This post may be a little dated, since I was gone for the last two weeks. I'm sorry if it is no longer germane.

I'm writing in response to the discussion over P-A bureaucracy and comparative studies. I'd like to pose a question to the participants, and then add a small comment.

My question concerns P-A models. As I understand it, P-A as originally developed did not focus on the principal's dominance -- in fact, P-A was an attempt to understand how the Agent could gain slack from the Principal. As John Bendor put it, "It is trivial to show that given enough information about subordinates and sufficiently powerful instruments, politicians can induce bureaucrats to act as they want; it is much more interesting to show how control inefficiencies can emerge despite a principal knowing a great deal." As such, isn't the question not whether P-A is correct, but what are the conditions in Japan that lead to Political or Bureaucratic dominance? That is, isn't it true that comparative cases won't necessarily prove or disprove the Japanese case, because their situations may be different?

My comment focuses on Korea and the Philippines. I've done all my work on these two countries.

Regarding Professor Hiwatari's post of June 10, Korea during the 1960s provides a polar case of monitoring and sanctions. The authoritarian government under Park Chung-hee regularly monitored the bureaucracy, and regularly purged members. Korea might be the "trivial" case referred to by Bendor. On the other hand, the Philippines provides interesting comparisons. The democratic period preceding Marcos saw two rival parties, the Nationals and Liberals, and weak executive direction. Bureaucracies did indeed receive a fair amount of slack, similar to Barbara Geddes's description of Brazil. Under Marcos there _was_ reform, however, because of political direction from the top, increased efficiency only made rent-seeking more efficient!

I hope my digression to a "non-Japanese" focus is not too onerous!