Science is confirming something successful politicians seem to know instinctively: Support your local football team.
The success of major college teams in the two weeks before an election can have a measurable impact on how well incumbent politicians do at the polls, researchers report in Tuesday's edition of Proceedings of the National Academy of Sciences.
"Events that government had nothing to do with, but that affect voters' sense of well-being, can affect the decisions that they make on Election Day," the researchers said.
That's why incumbent politicians try to score some good news just before elections, and their opponents try to block that effort.
The study looked at elections for president, governor and the Senate between 1964 and 2008 and compared them with football results for 62 major college teams. The researchers found that wins in the two weeks before an election boosted the vote share of incumbents in the county where a school is located by 1.05 to 1.47 percentage points - enough to make a difference in a close race.
For teams they termed "powerhouses," the impact was even greater, giving the incumbents between 2.30 and 2.42 percentage points more than in years when the local teams lost. Powerhouses were defined as teams that had won a national football championship since 1964, or were among the teams with average attendance of 70,000 or more from 1998 to 2008.
Neil Malhotra, an assistant professor of political economy at Stanford University, and colleagues decided to look at the relationship between football and politics because they wondered whether elections were affected by "irrelevant" information.
The rationality of the American public when it comes to voting has generated a lot of discussion, Mr. Malhotra said in a telephone interview, but studies also have indicated that people can be "predictably irrational."
For example, a study looked at disasters such as floods and hurricanes, and found that people tended to vote against incumbents in the aftermath of such traumatic events.
"Just because government didn't cause the problem doesn't mean people won't hold them responsible," Mr. Malhotra said.
But, in a disaster, governments do tend to get involved, at least afterward.
So the researchers turned to an area even more unrelated to government: sports.
It tends to be a subconscious response, Mr. Malhotra said, with voters' moods affecting their decisions. The effect tends to disappear when the bias is brought to people's attention, he added.
In large parts of the country, college football outcomes would never be considered "irrelevant."
The study did not report results by individual school or conference, but combined the results from 1964 to 2008 to incorporate winning and losing years and several elections.
Mr. Malhotra's team conducted a second experiment, during the 2009 NCAA basketball March Madness championship tournament. In that case, they surveyed 3,040 people who lived in areas with teams in the tournament, asking each to name their favorite team.
The respondents were divided into two groups, one of which knew the results of the team's play before being surveyed and the other, which did not know the results.
After the third and fourth rounds of the tournament, participants were asked to rate the performance of President Obama.
Mr. Obama, an avid basketball fan, received a 2.3 percentage point boost for each win from respondents who knew how their teams were performing.
The survey confirmed the football results, the researchers said, showing that well-being induced by game results affected people's evaluations of the incumbent.