News Focus2012 Election

Want to Tear Down Your Rival? Here's What Might Work Best

See allHide authors and affiliations

Science  26 Oct 2012:
Vol. 338, Issue 6106, pp. 465
DOI: 10.1126/science.338.6106.465

A new study of negative political ads shows that timing and audience may be the keys to success.

If you want to attack your political opponent, do it after voters have made up their minds.

That's the conclusion of a new and controversial study on the impact of negative advertising in political campaigns. October is the critical month for election propaganda, says Yanna Krupnikov, a political scientist at Northwestern University in Evanston, Illinois. She finds that people are most susceptible to negative ads late in the campaign if they attack the candidate they have chosen. The ads can discourage people from voting, she reported last year in work funded by the National Science Foundation.

Rising tide.

Purely negative TV campaign ads—those mentioning only a rival, like many now targeting Barack Obama and Mitt Romney—have risen sharply in the past decade just before the U.S. presidential election. This sample shows the breakdown of ad tone for September since 2000.


Negative ads are on the rise (see graph), and the idea that they discourage voting has been batted around for decades. Political scientists Stephen Ansolabehere and Shanto Iyengar reported in the mid-1990s that such ads significantly discourage voter turnout, and their work triggered scores of follow-up papers.

More recent analyses have been overwhelmingly skeptical of that conclusion, however. Some researchers have found the opposite effect: that negative ads actually boost participation. One prominent doubter, John Geer, a political scientist at Vanderbilt University in Nashville, assembled a database of 3 decades of ads created for presidential campaigns. His 2006 book, In Defense of Negativity, argues that such ads do not stifle voter turnout and are actually good for democracy because statements that challenge an opponent contain more factual information than feel-good image ads and, thus, promote a vigorous debate.

Northwestern's Krupnikov doesn't think it's that simple. She believes that other analysts may have missed the vote-suppressing effect of negative ads because they didn't look carefully at the timing and targeting of such ads. Had they done so, she argues, they might have found a similar effect.

Her work is based in part on how people make consumer and voting decisions. It's a two-step process, she says: People amass and evaluate information (including negative comments, which are helpful at this stage) before making a choice. Then they decide to act. A late blast of negative information may discourage voting without changing a person's choice, she says. If a voter thinks that a chosen candidate may not be any better than a disliked candidate, that voter “has no reason to turn out and vote.”

Krupnikov used the 2004 presidential election to test her theory, examining the tone of advertising and surveys about voter turnout. She analyzed similar surveys from elections in 1976 through 2000 and melded those responses with ads in those seven earlier elections, cataloged in a database and evaluated by Geer and colleagues.

Krupnikov found that only late negative ads (those aired after 1 October) were linked to a significantly lower likelihood that people would turn out to vote. The data for the seven other elections are less precise, but in the October 2011 American Journal of Political Science, she writes that her analysis of the earlier years yielded “the same pattern” as for 2004.

This year, researchers are waiting to see what happens in a presidential election that appears headed for a record wave of negative ads. “Most of President Obama's advertising has been negative,” says Kenneth Goldstein, a political scientist who last year became CEO of Kantar Media/Campaign Media Analysis Group, a Washington, D.C., consulting firm. “But so has Mitt Romney's advertising—and ads by all the allies and PACs [political action committees].”

Goldstein developed methods of tracking and analyzing political television ads as a professor at the University of Wisconsin, Madison, and at Wesleyan University in Middletown, Connecticut. The Wesleyan Media Project (WMP) reports that most ads being aired in the final stage of the 2012 presidential campaign are negative. September's TV ads fall into three “tone” categories, WMP says: 62.8% purely negative (“mentioning solely the opposition candidate”), 29.5% contrast ads (“those that mention both the favored candidate and the opposition”), and 7.8% positive ads (“mentioning solely the favored candidate”). The share of negative campaign ads in September has risen steadily throughout the decade, from 22.7% in 2000 to 30.5% in 2004 to 56.2% in 2008.

Goldstein isn't particularly concerned about the trend. “I've studied this every which way in many years and in many races. I don't always find positive effects [encouraging voter participation], … but I have never found negative effects.”

What may be the most thorough scholarly rejection of the idea that negative ads suppress voting comes from a meta-analysis by Richard Lau, a political scientist at Rutgers University in New Brunswick, New Jersey, and colleagues. Their review of 111 papers in the November 2007 issue of The Journal of Politics found no significant vote-discouraging effect. “I spent 15 years of my life reading every damn paper that was written on this,” Lau says.

His conclusion is that these ads, if they influence voters at all, actually stimulate participation in elections. Lau would be “a lot more concerned” about the ad blitz if one side heavily outspent the other. But in this year's presidential election, he says, “both Obama and Romney have more money than God.”

Navigate This Article