Why Labour MPs could still decide the final result

The backing of MPs could still have a decisive effect on the votes of Labour members.

Sign Up

Get the New Statesman's Morning Call email.

I followed the nomination battle for the Labour leadership in much the same way that I imagine fans of the English Premier League followed the U-21 European Championship: with sadness that the main event (the general election) had come to an end, and a sense of foreboding that the summer wouldn't provide me with much excitement, still less the daily fix I needed to sustain my habit.

Indeed, with the exception of some polls which have gamely attempted to gauge the mood within the Labour party (or Labour party sympathizers, the leadership campaign seems to have changed from a numbers game to a game of positioning (or labelling). The Staggers’ list of MPs supporting each candidate has remained frozen in time, forever stuck  at the close of nominations this year.

On one reading of the leadership election, that's entirely to be expected. "The parliamentary party, having decided the shape of the ballot paper, no longer matter". The tripartite electoral college which so complicated matters in 2010 no longer exists, and the votes of MPs and MEPs weigh the same as the votes of ordinary party members.

Yet if we look back to the results of the 2010 leadership election, we can see that MPs endorsements did matter, and not just for deciding who got on the ballot paper. In 2010, Labour released the results of the membership vote in each constituency. In almost all of the Labour-held constituencies, the sitting MP endorsed one of the candidates. Those endorsements mattered for how MPs voted -- but did they have any effect on members of the constituency party?

Figure: Difference between share in constituencies where candidate received endorsement (in black) compared to all other constituencies (in grey).


The graph above shows the average (first preference) vote share for 2010 candidates in constituencies where they did and didn't receive endorsements. In all cases, candidates received more votes in constituencies where the MP had endorsed them. The endorsement bonus varied -- endorsements for Andy Burnham carried a lot more weight than endorsements for Ed Balls -- but in all but one case (Diane Abbott) there's statistical clear blue water between endorsing and non-endorsing constituencies.

Of course, constituencies in which the sitting MP endorsed Diane Abbott are likely to be different in lots of respects from constituencies in which the sitting MP endorsed David Miliband. That is, these differences might be epiphenomenal -- the same things might be driving both the endorsement and the difference we see in vote share. But when I control for a number of obvious factors -- like how left or right-wing the constituency is -- endorsement still matters. The relative position of candidates changes: Diane Abbott gets on average six percentage points from her endorsements, compared to 10 percentage points for Andy Burnham, or just five for endorsements for David Miliband.

This matters, because according to the frozen-in-time list of endorsements, 22 MPs can still endorse but have not yet decided. By continuing to work on their parliamentary colleagues, leadership candidates may end up winning votes amongst ordinary party members.

I admit, the rewards are not huge. The average turnout in each constituency party is likely to be between two and three hundred voters, so each endorsement only brings twenty extra first preferences. And it may be that ordinary party members will care less about MPs' endorsements now that they now they won't have extra weight in the final tally. But in an election where small margins can affect not just the final result, but also the transfers of votes between candidates, those extra first preferences might well be worth playing for.