Copyright 2004 The New York Times Company

June 12, 2004 Saturday
Late Edition - Final

Bush Aide Watches Polls And Public Perceptions

By JIM RUTENBERG

Matthew Dowd, President Bush's chief campaign strategist, is not just the man who conducts the president's polling. He also works to control public perceptions about where the presidential race stands, perhaps more aggressively than many other campaign aides in his position.

When Mr. Bush has risen sharply in the polls, Mr. Dowd has stepped in pre-emptively with memorandums widely sent to Republican officials, supporters and journalists to dampen expectations and warn that the country remained closely divided. ''President Bush's approval numbers will again fall back to more realistic levels fairly quickly,'' he wrote in a publicly released memo when the President had particularly high ratings after major combat operations ended in Iraq last spring.

When campaign officials worry public polls make the President's situation look too grim, Mr. Dowd also steps in, most vociferously when he believes the grimness to be in error. The most recent example of that came this week, when a new poll from The Los Angeles Times showed Senator John Kerry of Massachusetts leading President Bush among registered voters by seven percentage points -- a lead just beyond the poll's margin of error. Mr. Dowd publicly and sharply called the poll a ''mess,'' prompting a public spat with the news organization's polling director about the nitty-gritty of polling methodology.

To be sure, polls are often the blood that flows through the body politic, helping set perceptions about the state of a campaign, and the Kerry camp also frequently sends out memorandums, usually to reporters, that try to put public polls in the most favorable terms. In one of his commercials Mr. Kerry, the Democratic presidential contender, even says that the nation is going ''in the wrong direction,'' picking up the language of a standard survey question used to measure public discontent in what seemed at least in part devised to get voters to say the same thing to pollsters.

Still, analysts said, Mr. Dowd is exceptional. They described him as creating a new role for a presidential campaign as an expert polling director offering a more aggressive running commentary on the various public polls, one that often goes out not just to reporters, but also to Web sites and to six million supporters via e-mail.

''I don't know if I've ever seen a campaign have somebody who was in charge of overseeing polling and at the same time play this public of a role,'' said Charles Cook, the editor of The Cook Political Report, a nonpartisan political newsletter, and also a connoisseur of the fine points of polling. That Mr. Dowd does play this role underscores just how much Mr. Bush's campaign believes public polls can affect public perceptions of the race and help shape its contours even five months before Election Day.

Mr. Dowd says he is simply, to a large extent, working to correct a campaign scorecard -- likely to get endlessly amplified in these days of the 24-hour news cycle -- that he sees as occasionally marred by flawed polls biased against his candidate. His goal, he says, is to prevent the demoralization of the faithful upon whom the campaign is counting to rally friends and neighbors.

''I just want to make sure people have a realistic view,'' said Mr. Dowd, whose official title is ''chief strategist,'' in an interview Friday. ''There are highs that are going to go down, there are lows that are going to go up. I'm not just trying to argue with news that is perceived as bad -- I'm trying to argue against wrong news, good or bad, like a newspaper journalist might.''

Underlying the strategy is the belief among political strategists with both parties that poll results can become self-fulfilling prophecies, contributing greatly to the direction of a campaign by causing enthusiasm or demoralization.

Mr. Dowd said perceptions that the president's poll standing was poor, for instance, could lead the news media to constantly cover the president in an unduly negative light. ''You don't want to let a bad poll stand so somehow in the coverage, no matter what happened, the president is still behind,'' he said.

Democrats said the Bush campaign's attention to poll results undercut the president's oft-stated assertion that his administration is less concerned with polls than those of past presidents. Some said that it also indicated that Mr. Bush's campaign aides were more concerned about his recent standing with voters than they were willing to admit .

''What they're doing now is they're trying to say to their people, 'We're not sinking -- let's discredit the polls because we're doing so badly,''' said Douglas Sosnik, the political director of Bill Clinton's White House during the 1996 campaign, when it was itself known for paying heavy attention to polling.

Bush campaign officials denied such a motivation, pointing out that they have also reacted publicly when they believe that assessments of the president's standing are too rosy.

Susan Pinkus, director of polling for The Los Angeles Times, dismissed the Bush campaign's sharp questions about her newspaper's poll as ''negative spin'' in a public response posted on the ABC News political website, ''The Note.'' The Los Angeles Times poll was similar to several others, showing Mr. Kerry as leading Mr. Bush in a direct match-up by seven percentage points. Mr. Kerry's lead was just outside the poll's margin of error of plus or minus three points.

Mr. Dowd complained during an interview that while 38 percent of voters who responded to the Los Angeles Times poll identified themselves as Democrats, 25 percent identified themselves as Republicans. He argued that the poll results should have been adjusted to diminish what he said was an overly Democratic skew. It is an argument he has made about other polls, as well, including a New York Times/CBS News poll in late April that found that Mr. Bush's approval rating had slid to 46 percent, the lowest level of his presidency in that poll.

Mr. Dowd's argument is part of a wider debate within the political community over whether such adjustments to a polling sample should be made if a random survey appears to count too many people of one party or the other. Those opposed to such adjustments, including Ms. Pinkus, say that doing so would possibly undercut accurate findings.

Lee Miringoff, president of the National Council of Public Polls and the director of the Marist College Poll, said he was less sure. Mr. Miringoff said he also did not know how much all of this would affect voters this far from Election Day anyway. ''I'm not sure that the polls are as influential on voting as the political folks, and maybe even some of the pollsters, think,'' he said.

URL: http://www.nytimes.com