Snap judgments are the best way to pick election winners

December 7, 2006
Contact:
  • umichnews@umich.edu

ANN ARBOR—Voters’ snap judgments of candidate charisma are more accurate in predicting the outcome of elections than economic conditions and nearly as accurate as candidate campaign spending levels, a new study finds.

“But you need to measure charisma with silent video clips rather than sound-on clips because knowing about candidate policy positions disrupts people’s ability to judge the non-verbal cues that really matter,” said Daniel J. Benjamin, a fellow at the University of Michigan Institute for Social Research (ISR) and the co-author of the study with Jesse M. Shapiro of the University of Chicago Graduate School of Business.

After watching 10-second silent video clips of competing gubernatorial candidates, participants in the study were able to pick the winning candidate 58 percent of the time” significantly better than chance. When the sound was turned on and participants could hear what the candidates were saying, accuracy dipped to 53 percent” no better than chance since the study had a two percent margin of error.

Researchers found that the accuracy of predictions based solely on silent video clips was greater than the accuracy of predictions based on knowledge of which candidate was the incumbent and information about the prevailing economic conditions at the time of the election, including the unemployment rate and any changes in personal income for the year prior to the election.

They also compared the predictive accuracy of snap judgments about candidate charisma to predictions based on knowledge of candidate campaign spending.

“Charisma is involved in the ability to raise large sums of money,” Benjamin said. ” So it wasn’t surprising that campaign spending turns out to be a slightly more accurate predictor than snap judgments of charisma” but some of the predictive power of campaign spending could be picking up charisma.

“The basic finding that adding policy information to visual information about candidates actually worsens voter judgment has some important implications,” Benjamin said.

“It may help to explain, for example, why expert forecasters, who are highly informed about and attentive to policy matters, have been found to perform no better than chance in predicting elections.”

The findings also underscore the importance of charisma as distinct from policy positions or party affiliations in winning elections.

“It may be difficult to describe the factors that determine a politician’s charisma,” Benjamin said. ” But it can be measured by how people react to a politician in the absence of information about policy positions. Our study clearly shows that reactions to even small amounts of visual information are highly informative about charisma.”

For the study, Benjamin and Shapiro showed 264 participants, virtually all Harvard undergraduates, 10-second video clips of the major party candidates in 58 gubernatorial elections from 1988 to 2002. The clips were taken from C-SPAN DVDs of gubernatorial debates. The participants rated the personal attributes of the candidates, guessed their party affiliation, and predicted which of the two would win. Most of the clips were silent, but some had sound, allowing participants to compare their policy positions. Participants were not shown clips from states they grew up in, and were asked if they were familiar with any of the candidates. If they were, their predictions were eliminated from the analysis.

“Although a large body of existing research shows that candidates’ policy positions have a powerful influence on individual voting behavior, our study finds that ratings based solely on short selections of silent video are highly predictive of the overall vote share in gubernatorial contests,” Benjamin said.

He notes the large literature in social psychology establishing that judgments about other people based on ‘thin slices’ “exposures as brief as a few seconds” are highly predictive of reactions to much longer exposures.

The study was supported by the Taubman Center for State and Local Government at Harvard University’s John F. Kennedy School of Government and by KNP Communications, a political consulting firm based in Cambridge, Mass, that intends to use the findings to help predict election outcomes.

Benjamin is also a faculty member at Dartmouth College.

 

Related Links: