The New York Times and CBS News, in partnership with YouGov, released the results of a controversial survey that shows if the election were held today, GOP Senate candidate Dan Sullivan would lose to Mark Begich by 12 points, and Lt. Gov. Mead Treadwell would lose by only two points. The survey provided fodder for Treadwell’s campaign to claim momentum.
“This new nationally-recognized poll and the recent attacks we’ve seen from Mark Begich and his liberal allies show we’re surging and that we’re best positioned to make Begich a one-term Senator,” Treadwell said in a release.
The poll, however, lends little credibility to his claim.
The survey interviewed 452 Alaskans as part of a national project that asked 100,000 voters nation-wide questions about every Senate race in the country. The survey says that Republicans have a 60 percent chance of winning the Senate back.
Beware: The survey doesn’t even conform to the New York Times’ standards of what it considers a “credible” poll. What’s so controversial about it is that it used “opt-in” respondents, meaning basically that people had to go to them online to get surveyed. Such survey methods fly in the face of what has been considered scientifically sound polling, because, among other things, people who opt in for such surveys tend to be more politically engaged and more educated.
The story accompanying the Times piece on the poll had a cautionary paragraph for the Alaska results, noting that in Alaska, “internet penetration is surely lower than elsewhere in the country.” However, the NYT’s apparently didn’t check its facts. Internet penetration in Alaska is among the highest in the country.
Indeed, the survey results in Alaska are outliers. Nearly every other poll has shown both Treadwell and Sullivan neck-and-neck with Begich.
The New York Times itself refused to publish online polls prior to this one, and it still has a policy that says that it won’t publish exactly the same kind of poll that it ended up publishing. A reporter for Politico Pro (paywall alert) caught that one. From the paper’s policy, which was still on the paper’s site on Monday:
Self-selected or ‘opt-in’ samples — including Internet, e-mail, fax, call-in, street intercept, and non-probability mail-in samples — do not meet The Times’s standards regardless of the number of people who participate…In order to be worthy of publication in The Times, a survey must be representative, that is, based on a random sample of respondents. Any survey that relies on the ability and/or availability of respondents to access the Web and choose whether to participate is not representative and therefore not reliable. The hallmark of any good poll is that the poll taker chooses and pursues the respondent.
I don’t know whether or not Treadwell is surging. He might be. He might not be. But I would think that if it’s true, and he’s the candidate to beat Mark Begich, then either Mark Begich or the super-PAC that supports him would have put some money into going after him. They haven’t.
What I do know is that this poll is highly suspect.
As Washington Post pollster Scott Clement put it:
— Scott Clement (@sfcpoll) July 27, 2014
Contact Amanda Coyne at email@example.com