Monday, November 19, 2007

WI Presidential Head-to-Head Poll

Lots to blog about now that my race for Chair of the Democratic Party of Dane County is over. I'll talk about my defeat later, but I spotted this Survey USA poll and it's nice to see Wisconsin getting at least a little attention. All the usual caveats apply, keep in mind this is registered, not likely, voters, and unfortunately Edwards wasn't included:

(Click on image for larger version)

Clinton vs. McCain - Clinton wins by 2%
Clinton vs. Giuliani - Clinton wins by 7%
Clinton vs. Romney - Clinton wins by 16%
Clinton vs. Huckabee - Clinton wins by 17%

Obama vs. McCain - Obama loses by 4%
Obama vs. Giuliani - Obama wins by 5%
Obama vs. Romney - Obama wins by 12%
Obama vs. Huckabee - Obama wins by 17%

At this point the results are probably due more to name recognition and the generic Democratic advantage than "real" factors that will come into play as we get closer to the election next fall. Dems win every matchup except Obama vs. McCain. Clinton is significantly stronger than Obama against McCain and Romney, although not against Giuliani or Huckabee.

As usual, the real entertainment is in the crosstabs. For instance, Clinton crushes the Republican candidates among female voters. The gender gap is alive and well! But the age distribution contains good news and bad news. The good news is that young voters in Wisconsin are becoming strongly Democratic. The bad news is that they tend not to actually vote, at least compared to older voters. This means that any Wisconsin poll of registered voters, as opposed to likely voters, is probably going to be biased Democratic.

The age factor is mitigated somewhat here because voters 55 and up tilt significantly Democratic in most of these matchups, and they vote at higher rates than middle-aged voters. McCain is the exception because he has strong support among older voters. His actual numbers are probably a few percent better than the poll indicates, meaning he would likely beat Clinton as well as Obama here in Wisconsin.


I have some interesting projects on tap. I've calculated the Partisan Voter Index* (PVI) for each Wisconsin Senate district, and I'm working on the Assembly. This sort of data is hard to find for state races because it requires an analysis of election returns for each of the more than 3500 wards in the state. A lot of work, but I've automated much of it so it's only slightly painful. Political professionals have these numbers, of course, but they tend to be rather reluctant to share them.

My loss in the Dane Dems Chair's race also puts me in the interesting position of no longer being a member of the Dane Dems executive board. As a board member I was largely ineffective in achieving the changes and reforms I wanted, yet due to my position on the board I was reluctant to bring many of those issues to public attention. Those constraints no longer apply, opening up some intriguing new possibilities for building a bigger, better, and stronger Democratic Party.

*Technically the PVI as developed by Cook only applies to federal races, so I'm taking some liberties with the term even though the data analysis for state races is essentially the same as for federal races.