The local and European elections do not suggest Labour is on course to win the general election, argues Peter Kellner

There is a case for Labour optimism. It goes like this: Labour could lag the Tories by three points in the popular vote and still be the largest party. No governing party has ever added to its vote share at the end of a full parliament, so the Tories are certain to fall from 37 per cent. And, as most people feel that the economic recovery is not helping them, an effective Labour campaign on the ‘cost-of-living crisis’ can force the Tory vote share down – especially as the United Kingdom Independence party provides a non-toxic alternative for right-of-centre voters. Just as the Social Democratic party crucified Labour in the 1980s by splitting the anti-Tory vote, so Ukip could be the agent of David Cameron’s defeat.

Meanwhile, so the optimists say, do not worry about voters preferring the prime minister to Ed Miliband – Margaret Thatcher was less popular than James Callaghan when she won the 1979 election. And do not fret either about Labour lagging the Tories on the economy. We were in the same position in 1997, yet still won a landslide victory.

That is the argument. Does it really hold up? Better to know if it does not and to take action than to pretend that all is well and to take none.

Let us start with the recent election results. According to Colin Rallings and Michael Thrasher of Plymouth University, Labour secured a one per cent lead over the Conservatives in the local elections. This is slightly less than the three per cent lead that Labour enjoyed in 1991 – another occasion when a Conservative prime minister (then John Major) was a year away from a general election. The Tories went on to win one year later with a seven per cent lead.

On that precedent, Labour’s prospects for next year look bad. Indeed, they could be worse than they were 23 years ago. Then the Conservatives monopolised the right-of-centre vote, while the Liberal Democrats ‘borrowed’ Labour votes for local elections. If we adjust for the votes that would have gone to Labour had the 1991 elections been for parliament, the party’s underlying lead was around eight per cent. These days, however, it is Ukip that is borrowing Conservative votes; few Labour supporters back Nick Clegg’s party in local elections. Labour should be miles in the lead, not a paltry one per cent ahead.

The lesson from the European elections is no more cheerful. For the third consecutive time, they are being held one year before a general election. In 2004, the opposition Conservatives enjoyed a four per cent lead over Labour (27 to 23 per cent) – yet Labour went on to win the 2005 general election. In 2009, the gap was 12 points (with the Tories on 28 per cent and Labour on just 16 per cent). A year later, the gap narrowed to seven points and the Tories failed to win an overall majority. These figures are confirmed by opinion poll data over five decades: the final year of a four- or five-year parliament usually sees a swing back to the government of the day.

The really worrying thing for Labour this year is not just that it came second to Ukip but that it secured only 25 per cent of the vote, and led the Conservatives by less than two per cent, just as it did in the local elections. Around two million people voted Conservative in 2010 and Ukip in this year’s European election. YouGov’s eve-of-election poll found that two-thirds of them said they would vote Tory in a general election if they lived in a Conservative-Labour marginal.

If history is any guide, then, Labour has a tough job if it is to win the popular vote next May. But does it have to do so in order to win the largest number of seats, and for Miliband to head a Labour-Liberal Democrat coalition government? That is true on the normal assumptions about uniform swing. Suppose the share of votes next May is Labour 34 per cent, Conservatives 37 per cent, Liberal Democrats 13 per cent, and Ukip 12 per cent. Compared with 2010, Labour would have gained four per cent, and Ukip nine per cent; the Tories would have stood still while the Liberal Democrats lost 11 per cent. If we apply those shifts to each constituency, Labour would have 299 members of parliament and edge ahead of the Tories on 297, with the Liberal Democrats reduced to only 26.

But that is not what would happen. Past general elections suggest that we must take two factors into account. The first is that first-time MPs seeking re-election tend to do slightly better than their party’s national average. In the United States, where the same thing happens on a larger scale in congressional elections, this is called a ‘sophomore surge’. I estimate this is worth 15 seats to the Tories, as almost all their marginal seats are ones they captured from Labour in 2010. It will be like 2001 in reverse, when Labour suffered a two per cent swing to the Tories, but lost hardly any of the MPs who had first been elected in 1997.

The second factor is that some Liberal Democrat MPs are likely to hold on to a higher share of their past vote than the party nationally. My guess is that the party will win around 10 seats more than any figure emerging from a uniform swing calculation, with six extra Liberal Democrat MPs withstanding a Tory challenge, and four holding off Labour.

When we adjust my scenario for both factors, we end up with 306 Tory MPs, 280 Labour MPs and 36 Liberal Democrat MPs. In a very close election, the party with the most votes might just come second in seats but the wisest assumption for now is that if Labour is to come first in seats, it must also come first in votes.

Finally, let us cremate the notion that Labour could win while being placed well behind by the voters in polling on both who would be best prime minister and which party is best able to grow Britain’s economy.

It is true that Callaghan enjoyed a 24-point lead over Thatcher six months before the 1979 general election. Had he called an election then, Labour might have won. But he did not, and the winter of discontent caused Labour’s support to tumble. Perhaps Cameron’s government will be laid low by a catastrophe of similar scale, but it is politically unwise to rely on it and morally indefensible to want it.

As for the economy: yes, the Tories did edge ahead of Labour in early 1997 on ‘managing the economy’. But that needs to be put in perspective. The top six issues were health, education, crime, jobs, pensions and taxes. Labour was well ahead on four of them, and narrowly ahead on the other two. And Tony Blair enjoyed a double-digit lead over Major when voters were asked who would make the best prime minister. Today, the economy is by far the biggest issue, and the Tories enjoy a bigger and more sustained lead than they did in 1997 – and now it is Cameron who enjoys a double-digit lead over Miliband.

The point of this analysis is not to say that Labour cannot win next year. Historical records are there to be broken. Never before have we headed into an election with a coalition government and with the right-of-centre vote badly split. Perhaps things will be different this time. But the only way to ensure that they are different is for the right people to take the right decisions and carry them out effectively. Labour still has a daunting mountain to climb.

———————————