USA TODAY

Wednesday, September 29, 2004, page 6A

Politics

 

Gallup defends results against MoveOn.org attack

 

Liberal critics argue that pollster's methods give Republicans advantage

 

By Mark Memmott, USA TODAY

 

This year's USA TODAY/CNN/Gallup polling on the presidential race is coming under attack from liberal critics who say Gallup's methods produce results that are biased in favor of President Bush.

 

Gallup says it stands by its work, which Tuesday showed Bush ahead of Sen. John Kerry by 8 percentage points among likely voters -- if the election were held "today." That was down from a 13-point lead for Bush two weeks earlier. Gallup's recent polls have consistently shown Bush further ahead than he is in other surveys.

 

USA TODAY said Tuesday it remains confident in Gallup.

 

The firm, USA TODAY Polling Editor Jim Norman said, "has been doing USA TODAY's political polling since 1988. ... I'm impressed by the constant testing and retesting they do of their methodology to make sure they get it right."

 

CNN spokesmen were unavailable for comment.

 

In recent weeks, complaints about the USA TODAY/CNN/Gallup Polls had mostly been aired in "blogs," or online diaries.On Tuesday, the issue spilled into the "mainstream media." MoveOn.org, a liberal advocacy group, paid $68,000 to run a full-page ad in The New York Times. The ad's headline: Gallup-ing to the Right. Why does America's top pollster keep getting it wrong?"

 

The ad goes on to say that "two media outlets, CNN and USA TODAY, bear special responsibility" because "they pay for many of Gallup's surveys."

 

At issue: Whether too many Republicans end up being counted as "likely voters" in Gallup's polls. In the past six USA TODAY/CNN/Gallup Polls this year, about 40% of the likely voters in the surveys said they considered themselves to be Republicans. By one measure, that's higher than might be expected: Exit polls after the past three presidential elections showed that about 35% of voters in those years said they were Republicans.

 

Peter Schurman, MoveOn.org's director, said Tuesday that Gallup "should admit its mistake and correct it by using samples that more closely reflect" likely turnout.

 

Frank Newport, editor in chief of the Gallup Poll, said the critics don't understand the science behind the polls. "This issue has been the subject of intense scholarly discussion and years of research. We're confident in what we're doing," he said.

 

Actually, it's what Gallup doesn't do that is at the heart of the debate. The polling firm does not adjust its "pool" of voters to add or subtract Republicans or Democrats in an effort to mirror those parties' estimated makeups.

 

Among the reasons Gallup doesn't try to do that:

 

--It believes there are no reliable data on which to estimate exactly how many Republicans or Democrats there are in the country. Some states, for example, don't require voters to register by party affiliation. Basing an adjustment on previous year's exit polls, "means you're 'weighting' one poll based on the results of another poll, which has its own built-in sampling error," Newport said.

 

--It believes that party affiliation "is an attitude, not a demographic trait" and that voters can change their minds about which party they identify with more than once during an election year, Newport said. That would explain, he said, why the number of people who identified themselves as Republicans went down during this year's Democratic primaries, when Kerry and his competitors were in the news.

 

Most polling firms use the same methods as Gallup when identifying party affiliations. Among those are the surveys done by the Pew Research Center for the People and the Press. Andrew Kohut, the center's director, said in a statement last week that "important shifts in voter sentiment" could be missed if pollsters tried to apply rigid party formulas to results.

 

But not all pollsters agree with Gallupís approach. John Zogby is CEO of the independent polling firm Zogby International.

 

He adjusts the voter pools in his surveys to mirror party affiliations expressed in earlier exit polls. "I am one of the heretics in the polling industry," he said Tuesday. He maintains that there are variations in peopleís party affiliations, but they aren't changing much daily, weekly, or even monthly."

 

Critics say the debate over Gallup's work is important because the media's reporting of polls can affect the dynamics of a campaign. "We need the most accurate information possible. Next week, the stories could be 'Kerry's surging in the polls,' but would that be true?" asks Markos Moulitsas Zuniga, who's dailykos.com blog is a popular site among liberals.

 

Norman, however, said that the Gallup's "overall record since 1988 on presidential elections, senate races and the national vote on congressional elections is as good as anyone's. And their record in the 2002 elections was clearly the best."

 

GOP represented in greater numbers

 

The Gallup Poll asks voters to identify their party affiliation after they tell pollsters whom they would vote for if the election were that day. How the party affiliations compare with poll results among likely voters:

 

Source: Gallup Polls; those marked with an asterisk were conducted independent of USA TODAY and CNN

 

Poll

Republican respondents

Independent respondents

Democratic respondents

GOP advantage over Democrats

Bush support

Kerry support

Jan. 9-11

39%

30%

31%

8%

55%

43%

Jan.29-Feb.1

36%

30%

35%

1%

46%

53%

Feb. 6-8

37%

31%

32%

5%

49%

48%

Feb. 16-17

32%

33%

36%

-4%

43%

55%

Mar. 5-7

35%

26%

40%

-5%

44%

52%

Mar. 26-28

44%

23%

32%

12%

51%

47%

Apr. 5-8*

41%

24%

34%

7%

48%

45%

Apr. 16-18

41%

25%

34%

7%

51%

46%

May 2-4*

38%

23%

38%

0%

48%

49%

May 7-9

37%

30%

33%

4%

48%

47%

May 21-23

38%

26%

34%

4%

47%

49%

June 3-6*

36%

27%

36%

0%

44%

50%

June 21-23

36%

28%

37%

-1%

49%

48%

July 8-11

38%

24%

38%

0%

46%

50%

July 19-21

41%

23%

35%

6%

47%

49%

July30-Aug.1

42%

23%

34%

8%

51%

47%

Aug. 9-11*

40%

25%

34%

6%

50%

47%

Aug. 23-25

38%

29%

32%

6%

50%

47%

Sept. 3-5

40%

27%

33%

7%

52%

45%

Sept. 13-15

40%

27%

33%

7%

55%

42%

Sept. 24-26

43%

25%

31%

12%

52%

44%