Bias In The Social Sciences Ctd

Apropos Andrew’s earlier post about the social sciences being dominated by self-identified liberals, I wanted to examine some of the reasons why the vast majority of psychologists in a room with Jonathan Haidt would self-identify as liberals. I suppose my first observation would be that this is a remarkably unscientific way to study anyone’s political views. Certainly Haidt is not going to use this as “research” in a serious academic paper – it is more an anecdote for popular publication (which is why it appeared in a New York Times article). Next we have the problem of definition – there are popular conceptions of what constitutes a “liberal” or a “conservative” but almost all of them are extremely malleable. Moreover people don’t always self-identify in the ways we expect them to do so. Musician Frank Zappa called himself a conservative while defending “offensive” rock music, but Ayn Rand – the current intellectual darling of the right – thought conservatives (particularly religious ones) were irrational defeatists.

Now one could still wave all this away and assert that even still, if there were a more reliable test for identifying liberals and conservatives as defined in our current political culture, there would still be a bias towards liberal politics in the social sciences (or at least in psychology). So where does this hypothetical bias come from? One or both of two things has happened: 1) People with liberal politics self-select for the social sciences, political conservatives are not as interested. 2) Something about pursuing education and/or a career in the social sciences makes people change there political views to more liberal ones. I suspect that it is in fact that both of these things – just as I suspect a mirror-image of this phenomenon happens in MBA programs or even most economics departments.