is, indeed, true that Mr. Trump served as inspiration for the character.)” (May 5, 2016)

• • “Donald Trump should be a late-night comedian’s dream—a bombastic reality TV star running for president with a distinctive vernacular (‘yuge,’ ‘loser’) and physical features (hair swoop, suspicious tan) that practically write their own jokes.” (May 27, /2016)

• • “If you admire a good story, fine, but why hire an amateur like Trump? Why not J.K. Rowling, who gave us Harry Potter? This would get millions of young people interested in politics, which could use some wizards and special effects, Supreme Court justices flying through the air, robes fluttering.” (September 27, 2016)

• • “With his flights of fancy and bouffant of weirdly styled hair, he now seems to be more desperate Bond villain than presidential contender.” (October 28, 2016)

Content Analysis Findings

Our content analysis shows the same high levels of negative coverage of Trump that other analyses reveal (Patterson Reference Patterson2016a, Reference Patterson2016b, Reference Patterson2016c), offering face validity to our data (see Appendix). Our purpose here, however, is to examine the idea that the press covered Trump in a disjointed fashion—sometimes taking him seriously, sometimes treating him as a clown, in ways that cannot simply be explained by their estimation of his electoral chances at the given moment.

We begin with figure 1, which shows the percentage of each moment’s coverage that treated Trump as a “serious” candidate and the percentage that treated him as a “clown.” This figure offers a direct demonstration of our argument that the journalistic conundrum we have described led the press to cover Trump in a schismatic fashion. We see that, at least in these two leading newspapers, coverage was nearly evenly divided between the two modes of coverage. Moreover, the tension between these two versions of Trump in the press was never reconciled, even as Trump secured the GOP nomination. Instead, we see a nearly even give-and-take, with the press framing Trump seesaw-style as someone to laugh at and someone to heed throughout the remainder of the election season.

Figure 1 Coverage of Trump as presidential material vs. clown

As noted earlier, there were no significant differences (at least with respect to the variables we consider here) between print and non-print (i.e., blog) coverage. Overall, 16% of print articles and 19% of non-print articles were coded as treating Trump “seriously.” A two-sample t-test confirms that this difference is not statistically significant (p = 0.242). Likewise, 17% of print articles and 15% of non-print articles were coded as treating Trump like a “clown”—a non-significant difference (p=0.526).

Given how colorful some of the language describing Trump is in the examples provided here, it seems plausible that the “clown” coverage appeared primarily on the editorial pages rather than in “straight” news stories. To explore this possibility, we further divided the articles into editorial/op-ed content and non-editorial content (including everything from Page A1 articles to features stories). Here, we find further evidence of the journalistic conundrum. We might have expected non-editorial content to contain more “serious” coverage, while expecting editorial content, where journalists are freer from strict objectivity norms, to contain more “clown” coverage. However, 23% of editorials were coded as “serious” compared to 13% of non-editorials (a statistically significant difference, p=0.017). And 24% of editorials were coded as “clown” coverage compared to 14% of non-editorials (again, a statistically significant difference, p=0.010). In other words, while dismissive coverage of Trump was more likely on the editorial pages, more serious coverage of Trump was found there as well, while clown coverage was sprinkled throughout “straight” news stories as well as editorials.

Figure 2, displaying the percentage of each moment’s coverage that cast Trump’s electoral chances in a positive or negative light, supports our argument that disjointed press coverage of Trump (as figure 1 documents) reflects the collision of journalism standards with celebrity aesthetics, rather than the press simply not thinking Trump had a chance of winning. We see here that even as Trump passed the delegate threshold in late May, press portrayals of his electoral chances became less positive—presumably as the press looked ahead toward the general election. When Trump attacked the Gold Star Khan family during the Democratic convention, these newspapers’ assessments of his electoral viability dropped sharply, recovering only when FBI Director James Comey released his letter a week before Election Day announcing the re-opening of the Hillary Clinton email investigation. And even though this development gave Trump an unexpected late-game advantage, media coverage of his chances became only slightly more positive than negative.

Figure 2 Coverage of Trump’s electoral chances

Note: Data includes signals about Trump’s positive/negative electoral chances for the GOP primary (prior to the reaching the delegate threshold) as well as his electoral chances for the general election (throughout).

We can test the idea that “serious versus clown” coverage was orthogonal to “positive versus negative electoral changes” more directly by looking at the co-occurrence of “positive electoral chance” codes (figure 2) with “serious” codes (figure 1) and, likewise, at the co-occurrence of “negative electoral chance” codes (figure 2) with “clown” codes (figure 1). The story-level correlation of positive electoral chance codes with serious treatment codes is a mere 0.16, and the story-level correlation of negative electoral chance codes with clown treatment codes is 0.11. Regardless of the press’s fluctuating estimates of Trump’s chances of winning the presidency, press coverage was schismatic between serious and clown portrayals.