Thailand Under PR
By Allen Hicken
Picture: Bangkok old parliament
Source: Wikipedia Commons
This post continues our series on the possible effects of alternative electoral systems on Thai politics. The first two posts looked at the possible effects of adopting a one province, one seat electoral system and the effects of adopting an Alternative Vote system. Today we look at the effects of a proposal to move Thailand in a more proportional direction by adopting pure proportional representation.
How would we expect electoral results to change were Thailand to adopt a system of pure proportional representation (PR) instead of the mixed-member system Thailand currently employs? The answer to this question depends in part on the size of the electoral constituencies and the specific method for translating votes into seats. I look at three different types of PR constituencies below: a single national constituency, regional constituencies, and provincial constituencies. For the seat allocation method I assume that Thailand would continue to use the D’Hondt method, which is a highest averages method for allocating seats.
How might have the 2011 election results been different if Thailand had used a purely PR system of choosing MPs? The easiest scenario to imagine is the case of a single national constituency—equivalent to what was used to elect the list tier in 2011. If we assume that the 2011 party list results represent how voters would have behaved if they were only voting for a party list under a pure PR system then we would expect the PR election results to mirror those from the party list election. As you can see in Table 1 below, Pheu Thai suffers under this scenario, losing its majority, though it remains the largest party in parliament. The Democrat Party benefits, moving from 31.8 percent of seats to 35.2 percent. Smaller parties on balance do slightly better as well.
Although single national constituencies are used in some PR systems, the use of smaller subnational constituencies is more common. Two of the most likely ways to divide Thailand into subnational PR constituencies would be to use either regionally-based or provincially-based constituencies. In the simulation below I assume that each province/region would be allocated a number of seats in proportion to its population—equivalent to the proportion of seats each province/region elected in the 2011 constituency election.
Let’s look first at the simulated PR results using four regional constituencies. If we divide Thailand into four regional constituencies based on the 2011 seat allocation the Central Region (including Bangkok) would elect 127 seats, the North would elect 69 seats, the Northeast 126 seats, and the South 53 seats. (Alternatively, we could use five regions by making Bangkok a separate region, with 33 seats, leaving the Central Region with 94 seats). To simulate each party’s vote share for each region I used the total number party list votes each party received in 2011, aggregated to the regional level. I then used the D’Hondt method to allocate the seats within each regional constituency. To simplify the calculations I used a vote threshold of 1 percent—meaning any party receiving less than one percent of the party list votes was not included in the seat allocation calculation. The results of the regional PR simulation can be seen in the third column of Table 1. Under this scenario Pheu Thai loses some ground, but retains its majority, while the Democrat Party gains seat share compared to the 2011 results. Table 2 shows the breakdown of party seat shares in each regional constituency. Not surprisingly, the Democrats dominate the South, Pheu Thai reigns supreme in the Northeast and North, and the two parties split support in the Central Region. The results are similar if we use 5 regions with Bangkok as a separate region. The Democrats have the edge in Bangkok and Pheu Thai captures the most votes in the Central Region (Table 3).
Seat Shares Compared under PR (2011)
M=Constituency magnitude, or the number of seats in a constituency
Party Seat Share by Regional PR Constituency (2011)
Another possible way to use PR would be to turn each province into a separate PR constituency. If we assume that each of Thailand’s 77 provinces would elect the same number of seats it elected in 2011 then the constituency magnitude (the number of seats in a constituency) would range from a high of 33 seats in Bangkok, to one seat each in Trat, Phang Nga, Mae Hong Son, Ranong, Samut Songkhram and Singburi. The average magnitude across all constituencies would be 4.87 seats. Under this scenario both Pheu Thai and the Democrats would improve their seat share over what they received in 2011 (see column 5 in Table 1). The Democrat Party seat share increases by 9 percentage points, while the seat share for Pheu Thai grows by more than 2 percentage points.
The big losers under the provincial scenario are the minor parties. This is precisely what we would expect. How favorable PR is to minor parties very much depends on the number of seats in the constituency. At the extreme, a PR constituency with only a single seat functions in practice just like a first-past-the-post contest—the party with the most votes gets the seat. As we increase the number of seats available, however, the chance of smaller parties winning a seat increases. In Figure 1 we can see how smaller parties fare compared to the two major parties under our different PR scenarios. Not surprisingly small parties capture the most seats under the current system or under national PR, where magnitude equals 125 seats. As we move to Regional PR and then Provincial PR the average magnitude decreases and the fortunes of smaller parties suffer as a result.
So, what is the bottom line? As with the other electoral systems we have examined thus far, adopting PR in 2011 would likely not have greatly altered the electoral results, though under a National PR system Pheu Thai would have been short of a parliamentary majority. In the next post we will examine the effects replacing the current mixed-member majoritarian system (MMM) with a German-style mixed-member proportional system (MMP).
 The D’Hondt is quota=V/(S+1) where V is the number of votes a party receives and S is the number of seats a party has already been awarded.