Turning Them Out

Kevin Lewis

November 05, 2021

How Social Desirability Response Bias May Lead to an Overestimation of Obama-Trump Voters
Christopher Stout, Keith Baker & Madelyn Baker
Public Opinion Quarterly, Summer 2021, Pages 694-705

In this research note, we argue that a segment of Trump voters may be motivated to demonstrate that they are not racially insensitive by saying they voted for Obama in 2012. As a result, what may appear as a substantial group of swing voters may be -- in part -- partisan voters who want to avoid perceptions of racial bias. Using data from a survey of a nationally representative nonprobability sample from YouGov, polling data from the 2016 election, and a survey experiment conducted with a convenience sample of American adults, we find that Obama-Trump voters score significantly higher on a social monitoring scale than others and that respondents are significantly more supportive of Trump when they are given the opportunity to mention who they voted for in 2012. The combination of results suggests that the number of Obama-Trump voters may be overestimated due to self-monitoring. 

Debunking the Non-Delegation Doctrine for State Regulation of Federal Elections
Mark Krass
Virginia Law Review, forthcoming

One objection to the conduct of the 2020 election concerned the key role played by state executives in setting election rules. Governors and secretaries of state intervened to change a host of regulations, from ballot deadlines to polling times, often acting pursuant to legislation granting them emergency powers. Some advocates, politicians, and judges cried foul. They argued that state legislatures may not devolve the power to set the "Times, Places, and Manner" of federal elections under article I, section 4 of the Constitution. This Article contests that view. Drawing on a survey of elections statutes in the thirteen original colonies, I argue that Elections Clause delegations were a pervasive part of early American practice. Executive officers like sheriffs, and local officials like selectmen, had enormous discretion to determine the time and place of elections, and sometimes also their manner. That discretion was repeatedly affirmed by Congress. As a matter of original meaning, this legacy suggests that legislatures today may indeed delegate under the Elections Clause. 

Auditing the 2020 General Election in Georgia: Residual Vote Rates and a Confusing Ballot Format
David Cottrell et al.
Election Law Journal, forthcoming

The 2020 general election in the United States took place against the backdrop of a pandemic and countless claims about voter fraud. The presidential race in Georgia was extremely close, and in this state both a hand and machine recount followed the initial promulgation of results. Armed with certified results, we conduct a statistical audit of the 2020 Georgia election by analyzing residual vote rates in statewide races. A race's residual vote rate combines the rates at which ballots contain undervotes (abstentions) and overvotes (when voters cast more than the allowed number of votes in a race). Anomalously high residual vote rates can be indicative of underlying election administration problems, and our analysis of these rates in Georgia finds nothing anomalous in the state's presidential race, a notable result given this contest's closeness. We do, however, uncover an unusually high overvote rate in Georgia's special United States Senate election. This overvote rate is concentrated in Gwinnett County and appears to reflect the county's two-column ballot design that led roughly 4,200 voters to select more than one candidate for Senate in the special election, in the process rendering invalid their votes in this contest. Gwinnett County's two-column ballot was not pivotal to the outcome of Georgia's Senate special election but nonetheless joins other ballot formats, like the infamous butterfly ballot used in Palm Beach County in the 2000 presidential race, that contribute to voter confusion and should be avoided in the future. 

The effects of racial status threat on White Americans' support for Donald Trump: Results of five experimental tests
Sheridan Stewart & Robb Willer
Group Processes & Intergroup Relations, forthcoming

Prior theory and research suggest that threats to the status of White Americans may increase support for Donald Trump. Consistent with this, one previous experiment conducted in early 2016 documented this effect (Major et al., 2018), finding that making salient the declining White majority in the United States increased support for Trump's presidential candidacy among White participants with high levels of ethnic identification. We report the results of five very similar experiments (total N = 3,076) also conducted in 2016, including one conducted on a national probability sample. The first experiment (conducted in January 2016) found that racial status threat increased Whites' support for Trump. The other four (conducted from February to October 2016), however, all found null results, and an internal meta-analysis of the five studies found no significant main effect overall. Additionally, none of the studies found an interaction of racial demographic shift and ethnic identification in which racial demographic shift increased Trump support among high-identifying Whites. We conclude by discussing a variety of potential explanations for our findings, including (a) that racial status threats did not increase Whites' Trump support, (b) that racial status threats increased Trump support early in the 2016 election cycle, but the role of this factor in Trump support declined over time, or (c) that this pattern is an example of a broader tendency of declining experimental treatment effects on candidate support over time in campaigns (Kalla & Broockman, 2018). 

Partisanship & nationalization in American elections: Evidence from presidential, senatorial, & gubernatorial elections in the U.S. counties, 1872-2020
Sharif Amlani & Carlos Algara
Electoral Studies, October 2021

Scholars argue that contemporary American elections are pronounced in their degree of partisanship and nationalization. While much of this work largely uncovers a heightened degree of nationalization in contemporary elections, little is known about how far back these patterns generalize. Given the limited availability of American electoral data, this work also generally focuses on a single office or during a certain segment of the post-war period since 1946. Moreover, this work largely focuses on states as salient units of analysis, masking potential variation found in U.S. counties, the smallest geographical unit constituting panel observations over time and across elections. In this note, we leverage a novel dataset of county-level election returns for President, U.S. Senate, and Governor, to specify a model assessing whether American elections are more nationalized and partisan than during any other period since the Civil War. We find evidence that presidential and Senate elections are more partisan today than any period since the Civil War, while gubernatorial elections are as partisan today as they were during the late 1800s. Our findings have implications for contemporary-based theories explaining the rise of partisanship in American elections and demonstrates the utility of county-level data in assessing electoral changes in America. 

Episodic master narratives through time: Exploring the temporal dynamism of 2016 election night stories
William Dunlop, Dulce Wilkinson & Nicole Harake
European Journal of Social Psychology, forthcoming 

Master narratives are culturally vetted stories that guide collective action. Over an 18-month period, we examined the dynamism inherent in master narratives of the 2016 American presidential election. In assessments separated six months apart, four independent samples of Clinton and Trump voters provided their stories of election night 2016 and rated the positivity and vividness of these experiences. Narratives were additionally quantified along both linguistic (I-talk and We-talk) and conceptual (redemption and contamination) dimensions. Over time: (a) election night experiences became less vivid, (b) the proportion of We-words in election night narratives increased (suggesting greater group identification), (c) Clinton supporters' evaluations of their election nights became marginally more positive (suggesting an increased optimism), and (d) Trump supporters' narratives became less redemptive (suggesting a reduced enthusiasm for this historical event). These results signal the enduring and dynamic elements of the narratives central to two of America's diametrically opposed sociopolitical groups. 

The Common Good and Voter Polarization
John Matsusaka & Chad Kendall
University of Southern California Working Paper, August 2021

Do voters see democracy entirely as a game of self-interest in which one person's gain is another's loss, or do they also view it as a search for the common good, as some democracy theorists have long conjectured? Existing empirical research that assumes entirely private interests cannot answer this question, by design. We develop an empirical model in which voters derive utility from both common-good and private considerations, and show formally how to disentangle the two preference components. We estimate the model on California ballot propositions from 1986 to 2020, and find that 46 to 87 percent of voters place significant weight on the common-good aspects of proposals. Common-good concerns mitigate the effects of voter polarization, which we find substantially increased over out study period -- particularly in the last six years. 

Is Voting Transformative? Expanding and Meta-Analyzing the Evidence
John Holbein et al.
Political Behavior, forthcoming

Voting is the foundational act of democracy. While thousands of studies have treated voting as a dependent variable, comparatively little research has studied voting as an independent variable. Here we flip the causal arrow and explore the effect of exogenous voting shocks on citizens' broader attitudes and behaviors. To do so, we first use two waves from a uniquely large survey of young people in the United States, pairing this with a regression discontinuity design. We augment these results with a new meta-analysis of all causally-identified studies exploring whether voting is transformative. We find that -- despite voting at much higher rates -- individuals induced to vote, regardless of the mode used to mobilize, are (precisely) no different from all-else-equal individuals that are not. Our results illuminate the (non)consequences of a vitally important -- and widely studied -- political behavior and speak to the broader importance of voting as an object of study. 

The Contingent Effects of Sexism in Primary Elections
Danny Hayes & Jennifer Lawless
Political Research Quarterly, forthcoming

Although the landscape for female candidates in U.S. politics has improved, research continues to find that many voters possess sexist attitudes. We rely on a standard political communication framework to help reconcile sexism in the electorate with increasingly favorable outcomes for women in primary elections. Based on two national survey experiments, we first demonstrate that in the absence of gendered campaign rhetoric, sexism is a weak predictor of support for female candidates on both sides of the political aisle. We then show, however, that when a male candidate attempts to activate sexism among voters by attacking a female opponent, gender attitudes become more salient -- but not to the woman's disadvantage. In a Democratic primary, gendered attacks backfire and lead to a significant boost in support for the female candidate. On the Republican side, a male candidate does not face the same backlash, but the attacks do very little to depress his female opponent's support. While the persistence of hostile attitudes toward women has slowed the march toward gender equality in society, our experimental results suggest that sexism exerts only contingent effects in primary elections and not systematically to female candidates' detriment. 

The Effect of Core Values on Support for Electoral Reform: Evidence from Two Survey Experiments
Sheahan Virgin
American Journal of Political Science, forthcoming

A central tenet in the electoral systems subfield is that parties seek desired outcomes via the strategic adoption of electoral rules. Such partisan self-interest, however, is merely one explanation for reform: A second is that an actor may attempt to maximize her core values, which constitute her perception of the "common good." Although the extant literature has demonstrated the motivational power of core values, their effect on electoral rule choice has not been tested. Using a factorial experimental design that manipulates the partisan- and values-implications of a fictitious reform proposal, I find evidence in favor of core values: Not only do they have an effect net of partisan concerns, but they also attenuate the effect of partisan self-interest when the two predispositions countervail. The results provide evidence that partisan self-interest offers an incomplete picture: Actors hold -- and pursue through electoral reform -- objectives that are not immediately partisan in nature. 

It's not me, it's you: Perceptions of others and attitudes toward a female nominee in the 2020 New Hampshire democratic primary
Jennifer Lucas & Elizabeth Ossoff
Politics, Groups, and Identities, forthcoming

While the 2020 Democratic field was touted as one of the most diverse in presidential nomination history, a white, male, heterosexual candidate ultimately won the nomination. This is, on its face, surprising, as we might expect less sexism and more enthusiasm for diverse candidates among Democratic voters. To help explain this outcome, we refocus attention on the "third-person effect" and the anticipated reactions of others to a female candidate, rather than voters own individual beliefs in two ways. First, we demonstrate that attitudes about female presidential candidates still follow third-person effect predictions; individuals attribute socially desirable attitudes to themselves and less so more distant others (i.e. "Americans"). Second, we analyze how voters take into account perceived potential gender bias by others, which in turn influences female candidates' perceived electability (likely support from other voters). Gendered electability then has measurable impacts on preference for female candidates, as they strategically choose the more electable candidate. Overall, these results demonstrate looking at the anticipated gender bias of others, rather than just an individual's attitudes, can be helpful in explaining the continued perception of female presidential candidates as less electable and, ultimately, why women have not yet broken the highest glass ceiling. 

Deception Detection in Politics: Can Voters Tell When Politicians are Lying?
Kyle Mattes, Valeriia Popova & Jacqueline Evans
Political Behavior, forthcoming

In this study, we investigate voters' unaided perceptions of whether politicians are lying. We conduct an experiment in which participants attempt to uncover politicians' dishonesty by watching videos of their speeches. We find that verbal cues (specifically, the amount of detail in the speech) and general demeanor cues explain the success (failure) of veracity judgments far better than paraverbal and nonverbal cues. We also find evidence of a truth bias -- people are more likely to judge statements to be true than false -- despite the political setting, where voters might have been more skeptical. However, gender plays a deterministic role for veracity judgments in political context; female politicians are more likely to be judged as honest.


from the


A weekly newsletter with free essays from past issues of National Affairs and The Public Interest that shed light on the week's pressing issues.


to your National Affairs subscriber account.

Already a subscriber? Activate your account.


Unlimited access to intelligent essays on the nation’s affairs.

Subscribe to National Affairs.