The best argument for changing the voting system

I am not in favour of the Lib Dems' proposals for a radical over-haul of the Commons, but I had not realised, until today, quite how uncorrelated the vote share can be with the seats share. It is not simply than the parties come in in the "correct" order but with the "wrong" number of seats: you can check for yourself that the UK Polling Report swing-o-meter will give the following figures, provided with hopelessly unrealistic figures for the national vote share:

	Con	Lab	LD
Vote (%)	31	25	37
Seats	213	215	189

The calculator assumes a uniform swing across the country, so it is safe seats which are skewing the results. And look at how skewed they are! You can see that the Lib Dems earned far and away the largest share of the vote, and Labour the smallest, yet Labour just pipped the Tories to first place in terms of seats. The Lib Dems managed to come in third, despite as I say gaining the largest share of the vote.

That is the key point about the electoral system: not that a party got 20% of the votes but only 10% of the seats, but rather that a party might get the largest vote but—of the three big parties—the smallest number of seats. It is not a matter of the electorate's will being imprecisely reflected (the first objection), but inaccurately reflected (the second).

I believe strongly in the constituency link, and proportional representation would entrench politicians who have oozed their way to the top of their party's list. Nevertheless, I can well see the case for reform, if this kind of result is possible. (Incidentally, Arrow's Impossibility Theorem proves that we cannot be perfectly accurate; but surely there must be a more accurate means of reflecting the people's will.)