LizF
06-22-2007, 08:04 AM
To help keep things in perspective about those pesky polls, check out this article:
http://www.nytimes.com/2007/06/20/us/politics/20web-toner.html?_r=1&oref=slogin
June 20, 2007
The Limits of Early Polling
By ROBIN TONER
For at least a generation, press critics have complained bitterly about the news media's obsession with polls and the horse race aspect of the presidential campaign. The constant focus on who's ahead is a pointless distraction, those critics argue, from what should be the real substance of the campaign – such as what the candidates would do as president.
But there is another reason to avoid putting too much stock in any assessment of the horse race, more than six months before the voting begins in the Iowa caucuses: It may have very little relationship to the ultimate outcome.
Events change, voters change, candidates and campaigns implode, new ones enter. (Such as, say, Mayor Michael R. Bloomberg of New York, who appeared to edge closer to a race-scrambling independent candidacy this week.) Even the activists, at this stage of the campaign, are not terribly focused or committed. Pollsters say they may be measuring passing whims and name recognition as much as settled judgments.
Yes, the national polls show certain patterns: Among the Republicans, former Mayor Rudolph W. Giuliani of New York tends to be at the head of the pack; former Senator Fred D. Thompson and Senator John McCain are right behind, and former Governor Mitt Romney is moving up behind them. And yet, consider this: Only 44 percent of the Republicans who said the abortion issue was "very important to them" could identify Mr. Giuliani as the "pro-choice candidate" in a recent Pew Research Center poll.
As Andrew Kohut, president of the Pew center, put it, "People aren't paying a lot of attention. Their attitudes are soft, so tests of their preferences are not reliable."
A little history backs him up: Four years ago at this point, Democrat Howard Dean's insurgent campaign was beginning to take off; his fundraising, especially over the internet, shocked his rivals in the end-of-June campaign finance reports. "In May of 2003, nobody thought Howard Dean was possible," said Geoffrey Garin, a Democratic pollster. "By the fall of 2003, the talk was that he was unstoppable. And the common wisdom was wrong at both points."
By the end of 2003, Mr. Dean was the clear frontrunner for the Democratic nomination in some polls. Then, in 2004, the voters entered the equation, and Senator John Kerry of Massachusetts, the eventual Democratic nominee, suddenly began to move. "Three weeks out, let alone seven months out, John Kerry was well behind in Iowa and New Hampshire – and the truth is, in the course of less than three weeks, he tripled his vote in Iowa and quadrupled his vote in New Hampshire," said Mark Mellman, Mr. Kerry's pollster that year. Most voters simply didn't make their minds up until after January 1st, Mr. Mellman added.
Bill McInturff, a pollster for Mr. McCain in 2000 and today, said that national polls are of limited utility at this stage of the contest, and state polls are subject to dramatic change. "Results in the early races ripple through the primaries," he said, recalling Mr. McCain's upset victory in New Hampshire seven years ago, just before the race moved to South Carolina. "The day before New Hampshire, McCain was 20 points down in South Carolina. The day after, he was tied."
The 1992 campaign offered another object lesson in the limits of early polls: That big, outside events can upend any tidy media narrative for how a campaign will proceeed. President George H.W. Bush seemed invincible after the 1991 war in the Persian Gulf, permanently lodged at the top of the opinion polls. But the memory of the war receded, and the economic slowdown settled in. And suddenly, in the fall of 1991, Mr. Bush's popularity plummeted, and the Democratic nomination seemed worth having.
Polls, of course, are only one way to handicap a race. Fundraising tells a lot, and so do more intangible standards — like a candidate's experience, their discipline and their ability to avoid mistakes, or to handle them once made. Some candidates, like Bill Clinton in 1991, have a strategic sense and a political rationale – where they want to take the party and the country -- that makes their campaigns seem sturdier from the start.
And polls can be useful beyond the horse race; for example, for exploring Senator Hillary Rodham Clinton's continued strength among women, or Senator Barack Obama's among affluent voters, or the intensity of conservatives in the post-Bush era.
But some analysts and political operatives have long complained that horse race polls, at this stage of the election, can become a self-fulfilling prophecy – a candidate does not do well in the polls, therefore does not get good (or much) coverage, therefore does not raise money, therefore does not do well in the polls. Mark Mellman, who does not have a horse in the current race, says, "What you've got is a lot of information – poll numbers, early endorsements – that don't necessarily have a lot of real meaning to the outcome of the race, but the campaigns peddle them as metrics of success."
Mr. McInturff said, "Polling affects press coverage, which affects your name i.d., which affects polling."
Even so, Mr. McInturff hastens to add that candidates can defy this by breaking out in early states, as his candidate did seven years ago.
This, in the end, is a source of comfort to the candidates who do not place near the top in the current horse-race polls, such as Senator Joseph R. Biden Jr., the Delaware Democrat. "Notwithstanding what you write about," he told a group of reporters this week, "nobody has made up their minds in the Democratic primary." He added, "Nobody."
James Carville, a principal advisor to Mr. Clinton in the 1992 campaign, puts it slightly differently: "I've never seen so many pay attention to so much that means so little."
http://www.nytimes.com/2007/06/20/us/politics/20web-toner.html?_r=1&oref=slogin
June 20, 2007
The Limits of Early Polling
By ROBIN TONER
For at least a generation, press critics have complained bitterly about the news media's obsession with polls and the horse race aspect of the presidential campaign. The constant focus on who's ahead is a pointless distraction, those critics argue, from what should be the real substance of the campaign – such as what the candidates would do as president.
But there is another reason to avoid putting too much stock in any assessment of the horse race, more than six months before the voting begins in the Iowa caucuses: It may have very little relationship to the ultimate outcome.
Events change, voters change, candidates and campaigns implode, new ones enter. (Such as, say, Mayor Michael R. Bloomberg of New York, who appeared to edge closer to a race-scrambling independent candidacy this week.) Even the activists, at this stage of the campaign, are not terribly focused or committed. Pollsters say they may be measuring passing whims and name recognition as much as settled judgments.
Yes, the national polls show certain patterns: Among the Republicans, former Mayor Rudolph W. Giuliani of New York tends to be at the head of the pack; former Senator Fred D. Thompson and Senator John McCain are right behind, and former Governor Mitt Romney is moving up behind them. And yet, consider this: Only 44 percent of the Republicans who said the abortion issue was "very important to them" could identify Mr. Giuliani as the "pro-choice candidate" in a recent Pew Research Center poll.
As Andrew Kohut, president of the Pew center, put it, "People aren't paying a lot of attention. Their attitudes are soft, so tests of their preferences are not reliable."
A little history backs him up: Four years ago at this point, Democrat Howard Dean's insurgent campaign was beginning to take off; his fundraising, especially over the internet, shocked his rivals in the end-of-June campaign finance reports. "In May of 2003, nobody thought Howard Dean was possible," said Geoffrey Garin, a Democratic pollster. "By the fall of 2003, the talk was that he was unstoppable. And the common wisdom was wrong at both points."
By the end of 2003, Mr. Dean was the clear frontrunner for the Democratic nomination in some polls. Then, in 2004, the voters entered the equation, and Senator John Kerry of Massachusetts, the eventual Democratic nominee, suddenly began to move. "Three weeks out, let alone seven months out, John Kerry was well behind in Iowa and New Hampshire – and the truth is, in the course of less than three weeks, he tripled his vote in Iowa and quadrupled his vote in New Hampshire," said Mark Mellman, Mr. Kerry's pollster that year. Most voters simply didn't make their minds up until after January 1st, Mr. Mellman added.
Bill McInturff, a pollster for Mr. McCain in 2000 and today, said that national polls are of limited utility at this stage of the contest, and state polls are subject to dramatic change. "Results in the early races ripple through the primaries," he said, recalling Mr. McCain's upset victory in New Hampshire seven years ago, just before the race moved to South Carolina. "The day before New Hampshire, McCain was 20 points down in South Carolina. The day after, he was tied."
The 1992 campaign offered another object lesson in the limits of early polls: That big, outside events can upend any tidy media narrative for how a campaign will proceeed. President George H.W. Bush seemed invincible after the 1991 war in the Persian Gulf, permanently lodged at the top of the opinion polls. But the memory of the war receded, and the economic slowdown settled in. And suddenly, in the fall of 1991, Mr. Bush's popularity plummeted, and the Democratic nomination seemed worth having.
Polls, of course, are only one way to handicap a race. Fundraising tells a lot, and so do more intangible standards — like a candidate's experience, their discipline and their ability to avoid mistakes, or to handle them once made. Some candidates, like Bill Clinton in 1991, have a strategic sense and a political rationale – where they want to take the party and the country -- that makes their campaigns seem sturdier from the start.
And polls can be useful beyond the horse race; for example, for exploring Senator Hillary Rodham Clinton's continued strength among women, or Senator Barack Obama's among affluent voters, or the intensity of conservatives in the post-Bush era.
But some analysts and political operatives have long complained that horse race polls, at this stage of the election, can become a self-fulfilling prophecy – a candidate does not do well in the polls, therefore does not get good (or much) coverage, therefore does not raise money, therefore does not do well in the polls. Mark Mellman, who does not have a horse in the current race, says, "What you've got is a lot of information – poll numbers, early endorsements – that don't necessarily have a lot of real meaning to the outcome of the race, but the campaigns peddle them as metrics of success."
Mr. McInturff said, "Polling affects press coverage, which affects your name i.d., which affects polling."
Even so, Mr. McInturff hastens to add that candidates can defy this by breaking out in early states, as his candidate did seven years ago.
This, in the end, is a source of comfort to the candidates who do not place near the top in the current horse-race polls, such as Senator Joseph R. Biden Jr., the Delaware Democrat. "Notwithstanding what you write about," he told a group of reporters this week, "nobody has made up their minds in the Democratic primary." He added, "Nobody."
James Carville, a principal advisor to Mr. Clinton in the 1992 campaign, puts it slightly differently: "I've never seen so many pay attention to so much that means so little."