What are authoritarians like?

Mixing Memory has an excellent post today reviewing some recent empirical work on patriotism. The post discusses a paper by Schatz, Staub, and Lavine, in which the authors distinguish between ‘blind’ and ‘constructive’ patriotism, with the argument being (as you can imagine from the names) that one is constructive and the other is… less constructive. While I’m not very interested in patriotism per se, I think the literature in question is heavily related to more general work on authoritarianism and political orientations which I do find pretty fascinating, so I thought I’d piggy-back on the Mixing Memory post and discuss some related findings in behavioral genetics and personality psychology.

One of the many interesting findings to come out of the behavioral genetics literature is that the heredity of political orientation (defined in terms of variables such as conservatism vs. liberalism, right-wing authoritarianism, etc.) is about as high as that of general intelligence and most major personality dimensions–typically around 50-65%. That’s to say, over half of the variance in questionnaires including items such as “Our country needs a powerful leader to overthrow the radical and immoral values that are present in today’s society” is attributable to genetic influences (most of the remainder is due to unique, or non-shared, environmental influences). This finding is, on the face of it, pretty perplexing: are supporters of G.W. Bush really born and not made? Should Democrats blame their loss in the 2004 elections on the genetic composition of middle America?

The short answer, of course, is no. There is no ‘Digging For Oil In The Arctic Is Okay’ gene, and one shouldn’t expect to find a locus on chromosome 13 coding for degree of support for the war in Iraq. But that doesn’t mean there isn’t any heritable component to such views. Rather, the general thinking among behavioral geneticists and personality psychologists is that political attitudes are highly heritable because they derive from more general personality factors that are plausibly coded for in our genome. Reflecting this assumption, the intersection between molecular genetics and personality is a rapidly growing area of research, with novel findings hitting the headlines almost every week (unfortunately, replications are harder to come by). A good example of this is the recent finding that a polymorphism in the DRD4 gene (which codes for the expression of the D4 dopamine receptor) is associated with variability in the human sexual drive, potentially reflecting a more general contribution of dopamine function to broad personality dimensions such as extraversion or novelty-seeking.

The idea that personality and political orientation are related shouldn’t strike anyone as surprising. If you’ve ever got into a political argument with relatives over dinner, you probably already suspect that people’s views on homosexuality, religion, science, and G.W. Bush aren’t completely independent. Empirically, that turns out to be true: a wealth of evidence accumulated over the past half-century documents the existence of highly reliable associations between political attitudes and personality traits. Put simply, what you believe isn’t independent of who you are. As Chris notes in his Mixing Memory post, people have a ubiquitous tendency to hold opinions that are consistent with their prior beliefs. This is the famous confirmation bias, and it applies not only to belief but also to personality: people of a certain type endorse views of a certain type, irrespective of any elaborately concocted rationalizations (after all, nobody ever admits they believe what they believe just because they happen to be extraverted or unconscientious!)

So, to the interesting question (today, at least): what personality traits do political attitudes predict? Naturally, the answer depends on which variables you look at. Since there’s little point in honing in on very specific attitudes (e.g., asking people exclusively about their views on the logging industry isn’t going to give you a very good measure of liberalism), I’m going to focus on a cluster of traits that have variously been referred to as authoritarianism, conservatism, and traditionalism. These dimensions are often operationalized somewhat differently (indeed, there are several dozens published questionnaire measures purporting to measure any one of these), but generally tap most or all of the following characteristics: willingness to follow authority and to impose it on others; preference for hierarchical rather than egalitarian social and political systems; negative views toward out-groups, including apprehensiveness toward non-normative behaviors (e.g., homosexuality). In general, measures of these 3 dimensions correlate fairly highly almost regardless of which measure you use. Put differently, there’s a single higher-order factor that most measures of authoritarianism, conservatism, and traditionalism load heavily on (though each individual measure also has a large portion of variance unique to it). For example, in a recent study, Bouchard and colleagues found that a measure of conservatism had correlations of 0.58 and 0.72 with measures of traditionalism and authoritarianism respectively (canonically, anything above around 0.5 is considered a large effect).

What about more general personality factors? As I mentioned earlier, personality taxonomists typically think of relatively specific attitudes as emanating from somewhat more general higher-level factors, which in turn depend on a still smaller set of dispositional variables. The most popular dimensional model of personality is the 5-factor (or Big 5) model, which includes dimensions of Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism (convenient acronym: OCEAN). While I’m not a huge fan of the 5-factor model (for reasons I won’t get into here), I think it’s achieved a critical mass in personality psychology and even pop culture, so it makes sense to try to relate its dimensions to somewhat more specific traits such as authoritarianism.

Across studies, 4 of the 5 dimensions appear to be reliably correlated with authoritarian/conservative tendencies: openness (negative correlation), conscientiousness (positive correlation), extraversion (positive correlation), and agreeableness (positive correlation). That’s to say, highly authoritarian individuals tend to be somewhat more conscientious and extraverted than non-authoritarian individuals, and somewhat less open-minded and agreeable. Note that this pattern of correlations makes it apparent that authoritarianism isn’t always a bad trait. Extraversion and conscientiousness are generally considered positive traits, and the fact that authoritarian individuals are likely to be more assertive and conscientious attests to the decisiveness and confidence with which authoritarians are capable of acting. On the other hand, the correlations with the other two traits reflect the down-side of that decisiveness and moral certitude: a failure to consider alternatives and contemplate speculative ideas (the dimension of openness is sometimes described as intellectual curiosity), and a relative disregard for people’s feelings. So the Big 5 model provides a potentially useful framework for understanding lower-order personality traits that give rise to a broad range of political and social attitudes. That said, the correlations between the Big 5 dimensions and authoritarianism scales are typically small to moderate (with the largest effect holding for Openness, typically on the order of around .25 – .3.) So it’s certainly possible to come across highly agreeable, intellectually curious authoritarians–it just doesn’t happen that often.

We can also ask about other traits and abilities. One of the most interesting and potentially most relevant domains to the traits of authoritarianism and conservatism is moral reasoning. Since authoritarians frame so many of their views in terms of morality, you might expect them to have some advantage in this arena. But the opposite turns out to be true. There’s a negative correlation between authoritarianism scores and moral competence, such that when presented with moral dilemmas, authoritarians’ responses tend to depend more on factors such as acquiescence with cultural norms than on an internally-coherent system of beliefs. Authoritarianism also correlates positively with almost all forms of prejudice–presumably not the best marker of a moral nature. Finally, there’s a reliable and moderately-sized inverse correlation between authoritarianism and general intelligence (this correlation weakens somewhat when you look at measures of conservatism, though it’s still present). Thus, ironically, authoritarians tend to be less open-minded and less intelligent than non-authoritarians despite being more assertive and decisive, leading to the rather undesirable state of affairs that the people who are most confident in their actions are often the people who are least qualified to do so. Insert your own snide remarks about the current state of US politics here: _________________________________________