. 21
( 35 .)


ment in cross-district coordination and an accompanying fall in the
effective number of electoral parties. The ¬nal section concludes with a

Explaining Aggregation in Thailand 117

few words about aggregation in light of the 2006 coup and the
subsequent constitutional reforms. The new 2007 Constitution was an
attempt by Thailand™s conservative forces to return Thai politics to the
pre-Thaksin era, and it thus reduced many of the new aggregation
incentives introduced in 1997. An analysis of the preliminary results
from the December 2007 elections demonstrates that, as expected,
aggregation declined in the wake of these reforms.

5.2 explaining poor linkage in pre-1997 thailand

5.2.1 The Size of the Aggregation Payoff “ Vertical
and Horizontal Centralization

The case of Thailand supports the argument that a concentration of power
at the national rather than the subnational level (vertical centralization) is
not enough to produce strong aggregation incentives. Thailand is a unitary
state. Governors appointed by the Ministry of Interior (MoI) headed the
country™s 76 provinces. The few locally elected of¬ces during the pre-1997
period “remained in the MoI™s sphere of power” (Nelson 1998, 34). Nearly
without exception Thailand is characterized by academics, bureaucrats,
and politicians as an extremely centralized state “ political and economic
power remain heavily concentrated at the center (Bangkok) (Nelson
1998).1 For example, over the 1980s and 1990s, the central government
controlled 92% of all government expenditures and 95% of all govern-
ment revenues (World Bank n.d.). As a percentage of GDP, subnational
government™s share of total revenues and expenditures was just a bit more
than 1% (World Bank n.d.). Much of the revenue subnational governments
would collect went directly to Bangkok “ for example, 90% of the sales and
excise taxes and 60% of the local surcharges collected by local governments
were transferred to Bangkok (Hunsaker 1997, 144).2
If we focus solely on vertical centralization, we would expect
Thailand to have excellent aggregation. However, as Chapter 4 demon-
strated, this is not the case. In short, the concentration of power and

For more information regarding the dominance of the central government over local
administrative units in Thailand, see Nelson (1998) and Missingham (1997).
To say that power is extremely centralized is not to say that local government units
have no power. Hunsaker (1997) demonstrates that municipalities can and do bargain
with the central government over revenue grants.
Building Party Systems in Developing Democracies

resources at the national level in Thailand does not translate into
suf¬ciently strong incentives for cross-district coordination. Instead, one
must look to the second component of the aggregation payoff “ the
concentration of power within the national government, or horizontal
Leaders of the largest party in pre-1997 Thailand faced signi¬cant
checks on their power, which reduced the perceived payoff to aggrega-
tion. In Chapter 2, I argued that the presence of bicameralism, reserve
domains, and party factionalism all reduce the size of the aggregation
payoff. Each of these was present in pre-1997 Thailand. First, Thailand
has a bicameral legislature consisting of a House Representatives
(saphaphuthaenratsadorn) and Senate (wuthisapha). The House of
Representatives was the primary legislative body “ its approval is re-
quired for all legislation. The Senate could delay a bill but the House can
eventually pass legislation over the objection of the Senate. Strictly
speaking then, the Senate was not a formal veto gate and theoretically
could have served as only a mild check on the authority of the largest
party in the House. However, while the Senate did not have the formal
power to block legislation, it was none-the-less dif¬cult for elected
governments to ignore the interests of the Senate. Until 1997, the Senate
was an appointed body. During the 1980s, these appointees were gen-
erally former military of¬cials and bureaucrats. Representing as it did the
interests of Thailand™s conservative forces “ who, recall, had a long
history of intervening to shut down democratic institutions “ the Senate™s
position on a matter carried a good deal of weight. This was especially
true when the Senators were relatively united on an issue. In effect, the
Senate functioned as a reserve domain. The existence of an unelected
Senate stocked with representatives of the military and bureaucracy
meant that the parties that controlled the House and cabinet still did not
hold all the reins of power.
Throughout much of the 1980s, another portion of the potential
aggregation payoff was off limits to party leaders “ macroeconomic and
budgetary policy. This reserve domain emerged as the result of
a compromise between elected politicians, Thailand™s conservative
forces, and Prime Minister (and former general) Prem Tinsulanonda.3

For more on this compromise, see Hicken (2001, 2005), Christensen et al. (1993), and
Doner and Ramsey (1997).
Explaining Aggregation in Thailand 119

As part of this “pork-policy compromise,” macroeconomic and
budgetary policy was shielded from elected politicians and run by Prem-
backed technocrats. In exchange, the political parties were given control
of the sectoral ministries (Commerce, Industry, Education, Agriculture,
etc.) and were allowed to run them as they saw ¬t provided they avoided
major scandal and respected the budgetary ceilings set by the techno-
crats (Hicken 2001, 2005).
By the early 1990s, both of these reserve domains were withering
away. In 1988, an elected politician became Thailand™s prime minister
for the ¬rst time since 1976 and immediately did away with the pork-
policy compromise by seizing control of macroeconomic and budget-
ary policy (Hicken 2001, 2005). In addition, the composition of the
Senate gradually changed so that over time business interests came to
make up a larger and larger portion of Senate appointees. Yet, while
these reserve domains were in place, they represented a signi¬cant
diffusion of political authority and a disincentive for aggregation.
A third reserve domain exists in Thailand, one that did not wane
during the 1990s “ the institution of the monarchy. While the Thai king
typically plays a largely ceremonial role in politics, Thai constitutions
have reserved signi¬cation powers for the king in the area of legislation.
Namely, by refusing to sign a bill into law, the king can effectively veto
that legislation, and it then requires a two-thirds vote of both houses to
pass the bill over the king™s objections. Although the king has rarely, if
ever, invoked this power, the fact remains that even a leader of the
large, majority party faced a potential check on his authority.4
Another factor contributing to the horizontal diffusion of power “
one that did not change over the course of the 1980s and 1990s “ was
the factionalized nature of Thai parties. As discussed in Chapter 4, Thai
political parties were extremely factionalized, due in large part to the
nature of Thailand™s pre-1997 electoral system. Because of the rampant
party factionalism, the leader of a political party was more like a ¬rst
among equals than the head of a political hierarchy. The fact that the
leader of the largest party might still ¬nd his power checked by rival

In addition, the revered status of Thailand™s current long-serving king means that on
the occasion when the king weighs in on a political or policy matter, politicians
disregard that advice at some cost. The clash between elected Prime Minister Thaksin
Shinawatra and the monarchy was a major factor behind Thaksin™s ouster from of¬ce
in a September 2006 coup.
Building Party Systems in Developing Democracies

factions within his own party discouraged greater attempts at
aggregation. Party factionalism was a major cause of the frequent
cabinet reshuf¬‚es and short-lived governments that so characterized
Thailand™s pre-reform system as disgruntled faction leaders actively
sought to bring down their rivals, including sometimes the nominal
head of the party “ the prime minister (see Chambers 2003).
Consider the history of the New Aspiration Party. The NAP vaulted
to power as the largest party in 1996 after luring several factions to the
party with promises of cabinet portfolios. The party, with its eight
factions, was still far short of a majority in parliament with only 32% of
the seats.5 It was therefore forced to invite ¬ve other parties to join with it
in coalition “ most of which were also composed of multiple factions.
During its term in of¬ce, the various factions within the NAP were
perpetually in con¬‚ict. Indeed, factional con¬‚icts were among the
factors behind the government™s slow and inconsistent response to the
economic crisis of 1997 (MacIntyre 1999; Chambers 2003). NAP
faction leaders clashed with each other over how to respond to the
growing crisis, what strategy to adopt regarding proposed constitu-
tional reforms, and most of all over the allocation of portfolios.
Chavalit Yongchaiyudh, the party™s leader, was often reduced to the
role of mediator and referee in a bid to keep all the factions on board
and the party together. Unhappy with their share of the payoff after
yet another cabinet reshuf¬‚e in October of 1997, some NAP factions
began looking for greener pastures in other parties. Faced with this
threat and unable to govern effectively while sitting astride the NAP
hydra and ¬ve other coalition partners, Chavalit ¬nally stepped down
in November 1997. Over the next few years, the party steadily disin-
tegrated as faction after faction abandoned the party. What remained
of the party was ¬nally folded into the Thai Rak Thai Party in 2002.
What is instructive about the NAP case is not how unusual it is but
rather how typical. Factional con¬‚ict within the ruling party and/or
within its coalition partners was a major cause of the collapse of nearly
every democratic government prior to NAP (Chambers 2003). Recall
that these were not large national parties. Each governing party con-
trolled only a relatively modest plurality of parliamentary seats and drew
support from only one or two regions. Yet even these moderately sized,

See Chambers (2003) for details on these eight factions.
Explaining Aggregation in Thailand 121

nonnational parties were unable to manage internal con¬‚ict between
factions. Historically, Thai parties that try to grow beyond a modest
number of MPs implode in relatively short order (Chambers 2003).6 The
lesson internalized by nearly all party leaders and politicians during the
1980s and 1990s was that attempts to better coordinate across districts in
an effort to build a larger, national party would not be worth the cost to
party cohesion. Indeed, in numerous interviews with party of¬cials, the
dangers and drawbacks associated with factionalism was frequently cited
as a major reason why coordination in a bid to form large, national
parties was rarely attempted (Party Interviews 1999, 2000).
In short, Thai party leaders and would-be-prime ministers under-
stood the challenges associated with the factionalized nature of Thai
parties. Even when a prospective prime minister had the opportunity to
create a nationwide majority party through aggregation, the specter of
factional con¬‚ict undermined the appeal of such an option.
To summarize, power was vertically centralized in pre-reform
Thailand, but within the national government itself power was dispersed.
The existence of party factionalism, an upper chamber, and signi¬cant
reserve domains, especially during the 1980s, reduced the payoff to being
the largest party in government. This small aggregation payoff in turn
contributed to weak aggregation incentives and the poor cross-district
coordination that characterized pre-reform Thailand.

5.2.2 Prime Ministerial Selection Method

In parliamentary systems like Thailand™s, the method of selecting the
prime minister determines the probability that becoming the largest
party will translate into control of government. If the rules or norms of
parliament are such that the leader of the largest party always has the
¬rst opportunity to form a government and usually succeeds, then
aggregation may be worthwhile. If, on the other hand, actors other
than the leader of the largest party often form or get a chance to form
the government, then aggregation incentives are weaker.
In Thailand, the leader of the largest party did successfully head a new
government after the September 1992, 1995, and 1996 elections, but this

The threshold of sustainability seems to be between 80 and 90 members of parliament
(Chambers 2003).
Building Party Systems in Developing Democracies

was not always the norm. After the 1979, 1983, 1986, 1988, and March
1992 elections, non-elected individuals (military ¬gures) were invited to
form a government either immediately after the election, or after polit-
ical party leaders failed in their attempts. In 1988, the man invited to be
prime minister, General Prem Tinsulanonda, turned down the invitation;
and the head of Chart Thai, the largest party, became prime minister.
After a non-elected individual was again invited to form the government
after the March 1992 elections (resulting in mass protests), a constitu-
tional amendment was passed requiring that the prime minister be a
member of the House of Representatives. Prior to the amendment,
however, the high probability that the leader of the largest party would
not get the opportunity to form a government undermined the incentives
to try to create a large national party.
To summarize, the pre-1997 institutional environment generated
weak aggregation incentives. Even though power was vertically central-
ized, within the national government power was less concentrated. The
existence of party factionalism together with a Senate and the presence of
reserve domains placed checks on the power of the largest party and kept
the potential payoff to aggregation low throughout the period. In addi-
tion, for much of the pre-1997 period the selection procedure for the
prime minister was uncertain. This uncertainty together with the small
aggregation payoff undermined incentives to coordinate across districts.
Note, however, that as reserve domains diminished and a constitutional
amendment was adopted prohibiting non-elected prime ministers,
aggregation did somewhat improve. The average in¬‚ation score for the
1980s, when reserve domains were strongest and the government was
regularly headed by a non-elected premier, is 58. After the 1992 consti-
tutional amendment, the average in¬‚ation score falls to 50 “ still high as
expected given the enduring factionalism, but noticeably lower than the
previous decade.

5.3 alternative explanation: social
heterogeneity and regionalism
It is worth taking a moment here to review some alternative explanations
for Thailand™s high in¬‚ation rate. In Chapter 3, I noted that when there is
social heterogeneity and cleavage groups are concentrated in separate
regions, cross-district alliances may be dif¬cult to build (though I also
Explaining Aggregation in Thailand 123

argued that heterogeneity interacts with the size of the aggregation
payoff). In general, Thailand lacks the pronounced and politicized
social cleavages found in many of its neighbors. While its ethnic
fractionalization score is a notable .43 (Fearon 2003), ethnicity has not
been a major basis of political or party competition. In fact, “ideology,
religion, ethnicity, and policy issues have generally played a minor role in
the [Thai] electorate™s voting behavior” (Surin and McCargo 1997, 135).
To the extent a signi¬cant cleavage exists in Thailand, it would be an
urban“rural cleavage, and some scholars do claim that the urban“rural
divide hinders the development of national parties. Due to their very


. 21
( 35 .)