Question: How many ways are there to configure the Group Representation Constituency (GRC) system?
Answer: Several hundred permutations, at least.
You could have most of the 87 seats as Single-Member Constituencies (SMCs) or go the whole hog with 14 six-member GRCs. Between these two extremes there are any number of combinations of SMCs and three-, four-, five- or six-member GRCs.
In the real world, though, the possibilities are more limited, but still Singapore has seen several combinations being tried since the scheme was introduced in 1988.
That year, it started with 42 SMCs and 13 three-member GRCs.
In 1991, the SMC numbers were halved to 21 and the size of GRCs increased to include four-member wards. Further changes took place in 1997, when six-member GRCs were created, with more tweaks made subsequently to reduce the average size of GRCs.
Now, with last Monday's announcement by the Prime Minister, more tinkering will take place before the impending general election.
Is there an ideal solution that will stop the experimenting?
An electoral system shouldn't be changed so frequently, and its underlying rationale should stand the test of both time and scrutiny.
It certainly shouldn't change at every election.
Looking back at the scheme's controversial past, there seems to be as many arguments put up to back the case for smaller or bigger GRCs, depending on what was being proposed.
The one argument which remains unchanged and has stayed true to the original purpose of the scheme: to allow minority candidates to stand a better chance of being elected as members of a GRC. This is a worthwhile aim and is the only one worth preserving.
What of the other considerations? Take your pick of any number of arguments for bigger or smaller GRCs.
You prefer bigger ones? When five- and six-member GRCs were proposed, the argument was that they would lead to more efficient town councils which could benefit from having economies of scale.
You could extend this logic if you wanted to make the case for ever larger GRCs.
What about this other argument: Bigger GRCs would make it easier for the ruling party to recruit candidates.
In today's less forgiving political environment, you will not hear it being repeated. But it was made as recently as 2006, when then Senior Minister Goh Chok Tong said: "Without some assurance of a good chance of winning at least their first election, many able and successful young Singaporeans may not risk their careers to join politics."
On the opposing side, critics had long complained that big GRCs make it even harder for the Opposition without the resources to put together sizeable teams.
Their opponents were fond of saying they complained too much.
And so the debate raged.
What about the case for smaller GRCs? Here's a fairly recent one: to help the voter identify better with the ward's MP.
Indeed, Prime Minister Lee Hsien Loong put it this way in 2009: "But, at the same time, there are some downsides to having too big a GRC because it becomes harder for voters to identify with the whole GRC or with the whole GRC team. The members split up, so the voter knows who is looking after his area, but for the other MPs, he may not have quite as close a relationship.
"Each MP has to look after his own ward in the GRC and, therefore, it is not easy for him or her to get to know the voters in all of the other wards."
Perhaps you are one of those who believe that bigger or smaller really does not depend on any of these arguments but on how the People's Action Party (PAP) feels about whether it improved its chances of winning.
In the old days, it was clearly the case that the big GRCs were seen as natural stomping grounds of the ruling party. When helmed by heavyweight ministers, they offered almost certain victory to fresh recruits headed for bigger things in government.
In many cases, victory was secured without a single ballot being cast as the opposition failed to show up.
And when they did, the polling results reinforced the belief that bigger was better for the PAP.
In the 2006 GE, the PAP's share of the votes in contested GRCs was 67 per cent compared to 61 per cent in SMCs.
But the 2011 GE turned this on its head with the PAP's vote share in GRCs (60.3 per cent) only very marginally higher than in SMCs (59.3 per cent).
More painful was its loss of two ministers and a minister of state when it was defeated by the Workers' Party (WP) in Aljunied, a big GRC.
Did this signal the end of its invincibility in these large wards?
It's a possibility that cannot be ruled out, given the desire for a stronger opposition presence, especially among younger voters.
In fact, as in 2011, one can expect the WP to concentrate its firepower on one or even two or three GRCs.
It can do this either by moving its brand name leaders to these wards, as Mr Low Thia Khiang did in 2011, or introducing strong candidates that impress voters, such as the likes of Mr Chen Show Mao.
Because it is the underdog, these moves will create a bigger impact than if the same type of candidates were to appear for the ruling party.
The WP does not need too many of these, just enough to contest in a few GRCs. By thus focusing its star candidates on a few big GRCs, the WP might gain more than if the fight was dispersed among smaller GRCs or SMCs.
Whether this will indeed happen in the coming GE, no one knows until the contest gets under way.
The point is that bigger GRCs no longer confer the same certain advantage of before on the ruling party.
So, however you analyse this issue, whether from the perspective of which configuration makes better sense, or in realpolitik terms, on what size best suits which party, the answers are not clear-cut.
The truth of the matter is that tinkering with different GRC sizes isn't a productive exercise for any party any more.
The only consideration worth retaining is that of ensuring adequate minority representation, and the best way to do this is to revert to how it was first conceived in 1988, with a combination of SMCs and three-member GRCs.
Making all GRCs one size is good for limiting the number of changes that can be made and ensuring that every voter in a GRC faces the same choice of choosing a three-member team. Adjacent GRCs can be combined to form town councils, thus retaining the advantage of size present in existing large GRCs.
With three MPs in every GRC, and with at least one of them from a minority race, the number of GRCs will be determined, not arbitrarily, but by deciding the required minimum number of minority MPs in Parliament.
Hence, if it is thought Singapore should have at least a quarter of them, or 22 in a House of 87 elected representatives, that would mean 22 three-member GRCs and 21 SMCs. This way, the number of GRCs is precisely determined, and not subject to extraneous considerations.
Such a transparent system has a better chance of enduring.
This article was first published on July 19 2015.
Get a copy of The Straits Times or go to straitstimes.com for more stories.