There is an air of unreality about the Labour Party's deliberations on electoral reform, which looks set to be the most controversial policy topic at the party's annual conference in Brighton next week.
Everyone knows that Labour's prime task is to win the next election under the existing first-past-the-post system; and everyone knows that it is unlikely that supporters of change will convince conference to back anything more specific, at least for the House of Commons, than keeping all options open, except for ruling out the multi-member constituency, single transferable vote system favoured by the Liberal Democrats.
That, in a nutshell, is the position taken in July by the interim report of the Labour Party's commission on electoral reform chaired by Raymond Plant, and it is the only one that will preserve party unity. This does not mean, however, that Labour Party members need not bother to work out what they think. Keeping options open is a temporary measure: soon after the election, however Britain votes, Labour will be forced to make up its mind.
Plant laid down a series of criteria for judging electoral systems, the most important of which were whether a system would yield parliaments in which the number of seats held by a party is proportional to the votes cast for it, whether a system would maintain a close link between MPs and their constituents, and whether a system would produce stable and effective governments.
In the past couple of months, three options have emerged with significant support: no change, the Alternative Vote, and the Additional Member System. On balance, the system that satisfies most of Plant's criteria is AMS, in which MPs elected in single-member constituencies are topped up from regional lists. Crucially, neither no change nor AV, in which single-member constituencies are maintained but voters list their preferences among candidates, adequately satisfy the criterion of proportionality.
Critics of AMS argue that it would hand disproportionate influence to centre or fringe parties, would create two classes of MP and would allow party machines too much influence in determining candidates for the regional top-up seats.
The last of these objections is the least serious: a democratic regional party selection process would ensure that parly members and not party bureaucrats would determine the party list. The "two classes of MP" argument is stronger, though hardly decisive: it would be easy enough to organise a system whereby regional list MPs were allocated constituency-type duties in areas of their region where no single-member constituency MP had been elected.
The most important argument against AMS is that it gives Centre or fringe parties too much influence, inevitably yielding coalition governments which emerge after secret wheeling and dealing behind closed doors in the immediate aftermath of an election. Against this, it is rightly said that political parties are themselves coalitions, that no system, even FPTP, rules out coalition government, and that AMS would not rule out creation of a single party forming a government if it really had majority support among the electorate.
Nevertheless, the issue of disproportionate centre or fringe party influence - which also applies to AV - cannot be ducked. The choice in the end is between a system in which the centre or the political fringe might have greater influence than their support warrants and one which has given the Tories 12 years of massive parliamentary majorities on a minority vote. On balance, the former has to be the lesser evil.