Arnold Kling queries what I meant on that point in a recent talk. I meant the following:
1. There is an entire class of economists — a large class at that — whose choice of problems to work on bears little relation to what they think are important issues in the real world. I would stress, however, that it is difficult to find such economists (though there are some) at the top tier schools. This is a bigger problem at lower tier and wanna-be institutions.
2. I commonly meet economists and other social scientists who will tell you about the implications of their latest research, yet if you ask them other questions they will respond in hushed tones of the most severe agnosticism. For instance they will refuse to answer Robin Hanson's question about identifying large inefficiencies in the contemporary United States.
Now, if such agnosticism truly represents their actual views as human beings, that is a perfectly defensible stance. Yet I find that many (most?) of these same people will hold very definite political views and act on them in their private lives. They will support candidates, donate money, condemn colleagues who don't hold similar views, and so on. In other words, they are not really agnostic on all those other issues, they just don't want their personal views subject to full analytic scrutiny. They bifurcate the personal and the political.
This is one of my pet peeves. It is defensible to be truly agnostic. It is also defensible to believe that general principles of economic theory and empirics and ethics allow us to have "all things considered" policy views on matters we have not studied closely. It is not defensible to hold such views but, under the cloak of a not-really-meant agnosticism, refuse to put them on the social science table, so to speak.
(I find that bloggers hardly ever suffer from this problem. In many ways the core of blogging is a willingness to apply what you know to every problem you encounter, and see how good a job you can do of it in a more or less integrated fashion.)
3. Most intelligent people, in their ordinary lives, believe that other people, especially less intelligent ones, make stupid mistakes all the time, including in their decisive choices (not just in their voting). Yet some of these intelligent people call themselves rational expectations theorists. I don't get it.