Tuesday, January 22, 2008

Polling News

As always, go to Pollster for the best selection of state and national polls, plus some of the most illuminating commentary on the ins and outs of political polling anywhere.

New Hampshire is going along with its hand-recount. Here is a page from the NH Secretary of State's office listing results from Hillsborough County. NOTE: the Weare results are not yet in so there is a funny glitch in everyone's recount. Just subtract Weare from the first column for a candidate and that takes it out as a factor.

It turns out that the machines did overcounts for everyone, even Obama. The margin of error was less than .001%. So much for the massive machine based fraud done on behalf of the Clinton campaign. For the wonks out there, here is a white paper discussing the various reasons for the higher vote margins for both Clinton and McCain, none of which have anything to do with rigged machines.

South Carolina polls are coming out (though none post-debate) and they are showing HRC with 15%-17% African American vote, which is what she has received in all other states except Michigan, where she got 30%. Obama is just under 75% of African American votes. After last night's debate performance, I expect that Edwards is going to come up sharply in South Carolina polls, though his gains will come more from HRC's side of the ledger than from Obama's. When I compare a pollster's number over time (such as all Rasmussen, all Survey USA, all PPP, etc.), I see that HRC was flattening out on her percentage of the vote in South Carolina back in December. Only one poll was done after Nevada, and that one shows declines or stalling for all three major candidates. The other polls do not reflect any effect the Nevada caucuses may have had - which I expect is not much. There are two things to note about South Carolina polls:
  1. Obama's gains are almost all immediately after the Iowa win, not after the race-baiting allegations. This points to African American voters being assured that they were backing a viable candidate, so moving from undecided to Obama. Overall, however, trend lines show HRC rising in the polls in SC. It will not be enough for a victory unless some very bad scandal hits the Obama campaign that SC voters would give two figs about. I don't see any.
  2. South Carolina has always been a tough primary for HRC. Her margins over the competition are narrower here than in any other state except Iowa, where she held the polling lead for a very short preiod of time, and actually took it back towards the end. The actual loss in Iowa, as I've explained elsewhere, is easily explained by Edwards' relative strength (drawing off her voters) and through excellent caucus "second-choice" strategy by Obama.

The campaign has made a strategic choice not to spend more political capital in South Carolina, which is probably a good decision. Obama will win reasonably handily, but it will be more like Michigan for Hillary than another Iowa win for him. No surprise, no real contest, and the other people are not giving him much fight.


No comments: