Derek da Cunha, an independent scholar in Singapore, suggested before that the ‘law of large numbers‘ in the GRC system makes it difficult for the opposition to win a GRC. Well, Aljunied GRC has been lost to the WP. Does the law of large numbers still stands?
According to da Cunha, the more voters there are in a GRC, the results will tend to reflect nation-wide results. So for GE 2011, the PAP has a popular vote share of 60.14 percent. From the table above, you can see 10/14 GRCs have a deviation of +5 or -5 from the PAP popular vote share. Only West Coast, Jurong and Ang Mo Kio have more than +5 or -5 deviation.
It’s easy to see why: West Coast and Ang Mo Kio were challenged by the newcomer Reform Party, resulting in above-average performances by the PAP. Jurong was contested by an unknown and weak NSP team.
From the table, you can also see out of the five largest GRCs, only Ang Mo Kio deviated more than 5 percent from the PAP popular vote share. Even for the five smallest GRCs, only East Coast deviated slightly, at -5.31.
Conclusion: GRCs have a tendency to reflect overall vote share. As long as overall support of the PAP remains above 50 percent, it is difficult to dislodge the incumbents. What happened in Aljunied was a combination of chance, strategic choices made by the WP, errors committed by the PAP and a general sense of discontentment. I’m not too sure if the WP or any other party can replicate this victory in a couple of GRCs in the next elections, if overall support for the PAP remains good.
Take a look at the SMCs:
In SMCs, the biggest ward is Punggol East – 33 281, 18 percent the size of the largest GRC and 38 percent the size of the smallest GRC. The smallest ward is Potong Pasir, with only 17 327 voters. Only three SMCs kept within the +5 or -5 deviation limit which I set for the PAP popular vote share of 60.14 percent. The deviation range can run from -24.95 to 10.47, with the results of most SMCs deviating from the popular vote share.
SMCs, unlike GRCs, have a propensity to deviate from the popular vote share. This is because of the smaller number of voters in each SMC, and the law of large numbers does not hold here anymore. From a statistical point of view, it is easier for the incumbent to be knocked out (or hold onto power) in a SMC than in a GRC, because of the much smaller size of the SMC. Hence Potong Pasir and Joo Chiat were won by tight margins, while candidates for Hong Kah North, Radin Mas and Yuhua scored huge margins (they were also facing weaker opponents). As SMCs do not necessarily reflect the popular vote share, if George Yeo, Lim Hwee Hua and Zainul Abidin had stood in SMCs, they might have won easily, instead of squaring off against the WP’s dream team in Aljunied.
Conclusion: SMCs might be easier to win (or lose), but with creation and destruction of SMCs in every elections, it is difficult for opposition candidates to work the ground and stand in a particular SMC. But the first-past-the-post system must also be considered, even winning by a 0.01 margin is also considered a win. SMCs tend to have such close margins, while GRCs do not. Hence the fear of such close fights might have led the PAP to enlarging the size of GRCs over the decades, while reducing the number of SMCs to a mere 12/87.
In all, while the WP victory in Aljunied inflicted a psychological blow to the PAP, it was an exceptional one. Whether it can win in multiple GRCs throughout Singapore to control 1/3 of Parliament or even form the next government is greatly dependent on the national mood. If support for the PAP falls tremendously, in all likelihood GRCs will reflect this loss of support easily, as compared to SMCs, and many GRCs will be captured by the opposition. The PAP has won big through GRCs – and it might lose big through GRCs too. The first-past-the-post system and the GRC system have so far favoured the PAP, but if one day a perfect storm converges, it would be the biggest loser.