Len Schoppa's latest comments on a series of wonderful posts regarding to policy networks made me realize ways to overcome the problems with (historical)
institutionalist analysis. Since I know almost nothing about the policy network literature but a little about institutional analysis, it was a discovery to me (and I would like to thank list contributors for directing my attention to this).

To put it crudely, institutional analysis, especially when applied to a longer time frame, tend to sound deterministic since institutions effect policy outcomes. Thus, for example, Desmond King's thorough research on the development of employment policies in the US and UK (Actively Seeking Work?, Chicago U.P., 1995) tend to give the impression that attempts to introduce active labor market policies in the US and UK in the 70s were doomed to fail and the retrenchment of such policies in the 1980s were destined to succeed because of the "liberal" nature of employment policy institutions set up in the UK (in the 1910s) and the US (in the 1930s).

Probably a more dynamic process could have been portrayed if we understand the obvious fact that there are several kinds (degrees?) of policy networks can be formed in relation to existing policy institutions. Thus, on the one end, there are policy networks that receive feedback and information from existing configurations of institutionalized interests (more or less the "subgovernment" model) while on the other pole there are policy networks that define issues in a way that cross cut existing institutions and/or aim to create new institutions and thus if the policy is implemented create new networks. In the latter case, it seems the issue is defined in a way to expand the scope of interests involved/mobilized and/or create new interests (the classical Schattschneider notion) by the power of ideas. This notion of networks allow researchers to incorporate the notion of institutional change (not only creation) as part of policy change into institutional analysis.

Although the above must have been obvious to many readers, the point I wanted to empathize is the dialectic relationship between the power of ideas/knowledge and the restraints of institutions with relation to policies can be made clearer, especially in terms of comparative analysis, if policy networks are considered with their relations to existing institutions, since the two approaches (network and institutional analysis) seem to be complementary.

Needless to say what I have in mind are domestic institutions and I am not sure to what extent this can be said of international institutions.

Nobuhiro Hiwatari