Must Labour lose? The question was so widely asked after the party's third consecutive general election defeat in 1959 that it became the title of a well-known work of political analysis.

Natural, then, that it was asked again, not least by Labour's own leadership, following a fourth consecutive loss in 1992. Some critics said British democracy was taking on a Japanese tinge - a de facto one-party state with the Conservatives in the role of Japan's seemingly eternal democrats. Japan's Liberal Democratic Party shortly afterwards contrived to lose power.

Such is the fate of instant analysis. Now the questionis "can Labour lose?" asTony Blair piles up massive poll leads. Canada - where the ruling Conservatives were reduced to two seats at the last general election - has become the fashionable international comparison.

But does anybody in the know really believe this? We asked 32 political and electoral experts three questions about the next general election: 1. Who will win and by what margin? 2. How many seats will the Liberal Democrats win? and 3. What, if anything is likely to upset this prediction? Polling took place between March 26 and April 9 - after the BSE scandal broke and before last week's by-election.

É«ºÐÖ±²¥

Forecasting elections is a thankless task and all predictions come with provisos. Eric Shaw of Stirling University recalled"Shaw's first and only law of politics - the unexpected always happens." Ben Pimlott, professor of history at Birkbeck College, warned: "Even the most confident prediction must be prefaced with the word 'probably'". As it happensDavid Butler of Nuffield College, Oxford, did not offer his famous dictum that: "Experts are no more likely to make accurate predictions, they just make wrong ones for more sophisticated reasons," - but it always applies.

The clear subtext to question one was: "Do you believe the polls?" Up to a point, yes. Of 30 replies, 29 said Labour was likely to be the largest party but only two predicted a majority of more than 50.

É«ºÐÖ±²¥

John Benyon, professor at Leicester University, tipping a Labour majority of 143, said:"I really think the floodgates have opened. It could be a landslide," while adding warily: "I thought Labour wouldwin in 1992."

Some 17 of the 29 Labour backers predicted a majority of 25 or less - with five believing that a hung Parliament was likely; in effect a rerun of Labour's narrow victories of 1964 and February 1974. Patrick Dunleavy, professorat the LSE said: "It is desperately difficult to get the Conservatives below about38 per cent. I suspect theories about their supporters being reluctant to declare themselves to pollsters."

Several respondents were extremely sceptical about the real depth of Labour's appeal: "Still much more anti-Tory than positively pro-Labour," suggested Neil Carter of York University. Some undoubtedly reflected on memories of 1992. Dominic Wring, of Nottingham Trent University, worked for the BBC during the last election and remembers being told by an election unit and computer programmer that the Conservatives would have a majority of 20, at a time when expert opinion pointed the other way. Andrew Taylor of Huddersfield, the one backer for the Conservatives, said: "I can easily see them returning as the largest single party," noting that it still might lead to a non-Conservative government.

The Liberals can feel reasonably encouraged by responses to question 2. Of the 28 answers, 15 expected them to win 30 or more seats and only 7 predicted a decline from their current 23. But nobody was tipping a decisive breakthrough - 40 to 45 was the highest prediction - from Justine Fisher of London Guildhall University who said a lot of disillusioned Conservatives "won't go all the way to Labour". Several expect the Liberals to gain in the Southwest but lose seats in Scotland.

É«ºÐÖ±²¥

And the minority who predicted a decline visualise the Liberals caught in a classic two-party squeeze. Richard Topf, of London Guildhall University said: "We are seeing a real polarisation. I think they could get hammered."

Given the preponderance of Labour predictions, the subtext of question 3 was "what can the Conservatives do to get out of their hole?"

Several noted that wars are generally good news for right-wing governments - one respondent nominated France as the ideal opponent. Others suggested a change of leadership, with several hinting that John Major's last service to his party might be to attract a sympathy vote by falling under the hypothetical London bus. One thought that David Owen joining the shadow cabinet might do the trick.

Tax cuts, promises of economic revival and smear tactics having prevailed in 1992, some respondents suggested they might do so again in 1997. But others were more sceptical: "Distrust is now so deep it's hard to see them being believed whatever they say," was the tenor of several replies. Others questioned whether Labour's historically unparalleled capacity for self destruction really was a thing of the past.

É«ºÐÖ±²¥

Most pointed to the traditional scenario of the Government recovering fast in the closing months of a Parliament. But that assumes it will get to the closing months. Few doubt that John Major will try to hang on until next year, but defections or a run of byelection defeats might rob him of the prime minister's last card, the ability to decide when the election is held. "The earlier he has to go, the better for Labour," said several respondents. Canada syndrome may yet have its day.