<<

. 20
( 35 .)



>>

system™s size can be attributed to poor aggregation (Figure 4.6).


39
In Chapter 3, I used a version of the in¬‚ation score that ranged from 0 to 1. Here I
multiply it by 100 to convert it to a 0 to 100 scale.
Thailand: Aggregation, Nationalization, Number of Parties 111

11
10
Number of Parties
9
National Level
The Effective



8
7
6
5
District Level
4
3
2
1
0
1986 1988 1992b 1995 1996
Election Year
¬gure 4.5. Effective Number of Parties: District Versus National




70
60
50
40
30
20
10
0
1986 1988 1992b 1995 1996
Election Year
¬gure 4.6. In¬‚ation of the National Party System




Thailand™s average in¬‚ation score over the period is 54, meaning 54% of
the size of the national party system is the result of poor aggregation while
46% is due to the average number of parties at the district level. This
in¬‚ation score is quite high by comparative standards (see Table 3.1).
Where speci¬cally are aggregation failures occurring? Does the
degree of cross-district coordination vary across different government
administrative levels? For example, is it the case that coordination
is good within provinces or regions, but not across them? To answer
these questions, I calculate the extent to which candidates in
provinces with multiple electoral districts link across those districts.
Next, I examine aggregation between candidates across provinces
within a given region. Finally, I determine how much aggregation
occurred across different regions. Figure 4.7 displays the results.
Figure 4.7 traces the in¬‚ation of the party across each administrative
level. I take the average effective number of parties and in¬‚ation scores for
Building Party Systems in Developing Democracies
112

I = 34




I = 16 I = 21 I = 30
National
Regional
Provincial
District
Level
Level
Level
Level
3.2 effective 3.9 effective 4.9 effective 7.0 effective
parties parties parties parties




I = 54
¬gure 4.7. Where the Thai Party System Gets In¬‚ated (Multidistrict
Provinces Only)



all elections and all levels of government and show step by step where
in¬‚ation occurs as one moves from the district to national level. Clearly,
Thai parties failed to coordinate across the various administrative regions
in Thailand. Speci¬cally, 30% of the size of the national party system is
due to poor aggregation across Thailand™s 76 regions (I ¼ 30). Measured
in terms of percentage change, the party system grows by 45% between
the regional and national levels. However, poor aggregation is even more
pronounced at the subregional level. Between the district and regional
levels, the party system expands by 52% for an in¬‚ation score of 34.
Broken down even further, it becomes clear that there are aggregation
failures both between the provinces (I ¼ 21) and between districts within
the same province (I ¼ 16).40


4.5 conclusion
Returning to the question posed at the beginning of the chapter: Why
were there so many parties in pre-1997 Thailand? The large number of
parties was not a function of a permissive electoral system or social
cleavages that produced coordination failures at the district level.
Thailand™s block vote plurality electoral system is not an extremely

40
The aggregation data from Figure 4.7 allow me to examine the extent to which
Thailand has evolved into a party system made up of several major parties that
dominate different regions of the country as some have suggested (see Surin and
McCargo 1997). For an analysis, see Chapter 5.
Thailand: Aggregation, Nationalization, Number of Parties 113

permissive electoral system (though it is more permissive than a single-
seat plurality system). As a result, candidates, voters, and parties were
able to coordinate on a modest effective number of parties in each
district. The level of intra-district coordination was such that Thailand
would have had between three and four parties in the House of
Representatives if parties had been able to perfectly aggregate across
districts. I found also that the average effective number of parties at the
district level varied by district magnitude in a manner consistent with
what one would expect given the electoral system. (This is also further
evidence of the weakness of Thai party labels.)
Rather than intra-district coordination failures, the source of
Thailand™s in¬‚ated party system was poor aggregation “ the failure of
candidates to better coordinate across districts. The data show that
aggregation was poor between the districts, provinces, and regions.
This ¬nding begs some obvious questions. Namely, why has aggrega-
tion been so poor in Thailand relative to other countries? This is the
subject to which I turn in Chapter 5.
4.6 appendix




114
table 4a. Thailand™s Constitutional and Electoral History

Nature of Elections Total number Number of District Election
Constitution Duration Assembly Helda of MPsb MPs Electedb Magnitude Method
70 0
1932 (I) 5 months 13 days Unicameral None “ “
1933 156 78
1932 (II) 13 years 4 months Unicameral 1 to 3 Plurality
29 days (indirect)
1937 182 91 1 Plurality
1938 182 91 1 Plurality
1946 192 96c 1 Plurality
1946 1946 178 178c 1
1 year 5 months Bicamerala Plurality
30 days
1947 1948 99 99
1 year 4 months Bicameral 1 to 4 Plurality
14 days
1949 1949d 120d 120
2 years 8 months Bicameral 1 to 4 Plurality
6 days
1952 1952e 246 123
6 years 7 months Unicameral 1 to 6 Plurality
283 160
12 days 1957 (Feb.) 1 to 6 Plurality
281f 160f
1957 (Dec.) 1 to 6 Plurality
1959 240 0
9 years 4 months Unicameral None “ “
23 days
1968 1969 219 219
3 years 4 months Bicameral 1 to 15 Plurality
28 days
1972 299 0
1 year 9 months Unicameral None “ “
28 days
1974 1975 269 269
2 years Bicameral 1 to 3 Plurality
1976 279 279 1 to 3 Plurality
1976 340 0
11 months 28 days Unicameral None “ “
1977 360 0
1 year 1 month Unicameral None “ “
13 days
1978 1979 301 301
12 years 2 months Bicameral 1 to 3 Plurality
1983 324 324
1 day 1 to 3 Plurality
1986 347 347 1 to 3 Plurality
1988 357 357 1 to 3 Plurality
292 0
1991 (I) 9 months 8 days Unicameral None “ “
360
360 Plurality
1 to 3
1991 (II) 5 years 10 months Bicameral 1992 (March)
360
360 Plurality
1 to 3
2 days 1992 (Sept.)
391
391
1995 Plurality
1 to 3
393
393
1996 Plurality
1 to 3
1997 2001 500 500 1
8 years 11 months Bicameral Plurality
2005
8 days 100 (national PR
party list)
2007 2007 480 480
Current Bicameral 1 to 3 Plurality
Constitution 10 (8 regional PR
party lists)
a
Elections listed are for the lower chamber (House of Representatives). The Thai Senate was an appointed body until the 1997 Constitution. The 2007
Constitution made the Senate into a partially appointed body, with half of the seats appointed.
b
Members of the lower chamber (House of Representatives) of parliament when there are two chambers.
c
Two rounds of elections were held in 1946. The ¬rst, held under the rules of the 1932 Constitution, was for the 96 elected seats in the assembly. After
a new constitution was adopted in 1946, a second, supplementary round of elections was held to bring the number of MPs up to 178.
d
A bi-election was held in 1949 to ¬ll the 21 new seats created by the 1949 Constitution.
e
The 1952 elections were held in February, one month prior to the formal adoption of the new constitution.
f
In February 1958 the number of appointed members was reduced to 95. In March 1958 a by-election was held to ¬ll an additional 26 elected seats,
bringing the total number of elected seats to 186.




115
5

Explaining Aggregation in Thailand




5.1 introduction
What does the theory of aggregation incentives as described in Chapter 2
tell us about the roots of cross-district coordination failures in pre-1997
Thailand? Why was aggregation so poor? Drawing on the theory pre-
sented in Chapter 2, I argue that poor aggregation incentives in Thailand
re¬‚ected a diffusion of power within the national government (due to
party factionalism and an appointed Senate) and uncertainty over the
procedure for selecting the prime minister. In short, the expected utility
associated with being the largest party in parliament was relatively low in
pre-1997 Thailand. This discouraged greater attempts at cross-district
coordination. I also discuss and evaluate possible alternative explanations
(i.e., social heterogeneity and regionalism). I then devote the rest of the
chapter to analyzing the effects of constitutional reform in light of the
theory. Since the new Thai constitution and electoral system adopted in
1997 altered some of the variables I claim help shape aggregation incen-
tives, this episode of institutional reform is an ideal opportunity for the use
of comparative statics. In short, the constitutional reforms present me
with a natural experiment that I can use to test the predictive power of the
theory. The theory helps explain how and why the Thai party system has
changed since the constitutional reforms. Speci¬cally, I show that
improvements in the aggregation payoff resulting primarily from new
tools to combat party factionalism contributed to a dramatic improve-

<<

. 20
( 35 .)



>>