Crunching decades of leadership election results suggests Labour should pick Keir Starmer

Historical analysis shows he would stand a better chance than the other candidates of leading his party to victory

January 24, 2020
A coalition of the left is needed to stand up for state support in this crisis. Photo: Empics/EMPICS Entertainment
A coalition of the left is needed to stand up for state support in this crisis. Photo: Empics/EMPICS Entertainment

Labour is not so much engaged in a leadership race as a leadership triathlon. We’re just out of the first stage, where the hopefuls have to gather MPs to nominate them, and into the second, where they seek the backing of local parties and affiliated organisations such as trade unions. Only then do we move to the third and final stage, where the ballots are printed and sent out to members plus registered supporters and union members to cast their individual votes.

Both the rule book and the twin experiences of 2015 and 2016—the only leadership contests run under anything even vaguely resembling the current rules—suggest that what matters for clinching the leadership is the final vote. In neither of the last two contests did Jeremy Corbyn have anything other than a small minority of MPs behind him, and yet in both he walked the final straight with an outright majority.

But the country might well be more interested in who is going to be its prime minister than in the identity of yet another opposition leader. And if so, history suggests we might do well to focus not on the final furlong, but instead on the first stage of Labour’s triathlon—because MPs’ votes in party-leadership elections have previously been a good guide to subsequent general election outcomes.

Until the mid-1960s, the Conservatives didn’t bother with leadership elections at all—instead the leader somehow “emerged” from a mysterious process involving a mix of the “men in grey suits” and Buckingham Palace. Therefore it makes sense to start our analysis in 1966, the first general election for which there is clear data on how many MPs had previously backed the leader in both parties.

There have been 15 general elections since then. Of those 15, as many as 12 have been won by the leader who previously enjoyed the biggest winning margin among MPs in their own internal party leadership election. (With Labour on a losing streak just now, it comes as no surprise to learn that the only exceptions came in three elections where a Labour leader with a bigger lead among their own MPs was defeated by a Tory with a smaller edge within their own parliamentary tribe.)

Given that Boris Johnson won all the rounds of MP voting in last year’s Conservative Party leadership election, the only Labour leadership candidate who could possibly have a bigger margin of victory among Labour MPs is the one who won the most nominations: Keir Starmer. Last year, Johnson’s margin of victory was 22.7 per cent in the first round. Starmer actually did better last week, with a lead of 25.9 per cent among Labour MPs’ nominations.

Given that 80 per cent of general elections since 1966 have been won by the leader with the biggest lead among their own MPs, can we now say that if Starmer becomes Labour leader he would have an 80 per cent chance of beating Johnson at the next election?

Not quite. A four out of five chance is the most generous interpretation of the data for Starmer. Considering only elections where the Labour leader had the bigger margin among MPs, Starmer would have a two out of three chance of winning the next general election. Still, that doesn’t sound bad.

Sadly for Labour, past experiences suggests that the scale of last year’s defeat reduces the chances of a Labour victory next time. If we factor into our analysis each party’s prior share of seats in the Commons, alongside the margin the leader enjoys among MPs, then Labour’s hopes of winning most seats under Starmer sinks to 50:50. But that still means a better than average chance of forming a government if most third-party MPs continue to be better disposed to Labour than the Tories.

Inevitably, there are different ways to crunch the numbers. How much weight, for example, do we give to the fact that Starmer’s lead among Labour MPs is only slightly greater than Johnson’s among Tory MPs? Factor this in to the analysis along with the numbers from the previous election and the correlations become very weak. There are further choices about which data to use, such as which rounds of voting to count and whether to update with no-confidence votes.

And yet in general elections the quality of the leaders is bound to matter, and it is better to assess this with an imperfect gauge than no gauge at all. While MPs may sometimes be out of kilter with the country, there are some good reasons for paying attention to them: their jobs and prospects are on the line, and they know the candidates best. In short, they have a strong incentive to pick a winner, and the best information with which to try and do so. And thus indeed it turns out that simply identifying which leader has the bigger lead among MPs has helped predict general election outcomes.

If Labour does pick someone other than Starmer, it will not be the first time it has chosen someone who did not win the first round of MP voting. Before Corbyn, Callaghan (1976), Foot (1983) and Miliband (2010) all eventually carried the day despite losing the initial round. None of them ever won an election. By contrast, Labour’s only post-war winners (Blair, Wilson and Attlee) all arrived in the top job with their MPs behind them.

Conservative leaders similarly have a much better record if they were the most popular candidate in the first round of their leadership election. Barring Michael Howard’s unopposed selection in the emergency that followed the defenestration of Iain Duncan Smith, all five of the leaders who won their initial round went on to win at least one general election (Heath, Thatcher, Major, May and Johnson). By contrast, of the three who lost their first round leadership contests (Hague, Duncan Smith and Cameron) only Cameron won a general election, and he had only missed out narrowly by six MPs’ votes. By contrast, the second-placed Labour leadership candidate, Rebecca Long-Bailey, won 49 fewer MP nominations than Starmer.

Labour will be going into the next election 162 seats behind the Conservatives. Whomever Labour picks as its next leader, the party has a lot of ground to make up. But, with an increasingly volatile electorate and with the Tories defending a long period in office, the opposition has a serious chance of winning the next election.

Historical experience is imperfect, open to interpretation, and is not always the best guide. But Labour MPs have sent a clear signal in favour of Starmer. And imperfect as it can be, the guide of historical experience therefore suggests that he would have a good chance of winning an election—and a better chance than the other leadership contenders.