Tuesday, April 29, 2008

A First Look At Indiana--Does Pennsylvania Teach Us Anything??

With a week to go before the double bill in North Carolina and Indiana, I thought I’d take a first look at the Hoosier state and size up Obama’s chances. Most of the recent polls I’ve seen up to this point show him trailing Senator Clinton, with varying margins. Assuming that the results in Pennsylvania will have some bearing on the Indiana vote (press coverage, Clinton momentum, etc.), I thought I’d use last week’s vote as a way to speculate on Indiana.

In the Pennsylvania primary, Obama won 7 counties (Centre, Chester, Dauphin, Delaware, Lancaster, Philadelphia, and Union) on the way to capturing 46% of the statewide total. What we seem to be seeing as these contests progress is a dynamic where Obama’s vote tends to be concentrated in urban areas and inner ring suburbs, plus more outlying counties that have an academic/university/research presence that brings with it young voters and highly educated/high income professionals. Basically what we’re finding is that Obama does better in areas with more African American voters, higher median incomes, higher education levels, and fewer seniors. Using these basic points of comparison, we see that Obama is perhaps at a disadvantage in Indiana, relative to Pennsylvania (data available here):

Indiana Vs. Pennsylvania:
African American %: 8.7 vs. 10.4
Median Income: $45,394 vs. $46,259
% w/College Degree: 21.7 vs. 25.4
% 65 Years & Over: 12.4 vs. 15.1

Only on the age variable, does he fare more favorably than he did in Pennsylvania.

Running the numbers on the counties Obama won in Pennsylvania, I found that these counties provided 32% of the total votes cast statewide. Philadelphia, with 19% of the statewide vote, made up the bulk of this. Looking at Indiana, I’ve tried to identify counties that might be favorable to Obama. I would identify first Marion County (Indianapolis). Marion County is 24% African American which should bode well for Obama. Next is Lake County (Gary, Hammond). Lake County, part of the Chicago metropolitan area, is 25% African American. Third is St. Joseph County (home to Notre Dame). This county is 12% African American plus has a sizable student/university population. Fourth is Monroe County, centered around Bloomington and Indiana University. Monroe is only 3% African American but, as noted, has a large student/university population. Finally, La Porte County is 10% African American. In the 2004 election, all of these counties gave Kerry a victory with the exception of St. Joseph which narrowly went to Bush.

Can we use this county level data, plus the Pennsylvania results, to make some projections? Because we don’t have a recent Indiana primary to use as a baseline for predicting turnout, I decided to look at the 2004 general. Using 2004, we see that the counties I’ve identified as most favorable to Obama provided 29% of the statewide turnout. Comparing this to Pennsylvania, we see Obama perhaps in a worse position. The more rural counties provide a greater share of the vote in Indiana than they did in Pennsylvania. So, making the assumption that Obama wins those counties I identified, he might end up further behind than he did in Pennsylvania (this also assumes Pennsylvania like margins in the remaining counties). This also assumes he wins with comparable percentages. It is perhaps a stretch to think that he’ll match or exceed the 65% of the Philly vote he got in Indianapolis. Whereas Philadelphia is roughly 45% African American, Indianapolis is only one-quarter black.

So, based strictly on this cursory demographic examination, things don’t look so good in Indiana for Obama. How might he fare better? One variable working in his favor, as several commentators have noted, is the “Midwest” factor. Whereas Obama has had a problem in more “Appalachian” regions (see here), Indiana’s geography and ethnic patterns are a bit different. Furthermore, bordering his home state of Illinois, there will be a greater level of familiarity with him, especially in the northwestern corner of the state (Lake, La Porte, St. Joseph Counties). A second variable is the fact that Indiana, unlike Pennsylvania, is an open primary. Obama has consistently outperformed Clinton among independent and Republican leaning voters who have participated in Democratic contests this year. So, to the degree that Obama is able to close the gap with Clinton in Indiana, it would seem that it's these voters that he’ll have to rely on.

No comments: