They believed that it was crucial to shake the electorate's perception of where Labour stood on the main issues, even if this involved leapfrogging right across the centre, and adopting centre-right positions on some key economic issues. The resulting party and policy revolution of 1994-96 was implemented with ruthless efficiency.
Four years, and one election landslide, later, two key questions need to be asked about this strategy. First, did it actually work in attracting new votes for the Labour Party from previously virgin territory? Second, to what extent did it cost votes among disaffected core Labour supporters?
A superficial look at the results of the last election would certainly suggest that the Labour revolution succeeded in attracting many new votes in previously hostile territory. A non-superficial look - based on post- election studies written by John Curtice, Michael Steed, Anthony Heath, Peter Kellner and Pippa Norris among others - suggests exactly the same thing. In fact, only rarely can an electoral strategy have hit its intended target so precisely.
First, there is no question that voters noticed extremely clearly that the nature of the Labour Party had changed. Data from the British Election Panel Study enable us to compare Labour's position on several key issues - such as tax and spend, and nationalisation - with that of the electorate as a whole. When we do this, we find that the electorate perceived Labour to be much nearer the centre, or even the right, in 1997 than they had in 1992. In particular, 57 per cent of the population believed they were to the right of Labour on tax and spend in 1992, whereas only 38 per cent believed they were to Labour's right in 1997.
Furthermore, the swing to Labour from 1992-97 was much greater among centre-right voters than among the left. Among those who considered themselves to Labour's left in 1992, the swing to Labour in 1997 was only 4 per cent. Among those who believed themselves in 1992 to be to Labour's right, the swing was around 10 per cent.
In addition, there is plenty of evidence that Labour's pledge to freeze income tax rates had a big electoral pay-off. In 1992, Labour trailed the Tories by 22 per cent as the party most trusted to take the right decisions on income tax, whereas by 1997 Labour actually led the Conservatives by 8 points.
Income tax seems to have been especially important among those people who decided to desert the Conservatives. In this group of voters, Labour led the Tories by 18 points on the tax issue, more than twice its lead among the rest of the electorate.
For whatever reason, Labour did not only attract centre-right votes in unprecedented numbers in 1997, but it did so in precisely the right geographical spots to win seats. The swing to Labour was 12-13 per cent in London and the South-east, but only 7 per cent in Wales and Scotland, and around 9-11 per cent in the Midlands and North.
Not only that, but Labour was now perceived to be sufficiently close to the Liberal Democrats to enable widespread tactical voting to take place. In Conservative seats where Labour was second in 1992, Labour's share of the vote went up by 13 per cent, while the Liberal Democrats' share declined by 3 per cent. The opposite pattern applied in seats where the Liberal Democrats were second.
These variations in swing for tactical reasons greatly increased the size of Labour's majority. According to Pippa Norris, a uniform national swing would have produced 46 more Conservative seats, 28 of which would have been subtracted from Labour and 18 from the Liberal Democrats. Of these extra seats, roughly half may have fallen directly to the phenomenon of tactical voting, while the rest may have been influenced by variations in regional swing.
Wherever we look, it is the same story of Labour making massive gains in previously untouched areas. The swing to Labour among mortgage-payers was 18 per cent, while that among council tenants was 2 per cent. The swing among graduates was a massive 19 per cent, and among share owners an even more remarkable 28 per cent. The swing among skilled non-manual workers was 25 per cent, while that among unskilled manual workers was only 2 per cent. To some extent, these phenomenal differences are no doubt explained by the fact that Labour was already dominant among its core support in 1992, but surely something much more significant was afoot.
The second question is whether core supporters were significantly turned off by Labour's shift to the centre. Here the verdict must be mixed. Although the swings to Labour were much smaller among its core supporters than in the rest of the population, Labour continued to enjoy huge absolute leads over the Conservatives in these core groups. So there was no major sign of erosion in the shares of votes actually cast.
However, there we clearly some signs of left-wing disaffection in the figures for turnout. On average, turnout in Labour seats was only 68 per cent, compared to 74 per cent in Conservative seats, the widest gap ever. It is possible that this can be partly explained by the fact that Labour's large opinion poll leads led to apathy in its safe seats, but we do not find that the drop in turnout was related at all to the closeness of the local race.
By contrast, it is clear that turnout fell most in seats with a high concentration of working class votes. In addition, the "loyalty factor"among working class voters - ie the propensity of working people to prefer Labour, over and above that which is explained by its ideological stance on the major issues - was significantly eroded in 1997. The working class continued to vote Labour in dominant numbers, but not to any greater extent than would have been expected from their views on the major issues. This is one indication that the special ties between Labour and its core supporters may have partially broken down.
However, there is no question that Labour made massive electoral gains from this trade-off between core and non-core support. The shift to the centre led to disproportionate gains in Labour votes in the most effective geographical areas, and permitted a much greater degree of co-operative or tactical voting with the Liberal Democrats than ever before. By contrast, any decline that there may have been in turnout among core Labour supporters probably cost very few seats.
So New Labour's electoral strategy definitely worked. But to what extent is the Labour Party in government now hoist with its own petard? Is it prevented from introducing a radical programme of policy reform, presumably involving a significant degree of income redistribution, by the fear of alienating the new supporters on which its electoral landslide was based?
This is a complicated question, since there is plenty of evidence that some forms of tax and spend - "competent tax and competent spend" - are still very attractive to the electorate. I will return to this topic next week.Reuse content