The Boston Globe just posted a write-up of their new poll of the US Senate race. The poll shows a statistical tie, with Scott Brown a nose ahead of Elizabeth Warren, 39%-37%. This means the race is unchanged since March, when the last Globe poll had Brown a nose ahead, 37%-35%. So, another poll shows a neck-and-neck race in which Warren's position has not been hurt by the controversy over her heritage. "The vast majority of voters" in the Globe poll say that issue is not significant and would not affect their vote.
But that's not what I want to talk about. I want to talk about why the Globe had their usual pollster, Andrew A. Smith of the university of New Hampshire, conduct the poll and provide the analysis quoted in the article.
See, I might have thought they'd have a little heart-to-heart with Smith after he was quoted, twice, in Boston Herald articles playing up the political damage being done to Warren -- without, incidentally, any mention that he does the Globe's polling on the Senate race.
Here he is in the Herald on May 21, in a story about whether Massachusetts Democrats will ditch Warren as their candidate:
“The Democratic Party is really stuck,” countered University of New
Hampshire political science professor Andrew Smith. “They essentially
cleared the path for her as a candidate, and they can’t get rid of her
now. She could conceivably drop out, but I doubt that will be the case,
and I doubt the party will try to push her aside."
....“They’re in a tough spot, but there’s not a lot they can do about it,” Smith said.
And the next day, May 22, in an article about how much damage may have been done to Warren even though polls were not showing it:
pollsters said it’s likely television ads will still mock Warren... on the issue.
“As a candidate, that’s the worst thing,” said pollster Andrew E.
Smith, director of the University of New Hampshire Survey Center. Smith
said he believes “this will definitely be used by the Brown campaign.”
Eventually, Smith said, voters will find out whether Warren really
does have Cherokee roots, and decide for themselves whether her claims
of minority status played a role in the jobs she landed at two top Ivy
League law schools.
“Any genealogist worth their salt is trying to investigate this and follow up on it,” he said.
“They’re in a great position in the fall,” he said. “There’s people all over the country trying to investigate this.”
So here is Smith telling one of the state's major media outlets -- absent any polling evidence -- that the heritage issue was terribly damaging to Warren, damaging enough to have put the party in a tough spot, damaging enough that the Brown campaign will certainly use it.
And we now know that three days later, Smith had a new poll in the field for another major media outlet, asking whether people believed what he just told them.
Which, incidentally, they did not. (So I suppose we can at least think that Smith did not, consciously or not, skew the poll to get the results he already believed.)
And yet... Smith still insists otherwise, in interpreting his new poll for the Globe:
‘‘Overall, this shows the strengths that Brown has and it shows the
problems, obviously, that the Warren campaign has had,’’ said Andrew E.
Smith, the Globe’s pollster.
Why does he say this? Because Warren's unfavorable ratings are up nine points in the past three months. Is there any reason for Smith to think that this increase -- to 32% -- had anything to do with the heritage issue? No, not really.
I don't know if Smith's poll is bad, or for that matter is his analysis is bad.
But I do think the Herald, the Globe, and UNH all have reason to wonder about Smith.
Latest Slideshows
