A modest proposal for GRC reform

1. Decouple the political and administrative dimensions of Group Representative Constituencies

While efficiencies of scale were never put forward as a justification for the GRC scheme, it is clear that these efficiencies do exist, enabling town councils in GRCs to lower costs and build up million-dollar rainy day funds.

Thus, we propose that GRCs be kept.

We also propose that GRCs be capped at a size of 4, and for all SMCs to be scrapped entirely.

(Minilee's proposal to increase the number of SMCs and to cap GRCs to a maximum of 5 seats when the average size of GRCs is 5.4 would mean that the absolute number of minority representation in Parliament may yet again decrease)

We note that these efficiencies of scale exist independently of the political dimension of GRCs. This is a purely administrative side-effect of GRCs that will continue to exist regardless of party control.

2. Ditch the block ticket system for GRCs

Mathematically, as our GRCs stand: in a constituency of n available seats, the block ticket system ensures x(n) number of candidates where x is the number of parties contesting, but only x unique solutions, because the 'team' either wins all, or loses all.

The block ticket system of GRC voting ensures that the entire party slate wins, regardless of the unpopularity or perceived incompetence, actual gaffes and weaknesses of certain nominees. It creates a disturbingly high rate of walkovers, unelected representatives, ruling party MPs who have never faced a real contest - a situation that increasingly delegitimises the government of Singapore and its ruling party. In a country where voting is compulsory, less than half its population voted the last general elections due to walkovers. A semi-permanent swathe of citizens have never voted in their lives - a situation that increasingly delegitimises the idea of Singapore-style democracy.

As noted earlier, the sole constitutional justification for the GRC scheme is to ensure a healthy minority representation in Parliament. Nothing in the constitutional justification of GRCs says that the candidates contesting for a GRC must be bundled together.

Reforming GRCs

We propose the following set of changes that will restore the outcome of GRCs as intended by the constitution, as well as remedy the glaring weaknesses and unintended consequences of the "team wins/loses all" implementation:

For a n-seat GRC (where n is equal to or less than 4), there will be n unique winners, regardless of x parties contesting.

Each political party is free to nominate any number of candidates to the GRC.

For a n-seat GRC, each voter gets to cast their ballot for n-2 candidates on their voting forms. They are to choose exactly n-2, not more, and not less.

The winners are chosen thusly:
Top n-1 candidates with the most votes are elected, following which
From the pool of remaining candidates, elect the minority candidate with the most votes if there are none in the top n-1 candidates. Otherwise, elect the candidate with the most votes.

Implications of proposed GRC reforms:

There will be no bundling of candidates on the party ticket.

Each candidate will be judged on their own merit, preventing weak and untested candidates from riding on the coattails of Ministers. Each candidate will have to WORK FOR THEIR SEAT and prove they deserve to be elected. This is the trial of fire sorely lacking in Singapore's political system since its implementation of GRCs.

All parties are still free to present their candidates as a team; all voters are still free to vote straight-ticket ballots.

This proposal does not contradict the stated and constitutional purpose of the GRC system, i.e. ensuring minority representation in Parliament.

If the GRC system was to ensure minorities have a voice in Parliament, then the block ticket system should not be a sacred cow.

Other GRC reform proposals on the net:
Kent Ridge Common: Contesting GRCs the SMC way