Tribune leader, 19 June 1992
The London School of Economics’ simulation of what would have happened in the general election if the electoral system had been different, conducted by the ICM polling organisation and published last week, is of course only a rough guide to the way that Britain would actually have voted if the first-past-the-post system had been replaced by the additional member system, the alternative vote or the single transferable vote.
Apart from any considerations of margins of error in opinion polls, even those which survey nearly 10,000 people, voters would almost certainly have behaved differently in a real-life election under a new system than they did when asked by the LSE’s pollsters to play a game of “what if?”.
Nevertheless, the LSE survey, commissioned in the expectation that there would be a hung parliament and that electoral reform would be at the top of the political agenda, is the best guide we have to the effects of electoral reform for Westminster elections. Its findings are directly relevant to Labour’s debates on electoral reform and, more surprisingly, on relations with the Liberal Democrats.
As far as electoral reform is concerned, the survey suggests that, of the two options for change currently being given serious consideration by Labour’s Plant Commission on electoral systems, the alternative vote and the additional member system, only the latter fully corrects the pro-Tory bias inherent in first past the post.
According to the survey, if the April 9 general election had taken place using AV, which retains single-member constituencies but requires voters to rank candidates in order of preference, the Tories would have emerged with a share of seats much larger than their share of the vote, a couple of seats short of an overall majority. Labour would have taken a seat less than it did under first-past-the-post.
By contrast, under AMS, in which MPs from single-member constituencies are “topped up” with MPs from regional lists, seats gained by all parties would have been close to proportional to votes cast. The Tories would have got 268 seats (down 68), Labour 232 (down 39) and the Liberal Democrats 116 (up 96), with Plaid Cymru and the Scottish National Party taking 18 between them (up 11).
These conclusions should reinforce the already strong case for Labour to reject the idea of changing to AV for elections to the House of Commons. The LSE survey shows that AV has no significant advantages over first past the post and all the disadvantages of non-proportionality.
The Plant Commission should now explicitly rule out AV just as it has effectively ruled out the single transferable vote, which is favoured by the Liberal Democrats and would do away with single-member constituencies.
The next step should be a strong recommendation of AMS, the only model to tackle the problem of proportionality at the same time as keeping single-member constituencies, in good time for a decision at 1993 party conference.
But the survey should also make Labour banish any notion that it should enter into an electoral pact with the Liberal Democrats before the next election. The reason that the Tories would have done so well under AV is that more Liberal Democrat voters would choose a Tory as their second preference vote than would choose a Labour candidate. This means that an electoral pact in which the Liberal Democrats stood down in Tory-held Labour target seats would do Labour no good at all, and would possibly save several Tories’ skins.
The only anti-Tory electoral pact that might work would be a unilateral decision by Labour not to stand in certain Tory-held Liberal Democrat target seats, but the political costs of such a gift to the centre, in terms of internal strife and Labour’s credibility as a national party, make such generosity distinctly unappealing.