RealClearPolitics - Tea Partiers Racist? Not So Fast

RealClearPolitics - Tea Partiers Racist? Not So Fast

Here is an interesting article where the Tea Party is shown to be racist by the way it answers questions which ask the respondent to generalize about blacks. The linked article is a defense of the Tea Party respondents, but even the article concedes that the existence of racial stereotypes is disturbing. But is that really fair? The questions ask the respondent to generalize about different races. The respondents do so. That one is willing to have beliefs about the different averages for different groups of various traits is not, it seems to me, to be evidence of what we call having stereotypes. When we say someone has a stereotype it is not to say that they have a belief that one group, on average, varies from another on a particular trait. It is that they are unwilling to evaluate new evidence or are unwilling to examine individual level evidence. These two seem hardly the same thing. If asked how Asians, on average, compare with whites on math ability I would have little trouble saying that Asians on average have better math abilities. But that hardly means that I am unwilling to entertain the possibility that the person in front of me who happens to be Asian is not good at math, or that the white kid standing in front of or in my class might not be good at math. Still less does it meant that I would ignore evidence that the gap is closing, or be resistant to believing that a particular white kid is better at math than a particular Asian kid, or that my evaluation of their individual math skills would be influenced by my knowledge of their race (though this last assertion would seem to be irrational from a Bayesian point of view. Does that mean that a Bayesian who wishes to be fair to other races must not have prior beliefs about different means among races? Surely it cannot. The theory of rationality that implies Bayesian updating is one that requires a rational decision maker to take into account all evidence, including information about averages of different groups).

It is like asking one whether he thinks that men are taller than women and then, when he answers men are taller berating him for thinking about categories of individuals rather than individuals and ruling out of bounds any discussion of whether men are, in fact, on average, taller than women. Of course they are. And of course believing this general statement in no way obligates one or even makes it more likely for one to assume that any particular man is taller than any particular women. I should know. At 5' 6" I believe that men are taller than women but have no trouble understanding that there are plenty of individual exceptions--I find myself looking up at them all the time.