Santa Clara University

Going Negative

It Affects Voters’ Choices More Than Previously Believed

Negative advertising has become a pervasive part of America’s political landscape, yet there are substantial gaps in knowledge of how and whether it works. Most of the research has focused on its effect on voter turnout, generally concluding that negative ads tend to lower participation in elections.

Ravi Shanmugam, assistant professor of Marketing, has applied standard market- research techniques to a rich vein of election data and concluded that negative advertising has a statistically higher impact on voters’ choice of a candidate than on turnout.

“We looked a lot more at the effect of negative advertising on voter choice,” says Shanmugam. “We took the research techniques used to determine how people make choices in basic product categories, such as toothpaste, and applied them to the broader choice voters make in deciding which candidate to vote for.”

Those findings are reported in “Negative Advertising and Voter Choice,” a working paper by Shanmugam, Hai Che and Ganesh Iyer.



Shanmugam and his colleagues were able to apply the market-research techniques to a wide range of data collected by the American National Election Studies Project at the University of Wisconsin during the 2000 presidential and congressional elections. That study provided detailed information on 1,807 individual voters around the country who were surveyed before and after the election. The ANES study also compiled detailed information on different types of advertising deployed in election markets, which made it possible to categorize nearly every ad in every studied campaign as either positive or negative.

Several clear trends emerged from the study. One was that, as a rule, negative advertising tended to be more prevalent in the presidential campaigns than in congressional races. Republican candidate George W. Bush’s ads were 68 percent negative, whereas Republican Congressional candidates’ ads were only 49 percent negative.

The most likely explanation for this, Shanmugam says, is that many of the Republican Congressional candidates were incumbents, some in safe seats, who saw less need to go negative.“In many Congressional races, name recognition is the key hurdle a challenger has to overcome,” Shanmugam says. “In those cases, incumbents might feel that negative advertising would give an opponent more attention.”

Detailed analysis of advertising strategy also showed that negative advertising is more likely to be used in campaigns for closer races and in Congressional districts where the cost of advertising is relatively low. Campaigns are less likely to take a negative approach in markets with a more educated electorate, where advertising exaggerations or errors are more likely to be spotted, resulting in a backlash in favor of the candidate being attacked.

The most critical finding, however, was that when market-research techniques were applied to the data, it turned out that 80 percent of the impact of negative advertising was on the choice of candidate, and only 20 percent was reflected in voter turnout.

“This uncovering of the relatively high impact on voter candidate choice as compared to that on turnout is missing in existing political advertising studies,” the authors write.

Shanmugam says he was drawn to this research by a long-standing interest in American politics and that he hopes to do additional research building on the findings of this paper.

“Now that we have this information on the outcome of negative advertising, we have the potential to look at the actual traits of the ads,” he says. “The main way we can extend our research on political advertising would be to look at the content of negative ads and try to determine why some work and others don’t.”

Printer-friendly format