Professor Kyogoku,

Your question about the implications of the new electoral laws is a good one. At the risk of preempting Frances, let me take a shot at answering it.

First, I think your understanding of Duverger is a bit different from the conventional reading, and is certainly different from what Prof. Rosenbluth was suggesting. You paraphrased Duverger correctly, but misinterpreted the implications for campaign appeals. Party-based, "programmatic appeals" are appeals based primarily upon public-goods-type policy platforms. They rely heavily upon the implicit content of the "party label." For example, in Britain, voters vote Tory or Labour (or Liberal in some places) based upon the perceived platforms of the two rivals for government. They may not even know about, and certainly pay less attention to, the particular characteristics of the individual Tory and Labour candidates running in their specific districts. Thus, Cain, Ferejohn, and Fiorina (1987: HUP) found precious little "personalistic voting" in the UK. In the U.S., despite the single-member district electoral system, the inability of parties to control who runs under their banners (because of open primaries) means that candidates can still supplement their partisan affiliation with personalistic campaigning.

The argument concerning the links between Japan's SNTV [i.e., Single Non-Transferable Vote] system and "vote gathering" based upon personalistic appeals was that LDP candidates had to find a way to compete with each other that maintained the integrity (in terms of its connotations for policy) of the party label. The LDP did not want its two or three candidates in a given district to run against each other as "hawks" or "doves" or as "conservative, moderate, or liberal" LDPers because to do so would confuse voters as to what voting LDP implied. All polls of Japanese voters indicate that they first decide which party to support, and then, if necessary (i.e., if LDP) decide which specific candidate to vote for. The first decision requires some notion of how the LDP is different from other parties, and intraparty competition of an ideological sort would hurt the LDP's ability to draw that distiction.

On the other hand, personalistic appeals, based on pork and favors could help to divide the vote efficiently without diluting the policy-import of the party label (although it certainly added "pork provider" as one aspect of what it meant to be a LDPer).

A switch to single-member districts, then, has two effects. First, it eliminates the intraparty competition AT THE DISTRICT LEVEL, and hence removes the need for same-party candidates to find a way to divide the vote. As long as the parties maintain control over who runs under their party label, this implies a move away from personalistic campaigning and toward party-based appeals.

The second, important effect is that the threshold of exclusion (the percentage of votes a candidate must win in order to guarantee victory) increases dramatically. Now a candidate must win 51 percent of the vote to be assured of a seat (a smaller number might be enough if there are more than two candidates, but Duverger's point was that the field should eventually be whittled to two). Under SNTV with multimember districts, one could be assured of a seat with as little as 25% of the vote in 3-member districts, and as little as 16.7% of the vote in 5-mbr districts. With a threshold this low, a strategy based on narrow, parochial appeals made sense: just add together enough local special interests until you've garnered "hard votes" totalling 1/6th of the district, and you win. Prof. Rosenbluth's point (taken from Duverger and Downs 1957) was that with the need to win a MAJORITY (50%+1) of the voters, candidates would be obliged to cast their nets more broadly, to appeal to "lowest common denominator" interests in the electorate. This could imply more consumer-based appeals (everyone is a consumer) and fewer "producer-based appeals" (i.e., less catering to special interests).

Hence, the elimination of district-level intraparty competition, in combination with the higher threshold of votes necessary to get elected should over time lead to a reduction in personalistic appeals to specific groups, to a reduction in the cost of elections in terms of pork and campaign financing, and to a restructuring of candidate-support organizations as a result.

One final comparative note in an overly-long contribution. In Germany, whose lower house electoral system Japan has adapted and adopted (there is one crucial difference concerning how seats are allocated between the PR and SMD sections), there is very little if any personal voting of the type common in Japan under SNTV. So little, in fact, that a colleague found it necessary at the last Midwest Political Science Association meetings to present a paper that tried to determine whether there is any "Personal Vote" in Germany at all. It is also true, however, that the still-low threshold of exclusion in the PR section (although it wil be higher in Japan than it is in Germany) permits some small parties to win a few seats in the national legislature, so the move to a 2-party systems stops a bit short at the national level.

Hope that helps more than it confuses.

Mike

---------------------------------------------------------------- Michael F. Thies Department of Political Science
(ph) 310-825-1976 UCLA -- Box 951472
(fax) 310-825-0778 405 Hilgard Avenue
thies[atx]nicco.sscnet.ucla.edu Los Angeles, CA 90095-1472
----------------------------------------------------------------