I found the summary of Prof. Kono's talk very informative. And I find his critique of sweeping attacks on Rational Choice convincing.

My question is, if who is principal and who is agent is a given as an assumption in P-A theory, then how can that assumption be verified in the specifically Japanese context?

The trouble I had with the Ramseyer/Rosenbluth book is that it seems to be using descriptive material previously used to conclude that the bureaucrats are dominant actors to draw the opposite conclusion without providing additional information.
If P-A theory is indeed more than an assumption but an alternative method to descriptive analysis, should it not have its own distinct way of proving/testing its assumptions?

I would appreciate if anyone can provide input on how P-A theory can show convincingly that the electorate/party is controlling the bureaucracy, and not the other way round.

Most of the list members might be aware of this but even descriptive analysis has questioned the "bureaucrats in control" thesis with mixed results----some negative, some affirmative, and mostly affirmative with conditions. It would be nice if P-A theory, or any other theory, could shed new light on this old but important debate.

Nobuhiro Hiwatari
Cambridge, UK

(Moderator's note: the book under discussion is _Japan's Political Marketplace_ by J. Mark Ramseyer and Frances Rosenbluth (Cambridge, Mass: Harvard U. Press, 1993).

----------------