Are Pollsters the Biggest Losers in NH?

How much should we blame predictors for getting it wrong?
By Rob Quinn,  Newser Staff
Posted Jan 9, 2008 1:17 PM CST
Democratic presidential hopeful, Sen. Hillary Clinton, D-N.Y., embraces her husband former President Clinton during a victory party after winning the New Hampshire Democratic primary in Manchester, N.H.,...   (Associated Press)
camera-icon View 5 more images

(Newser) – The New Hampshire primary was the biggest high-profile poll upset in memory, says Talking Points Memo’s Josh Marshall—but don’t jump too far down pollsters’ throats. Marshall has a hunch the polls weren’t “wrong,” but rather couldn’t capture a “late swing,” a last-minute “transcendent moment.” Pundits my have been pulling for Obama, but collusion with pollsters doesn't track, he argues. “Polls are usually right.”

It might be that Hillary Clinton connected in the final moments, or that Obama’s bounce dissolved naturally, but it’s no good to second-guess pollsters who missed “a dramatic, unexpected result.” It's convenient, adds ABC’s Gary Lawler, for pollsters to “fault their respondents” rather than “own up” to sampling and modeling failures—but last-minute decisions and two “groundbreaking candidates” may just have overwhelmed the wisdom of surveys.