Last week, we looked at the relationship between class and happy states. This week we look at the effects of class on the 2008 Obama-McCain presidential election. Since states are key units in the U.S. electoral college system, we took states as the units of analysis as opposed to individuals. So under the watchful analytical eye of Charlotta Mellander, we looked at how the class composition of states (as opposed to say the class membership of individuals) effected votes for Obama versus McCain. To make this a little bit more interesting and more fun, we also looked at the effects of factors like income, housing prices, and human capital, as well as the gay index and personality on state voting patterns. Basically, we wanted to identify the kinds of states that voted for Obama or McCain.
CLASS AND THE ELECTION: First things first. There was undeniable class pattern to state voting in the 2008 election. States with large concentrations of two classes – the creative class and the service class were strongly associated with Obama, while states with large working class concentrations went for McCain.
The Creative Class: The correlation between creative class states and Obama was positive and significant (.425), while it was negative and significant for McCain (-.442).
The Service Class: The same basic pattern was true of the service class, the largest class and also the class with the lowest average level of wages and salaries. Service class states were positively associated with Obama (.390) and negative for McCain (-.415).
The Working Class: Now check out the pattern with the working class. Hasn’t the conventional wisdom long been that working class voters tilt heavily toward the Democrats? When it comes to the concentration of working class jobs in a state, not so much. Working class states were even more strongly associated with McCain (.658) than creative class states for Obama; and working class states were quite negatively correlated (-.623) with Obama. This state level pattern contrasts with individual voting. While Obama won lots of working class voters, working class states went strongly for McCain.
Income and Economic Output: We also looked at the association between state voting and income and economic output (measured as GDP per person). Obama states were those with higher levels of GDP per capita (.375 Obama vs. -.388 McCain) and higher incomes (.516 Obama vs. -527 McCain). These are in line with earlier findings of Columbia University political scientist and Rich State, Poor State author, Andrew Gelman.
Housing Prices: States with high housing prices also were in the Obama camp. Housing prices were strong correlated with Obama states (.672) and negatively associated with McCain states (-.725).
Human Capital: Human capital – that is the percentage of adults with a college degree – was strongly positively associated with Obama states (.458) and negatively associated with McCain states (-.492). These patterns contrast somewhat with more nuanced data for individual voters. Gelman pointed out in an e-mail that: “At the individual level, Obama did best among people without h.s. degrees and people with postgraduate degrees. McCain did best among people with some college and people with college degrees (but no postgrad degrees).”
The Gay Index: Obama states were also those with greater concentrations of gays and lesbians. The Gay Index was positively associated with Obama states (.532) and negatively associated with McCain states (-.544).
Personality Factors: Psychologists have long been interested in the connection between personality and individual voting and ideology. So, once again using data originally collected by Cambridge University psychologist Jason Rentfrow and his collaborators, we compared state voting patterns to the concentration of the five major personality types – extroversion, agreeableness, conscientiousness, openness-to-experience, and neuroticism. While three of the types had little relation to state voting patterns, two were significantly associated with Obama vs. McCain votes.
Open-to-Experience: Obama states were associated with high concentrations of open-to-experience (.409) – that is highly creative and innovative people; openness was negatively associated with McCain states (-.371).
Conscientiousness: McCain states were associated with high concentrations of conscientious (or dutiful) personalities (.311).
In an e-mail, Rentfrow says these findings are in line with his own: “The links with class, openness and voting is consistent with what we found in the 1996, 2000, and 2004 presidential elections. Although we didn’t look at the class groupings you looked at, each is related to openness, and we found that openness was strongly related to voting patterns after controlling for income, education, race, and sex. I ran the same analyses for the 2008 election and the results are very similar…”