WASHINGTON (AP) – If John McCain and Barack Obama think their ads blasting each other are persuading undecided voters, they’re probably wrong. But negative ads do have an impact, an AP-Knowledge Networks poll suggests, even if it’s just to neutralize the other guy’s attacks.

They can also solidify support – or simply turn voters off to both candidates.

In a new survey, voters were asked to watch two of the presidential candidates’ negative ads, an Obama spot that says McCain would tax health benefits and a McCain ad that claims Obama wants “massive government.” The campaigns have spent millions of dollars on such ads with millions more committed for the last two weeks before Election Day.

On the whole, adwatchers who went into the experiment undecided were unmoved. About 60 percent of so-called “persuadable” voters said the ads made them no more or less likely to vote for McCain or Obama. And about a third appeared to throw up their hands, saying they were less likely to vote for either candidate after watching the ads.

Are this year’s ads fair?

More than half the voters polled believe presidential campaign commercials have been unfair or somewhat unfair. And the more ads they said they had watched, the less fair they found them. People who had seen 10 or fewer ads mostly thought they were fair, 62 percent. But people who had seen 30 or more in the past week said the opposite – 63 percent said most of the ads were unfair.

People who had seen ads by both candidates tended to think Obama’s ads were more fair than McCain’s, 39 percent to 16 percent. That may have helped Obama neutralize McCain’s critical ads.

Among partisans, 39 percent of strong Obama supporters said the ads made them more likely to vote for him, while 29 percent of McCain’s strong supporters said the same for him.

As for poll respondents’ views about what they see on TV, about four of 10 said Obama’s ads mostly attack, while about seven of 10 said that of McCain’s.

A study by the Wisconsin Advertising Project at the University of Wisconsin found that at the beginning of October McCain’s ads were almost all negative, whereas only 34 percent of Obama’s ads were. But the ratio has been much closer though the campaign, with seven of 10 McCain ads and six of 10 Obama ads criticizing opponents. McCain is beginning to mix his ads, much as Obama has, into a blend of positive and negative.

The Associated Press-Knowledge Networks poll was conducted Oct. 10-12, a time when McCain was drawing particular attention for questioning Obama’s relationship with 1960’s militant William Ayers.

“It’s not just McCain’s negative ads, it’s also the news media’s coverage,” said John Geer, a political scientist at Vanderbilt University and author of “In Defense of Negativity: Attack Ads in Presidential Campaigns.” “The news media have given McCain a bit of a hard time for running negative ads, and I think that’s also shaping the public’s perception.”

The two ads viewed by poll respondents aimed to cast the rival candidate in a negative light.

The McCain ad argues Obama and “his liberal congressional allies” will produce “painful income taxes.” Obama’s ad strikes at McCain’s health care plan, saying it would tax workers for their employer-provided health coverage. In fact, Obama would raise personal income taxes only on families making over $250,000 and individuals making more than $200,000. And McCain’s health care plan offers a tax credit that would be more generous for the vast majority of people than the current tax break on benefits.

Researchers who study political advertising say there is no empirical evidence negative ads alone drive voters away.

“Very many negative ads are perfectly accurate and perfectly factual and perfectly relevant. And they don’t always work,” said Ken Goldstein, a professor at the University of Wisconsin and director of the Wisconsin Advertising Project.

The AP-Knowledge Networks poll of 1,287 registered voters has an overall margin of sampling error of plus or minus 2.8 percentage points. It was conducted over the Internet by Knowledge Networks, which initially contacted people using traditional telephone polling methods and followed with online interviews. People chosen for the study who had no Internet access were given it for free.

—

AP Director of Surveys Trevor Tompson contributed to this report.

AP-ES-10-21-08 1740EDT