It is not necessary to wait until there is an election under the new election system in Japan to get a sense of what its impact is going to be on party organization, campaign strategies, party policies, and so on. We can make some preliminary assessments by looking at how politicians and parties have been trying to position themselves for the coming election. Doing so raises serious doubts about almost all of the assumptions about the likely consequences of the change in the electoral system made by many political scientists as well as by the Japanese politicians who advocated this electoral reform.
First, there is no evidence that the new system is producing a party system characterized by clear policy differences between the main parties. Quite the opposite is occurring. The reason simply is that the major parties embrace the same perception of the median voter. If Japan were characterized by a sharp class division, one could imagine parties in a single member system identifying different core constituencies and shaping their policy agenda to solidify their support in their core constituency while trying also to reach out to the floating voter whose swing from one party to another could spell the difference between victory and defeat. In other words, if Japan looked more like Britain the single member system would operate more like it does in Britain. But the LDP and Shinshinto are after the same voter and their policy agenda is driven more by a desire to avoid alienating this voter than by proposing policy alternatives to pull him away from supporting the other party. How much of this blurring of policy differences is a result of the new election system is another question. Obviously the end of the Cold War and other factors have a lot to do with it. Yet one might think that after five years of no growth there might be a debate about fiscal policy between the parties, but it is not there, and the reasons include concerns about alienating voters when now a much larger number of voter are needed to win than under the old system.
There is an important lesson imbedded in this reality: one should not treat any variable, however important, in isolation. The election system does not cause any effects in and of itself but only in combination with other factors. Unfortunately, Japanese political reformers for reasons that go back to the prewar origins of modern parties in Japan, are captivated by an old British model of party politics and believe that if only the institutions can be made to look more British so will the behavior. But what puzzles me is why political scientists, especially American ones, have embraced the same misconception.
Second, the new system is not weakening the links between politician and interest groups but seems to be intensifying them. It puzzles me why anyone should believe that a single member system would reduce the power of interest groups. Japanese politicians are desperate to expand their support base to win an election in which they will need in many cases more than fifty percent of the vote rather than twenty percent or so as under the old system. So they are reaching out for every hook into the electorate that they can find, and interest groups provide that access. Would anyone seriously argue that special interests in the US with its single member district Congress are weak? Why should they be any less important to politicians in Japan?
This is not the same question as whether interest groups are effective in mobilizing the vote. They are almost certainly less effective today than in years past, but that has nothing to do with the electoral system. After all, the Nihon Shinto won a lot of seats in the last election under the old system without any significant organizational backing. Perhaps in the next election, new candidates running without major party and interest group backing will win a lot of seats. But if one focuses on the impact of the electoral system on interest group-politician and party linkages, the new system is expanding those ties rather than reducing them.
Third, the reform supposedly was to create more party- centered and less politician centered election campaigns. The reasoning here was that with only one candidate per district, party headquarters could impose conformity among its candidates by its control over nominations and over money. But this is a mistaken assumption. Incumbency, not party headquarters, controls nominations. The new system will make it even more difficult to deny nominations to incumbents than the old one. The reason is that the incumbent from the LDP or the Shinshinto will not only be in office but he will invariably control the district party organization. Unless the headquarters decides to deny a role to the local constituency organization in determining nominations, the incumbent will have a strangle hold on the process. In any case, I do not know of any politician in Japan who worries about having his nomination denied after having been successfully elected. So far at least there is no evidence that the new system has increased pressure on individual politicians to adhere to policy lines defined by the party headquarters, though this is complicated by the fact, as mentioned above, that party headquarters seems determined to avoid defining policy lines.
Money is a more complicated matter. Let me just note that contributions to the party do not need to be made to central party headquarters under the revised law; they can be given to local party headquarters, i.e. to individual Diet members in their capacity as shibucho of the local party branch. Public financing apparently will be channeled through the party headquarters, which is going to give the Secretary General even more power than in the past, but it is virtually inconceivable that this will significantly compromise the pattern of automatic renomination of incumbents.
Fourth and a related point is the failure, so far at least, of the new system to reduce the importance of the koenkai. As far as I know, every politician in Japan is scrambling to expand his koenkai to meet the need for an expanded base within the electorate. Politicians do what they know how to do and what Japanese politicians know is how to build mass membership personal support organizations. There is no evidence that either the LDP or the Shinshinto has made any progress whatsoever in building local organizations that are distinct from the organizations of their candidates. Again, perhaps the koenkai won't be effective under the new system and if isn't it will decline in importance over time as other more effective methods of mobilizing support are discovered. But the point is that adoption of the new election system has not produced a move away from reliance on koenkai as a key strategic approach to mobilizing support.
Fifth, the new system does not seem to be fundamentally changing candidate recruitment patterns. Reformers argued that the new system would park the most senior politicians in the PR constituency and recruit new blood into the single member districts. But it is not working out that way: look at the success of Nakasone or Watanabe Michio or other senior leaders in resisting pressure to run in the PR and forcing themselves into the single member districts. One reason Nakasone apparently was anxious to run in the single member district rather than the PR constituency was to hold his koenkai together so that his son will have a base of organized support to use to succeed his father. In other words, the second generation phenomenon seems to be alive and well under the new system as under the old.
Now it is true of course that we won't know how the new system will affect behavior until there are a few elections held under it; and it is possible that it will be revised before its effects can be fully felt. But we know enough to know that this case of political engineering is not having the clear effects the engineers anticipated or that many political scientists assumed. The only to find out what effects it is having is to go out and look. Japan right now offers a unique opportunity to analyze the impact of institutional change on political behavior. I would urge scholars interested in electoral politics to organize themselves to do the field work needed to move this discussion of electoral reform into the real world of Japanese politics.

Gerry Curtis, Columbia University