Left-right ideology is the single most powerful determinant of party identification and issue positions, but ideology itself is almost impossible to predict.  It does have a few mild correlates – for example, the well-educated are a little more liberal, the high-income are slightly more conservative.  Even if you throw in a kitchen sink full of variables that we assume are politically salient, though, positions on the left-ring spectrum are close to random.  That’s why I’ve often been tempted to call ideology the Unmoved Mover of politics.

Another possibility, of course, is that researchers have been looking in the wrong place.  Instead of trying to use traditional political variables to predict ideology, maybe we should be looking at psychological variables.  As a fan of personality research, this seems like a plausible hypothesis.  When I first looked into this possibility about eight years ago, though, I didn’t find much, and gave up.

My bad.  A compelling recent paper by Alan Gerber and co-authors shows that personality and ideology are closely linked.  Liberals are markedly higher in Openness and lower in Conscientiousness.  Gerber et al’s real contribution, though, comes when they distinguish social ideology from economic ideology.  (It’s almost as if they were inspired by the World’s Shortest Political Quiz!)

For social liberalism, they once again find that it is associated with higher Openness and lower Conscientiousness.  For economic liberalism (in the American sense of lower support for free-market policies), though, they get a much more thorough profile.  Economic liberals are less Extroverted, more Agreeable, less Conscientious, less Stable (i.e. more Neurotic), and more Open.  Or if you flip the perspective, free-marketeers are more Extroverted, less Agreeable, more Conscientious, more Stable, and less Open. 

The authors insist that the distinction between social and economic ideology is very robust:

While the relationships between Conscientiousness and Openness and attitudes are similar across domains, the other three Big Five traits are only related to economic attitudes.  This is not a matter of coefficients falling just short of conventional levels of statistical significance in one domain and barely exceeding them in the other. In the social domain, the magnitude of the t-tests on these three variables range from .02 to 1.67; in the economic domain the magnitudes range from 7.13 to 15.19.  Thus, while previous research has rightly identified Conscientiousness and Openness as the traits most consistently related to ideology, our analysis shows that the other Big Five traits–particularly Emotional Stability and Agreeableness–significantly and substantially affect political attitudes.

They also provide a neat diagram to summarize their findings – check it out.

My main empirical reservations:

1. Instead of measuring social and economic ideology with self-placement, they measure it by averaging a few issue positions.  Social ideology is an average of positions on gay marriage, federal funding of stem cell research, and abortion.  Economic ideology is an average of positions on health care, social security privatization, the minimum wage, and preferences for taxation vs. spending.  This seems pretty reasonable, but it does make me wonder if they’re cherry picking issues where personality has strong predictive power.

2. There’s a lot of evidence that more educated people are more economically conservative.  More educated people do much better on tests of political knowledge, and as Scott Althaus shows in his exhaustive literature review, high political knowledge predicts greater economic conservatism.  But Gerber et al find no connection between education and economic conservatism.

Admittedly, in my work on education, IQ and economic beliefs, I found that the connection between education and pro-market views was a little spotty: While education never predicts anti-market views, it only predict pro-market views two-thirds of the time.  In Gerber et al’s Table A1, however, education doesn’t seem to matter much for any of the economic questions.  This makes me wonder how representative their questions are.

3. Openness is the one Big 5 personality trait that is clearly correlated with IQ: smart people are more Open.  In my research, I’ve found that higher IQ predicts more pro-market views.  It’s puzzling, then, that more Open people would be less pro-market, especially since Gerber et al don’t control for IQ.

Despite these doubts, I’m really impressed by this piece.  Precisely because voters can vote against their objective self-interest at near-zero cost, it makes sense for psychological variables like personality to strongly affect political positions.   If these results hold up, they should sharply change the way that social scientists think about politics.  Unfortunately, I suspect that no matter what the data show, political economy is not yet ready to give personality its due.  How this will all play out?