Let me put my two cents in on the questions of party platform, electoral systemand the like. First, there seems to be very clear evidence that (other things being equal), electoral systems do have consequences. Japan's multi-membersystem clearly made it possible for, and indeed actually encouraged the continuation of, many smaller niche parties. They could contest eletions successfully due to the fact that successful parliamentary candidates needed only 12-15% of the vote from a constituency to get elected. This was clearly not the case in single member district systems. And to the extent that Japan's new system comes close to that of Germany's we can probably look for a consolidation in the number of parties, but given the continuation of some p.r. seats, we can also expect that a few small parties will continue. And this multimember system also helped encourage factions within the LDP for the well- quoted reason that any party wishing to win a parliamentary majority neeed to elect more than one candidate from a district. This will be less true in the new system as regards the smds (but it will continue in the pr seats).
None of this is to negate Gerry's point about the importance of incumbents. And we can surely expect that any adjustments in party system, election platforms, voter support patterns and the like will take time to happen and to shake out.
Remember, for example, the importance of personality factors involving Yoshida and Hatoyama and their supporters in the 1950s and into the 1960s--factors which were all but dead a decade or two later, even though the specific factions they created continued as institutions, long after the personality factors that helped generate these factions had disappeared.
But over and above the formal institutional setup is the question of party divisions. Gerry's point about Japan being far less class-riven than Britain, combined with the fact that the LDP and Shinshinto politicians are trying to capture the same voters are surely true. At the same time, when there are not too many "real" issues to divide voters (and hence for politicians to try tapping into), the temptation for politicians is to create more "artificial"issues. Consider only the situation in the US with for the most part the parties are looking for rather similar voters--the middle class (however defined), suburban, largely white voter, etc. etc. But each starts wtih enough history, ideological baggage, easily-identifiable labels and so forth, so that by and large, far more lower class and minority voters gravitate toward the Dems. and far more well-to-do upper class voters lean Republican. But the parties (and individual candidates ) are always looking to identify wedge issues that can cut into their opponents' bases and perhaps broaden their own. Thus, even on issues like gun control or abortion where public opinion suggests that 60-80% of the voting public has a pretty clear position in favor of gun control and for relatively free choices for women on abortion, the parties do not both gravitate to such "majoritarian" positions. Instead, the Republicans in particular take a hard line on no gun control and no abortion. Why? Simply because these two issues are of such strong interest to a number of voters that the party stalwarts know that they can win (or lose) significant numbers of voters by taking a strong stand that is against the views of the majority (who may not care enough about either of these single issues to actually "vote" them to the exclusion of other issues).
When one turns to Japan, the conclusion to me seems to be simple: in a few years, there will be a gravitation toward a relatively limited number of larger parties which will try to distinguish themselves from one another on various issues. Class a la Britain will almost certainly NOT be one of these; nor will race or ethnicity. These simply are not salient social divisions in Japan today.
But that doesn't mean that divisions don't exist, and that doesn't mean that politicians and parties won't try to find and extend these divisions. Among the more important divisions in my mind are such things as "open vs closed" economy
(different groups will benefit or not by freer or less free trade and investment policies); "national" vs "international" posture for Japan (again, the ideological divisions from the past, combined with new nationalist sentiments mean that many Japanese differ in real ways on the role of the SDF, PKO forces, close coopertion with the US vs. ASEAN, etc. etc.)
In effect, therefore, I think we can look forward to some real changes in the party system, and in the issues that divide parties in Japan once the new electoral system takes place. But it probably will not be in the direction of any of the sillier directions suggested by American journalists who are misinterpreting U.S. or British politics. Rather it will be in the direction of many of the longstanding postulates of serious students of electoral systems and comparative party systems.
But as a final caveat, these changes will clearly take place within a Japanese socio-economic context. And hence, we need to look, not at any simple single variable--such as the elctoral changes themselves--but at the broader context, or the tableaux, on which that set of electoral changes is super imposed.
T.J. Pempel