Berry Boessenkool asks:
What you think about this article that basically implies that political views are not so much actively chosen or the product of upbringing, but based more on inherited traits.
I [Boessenkool] at first even thought it was satirical. . . .
The article, which appears on the Web site of Mother Jones magazine, is by Chris Mooney and is called, “Scientists Are Beginning to Figure Out Why Conservatives Are . . . Conservative.”
I think the author of the article means well but this does seem a bit of a case of gee-whiz science reporting. Evolution and genetics often seem to be an excuse for scientists and reporters to shut off their B.S. filters!
In particular, there’s this bit:
A large body of political scientists and political psychologists now concur that liberals and conservatives disagree about politics in part because they are different people at the level of personality, psychology, and even traits like physiology and genetics.
OK, so far, no problem. This works as long as we take the terms “physiology and genetics” generally enough. For example, African Americans and European Americans differ on average on their attitudes toward income redistribution and they also have some physiological and genetic differences. So the statement is literally true and not remarkable.
But then Mooney continues:
That’s a big deal. It challenges everything that we thought we knew about politics—upending the idea that we get our beliefs solely from our upbringing, from our friends and families, from our personal economic interests, and calling into question the notion that in politics, we can really change (most of us, anyway).
And that is just ridiculous. “We” never thought that “we get our beliefs solely from our upbringing, from our friends and families, from our personal economic interests.” That’s a statement that’s wrong on so many levels. First, of course there’s variation, it’s easy to find pairs of people with similar upbringing and economic status with different political attitudes. Second, talking about “personal economic interests” is missing so much because it ignores the issue of economic ideology. Third, political science research has found that voters are influenced by national political conditions when they are young (not just by their upbringing). Finally, the last sentence quoted above fits into the horrible attitude of science reporting in which any correlation of a behavior with a gene is interpreted as implying some sort of deterministic relationship. Why must it be that, just because certain political attitudes or behaviors are correlated with some genes, that this calls into question the notion that we can really change?
Finally, there’s the title of the news article (“Why Conservatives Are . . . Conservative”), which I guess I can blame on Mooney’s editors at Mother Jones. This title gets the message of the original article backward! Hibbing et al. argue that conservative attitudes are most natural (“a strong negativity bias was extremely useful in the Pleistocene”) and that liberalism could be a more recent development.
Anyway, just to be clear, I’m not saying there’s anything wrong with the review article by John Hibbing, Kevin Smith, or John Alford, or with much of what Mooney is writing in his news article. This is indeed interesting (if speculative) work and I’m glad to see people talking about it.
I just think there should be a way to report on new research without trying to oversell it by contrasting it to a drastically simplified presentation of our prior understanding. Researchers have been studying the correlations between political attitudes and personality types for over 50 years. It’s good to see continuing work in this area, and it’s also important to recognize that some of this new work is useless (for example, this study that Dan Kahan pointed me to last year). All this helps to improve our understanding, even more so if we don’t fall into the trap of thinking that genetic correlation = determinism.
P.S. To illustrate this post I wanted to include a graph from the Hibbing et al. paper, but it had no graphs! That’s not good. So I found the nearest graph at hand, which came from the discussion by Christopher Federico, Christopher Johnston, and Howard Lavine. Actually, all these discussions together only contained a single graph of data, but I don’t want to reproduce it here because the authors (Jacob Vigil and Chance Strenth) use it to make a this-is-statistically-significant-and-that-is-not sort of comparison that I generally don’t trust. The graph is on page 332 of the journal and you can look at it yourself if you’d like.