January 10, 2008
It looks like the pundits and the polls were wrong on this one. Clinton was able to grab a narrow victory in New Hampshire, and consequently restore confidence in her campaign. Unfortunately, it looks like the news media is unwilling to evaluate its methods, and are busy spinning a new narrative. Here’s a few headlines from the past 24 hours:
CNN: Analysis: New Hampshire crowns two new ‘comeback kids’
CBS: Obama Supporters See Triumph Then Disaster
WaPo: Back From The Brink
From the perspective of the mainstream media, there have been two seismic shifts in the Democratic race for the nomination. From a more sensible perspective though, nothing big has happened at all. Clinton - heading a powerful political organization - won a relatively small victory in a state she had been leading for some time now. Obama’s performance was as predicted; he did extremely well with independents and younger voters but had less support among older women, who were the largest group of primary voters.
Clinton hasn’t made a “comeback” and Obama isn’t “faltering”; Clinton is still the frontrunner, and Obama is still the “underdog.” New Hampshire showed us that any attempt to prematurely end the race, will be “foiled” by the voters themselves. It’s still a wide-open race, and considering that this is supposed to be a democratic process, that is nothing but a good thing.
January 10, 2008 at 12:40 pm
“It looks like the pundits and the polls were wrong on this one.”
Nah, just the pundits. Polls are by definition out of date, since by the time a pollster calls someone, the last person they called may have changed their mind. This is especially true in places filled with undecided or cantakerous people, which by all accounts New Hampshire is, and in practice most polls are conducted over 2-3 days and then released a day or two later. Polls aren’t a “snapshot in time”; they’re more like the old blurry photographs back from when exposure times were measured in minutes.
The fault lies with pundits and journalists (and pollsters who, after all, need to continue making a living) who leave the impression that a poll released on Monday indicates what people were thinking on Monday. If anything*, it reflects what people were thinking on, oh, Saturday or Sunday. And maybe not even that, if something happens during the calls which can change a voter’s mind. And such things — let’s call then “campaigns” — are of course happening all the time.
* “If anything,” because aside from the timing problem, polls can have a slew of other methodological problems.