Wisconsin by the Numbers
Scott Walker cruised to a 53-46 win in the Wisconsin Gubernatorial Recall, stunning Democrats expecting a tight race after early exit polls.
Walker's victory was built on a GOP-friendly electorate, even whiter, older, richer, and less Democratic than the 2010 midterms, let alone 2008. Seniors represented 18% of the electorate, up from 16% in 2010 and 14% in 2008. The non-white share of the electorate fell from 11% in 2008 to 10% in 2010, to 9% in 2012. 20% of voters made more than $100,000/year, up from 16% in 2010 and 19% in 2008. Only 34% of voters identified as Democrats, compared to 37% in 2010 and 39% in 2008.
Although Barrett exceeded Kerry's margins in densely populated and heavily Democratic Milwaukee and Dane (Madison) Counties, Barrett fell well behind Democratic benchmarks across the rest of the state, outright losing many Kerry/Gore counties in the state's populist/progressive southwestern quadrant. In the traditionally Republican Milwaukee suburbs, Walker won more than 70% of the vote, including 72% of the vote in populous Waukesha County. No GOP presidential candidate has ever received more than 72% of the vote in Waukesha County, or at least not since 1892.
Despite Walker's impressive showing and the GOP-friendly electorate, exit polls showed Obama leading Romney among recall voters by 7%, 52-45. Although the large contingent of Obama-Walker voters seems to have surprised at least a few analysts, the exit poll was highly consistent with pre-election polling. According to the Washington Post, 59% of Obama-Walker voters were independent and 52% considered themselves moderate.
The Walker-Obama phenomenon might seem counterintuitive, but recall that independent voters support candidates with exceptionally different ideological perspectives all the time. Many independent voters are independent precisely because they are less invested in the ideological battles that dominate partisan politics, so we can't be surprised that they're willing to cast ballots for candidates with seemingly contradictory ideological positions.
Since independent voters don't necessarily cast their ballots on firm convictions about the appropriate role of government, likability and personal characteristics might provide a simple explanation. A pre-election Marquette poll showed that Obama and Walker were viewed favorably, while most voters were skeptical of Barrett and Romney.
Some have argued that Obama's 7% advantage is actually a sign of weakness, as Obama won the state by 14% in 2008. A similar shift nationally, these analysts contend, would put Obama's reelection in jeopardy since he won nationally by 7% in 2008. This argument has three weaknesses.
First, the Obama-Romney numbers should be considered in the context of a GOP-friendly electorate. The general election is likely to bring an increase in youth and non-white turnout, even if Obama can't replicate his 2008-showing. A 7% advantage yesterday might be indicative of a much larger advantage with a more neutral electorate.
Second, Wisconsin is an extremely white state. Since 2008, Obama has bled considerable support among white voters, even while maintaining the overwhelming support of African American and Hispanic voters. As a result, we would expect Obama to lose a larger share of his 2008 support in Wisconsin than the rest of the country.
Third, Obama's 14% victory in 2008 is not representative of Wisconsin's political leanings. Kerry and Gore won Wisconsin by less than 1% in 2004 and 2000, and Obama's 7% lead is somewhat more impressive in that context. Relatedly, it would not be surprising if Obama's standing fell more in Wisconsin than elsewhere, since his big win was built on a large number of Bush voters. That wasn't nearly so true nationally, as much of Obama's gains were the result of changes in turnout and big improvements among non-white voters.
On the other hand, weak turnout among core Democratic constituents illustrates the challenges ahead for the Obama campaign. There is no guarantee that young and non-white voters turnout to the same extent they did in 2008 and conservatives are clearly energized. In this context, the Obama campaign's decision to invest millions of dollars in the ground game and turnout operation is unsurprising.
11:52PM: Alright, I'm signing out. I'll probably breakdown the results over the next few days.
11:47PM: Apparently the newest exits show Walker up by 14% and Obama leading Romney by 7%, 51-44. Given that Walker leads by about 9% with 87% of precincts reporting, this might wind up adjusted in a pro-Obama direction.
11:26PM: I'm observing quite a bit of effort devoted to explaining the Obama-Walker phenomenon.
My advice: remember that independent voters support candidates of exceptionally different ideological perspectives all the time. Many independent voters are independent precisely because they are less invested in the ideological battles that dominate partisan politics, so we can't be surprised that they're willing to cast ballots for candidates with seemingly contradictory ideological positions.
The simplest explanation might be the best: a majority of independents like both Walker and Obama. A pre-election Marquette Poll showed Walker up 7 and Obama up 8, with both candidates sporting favorability numbers over 50%. In contrast, Barrett and Romney were mired at 41 and 40%. For independent voters unswayed by contemporary ideological clashes, likability might just be the decisive factor.
11:22PM: Although the Post didn't breakdown Obama-Walker supporters geographically, it's pretty clear that most of these voters are out in the countryside, or at least outside of Milwaukee and Madison. Barrett is matching Kerry's performance in those two Democratic strongholds, but he's far, far behind Kerry's results elsewhere.
11:19PM: It's too bad the exit poll didn't ask about Obama's approval. It would be nice to confirm that most undecided voters harbor reservations about Obama's performance.
11:05PM: Who are the Walker-Obama voters? The Washington Post has the raw exit poll data necessary to breakdown this small subset of the electorate:
"The exit poll — a survey of 2,457 randomly selected recall voters — has barely enough such Walker-Obama to analyze. But some breakdowns are clear: 59 percent are independents, far above the rate in the overall electorate. More than half — 56 percent — described themselves as moderates, again well above the number in the full voter population. Some 52 percent are male; 23 percent are from union households."
11:00PM: Walker's margin is beginning to shrink after a wave of Barrett-friendly returns in Milwaukee and Dane Counties. With 76% of precincts counted in Dane County, home to Madison and thousands of public-sector employees, Barrett holds 67.7% of the vote, just slightly less than the 67.9% he received in 2010. It will be interesting to see whether Barrett's lead continues to grow in Dane County, as it so far has throughout the evening.
10:51PM: TV coverage is harping on the Obama-Romney numbers in the exit poll, but the Obama-leaning pundits might want to be careful. The adjusted exit poll reflects a 5% Walker advantage, but Walker's on pace for a much more substantial victory. New numbers seem likely to tighten the horserace, even if Obama is still likely to hold a lead. Those pundits should hedge their bets by emphasizing the composition of the electorate, which was older, whiter, richer, and less Democratic than even the relatively GOP-friendly 2010 midterms.
10:47PM: So far, Walker has flipped Crawford and Green Counties from Barrett to Walker. He also leads in Trempealeau, Eau Elaire, and Portage, all won by Barrett in 2010, despite a large number of votes. Barrett trails in Milwaukee, although that's sure to change once more votes from the city are tabulated.
10:32PM: How much might Madison and Milwaukee reduce Walker's lead? Quite a bit, but not nearly enough. In 2010, Barrett netted 162,000 votes in Dane and Milwaukee Counties, but Barrett has already picked up about 50,000 of those votes, mainly in Madison. However, Walker currently leads by 195,000 votes, and plenty of Republican votes are left to be counted in the Milwaukee suburbs and heavily Republican stretches of the eastern Wisconsin countryside.
10:25PM: The adjusted exit polls show that Walker's victory relied on a richer, older, whiter, and less Democratic electorate than 2010, let alone 2008.
Seniors represented 18% of the electorate, up from 16% in 2010 or 14% in 2008. The non-white share of the electorate fell from 11% in 2008, to 10% in 2010, to 9% in 2012. 20% of voters made more than $100,000/year, up from 16% in 2010 and 19% in 2008. Only 34% of voters identified with the Democratic Party, down from 37% in 2010 and 39% in 2008. Again, these numbers are subject to revision.
10:20PM: Four counties have finished counting their votes, and every county shows Walker winning a larger share of the vote than he did in 2010.
10:14PM: It's worth noting that these updated exit polls reflect a modest Walker lead of about 4/5%.
10:12PM: The re-tabulated exit poll numbers - which still might not be final - show Obama leading Romney by 11%, 53-42%. According to the exit poll, 17% of Obama supporters voted for Walker. While that might seem counter-intuitive, it's actually been a pretty consistent pattern in pre-election polling.
10:07PM: Revisiting the great chart from the Huffington Post, you can see that Walker is outperforming his 2010 performance across the board. While it's possible that a strong Barrett showing in Milwaukee and Madison could keep Walker's margin near 2010 levels, it looks like Walker will win by a wider margin than last time.
10:03PM: 95% of precincts have reported in Door County, a middle class, swing county jutting into Lake Michigan. Bush received 51% of the vote and Walker received just 50% in 2010, making it one of Barrett's stronger rural counties. So far, Walker has 56% of the vote.
10:00PM: Walker leads by 20% with 27% of precincts counted, but the city of Milwaukee is largely outstanding and Dane County, home to Wisconsin, also appears to have disproportionately reported votes from the countryside. As a result, Walker's margin can be expected to shrink as these Democratic areas finish their counts.
9:57PM: Other than the eventual margin, the most interesting outstanding question is the re-tabulated Obama-Romney head to head match-up. The exit polls will be adjusted to reflect the actual results. It will be interesting to see the extent to which Obama's 12% margin narrows after those modifications are made.
9:54PM: While I can sense the disbelief on twitter, remember that the exit polls have a long history of overstating the eventual Democratic performance. The much more reliable pre-election polls consistently showed Walker above 50% of the vote.
9:49PM: NBC News projects that Governor Scott Walker has reclaimed the governorship. Walker appears to be outpacing his 2010 performance across the state, even in the counties where the most votes have been tabulated. Even though relatively few votes have come in from Milwaukee and Madison, there just aren't enough votes in these Democratic bastions to overcome Walker's strong showing in the Milwaukee suburbs and across the Wisconsin countryside.
9:42PM: With 20% of precincts reporting in Waukesha County, home to Milwaukee's most populous suburbs, Walker holds 75% of the vote, up from 71% in 2010. It is difficult to envision that a strong Democratic showing in Milwaukee or Madison would be enough to overcome that type of margin in the Milwaukee suburbs, especially if these numbers across the countryside are representative.
9:40PM: The exit polls have been revised to reflect a 52-48 Walker advantage. Even that tally looks better for Barrett than the returns reported so far.
9:36PM: This fantastic chart from the folks at the Huffington Post illustrates the broad trend we've been observing: Walker is doing quite well in most counties.
9:33PM: 66% of precincts have reported in Dodge County, a Republican county in southeastern Wisconsin, and Walker is still doing somewhat better than he did in 2010. If the exit polls didn't show a close race, I'd say Walker had the advantage. Wait for more votes.
9:28PM: Walker looks like he's out performing Bush - and even his 2010 showing - across rural Wisconsin. Bush barely lost Wisconsin in 2004 and 2000, so if those margins hold, Barrett will probably need to make up ground in Milwaukee and Madison.
Perhaps most significantly, Barrett trails in the early count in several traditionally Democratic counties (Lafayette/Grant) in southwestern Wisconsin. On the other hand, relatively few votes have been counted in these areas, and it's possible that the returns undersample more densely populated Democratic towns, or oversample early/absentee voters who might tilt toward Walker. We'll watch careful for look for evidence that Barrett improves as more votes get counted.
9:24PM: Wondering how Walker has an early lead in heavily Democratic Milwaukee County? Milwaukee County actually includes many Republican suburbs outside of city limits. For an idea of how the Milwaukee suburbs vote, consider Waukesha County, which offered Walker more than 70% of the vote in 2010.
9:18PM: Walker has opened up a lead as rural, traditionally Republican counties in eastern Wisconsin begin to report. Fond du Lac has tabulated 44% of precincts, more than any other county, and Walker has 67% of the vote, up slightly from his 64% tally in 2010. Of course, rural precincts with fewer voters often count their votes faster than the towns where Democrats tend to do somewhat better. My point: wait until counties have near completion before judging.
9:14PM: The first votes have begun to trickle in from heavily Democratic Milwaukee County and Republican Jefferson County. Walker now holds a 57-41 lead with a few thousand votes counted, but one should be very careful about judging and extrapolating early returns. Sporadic returns are unlikely to provide a holistic view of diverse counties. Early and absentee votes, which aren’t necessarily representative of the electorate as a whole, are often reported before Election Day votes are tabulated.
9:11PM: Barrett pulled off an Obama-esque 95% among African American voters, up from 87% in 2010. African Americans represent only a sliver of the Wisconsin electorate (5% in 2008, 4% in 2010), but that can make a difference in a close election. Their impact will be felt most acutely in Milwaukee County, where 26% of residents are African American.
9:08PM: The exits also show Obama leading by 12%, 54-42. Given the low youth turnout, that's very impressive. Obama won 56% in 2008.
9:02PM: Democrats didn't get the youth turnout they wanted, with 18-29 year olds only representing 13% of the electorate, compared to 15% in 2010 and 22% in 2008.
9:00PM: The polls have closed in Wisconsin and Walker and Barnett are neck and neck. The exit polls shows a 50-50 split. It's important to remember that these results aren't final, and don't include a final wave of respondents and the fine-tuning that goes on throughout the evening.
8:50PM: A clear picture of the race won't emerge until we start getting actual results, but we will get a better look at the exit polls results when voting ends in ten minutes. The 9PMEST exits aren't final, since a few waves of late voting respondents haven't been incorporated and the exits are usually adjusted throughout the evening. Even so, we'll get initial horse race numbers and the best guess yet at the composition of the electorate.
8:00PM: Want real-time analysis of the Wisconsin election results? Sure you do! Return to The Plank at 9 pm, when Nate Cohn (aka: Electionate) will start live-blogging the recall.