Learning From the 2016 U.S. General Election Presidential Debates

Abstract

Over four decades of research show that presidential debates make it possible for viewers to learn about the issues and form opinions about the character of the candidates running for office, but whether this long-lived finding would hold true in 2016 was open to doubt. In this article, we draw on panel survey data gathered from viewers of the first and third 2016 general election presidential debates to document across time changes in knowledge of issue positions, beliefs about candidate qualifications and character, and perceptions that, if elected, a candidate would or would not pose a threat to the well-being of the nation. We find that viewers of postdebate coverage of the 2016 general election debates exhibited increased knowledge, but viewing the debates or the postdebate coverage had little effect on assessment of either candidate qualifications or perceptions of whether candidates would threaten the nation’s well-being, if elected.

Keywords

presidential debates 2016 presidential elections public opinion candidate evaluation panel study

Overview

The 2016 general election presidential debates were the most viewed in U.S. history.¹ On stage were Hillary Clinton, the first female nominee of a major political party, and Donald Trump, the first major party nominee to step before the podium with neither elected nor military leadership on his resume. Where one was a billionaire businessman and showman extraordinaire with a successful television franchise to his credit and buildings around the world emblazoned with his name, the other was the only candidate in the television era to bring both executive and legislative branch experience to the table. Where he was shadowed by accusations that his tenuous command of fact, disposition to lash out at opponents with ad hominem, and lewd comments about women disqualified him from high office, she was plagued by allegations that her use as secretary of state of a private e-mail server put the country at risk and merited prosecution.

At the same time, the surprises that contextualized the 2016 debates were unprecedented. Prior to the first one, the Republican nominee alleged that his opponent’s supporters had rigged the debate process. The second debate was shadowed by the release of a tape on which Trump was heard making comments about women cast by some as an admission that he had engaged in “sexual assault.” To counterbalance the resulting coverage and reframe that issue heading into the debate, the Trump campaign not only issued debate invitations to women who had allegedly been sexually assaulted by Hillary Clinton’s husband, former president Bill Clinton, but also tried to seat them in the space allocated to the Trump family, at the front of the auditorium, a placement that ensured that they would be able to confront the Democratic party nominee’s spouse. Moreover, the final debate was contextualized by the Republican nominee’s suggestion that, were he not the victor, he might not accept the outcome of the election.

Although past debates have reliably increased viewers’ knowledge of candidate positions discussed in the debate (Gottfried, Hardy, Winneg, & Jamieson, 2016), the unique nature of these circumstances and of the debate exchanges themselves raised the possibility that that finding would not prove durable in 2016. Instead it was plausible to suppose that the theatrics that characterized the closing moments of the first and third presidential debates, on which we focus here, would wipe out any issue learning that might otherwise have occurred.

Both sets of concluding exchanges were unprecedented and memorable. Near the conclusion of the first debate, Clinton noted,

One of the worst things he [Trump] said [about women] was about a woman in a beauty contest. He loves beauty contests, supporting them and hanging around them. And he called this woman “Miss Piggy.” Then he called her “Miss Housekeeping,” because she was Latina. Donald, she has a name: Her name is Alicia Machado.

Trump interrupted:

“Where did you find this? Where did you find this?”

Clinton:

“She has become a U.S. citizen, and you can bet she’s going to vote this November.” (Blake, 2016a)

In the third and final debate, the unexpected moment was initiated by a question from the moderator, Fox’s Chris Wallace, about whether the Republican nominee would accept the outcome of the election if he lost.

When Trump refused to answer the question, Wallace persisted,

There is a tradition in this country, in fact one of the prides of this country is the peaceful transition of power, and that no matter how hard-fought a campaign is, that at the end of the campaign that the loser concedes to the winner. Not saying that you’re necessarily going to be the loser or the winner. But that the loser concedes to the winner, and that the country comes together in part for the good of the country. Are you saying you’re not prepared now to commit to that principle?

In response, Trump declared, “I’ll tell you at the time. I’ll keep you in suspense, okay?” (Blake, 2016a).

The general election presidential debates of 2016 also brought two candidates with historically low perceived trustworthiness face-to-face not only with each other but with the most polarized electorate of the television age. This sad state of affairs invites the question, did the debates alter perceptions of the candidates’ fitness for office or the perceived likelihood that the presidency of either would threaten the well-being of the country, the latter a question not plausible in years past.

In this article, we draw on survey data gathered from viewers of the first and third 2016 general election presidential debate to document across time changes in knowledge of issue positions, beliefs about candidate qualifications and character, and perceptions that, if elected, a candidate would or would not pose a threat to the well-being of the nation. Our research focuses on three questions: (a) Do debate viewers become more knowledgeable after viewing the presidential debates and the coverage following them? (b) Is there a change in the proportion of those who say the candidates are qualified to be president following the debates? and (c) Does debate viewing contribute to voters’ perception that, if elected, the candidates would pose a threat to the nation’s well-being?

To address these questions, we recruited a probability-based panel of U.S. adults to watch the first debate and the third presidential debate. In the next section, we set out the study hypothesis and research questions in greater detail.

Hypothesis and Research Questions

Knowledge Gain From Presidential Debates

Since the 1976 presidential election, televised debates have been a regular fixture of the process of electing a U.S. president. Across the decades, researchers have found a link between debate viewing and higher political knowledge (e.g., Benoit & Hansen, 2004; Chaffee, 1978; Gottfried et al., 2016; Holbert, Benoit, Hansen, & Wen, 2002). Indeed, a meta-analysis on data from 18 studies of the 1976 to 2000 debates (total N = 7,202) found a “mean weighted correlation coefficient of 0.256 between debates and political knowledge” (Benoit, Hansen, & Verser, 2003). Included in that meta-analysis were studies that found that viewers learned about issue positions that were discussed in the debates (Jamieson & Adasiewicz, 2000; Zhu, Milavsky, and Biswas, 1994). Consistent with these findings, we hypothesize:

Hypothesis 1: Viewers exposed to the 2016 presidential debates will exhibit greater accurate knowledge of the issues and matters discussed in the debates themselves.

An important element of the presidential debates is the news coverage following them. There is evidence that viewing postdebate coverage may affect what voters learn (Jamieson & Adasiewicz, 2000; Johnston, Hagen, and Jamieson, 2004). If postdebate coverage includes substantive coverage of issue positions, the effect on knowledge could be positive. If instead, the coverage is tactical, we would expect it to decrease learning and activate cynicism (Cappella & Jamieson, 1997). Therefore, we ask,

Research Question 1: How, if at all, did postdebate viewing of broadcast or cable news coverage affect learning?

Debates and Assessment of Whether a Candidate Is Qualified to Be President or Considered a Threat to the Nation

While the meta-analysis conducted by Benoit et al. (2003) found no significant effect of debate viewing on perceived candidate competence, the 2016 election pitted a candidate with many years of government experience against one who had never run for or held an elected office. The contest also involved charges by each nominee that the other was not only not up to the office but would pose a threat to the nation. So for example, on November 1, 2016, Donald Trump disparaged Clinton by tweeting: “Look at the way Crooked Hillary is handling the e-mail case and the total mess she is in. She is unfit to be president. Bad judgement(sic)!”

In a similar vein, in October, 2016 Clinton told an audience,

I mean, his latest Twitter meltdown is unhinged, even for him. It proves, yet again, that he is temperamentally unfit to be president and commander-in-chief. I’ve said it before and I’ll say it again, a man who can be provoked by a tweet should not be anywhere near the nuclear codes. (Luscombe, 2016)

Indeed, throughout the 2016 campaign, both Trump and Clinton were accused of actions calling into question their competence, integrity, honesty, and temperament. In our study, we did not capture specific evaluations of these individual traits, but rather recorded perceptions of whether either of the candidates would threaten the well-being of the nation. Since presidential debates provide a unique opportunity for the electorate to see the candidates not only tackle the issues but show their habits of mind and temperament, we ask,

Research Question 2: Did viewing a debate change assessment of candidate qualifications?

Research Question 3: Did viewing a debate change assessment of whether the viewer sees either of the candidates as a threat to the nation’s well-being?

Method

Our data come from the Annenberg Public Policy Center’s (APPC) 2016 Presidential Debate Study, a four-wave national online probability- based panel composed of those recruited to view the presidential debates. Surveying took place before and after the September 26 and October 19 presidential debates. The main sample for this survey consisted of U.S. adults, 18 years and older, English and Spanish language survey takers. For the first debate survey GfK, under contract to APPC, surveyed 6,115 eligible respondents recruited from its Knowledge Panel who agreed to watch the first debate (Wave 1) and 5,145 of those respondents (84.1%) reported watching the debate and completed the postdebate survey (Wave 2). For the third debate survey, we surveyed 6,323 eligible respondents recruited from GfK’s Knowledge Panel prior to the final debate (Wave 3). Among those completing Wave 3, 5,241 (82.9%) said they watched the third debate and completed the postdebate survey (Wave 4).² Seventy-two percent of the panelists (3,776) viewed both the first and second debate.

In order to test the effect of campaign learning, and control for priming effects, half of those in the sample in the predebate surveys were exposed to the candidate knowledge questions and half were not exposed. Respondents in the unexposed sample were asked only about political behavior and intention and whether they would view the debate. Respondents in the exposed group were asked those questions as well as the knowledge ones. Following the debate, those in both conditions were asked the candidate knowledge questions which enabled us to measure whether candidate learning increased as a result of prior exposure to the knowledge questions. As the results will show, we found no evidence of priming. In total, 2,520 of the 5,145 Debate 1 viewers were randomly exposed to the knowledge questions before and after the first debate and 2,546 of the 5,241 debate viewers were randomly exposed to the knowledge questions before and after the third debates.

The postdebate surveys were opened to respondents on the day following the debates, so they could read and view postdebate analysis immediately following the debate or the next day. The field period was closed 48 hours after it was opened to limit the effect of postdebate news exposure.

Knowledge Measures

Our research interest is in ascertaining whether panelists’ knowledge of the two major party candidates’ stances on campaign issues changed following viewing of the first and third presidential debates. Consequently, we relied on a series of knowledge items discussed in those debates, respectively. Prior to each debate wave (Waves 1 and 3), we attempted to anticipate what the debate moderators would ask and what the candidates would discuss. During the predebate waves, panelists responded to some questions which were not discussed during the debates. These were dropped from the postdebate surveys. In Debate 1, six items were included in both the pre- and postdebate, while in Debate 3, five were asked before and after that final debate. Two questions overlapped from the first and third debates—support for the Trans Pacific Partnership (TPP) and favoring a minimum wage increase. While most knowledge items were uncontested in the debates—that is, the rival candidate did not question the other’s stated position, two positions captured by the survey were challenged by the opposing candidate. These were the support for the TPP and support for the Iraq war from the beginning (see Table 1). Additionally, four items on the Wave 1 predebate survey were dropped from the Wave 2 postdebate follow-up survey because results showed that predebate knowledge of the candidate positions was very high. These included correct knowledge about which candidate would build a wall at the Mexican border (94.7% Trump), which candidate would temporarily suspend immigration from areas of the world where there is a proven history of terrorism against the United States, Europe, or our allies (86.0% Trump), which candidate favors repealing and replacing the Affordable Care Act, also known as Obamacare (82.4% Trump), and which candidate would give a path to U.S. citizenship to those brought to the U.S. illegally as children, and who are sometimes called “Dreamers” (74.2% Clinton).

Table 1.

Knowledge Questions Asked Before and After Debates 1 and 3.

Question wording^a: Please answer the following questions by telling us as best you can what the candidate says he or she will do if elected president.	Correct answer	Debate
Which candidate favors an increase in the minimum wage? Hillary Clinton, Donald Trump, both, or neither?	Clinton	1 and 3
Which candidate says that climate change is an urgent issue? Hillary Clinton, Donald Trump, both, or neither?	Clinton	1
Which candidate supports a federally funded family leave program that covers at least two-thirds of pay for up to 12 weeks for a mother or a father after the delivery of a child? Hillary Clinton, Donald Trump, both, or neither?	Clinton	1
Which candidate favors increasing taxes on the wealthiest Americans? Hillary Clinton, Donald Trump, both, or neither?	Clinton	1
Which candidate supports allowing students to attend in-state, 4-year public colleges tuition-free, if they came from families with annual incomes of under $125,000? Hillary Clinton, Donald Trump, both, or neither?	Clinton	1
Which candidate’s plans would increase the national debt most? Hillary Clinton, Donald Trump, both, or neither?	Trump	3
Which candidate would consider not automatically defending NATO allies that do not pay their full share of defense costs? Hillary Clinton, Donald Trump, both, or neither?	Trump	3
Contested
Which candidate supports a trade agreement known as the Trans Pacific Partnership, or TPP? Hillary Clinton, Donald Trump, both, or neither?	Neither (but Clinton supported as Secretary of State)	1 and 3
Which candidate opposed the Iraq war from the beginning? Hillary Clinton, Donald Trump, both, or neither?	Neither (but Trump claims he always opposed)	3

The order of the names, Hillary Clinton and Donald Trump, were rotated to minimize order bias.

Viewing Postdebate Coverage

Not only did the survey measure debate viewing but it also asked whether viewers watched postdebate coverage immediately following the debate or the next morning. In our first research question, we ask whether knowledge gains are different for those who reported viewing the postdebate coverage compared with those who did not. Postdebate coverage viewing is based on the responses to the question: Did you watch, follow, or listen to any of the news discussions right after the debate or the next morning? About two thirds of debate viewers reporting attending to some postdebate coverage; 65.4% after the first debate and 64.3% after the third debate. Those who reported attending to postdebate coverage were better educated and older than those who said they did not follow postdebate coverage. Finally, it is important to note that those more likely to watch the entire debate were also more likely to say they tuned in to postdebate coverage. Seven in 10 (71.0% Postdebate 1 and 70.7% Postdebate 3) who reported watching the entire debate, also said they watched, listened to, or followed postdebate coverage. Only about half (51.3% Postdebate 1 and 50.2% of Postdebate 2), who said they watched just some or most of the debates, also tuned in to or followed postdebate coverage.

Qualified to be President and Threat to the Nation’s Well-Being

For Research Questions 3 and 4, our goal is to assess changes, if any, in how debate viewers evaluate the candidates on two attributes: qualification for the position, and perceived threat to the well-being of the nation. In both debates, the survey asked: Which of the two major party candidates is qualified to be president of the United States: Hillary Clinton, Donald Trump, neither, or both equally (pre- and postdebates)? In the pre- and post–third debate, only, the survey asked about the impact each candidate would have on the nation’s well-being: If Trump/Clinton wins the election, how much, if at all, would his or her presidency threaten the nation’s well-being? A great deal, a good amount, not much or not at all.

Analysis

For the tests of the knowledge hypothesis and research questions, we ran paired samples t-tests after converting the categorical responses into dichotomous correct (1) and incorrect (0). The percentage correct was treated as a mean score for pre-post comparison. We conducted a logistic regression to control for education to further answer the questions about the effect of viewing post-debate coverage. We conducted paired samples t-tests to measure any changes in the evaluation of presidential qualification and potential threat to the nation’s well-being.

Results

Political Knowledge Gain

We found support for Hypothesis 1. In the first debate, we found a significant and positive knowledge gain among viewers on all issue stances. The greatest increase in the proportion who correctly identified an issue stance occurred for the question asking which candidate “favors increasing taxes on the wealthiest Americans.” The correct answer is Hillary Clinton who said in the debate:

What I have proposed would be paid for by raising taxes on the wealthy, because they have made all the gains in the economy. And I think it’s time that the wealthy and corporations paid their fair share to support this country. (Blake, 2016a)

The same information was found in a question posed by the debate moderator, Lester Holt, who stipulated that “Secretary Clinton, you’re calling for a tax increase on the wealthiest Americans. I’d like you to further defend that. And, Mr. Trump, you’re calling for tax cuts for the wealthy. I’d like you to defend that” (Blake, 2016a). The smallest increase in knowledge occurred among those who correctly identifying the candidates’ murkier position on the TPP. While significantly higher, the mean difference between viewers’ responses before and after the debate was 4.6%. At the time of the debate, neither candidate publicly supported TPP. However, as secretary of state, Hillary Clinton was on record as doing so (Bradner, 2016). During the debate, about 21 minutes in, Trump suggested that her position might not be firmly held when he reminded viewers that she had once called the TPP the “gold standard” of trade deals,

TRUMP: You called [the Trans Pacific Partnership] the gold standard of trade deals. You said it’s the finest deal you’ve ever seen. . . . And then you heard what I said about it, and all of a sudden you were against it. (Blake, 2016a)

As Table 2 shows, only about a quarter knew the candidates’ TPP position. However, while there was significant knowledge gain, there also was a significant increase (13.5%) in the number who surmised, incorrectly, that Clinton supported TPP (37.8% to 51.3%), a possible effect of Trump featuring her past position. On the other issues, increases in correct knowledge were much higher than for TPP.

Table 2.

Pre- and Postdebate 1 Knowledge Responses With Significant Mean Differences Noted.

Question wording	Correct answer	% Correct Predebate 1 (n = 2,520)	% Correct Postdebate 1 (n = 2,520)	% Correct Postdebate 1; Unexposed Predebate (n = 2,625)
Question wording	Correct answer	a	b	c
Which candidate favors an increase in the minimum wage?	Clinton	60.1	74.4***	73.6
Which candidate says that climate change is an urgent issue?	Clinton	67.2	83.5***	80.4
Which candidate supports a federally funded family leave program that covers at least two thirds of pay for up to 12 weeks for a mother or a father after the delivery of a child?	Clinton	47.9	67.4***	63.3
Which candidate favors increasing taxes on the wealthiest Americans?	Clinton	60.5	85.8***	85.4
Which candidate supports allowing students to attend in-state, 4-year public colleges tuition–free, if they came from families with annual incomes of under $125,000?	Clinton	57.6	79.6***	71.2
Contested
Which candidate supports a trade agreement known as the Trans Pacific Partnership, or TPP?	Neither (But Clinton supported as Secretary of State)	17.6	23.2***	22.2

***

p < .001 for predebate exposed (a) versus postdebate exposed (b).

In the third debate, knowledge gains were somewhat modest relative to those observed in the first one. The smaller increases were evident in both the questions carried over from the first debate and the ones asked for the first time prior to and following the final debate. Keeping in mind that 72% were in both samples, one would expect knowledge gains to be less for the repeated stances, but not small for the newer ones. We compared both those who were in the prior postdebate wave to those who were only in the third and fourth waves and found no significant difference between each group’s responses. Correctly identifying Trump as the candidate whose plans would increase the national debt most showed the greatest increase pre- to postdebate. During the debate, moderator Chris Wallace cited the nonpartisan Committee for a Responsible Federal Budget’s assessment that Clinton’s budget proposals would increase the debt from 77% of GDP to 86% and Trump’s proposals would increase it to 105% of GDP.³ His question to both candidates was, “Why are both of you ignoring this problem [national debt]?” Trump pushed back at the estimates,

Well, I say they’re wrong, because I’m going to create tremendous jobs. And we’re bringing GDP from, really, 1 percent up to 4 percent. And I actually think we can go higher than 4 percent. I think you can go to 5 percent or 6 percent. And if we do, you don’t have to bother asking your question, because we have a tremendous machine. (Blake, 2016b)

In response to Wallace’s question, Clinton said her proposed plans would not “add a penny to the national debt” and she said at one point, “but what Donald is proposing with these massive tax cuts will result in a $20 trillion additional national debt” (Blake, 2016b).

Finally, Tables 2 and 3 show no apparent effect of priming or preexposure to the knowledge questions. The results for those exposed to the knowledge questions only after the debates were similar to those exposed both before and after the debate.

Table 3.

Pre- and Postdebate 3 Knowledge Responses With Significant Differences Noted.

Question wording	Correct answer	% Correct Predebate 3 (n = 2,546)	% Correct Postdebate 3 (n = 2,546)	% Correct Postdebate 3; Unexposed predebate (n = 2,695)
Question wording	Correct answer	a	b	c
Which candidate favors an increase in the minimum wage?	Clinton	68.0	74.5***	77.3
Which candidate’s plans would increase the national debt most?	Trump	34.4	46.9***	46.1
Which candidate would consider not automatically defending North Atlantic Treaty Organization allies that do not pay their full share of defense costs?	Trump	71.2	80.4***	77.9
Contested
Which candidate supports a trade agreement known as the Trans Pacific Partnership, or TPP?	Neither (but Clinton supported as Secretary of State)	22.1	26.0***	23.4
Which candidate opposed the Iraq war from the beginning?	Neither (But Trump claims he always opposed)	38.3	38.6^ns	38.4

***p < .001 for predebate exposed (a) versus postdebate exposed (b).

Effect of Viewing Postdebate Coverage on Knowledge

In the first debate, results show that knowledge increased among reported viewers regardless of whether they indicated that they had viewed postdebate coverage. In the third debate, increases in knowledge were significant for those who said they tuned in to or followed postdebate coverage, except for the candidates’ position on supporting the Iraq war. For those who said they did not tune in to or follow post–third debate coverage, results show significant knowledge gains for all the items, except for Iraq war support and TPP support. While both followers and nonfollowers of postdebate coverage showed an increase in knowledge, our analysis shows that those who followed postdebate coverage had higher levels of knowledge compared with those who did not follow postdebate coverage (see Tables 4 and 5).

Table 4.

Pre- and Postdebate 1 Knowledge Responses With Significant Differences Noted—Viewed Postdebate Coverage Versus Did Not View Postdebate Coverage.

Question wording	Correct response	Viewed postdebate coverage (N = 1,707)		Did not view postdebate coverage (N = 810)
		% Correct Predebate 1	% Correct Postdebate 1	% Correct Predebate 1	% Correct Postdebate 1
		a	b	c	d
Which candidate favors an increase in the minimum wage?	Clinton	65.3	78.2***	50.4	67.4***
Which candidate says that climate change is an urgent issue?	Clinton	72.9	87.1***	56.7	76.9***
Which candidate supports a federally funded family leave program that covers at least two-thirds of pay for up to 12 weeks for a mother or a father after the delivery of a child?	Clinton	49.8	68.9***	44.5	64.5***
Which candidate favors increasing taxes on the wealthiest Americans?	Clinton	64.6	89.3***	52.9	79.4***
Which candidate supports allowing students to attend in-state, 4-year public colleges tuition-free, if they came from families with annual incomes of under $125,000?	Clinton	62.5	83.1***	48.7	73.0***
Contested
Which candidate supports a trade agreement known as the Trans Pacific Partnership, or TPP?	Neither (but Clinton supported as Secretary of State)	22.2	27.7***	9.2	14.9***

***

p < .001 for predebate versus postdebate (a vs. b; c vs. d).

Table 5.

Pre- and Postdebate 3 Knowledge Responses With Significant Mean Differences Noted—Viewed Postdebate Coverage Versus Did Not View Postdebate Coverage.

Question wording	Correct response	Viewed postdebate coverage (N = 1,754)		Did not view postdebate coverage (N = 783)
		% Correct Predebate 3	% Correct Postdebate 3	% Correct Predebate 3	% Correct Postdebate 3
		a	b	c	d
Which candidate favors an increase in the minimum wage?	Clinton	73.6	77.4***	56.7	68.6**
Which candidate’s plans would increase the national debt most?	Trump	39.5	53.4***	24.4	34.3***
Which candidate would consider not automatically defending NATO allies that do not pay their full share of defense costs?	Trump	77.2	84.0***	59.6	73.1***
Contested claim
Which candidate supports a trade agreement known as the Trans Pacific Partnership, or TPP?	Neither (But Clinton supported as Secretary of State)	26.1	30.3***	14.4	17.4^ns
Which candidate opposed the Iraq war from the beginning?	Neither (but Trump claims he always opposed)	41.8	42.6^ns	31.1	30.7^ns

***p < .001, **p < .01 for predebate versus postdebate (a vs. b; c vs. d).

We recognize the research that suggests that people overreport their media and presidential debate consumption (e.g., Prior, 2009, 2012). However, since we examine the effect of debate viewership on knowledge, overreported debate viewership would understate effects because those who falsely reported viewing the debates were analyzed as viewing them, diminishing the actual knowledge effects of those who accurately reported viewing the debates.

Was the knowledge gain by those following postdebate coverage a result of the act of following the debate or was it due to higher education? While we generally found that viewers learn from watching the debates, attending to postdebate coverage was a stronger predictor of such learning. In a logistic regression model, following the coverage after the debate was a significant predictor of post-debate knowledge in the presence of education controls for all items except three across both debates. Education, rather than following postdebate coverage, was a significant predictor for knowledge that Clinton favored an increase in the minimum wage in the first and third debates. In the first debate, neither education nor following postdebate coverage was a significant predictor of knowing that Hillary Clinton supported federally funded family leave. So, education may account for some of the knowledge gain, but the data suggest that following postdebate coverage acted to enhance knowledge (see the appendix for regression tables).

Debate Viewing and Evaluation of Candidate Qualifications

Debates offer the viewer the opportunity to directly assess whether each candidate is qualified to serve as president. Interviewing before and after each encounter showed little movement in viewers’ assessments on whether each of the candidates is qualified to serve. In each of the waves, we asked, Which of the two major party candidates is qualified to be president of the United States: Hillary Clinton, Donald Trump, neither or both equally? Going into the debates Clinton was seen by a 2 to 1 margin as the more qualified candidate (54% vs. 26% in the first debate and 52% vs. 24% in the third debate). Assessments changed little between the first and third debates and our data show minor change in those numbers immediately following the two debates. After the first debate, the proportion of those who saw Clinton as qualified increased significantly by 3 points, while the assessment of Trump’s qualifications was unchanged. Following the final debate, viewers’ evaluation of Clinton as qualified rose significantly, though slightly, to the level observed just before the first debate. Trump viewers’ assessment that he is qualified also increased significantly by two points. The debates did little to alter viewers’ assessment of whether both, neither, or one or the other of the candidates was qualified for president.

Debate Viewing and Evaluation of Candidate as Threat to the Nation’s Well-Being

The question of whether one or the other candidate, if elected, would threaten the nation’s well-being was part of the survey only before and after the final debate. Prior to that face-off, half of the debate viewers (50.3%) saw Clinton and nearly two-thirds (63.4% great deal/good amount) viewed Trump as a threat to the nation’s well-being, if elected president. Viewing the debate had no effect on the assessment of either candidate with 50.2% of viewers saying Clinton would be a threat and 63.2% viewing Trump as threat following the debate.

Discussion

Because of the unprecedented numbers of viewers, the 2016 presidential debates can be considered what Dayan and Katz (1994) call historic media events, attracting nearly 100 million viewers in the United States. Therefore, the debates presented an opportunity to inform a vast number of voters about the candidates. This study shows that viewers of the first and third 2016 general election presidential debates increased their knowledge of the candidates’ issue stances discussed in the debates. Watching and then following the postdebate coverage immediately or the next day was a strong predictor of postdebate knowledge gain. At the same time, Trump’s attack, in the first debate, on Clinton’s position on the Trans Pacific Partnership may have had the effect of increasing inaccurate information since we saw a significant increase in the number of viewers who believed that Hillary Clinton supported TPP. Indeed, the “Gold Standard” clip in the first debate was the third most broadcast by the major cable news channels and morning news programs on the day following the debate (Internet Archive, 2016). In the final debate, TPP and Trans-Pacific Partnership were mentioned once by each candidate, 47 minutes into the debate, and were not among the top clips rebroadcast following the debate. We found no significant increase among third debate viewers who believed Clinton supported TPP, and just a small increase in the difference in correct response between those who did and did not watch postdebate coverage. In both debates at least half believed the incorrect or unclear claim. For another contested claim about support for the Iraq war from the beginning, we found no knowledge gain following the third debate. This topic came up near the end of the debate in the 75th minute (Internet Archive, 2016). Increases in knowledge after the third debate were lower compared with such gains after the first debate for questions repeated in all four waves. The smaller increases were possibly due to the nature of the panel sample where 72% participated in both the post–first and post–third debate surveys. This suggests panelists retained their knowledge from the prior debate, with the third debate affirming that knowledge. Yet, those only participating in Waves 3 and 4 (pre– and post–third debate) showed similar levels of knowledge.

Importantly two of the questions on which significant change in knowledge occurred (i.e., increasing taxes on wealthy Americans—Clinton, and increase the debt most—Trump) were ones in which the accurate answer could be found both in a moderator’s question and in a candidate’s answers. This finding suggests that value of moderator’s embedding clear accurate information in the questions themselves. A limitation of this study is the lack of comparison to nonviewers of the debates; therefore, we cannot definitively say debate viewing led to increased knowledge, but following postdebate coverage was a predictor of increased knowledge.

The debates neither changed assessments of whether either candidate was qualified for president nor altered perceptions that either candidate, if elected, would be a threat to the nation, as a whole. Perhaps a full battery of personality traits and qualification measures would lead to a different conclusion, but our findings are consistent with Benoit et al.’s (2003) meta-analysis.

Footnotes

Appendix

Debate 3: Effect of Viewing Postdebate Coverage on Postdebate Knowledge Controlling for Education (N = 2,537).

	Model 1 (No education control) Exp(B)	Model 2 (Education control) Exp(B)
Which candidate supports TPP?
Predebate knowledge	48.45***	43.42***
Followed postdebate coverage	1.68***	1.65***
Education	NA	1.19***
Pseudo R²	0.56	0.57
Which candidate favors increase in minimum wage?
Predebate knowledge	15.04***	14.49***
Followed postdebate coverage	1.06^ns	1.04^ns
Education	NA	1.14***
Pseudo R²	0.37	0.38
Which candidate’s plan would increase debt most?
Predebate knowledge	32.11***	31.40***
Followed postdebate coverage	1.84***	1.82***
Education	NA	1.12***
Pseudo R²	0.5	0.51
Which candidate opposed Iraq war from beginning?
Predebate knowledge	23.56***	21.64***
Followed postdebate coverage	1.69***	1.66***
Education	NA	1.15***
Pseudo R²	0.5	0.51
Which candidate would consider not defending NATO allies that do not pay fair share?
Predebate knowledge	8.59***	7.94***
Followed postdebate coverage	1.46**	1.45**
Education	NA	1.18***
Pseudo R²	0.23	0.25

p < .05. **p < .01. ***p < .001.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Kenneth Winneg is the Managing Director of Survey Research at the Annenberg Public Policy Center of the University of Pennsylvania. His research focuses on political communication, emphasizing both online and offline political participation.

Kathleen Hall Jamieson is the Elizabeth Ware Packard Professor of Communication and Director of the Annenberg Public Policy Center at the University of Pennsylvania.

References

Benoit

W. L.

Hansen

G. J.

(2004). Presidential debate watching, issue knowledge, character evaluation, and vote choice. Human Communication Research, 30, 121-144.

Benoit

W. L.

Hansen

G. J.

Verser

R. M.

(2003). A meta-analysis of the effects of viewing U.S. presidential debates. Communication Monographs, 70, 335-350.

Blake

(2016, September 26). The final Trump-Clinton debate transcript, annotated. The Washington Post. Retrieved from https://www.washingtonpost.com/news/the-fix/wp/2016/09/26/the-first-trump-clinton-presidential-debate-transcript-annotated/?utm_term=.ba400f7c3093

Blake

(2016, October 19). The final Trump-Clinton debate transcript, annotated. The Washington Post. Retrieved from https://www.washingtonpost.com/news/the-fix/wp/2016/10/19/the-final-trump-clinton-debate-transcript-annotated/?utm_term=.90dc89b448b8

Bradner

(2016, July 21). Clinton’s TPP controversy: What you need to know. CNN Politics. Retrieved from http://www.cnn.com/2016/07/27/politics/tpp-what-you-need-to-know/

Callegro

DiSogra

(2008). Computing response metrics for online panels. Public Opinion Quarterly, 72, 1008-1032.

Cappella

J. N.

Jamieson

K. H.

(1997). Spiral of cynicism: The press and the public good. New York, NY: Oxford University Press.

Chaffee

(1978). Presidential debates: Are they helpful to voters? Communication Monographs, 45, 330-346.

Clinton’s misleading debate claims. (2016, October 20). FactCheck.org . Retrieved from http://www.factcheck.org/2016/10/clintons-misleading-debt-claims/

10.

Dayan

Katz

(1994). Media events: The live broadcasting of history. Cambridge, MA: Harvard University Press.

11.

First presidential debate of 2016 draws 84 million viewers. (2016, September 27). Nielsen. Retrieved from http://www.nielsen.com/us/en/insights/news/2016/first-presidential-debate-of-2016-draws-84-million-viewers.html

12.

Gottfried

J. A.

Hardy

B. W.

Winneg

K. M.

Jamieson

K. H.

(2016). All knowledge isn’t created equal: Knowledge effects and the 2012 presidential debates. Presidential Studies Quarterly, 44, 389-409.

13.

Holbert

R. L.

Benoit

Hansen

Wen

(2002). The role of communication in the formation of an issue-based citizenry. Communication Monographs, 69, 296-310.

14.

Internet Archive. (2016). 2016 Presidential debates: Giving the public tools to track and analyze debates. Retrieved from http://politicaladarchive.org/debate-project/

15.

Jamieson

K. H.

Adasiewicz

(2000). What can voters learn from election debates? In Coleman

(Ed.), Televised election debates: International perspectives (pp. 25-42). New York, NY: St. Martin’s Press.

16.

Johnston

Hagen

M. G.

Jamieson

K. H.

(2004). The 2000 presidential election and the foundations of party politics. New York, NY: Cambridge University Press.

17.

Luscombe

(2016, September 30). Hillary Clinton calls Trump “temperamentally unfit” to lead after Machado spat. The Guardian. Retrieved from https://www.theguardian.com/us-news/2016/sep/30/clinton-trump-temperamentally-unfit-lead-alicia-machado

18.

Prior

(2009). Improving media effects research through better measurement of news exposure. Journal of Politics, 71, 893-908.

19.

Prior

(2012). Who watches presidential debates? Measurement problems in campaign effects research. Public Opinion Quarterly, 76, 350-363.

20.

Stelter

. (2016, September 27). Debate breaks record as most-watched in U.S. history. CNN Money. Retrieved from http://money.cnn.com/2016/09/27/media/debate-ratings-record-viewership/

21.

Third presidential debate of 2016 draws 71.6 million viewers. (2016, October 20). Nielsen. Retrieved from http://www.nielsen.com/us/en/insights/news/2016/third-presidential-debate-of-2016-draws-71-6-million-viewers.html

22.

Trump

D. J.

(2016, November 1). Look at the way Crooked Hillary is handling the e-mail case and the total mess she is in. She is unfit to be president. Bad judgement! [Tweet]. Retrieved from https://twitter.com/realdonaldtrump/status/793400131525677056?lang=en

23.

Zhu

J. H.

Milavsky

J. R.

Biswas

(1994). Do televised debates affect image perception more than issue knowledge? A study of the first 1992 presidential debate. Human Communication Research, 20, 302-333.