19
results for the LDP in Table 2. Unlike the LDP, however, rival expenditures on DPJ vote
shares have their expected negative effect.
Table 3. Regressions of Expenditures on Vote Shares (DPJ 2000)
Variables
Model I
Model II
Model III
Own campaign-period spending
0.08(0.03)**
--
0.03(0.03)
Rival campaign-period spending
-0.03(0.02)*
--
-0.04(0.02)*
Own fund spending
--
0.03(0.01)**
0.03(0.01)***
Own party branch spending
--
0.02(0.01)**
0.01(0.01)**
Own koenkai spending
--
0.01(0.01)**
0.02(0.01)**
Rival fund spending
--
0.00(0.00)
0.00(0.00)
Rival party branch spending
--
-0.00(0.00)
-0.00(0.00)
Rival koenkai spending
--
0.00(0.00)
0.00(0.00)
Incumbent
2.77(0.95)**
2.22(0.91)**
1.95(0.92)**
Previous terms
0.62(0.22)**
0.42(0.21)*
0.41(0.21)*
Number of support groups
0.01(0.05)
-0.03(0.05)
-0.02(0.04)
Competition -0.36(0.05)***
-0.34(0.05)***
-0.33(0.05)***
Number of candidates
-2.03(0.35)***
-2.04(0.33)***
-1.95(0.34)***
% Densely inhabited district (DID)
-0.04(0.01)***
-0.05(0.01)***
-0.05(0.01)***
Number of cases
242
242
242
Intercept
30.35
29.83
30.18
Adjusted R-squared
0.48
0.53
0.54
Note: Dependent variable = DPJ percentage of total district vote.
* = significant at .10 level. ** = significant at .05 level. *** = significant at .001 level.
In Model II, the effects of the year-based figures on votes are reported without
campaign-period spending. Both own party branch and fund agent expenditures are
statistically significant in the expected direction with a small effect on vote shares (1%
and 2% for ¥100 per elector). However, it appears that none of the rival expenditures
have the expected negative correlation with DPJ vote shares. Unlike the results for the
LDP, this suggests that DPJ vote shares, as a whole, are not pivotally shaped by the
Convention