PPP poll confirms- with high turnout, Barrett wins

Lot of attention from yesterday's PPP poll on the recall election. And like other previous polls, it showed a slight Walker lead at 50-47. However, like most other polls, it has its flaws and skews, and a closer look will show this thing is dead even.

Like most other polls, the crosstabs tell the real story, and a couple of figures jump out at you. First of all is the traditional questions on Party ID and ideology.

Party ID, PPP Poll
Democrat 32%
Republican 34%
Independent/Other 34%

Ideology, PPP Poll
Very liberal 9%
Somewhat liberal 17%
Moderate 35%
Somewhat conservative 24%
Very conservative 16%

Both of these numbers do not jibe with the reality of the 2010 Wisconsin exit poll, and this time, I'll use CNN's 2010 Governor exit poll, which was right on with a prediction of Walker by 6.

Party ID, 2010 Wisc exit poll
Democrat 37%
Republican 36%
Independent 27%

Ideology, 2010 Wisc exit poll
Liberal 21%
Moderate 42%
Conservative 37%

Both show a smaller GOP turnout, but a mixed picture with a lower Dem number, but a higher liberal number. But there's a second variable in play here. Given that the GAB has indicated that turnout will be well above the 2010 turnout of 50%, and closer to the 2008 turnout number of 69%. If that's the case, you get a very different electorate, as the 2008 Wisconsin presidential election poll shows.

Party ID, 2008 Wisc exit poll
Democrat 39%
Republican 33%
Independent 29%

So instead of the GOP +2 in the PPP Poll, you have 2010 polls with Dem +1, and 2008 with Dem +6. With this in mind, and with turnout expected to be closer to the 2008 turnout than even the 2010 turnout, that poll sample should be adjusted for who will actually be in the polls. So let's use figures between the 2008 and 2010 turnouts as a general rule of thumb, and use that as a predictor for tomorrow. In Party ID cases, this would make the turnout Dem 37, GOP 34, Independent 28. So put that in with the same PPP crosstabs and what do we get?

Barrett 48.38, Walker 47.45

And even though we don't have a 2008 ideology stat, let's use the 2010 figures and bump liberal up 2 and moderate down 2, turning it to 23-40-37 liberal-moderate-conservative. Then we'll run PPP's numbers on that., and we get...

Walker 49.25, Barrett 47.80

So even though Walker leads, it's cut in half to a virtual toss-up. In addition to political preferences, there are demographics that the PPP was off in, and a regular turnout would swing these votes toward Barrett. Here's a couple to go over.

Age of respondents, PPP Poll
18 to 29 10%
30 to 45 26%
46 to 65 39%
65 and over 25%

Now compare it to turnout history.

Age of voters, 2010 exit poll
18 to 29 15%
30 to 44 23%
45 to 64 46%
65 and over 16%

Age of voters, 2008 exit poll
18 to 29 22%
30 to 44 29%
45 to 64 35%
65 and over 15%

So the PPP poll oversamples young voters and oversamples elderly voters. This is a big difference that favors Walker in the poll, as Scotty's base of support in the PPP Poll are people on Social Security.

18 to 29 53-39 Barrett
30 to 45 52-46 Barrett
46 to 64 51-47 Walker
65 and over 58-39 Walker

So let's use an average of age turnout for the 2008 and 2010 exit polls, which basically slides 8% of the voters from the old to the young, and see what the results become.

Walker 48.66, Barrett 48.10

I also want to bring up minority voting. The PPP poll shows Walker getting a surprising number of minority votes, with Barrett only winning 58-36 among non-white voters (who are correctly listed as around 11% of the vote). Except that Barrett won the non-white vote 69-31 in 2010, and Obama won it approximately 75-25 in 2008. I'd guess 70-30 is a more likely number tomorrow, particularly given Walker's Milwaukee-bashing and J.B Van Hollen's race-baiting "voter supervision" acts, as well as with people like Jesse Jackson and Al Sharpton holding turnout rallies in Milwaukee's inner city. That 10-to 12-point addition in minority voting share would allow Barrett to gain 2-3 points on this alone.

Also, the PPP poll has an unusual response to the following question:

Who did you vote for in 2008?
Obama 49, McCain 42. Actual vote: Obam 56, McCain 42.

It doesn't ring true that all 7% that drop off from the 2008 election would be Obama voters, especially when it likely that some of McCain's voters WOULD BE DEAD, given how old the average GOP voter is. A more realistic figure would be 51-40, if you think about 9% have dropped off. Turn that change into the PPP Poll and you get:

Barrett 48.50, Walker 47.98

The common thread in all of this, Barrett gains when you look at how the demographics tend to play out on Election Day, and that this race is as close as you can get. And if Barrett and the Dem groups are successful in turning out their voters, they have far more upside vs. Walker, who's pretty much going to get the same amount of votes regardless of the total turnout, just based on the type of person that votes for Walker (old white men outside of Milwaukee and Madison- Barrett wins practically every other demo).

With that in mind, my prediction for tomorrow is straightforward. If there are more voters in the City of Milwaukee than Waukesha County, and if the youth and minority turnouts are anything close to their 2008 levels, Tom Barrett will be the new govenror of Wisconsin. That's not a guarantee, but it's also an outcome that I think has a good chance of happening. I'd be shocked if the Obama people, unions, and Barrett's campaign do not execute tomorrow, regardless of the tricks that the GOP will try to pull.