Saturday, July 14, 2012

Who Votes Democrat/Republican/Third Party Today?

Two weeks ago I posted the results of a demographic analysis of Ohio voting for a recent election where there was a state-wide Socialist candidate. Here is a follow-up post about demographics of voting patterns for the major parties, as well as compared to "Other" (non-Rep, non-Dem). Both analyses utilize Census data from the 2010 American Community Survey (5-year estimates) and 2000 Census.

This analysis explores the last 3 presidential elections (2000, 2004, 2008) to look for demographic patterns. I do not take into account national factors such as incumbency, unemployment, etc. The Census has county-level demographic data for all US counties, and county-level voting data for all 3 of these elections is publicly available. I averaged the "percent" that voted for each candidate for these 3 elections and constructed a simple correlation using data mining of about 250 socio-economic variables. Below are the strongest correlations for who voted Democrat (almost identical, but opposite those who voted Republic, so I did not include them as a column), and those who voted for a third-party candidate--few of the correlations were above 0.40, and I include only those correlations greater than 0.25. These correlations are for counties high in these populations, and do not necessarily mean these groups themselves are more or less likely to vote in these patterns.

Briefly, those counties more likely to vote Democrat have higher percentages of the following: worked in professional jobs (science, education, management) and the arts; the unemployed; renters; female-headed, single-parent homes; those never married; those who have needed to use public services (food stamps, or public transportation to get to work); those with some member of the household in college or graduate school; and some race minorities (especially counties with high numbers of Blacks and Chinese).

Those counties more likely to vote Republican have higher numbers of the following: the self-employed; those in who work in "natural resources" (mining, agriculture) and construction; the married; those who own their own homes; whites; those with some college but failed to graduate; those with children in elementary school.

Interestingly, those counties whose residents have needed to rely on public services are unlikely to vote for a third party, even more-so than their reluctance to vote Republican, as are those separated from their spouse, female-headed-single-parent households, and those with some high school but who failed to graduate. Those more likely to vote third-party: those who walked to work or worked at home; those who are consistently employed; the never married; whites; those with a high school diploma or greater.

DemocratThird-Party/ IndependentDemographic
-0.35Self-employed in own not incorporated business workers
0.33Public transportation to work
0.43Walks to work
-0.38Car, truck, or van -- drove alone to work
0.36Worked at home (no commute)
0.29-0.31Requires food stamps/SNAP
0.28Professional, scientific, and management, and administrative and waste management services
0.24Arts, entertainment, and recreation, and accommodation and food services
0.25Educational services, and health care and social assistance
-0.43Agriculture, forestry, fishing and hunting, and mining
-0.43Some high school, no degree
-0.30Some college, no degree
0.41HS diploma or greater
0.48-0.36Female, Single-parent
-0.34Separated from spouse
0.520.33Never married
-0.57Currently married (not separated)
-0.30Have kids in elementary school
0.31Someone in household in college/graduate school
-0.28Owner-occupied housing

No comments:

Post a Comment