Singapore Election 2011: The Redistricting
**Update: **Hi! If you enjoyed this post, I’ve got a series on Singapore’s election going up over the next week:
Election 2011: the Redistricting (this post)
…and there’s more to come.
This is the first in a series of pieces on Singapore’s General Election, which was announced yesterday after a month of phoney-war campaigning. 2.4 million registered voters will go to the polls on May 7th, and - let’s not mince words here - the result is a foregone conclusion. The People’s Action Party will win the vast majority of the 87 seats up for grabs (if you asked me a two-way I’d make you 82-84 and I’d probably get paid) - but that doesn’t mean it’ll be boring and un-news-worthy.
Elections in Singapore are traditionally foreshadowed by a couple of events. A few months before the election, the electoral rolls are opened for inspection; not long after this, the Elections Department (which, it should be noted, reports directly to the Prime Minister’s office) announces the results of redistricting, and the election is typically called within a couple of months of the redistricting. Some uncharitable souls have criticised this, saying that the short time lag between redistricting and the election itself leaves precious little time for candidates to make themselves known in the electorate.
The new “election battle-map” for 2011 was released late in February, giving the candidates a whole two-and-a-half months to get used to their new electorates. The new map, from the ST’s comprehensive elections portal, looks like this:
There are two different types of seats in Singapore: single-member constituencies (SMCs, pink on the map) are regular seats, where voters elect a single candidate; and group representative constituencies (GRCs, green, purple and orange on the map), where voters elect a slate of four, five or six candidates instead of a single candidate.
GRCs were introduced to Singapore’s electoral landscape in 1988, with the stated aim of increasing the representation of minority ethnic groups (Malay and Indian) in Parliament. Each team contesting a GRC is required to field at least one minority candidate in their lineup, and the required ethnicity depends on the seat - for example, parties contesting Chua Chu Kang GRC are specifically required to field at least one Malay member in their five-person team; in Jurong GRC, the requirement is for at least one Indian member in the team of five. (Wikipedia claims, without citation, that minority representation in Singapore’s parliament has actually decreased since GRCs were introduced.)
Update: a sharp-eyed reader has done the numbers on this claim, and it appears to be busted: the number of minority MPs - from these lists of former MPs - remained pretty consistent around 20-25% from 1984 (before GRCs) through 1997.
They work slightly differently from the mixed-member proportional systems that exist in Germany, New Zealand, Scotland and Wales: in an MMP system, a 60-40 vote in a five-representative district would lead to three representatives in one party and two from the other; in Singapore, that vote count would deliver five representatives from the winning party and none from the losing party (which would make the GRC system more of a “mixed-member _dis_proportional”). GRCs are winner-take-all, making it easier for a party to dominate the resulting parliament without necessarily dominating the vote.
In fact, not all seats are always contested, partly because running for office in Singapore is an expensive exercise. The fee to get on the ballot paper in Singapore is pretty steep - this year it’s $16,000 per person, so running a full slate of candidates in a six-member GRC like Ang Mo Kio will cost a competing party $96,000. This occasionally leads to situations where seats go uncontested for years - this “first-time voter” who wrote to Today Online’s letters page has been eligible to vote for thirteen years, but has never had anyone other than the PAP contesting his district.
Another interesting feature of Singapore’s electoral map is the wildly varying number of voters in each district, which leads to wildly varying voter power between electorates (and a seeming disregard for the principle of “one man, one vote”).
As a control point, Australia’s electoral commission has extremely tight rules on the variability allowed between electorates; they target a maximum of +/- 10% variation at the time of the redistribution, and +/- 3.5% three-and-a-half years afterward.
Singapore’s boundaries review committee targeted “a range of 20,000 to 36,000 voters per MP, with a 30 per cent variation as in past practice”, but still managed to miss the side of that particular barn. For example, the opposition stronghold of Potong Pasir SMC has 17,000 voters electing one MP; right next door, Marine Parade GRC has nine times as many voters (154,000), but only elects five MPs - in effect, one vote in Potong Pasir has almost as much power as two votes in Marine Parade.
Once the committee handed down its recommendations, the government (which has the final say on redistributions) accepted the report and issued the new electoral map back on the 24th of February; seven weeks later, here we are, counting down the sixteen days to polling day.