ArticlePublisher preview available

An Empirical Test of the Comey Effect on the 2016 Presidential Election

Wiley
Social Science Quarterly
Authors:
Dennis Halcoussis at California State University, Northridge
Dennis Halcoussis
Anton D. Lowenberg
Anton D. Lowenberg
Anton D. Lowenberg
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Michael Phillips at California State University, Northridge
Michael Phillips
Read publisher preview
Download citation
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Objectives The contentious 2016 U.S. presidential election was marked by acrimonious televised debates between the two major candidates, Hillary Clinton and Donald Trump, federal investigations of Clinton's emails that were sent from a personal server when she held office as Secretary of State, and the release of a videotape of lewd remarks by Trump about his behavior toward women. Financial market uncertainty also played in role in the election campaign. The objective of the article is to examine the impact of these different factors on the election. Method The present article uses a popular vote prediction market to test the impact of these factors on the probability of Trump winning the election. Results Results indicate that the debates and videotape release were not statistically significant, but that a letter to Congress released by FBI Director James B. Comey on October 28, 2016, substantially decreased Clinton's probability of winning the popular vote and simultaneously increased Trump's probability. Financial market uncertainty is found to have some marginal positive effect on Trump's probability of winning. Conclusion Trump's probability of winning the election received a substantial boost from FBI Director James B. Comey's “last‐minute” announcement on October 28, 2016.
A preview of this full-text is provided by Wiley.
Learn more
Content available from Social Science Quarterly 
This content is subject to copyright. Terms and conditions apply.
An Empirical Test of the Comey Effect on the
2016 Presidential Election
Dennis Halcoussis ,California State University
Anton D. Lowenberg, California State University
G. Michael Phillips, California State University
Objectives. The contentious 2016 U.S. presidential election was marked by acrimonious televised
debates between the two major candidates, Hillary Clinton and Donald Trump, federal investiga-
tions of Clinton’s emails that were sent from a personal server when she held office as Secretary of
State, and the release of a videotape of lewd remarks by Trump about his behavior toward women.
Financial market uncertainty also played in role in the election campaign. The objective of the
article is to examine the impact of these different factors on the election. Method. The present
article uses a popular vote prediction market to test the impact of these factors on the probability
of Trump winning the election. Results. Results indicate that the debates and videotape release
were not statistically significant, but that a letter to Congress released by FBI Director James B.
Comey on October 28, 2016, substantially decreased Clinton’s probability of winning the popular
vote and simultaneously increased Trump’s probability. Financial market uncertainty is found to
have some marginal positive effect on Trump’s probability of winning. Conclusion. Tr u m p ’ s p r o b -
ability of winning the election received a substantial boost from FBI Director James B. Comey’s
“last-minute” announcement on October 28, 2016.
The 2016 presidential election was one of the most contentious in recent U.S. history.
Donald Trump, running on a populist platform, proposed to abandon or reverse many
key policies of previous administrations in the areas of healthcare, financial regulation,
taxation, international trade, migration, defense, and foreign policy. His anti-globalization
stance appealed to constituents who felt their standard of living threatened by interna-
tional competition and deindustrialization forces within the U.S. economy, which Trump
promised to roll back. Hillary Clinton, on the other hand, advocated a continuation of
more traditional, liberal policies that were perceived as less disruptive to the political status
quo within Washington.
During the campaign, Clinton was viewed as outperforming Trump in the presidential
debates, but Clinton’s public image was potentially harmed by announcements of FBI
investigations of sensitive emails sent from her personal server when she was Secretary
of State. Trump’s chances of electoral success were seemingly set back by the release of a
videotape by the Washington Post in which Trump boasted of sexually assaulting women.
Financial markets appeared to respond quite differently to the prospect of a Clinton
victory at the polls than to that of a Trump presidency. In almost every U.S. presidential
election since 1880, equity markets have risen on the expectation of a Republican win and
fallen when Democrats were expected to win (Snowberg, Wolfers, and Zitzewitz, 2007a).
Direct correspondence to Dennis Halcoussis, Department of Economics, California State University,
Northridge, 18111 Nordhoff Street, Northridge, CA 91330-8374 dennis.halcoussis@csun.edu.
SOCIAL SCIENCE QUARTERLY, Volume 101, Number 1, January 2020
C2019 by the Southwestern Social Science Association
DOI: 10.1111/ssqu.12729
... We use the Iowa Electronic Market for these prices (as does Halcoussis, Lowenberg, and Phillips (2020) and Halcoussis, Lowenberg, and Phillips (2009).) 4 The use of these prices is common in the literature (Erikson and Wlezien, 2012;Rhode and Strumpf, 2004). Specifically, as our dependent variable, we use the average price per day of a "winner-take-all" contract that pays $1.00 if the Democratic nominee wins the popular vote. ...
... Polls taken immediately after the debate found that voters thought Clinton had won the debate by a clear margin(Wolfers and Zitzewitz, 2016, p. 1).3 The foregoing discussion draws fromHalcoussis et al. 2020. ...
The COVID effect: an empirical analysis of the pandemic and the 2020 U.S. presidential election
Article
  • Sep 2024
  • J Econ Finance
The impact of the COVID pandemic on the 2020 election outcome is analyzed using Iowa Electronic Market data, measures of socially and economically driven market volatility, a measure of COVID severity, and selected election-related events. Building on research regarding two previous U.S. presidential elections, we find that the pandemic helped the incumbent in two ways. The largest impact supporting the incumbent came from the apparent medical severity. A secondary impact came from social and economic volatility with the surprising finding that both risks helped the incumbent relative to the challenger. However, these impacts were not adequate to overcome the relatively large advantage of the challenger.
View
... Specifically, it may be that a fraction of Moderates were actually Democrats, who are known to have highly negative (and therefore not ambivalent) attitudes towards President Trump, whereas Independents would have greater levels of ambivalence as illustrated in Fig 6B. Returning to where this paper began (i.e., the 2016 presidential election), we offer a novel argument regarding ambivalence's role in Trump's 2016 election win: (a) a large fraction of the electorate identified as NRD [31]; (b) of these voters, a large fraction had some level of ambivalence regarding Donald Trump; (c) these ambivalent voters likely made their final decision just days before the election [5], and, consequently, were likely influenced by the political context at that time [32]; (d) the political context just days before the election was negative for Hillary Clinton (i.e., Comey's re-opening of the Hillary Clinton email controversy). Though we believe that the potential role of ambivalence in the 2016 presidential election is interesting, and perhaps worthy of further study, regardless of its veracity the "story" serves as a cautionary tale for the importance of ambivalence in political opinion polling. ...
Are political-opinion pollsters missing ambivalence: “I love Trump”… “I hate Trump”
Article
Full-text available
Given the increasing attention ambivalence is receiving from the psychological community, it must be asked if pollsters’ (routinely) dichotomous political opinion surveys are missing something crucial. To determine if there is any legitimacy to this question, undergraduates attending a Liberal Arts college in Southern California were asked to rate their level of agreement/disagreement to 28 statements regarding President Trump in two studies, with the items drawn from actual Quinnipiac (Q) and Brookings Institute (BI) surveys. To quantify ambivalence participants were told they could mark one or two responses per item, with double-responses serving as a measure of ambivalence. In Study 1, mean Trump approval ratings divided along party lines, and were consistent with the Q and BI findings. Nonetheless, approximately 40% of participants registered some level of ambivalence across all political-party affiliations, with those defining themselves as Neither Democrats (DEMs) nor Republicans (REPs) showing the greatest degree of ambivalence. In Study 2, ambivalence towards President Trump was examined looking at both party affiliation and political ideology (Conservative, Moderate, and Liberal). Again, roughly 40% of participants displayed some level of ambivalence, with greater degrees of ambivalence for Independents relative to DEMs and REPs, and Moderates relative to Liberals. Given research indicating that ambivalence is associated with delayed decision making and decisions based on “in the moment” contextual information, our findings our suggestive: if political opinion pollsters do not assess ambivalence, they may be missing information on a fair-sized demographic that could influence an election based on negative information (real or fictitious) surfacing only days before an election… as it did in 2016.
View
... Perhaps unsurprisingly, betting markets generally adjust quickly to politicians' behaviors and the perception thereof. For example, betting odds shifted substantially after the first U.S. presidential debate in September 2016 (Wolfers and Zitzewitz, 2016), as well as when FBI Director James B. Comey made his announcement in October 2016 regarding the status of Hilary Clinton's private email investigation (Halcoussis et al., 2020). While presidential betting markets may seem novel, they have been ''large and well-organized markets for betting on presidential elections" stretching back to at least 1868 (Rhode and Strumpf, 2004). ...
Later bedtimes predict President Trump’s performance
Article
  • Dec 2020
  • ECON LETT
Technology and social media use are increasingly associated with delays in nightly sleep. Here, we consider the timing of President Trump’s official Twitter account posts as a proxy for sleep duration and how it relates to his public performance. The President wakes around 6am, a routine which has not changed since early 2017. In contrast, the frequency of Twitter activity 11pm–2am increased 317% from under one day per week in 2017 to three days a week in 2020. The President’s increased late-night activity is not accounted for by increases in the frequency of his use of social media over time, his travel schedule, or seasonality. On the day following one where he posts late at night, his Twitter followers interact less with his posts, described as “official statements by the President of the United States”.¹ He receives 7400 fewer likes per tweet, 1300 fewer retweets per tweet, and 1400 fewer replies per tweet after a late night (drops of 6.5%–8%). Tweets aside, the President’s speeches and interview transcripts have previously been coded for their dominant emotion through text analysis. On the day following a late night, the President’s inferred emotion is less likely to be “happy” and nearly three times more likely to be “angry” in his interviews and speeches. Finally, the 2020 election odds of the President’s chief opponent also increase after a late night, while the President’s are unchanged. The pattern we document is consistent with a progressive shortening of the President’s sleep over his first term and compromised performance from sleep deprivation.
View
View
Analyzing event statistics in corporate finance: Methodologies, evidences, and critiques
Book
  • Jan 2015
Analyzing Event Statistics in Corporate Finance provides new alternative methodologies to increase accuracy when performing statistical tests for event studies within corporate finance. In contrast to conventional surveys or literature reviews, Jeng focuses on various methodological defects or deficiencies that lead to inaccurate empirical results, which ultimately produce bad corporate policies. This work discusses the issues of data collection and structure, the recursive smoothing for systematic components in excess returns, the choices of event windows, different time horizons for the events, and the consequences of applications of different methodologies. In providing improvement for event studies in corporate finance, and based on the fact that changes in parameters for financial time series are common knowledge, a new alternative methodology is developed to extend the conventionalanalysis to more robust arguments.
View
View
Does market volatility impact presidential approval?
Article
  • Nov 2011
  • J Publ Aff
Recently, market volatility has been used as an explanatory variable for presidential job approval. Our research builds on such an approach by first extracting the economic and non-economic components of market volatility using the Eta® (c4cast.com, Inc., 750 E. Walnut St., Pasadena, CA 91101) model, which considers 18 economic factors, far more than the unemployment and inflation factors used in most similar studies. Further, our results suggest that it is not aggregated market volatility but rather disaggregated market volatility that has a causal effect on presidential job approval ratings. In addition, the market's expectation of economic and non-economic market volatility exhibits different patterns, in terms of direction and timing, in their relationship to presidential job approval ratings. Copyright © 2011 John Wiley & Sons, Ltd.
View
Markets vs. Polls as Election Predictors: An Historical Assessment
Article
  • Sep 2012
  • ELECT STUD
Prediction markets have drawn considerable attention in recent years as a tool for predicting elections. But how accurate are they? Do they really outperform the polls, as some would have us believe? Do prices in election markets carry information beyond the horserace that is portrayed in the latest polls? And how well do election markets perform when there are no public opinion polls to provide the cues? This paper compares the accuracy of presidential betting markets in years before and after public opinion polls were introduced. For data, we rely on prices from modern internet-based election markets since 1992 but also from the vigorous election markets that thrived on Wall Street going back from at least as far as 1880 up to 1960. Our results comparing the performance of markets and polls are provocative. First, we find that market prices are far better predictors without polls (1880-1932) than with polls available (1936-2008). Second, we also find that market prices of the pre-poll era predicted presidential elections at least as well as polls have done following the introduction of scientific survey research. Finally, we find that last-minute market prices add nothing to election prediction once we control for trial heat polls during the final week of the campaign. The ordering of these results is not transitive.
View
Suckers Are Born but Markets Are Made: Individual Rationality, Arbitrage, and Market Efficiency on an Electronic Futures Market
Article
  • Mar 2004
The Iowa Electronic Markets are specially designed futures markets that appear to aggregate information efficiently to predict events such as election outcomes. Yet, in theory, perfect information aggregation is impossible. Further, the markets are populated by a nonrepresentative sample of mistake-prone and biased traders. That is, traders are prone to the behavioral anomalies predicted by behavioral finance. How can this be reconciled with market efficiency? Here, we take a first step by analyzing the behavior of two self-selected types of traders. Dramatic differences in mistake rates across traders can help us answer the question. Market-making traders who set prices are less mistake prone and appear to be more rational than price-taking traders. This highlights an important feature of markets: marginal (in this case, market making), not average, traders set prices. This can drive the efficiency of market prices in spite of large numbers of traders who display patently suboptimal behavior.
View
The Obama Effect
Article
  • Jul 2009
Many observers argued that Barack Obama’s candidacy in the U.S. presidential election of 2008 benefited from the financial crisis and recessionary economic conditions which voters blamed on the Republican administration. However, an empirical examination of stock price and public opinion data indicates that improvements in Obama’s electoral prospects led stock price declines and that gains by Obama were more likely to be followed by falling stock prices than by rising prices. This evidence suggests that the poor performance of the stock market in the days leading up to Obama’s victory was partly caused by, rather than causing, Obama’s success.
View
Did Steve Forbes scare the US municipal bond market?
Article
  • Feb 1999
  • J PUBLIC ECON
This paper investigates the relationship of the daily spread between taxable and municipal securities to a daily indicator of the likelihood of a tax reform that would eliminate the preference now accorded to municipal securities, the ‘flat tax.’ The probability of tax reform is proxied by the market price of the shares of Steve Forbes, a champion of the flat tax, in the Iowa Electronic Market. We find evidence that the indicator of tax reform’s prospects was correlated with a decline in the spread on five and ten-year maturity bonds, although not for 30-year maturity bonds. The correlations do not, however, survive in first-difference form, suggesting either that Forbes’ prospects did not immediately affect expectations of future tax policy, that the correlations are coincidental rather than causal, or that the decreased signal-to-noise ratio obscures a real, but small, causal relationship.
View
Partisan Impacts on the Economy: Evidence from Prediction Markets and Close Elections
Article
  • Jan 2006
Analyses of the effects of election outcomes on the economy have been hampered by the problem that economic outcomes also influence elections. We sidestep these problems by analyzing movements in economic indicators caused by clearly exogenous changes in expectations about the likely winner during election day. Analyzing high frequency financial fluctuations following the release of flawed exit poll data on election day 2004, and then during the vote count we find that markets anticipated higher equity prices, interest rates and oil prices, and a stronger dollar under a George W. Bush presidency than under John Kerry. A similar Republican-Democrat differential was also observed for the 2000 BushGore contest. Prediction market based analyses of all presidential elections since 1880 also reveal a similar pattern of partisan impacts, suggesting that electing a Republican president raises equity valuations by 2–3 percent, and that since Ronald Reagan, Republican presidents have tended to raise bond yields.
View
What Do Financial Markets Think of War in Iraq?
Article
  • Apr 2003
We analyze financial market data in order to produce an ex-ante assessment of the economic consequences of war with Iraq. The novel feature of our analysis derives from the existence of a market for Saddam Securities,' a new future traded on an online betting exchange that pays only if Saddam Hussein is ousted. A variety of tests suggest that this future's price provides a plausible estimate of the probability of war. The spot oil price has moved closely with the Saddam Security, suggesting that war raises oil prices by around $10 per barrel. Futures prices imply that markets expect these large immediate disruptions to dissipate quickly, with prices returning to pre-war levels within about a year and a half. Evidence on the long-run effects is fragile, and while prices are probably expected to fall a little as a result of war, any oil dividend' will be minimal. We find large effects in equity markets: and war lowers the value of U.S. equities by around 15 percent. This effect is concentrated in the consumer discretionary sector, airlines and IT; the prospect of war bolsters the gold and energy sectors. Analyzing option prices, we find that the large estimated average effects of war reflect the market pricing in a range of different scenarios - a 70 percent probability that it will lead to market declines of 0 to 15 percent, a 20 percent chance of 15 to 30 percent declines, and a 10 percent risk of a fall in excess of 30 percent. Across countries, the most extreme effects are on the stock markets of Turkey, Israel, and several European nations. Countries that are highly enmeshed in the world economy, or net oil importers, are most likely to experience adverse effects from war.
View