Wednesday, June 18, 2008

Does Obama Have a Polish Problem???

With the primaries over and the nomination decided, we have some time to go back and look at some data that got lost in the bustle of the unfolding contests. I’ve done a lot of reading over the last few years on Chicago politics and history, including some recent study of the Great Migration. I thought I’d go back to the Illinois primary and see if anything jumped out about the voting in the Windy City. Given that this is Obama’s backyard, any strong Clinton showing could highlight potential problems for the Illinois senator.

Indeed, when we start looking at smaller units of analysis, we see that the Chicagoland voting was not uniformly for Obama. Statewide, Obama received 64.7% to Clinton’s 32.8%. In Cook County, he received 68.9% to Clinton’s 29.2% and in Chicago proper Obama got 72.8% to Clinton’s 25.6%. Within both Cook County and the city, we see some areas of Clinton strength.

Cook County, for tax assessment purposes mainly, organizes itself into a number of townships. The County also uses these townships to report election returns (get the full primary election returns, by township, here). Of the 30 townships outside the city of Chicago, Clinton actually won 9 (Berwyn, Cicero, Lemont, Leyden, Maine, Norwood Park, Orland, Palos, and Stickney). Mapping these (Clinton in red; Obama in blue), one sees a clustering effect. Wondering if there was any correlation between the demographics of these inner suburbs and the vote (as we saw in virtually every primary nationwide), I got some census data and found that Clinton's wins tended to be in townships with either 1) a very large white electorate or 2) a very large Hispanic electorate. The correlation is not perfect but there is a general trend. Of the four townships that are over 90% white, Clinton won three (Lemont, Norwood Park, and Orland). The Obama win in this category, New Trier, comes in a township of very high income and education levels and conforms with trends we found earlier in the primary season. New Trier Township includes such villages as Wilmette, Winnetka, Kenilworth, and Glencoe. To use a pop culture reference, this was the area where Ferris Bueller’s Day Off and a number of other John Hughes movies were filmed and set.

For the Hispanic dimension, Cicero stands out. With a Hispanic population of 77%, it gave 65%--her largest total in any township. Also Berwyn (38% Hispanic) and Leyden (23% Hispanic) went Clinton although Hanover (23% Hispanic) and Calumet (23% Hispanic) went for Obama. In the case of Calumet, though, it is also the township with the largest African American percentage, explaining the tilt to Obama. I’ve produced a rudimentary graphic to show the relationship between each township’s white population and Obama vote (as I get more skilled in chart making the visuals should improve here).

When we look at the vote in Chicago proper (see results here), we see a similar ethnic dimension to the voting. Chicago is divided into 50 wards. Of these, Clinton won 14. Like with the township vote, there is a definite clustering of her support. I’ve highlighted the Clinton wins in the below map.

To get a sense of the correlation between ethnicity and the vote, compare the map on the right with this amazing map produced by the Newberry Library and the Chicago Historical Society, entitled “Chicago’s Ethnic Mosaic in 2000.” Two things jump out. As with the township voting, the heavily Hispanic areas of the city went for Clinton. Of perhaps more interest, one sees a rough correlation between the Polish vote and Clinton. Most notably, the areas around Midway Airport in the southwestern part of the city and O’Hare Airport in the northwestern part are heavily Polish (shaded purple). They also were areas of strong support for Clinton. I'd also note that these were areas where Harold Washington fared very poorly in his historic 1983 mayoral victory. In the southwestern part, one sees how the Polish population extends outside of the city limits into the neighboring townships that also went for Clinton.

Come November, all of this may not mean much. Obama is expected to win Illinois easily. Nonetheless, even in his home state, his support was not universal and may signal broader, more national, areas of weakness. While recent polling on the Hispanic vote seems to show Obama making tremendous gains, his white ethnic support isn't yet where he'd like it to be.

No comments: