How would PR have changed the face of Parliament?: Labour's Plant inquiry into electoral reform was finalised last night. Alex Renton looks at ways of voting, and how they could have changed history (CORRECTED)

CORRECTION (PUBLISHED 22 APRIL 1993) INCORPORATED INTO THIS ARTICLE

How the survey was carried out

ICM, funded by the Joseph Rowntree Reform Trust, interviewed 9,600 people immediately after the general election. Results were analysed by the Public Policy Group at the London School of Economics. Respondents were given samples of ballot papers, as below, and asked 'to act as if you were voting in the general election again'. They were not questioned on Supplementary Vote.

The SV figures on this page are calculations by Dale Campbell-Savours, MP for Workington, based on the 1992 election results. Northern Ireland was not included in the survey.

THE supplementary vote system of proportional representation - the one the Plant commission is to suggest the Labour Party backs when the party's National Executive Committee considers it on 19 May - is a new idea, but hardly a radical one. In fact - is it PR at all?

No, according to Simon Osborne of the Electoral Reform Society: 'It is not proportional to anything - any proportionality you'd get from it in practice would be by luck or fluke.'

A truly proportional system is one that gives a party the same proportion of seats as votes it gets. The least proportional system is the current one, first past the post - its 'unfairness' most glaringly indicated in the south- east of England.

In the 1992 general election the Conservatives received 55 per cent of the votes in the South-east and took 97 per cent of the seats. Labour had three seats for 21 per cent of the vote and the Liberal Democrats no seats at all, in spite of winning 23 per cent of the votes. In the north of England Labour got 81 per cent of the seats for 51 per cent of the vote.

But supplementary vote would, even on the most optimistic projections, give the Liberal Democrats only 35 per cent of the seats they would gain under a strictly proportional system. 'It's simply an amendment to first past the post - as little change as can be potentially stomached by the Parliamentary Labour Party,' Mr Osborne said.

Of course, that is just what the system's inventors, the MP Dale Campbell-Savours and Patrick Dunleavy, Professor of Politics at the LSE, wanted it to be. The attraction of SV - under which voters choose their first and second choice of candidates (the system is explained in full below) - lies in its simplicity and the fact that it is unlikely to redraw the electoral map in any grand way. Since it is the system that offers least change, it is most likely to give majority governments.

And that means it is least likely to do any substantial harm to the Labour Party - either in the debate during the coming months, or in the early years of the next century, when - or if - it is in place.

Philip Norton, Professor of Government at Hull University, said it would give disproportionate weight to people's second choices. 'You're in effect giving the balance of power to the lowest common denominator - allowing the tail to wag the dog.'

For Labour, the downside is, of course, that SV is unlikely to do them much good. Mr Campbell-Savours's own projected figures below show Labour making not gaining at all.

But winning Labour seats was perhaps never the point. Professor Dunleavy said: 'SV is an engine for tactical voting.' Under the similar alternative vote system, where people mark all the candidates in order of preference, there is a risk of the left vote fragmenting. A Liberal Democrat will put the Greens second rather than Labour. But SV, which only offers the chance to make a first and second choice, allows the Labour voter, say, to record his or her allegiance with the first vote and then make a genuinely potent choice - for the candidate who will oust the Tory - with the second. 'If you take the most optimistic scenario - that every Tory or Labour voter who was not going to give his second choice to the other major party, gave it to the Liberal Democrats, they would gain 48 or 50 seats,' Professor Dunleavy said.

That is a scenario Labour could sell to the Liberal Democrats at the next election - to persuade them to vote tactically in the hope of gaining, at worst, a hung Parliament. After that, a Lib Dem-Labour coalition would hammer out a proportional system that would please both of them. The Liberal Democrats are currently wedded to the single transferable vote system, which gives results much closer to true proportionality.

'The astounding thing is that John Smith has not seen the potential of the system,' Professor Dunleavy said. But one thing that will hold many Labour party members back is the implicit admission in Plant that Labour's future is irretrievably bound up with the Liberal Democrats.

Justice, it has often been said, is the only defensible argument in favour of PR. So, in the eyes of the Electoral Reform society, the Plant Commission has made an error in not choosing a system, like single transferable vote, that can be defended on moral grounds. Simon Osborne said: 'At the end Labour have yet to decide they like electoral reform because it's a good idea. The argument is still over whether it is good or bad for us.'

Replaying the 1992 General Election: How Britain would have voted under alternative electoral systems; by Patrick Dunleavy, Helen Margetts and Stuart Weir; available from Department of Government, LSE, Houghton Street, London WC2A 2AE; pounds 5.

----------------------------------------------------------------- GENERAL ELECTION RESULTS 1992: ----------------------------------------------------------------- First Past The Post ----------------------------------------------------------------- Con Lab LibDem SNP/PC MPs 336 271 20 7 Proportion of vote 45.5% 34.4% 17.8% n/a Proportion of seats 51.6% 41.6% 3.1% 1.1% Under a system giving the same proportion of seats as votes the result would have been: Con Lab LibDem SNP/PC Green/Other MPs 273 222 114 18 7 ----------------------------------------------------------------- NB: Northern Ireland, with 17 seats, is not included in the survey. -----------------------------------------------------------------

(Photograph omitted)