Nearly all the assumptions going into the Iowa caucuses last week were wrong. Many a pundit wondered aloud as to whether Kerry’s predicted third place finish would doom his candidacy, right up until he won easily. An equal number expressed their awe and admiration for Dean’s incredible Iowa field operation, right up until he finished a distant third.
OK, so nearly all of the pundits were terribly, almost embarrassingly, mistaken, but how did the pollsters do? After all, most talking heads base their analysis on the latest polling data, so if the polls are wrong, the pundits are wrong.
The afternoon of the Iowa caucuses, just a few hours before things actually got started, Daily Kos ran a compilation of all the latest data from the five most widely read polls available. None of the five were particularly close. Three of the five at least got the winner right, while only one — the Des Moines Register — actually got the final order right.
Two of the pollsters — Survey USA and Research 2000 — showed Dean with a slender first-place lead going into Jan. 19, which was obviously very wrong. Survey USA even had Kerry running third.
All five of the pollsters showed Gephardt at or near 20 percent and within striking distance of a victory. (Actual results for Gephardt — 11 percent and a distant fourth.) All five also showed the margin between Kerry and Dean to be no more than six percentage points. (Actual margin after the caucuses — 20 percentage points.)
I’m mentioning all of this not to make fun of pollsters, but to remind you (and myself) that even today’s New Hampshire primary is still unpredictable. All of the clichés I can possibly think of apply: anything can happen, it’s up in the air, don’t count anyone out, it ain’t over ’til it’s over, looks can be deceiving, expect the unexpected, the only certainty is uncertainty, etc.
Polling in New Hampshire tends to be a little more reliable than in Iowa, because it’s more of a traditional voting process as opposed to the convoluted caucus system. Nevertheless, the road is paved with many embarrassed pollsters in the Granite State. In 2000, a New Hampshire tracking poll from the American Research Group, an outlet I generally trust, showed George W. Bush with a narrow 2-point lead over John McCain going into the primary. Once people actually voted, McCain had trounced Bush by 16 points.
All of the polls I’ve seen in the past few days have shown Kerry with a lead in New Hampshire (ranging anywhere from three points to 18), Dean in second, with Clark and Edwards battling it out for third place. Some polls even show Lieberman making a strong showing in the waning days, with one poll putting him third.
Could the results actually turn out this way? Of course, but don’t be surprised if the pollsters go 0-for-2.
For what it’s worth, here are the most recent polls I’ve seen, as of this morning. We’ll know in about 12 hours if they’re right.
MSNBC/Reuters/Zogby
Kerry 30 (28)
Dean 23 (25)
Clark 13 (11)
Edwards 9 (10)
Lieberman 9 (9)
CNN/USA Today/Gallup
Kerry 36 (38)
Dean 25 (25)
Clark 13 (10)
Edwards 10 (9)
Lieberman 10 (12)
American Research Group
Kerry 35 (38)
Dean 25 (20)
Edwards 15 (16)
Clark 13 (15)
Lieberman 6 (5)
Marist Institute
Kerry 37
Dean 24
Clark 11
Edwards 11
Lieberman 9
Boston Globe/WBZ
Kerry 37 (37)
Dean 20 (17)
Edwards 12 (12)
Clark 8 (11)
Lieberman 7 (7)