Insightology

It is election day and you do your civic duty by going to your designated polling place, standing in line, chatting with a couple of nice people, drawing the curtain and pulling the lever. Do you notice where you have voted? Of course, it's at your local school (or church or firehouse). Did that have any influence on how you voted? Of course not, right? Not so fast. New research (by Jonah Berger, Marc Meredith and Christian Wheeler) indicates that the type of polling place can have a subtle effect on how people vote. The impact is small but it is there in both a controlled lab experiment and in a noisy real-world environment.

In the relatively straightforward lab experiment, the researchers exposed people to various pictures. One group saw predominantly school pictures, another saw church pictures and a third group saw neutral pictures. Then each group is asked to "vote" on a series of ballot measures as they would in a real election in their state. Of course, the real objectives of the study are masked from the participants. The results show that participants shown church pictures are less likely to support a stem cell initiative compared to those shown school pictures or neutral pictures. In contrast, those shown the school pictures are more likely to support an education initiative compared to those shown church or neutral pictures.       

Okay, so that was a lab experiment and people aren't really voting. Would such an effect show up in a real voting situation? To test this, the researchers used data from the 2000 election in the state of Arizona. Along with federal and state level races, there were 14 ballot initiatives on which people voted. Of particular interest to the researchers was Proposition 301 (Raise state's sales tax rate from 5.0% to 5.6% to increase educational expenditure). Would voting in a school result in more votes for this initiative than voting in other places?

The researchers collected precinct level polling location data from all 15 Arizona counties. About 40% of the votes were cast in churches and 26% in schools. In order to study the true effect of the polling place on voting behavior, they had to control for a variety of other factors. This included political views (vote for Bush or Gore), demographics and nearby location of a school when the vote was not in a school. After controlling for all these effects they still found a small but clear difference. People who voted in schools tended to favor the initiative more than those who did not. This effect did not manifest for any of the other 13 ballot initiatives, but then none of them was directly related to a school or religious issue. Further it wasn't like voters were making these choices consciously. If they were, it is likely the condition of the school they voted in (as measured by age) could have influenced their vote. It didn't.

Votes have real consequences and some of that is directly monetary, as in this case. Voting places, however, are chosen more for convenience than any other reason. In close elections it is possible that the results could turn on the actual location of the polling place. So the next time you go to vote, pay attention to where you are voting and think about whether that has any impact on your vote.    

Jonah Berger is an Assistant Professor of Marketing at the Wharton School in the University of Pennsylvania, Marc Meredith is a Visiting Lecturer in Political Science at MIT and  S. Christian Wheeler is Associate Professor of Marketing at Stanford University. This research was published in the Proceedings of the National Academy of Sciences.