. 25
( 35 .)


their preferences with the drafters (The Nation 2006).29 The drafting
assembly completed its work in July of 2007, and on August 19, 2007,
the charter was adopted in a national referendum with 57.8% of
the vote.
The 2007 Constitution represents an attempt to undermine the
capacity of political parties and elected leaders to challenge Thailand™s
conservative forces in the future. In short, a major goal behind the charter
was to prevent the rise of another Thaksin “ a powerful prime minister at
the head of a relatively cohesive, nationally oriented party. The new
constitution contained a number of important reforms (including the
introduction of a partially appointed Senate and redesigned party list
tier), but my focus here is on the reforms that bear on intra-district
coordination, aggregation, and ultimately nationalization.
The new constitution retains Thailand™s mixed-member system but,
in a nod to the pre-1997 electoral system, replaced the 400 single-seat
districts with multiseat districts elected using the block vote. Most
nominal tier districts once again contain two or three seats “ only a
handful have a single seat. Voters have as many votes as there are seats in
a district and are allowed to vote for the candidates of their choosing.
The top vote-getting candidates will receive the seats in each district. The
return to multiseat districts should increase the average effective number
of parties in each district. However, even with this increase, we would
still expect the number of parties in each district to be relatively modest,
given that the number of seats in each district is capped at three.
In addition to the abandonment of single-seat districts, a major
impetus toward greater party system fragmentation and less national-
ization will likely be the undermining of the incentives for cross-district
aggregation and national party building. Speci¬cally, the new consti-
tution dramatically reduces the relative power of the prime minister.
This means future prime ministers will ¬nd it more dif¬cult to build and
maintain anything close to a large, cohesive national party. To begin
with, the constitution strips the prime minister of much of his leverage
over factions within his own party. Politicians are no longer required to
give up their House seats in order to join the cabinet, meaning the costs

It is important to note that the junta leaders appear not to have gotten all that they
wanted. Some controversial proposals for which coup leaders expressed support,
such as eliminating the party list, instituting a “crisis council,” or allowing for a non-
elected prime minister, were defeated after much debate and public criticism.
Building Party Systems in Developing Democracies

of breaking with the prime minister are lower than under the 1997
constitution. The new constitution also effectively removes barriers to
party switching. Candidates must still belong to a party at least 90 days
prior to a general election. However, an exception is now made for any
sudden or unexpected dissolution of the House. In the case of such an
early dissolution, candidates must belong to a party for only 30 days in
order to be election eligible. Since elections must be held within 45 to
60 days after a House dissolution, the effect of this change is to take
away the prime minister™s ability to threaten an early election as a way
to keep potentially promiscuous party members from jumping ship.
The 2007 Constitution reduces the size of the prize associated with the
premiership in two additional ways. First, new constitution includes
provisions to ensure that future governments cannot guarantee their
security in of¬ce by capturing a supermajority of the parliament. Thaksin
was able to achieve effective immunity from no con¬dence challenges
after winning more than 75% of the seats in the 2005 elections, leaving
the opposition with less than the 40% of the seats required by law to
launch a no con¬dence debate. The new constitution lowers the seat
requirement for launching a debate to 20% (Section 158). More funda-
mentally, the constitution also allows half of the opposition MPs to join
together to launch a censure debate against the government if the total
number of opposition MPs is less than the 20% of the total House
membership typically required (Section 160). In effect, what this means is
that, short of winning every seat, no party can ever again secure immunity
from censure debate.30
Second, the new constitution limits the potential power of the
premiership by placing a two-term limit on the of¬ce. This measure was
designed to make it impossible for Thaksin to return as prime minister
and to prevent the rise of a future Thaksin. Term limits are a rare thing in
a parliamentary context, and Thailand is one of only a few countries to
have adopted term limits for its prime minister. At most, Thai prime
ministers will be able to serve 8 years in of¬ce before being forced to step
down. In reality, most governments will likely not survive the full 4-year
term in the new institutional environment, and hence even very popular
prime ministers will likely serve less than the possible 8 years.

Section 160 cannot be invoked until the government has been in of¬ce at least 2 years.
Explaining Aggregation in Thailand 143

If the theory outlined in Chapter 2 is valid, we should expect
an increase in the number of parties and a decrease the degree of
nationalization as a result of the 2007 reforms. To begin with, the number
of parties in each district should grow modestly due to the increase in
district magnitude. However, the number of parties nationally should
increase by a greater amount due to the decrease in aggregation incentives
and resulting deterioration of cross-district coordination.
The ¬rst election under the new constitution was held on December
23, 2007. As of publication, the of¬cial district-level results from the
election were not available. However, preliminary unof¬cial district-
level results were available via various newspaper Web sites.31 The
inferences from these data should be treated as indicative rather than
authoritative, both because the results are still unof¬cial, and because
data from about 20% of the districts are incomplete. We should
also be cautious about drawing conclusions from a single election “
particularly one that follows such a major political shock as the 2006
coup was. However, to the extent the 2007 election was unusual, the
bias should work against the hypotheses of declining aggregation and
more parties. The 2007 elections took place in a highly polarized
environment with the population divided into two camps. The ¬rst
group consisted of the supporters of ousted Prime Minister Thaksin
Shinawatra, who remains very popular in Thailand. Though the Thai
Rak Thai Party was disbanded and Thaksin was barred from standing
in the election, the Palang Prachachon Party (PPP) took up the banner
of the pro-Thaksin forces and received active support (¬nancial and
otherwise) from Thaksin. The anti-Thaksin forces made up the second
camp in the 2007 race, and their primary standard bearer was the
Democrat Party, Thailand™s oldest party and the party widely viewed
as the most viable alternative to Thaksin and the PPP. Given this
highly polarized environment, it would not be surprising to observe
no measurable change in aggregation or the number of parties. The
imperatives of polarization would potentially be more than enough to
compensate for the weaker aggregation incentives generated by the
2007 Constitution.

Matichon and The Nation, two daily newspapers, both reported district-level results.
The analysis here uses data from Matichon (http://info.matichon.co.th/election/
Building Party Systems in Developing Democracies

1986 “1996 (Average) 2001 “ 2005 (Average) 2007
¬gure 5.4. Effective Number of Parties: District Versus National

It is therefore interesting to see that despite the potentially mitigating
effects of polarization, the results of the 2007 election support the
hypotheses outlined earlier. As expected, we see a modest increase in the
average effective number of parties at the district level, from 2 to 2.9 “ an
increase of 45% (see Figure 5.4). The number of parties nationally also
increases from 2.6 to 4.1 “ an increase of 58%. As expected, with weaker
aggregation incentives, there was a deterioration of cross-district coor-
dination.32 This is re¬‚ected in a 24% increase in the in¬‚ation score, from
an average 23 in 2001 and 2005 to 29 in 2007.

5.6 conclusion
In any polity, the number of political parties is shaped by the extent to
which voters and candidates are able to coordinate their behavior. Most
analyses of strategic coordination have focused on the strategic entry
decisions of parties/candidates and the strategic voting decisions of voters “
all at the district level. As a result of this work, researchers are relatively
well equipped to make sense of district level electoral outcomes. Indeed,
analysis of the district-level party systems in Thailand revealed that the
effective number of parties did vary by district magnitude prior to 1997, in
line with theoretical expectations (Chapter 4), and that a shift to single-seat
districts in 1997 caused the effective number of parties to fall (this chapter).

The in¬‚ation score does not rise to pre-1997 levels, consistent with the higher po-
larization in 2007 compared to the pre-1997 elections.
Explaining Aggregation in Thailand 145

However, theoretical tools such as Duverger™s law and the M þ 1 rule,
restricted as they are to predictions at the district level, cannot suf¬ciently
account for the effective number of parties nationally. Instead, we need
to understand when candidates have incentives to coordinate across
districts “ to link together under a shared party label.
I have argued that these aggregation incentives are a function of (a) the
payoff to being the largest party at the national level (the aggregation
payoff) and (b) the odds that the largest party will capture that payoff.
In this chapter, I used the theory developed in Chapter 2 to help explain the
lack of aggregation and large number of parties in pre-reform Thailand
and the effects of the 1997 constitutional reforms. In pre-1997 Thailand,
an appointed Senate, factionalized party system, and reserve domains
combined to limit the size of the potential aggregation payoff. The practice
of selecting someone other than the leader of the largest party as premier
for much of the period also reduced the expected utility of aggregation.
Given the weak aggregation incentives, cross-district coordination in pre-
reform Thailand was poor, and the result was a large number of parties at
the national level. The 1997 constitutional reforms magni¬ed candidates™
aggregation incentives chie¬‚y by increasing the premier™s leverage over
internal party factions. This increased the potential size of the aggregation
payoff resulting in better aggregation and fewer national parties. The
2007 Constitution once again reduced the power of the prime minister,
thereby undermining aggregation incentives and leading to a decline in
aggregation and an increase in the number of parties nationally.

5.7 appendix

The Effect of the Senate and Party List Tier on Aggregation

Two other 1997 constitutional reforms worth noting are the switch to a
fully elected Senate and the addition of a national party list tier for
House elections. As I discuss here, neither of these reforms (or the
counterreforms in 2007) generates strong, consistent predictions about
changes to aggregation incentives.

Changes to the Senate
The 1997 Constitution replaced the appointed Senate with a fully elected
body. However, the effect of this change on aggregation incentives was
Building Party Systems in Developing Democracies

negligible. First, the Senate™s formal powers did not change a great deal “
the Senate still had only delaying power.33 Second, the Senate was
now directly elected but it remained outside the direct, formal control of
Thailand™s political parties, as was the case before the reforms.34 Senators
were constitutionally prohibited from belonging to a political party and
were allowed only one term “ drafters wanted to create a legislative body
that would remain above the petty political squabbling that in their view
characterized the House (Suchit 1999). Thus, even where a party con-
trolled the House of Representatives, the Senate could potentially remain
outside of the control of the prime minister.35 Third, recall that much of
the power of the appointed Senate stemmed from its role as a reserve
domain for Thailand™s conservative forces. In short, the authority of the
old Senate was a function of whom it represented (military and bureau-
cracy) rather than its formal powers. By the time of the 1997 constitutional
reforms, the Senate had already largely ceased to operate as a reserve
domain through the appointment of nonmilitary/bureaucratic individuals
to the Senate. In short, the change to an elected Senate in 1997 may have be
an important step forward for Thai democracy, but we would not expect it
to have a substantial impact on aggregation incentives.
The changes to the Senate under the 2007 Constitution may not have
such an innocuous effect on aggregation incentives.36 The 2007 charter
replaced the fully elected Senate with a new, partially appointed body.
Speci¬cally, the new Senate consists of 150 seats: 76 of those seats are to
be ¬lled via elections in Thailand™s provinces “ one seat for each prov-
ince. The remaining 74 seats will be selected by a special committee that
will choose members from among experts and prominent ¬gures in a

The new Senate did have added responsibilities for appointing members of the new
superintendent and oversight institutions the constitution established.
The ¬rst Senate elections were held in 2000.
There are, of course, many informal means for parties and the prime minister to exert
in¬‚uence on the Senate, and in practice there were signi¬cant ties between many
Senators and political parties. For example, many Senators were relatives of promi-
nent party politicians (Nelson 2000). Thaksin was also accused of trying to bring
the Senate under his thumb through various means (though it is hard to say whether
his leverage over the Senate was a result of it becoming an elected body or rather the
advent of stable majority party government (and the end of short-lived coalition
governments.)) To the extent an elected Senate did provide for more in¬‚uence for
political parties and the prime minister on the Senate, this should have improved
aggregation incentives, which is consistent with the data.
This discussion draws on Hicken (2007c).
Explaining Aggregation in Thailand 147

variety of ¬elds. The selection committee includes the president of the
Constitutional Court, the president of the Election Commission, the
president of the Of¬ce of Auditor General, the president of the National
Counter Corruption Commission, the parliamentary ombudsman, a
judge from the Supreme Court of Justice, and a judge from the Supreme
Administrative Court.37
There is an important distinction between this new Senate and


. 25
( 35 .)