Howard Kurtz Eleanor Clift Fred Barnes Arianna Huffington Herb Stein
9:27 a.m. Monday 9/23/96
It would be hard to escape the journalistic verdict that Bob Dole is "dead meat," in Jack Germond's elegant phrase. That conclusion is built into virtually every story, a steady drumbeat that grows more deafening day by day. Dole can't do ANYTHING right. He puts the Dodgers in Brooklyn. He's trailing in the polls. He falls off the stage. He's trailing in the polls. He talks of a bridge to the past. He's TRAILING IN THE POLLS.
The point is not that Dole is stronger than he looks; he's behind in Indiana, for crying out loud. Nor do reporters need to make the race seem closer than it is out of some journalistic obsession with balance. But the fixation with Dole's double-digit deficit has overwhelmed the campaign. The margin has become the message.
The weekend punditry is one thing. But when hundreds of newspaper and network stories contains some version of "Dole, desperately trying to jump-start his faltering campaign ...," it tends to drown out everything else. No matter what Dole says, no matter what Dole does, it's an act of desperation. "Dole, locked in a seemingly hopeless gap of as much as 17 points ... came to Villanova University yesterday in his latest bid to tunnel his way out," the Philadelphia Daily News declares. Hand that man a shovel!
This has nothing to do with what Dole calls "the liberal press" (we were more than happy to pronounce last rites for Dukakis and Mondale) and everything to do with the primacy of the horse race. The polls are everywhere, trumpeted on the front pages of the L.A. Times, Washington Post, USA Today. If it were permissible under the First Amendment, I'd ban half the polls. No, make that 90 percent. USA Today runs a presidential tracking poll EVERY SINGLE DAY, for those who need the fix.
Occasionally, of course, the polls prove a momentary illusion (Bush at 91 percent, Clinton in the doghouse after the '94 wipeout). But increasingly, this election doesn't seem to be about very much except tactics, polls and attack ads. Maybe that's partly the fault of Clinton and Dole, who ran prefabricated conventions and aren't saying much of consequence. But the journalistic consiglieres who declared Dole dead in early September aren't helping much. The trouble with horse-race journalism is that the press gets bored as soon as one horse pulls a couple of lengths ahead. And it shows.
10:02 a.m. Monday 9/23/96
I think we should quit whining about the polls. We can no more box them up and stash them away than we can any other modern development in politics. I remember in 1984 when Walter Mondale refused to wear makeup on television. This was before men routinely got patted with pancake. The dark circles under Mondale's eyes made him look like a raccoon under the bright glare of television lights. Like pancake, polls are here to stay. And there are even examples where the media behave responsibly in reporting the results of polls.
The networks routinely hold back the information from their exit polls on election day so as not to influence voters who have not yet been to the polls. Granted, the media moguls took this action in response to an enormous outcry primarily from Democrats who felt their congressional candidates were hurt by Jimmy Carter's premature concession speech in 1980, which he delivered before the polls had closed in California. In 1988, ABCanchorman Peter Jennings used poll results to bolster the visual image of an electoral map that showed Michael Dukakis hopelessly behind. Dukakis, watching television in his hotel room as he prepared to debate George Bush later that same evening, cited the poll as unfairly dampening his spirits--and those of his supporters--at a crucial moment in the campaign. It is my impression that television has since been more careful about appearing evenhanded at such key times.
The public, perhaps deadened by the onslaught of polls, is more sophisticated and doesn't take every poll to heart. Polls are political white noise. Voters discount them. In his book, The Reasoning Voter, political scientist Samuel Popkin argues that enough information gets through for voters to make rational decisions about candidates. Voters know that Bob Dole is behind. You could not suppress enough polls to conceal that information.
The polls do influence the media in the sense that we never take anything at face value. What candidates do is filtered through our knowledge of how it helps them in the polls. Be it ever thus. Let's just try to restrain ourselves and put it in the second paragraph instead of the lead.
1:27 p.m. Monday 9/23/96
I love polls. They're the first thing I look for in the newspaper every morning. They tell me (and everyone) a lot of what I want to know about a campaign or a candidate, chiefly trends like who's gaining, who's slipping. No, they are not warping the political process--please, spare me the hysterics. True, there are folks who think polls are having this effect. Guess who they are? Candidates trailing in the polls, their staffers, and supporters.
There is a problem with polls, however. They are overinterpreted or wrongly interpreted, particularly by the dopes at newspaper and TV networks who ought to know better. Upon looking at polls showing President Clinton with a 15- to 20-point lead over Bob Dole, media types want to declare the race over. Nonsense. The polls in the presidential race show something other than that the race is over, namely that Clinton's support is volatile, oscillating from the mid-40s to the mid-50s, and that Dole's support isn't volatile, sticking in the mid-30s to high-30s. My conclusion: Dole hasn't sold himself as a credible alternative to Clinton in a period of relative peace and prosperity, but that if he does, he may be competitive with Clinton.
One more concern: Journalists rely on polls too much. Often a good political reporter's instincts are better than a poll result. Smart reporters figured out in one Tennessee Senate race in 1994 that Fred Thompson was a strong contender even when he trailed Jim Cooper in polls early on. However, reporters frequently let polls govern their story line. The candidate who's behind must be doing something wrong. Well, maybe, maybe not. Perhaps his candidacy hasn't jelled yet, or perhaps his campaign plan has been played out, as was the case with George Bush in mid-summer 1988. Polls showed him way behind, but he won.
The biggest trouble with polls is that people think they reveal more information than they really do. Polls in the spring, summer, and early fall don't take into account what routinely happens in the closing week or two of a campaign. In 1982, virtually all Republican House incumbents led in polls before the final 10 days, then, boom, the bottom fell out and the GOP lost 26 seats. Few Republican challengers led in polls ever in 1994, then, boom, the bottom fell out for Democrats and the GOP won 52 seats. A week out, Republican Bill Frist held a 4-point lead over Sen. Jim Sasser in 1994 in Tennessee. On Thursday, his lead jumped to 6, then 9 on Saturday. He won the next Tuesday by 14.
My advice: Pay serious attention to polls, but don't overinterpret them.
2:26 p.m. Monday 9/23/96
Yes, polls are unequivocally polluting politics--not so much because they create the reality of winners and losers, but primarily because they turn political leaders into slavish followers of the most shallow reading of the electorate's wishes.
Every pollster declared balancing the budget one of those golden issues with a 75 percent approval rating--as high as Mother Theresa's. Republicans made it the centerpiece of the Contract with America. And in order to meet their poll-driven central objective of balancing the budget, they tackled Medicare before they had created a public consensus as they had on welfare reform.
President Clinton called the GOP pollsters' bluff by proving that even though 75 percent of the public wanted to see a balanced budget, they did not want to balance it "on the backs of the children, the elderly and the infirm." This became Clinton's tune and the public began to hum it with him. How come the pollsters forgot to tell the GOP leaders that?
Polling really started to infect campaigning and governing with John F. Kennedy. "We were not unlike the people who checked their horoscope each day before venturing out," wrote Evelyn Lincoln, Kennedy's secretary, providing an appropriate epitaph for modern political leadership.
The media have helped build the reputations of pollsters in exchange for being fed a daily grist for their mills. Candidates come and go, but pollsters are an enduring source.
As with medicine men and rain dancers, there is little correlation between the pollsters' legends and their successes. "Jimmy Carter is going to be president because of Pat Caddell--all because of Pat Caddell," Hamilton Jordan said in 1976, the first time one pollster had been given so much credit for a campaign. Caddell was also responsible for the greatest miscalculation of the Carter presidency--his malaise speech of July 1979--but the downfall was Carter's.
Bob Teeter was the first pollster to actually be named manager of a presidential campaign. The vacuity of the Bush campaign in 1992 owes a lot to Teeter's determination to poll every question before taking a decision.
But Dick Morris, who started as a pollster, is the apotheosis of the poll-driven consultant. His trademark is identifying piddling legislative solutions, short on substance but long on symbolism--like metal detectors in schools or school uniforms--test them through polls and then turn the ones with 75 percent approval into campaign centerpieces.
Morris has been hailed as the architect of Clinton's comeback. There is in fact a long list of poll-driven advice Morris gave the president which, unfortunately for the GOP, the president ignored--like introducing an across-the-board capital gains tax cut, raising Medicare premiums, or signing the balanced budget agreement.
To believe that Clinton is 20 points ahead in the polls because he listened to Morris and introduced his puny initiatives is to believe that trees moving make the wind. In fact, Clinton is poised to win in November not because of what Morris did, but because of what Gingrich and the GOP failed to do when they listened to THEIR pollsters and ceded the moral high ground to the Democrats.
Perpetuating the mystique of pollsters further pollutes modern politics by encouraging a breed of politicians pathologically unencumbered by principles, promises, or concern for the country's future.
4:06 p.m. Monday 9/23/96
I see that the subject of this panel--"Are the Polls Polluting Politics?"--is polluting at least one panelist. The last sentence of Huffington's Monday comment contained eight words starting with the letter "P."
The comments reveal three aspects of the potential significance of polls--as influence on election outcomes, as predictors of election outcomes, and as influence on politicians, either as candidates or as policy makers. I would like to focus for the moment on the first two of these, which are closely related. That is, I suppose that if people did not think that the polls were good predictors they would not have effects on the behavior of campaigners or of voters.
Kurtz emphasizes the prominent role that polls have in media reporting on campaigns and the effect of drowning out the candidate's--in this case Dole's--message. While the prominence of the polls is obvious, the effect is less so. Is there a message that is being drowned out? Or do the polls get prominence because there is no news in the message? How many news stories can begin with a sentence like, "Robert Dole today proposed a 15 percent tax cut"? Is it impossible for a candidate to create a message that is more interesting and newsworthy than the day's polls? Or are the reporters too lazy or ignorant to recognize a message other than the poll?
Do we have any evidence on the connection between poll results and voter turnout? One hypothesis is that when the polls show a big gap voters stay home. Do we know this from the historical record?
Another hypothesis is that when the gap is large the trailer has difficulty raising money. Apparently the campaign managers believe that. The other day I received a call from the Republican National Committee to inform me that Clinton's lead had been cut in half (which I have been unable to verify) and asking me if I wouldn't therefore contribute some more money.
Barnes suggests that the polls are not good predictors, and that therefore neither the voters nor the campaigners should be much influenced by them. He cites a few examples to support that. What can we say about the reliability of polls? If their reliability is commonly overestimated, whose responsibility is it to correct that?
Public opinion about public opinion polls is of some interest. These figures are from the Gallup Poll Monthly, May 1996. Of the persons polled in 1996:
Eleanor Clift Fred Barnes Arianna Huffington Herb Stein
Fred Barnes Arianna Huffington Herb Stein