Long-time readers of Yawning Bread will know that I have been arguing for a mixed model for Singapore’s parliament, since at least ten years ago. If my idea had been in effect, what would the 2011 parliament look like? Would it better reflect the people’s will?
Using the vote counts from this general election, I have estimated the composition that would result. Opposition members would make up about 24 percent of the new parliament, far improved from the current 81:6 ratio (or 6.9 percent), or 81:9 ratio (10 percent) if one adds in non-constituency members of parliament.
Briefly, to recapitulate my model:
I have long proposed that Singapore’s parliament should be comprised of members elected via two routes: (a) first-past-the-post from single-member constituencies (SMC) and (b) an equal number through proportional representation. The total number of members of parliament (MP) is not essential to fix, but in this exercise, for convenience I assume it is 87 x 2 = 174. That is, there are 87 SMC MPs and 87 MPs elected by proportional representation.
Singapore desperately needs some form of proportional representation. This is because as a single city-state with ethnic quotas for public housing (in which 85 percent of the population lives), there is considerable homogeneity across constituencies. A minority is a minority everywhere; it has no precinct of concentration. As a result, any election using a first-past-the-post system will tend to produce super-majorities in parliament from relatively small majorities vote-wise.
Look at the actual results of the 7 May 2011 poll. The People’s Action Party (PAP) got 60.1 percent of ballots cast but swept 93.1 percent of elected seats.
Many people would consider the result and the system that produced it, inherently unfair. I do.
One solution would be to go for a pure proportional representation system, one where the PAP would have gotten only 60.1 percent of seats for 60.1 percent of the votes that it obtained. A valid criticism of pure proportional representation systems is that it tends to produce hung parliaments, i.e. parliaments with no party having a majority, though this is not as terrible as it sounds. Coalition governments have been known to be effective. In any case, even if a party has a simple majority and forms the government with no demurral from other parties, it doesn’t guarantee smooth sailing either. Parties can contain factions and factional disputes can be more bitter and protracted than any dispute between coalition partners.
So I’ve long said, let’s have a blend of the two. 50 percent of the seats from SMCs and 50 percent from proportional representation.
I can well anticipate comments to this article suggesting all sorts of variants, but let me say this: Electoral systems need to be simple enough for laymen to understand. The more complicated it is, the more opaque the electoral process will appear to the average Joe or Jill, and the less legitimacy it may have in their eyes. So keep it simple.
You would notice that in my system, there are no group representation constituencies, there are no non-constituency members of parliament. And definitely no nominated members of parliament, which to me is a real bastardisation of democracy if ever there was one.
* * * * *
OK, so we have 87 SMC seats and 87 proportional representation seats. Based on the results from the poll earlier this month, how would they have been allocated? Let’s take the SMC seats first.
Hougang SMC went to the Workers’ Party (WP) on 7 May 2011. Naturally, one can assume it would go to the WP if my model had been in effect. Aljunied GRC also went to the WP, but in my system, GRCs would no longer exist. Instead, Aljunied would be five SMCs corresponding to the five precincts of Aljunied GRC. On 7 May 2011, each of these five precincts gave a majority to the WP, based on what I’ve heard from WP insiders. In other words, if they had been separate SMCs, the WP would have won all five.
One of the precincts voted almost 60 percent in favour of the WP. The last precinct however, was a close shave, just a tad over 50 percent. In other words, the WP’s vote-share of 54.7 percent in Aljunied GRC was an average of precincts with vote-shares ranging from about 50 to about 60 percent, a ten-percentage point spread.
This is roughly consistent with what I was told about East Coast GRC, the GRC with the next-highest opposition vote share. The WP gained 45.2 percent of valid votes cast. East Coast GRC also has five precincts, of which WP scored a majority in one., roughly consistent with the expected top end of the ten-percentage point spread. The People’s Action Party (PAP) retained a majority in the other four precincts. So if East Coast GRC had instead been five SMCs, one would have gone to the WP and four to the PAP.
The GRC with the third-highest vote for an opposition party was Marine Parade GRC with 43.4 percent for the National Solidarity Party (NSP). I do not know the results by precinct, but applying the estimated ten-percentage point spread, even its best-performing precinct would only be 48.4 percent. Thus I would not be giving the NSP any seats even if all five precincts of Marine Parade GRC had been SMCs.
To summarise, if there had been 87 SMCs, the WP would have won 7 and the PAP 80. See the second column from left.
Proportional representation seats
To allocate the proportional representation seats, we need first to look at total vote-shares. 60.1 percent gave their vote to the PAP, so I will give this party 60.1 percent of 87 proportional representation seats, i.e. 52 seats.
The remaining 35 seats would go to opposition parties. In what proportion?
One’s first instinct might be to apply the relative vote-shares obtained in the recent election, but to do so would be flawed. This is because which opposition party a voter voted for was not really determined by the voter. His choice was constrained by the fact that, except in Pungol East SMC, he had only one opposition party standing in his ward. Hence, the vote shares obtained by respective opposition parties were largely determined by how many constituencies they put candidates into.
In a proportional representation poll, a voter is presented with a complete list of all the parties and he has a free choice of party. Therefore, a better indicator of voting outcome would be the internet poll I did, and which is described in the article Talk at the Post Museum: election perspectives. One of the questions in that survey was: If you could choose among all these parties, which would you consider your first choice? The results:
Based on these percentages, I would allocate 25 seats to WP, 9 seats to SDP and 1 seat to the Singapore People’s Party (SPP). The exact calculation is shown in the table above.
In total therefore, parliament per my electoral system would comprise 132 PAP MPs, 32 from WP, 9 from SDP and 1 from SPP. The seating plan in the Chamber would look something like this — which is a lot more “normal”, with a clear opposition block facing the ministers’ bench.
* * * * *
The Workers’ Party, in its campaign, asked voters to demonstrate with their ballot their desire for a First World Parliament. Despite scoring an average of 46.7 percent in the areas it contested, the party earned a mere six seats. For the opposition as a whole, despite scoring an average of 39.9 percent of all valid votes cast, they still ended up with the same six seats.
We can’t get a First World Parliament because the system is designed against one. Reform, in my view, must include reforming the electoral system, and I hope I have demonstrated how simple it is to produce an electoral outcome that is miles fairer and more representative than what we actually got.