In the 2006 general election, the People’s Action Party (PAP) held 82 out of 84 seats even though they got 66.6 percent of total votes cast. This disproportionate result arises from the demographic homogeneity of electoral divisions, the result of a tight urban environment built up with cookie-cutter flats and further evened with ethnic housing quotas. Nowhere are minority interests and minority political opinions concentrated. So long as a party enjoys a comfortable margin of majority support overall, its candidates will coast to victory virtually everywhere.
This logic remains unchanged for the upcoming election.
After the recent redrawing of electoral boundaries, there will be 87 elected seats in the next parliament. Hougang and Potong Pasir, the single-member constituencies (SMC) that withstood the PAP juggernaut in 2006, remain intact. Will they return opposition members to the next parliament? Will other places turn their backs on the PAP and vote in opposition members for a change? Will a group representation constituency (GRC) fall to the opposition for the first time ever?
In every election cycle, it is at about this stage that pro-opposition optimism starts to climb. It reaches delusionary heights during the campaign itself and almost always comes down to earth with a crash the morning after polling day. Will the coming crash take the form of 86:1?
* * * * *
In 2008, I reviewed the vote shares from four elections (1991, 1997, 2001 and 2006) for the article The mathematics of elections 2. I have now refined the data a little more to give you two important numbers:
1. The standard deviation for PAP vote-share in GRCs through those four elections was 6.3 percentage points.
2. The standard deviation for PAP vote-share in SMCs was 13.2 percentage points.
For the layman, let me explain the significance of the standard deviation. In normal distribution, about two-thirds of data points in any set of data points will fall within the standard deviation. 95 percent will fall within 2 x standard deviation. It is not a hard and fast rule, there are ifs and buts, with the biggest reservation being whether election results naturally take the shape of normal distribution. However, for the purpose of this post and in the absence of any other model, we can use it as a working rule of thumb.
What it means for the upcoming election, in which we have fifteen GRCs, is this: Assuming all fifteen are contested, then about ten of them will likely yield results within plus or minus 6.3 percentage points of the mean PAP vote-share country-wide.
In about two or three GRCs, the PAP vote share will be more than six percentage points below the national mean. In other words, opposition victories in two or three GRCs become good possibilities should the overall PAP vote share fall to 56 percent (because 56 – 6 = 50%)
How likely is that? Well, I’m no pundit, but I did take a straw poll among several friends (I took the trouble to avoid asking those who were starry-eyed about the angelic qualities of opposition parties and their manifest destiny) which produced a consensus that the PAP’s vote share was more likely than not to stay above 60 percent. If they are right, we could well wake up after polling day to learn that once again, the PAP retained all GRCs.
For SMCs, the standard deviation is 13.2 percentage points. In the coming election, there are 12 SMCs, and almost surely, all of them will be contested. Two-thirds of them (i.e. eight) will likely yield results within 1 x standard deviation. I’d guess that two SMCs will see the PAP doing worse than mean minus 13.2 percent, and two SMCs will see the PAP doing better than mean plus 13.2 percent.
So, if the PAP vote share overall (i.e. the mean) is somewhere between 62 – 65 percent, perhaps two (perhaps one, perhaps three) SMCs will go to the opposition. If the PAP’s overall vote share hovers at or slightly below 60 percent, perhaps four SMCs will go over.
And that was what happened in 1991. That year, the overall PAP vote share was 61 percent. That year too, four SMCs fell to the opposition.
* * * * *
But these are very rough numbers from technical charting. It does not take into account local and personality factors.
I think most people will expect Low Thia Khiang, leader of the Workers’ Party, to defend his Hougang seat and to retain it. His support in the last election was a solid 62.74 percent.
It’s a lot more iffy in the case of Potong Pasir. Current member of parliament for this constituency, Chiam See Tong, has announced that he will not be defending this seat; instead his wife Lina Chiam will be standing for election there. In 2006, Chiam held the seat with 55.82 percent of the vote. Will all of these voters loyally give their vote to Lina Chiam this time around?
Somehow, I don’t think so. I think at least a small number will say: All right, I’ve supported Chiam long enough, at some cost to myself — delayed upgrading and all — and it’s time to look after my own selfish interests now that he’s retiring from my ward. All it takes is for one in eight of Chiam’s 2006 supporters to think this way and Lina Chiam will fail to win Potong Pasir, getting only 48.8 percent. If less than one in eight defect, she’ll squeak through.
I’m not a pundit. I’m not making any prediction for Potong Pasir or any other constituency for that matter. I only aim to point out certain realities to better temper wild optimism.
As for other SMCs, it’s too early to make any assessment. We don’t even know who is standing where.
* * * * *
The bottom line is this. Unless overall PAP vote share falls to 60 percent or less, we cannot expect significant opposition gains, least of all in GRCs. If you are like my straw poll friends and do not think such a large shift is likely, then I’d say keep the possibility firmly in mind — that you get out of bed after polling day to find headlines that say 86:1.