Is Trumpgate the Downfall of Donald Trump?
Political scientists will no doubt study the 2016 US Presidential elections for years to come. Although much about the campaign has been remarkable, the leaked video showing Republican nominee Donald Trump talk in a denigrating way about women has dominated the media coverage in the final weeks of the campaign. Many Republican Party officials distanced themselves in the wake of the release of the video. The question is therefore whether the discovery of the video, now referred to on Twitter as #Trumpgate, ended Trump’s hopes of winning the Presidency. In this blogpost I reflect on this question via an analysis of the existing literature on political scandals in American politics. Although the 2016 campaign is unique in many ways, political scandals have been an important part of many elections in US political history and may thus inform us on the relevance of #Trumpgate to the 2016 campaign.
Clinton-Lewinsky and Gary Hart
A first scandal we may look at is the Clinton-Lewinsky scandal of 1998. Trump himself referred to this scandal in the second debate to illustrate that his behavior was no different from that of other high-profile political officials. The Clinton-Lewinsky scandal emerged in 1998 when it became clear that incumbent President Clinton had engaged in an extra-marital affair with White House employee Monica Lewinsky. The scandal led to the impeachment of President Clinton by the House of Representatives, but was down-voted by Congress. Political scientists have studied the effects of the scandal on Clinton’s popularity and support among the public. Surprisingly, many studies have found that Clinton’s approval ratings increased in the wake of the scandal (Fischle 2000; Lawrence and Bennett 2001; Owen 2000; Zaller 1998). Several explanations have been provided for this. First, it has been argued that Clinton’s strong performance in office offset any negative feelings toward the President. Specifically, Clinton had managed to secure Peace and Prosperity for the country which made voters discount his marital indiscretion (Zaller 1998). Second, studies have shown that many voters saw the scandal as a `private’ rather than a `public’ or `political’ matter. This response was particularly prominent among Democratic partisans and was enhanced by media-reportings, which relied heavily on the salacious details of the event, suggesting to public that the political content of the affair was minimal (Fischle 2000; Owen 2000).
A second scandal of a similar nature occurred in 1988 and involved Democratic candidate Gary Hart who was set to win the Democratic primaries that year. Yet, over the course of the primaries, it became apparent that Hart was engaged in an extra-martial affair with a fashion model. An in-depth study of the scandal shows that, contrary to the Clinton-Lewinsky scandal, the revelations severely damaged Hart’s approval ratings and led to his withdrawal from the primaries (Stoker 1993).
One potential explanation for the difference in outcomes between the two scandals is the incumbency status of Clinton and the fact that his approval ratings were determined not just by a judgment of his character, but also by his actual (observed) performance in office. Furthermore, other studies on corruption scandals in the House of Representatives have shown that corruption allegations have a stronger effect in primaries when compared to general elections (Alford et al. 1994; Jacobsen and Dimock 1994). One important explanation for this is the fact that response to a scandal is strongly moderated by a voters’ partisan preferences. Thus, in general elections, partisan loyalty provides an important `cushion’ against punishment by in-party voters. In party primaries however, such a buffer of partisan loyalty is not present.
A final scandal that I explore in my own research is the Watergate scandal. The Watergate scandal revolved around a very different issue (the misuse of public office), but is similar to the scandals referred to above in the sense that it informed voters on the character and sense of judgment of the President. Figure 1 shows that there was a clear decline in the approval ratings of President Nixon over the course of Watergate (1972-1976), the left panel shows this for the entire electorate, the right panel of figure 1 shows the decline per partisan group. Here we see clearly that Nixon’s approval ratings are most robust among Republican partisans, but that even with this group, his approval ratings declined in the wake of Watergate.
Figure 1 - Nixon Thermometer Score 1972-1976
The decline in approval ratings suggests that the public responded negatively to the Watergate scandal. Yet, this does not automatically imply that Watergate led to a change in voting behavior. I thus explore this specific question and find that the electorate as a whole responded negatively in the 1974 mid-term elections in response to Watergate. Negative opinions on Watergate lead to a reduction in the likelihood of voting for the Republican Party in the 1974 mid-term elections. Yet – in line with the findings discussed above – I find that this response is moderated by partisan loyalty: the events of Watergate have a lower effect on Republican voters. However, my research finds that it is voters with moderate (as opposed to weak or strong) levels of partisan loyalty who turn away from the party. Figure 2 shows this by way of a calculation of the predicted likelihood of a vote for the Republican Party in the 1974 mid-term elections. In this plot I hold all other variables (such as race, gender, education, financial situation) constant and only vary the level of partisan identification of a voter. These predicted likelihoods are for Republican voters who voted for the Republican Party in 1972. We see in Figure 2 that especially voters with moderate attachment to the Party are less likely to vote for the party in 1974 (63% likelihood) when compared to weak (95% likelihood) and strong partisans (75% likelihood).
Figure 2 Predicted Likelihood Voting for the Republican Party in 1974 Mid-Term Elections
These findings are confirmed by other research which shows that voters with low and high interest in politics are most loyal to their preferred party (Lodge and Taber 2013; Stoker 1993; Taber and Lodge 2006; Zaller 1998) Voters with low levels of interest tend to rely on party cues rather than new information, whilst voters with high levels of interest reject information that does not correspond with their prior held beliefs.
I extend my analysis at the individual level with an analysis of the 1974 elections at the district level. Here l show that voters targeted their response to Republican House members who had protected Nixon against impeachment. These candidates experienced a larger decline (~6%) in their vote-share compared to House members who opposed Nixon. This is shown in figure 3 with the dotted line representing districts in which the House Representative was a member of the Judiciary Committee and voted against Nixon’s impeachment. The solid line represents members of the Judiciary Committee who voted in favor of Nixon’s impeachment. We see that the dotted line shows a larger decline in vote-share between 1972 and 1974 in response to Watergate.
Figure 3 Watergate Effect at District Level
The 2016 Presidential Campaign
Now, what does this brief overview of a history of political scandals in American elections tell us about the fate of Donald Trump? First, it is interesting to note that the literature suggests that scandals tend to have a stronger effect in primaries compared to general elections due to the absence of a partisan buffer in the former. Thus, had the video been discovered in the context of the Republican primaries, it might well have cost Trump the nomination.
Second, in the context of the general election, we might expect that the video confirms voters in their prior beliefs. Thus, those who disliked Trump see their opinion confirmed by the video. Those who like Trump either ignore the video, or re-interpret its meaning. For instance, similar to the Clinton-Lewinsky scandal, a Trump supporter might ‘decide’ (consciously or not) that Trump’s remarks about women are part of Trump’s personal views and do not reflect on his suitability to become President. Furthermore, a Trump supporter might see the video as part of a larger ‘conspiracy’ of the media or establishment to prevent Trump from becoming President. As such, the video might simply reinforce the existing divide between Clinton and Trump supporters without persuading or turning any voters. Importantly however, as the research discussed above shows, moderate partisans are most susceptible to change and persuasion. This particular fact might be problematic for Trump because, unlike most candidates, he does have a very strong base of support in the Republican Party. It is likely that the video dissuades moderate Republicans as well as hesitant Trump supporters from supporting him.
A further consequence of the video is the fact that many Republican Party officials have now distanced themselves from Trump. As the case of Watergate shows, this is a wise decision on their part. National-level events have an effect on lower level elections and voters do punish candidates who end up ‘on the wrong side of history’. As such #Trumpgate might well have signiticaly altered the course of the 2016 Presidential election campaign.
Alford, John R., Holly Teeters, Daniel S. Ward, and Rick K. Wilson. 1994. “Overdraft: The Political Cost of Congressional Malfeasance.” The Journal of Politics 56(3): 788–801.
Fischle, Mark. 2000. “Mass Response to the Lewinsky Scandal : Motivated Reasoning or Bayesian Updating ?” Political Psychology 21(1): 135–59.
Jacobsen, Gary C., and Michael A. Dimock. 1994. “Checking Out: The Effects of Bank Overdrafts on the 1992 House Elections.” American Journal of Political Science 38(3): 601–24.
Lawrence, Regina G., and W. Lance Bennett. 2001. “Retinking Media Politics and Public Opinions: Reactions to the Clinton-Lewinsky Scandal.” Political Science Quarterly 116(3): 425–44.
Lodge, Milton, and Charles S. Taber. 2013. The Rationalizing Voter. Cambridge University Press.
Owen, Diana. 2000. “Popular Politics and the Clinton/Lewinsky Affair: The Implications for Leadership.” Political Psychology 21(1): 161–77.
Stoker, Laura. 1993. “Judging Presidential Character: The Demise of Gary Hart.” Political Behavior 15(2): 193–223.
Taber, Charles S., and Milton Lodge. 2006. “Motivated Skepticism in the Evaluation of Political Beliefs.” American Journal of Political Science 50(3): 755–69.
Zaller, John. 1998. “Monica Lewinsky`s Contribution to Political Science.” PS: Political Science & Politics 31(2): 182–89.
Roosmarijn De Geus