Here are some observations from the election results in the Wisconsin recall race. For the results I’ve used the convenient interactive maps provided by the Milwaukee Journal Sentinel and the Huffington Post. When you look at the map, you notice that Democrat Tom Barrett won only in a relatively few counties, 12 of 72. He was reduced to the Democratic base: Madison and a few surrounding counties, the central city of Milwaukee but definitely not the suburbs, a few old factory towns like LaCrosse and Stevens Point, three counties up on Lake Superior near Duluth, Minnesota, and the Menominee Indian Reservation. Everything else went for Scott Walker.
Turnout in this election was really high: 2,507,269, compared to 2,160,832 in November 2010 and closer to the 2,984,417 in November 2008. Much has been made of the exit poll finding that union members were one-third of the total, up from 2010, but we see evidence of this in county returns as well. Turnout was up 16% statewide, but it was up 20% in Kenosha County and in Douglas County (Superior) which have had lots of blue collar voters. Walker’s percentage as compared to 2010 declined 6.2% in Douglas County, more than in any other county in the state, and Kenosha County was one of the few counties Walker carried in 2010 but lost in 2012. Turnout was up 16% in Milwaukee County, suggesting that black turnout was fairly robust, and up 15% in Dane County (Madison), the epicenter of anti-Walker forces and up 22% and 25% in adjacent Columbia and Dodge Counties, which were two of the 16 counties where Walker’s percentage fell from 2010 to 2012. One has visions of Madison Occupy-types heading out to canvass in rural areas nearby. Conclusion: the union and leftish Democrats did a good job of turning out their voters. It was like the 2004 presidential race in Ohio, where the Kerry forces did a great job turning out voters in central cities, but were still beaten because there was also heavy turnout in small and medium-sized counties of strongly motivated Republican voters. Such was the case in Wisconsin.
Walker’s percentage rose most in small counties in the northwestern quadrant of the state and near Green Bay. These were mostly Barack Obama territory in 2008; the rise in Walker percentage suggests that confrontational tactics hurt Democrats there (as I speculated in a blogpost last night) and they might hurt Obama there as well.
The strongly committed areas remain very strongly committed. Turnout in Dane County, as noted, was up 15%, but Walker’s percentage declined by just 0.5%. In Waukesha County, just west of Milwaukee, Washington County, to the northwest, and Ozaukee County, directly north, turnout was up 13%, 17% and 12%. Voters in these high income areas usually vote in large numbers, so perhaps there wasn’t much room for improvement. Walker’s percentages increased only slightly in these counties, 0.9%, 0.6% and 1.8% respectively, but they were very high. In Milwaukee County, Walker’s percentage was up 3.6%, suggesting pretty robust turnout in black areas but even greater increases in white suburbs. The result in all cases was greater popular vote majorities than in the 2010 election, which I’ve shown in the table below.
County 2012 margin 2010 margin
Dane 98,485 Barrett 81,461 Barrett
Waukesha 96,036 Walker 81,924 Walker
Washington 35,656 Walker 29,946 Walker
Ozaukee 20,206 Walker 16,646 Walker
Milwaukee 106,660 Barrett 81,320 Barrett
Rest of state 224,912 Walker 167,956 Walker
WISCONSIN 171,665 Walker 124,638 Walker
In other words, these big counties voted pretty much as before, but in larger numbers. The net vote-shifting to Walker was greater in the rest of the state.