NOTE: BP is still on hiatus. The technical post(s) about the electoral system are an exception. The post(s) are designed to explain the proposed electoral system and its implementation. The posts do not endorse any electoral system or represent the views of anyone else.
On 29 January, the second group of constitutional drafters appointed by Thailand’s NCPO released the draft of another new Constitution. Like its predecessor the current draft proposes to change the way Thailand elects its representatives. Recall that the electoral system Thailand used in 2011 was a mixed member majoritarian system (MMM) with 375 single-seat constituencies, and 125 party list seats, elected from a single national constituency using proportional representation. The previous constitutional draft (rejected by the National Reform Council last September) called for a switch to a German-style mixed member proportional system (MMP). The new draft proposes yet another type of mixed member system—what is being called a mixed member apportionment system (MMA).
Under the proposed MMA system, Thailand would have 350 constituency seats and 150 party list seats (Section 78). Instead of voters casting two separate votes, one for a candidate and one for a party list, under MMA voters will cast a single, fused ballot for a candidate (Section 80). That vote will count as both a vote for the candidate, and simultaneously a vote for that candidate’s party for purposes of the party list seats. The total number of votes a party receives nationwide via this single vote will determine the total share of seats a party is entitled to (Section 86). Party list seats will be added to a party’s constituency seats until this total is reached (Section 86). In this way, MMA is more proportional than the 2011 MMM system which simply added the party list seats to the constituency seats already won by a party.
The Effects of MMA
What would the likely effects of MMA on Thailand’s election results be? To determine this we ran a simulation using 2011 election results. The biggest assumption we need to make in order to conduct this simulation is that voters would behave the same way under MMA as they did under MMM. In other words, the simulation assumes that voters would treat their vote as primarily a constituency vote, i.e. they would cast the same constituency vote under the new rules as they did under the old rules.
Losers and Winners
Table 1 compares the number of seats each party captures under MMA with the number of seats they won in 2011.  Figure 1 compares each party’s seat share in 2011 with the seat share under MMA and under the previously proposed MMP.
As Table 1 and Figure 1 indicate, Pheu Thai is the clear loser under MMA. While Pheu Thai remains the largest party, they go from winning 265 seats (53%) to winning only 225 seats (45%). The new system, MMA, is even worse for Pheu Thai than the previously proposed MMP. This is because Pheu Thai (and its predecessor, Palang Prachachon) always captured a higher percentage of party list votes than they did constituency votes, as Table 2 shows. In effect, the separate party list vote gave Pheu Thai a seat bonus. This disappears under our simulation.
The other big loser under the proposed system is small parties who don’t have the resources to compete in a large number of constituencies. Regardless of Chuwit’s jailing, Rak Prathet Thai would be in trouble under MMA because they were competing for the party list vote while not simultaneously competing in constituencies. Under the previously proposed MMP they would have been one of the big winners (nearly 3% of the seats), but they get no seats under MMA (see Figure 1). In order to compete under MMA, parties like Rak Prathet Thai will have to adjust their strategies and begin competing for the constituency vote—ideally fielding candidates for all seats throughout the country. However, such a campaign strategy requires resources—resources that may be beyond most small parties. The option to compete solely on the party list, allowed under MMP and the 2011 electoral system, would disappear under MMA. Thus the irony of a system that has purportedly been designed to help smaller parties: in theory, a proportional electoral system with no threshold would be good for small parties, but no separate party list vote will make it difficult for new small parties with no local base to compete.
The biggest winners under the new system are the medium-sized parties who can compete in a large number of constituencies nationwide, but who have been unable to capture a large share of party list votes in the past. These include Bhum Jai Thai, Chart Pattana Pheu Paendin, and to a lesser extent, Chart Thai Pattana, who pick up an additional 22, 13, and 5 seats respectively under MMA. Notably, all three of these medium sized parties would do poorly under the previously proposed MMP than under both MMA and MMM. Again, the reason is their basis of support. These parties tend to be competitive in a handful of constituencies, but do not command much of a national following. They are therefore at a competitive disadvantage under MMP, and to a lesser extent, MMM. On other hand, MMA, by calculating the party list seats from the constituency votes, works to these parties’ great advantage.
What of the Democrat Party?
The Democrats could be considered partial losers or partial winners depending on how we look at it. On the surface, it appears that the Democrats receive essentially the same number of seats under MMA compared with the 2011 election results. However, they could be considered “winners” because Pheu Thai does worse under MMA, and loses its majority. This makes it more likely that the Democrats would be part of or even lead a coalition government
Nevertheless, the Democrats seem to consider themselves partial losers. For example, see the concerns raised by Abhisit in comments to the Bangkok Post. Under almost any proportional system that uses the party list vote to determine the overall percentage of seats a party wins, the Democrats would do better than under the 2011 MMM system. This includes simple proportional representation or the previously proposed MMP. However, the MMA system chosen by the drafters is one of the few proportional systems where the Democrats actually don’t do better. The reason is the Democrats, like Pheu Thai, always do better on the party list vote compared with the constituency vote. As Table 2 above shows, the Democrats did 10.2% better in 2007 and 2.8% better in 2011. The Democrats don’t profit from this competitive advantage under MMA, and as it result it undermines the advantages of moving to a more proportional system.
To summarize, like the previous draft, the new electoral system in the draft constitution, MMA, reduces the electoral fortunes of Pheu Thai. MMA also boosts the prospects of medium-sized political parties with strong constituency bases (something the previously proposed MMP failed to do), while limiting the gains for the Democrats and small parties that compete mainly through the party list.
Recall that we assume in all of this that voters will treat their vote as primarily a constituency vote, i.e. they would cast the same constituency vote under the new rules as they did under the old rules. But there are good reasons to believe that with that one vote voters would be more likely to think party first (given the current level of polarization and given how we expect parties to campaign) and that would come at the expense of small and medium-sized local parties. Voters, in other words, would no longer have the luxury of voting for their local favorite AND picking a side in the larger political conflict. There are some indications that voters in other places that use a single ballot do tend think of their vote as primarily a party vote (for example, Mexico). As a result we expect the seat share for small and medium-sized parties is probably overstated (provided the two big parties don’t collapse).
 Our understanding from the wording in Section 86 of the draft constitution is that the allocation seats is calculated using the Hare quota. We make the following additional assumptions. 1) The 2011 election had 375 constituency seats while the new system reduces this to 350 seats. We assume that in the reducing the seats from 375 to 350 the proportion of seats across parties remains the same. 2) Section 86 does not specify how unassigned seats are too be allocated. We used the common Largest Remainder method. 3) Section 86 was also not clear about what do in the case of overhang seats (of which there were 2 from Palang Chon). Since the number of seats is fixed at 500 we took the seats away from the two smallest parties, but the basic results don’t change if we take them away from the two biggest parties. 4) Since the draft mentions no electoral threshold we assume the threshold is 0.
 For other analysis that reaches conclusions different than ours see http://www.tnews.co.th/html/contents/178013/