The difficulty for Labour advocates of the Alternative Vote (such as Peter Mandelson and Peter Hain) is to show how this system can be reconciled with Labour's manifesto pledge to offer a proportional system. In our recent report, "Making Votes Count" (published this week by Scarman Trust Enterprises) we show that this system would be far less proportional than first-past-the-post elections - for instance, cutting Tory MPs at the 1997 election from 165 to just 110, half the number they would be entitled to under a proportional system.
We also showed that the Liberal Democrats' favourite system (the single transferable vote) would have worked in an unproportional way in 1997, so that Paddy Ashdown also confronts a problem in reconciling his preference with the promises already made to voters.
By contrast the approach which Donald Macintyre sees as a "possible cloudy basis for compromise" is a variant of a third system called the Additional Member System, which does operate in a reliably proportional way, so long as it is designed right. It involves electing some MPs locally, and some regionally through party top-up lists.
With a 50:50 split between the two types of MPs, or slightly more local MPs, as in the Scottish Parliament, this method works very proportionately.
But the more local MPs there are (and hence the fewer top-up MPs to correct distortions) the more dis- proportional this system would become.
These results, established by the most rigorous political science methods, are not just "voting systems nerdery" (as Donald Macintyre terms it). They go to the heart of the problems which both Labour and Liberal Democrats confront in the future. You cannot inaugurate a new age of better politics by betraying firm manifesto promises on the choice that citizens will get in the referendum on voting reform.
Patrick Dunleavy (LSE), Helen Margetts (Birkbeck College), Stuart Weir (University of Essex)
Department of Government