Are lefties really smarter than conservatives?

Hum de hum. You may know that a number of studies have demonstrated that higher intelligence is correlated with left-wing political beliefs. And that might be so. But I've always been a bit skeptical because (a) it seems unlikely, and (b) maybe what's really going on is that smart people tend to go to college and it's actually college education that's correlated with lefty beliefs.

A new study from a grad student at the University of Minnesota tries to tease this apart. Here's the basic result:

This chart shows how much political beliefs shift leftward in response to an IQ increase of one standard deviation (15 points). The raw result is 0.54, which is—a lot? A little? It's hard to say because the paper doesn't explain the scale they use to measure political orientation.¹ In any case, if you control for income and education about a third of the effect goes away: the new result is 0.38 and is no longer statistically significant.

Now let's take a look at a different table:

This is the correlation between political beliefs and educational attainment (EA). The raw result is 0.176, which is about a third of the raw effect of IQ. That fits with the first result.

I have no idea if the authors' methods are reliable. For technical reasons, their measures of intelligence and educational attainment are based on clusters of genetic markers from other studies: One cluster is associated with intelligence and the other is associated with education. So a big part of the correlations depends on whether these genetic clusters are reliable in the first place.

There's also the fact that once you control for education you (barely) lose statistical significance for the effect of IQ. This may not be a big deal, but it's not especially promising either.

In any case, if the genetic clusters are reliable and the methodology is sound, it suggests that intelligence really does have an impact on political beliefs regardless of educational level. Which I still find odd. I know that we lefties want to say "Duh, of course smart people are more liberal," but that's really not persuasive. It's very hard to conceive of why a high IQ should make you more likely to support abortion or social welfare or a smaller defense budget. But maybe that's just a failure of imagination on my part.

¹Their scale is a "summed composite" of five other scales, which I'd guess are each scales of 0-1. That makes the composite a scale of 0-5, so an increase of 0.54 is about 10%. However, this is just a wild guess on my part. I have no clue why the authors present a bunch of numbers without explaining what they mean.