The recent discussion of policy networks has been very helpful as I am trying to use the concept in my dissertation research on agricultural in the U.S., France, and Japan.

I would like to briefly describe my use of policy networks to the moderators as they are familiar with this literature.

I use the term policy networks to describe types of relations between bureaucrats, politicians, and producers. Using the concept of resource dependency put forward by R.A.W. Rhodes, we can evaluate different networks according to whose interests are more heavily weighted in the formation of policy. That is, by virtue of institutions such as executive agencies, political parties, and interest groups, bureaucrats, politicians, and producers have varying capacities to supply resources needed by other actors in the policy process such as votes, policy outcomes, or support in inter-ministerial conflicts. Different combinations of institutions create different relations of resouce dependency and, therefore, different degrees of influence over policy outcomes by politicians, producers, and bureaucrats.

I describe political parties, interest groups, and executive agencies as dichotomous variables according to whether or not they afford respective actors in the process with the resources that provide influence. Strong political parties that enforce discipline and increase chances of reelection are contrasted to weak political parties that require individual politicians to construct their own electoral coalitions. Unitary executive agencies that have clearly marked jurisdictional boundaries are contrasted to fragmented agencies that suffer from bureaucratic competition over jurisdiction. Finally cooperative interest groups that organize producers into a single bloc are contrasted to competitive groups that are rivals for influence in the policy process.

This combination of 2 x 2 x 2 variables yields eight distinct networks (the dominant actors in each network are enclosed in parentheses): statism (bureaucrats), clientelism (bureaucrats and producers), corporatism (bureaucrats, politicians, and producers), parentela pluralism (politicians and producers), sponsored pluralism (bureaucrats and politicians), pressure pluralism (politicians), iron triangle (producers), and issue network (none). In other words:

unitary agency + weak party + competitive int. group = statism

unitary agency + weak party + cooperative int. group = clientelism

unitary agency + strong party + cooperative int. group = corporatism

fragmented agency + strong party + coop. int. group = parentela pluralism

unitary agency + strong party + competitive int. group = sponsored plur.

fragmented agency + strong party + competitive int. group = pressure plur.

fragmented agency + weak party + cooperative int. group = iron triangle

fragmented agency + weak party + competitive int. group = issue network

These networks form around sectoral policy areas such as agriculture. They facilitate comparative policy analysis by providing a straight-forward, yet nuanced approach to how institutions affect the relations between actors in the policy process. Although bureaucrats, politicians, and producers are all engaged in policy formation, I argue that policies will bear the distinct mark of those interests which have greatest influence over outcomes by virtue of the resource dependencies that follow from institutional endowments.

Policy analysis may proceed through two steps. First, we analyze the relevant institutions in each country to determine the policy network. With this in mind, differences in national policies may be partially explained by the interests of those actors who most heavily influence policy outcomes.

In the U.S., France, and Japan during the 1950's and 1960's regulatory regimes for agricultural price and income support policies were formulated. At the same time, important institutional developments were taking place in the agricultural policy networks.

In the U.S., party competition over farmers in the late 1940's and early 1950's gave way to bipartisan policy formation within the confines of congressional agricultural committees. With parties weakened during the 1960's and the Department of Agriculture competing over jurisdiction with the Departments of Interior, State, and others, a cooperative logroll of specialist commodity organizations could control policy under the iron triangle.

In France, the Fifth Republic constitution shifted power from the National Assembly to the executive ministries by curtailing the powers of the Assembly over the legislative process. At the same time, important reforms in the Ministry of Agriculture solidified control over traditionally independent technical corps in the countryside as well as resolve longstanding rivalries with the Ministries of Education and Public Works over agricultural education and rural construction respectively. The peak farm organization, the FNSEA, developed close ties with the Ministry and worked to marginalize competing farm interests through clientelistic relations with the state.

In Japan, a corporatist policy network emerged in agriculture as the cooperative association movement recreated in 1947 became essential to the conduct of government food control policy, especially after the introduction of the subscription system for rice control in 1955. The only competition to the political arm of Nokyo, the Japan Farmers Union, declined in importance after the land reforms had been institutionalized. Meanwhile, the party instability of the immediate postwar years gave way, of course, to the nearly forty year control by the LDP. LDP party leaders such as Kono Ichiro established party control over ministries such as Agriculture and policy-oriented party institutions such as the agriculture Policy Affairs Research Committee became central to policy formation. In clear contrast to France, the agricultural policy network in Japan is distinct because of the role both the party and the relevant ministry play in decision-making.

Given these differences in what interests dominated policy during the late 1950's and 1960's (producers in U.S., producers and bureaucrats in France, and producers, politicians, and bureaucrats in Japan), there are corresponding differences in policy that bear the mark of these dominant actors in the policy process.

Without getting into too much detail, let me briefly discuss some of these policy differences. In the U.S., mandatory acreage restrictions and price supports were replaced in the 1960's with voluntary programs that gave farmers the freedom to decide whether to participate in government programs or not.
Specific allotments of crop acreages were replaced by set-asides in which farmers were only asked to remove a percentage of total acreage with the feedom to plant whatever they wished on the remaining acreage.

In France, the Common Agricultural Policy, which began for wheat in 1962. assured high prices for major products such as wheat and dairy while French national policies assisted with structural improvements such as increases in farm size, subsidies to young farmers, and retirement plans for elderly farmers. As a result, the FNSEA became an important medium between farmers and the state (see Keeler). At the same time, the relocation of price-support decisions from Paris to Brussels assured that the executive ministries and not the politicians in the National Assembly would control policy.

Finally, in Japan agricultural policy making centered around the yearly rice pricing system and the maintenace of the 1942 Food Control Law as the basis of policy. As a result, Nokyo continued to grow as the flow of rice to the government and the flow of subsidies and payments to farmers helped make the banking arm of Nokyo, Norinchukin, one of the largest banks in the world. For the Ministry of Agriculture, Forestry and Fisheries, the Food Control Agency responsible for policy grew to nearly 12,000, about a third of the entire agency (MITI total is about 8,000). Finally, the near-exclusive focus of policy debates on the price of rice turned agricultural policy into what Mayhew calls a particularized benefit. A focus on price made it easy for politicians to point to concrete benefits that they helped to secure for rural constituencies.

As Rhodes and others point out, the advantage of policy networks is that they focus attention on the relations between individuals and not just the structure of institutions they inhabit. In addition, policy networks help us conduct more nuanced comparative studies. As opposed to lumping together Japan and France as "strong states" for example, policy networks help us make distinctions that the previous focus on state structures caused us to miss.

Adam Sheingate
Yale University
sheingat[atx]minerva.cis.yale.edu