The small effects of political advertising are small regardless of context, message, sender, or receiver: Evidence from 59 real-time randomized experiments
Alexander Coppock, Seth Hill & Lynn Vavreck
Science Advances, September 2020
Evidence across social science indicates that average effects of persuasive messages are small. One commonly offered explanation for these small effects is heterogeneity: Persuasion may only work well in specific circumstances. To evaluate heterogeneity, we repeated an experiment weekly in real time using 2016 U.S. presidential election campaign advertisements. We tested 49 political advertisements in 59 unique experiments on 34,000 people. We investigate heterogeneous effects by sender (candidates or groups), receiver (subject partisanship), content (attack or promotional), and context (battleground versus non-battleground, primary versus general election, and early versus late). We find small average effects on candidate favorability and vote. These small effects, however, do not mask substantial heterogeneity even where theory from political science suggests that we should find it. During the primary and general election, in battleground states, for Democrats, Republicans, and Independents, effects are similarly small. Heterogeneity with large offsetting effects is not the source of small average effects.
Reconsidering Lost Votes by Mail
MIT Working Paper, September 2020
A “lost vote” occurs when a voter does all that is asked of her, and yet her vote is uncounted in the final tally. Estimating the magnitude of lost votes in American presidential elections has followed the work of the Caltech/MIT Voting Technology Project, which initially estimated the magnitude of lost votes in the 2000 presidential election, due to failures of voter registration, polling-place management, and voting technologies, to be between 4 and 6 million out of 107 million cast that year. Because of data and conceptual limitations, lost vote estimates have tended to focus on in-person voting, ignoring lost votes due to mail voting. This paper revisits the one previous effort to estimate lost votes, by considering data available from the 2016 presidential election. Conceptually, the paper highlights how differing mail-ballot legal regimes produce lost mail votes in different ways, and at different rates, on account of differing laws, regulations, and practices. Empirically, the paper draws on administrative records and surveys to provide an estimated number of lost mail votes in 2016. That estimate works out to approximately 1.4 million votes in 2016 -- 4.0% of mail ballots cast and 1.0% of all ballots. These estimates are relevant in light of efforts to expand mail balloting in the 2020 presidential election. States that will see the greatest growth in mail ballots tended to have higher lost vote rates than those with vote-by-mail systems. This implies that a doubling or tripling of the number of mail ballots in 2020 will result in a disproportionate growth in the number of lost votes due to mail ballots.
Voting by Mail and Ballot Rejection: Lessons from Florida for Elections in the Age of the Coronavirus
Anna Baringer, Michael Herron & Daniel Smith
Election Law Journal, September 2020, Pages 289-320
The COVID-19 pandemic and its concomitant need for social distancing have increased the attractiveness of voting by mail. This form of voting is nonetheless not a panacea for election administration in the time of a public health crisis, as a widespread move to ballots cast by voting by mail risks exacerbating existing inequities in mail-in ballot rejection rates across voters and jurisdictions. This motivates our examination of the roughly 9.6 million and 8.2 million ballots cast in the 2016 and 2018 general elections in Florida, respectively, including over 2.6 million vote-by-mail (VBM) ballots cast in each. Using a selection model that analyzes all ballots cast and those VBM ballots not counted in Florida in these two elections, we find that younger voters, voters not registered with a major political party, and voters in need of assistance when voting are disproportionately likely to have their VBM ballots not count. We also find disproportionately high rejection rates of mail ballots cast by Hispanic voters, out-of-state voters, and military dependents in the 2018 general election. Lastly, we find significant variation in the rejection rates of VBM ballots cast across Florida's 67 counties in the 2018 election, suggesting a non-uniformity in the way local election officials verify these ballots. As interest in expanding mail voting swells as a consequence of the novel coronavirus, protecting the rights of all voters to participate in electoral politics requires a characterization of the correlates of VBM ballot rejection with an eye toward considering how disparities in ballot rejection rates might be rectified.
Mobilization and Counter-Mobilization: The Effect of Candidate Visits on Campaign Donations in the 2016 Presidential Election
Boris Heersink, Brenton Peterson & Jordan Carr Peterson
Journal of Politics, forthcoming
Political scientists studying the impact of campaign visits by presidential candidates have come to conflicting conclusions on whether campaigns change voter behavior in even small ways. In this paper, we argue that, while scholars have generally interpreted campaign effect results as being uni-directional, the traditional metrics of such effects - polls and aggregate vote results - inherently reflect a net effect combining any potential mobilization of a candidate’s supporters, offset by any counter-mobilization of their opponents. If such counter-mobilization occurs, weak or null findings in the campaign effects literature may understate or miss the true impact of campaign activities on voter behavior. To assess whether campaign visits produce mixed effects, we measure the extent to which visits by presidential and vice-presidential candidates in the 2016 presidential election produced increases in campaign donations in the immediate aftermath of a visit. Our results show that visits by Donald Trump and Hillary Clinton inspired their supporters to donate more money than they otherwise would have. However, we also find a considerable level of counter-mobilization: visits by both Trump and Clinton resulted in an increase in donations to the opposing presidential campaign.
The Race-Blind Future of Voting Rights
Jowei Chen & Nicholas Stephanopoulos
Yale Law Journal, forthcoming
A critical issue in any racial vote dilution case is the proportionality (or lack thereof) of a minority group’s representation: how well (or poorly) minority voters are represented relative to their share of the population. In an important recent opinion, Judge Easterbrook proposed replacing this proportionality benchmark with what we call the “race-blind baseline.” Under this approach, minority voters’ representation would be compared not to their population share but rather to the fraction of seats they would control if districts were drawn randomly and without the use of racial data. Unsurprisingly, conservative advocates have been quick to embrace Judge Easterbrook’s idea. The current Supreme Court, which has already dismantled part of the Voting Rights Act, may also be interested in adopting the race-blind baseline. Yet until now, no one has explored this benchmark’s implications: how it would affect minority representation as well as the partisan balance of power. In this Article, we tackle these questions for the first time. We do so using a technique - the random generation of district maps by a computer algorithm - that has become the gold standard in partisan gerrymandering cases, but that has not previously been deployed in the context of race and redistricting. We find, first, that in most states, a non-racial redistricting process would yield substantially fewer districts where minority voters are able to elect their preferred candidates. Judge Easterbrook’s proposal would thus cause a considerable drop in minority representation. Second, we show that the minority opportunity districts that arise when lines are drawn randomly are quite different from the ones that now exist. They are less likely to pack minority voters and more apt to represent them through coalitions with white voters. And third, contradicting the conventional wisdom about the link between minority and partisan representation, we demonstrate that Democrats would not benefit from the elimination of opportunity districts under the race-blind baseline. Rather, in the southern states where the benchmark would have the biggest impact, it is Republicans who would gain a partisan edge.
The Heightened Importance of Racism and Sexism in the 2018 US Midterm Elections
British Journal of Political Science, forthcoming
In 2016, attitudes related to racism and sexism were strong predictors of vote choice for president. Since then, issues related to race and gender have continued to be an important part of the political agenda. This letter shows that hostile sexism and denial of racism emerged as stronger predictors of the House vote in the 2018 cycle than they had been in 2016. The results show that the increased importance of these factors came largely from the shifting of less sexist and less racist voters from voting Republican in 2016 to voting for Democrats in 2018. Overall, the results suggest that Trump's hostility towards women and minorities is becoming part of the Republican Party's brand, and that this appears to have created an electoral penalty for Republican candidates in 2018.
The Macedonian Fake News Industry and the 2016 US Election
Heather Hughes & Israel Waismel-Manor
PS: Political Science & Politics, forthcoming
During the 2016 US presidential election, Americans were exposed to an onslaught of disinformation on social media. Many of the most viral posts originated from Veles, a small town in central Macedonia. During fieldwork in Veles, where we interviewed several residents and disinformation creators, we found that the epicenter of this viral phenomenon was Mirko Ceselkoski, an autodidact social media expert, teacher, and mentor to Veles’ fake news operators. We interviewed Ceselkoski and registered and attended his online course - the same course numerous Veles residents took offline. Our research confirms (1) the pivotal role Ceselkoski had in the creation of this industry; (2) the economic motivation driving the fake news disseminators; and (3) the manner in which the mostly young people in their early twenties with little English fluency were able to generate so much traffic and disseminate so much disinformation.
The cost of gendered attitudes on a female candidate: Evidence from Google Trends
Raphael Corbi & Pedro Picchetti
Economics Letters, forthcoming
How much can negative attitudes towards women affect voting for a female candidate on a major election? We measure gender animus by calculating a proxy based on Google search queries that include gender-charged language. Such approach likely elicits socially sensitive attitudes by limiting the concern of social censoring, circumventing usual difficulties associated with survey-based measurements. We compare the proxy to Hillary Clinton’s vote share in the presidential election of 2016, controlling for the vote share of the previous Democratic presidential candidate, Barack Obama. Our results indicate that a one standard deviation increase in our proxy is associated with a 2 percentage points relative loss for Hillary and suggest that online-based observable behavior can be useful for measuring different kinds of hard-to-measure social attitudes.
Does Distance Matter? Evaluating the Impact of Drop Boxes on Voter Turnout
William McGuire et al.
Social Science Quarterly, forthcoming
Methods: The placement of five new ballot drop boxes was randomized among six potential sites identified based on similar criteria. The randomization of the five boxes across the six sites created natural Treatment (those sites that received a new box) and Placebo (the site that did not receive a new box) groups. We then employed a difference‐in‐difference design to determine whether voters in the Treatment group were more likely to vote in the 2017 general election compared to those in the Placebo group.
Results: We find that a decrease of one mile to the nearest drop box increased the probability of voting by 0.64 percent.
Whither Presidential Approval?
Raphael Small & Robert Eisinger
Presidential Studies Quarterly, forthcoming
Until recently, presidential job approval has been positively correlated with a healthy economy. Despite pre‐COVID‐19 sanguine macroeconomic indicators, President Donald Trump's job approval numbers appear to be quite low relative to those of his predecessors. Using a Bayesian model to infer latent presidential approval, we regress approval on consumer sentiment, finding strong evidence of a structural break commencing with the inauguration of President Barack Obama. We also find that the positive associations between presidential approval and consumer sentiment have weakened. Finally, we posit that a polarized political landscape may contribute to the changing of the meaning of presidential approval, which now appears to be a proxy for partisanship.
Assessing Group Incentives, Independent Spending, and Campaign Finance Law by Comparing the States
Charles Hunt et al.
Election Law Journal, September 2020, Pages 374-391
Independent expenditures (IEs) in U.S. elections have increased substantially at nearly all levels of government over the past decade, but judicial decisions are only a partial explanation for this growth. Using a descriptive difference-in-differences approach, we show that the growth has been uneven across types of elections and spenders under different legal regimes. This finding highlights the importance of disaggregating spenders, elections, and laws in order to explain IEs more fully. This article analyzes IEs in state gubernatorial and legislative elections from 2006 through 2018 across states with differing campaign finance laws and political contexts. It uses an original and detailed classification of spenders, along with data on IEs from the National Institute on Money in Politics, the Campaign Finance Institute's historical database of state campaign finance laws, and other sources. The legal variations on which the article focuses are the various states' laws limiting contributions to candidates and political parties. It concentrates on these because of an oft-stated expectation that removing contribution limits will sharply reduce the level of IEs. In addition to contribution limits, we also assess partisan competition as a primary explanation for the level of IEs in various states, and across the sectors of spenders. We find, using multi-variate analysis, that increased partisan competition (at both the candidate level and chamber level) is in most cases a significant driver of IEs. In contrast, the associations between IEs and contribution limits are inconsistent and generally not significant. Importantly for ongoing policy debates, ideological and issue-driven spending appears to have weak association (or none) with contribution limits. Therefore, if the recent increase in IEs is in fact a normative problem, the solution may be more elusive than once thought.
Southern Accents and Partisan Stereotypes: Evaluating Political Candidates
Kathleen Ash et al.
Social Science Quarterly, forthcoming
Methods: We designed an experiment using actors to record candidate stump speeches, manipulating the accent of the candidate and the partisan issue position of the candidate.
Results: Democrats and Republicans in our sample were more likely to believe candidates with southern accents were Republican as opposed to Democratic, and were also more likely to negatively evaluate candidates with a southern accent. This was true in both instances regardless of whether the candidate espoused a typical Democratic or Republican issue position. Democrats’ judgments of candidates with southern accents, though, were harsher than those of Republicans and, again, this was the case across the partisan positioning of the candidate. Finally, both Democrats and Republicans in our pool were less likely to say they would vote for the candidate with a southern accent regardless of the partisan position of the candidate.
The Past as Prologue: How to Forecast Presidential Elections
Megan Czasonis, Mark Kritzman & David Turkington
MIT Working Paper, August 2020
The authors apply a novel forecasting technique called Partial Sample Regression to predict the outcomes of U.S. presidential elections. This technique first measures the statistical relevance of past elections. It then employs an obscure mathematical equivalence - that the prediction from a linear regression equation equals a relevance-weighted average of the values for the dependent variable - to forecast election outcomes from a subsample of prior relevant elections. This technique has been applied successfully in finance to predict factor returns and the correlation of stock and bond returns. The authors apply Partial Sample Regression to predict the outcomes of the past five presidential elections as well as the 2020 election. They also report which past elections were identified as being statistically most relevant for each of the elections they predict.
Price discrimination in political advertising: Evidence from the 2012 presidential election
RAND Journal of Economics, Fall 2020, Pages 615-649
In 2010, the US Supreme Court loosened contribution limits to Political Action Committees (PACs), sparking fears that big donors could exert outsize influence on elections by funding PAC advertising. However, PACs are potentially handicapped when buying advertising time; data from 2012 reveal that PACs pay 32% above regulated campaign rates. I estimate a model of demand for advertising by PACs, exploiting the misalignment of state and media market borders to address price endogeneity. I find that prices reflect willingness‐to‐pay for viewer demographics rather than media bias. The estimates further suggest that network‐owned stations discriminate more successfully than do local affiliates.