Electoral reformers looking for a hero to celebrate should steer away from the Liberal Democrats, despite their enthusiasm for the cause. They should also discard Enid Lakeman, for many years the redoubtable leader of the Electoral Reform Society. Instead they should alight on AP Herbert.
“AP who?" you may ask. Older NS readers may recall him as a genial but sharp-minded satirist and legal gadfly. For 15 years he was also an independent MP, one of the few ever to be elected under the single transferable vote (STV) - see the key to voting systems on page 6. He first won his seat, to represent Oxford University, on the third count in 1935. Among his achievements was persuading parliament to reform our antiquated divorce laws. So, one thing first-past-the-post advocates cannot say is that proportional representation is an alien concept, only recently imported into British politics to elect people to the Scottish, Welsh, London, Northern Ireland or European parliaments and assemblies.
How huge a step would it be to change the way we elect MPs? What would the effects be? The answer comes in two parts: it depends on the system we choose and on the way parties and voters then react to it. Its final effect may be rather different from its initial impact.
Imagine what might have happened had the last general election been fought under different voting systems. The Electoral Reform Society attempted this exercise; its main projections are summarised in the table on page 6. Under our current system, Labour enjoyed a majority of 66, despite winning only 36 per cent of the national vote. Under the alternative vote (AV), its majority would have been even higher. This is because most Liberal Democrats preferred a Labour to a Conservative government; their second-preference votes would have helped Labour retain a handful of seats that, under first-past-the-post (FPTP), it lost to the Tories. The Lib Dems would also have benefited by capturing more Tory seats with the aid of the second preferences of Labour supporters. The Tories would have been the big losers, not because they were Tories per se, but because they were generally unpopular. The same process would have punished Labour in 1983 and 1987 had those elections been fought under AV.
The other, more proportional, systems would have produced hung parliaments - with the additional member system (AMS), the most proportional of all, leaving Labour furthest from the winning post. The Lib Dems would have won more than twice as many seats as they achieved under FPTP. Under STV or AMS, although not under the alternative vote-plus (AV+), the arithmetic would have allowed the Lib Dems to choose whether to combine with either the Tories or Labour to form a two-party coalition government with a secure majority.
Whether or not that represents an improvement or a deterioration in our democratic arrangements depends on whether you regard the main function of general elections to encourage cross-party consensus or to give voters a clear choice of government. As a technical note, elections fought under a regional list system - the one we use to elect our MEPs - would produce much the same outcome as AMS.
There is no panacea
Ukip, the BNP and the Green Party would stand the best chance of winning seats under AMS or regional lists. They would have got nowhere in 2005 under AV or AV+; the Green Party might have won just a single seat under STV. The Electoral Reform Society projects that, under AMS, the three parties would have won 15 seats altogether.
All that said, such counter-factual exercises do more to illuminate the different properties of the various systems than to provide us with a reliable guide as to what would happen in reality if we changed our system for electing MPs. It's rather like changing the offside rule in soccer, or the LBW laws in cricket. The players would change the way they played the game. In politics, the "players" would be voters as well as parties, and it would not be just a one-off change: adaptation would continue as we came to understand and operate the new system.
This can be seen from the European Parliament elections we have held. In 1994, the last to be fought under first-past-the-the-post, Labour, Conservatives and Liberal Democrats together won 89 per cent of the votes cast in Britain; minor parties such as the Greens, BNP and Ukip, won a total of 11 per cent. In 1999, we switched to a regional list system, which helps smaller parties. The minor-party vote doubled to 23 per cent and Ukip and the Greens won their first seats. The really significant thing is that this process has continued. In 2004, the total minor-party vote rose to 36 per cent, and this year to 43 per cent, with Ukip coming second in the popular vote and the BNP winning its first seats.
Something akin to this process would be likely to happen if we changed the way we elect our MPs. AV would continue to punish minor parties. However, it is probable that Ukip and the Greens would win seats at some point under AV+, and so might the BNP, eventually.
As for STV, my guess is that the Greens and Ukip would eventually win seats (as the second-choice parties of some Lib Dem and Tory voters respectively), although the BNP might find it harder. Incidentally, for those who wish to keep the BNP out of council chambers as well as out of the House of Commons, the best advice is to campaign for AV for local elections. When BNP candidates win 30 per cent in a local four-cornered contest under FPTP, they often win the seat, but they would be unlikely to gain the extra second-preference votes they would need to reach the 50 per cent winning post under AV.
In the much longer run - say, 20 years or more - the effect of changing the voting system might be even more profound, as existing parties change, merge or split, and as new parties grow. Would this be more democratic, more dangerous, or both? Would it give us a parliament that is better or one that is worse? Would that parliament be more or less able to take tough decisions?
I make only one certain prediction: it will offer no panacea, for it will have both good and bad effects, and the net impact will depend on both what actually happens and on the different, subjective criteria by which each of us judges the way we are governed.
Peter Kellner is president of YouGov