Skip to content
 

Personality profiles and voting preferences: should liberals or conservatives like research in this area?

In his American Psychologist article reviewing studies of personality profiles and political affiliation, and the followup article here, John Jost writes,

Compared with liberals and moderates, conservatives score significantly higher on psychological instruments designed to measure epistemic needs for order, structure, simplicity, certainty, and closure, and they score significantly higher on instruments designed to measure the intensity of existential concerns such as fear of death and perceptions of a dangerous world. In terms of basic personality dimensions, liberals (and leftists) score significantly higher on Openness to New Experiences, and their greater open-mindedness manifests itself in terms of creativity, curiosity, novelty, diversity, and interest in travel. By contrast, conservatives (and
rightists) score higher on Conscientiousness, and they are generally more orderly, organized, duty-bound, conventional, and more likely to follow rules. The evidence strongly contradicts
the commonly held assumption that political orientation is “consistently and strikingly unrelated to personality and temperament factors.”

Like much social science, the above statement seems either obviously true or a ridiculous distortion, depending on how you look at it. But that’s the purpose of doing research, to try to evaluate such hypotheses. There are usually many different interpretations, but it’s good to establish the facts.

Taboo research?

In any case, what interests me here, beyond the importance of the topic itself, are the political reactions to such work. Jost’s article is called “The End of the End of Ideology” and it describes how studies of personality characteristics, political orientation, and authoritarianism were popular in the 1950s but had fallen into disfavor for several decades after. One claim that Jost makes is that much of the opposition to this research has been, implicitly or explicitly, political: findings such as, “conservatives have more authoritarian personalities, and liberals have more openness” are not value-neutral and do not fit into the mainstream of modern political science. (In contrast, a psychologist can feel more free to do such work, since psychologists do not have the same occupational inclination toward treating different political orientations symmetrically.)

The short version of the argument is: the data show correlations between personality types and political ideologies, but these results don’t fit well into the usual framework of political science, so this line of research is less well developed than it should be. It’s what Steven Pinker would call a taboo question. In this case, it’s people on the right who object to this research, who find it silly.

More recently, there’s been research on genetics and political behavior (see here for some of James Fowler’s work in this area), and I imagine there’s some resistance from people on the political left, considering the general association of genetics with racism or, more generally, social determinism (the idea that our positions in society are basically determined by our genes and thus (a) aren’t anybody’s fault, and (b) can’t easily be fixed.

Political cover?

I’m hoping that these two strands of research can provide political cover for each other. The people who study personality types can connect their work to genetics (or “human nature,” for the nonbelievers in evolution) to placate the conservatives, and the people who study genetics can discuss the personality research to keep the liberals at bay.

2 Comments

  1. paulse says:

    Some political beliefs appear to be shaped by environment.

  2. yolio says:

    What proportion of political affiliation are they attributing to personality/genes etc? Because I thought that political affiliation was closely related to group and region affiliation. For example, that people who move from Alabama to California become more liberal within a few years, and vice-versa.