Progress | Centre-left Labour politics

Incumbency matters

The ‘sophomore surge’ could stymie Labour’s path back to power, warns Peter Kellner

At the time few people noticed something odd about Labour’s second landslide in 2001. Nationally, there was a two per cent swing from Labour to the Conservatives. The Tories should have gained 17 seats that Labour had won in 1997. In fact they gained only four. Moreover, as Jim Knight overturned a tiny Conservative majority in Dorset South, the Tories’ net gains from Labour amounted to just three.

What happened was that in Labour’s tightest marginals there was, on average, no swing at all. Members of parliament who had gained seats in 1997 and were now defending them as first-time incumbents won an average of 1,000 votes more than would have been predicted from the national swing.

This phenomenon was particularly marked because Labour had gained so many seats in 1997. There were far more first-time incumbents than normal. The ominous thing for Labour is that 2015 could be like 2001 in reverse. This time there will be around 80 Conservative MPs who will be defending seats they captured from Labour in 2010. So, was 2001 exceptional, or par for the course?

The answer is that the incumbency factor is normal. Indeed, it applies to other countries. In the United States, first-time congressmen and congresswomen tend to do so well that they are said to enjoy a ‘sophomore surge’.

In Britain, evidence of sophomore surges can be found at every election for the past 35 years. It has been logged by the Nuffield series, ‘British General Election of …’, from which the information in the box opposite is taken.

The pattern is clear and remarkably consistent. If it persists next year then the average swing in Conservative-Labour marginal is likely to be around two percentage points less than the national average. Let’s suppose the two parties end up with the same national vote share. This would represent a 3.5 per cent swing to Labour. On a uniform swing Labour would gain 44 seats. This would be enough to take Labour to 302 seats, even before we add in any seats the party takes from the Liberal Democrats. The Tories would be down to 263. Even if Labour loses some seats in Scotland to the Scottish National party, Ed Miliband would still lead the largest party in the new House of Commons.

But if the sophomore surge reduces the swing in Tory marginals from 3.5 per cent to 1.5 per cent, Labour would gain just 20 seats. Labour would then have 278 seats, compared with the Conservatives’ 287 (though once again the figures would need to be adjusted to take account of the performance of the Liberal Democrats and SNP). On equal votes the sophomore surge might be enough to keep David Cameron in Downing Street.

For Labour to do significantly better than this, one or more of three things must happen. First the United Kingdom Independence party must hold on to more former Tory voters in Tory-held marginals than elsewhere. Second, Labour’s ground campaign must be especially effective in these marginals. Third, more first-time Conservative MPs must stand down. So far six of the 44 MPs vulnerable to a swing of 3.5 per cent have decided to stand down. One, Louise Mensch, has already quit, and Labour has gained her seat, Corby. The other five will have no incumbency bonus. Labour will hope that more Tory MPs with small majorities will decide to leave parliament next year.

As for the Liberal Democrats, history is less useful. We have never before seen such a haemorrhage of their votes, nor have they had to go into an election defending their record as a coalition party. But what evidence we do have suggests that they may enjoy an even greater incumbency bonus. If support for their MPs does hold up better than average, this could mean that they lose, say, 20-30 seats rather than 30-40. However, Labour will hope to take Brent Central and Redcar on big swings, as their Liberal Democrat MPs, Sarah Teather and Ian Swales, are standing down.


Peter Kellner is president of YouGov



Twelve Labour members of parliament defended seats they had gained from the Tories in October 1974. They suffered an average swing to the Conservatives around four per cent less than other Labour candidates in their region.


Labour did so badly that there was no extra incumbency bonus for Tory MPs first elected in 1979. However, five Labour MPs who replaced Conservative incumbents in 1979 all did better (or less badly) than expected in 1983.


Sixteen Tory MPs defended seats gained from Labour four years earlier where the former Labour MP was not standing again. They ended up an average of four per cent further ahead of Labour than would have been expected.


Twenty Labour MPs had ousted incumbent Conservative MPs in 1987. Their support was 1.7 per cent higher than in those seats which Labour held before 1987 and where the incumbent MP was standing again; the Conservative vote fell on average by one per cent more.


This time the incumbency effect in those seats that Labour had captured in 1992 was worth 2.5 per cent compared with comparable seats elsewhere.


In 80 seats that Labour gained in 1997 and where the new MP was defending it, Labour’s vote was down just 0.1 per cent; in those seats where the same Labour MP defended a seat in both 1997 and 2001, Labour’s vote was down 4.1 per cent.


In the two seats that Labour gained in 2001 (Dorset South and Ynys Môn) its share was down just 0.3 per cent and 0.4 per cent, compared with a loss of almost six per cent nationally. In seven seats that the Conservatives gained in 2001 from all other parties, their share increased by four per cent, compared with 0.5 per cent nationally.


Incumbent Conservative MPs who first won their seats in 2005 secured an average increase of 5.6 per cent, while for those who had been in the Commons longer the figure was just three per cent.

Progressive centre-ground Labour politics does not come for free.

It takes time, commitment and money to build a fight against the forces of conservatism. If you value the work Progress does, please support us by becoming a member, subscriber or donating.

Our work depends on you.

Print Friendly, PDF & Email

Peter Kellner

is president of YouGov

1 comment

  • Is Mr Kellner pro- or anti- Ed Miliband? Some recent articles in newsprint media have him slagging off Labour’s leader and their chances of a ‘win’ next May. If you can’t say something nice abut Ed [espcially on this site] then don’t say anyting at all. PS: As he is an analyst with a pretty impressive cv in the journo’ world he always appears very pro- Cam/ Con to me. Who he votes for is his choice entirely – to have his comments [and Dave Cs pic] appear in the Progress mag’ is somewhat mystifying to me.
    Politics makes for strange bedfellows. I ask Progress new director and editorial staff to leave articles which are [I contend] slagging off Labour & Ed’ to the DM & Murdoch’s rags.
    Or maybe I have missed something here?

Sign up to our daily roundup email