It is risky to forecast the result of the general election with the contest perhaps more than one year away. Unlike 1992, the economy will probably not be in recession and the recent past should warn us against discounting unpredictable, "random shock" events which can rapidly transform the allegiances of a volatile electorate. However, evidence we have compiled makes clear how easy it will be for the opposition parties to unseat the Government.

The election will be fought on new boundaries, designed to equalise the number of electors between seats, in more than 400 constit-uencies. "Notional" results for these new constituencies suggest that although the Conservatives would have emerged from the last election with a larger majority ( instead of 21), a swing against them of less than one per cent will still be enough to strip away their overall majority. In other words, unless they perform in 1996/97 as well as they did in 1992, they will be condemned to a period in opposition. They have no cushion against electoral unpopularity.

It is more difficult for Labour to win an overall majority than for the Conservatives to be unseated. However, despite the Conservatives enjoying a huge 7.6 per cent lead over Labour in votes cast in 1992, the balance of strength within individual constituencies means Labour may win an overall majority at the next election with just a one per cent lead in the popular vote. Assuming a roughly equal swing across the different types of constituency, Labour could have a majority of three, even if they polled just 40 per cent of the vote to the Conservatives 39 per cent. By contrast, the Conservatives will need a lead over Labour of at least 4 per cent and maybe as much as 8 per cent to cling on to office.

These figures put into context the gap between Labour and the Conservatives. The Conservatives need to do more than peg back Labour's lead; they must overtake Labour and get some way ahead. Our model, extrapolating a national equivalent share of the vote from the results of local government by-elections, suggests that Labour has a more modest lead than polls imply.

ÖйúAƬ

The by-elections model has successfully projected the result of the annual local elections to within 2 per cent of the support gained by the three major parties for the last three years. It remains untested at a general election. That event is too far ahead for convincing extrapolation.

However, Labour has consistently polled 5 per cent better in local elections in this parliament than it did in the previous one. The Conservatives are 7-8 per cent worse off. Even if the gap between the parties continues to close as it did last time, Labour may poll 40 per cent of the vote at the general election and be some 2-3 per cent ahead of the Conservatives. A much narrower win than looks likely, but enough to give Tony Blair an overall majority well into double figures.

ÖйúAƬ

Colin Rallings and Michael Thrasher, Local Government Chronicle Elections Centre, University of Plymouth