Democracy Without Competition:
Opposition Failure in One-Party Dominant Japan
Ethan Scheiner
Stanford University
Party Competition Failure:
Challenges to Democracy
• Problem of one party dominance
– “Uncommon Democracies”
• If the party is popular Æ not a problem
• If the party is unpopular Æ failure of
democracy
2
Ruling Party Unpopularity:
Failure of Democracy, the Japanese Case
Japan
•
•
•
•
Liberal Democratic Party (LDP) Rules
Economic Collapse
Political Corruption
Voter anger
–
–
–
–
55% of public: no party affiliation
Only 20-30% support for ruling LDP
Typically low cabinet approval
44% dislike LDP
¾But no successful challenger to LDP
3
The Puzzle: Party Competition
Failure in Japan
What can explain opposition party
failure in a democratic system where
the ruling party is very unpopular?
4
Outline
I.
Introduction – Party Competition Failure: Challenges
to Democracy
II.
Framework
III.
IV.
Background on Japan
Candidate Experience as Key to Party Success
V.
Analysis of Local Opposition Failure
A. Japan
B. Comparative Typology
VI.
Implications for New Democracies
5
Part II –Framework
The Impact of Candidacies:
Explaining Party Competition Failure in Japan
• Key to party success: strong candidates
– Japan: Weak opposition candidates
• Underlying problem: opposition weakness at
subnational level
• But, WHY subnational opposition failure?
6
Central Argument
Clientelism
+
Financially Centralized Government Structure
Failure in Subnational Office Elections by Parties
not in the National Government
7
Part III – Background on Japan and Existing
Explanations for Opposition Failure
A History of LDP Dominance
1955-1990
• LDP: majority of the seats in every House of Representatives (HR)
election
• But slow decline in LDP support
1990-1995
• Growing Anti-LDP sentiment
• LDP split and temporary loss of power (8/93-6/94)
1996-Present
• Anger toward LDP remains
• Birth of centrist Democratic Party of Japan (DPJ)
• BUT LDP dominance continues
8
Past Failure to Explain
LDP Dominance/Opposition Failure
I. Party Popularity
• Miracle Economy
• Japanese Culture
• LDP Policies
Problem:
¾ LDP is not popular. It has not won majority of the vote since 1963.
II. Electoral Institutions
• Opposition coordination problems
Problems:
¾ Electoral system affected opposition and LDP.
¾ Opposition failure continues under new electoral system.
9
Part IV – Candidate “Quality”
or Experience is Key to Party Success
Chart 1: LDP Success: It’s the Candidates, not the Party
P r o p o r t io n o f P a r t y V o t e s W o n
45
40
Percentage of PR Vote
35
LDP
30
25
20
L e a d in g
O p p o s i t io n
P a rty
15
10
5
0
1 9 9 5 (H C )
1 9 9 6 (H R )
1 9 9 8 (H C )
2 0 0 0 (H R )
E le c tio n
P r o p o r t io n o f C a n d id a t e V o t e s W o n
Percentage of District Vote
45
40
LDP
35
30
25
20
L e a d in g
O p p o s it io n
P a rty
15
10
5
0
1 9 9 5 (H C )
1 9 9 6 (H R )
1 9 9 8 (H C )
E le c t io n
2 0 0 0 (H R )
0
10
Implication: The Importance of Candidates
LDP Has Candidate Advantage
• Confirmed by opposition
• LDP has more incumbents
The Importance of “Quality” New Candidates
• Jacobson (1990): “Quality”/experienced U.S. Congressional
candidates more likely to win
• Best source of “quality”: subnational level office
• LDP advantage in “quality” of new candidates too?
11
What is a “Quality” Candidate?
• Former local office holders
• Other: Former member of Upper House of
parliament, former bureaucrat, former television
newscasters, those who “inherited” seat from
family member
12
Chart 2: All “Quality” Candidates Do Well,
But LDP Has Higher Proportion of Quality Candidates
LDP
DPJ
Total Candidates
280
242
New Candidates
56
139
% of New Candidates
Who Were “Quality”
59%
18%
% of “Non-Quality”
New Candidates
Who Won
22%
12%
42%
48%
% of “Quality”
New Candidates
Who Won
“Quality” LDP and DPJ Candidates in 2000
(300 total single member districts)
13
Chart 3: LDP As A Party Is Not More Popular
LDP Success Is Due To Its Candidate Advantage
Former Local Office Holders More Likely To Win
(Predicted Probabilities of Victory)
Runs against
DPJ or LDP
Incumbent
No Incumbent
Opponent
Non-Quality
Former Local
Office Holder
5.8%
19.2%
31.1%
58.2%
¾ LDP advantage is in its higher proportion of quality
candidates and former local office holders.
14
Part V – The Underpinnings of the Recruitment
Problems of Japan’s Opposition
Q: Why doesn’t the opposition run more candidates with local
office experience?
A: Few office holders belong to opposition parties at the
subnational level.
The Reason
Clientelism
+
Financially Centralized Government Structure
Failure in Subnational Office Elections by
Parties not in the National Government
15
Definitions
Clientelism
• Contrasts with “issue-based” politics
• Patronage and pork barrel
Financially Centralized Systems
• Subnational reliance on central government
financing
• Transfers to localities: politicized
16
Clientelism + Fiscal Centralization Encourages
Local Pols To Affiliate With Nat’l Ruling Party
Clientelist Systems
• Mainstream local politicians must show they can bring in
patronage
Financially Centralized Systems
• Local level politicians’ primary function: help deliver
benefits from center
Clientelist + Financially Centralized Systems
• To gain central funding, local politicians have incentive to
ally with national ruling parties
17
Chart 4: Local LDP Hegemony,
Utter Opposition Failure at Local Level
P rop ortion of L egislative Seats H eld by th e L D P (1970-1997)
70%
L D P P re fe c tu ra l A ss e m b ly S e a ts
60%
Percentage
50%
L D P N a tio n a l H R S e a ts
40%
30%
20%
10%
0%
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Year
P rop ortion of Seats H eld by N on-L D P P arties (1970-1997)
70%
Percentage
60%
O p p o s itio n N a tio n a l H R S e a ts
50%
40%
30%
20%
O p p o s itio n P re fe c tu ra l A s s e m b ly S e a ts
10%
0%
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
Year
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
18
National-Local Pipelines of Pork:
The Reason for Local Opposition Failure
• LDP local hegemony due to efforts to maintain
“pipeline” between center and localities
• LDP patron-client relationships at core of pipelines
• Pipelines
– Discourage local party defection from LDP
– Encourage local party defection to the LDP
19
If Pipelines Are Important,
What Should We See?
• If fiscal dependence is important to local elections
¾ Opposition most successful where the pipeline
is less important
The opposition’s greatest success should
occur in the most autonomous prefectures.
20
Definition of “Autonomy”
Local Fiscal Capability Index
Locality' s Revenues (i.e., Local Taxes)
Autonomy Index =
Locality' s Spending " Needs"
Note: Central government caps local tax rates across country
21
Opposition Wins More Assembly Seats In
Autonomous Prefectures
Chart 5: Opposition Prefectural Assembly Success
by Level of Autonomy (1967-1991)
Proportion of Prefectural Assembly
Seats Held by Opposition Parties
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Local Autonomy Index
R=.57
22
Greater Opposition Local Success in Autonomous
Prefectures: Review of Statistical Results
Prefectural Assemblies
• Even controlling for other
variables,
– More opposition assembly
members in places and
times of greater autonomy
Mayors
• During periods of greater
autonomy:
– More opposition mayors
• During periods of lesser
autonomy:
– Fewer opposition mayors
– Rise in number of
opposition mayors who also
sought LDP endorsement
23
Comparative Perspective:
Japan Is The Rule, Not The Exception
Chart 6: Comparative Typology
Party-Voter Linkages
Level of Financial Centralization
ProgrammaticIssue Based
Clientelist
•
Decentralized
Centralized
Germany
UK
Brazil
Italy
Austria
Mexico
Japan
Local one-party dominance common in Clientelist/Financially
Centralized cases.
–
Exceptions similar to Japan’s
24
Part VI: Conclusion
Summary of Key Points
(1) Importance of “quality” candidates
(2) Major Contribution:
Clientelism
+
Fiscal Centralization
Local Opposition Failure
(3) Explanation for opposition failure in Japan at the
national level: A combination of (1) and (2).
25
Final Thoughts
• A Vicious Circle in Japan
– Parties cannot gain strength at national level without gaining at local.
– Cannot gain strength at local level without holding power at national.
• Hope for Japan’s Opposition?
– National party developments (new LDP defection?)
– Decentralization movement
– Growing anti-clientelist sentiments
• Implications for New Democracies
– New democracies likely to be clientelistic
– Important to create institutions that decentralize fiscal power
¾Decentralization can raise the quality of democracy
26
LDP As A Party Is Not More Popular
LDP Success Is Due To Its Candidate Advantage
Chart 3: Probit Model of New Candidate Success in 2000 (LDP and DPJ)
Variables
Quality
Former Local Politician
HC
Bureaucrat
TV
Inherit
LDP
Campaign Expenditures
Opponent Inherits
Weak Inherit
Ran in 96
Urban
Incumbent Opponent
Major Incumbent Opponent
Aggregated Model
Disaggregated Model
Coef.
Coef.
(SE)
0.702
1.506
1.549
1.729
a
-0.476
0.405
b
1.109
0.527
0.338
-0.294
-0.784
(0.333)**
(0.708)**
(0.468)***
(0.860)**
1.077
-0.457
0.851
b
1.023
0.525
0.293
-0.441
-0.839
(SE)
(0.282)***
(0.297)
(0.851)
(0.822)
(0.288)*
(0.155)*
(0.360)
(0.296)**
(0.582)**
Constant
-1.189
191
N
79.0
Percent Correctly Predicted
.244
Goodman-Kruskal λ (PRE)
55.52
Chi-sq
0.0000
Prob>chi-sq
.2662
Pseudo R-sq
-76.520
Log Likelihood
*p<.05 (one-tail), **p<.05 (two-tail), ***p<.01 (two-tail)
•
•
•
-1.294
184
88.2
.244
42.77
0.0000
.2282
-72.329
(0.313)
(1.187)
(0.824)
(0.295)*
(0.162)**
(0.442)
(0.301)**
(0.667)*
“LDP”: negative and non-significant
“Quality” & “Former Local Politician”: positive and significant
27
Former Local Pols: 15-30 percentage points more likely to win than non-quality cands
Opposition Wins More Assembly Seats In
Autonomous Prefectures: Statistical Evidence
Opposition Party Success in Prefectural Assembly
Elections (1971-1991): Panel Data Estimation
D ependent V ariable= Proportion of Seats W on by O pposition in
Prefecture i in E lection t (w ith logit transform ation)
V ariables
C onstant
A utono m y
G D P G row th
L ag of D ependent V ariable (Pref. i, E lection t-1)
1971 (dum m y variable)
Proportion of Seats W on by O pposition at H R
L evel in Prefecture i in last H R election before t
N
F (5, 46)
Prob > F
R -Sq
N um ber of clusters (prefectures): 47
C oef.
-0.543
0.182
-0.033
0.718
0.254
0.588
(Std. E rror)
(0.129)***
(0.055)***
(0.008)***
(0.051)***
(0.055)***
(0.187)***
280
463.95
0.0000
0.755
*p<.05 (one-tail), **p<.05 (tw o-tail), ***p<.01 (tw o-tail)
•Autonomy is statistically significant and positive
28
Greater Autonomy Leads to a Larger Proportion
of Local Executives Who Are Progressive
Chart A: Mean Levels of Autonomy and Proportion of
Local Executives Who Are Opposition
1.05
Mean Autonomy
Score
for All Prefectures
0.95
0.9
0.85
Proportion of All Mayors W ho
Are Opposition
79
19
78
19
77
19
76
19
75
19
74
19
73
19
72
19
71
19
70
19
69
19
68
19
67
19
66
19
65
19
64
19
63
0.8
19
Standardized Values
1
Year
29
Chart B: Rise in Proportion of LDP-Affiliated
Mayors, While Decline in Opposition-only Mayors
Once Greater Dependence on Central Government
0.55
LDP-Opposition Jointly
Endorsed Mayors
0.50
0.45
Proportion
0.40
0.35
0.30
0.25
Opposition-only Mayors
0.20
0.15
0.10
0.05
0.00
1976
1979
1983
1987
Year
30
Chart C: Fewer Progressive Mayors When
Fewer Cities Operate at a Deficit
0.25
Proportion of Localities with Opposition Mayor
0.15
0.1
0.05
Proportion of Localities Operating at a Deficit
0
19
75
19
76
19
77
19
78
19
79
19
80
19
81
19
82
19
83
19
84
19
85
19
86
19
87
Proportion
0.2
Year
31
Chart D: Correlates of Opposition Party Success or Failure
TYPE O F SYSTEM
C L IE N T E L IS T ?
No
G erm any
S w eden
UK
No
C anada
US
No
M e x ic o
T a iw a n
Y es
C E N T R A L IZ E D ?
Y es
P A R L IA M E N T A R Y ?
Y es
C A N D ID A T E CENTERED
ELECTORAL
SYSTEM ?
No
A u s tr ia
Israel
No
I ta l y
C O M P E T IT IV E
O P P O S IT IO N
Y es
IN S T IT U T IO N A L
P R O T E C T IO N O F
C L IE N T E L E ?
Y es
Japan
O P P O S IT IO N
F A IL U R E
32