NOTE: BP is still on hiatus. The recent posts about MMP and the electoral system are an exception. The posts are designed to explain MMP and its proposed implementation. The posts do not endorse any electoral system or represent the views of anyone else.
We have been looking at the proposals for Thailand’s new 6-region party list. In part one we discussed the high degree of malapportionment in the distribution of seats across the 6 regions. In part two we showed that the primary motivation behind the malapportionment does not appear to be giving one of the two largest parties an advantage over the other.
Again, a couple of caveats. Neither of us have any first-hand knowledge of the drafter’s and/or the EC’s motivations. We are merely inferring intent from the nature of the proposal—and it is certainly possible that we are wrong. It is also possible that the current proposal will be modified before becoming codified.
So, what could be behind the distribution of seats in the current party list plan? A common goal expressed by the current authorities is increasing the power and importance of small and medium sized parties, while reducing the power of the two largest parties (see here and here for some examples). The current proposal appears to be consistent with that goal. Table 1 below shows the party list vote shares and the total seat shares for small and medium-sized parties in the malapportioned 5 regions. The numbers come from the 2011 election. It is striking that the average share of party list votes for smaller parties is 61 percent higher in the regions that are over-represented (13% in in under-represented regions vs 21% in over-represented regions). The disparity in seat shares is even starker. In under-represented regions smaller parties managed to capture only 6 percent of the seats, while in the over-represented regions they captured nearly a third of the total seats. Figure 1 presents another view of the same data. Both Table 1 and Figure 1 suggest a clear pattern: party list seats have been distributed in such a way as to try to give more weight to regions that in the past have supported small and medium-sized parties.
If indeed one of the motivations behind the distribution of seats across regions is to boost the fortunes of smaller parties, would we expect this effort to be successful? Our analysis using the 2011 election results suggests the malapportioned electoral system will not actually benefit the smaller parties. To simulate the effect of these proposed rules we re-calculate the 2011 election results using the proposed changes. Specifically:
- We award seats using the German-style mixed-member proportional system rather than the mixed-member majority system Thailand has traditionally used.
- We reduce the total number of seats from 500 to 455— 250 constituency seats and 205 party list seats, split across 6 regions, as described here.
- We incorporate overhang seats by increasing the size of the parliament by one seat for every overhang seat. (For more on overhang seats see here.)
- Since there is some debate over whether the party list will use a .5% threshold or no threshold at all, we run the simulation for both.
The results are summarized in Figure 2 below.
The first set of columns shows the actual results from the 2011 election, where Pheu Thai won 53% of the seats, to the Democrats 31.8%. Small and medium-sized parties collectively captured 15.2% of the seats.
The next two sets of columns show what the results would be under the proposed rules, using first no threshold (second set of columns) and then a .5% threshold (third set of columns). Despite all of the changes to the electoral system, and the high degree of malapportionment in the distribution of the party list seats, small and medium-sized parties gain no ground on the two largest parties. Their share of seats remains the same, regardless of which threshold is used.
Ironically, small and medium-sized parties would do a bit better under a system where each region received a proportional share of the party list seats (fourth set of columns). For example, supposed that we made the number of party list seats in each region proportional to the region’s population (36 seats for the Upper Central region, 33 seats for the North etc.). Small and medium-sized parties actually capture a few more seats under this proportional allocation. Figure 3 shows why this is the case. Notice that Pheu Thai and the Democrats win exactly the same number of seats (225 vs 166 respectively) regardless of whether there is a high degree of malapportionment in the distribution of seats or no malapportiontment. However, the small and medium-sized parties do a bit better. With more seats available in the previously under-represented North and Upper Northeast regions, four smaller parties are able to pick up extra seats: Chart Pattana Pheu Paendin (2 seats), Social Action Party (1), and Bhum Jai Thai (1). Fewer seats in the Lower Northeast and Upper Central regions translates into a loss of two seats for Rak Prathet Thai (partially offset by an additional seat for the party in the North). After the proportional adjustment in the Lower Northeast and Upper Central regions Bhum Jai Thai and Chart Thai Pattana are also entitled to fewer seats. However, since the seats each party won in those regions were constituency seats, they are allowed to keep those seats as overhangs and the number of seats in parliament adjusts upwards from 461 to 464.
All in all, the results are consistent with our earlier simulations of the effects of MMP. Small and medium-sized parties do not do much better under MMP, while the biggest effect of MMP is to reduce the gap between the largest party (Pheu Thai) and the second place party (Democrats).
Why don’t small and medium-sized parties fare better in our simulations even though at face value the electoral system seems to be stacked in their favor? There are two answers to this question. First, adjusting the distribution of party list seats actually does very little to help smaller parties because most small parties don’t win their seats via the party list. MMP is designed to compensate parties that get a lot of party list votes but few constituency seats. Since small and medium-sized parties, such as Chart Thai Pattana and Bhum Jai Thai, have traditionally won their seats in constituency elections, and done poorly on the party list, then MMP, even with a high degree of malapportionment in the distribution of seats, brings little benefit. In fact, the reduction of constituency seats from 375 to 250 likely offsets any benefits that MMP brings to smaller parties. Second, electoral systems set a ceiling on the number of parties that can be supported, but they do not set a floor. More permissive systems, like MMP, have higher ceilings and provide opportunities for more parties to win seats, but voters do not have to take that opportunity. Indeed, as long as most Thai voters continue to separate into two competing electoral blocks, as they have since 2001, the prospects for smaller parties will continue to be dim, regardless of what electoral system Thailand chooses.
NOTE: “2011-proposed” refers to a simulation using the 2011 election results with the proposed malapportioned distribution of party list seats as outlined in previous posts.
“2011-No Mallapportionment” refers to a simulation using the 2011 election results where we have adjusted the party list seats so they are in proportion to the population.