The Chronicle of Higher Ed has an article on implicit bias which raises questions a number of philosophers have pursued. The general question is about the relation between implicit bias and biased behavior, with specific reference to the IAT (Implicit Association Test). The conclusion that is emerging is that getting rid of implicit biases will not get rid of the biased behavior. The point is made that from the fact that bias causes biased behavior, it does not follow that getting rid of bias gets rid of the behavior.
While that sounds true, we should know more. One reason could have to do with what ‘implicit bias in a person’ really is. Edward Machery has a lot of work on this question. But in general we have known for sometime that that beliefs tend to occur in networks, and changing the belief in one network may leave it more or less intact in another part. Another could be that the biased actions or their results are held in place by structural factors in the society. I am not sure who all have pursued this line of investigation, but Sally Haslanger has a lot of excellent work in this area. In my contribution to the 2-volume OUP work on implicit bias edited by Brownstein and Saul, I raised a question about this, and look at some recent psychological literature. Once one starts to look at the structural factors underpinning racist actions, for example, one can see that the needed change has got at least to include a change in social structures. (You can find a good article by Machery in the same 2-vol edition. Google will reveal lots of work by Haslanger.)
It would be great to get more reading suggestions on the structural side in the comments.
For people who have been following this literature at all closely, one of the biggest surprises is that one of the originators of the supposedly fundamental IAT seems to have changed sides! That’s Brian Nozick of the University of VA.