Jan 26 2008
Even Pollsters Have Their Doubts
The polls are very fickled this year. It’s as if the public is saying ‘catch us if you can, but we are going to make some changes’. I predicted a while back Obama would win SC due to the Clinton’s screw up in playing the race card (why take a fake first black president when you can have the real thing?). But what I found really interesting today is a major polling outfit (SurveyUSA) coming out and admitting their numbers may be suspect:
Last time 14 pollsters all told you that Barack Obama would win a Primary, all 14 were wrong. That was the day before New Hampshire, 18 days ago.
History has an opportunity to repeat itself this weekend:
7 89 pollsters are working the South Carolina Democratic Primary, and all7 89 have Barack Obama ahead.
It is a refreshing moment of honesty about samples and turnout models which can be seriously flawed. I still think Obama takes SC in a storm today, crippling Clinton for Super Tuesday. But it should be noted that people cannot take polls too literally – you have to understand the inherent errors and the fact they are vague snapshots in time, not precise measurements.
Dick Morris has some interesting insight on the Clinton/Obama race in SC that are worth the read.
Polls only matter when people vote on election day. The way things go now, people can vote in some places starting a month before the election. So the final election day count is going to be nothing at all like the final election day polls. Also, as more ballots are cast by mail, the less meaningful are the exit polls. If 50% of the ballots were cast already, the exit polls on election day may not reflect the actual vote count at all.
I think Obama will win in SC. If I remember correctly Jesse Jackson won in SC a few years ago. It is down the road I am not sure about.