[Moderator's Note: The following is an excerpt from the August 1995 edition of Social Science Japan, the newsletter of the Institute of Social Science, University of Tokyo. The paper edition will be sent out to subscribers at the end of this month. Air mail subscription to SSJ is free, and can be arranged by
e-mailing your surface-mail address to ssjinfo[atx]iss.u-tokyo.ac.jp -- Please write "Sub SSJ" in the subject header of your message.]
----------------------

Copyright Institute of Social Science, University of Tokyo

An Unconsummated Marriage: Policy Networks and Two-Level Theory

Jonathan LEWIS

THE focus on Japan's domestic political economy in international trade negotiations prompts us to reflect on how to study policy processes simultaneously taking place on the domestic and international stages. One possible approach is a combination of policy network analysis and two-level theory. Both methods have been used separately in studies of Japan, but a combination of the two could tell us more about how Japan's international relations interact with domestic government- business relations, bureaucratic rivalry, and the role of politicians.

Policy network analysis is a set of analytical tools for describing which actors are involved in deciding particular policies, the strategies they follow, and how they depend on each other for funding, information, authority, and other resources. Policy networks span both institutionalized policy-making processes and informal channels of communication such as old-boy networks. By means of systematic description of such variables as numbers of private and public actors, levels of resource dependency between actors, and levels of professionalization, policy processes in different sectors and countries can be compared. Policy network analysis rejects state-level generalizations of "strong-state, weak-state", and makes neither elitist assumptions of restricted policy processes nor pluralist assumptions of open ones.

Two-level theory addresses the relationship between international negotiations and domestic politics. A branch of game theory, it depicts a national representative simultaneously negotiating at the international table with representatives of other nations and, at the domestic table, with those who must ratify and implement any agreement reached. Each representative has a "win-set", the range of possible international agreements which would be acceptable to domestic constituents. If the two sides can agree on an outcome which lies within both sides' win-sets, the negotiations are successful. If the win-sets do not overlap, the negotiations break down. Two-level theory is important because it cracks the conundrum of whether to consider countries as unitary actors pursuing diplomatic strategies or as mere aggregates of domestic disputes and alliances. The two methods would seem to have little in common. One is a framework derived from sociology and organizational analysis to facilitate the description and comparison of policy processes, the other a theory derived from economics regarding how to achieve particular policy outcomes . What is the case for combining the two?

First, the gap between processes and outcomes disappears when one studies particular cases. This fact is reflected in the growing sophistication of both methods: Policy network analysts are advancing theories about the kinds of policies likely to result from particular types of networks, while two-level theorists are investigating the formal and informal processes by which particular outcomes are reached.

Second, policy networks offer a useful extension to two-level theory's "domestic table". I suggest we can identify three categories of foreign pressure on domestic policy networks: those which aim to change the networks dealing with an existing issue; those which create a new issue to be dealt with by existing networks, and those which create new issues requiring new networks.

In the first category, Schoppa's study of the Strategic Impediments Initiative talks suggests that Japanese agreement on changes to domestic market structures is more likely to be forthcoming if the US could arouse the interest of a previously uninvolved Japanese constituent. However, Schoppa omitted to discuss how, even if new interest groups acquired a direct interest in an issue, they could gain access to policy networks. In Japan as elsewhere, not all members of a policy community necessarily have access to the policy network.

An example of an initiative in the second category was the invitation to Japan to participate in the International Space Station. US proposals were immediately taken up by the Japanese aerospace policy network, and the subsequent pattern of resource dependencies - government funding, contracts for all the major aerospace manufacturers, and a minimal role for space science despite its formal prime importance - followed that of most other aerospace projects. Participation in the space station fitted in neatly with existing Japanese space policy, and the network already in place was able to coordinate policy across many institutions very quickly.

Similar requests for Japanese cooperation on the Texas Supercollider, however, met with silence and prevarication, suggesting that the case falls into the third category. Japanese participation would have required the creation of a new policy network to deal with the new issue and fight for government funding of it. Moreover, Japanese equipment companies had little incentive to form such a network because the Supercollider offered their potential US competitors in the superconducting magnet market an opportunity to catch up.

While there is some evidence of loose global policy networks emerging in areas where international private capital or shared professional expertise are the chief resources, most policy networks are still emphatically national; the more so, the greater the dependence on public funding, protection and regulation. The time is therefore ripe for a combination of policy network analysis and two-level theory. The union of bedfellows from different ends of the social science spectrum could produce unexpectedly robust offspring.

FURTHER READING:
KNOKE, David; PAPPI, Franz; BROADBENT, Jeffrey and TSUJINAKA, Yutaka (forthcoming): Comparing Policy Networks: the Politics of Labor Policy Formation in the US, Germany and Japan. Cambridge University Press
KOIKE, Osamu (1995): Seisaku network to seifukan kankei ("Policy Networks and Intergovernmental Relations"),Chuo daigaku shakai kagaku kenkyuujo houkoku no.16, February, pp.27-46
PUTNAM, Robert D. (1988): "Diplomacy and domestic politics: the logic of two-level games", International Organization, Vol.42 No.3 (Summer), pp.427-460
SCHOPPA, Leonard J. (1993): "Two-level games and bargaining outcomes: why gaiatsu succeeds in Japan in some cases but not in others", International Organization, Vol.47 (Summer), pp.353-386
WILKS, Steven and WRIGHT, Maurice, eds. (1987): Comparative Government-Industry Relations: Western Europe, the United States, and Japan. Oxford: Clarendon Press

Jonathan LEWIS is a Research Associate at the Institute of Social Science,
University of Tokyo, and a PhD candidate at the University of Sheffield, UK.
Electronic mail to: jonathan[atx]iss.u-tokyo.ac.jp