Hoosiers
by techfidel, Mon Mar 24, 2008 at 07:03:02 AM EDT
Last week I posted a diary analyzing the Ohio Democratic primary with an eye to predicting the Pennsylvania results on April 22. The bottom line of that analysis was a 57-43 Clinton victory in the Quaker State.
Indiana Democrats are slated to weigh in two weeks later on May 6, and insofar as Indiana might also share demographic characteristics with Ohio, a similar analysis in presented here for the IN contest.
First, recapping the analysis of the actual results in Ohio's March 4th primary, a cluster analysis (k-means) of Ohio's 88 counties on a set of five non-political factors (population density, median household income, percent blacks, percent high school grads, percent college grads) yielded the following map:

More...
The six clusters are defined roughly as follows:
Hillary-billiesLow income, low education, low population density, 99% white; supports Hillary at 80%. These counties have often leaned Republican in presidential contests, although it should be noted that they largely went for Bill Clinton in 1992, so it's kind of a traditional "Clinton Country."
Ed-necks
Like the Hillary-billies, just a little less so (slightly more income, education, etc), 97% white; supports Hillary at 70%. "Ed-necks" is a pun on PA Gov. Ed Rendell, who noted that some whites are not ready to vote for a black candidate for president. The support for Obama in the nearly 100% white Hillary-billary and Ed-neck counties of only 20-30% would support the idea that there is a good bit of racial voting there.
Proletariat
Medium income, medium education, exurban, 94% white; supports Hillary at 61%.
Hoosierdom
Medium income, medium education, rural, 98% white; supports Hillary at 61%.
Crunchy-cons
High income, very high education, medium population density, 97% white; supports Hillary at 55%.
Obama-crats
Medium income, high population density, high education, urban, 21% black; supports Obama at 54%.
Overall, Hillary prevailed 55-45 in Ohio, and Obama only came out on top in five counties:

To project the results in Indiana from these Ohio results, it should be noted that the following demographic analysis presumes all other "news" factors are equal. On March 4, for example, the Rev. Jeremiah Wright flap had not become widespread. At that time, the biggest issue in the campaign was perhaps the story about the Obama campaign's duplicity on NAFTA, widely considered to have hurt him in Ohio. The PA results on April 22 should be an indication of whether these IN projections will play out similarly two weeks later.
Here's a map of Indiana's 92 counties, subjected to the same cluster analysis described above for Ohio:

IN has a population of 6.3 million compared to 11.5 million for OH. IN has a smaller fraction of blacks in the population, 8.4% compared to 11.4% for OH. Hispanics are 4.8% in IN, over double that in OH, 2.3%.
Here's how the six clusters in the two states compare in terms of fraction of the Democratic primary electorate (projected in the case of IN):

The Obama-crat cluster is reduced in IN (more spread out population and fewer blacks. Ed-necks, Proletariat and Crunchy-cons are somewhat greater in proportion. These factors would indicate that Hillary will do even better in IN than OH's 55-45 result.
The Democratic primary turnout estimates for each group in IN are based on the turnouts for the analogous groups in OH:

In a similar way, the total voting-age population in each cluster in IN could be multiplied by the Democratic turnout percentages above to get the projected turnout in each cluster (i.e., represented by the pie chart above), and then the Democratic turnout could be multiplied by the expected Hillary percent above in each cluster to get the projected IN final result. However, as there are some minor demographic differences between corresponding clusters in the two states, it's best to do a little tweaking first of the expected Hillary percentages.
Here is the demographic comparison of the six clusters in the two states:

The expected Hillary percentage in each cluster for IN has been adjusted slightly based on the demographic differences. For example, Obamacrats in OH have 21% blacks compared to 26% for IN. That would mean less votes for Hillary in IN. Offsetting that though somewhat is that the Hispanic percentage (not shown) is higher in this group in IN. Lake County (Gary), for example, has 13.9% Hispanics and 26% blacks.
Given the projected Hillary percentages in the table, the projected final result comes out to a 58-42 victory for her in IN.
As for the projected turnout and vote totals, it comes out to a turnout of 1,120,450 Democratic voters, Hillary 643,396 to Obama 477,055--a margin of 166,341 votes. These totals may be somewhat of an overestimate, as IN tends more Republican than does OH. On the other hand, it is an open primary, and there could be a good bit of crossover voting from Republicans and independents.
Tags: 2008, Barack Obama, Election, Hillary Clinton, Indiana, Ohio, Pennsylvania, Primary (all tags)









39 Comments