The relationship between urbanization, political reform and party voting in
Japan still seems unclear.

If urbanization was the "cause" of the difficulties of the LDP and the basis for
political reform, then why did reform occur so long after urbanization had
largely been completed? The question of timing of reform is problematic even if
we assume that urbanization played a role.

Another question is how did urbanization make personalistic voting more
difficult for the LDP? The LDP did carve out enclaves of support even in the
largest cities. In the rural areas, it is not clear that personalistic
strategies, such as taking credit for pork barrel projects, declined over the
past 30 years, and may have even increased as the LDP intensified its level of
support in rural areas to compensate for decline elsewhere.

Even if urbanization did cause a decline in the use of personalistic strategies,
what was it about urbanization which caused this decline to occur? Does this
phenomena support the modernization school view that rural Japan represented
traditional society and that modern urban voters would be more sophisticated and
less likely to vote for the LDP? What about "crisis and compensation" and all of
the arguments about the responsiveness of the LDP and their ability to co-opt
popular policies?

The U.K. analogy is also unclear. In the U.K., electoral reform was often
accompanied by the expansion of the franchise, and the politics of the
incorporation of the working-class was often a factor in party competition.
However, when the reform process was complete, the Labour Party began to emerge
from the Liberal Party creating three party competition (and four with the rise
of Irish nationalism). It would be difficult to predict these party system
outcomes from electoral rules alone.

It is true that the average size of electoral disticts in the U.K. increased,
though not only because of the elimination of 2 member constituencies, but also
due to the end of "rotten boroughs" which sometimes consisted of a very small
number of electors sending a member to parliament. Ironically, the size of
electoral districts is now much smaller than before in Japan and one might
imagine that they are very conducive to personalistic strategies, especially if
the LDP can continue to deliver benefits to the constituencies of their members.
(I am sure the New Frontier Party would aim to do the same if in power).

It is true that single member constituencies in theory lead to two party
competition for the median voter along an ideological continuum. Yet, even
single member constituency based systems have been known to produce predominant
party systems. Many scholars and commentators may have prematurely abandonned
the notion of a predominant party in Japan based on the dramatic events of
1993-94.

Of course, only empirical work will tell us if party oriented voting will
increase, or if electoral reform will produce effective two party competition
Japan. One problem is that different theories will search for different sets of
empirical evidence. Even so, the real test of any theory with claims to
scientific rigour would be propose a set of hypotheses derived from the theory
to test against the results of the next few elections. It would be interesting
if the hypotheses of the research planned to determine the impact of electoral
reform were made available in advance of the field studies.