By Beth Roberts
beth@uga.edu

Spencer Tinkham and Ruth Ann Lariscy

Ruth Ann Lariscy and Spencer Tinkham, both professors of advertising and public relations in the Grady College of Journalism and Mass Communication, have been studying the use of negative advertising in congressional election campaigns since 1982. They talked with Columns about their methods and their findings.

Columns: Maybe we should define negative campaigning first.

Tinkham: A negative ad mentions the opponent by name, and it focuses the content of the message on the characteristics of the opponent rather than the characteristics of the sponsor of the ad.

Lariscy: Not all negative ads are equal. We say there’s “good negative” and there’s “bad negative.” “Good negative” is not a personal attack—it might deal with issues or voting record. “Bad negative” is attacking the opponent’s character.

Tinkham: When we say good and bad we’re not necessarily saying effective or ineffective. Negative ads which are viewed by the voter as nasty—below the belt, unfair—are likely to boomerang.

Lariscy: Those are the most likely to have a backlash—they’ll hurt the person who initiated the attack more than they hurt the victim. But that’s rare—for the most part, negative ads are very effective.

Columns: That’s not the usual public assumption. Why effective?

Lariscy: Negative information is more complex, because you’re talking about two people. You think about negative information more than you think about positive information; it takes longer to process. It’s more memorable.

Tinkham: It grows in power over time—there’s a sleeper effect.

Lariscy: And you forget where you heard it. When you first encounter a negative message you might discount it, but then later you go into the voting booth and you remember there was something really nasty about that guy.

For the last year we’ve been studying the new legislation called “Stand by your ad”—where you see the candidate saying he approved the message.

Tinkham: It became federal law as part of the McCain-Feingold campaign finance reform act.

Lariscy: The intention was to decrease the amount of negative advertising, particularly in federal campaigns. So the candidates are not doing it as much, but the independent groups are doing the nasty stuff. A lot of people believe that the candidate is behind them, but the candidate is able to say I have nothing to do with those.

Tinkham: Whether that’s believed or not is still in question. For example, in 1988 the first President Bush was associated with the Willie Horton ad and people believed that was his campaign ad. It was actually an independent group sponsoring that ad, but the president benefited from it. So we might say it almost doesn’t matter if your campaign is perceived as affiliated with a nasty ad—it’s the content that remains. The sponsorship tends to be dissociated from the ad.

Columns: I wonder how you do this. Can you give me a 30-second methodology lecture?

Lariscy: The race that we’ve studied most consistently is the congressional campaign, starting in 1982. First we survey everyone who ran for the U.S. Congress in all 50 states that year. We’ve had remarkable response rates—ranging from 38 percent to as high as 60 percent—and always balanced between winners and losers, Republicans and Democrats. We’re very proud of these findings. The candidate or the campaign manager tells us how much money they spent, what kind of advertising they did. We supplement that with secondary data resources, like the federal election commission and the secretaries of the 50 states, from which we get all the official vote counts.

Tinkham: Even almanacs, for things like official district populations.

Lariscy: Then we do qualitative interviews—to help us understand the survey results. After an election we go to Washington and conduct in-depth interviews with campaign consultants, the people who create the ads and manage the campaigns.

Tinkham: The interviews always lead to new research questions, new ideas.

Lariscy: In the early ’90s, we began some experimental studies. I think it’s from our experiments that we’ve really learned how negative advertising works. We recruit registered voters—we’ve recruited from elderly populations, we’ve recruited from junior colleges. We don’t use UGA students, except for pre-testing. We bring them into laboratories and show them a political ad and ask questions about the ad that they just saw. Then we show them another ad and we repeat the process.

The sleeper effect which Spencer was talking about just a few minutes ago—the idea that over time the negative becomes more powerful—that came from the experimental studies. We had a fictitious candidate in the experimental lab and then we called everyone two to six weeks afterwards and asked them for whom they would vote in this fictitious campaign. We found that the negative message in fact increased in power.

Tinkham: And we found that a defensive message worked initially to decrease preference for the attacker, but over time the negative ad had a stronger impact.

Columns: So on the whole voters are not as offended by negative ads as they claim to be?

Tinkham: There is some evidence that negative campaigning depresses turnout. But it’s a very weak influence compared to other factors that influence turnout, like changes in population, ethnic distributions of population in the district, education level.

We’re still doing this analysis. Just yesterday I was examining an interaction effect. Challengers who are campaigning effectively give a reason for the incumbent being turned out of office, and that’s a negative message. When challengers do well, turnout seems to be lower, because a challenger has to raise doubt among the voters that the incumbent is the right choice. That stage of uncertainty seems to be the reason for a low turnout. Incumbents seem to benefit from high voter turnout, and if a challenger is going to win, it’s likely to be in a relatively low turnout race.