No man or woman is an island, not even in that sacrosanct chamber of democracy known as the ballot box. That’s because the nature of an assigned polling place can, without people knowing, sway how they vote — enough to swing a close election, a new study indicates.
In a 2000 statewide election in Arizona, a greater proportion of people who voted at schools supported a school-funding initiative than did people who voted at churches or other polling places, say marketing professor Jonah Berger of the University of Pennsylvania in Philadelphia and his colleagues. The researchers statistically controlled for a variety of factors that might have affected this finding, which they report online June 23 in Proceedings of the National Academy of Sciences.
A follow-up experiment conducted online found that people more often supported a school-funding initiative if they had just seen images of school settings during an ostensibly unrelated task, as opposed to images of office buildings and other structures. Participants reported no awareness of the school images having influenced their support for the initiative.
Science News headlines, in your inbox
Headlines and summaries of the latest Science News articles, delivered to your email inbox every Thursday.
Thank you for signing up!
There was a problem signing you up.
“Consequential real-world decisions can be influenced by subtle environmental features outside awareness,” says political economist and study coauthor Marc Meredith of StanfordUniversity.
This study is the first to probe how polling places influence the vote, Meredith adds.
Berger’s team offers one of the best demonstrations to date of situational influences on real-world behavior, remarks YaleUniversity psychologist John Bargh.
Laboratory studies have indicated that background features of a person’s immediate environment, such as a church, seamlessly activate knowledge and attitudes related to churches. This same mental process allows people to recognize acquaintances instantly and to know effortlessly how to classify various entities, such as chairs or trees, Bargh says.
It’s too soon to make any policy recommendations based on the new study, in Meredith’s view. “All voting systems are likely to have unappealing consequences,” he says. Earlier research found that voter turnout depends heavily on the proximity of polling places to people’s homes, the stability of polling locations and parking availability at polling sites.
Given the possibility that unscrupulous politicians might now try to manipulate polling locations to increase support for favored causes, Bargh recommends careful deliberation about where people should vote. “It may not be possible to find a perfectly neutral polling place, but a city hall makes more sense than a church or school, at least for me,” he says.
Subscribe to Science News
Get great science journalism, from the most trusted source, delivered to your doorstep.
Berger’s team analyzed precinct-level election results from Arizona’s 2000 general election. An initiative on that ballot proposed raising the state sales tax to increase education spending.
About 56 percent of people who voted at schools supported the initiative, compared with 54 percent of people who voted at other locations, the scientists report. That difference, although small, was statistically significant, even when school-based voters were compared only to those who voted at sites located near schools.
The analysis controlled for factors that could have affected the results, such as an observed tendency for a greater number of education-friendly people to live near and vote at schools and a propensity for affluent, liberal people to use schools as polling places.
Voting at schools had no effect on the outcomes for ballot measures unrelated to education.
Berger’s group then administered political attitude surveys to 327 adults as part of an online experiment. Two weeks later, participants rated the brightness of a series of images. Half perused images of schools mixed with those of office buildings. The rest examined images of churches and office buildings.
Immediately afterward, participants read a description of the Arizona education initiative and indicated whether they would support it. Nearly 64 percent of those shown school images endorsed the measure, compared with 56 percent of the rest. People who initially supported the need for new taxes and who had children were most likely to support the initiative. These influences had an effect on volunteers who had seen images of the schools but not as much as they had on those who saw other images.
When asked later, no one who viewed school images thought the scenes had boosted their receptivity to the education initiative.
The polling place effect joins other subtle environmental influences on voting, such as the tendency for people to vote slightly more often for candidates whose names appear first on a ballot.
The new findings raise important questions for further research, Meredith says, such as whether voting in a church influences support for gay marriage or stem cell research and whether voting in a school boosts preferences for candidates strongly associated with education.