Professor Challenges Theories on Elections
Harvard Crimson, November 28, 2012
"Election outcomes have very significant policy consequences, but the election outcomes themselves are largely random," Vanderbilt professor Larry M. Bartels declared.
Bartels used the annual social sciences lecture at the Radcliffe Institute on Tuesday to challenge commonly held beliefs about elections—all of which, he says, are false.
Throughout most of the talk, Bartels responded point by point to "The Responsible Electorate," political scientist V.O. Key Jr.'s 1966 book that argues that the American electorate collectively makes rational decisions based on policy concerns. Bartels argued that the average American voter is instead uninformed about relevant policy issues during election cycles.
He pointed to evidence including a poll showing that only 40 percent of undecided voters this past summer knew that Republicans controlled the House of Representatives.
"There are very few people who decide what they think about the issue and then went out and voted for the guy [that campaigned on that issue]," he said, disputing Key's assertion that the electorate votes based on policy issues.
Addressing the claim that, despite that lack of knowledge, the American electorate collectively arrives at a reliable decision because individual voters cancel out each other's errors, he said that this assumption fails to consider that voters' decisions influence other voters.
"People don't sit in isolation and think about the political process unconnected to how everyone else is thinking about it," he said.
Bartels said that his research, which focused on the issue of privatization of Social Security in the 2000 election, suggests that voters choose their preferred candidate independent of any policy concerns. Voters then adopt the policies that candidate supports, he said.
"Paying a lot of attention to the political process in our current political climate mostly consists of learning the talking points on your side of the partisan divide," he said.
Bartels also countered the argument that the electorate's decisions act as a referendum on the incumbent candidate.
He said his research suggests that much of the electorate decides whom to vote for based on short-term conditions in the months leading up to the elections, with little regard to the overall performance of the past administration.
Radcliffe Dean Lizabeth Cohen introduced Bartels' lecture as an opportunity to reflect on the 2012 election from an academic perspective.
"Now that the outcome has been decided and journalists have moved on to other stories, the election belongs to scholars," Cohen said. "Today's lecture is a good way to launch this next stage of deeper analysis."