Polls, polls, polls

As the election countdowns begin, we’re starting to see some more frequent polling data, which is a relief after the snazzy, yet frustratingly out-of-date 2006 Election Guide the New York Times has had up for the past few months. Sure it is color-coded and has interactive features which allow users to act out their fantasies regarding the November results, but they’re using polls from March.

The Wall Street Journal and John Zogby released polling data on Monday showing Nussle leading Culver 46-43. The Nussle campaign’s email, after employing a bizarre analogy comparing the gubernatorial race to the upcoming Cy-Hawks game (is Nussle ISU or UI? I have no idea…), was cautiously optimistic:

We are prepared for this race to go down to the wire - and we know the only poll that matters in the one on Election Day.

Even so, it’s worth looking inside the WSJ numbers, and the news is not good for Chet Culver.

(The IRP offers a more triumphant and less credible analysis.) Of course, looking inside the numbers would reveal that they are within the margin of error, just as has been reported in a KCCI poll released Wednesday that shows Culver up 48-43, with a 4 point margin of error.

Given that is seems unlikely that Nussle would lose three points while Culver gained five in the span of a few days (would that it were so), let’s assume that the differences arose from the fact that Zogby is a conservative pollster, KCCI being more liberal. So what’s the take home lesson? Surprise, surprise, it is going to be a close race.

Update: the Register just released a poll today (September 17) that shows Nussle and Culver tied 44-44. It offers some interesting demographic analyses.

Add comment September 15th, 2006