Because I myself am trying to cope, in my dissertation research, with the intersection of institutional environments that help to shape interests, I found Professor Schoppa's response regarding ideas and interests to be extremely helpful. I think he is probably right: the relationship would depend in large part on the issue area.

His empirical examples are also useful, but I have a question. In the case of land tax reform, Professor Schoppa notes that the enormous issue complexity, as well as number of actors, as well as complexity of interests involved, provided some space for ideas to make the difference. This seems to make sense, especially when compared to his earlier case in which interests were relatively straightforward. But I'm not particularly sure that this is logically necessary.
It seems to me that issue complexity, numbers of actors, etc., could combine to make ideas even harder to use and to manage, since recommendations by epistemic communities would likely face so many diverse perspectives. How can we be certain of the effect of epistemic communities by examining only clear policy outcomes?

One possibility for research (and this is one that I've long hoped we would see from the "epistemic communities" community) would be an additional comparison to an issue area in which we find a large number of interests, issue complexity, etc., and yet no substantive agreements or clear policy outcomes, largely because of the absence of epistemic communities. I suspect that many complex issue areas with a lot of actors fail to show any clear result, even though one might be able to find groups that approximate epistemic communites or policy networks working feverishly to make suggestions and to convince decisionmakers.
One possible case study might be the current financial crisis in Japan, in which we see a great number of opinion pieces from both Japan and abroad, and yet bureaucrats still look like deer frozen in headlights. I mean, someone had had to have been telling them that their response to the whole Cosmo collapse was a bit terrifying.

This makes research, of course, a bit tricky, since it requires that scholars look at issue areas that don't provide clear policy outcomes, or trying to figure out why the dog didn't bark. It seems to me that if we see the absence of an agreement at the same time that we see an absence of an epistemic community that is similar to those that help to produce agreements, we have a pretty good piece of evidence supporting this particular perspective. Again, however, I want to thank Professor Schoppa for his helpful comment; I look forward to the book, which will probably make my own dissertation much easier to write.

David Leheny
Department of Government, Cornell University