The University Record, November 2, 1992

Presidential candidates, issues pique voter interest

By Mary Jo Frank

Normally, voters are bored with presidential politics by mid-October in an election year, says Holli A. Semetko, assistant professor of communication and of political science. Not so this year.

“This year people are very, very interested in this campaign. And the campaign is dealing with serious issues. We also have a third candidate whose one-liners are often very funny, but he is dealing with serious issues,” Semetko says.

Ross Perot’s paid political advertisements are even attracting more television viewers than some sports and entertainment shows, she adds.

Television networks, along with newspapers and radio stations, that have relied heavily on polling results to supplement their coverage, may be in for a surprise, Semetko says, if the polls are as unreliable here as they were in the April 1992 British national election.

Comparing the two elections, Semetko says she can’t help but notice similarities.

British polls were predicting a Labor Party victory early in the campaign. By the end of the campaign, polls showed that the two main parties—Labor and Conservative—were within the margin of error. On election night, broadcasters reported the results were too close to call. Conservatives won.

“This has to be in the back of the minds of many, including George Bush, who has taken advice from the British Conservative Party,” Semetko says.

Bush, like Conservative leader John Major, is using taxes as a campaign issue. Conservative Party campaign advertising warned that if the Labor Party won, taxes would increase by 1,200 pounds (about $2,000) for the average family. The Labor Party didn’t rebut the ad, Semetko says, and that was a serious mistake.

Therefore, Democrat Bill Clinton responded immediately to the Republican’s negative ads that depicted tax hikes for average families if Clinton were


The “trust” issue also was used by the Conservatives against Labor Party leader Neil Kinnock, Semetko notes.

Polls may fail to correctly forecast the outcome, as in Britain. They also may affect voters’ behavior, Semetko says. In a two-party race, polls can inspire voters to join the bandwagon of the predicted winner or vote for the apparent underdog.

In this year’s race with three candidates it is unclear how Perot supporters will fall out, Semetko says.

Noting that polls taken by various news organizations within a few days of each other often show different results, Semetko explains that the sampling period in the field may be at slightly different times.

“People also could be changing their minds as things unfold,” Semetko adds.

“Tracking” polls—where pollsters continue to add fresh cross sections that make up 25–30 percent of the survey sample—are an interesting way to try to gauge shifting opinion, Semetko says, and represent a moving average.

Semetko has a special interest in this campaign. She is studying how the national networks are covering the presidential election and participating in a National Science Foundation-funded study of the impact of media and social networks on the electoral process in the presidential campaign.

Sensitive to criticism that in 1988 the networks engaged in “Tarmac journalism”—a shorthand for journalists’ readiness to take up and report staged media events, even on fields of airport runways—the networks this time have made a serious effort to cover issues, with expert reporters providing analysis, she says.

Semetko will provide some post-election analysis of her own this week—in Germany. She has been invited by the U.S. ambassador there to speak to the Bonn political community on the outcome of the U.S. elections and also will lecture at several German universities.