The picture in the Iowa primary has changed dramatically over the last week. Hillary Clinton, who had been trailing badly, has completed a turnaround and now leads Obama, 46.8% to 40%.
Here’s her performance over the last month:
After reaching her nadir on the 19th of December, the last ten days have seen almost constant improvement. It’s perhaps too early to tell what the assassination of Bhutto yesterday will do, but if it does anything it’s hard to see it hurting her already brightening chances.
You can see where this support comes from. She’s taken it from Obama:
Pollster.com agrees and sees the two candidates as essentially tied, a big setback for Obama after he had caught up very well with Clinton:
It’s remarkable how much things have changed, but remember, this is not due to the Bhutto assassination. That may eventually be a factor (as it may help Guiliani on the GOP side) but this change started before Christmas. As I blogged on the 21st:
And what of Edwards? There’s some speculation that anything on the foreign scene may help him as well. Musharref phoned him yesterday. So far, not much though:
We also should keep in mind that these “polls” are still showing a lot of movement and that they only cover Iowa. After that, there’s New Hampshire. Here. Clinton was already strong and given the lead by Intrade (anything over 50% means you are the most likely to win a two horse race):
But even here she shows improvement, again on or around the 19th of December. If she wins these two early primaries, it’ll be extremely hard to dislodge her.
Note: I prefer to use Intrade prediction markets rather than polls for analysis. Partly this is because I am very interested in crowdsourcing and its effectiveness, so I want to track what they say in some detail as we close in on the voting (the Iowa caucus is less than a week away remember).
But also polls are covered quite well on many other sites and attract a lot of attention (eg the LAT poll yesterday). I think there are quite a few problems with polls and more so with how they are interpreted. The LAT poll is a case in point: the headlines were “statistical dead heat in Iowa” but today that is revised to show Clinton ahead.