On that memorable election night in May 1997, what made people rub their eyes in disbelief was not simply the size of Labour's majority, but where it came from. The drumbeat of Labour gains was a roll-call of the suburban and small-town Home Counties: Braintree, Hastings, Harrow, Hendon, Hemel Hempstead, Hove, Putney, St Albans, Enfield Southgate - Southgate! - Watford. Only Surrey stayed solidly blue. The rest of John Betjeman country turned red.
Famous Labour victories were not always made this way. In 1964, when Harold Wilson ended 13 years of Conservative government, Labour's share of the British vote (44.8 per cent) was almost identical to that in 1997 (44.3 per cent). But in 1964, the electorate was polarised by class. Labour built its victory on 68 per cent of the working-class vote but a mere 19 per cent of the salariat vote. By 1997, Labour's share of the working-class vote had fallen slightly to 64 per cent, but its share of the salariat vote - 36 per cent - was almost double. Among all white-collar/blouse workers, Labour won 22 per cent of the vote in 1964, but 39 per cent in 1997.
Normally, the fluctuation in the Labour vote from one election to the next is pretty uniform across the social spectrum: the popular notion of a critical swing group (Worcester Woman, Essex Man and so on) is a myth. But in 1997, the middle classes switched to Labour in much larger numbers than the rest of the electorate. The 1997 British Election Study found that the Labour vote increased by 14 percentage points among all respondents to the survey, but by 18 points among routine non-manual workers, 19 points among the salariat and a huge 24 points among the self-employed.
Why were the professionals, managers and small-business people won over in such unprecedented numbers? Social commentators pointed to the new insecurities afflicting the middle classes - "delayering", takeovers, asset-stripping, computerisation, negative equity - and to their resentments at the telephone-number salaries in the City and utilities boardrooms. However, the verdict of the 1997 British Election Study is that the triumph of the Blairite project is at least as good an explanation. Among the middle classes who voted Conservative in 1992, the economically confident defected in the same proportions as the economically insecure. But in 1987, Labour was still seen as a sectional party, concerned above all with looking after the interests of trade unions. By 1997, 85 per cent saw new Labour as a party that looked after the interests of the middle classes "very" or "fairly" closely - well up on the 59 per cent of 1987, and fractionally more than said the same of the Conservatives. Unlike the Conservatives, it was also credited - by 93 per cent - with looking after working-class interests. It had become the classless, one-nation party.
Labour appears to have consolidated its middle-class support since 1997. MORI's aggregated polls for the second quarter of 2000 reported a further 3.5 per cent swing to Labour since the general election. This was almost entirely due to a surge of middle-class support: the swing was 6.5 per cent among the ABs (professionals and managers) and 5 per cent among the C1s (routine non-manual workers), but only 1 per cent among manual workers and welfare dependants. These figures have been borne out in by-elections since 1997. In the four middle-class Con-Lab seats (Uxbridge, Beckenham, Eddisbury, and Kensington and Chelsea), the Labour vote fell by a mere 1.1 per cent. In the four by-elections in working-class Lab-Con seats (Leeds Central, Wigan, Tottenham, Preston), it haemorrhaged by 15.3 per cent.
All the evidence points to Labour increasing its middle-class support on 7 June - an extraordinary achievement after a full term in office. According to MORI, the narrow 39-34 per cent Conservative lead among ABC1 voters in 1997 will be converted to a 45-36 per cent Labour lead. It would be the first time that the British middle classes had preferred the historic party of the left.
How has this remarkable transformation come about? Tony Blair appears to suit the professionals and managers particularly well. In 1997, they only narrowly preferred him - by 42 to 40 per cent - to John Major as a prime minister. This time, they prefer him to William Hague by 59 to 17 per cent - a larger lead than in any other social group. Blair's impeccably middle-class credentials (Fettes, Oxford and the Bar) can hardly be the explanation, as Hague went to Oxford, while Major comes from Brixton and left school at 16.
More important, almost certainly, is Labour's handling of the economy and its slaying of the three middle-class fears: rampant inflation, penal income tax rates and over-mighty unions. Black Wednesday of September 1992, when overnight the Conservatives lost a 40-year reputation for economic competence, was a turning point. Ever since, Labour has assumed the mantle of reliable stewardship of the economy. Ever since, the middle classes have voted Labour in unprecedented numbers.
Barring accidents, two records will be set on 7 June. The majority of the middle classes will vote Labour rather than Conservative. Even more significantly, the majority of Labour voters will be middle class. In 1964, about 20 per cent of the Labour vote was non-manual. In 2001, it will be about 53-54 per cent, mainly because of changes to Britain's class structure, but partly due to Blair's modernisation of the party.
In the 1960s, Labour MPs became middle class as teachers outnumbered the miners; in the 1980s, the party membership became middle class, as the polytechriat took over from the proletariat. In 2001, the embourgeoisement of the party will be complete. The balance of interests and values it represents will have tilted decisively to the median-income, owner-occupying, tuition-fee-paying, private-pension-contributing, ISA-holding, qualification-bearing knowledge and service workers of Middle England.
Critics on the left argue that Labour will lose its traditional and still vital working-class vote, now that it has become the managers' party. But caretaker, secretary and manager share a common interest in high-quality, free or reasonably priced public services and a stable, low-inflation, low-interest, high-employment economy. What strains this social coalition is not economic interest, but moral and cultural outlook - on minority rights, race, crime and punishment and, perhaps above all, Europe.
So long as a Labour government can deliver public services and a growing economy, the Conservatives - not Labour - face strategic dilemmas. They can offer an authoritarian/nationalist alternative take on the cultural and moral agenda: anti-Europe, anti-gay, anti-black (suitably coded), pro-jail. This might win over some working-class Labour supporters, but at the price of consolidating the middle-class Labour vote and dividing the Conservative Party. Or they can offer a free-market alternative on the social and economic agenda - low tax, privatisation of public services, less welfare. But this is unlikely to detach Labour supporters in any class, so long as the economy remains strong. A third option is simply to wait for the economy to go into long recession and for the government gradually to exhaust itself. None looks promising.
From the 1960s to 1980s, social scientists wrote books with titles like Must Labour Lose?, pointing out that, unless Labour modernised its organisation and ideology, it faced electoral decline as its social base eroded. Who would have thought that similar books will be published about the Conservative Party, even though its social base has expanded?
Ivor Crewe is vice-chancellor and professor of government at the University of Essex