There are four ways Cameron could keep power. Photo: Getty.
Show Hide image

Who will win? Miliband is favourite, but here's how Cameron can survive

Mathematically, Ed Miliband remains the most likely post-election PM. But we think there are four ways David Cameron could keep power.

Two weeks ago, May2015 suggested that Ed Miliband was headed for Number Ten. “It is hard to see how David Cameron can cobble together the 323 seats he will need for a majority,” we argued.

That piece examined the most optimistic seat scenarios for Cameron, and showed how he would struggle to keep power even if they came true. We concluded that Cameron could still win, but his path to victory was clearly more difficult than Miliband’s.

We then introduced what we think is the key graph for this election. It maps all the crucial Tory-Labour marginals. These are the 80 or so seats that the Tories won in 2010 and Labour are trying to win back. We ran through the election’s other 570 seats and showed how if we are right about them, Miliband will only need to win 35 of these Tory seats to dethrone Cameron.

To work out how likely that was, we compared the probability of a Labour victory as suggested by Election Forecast, whose predictions are reproduced by Nate Silver and Newsnight, with the margin of victory in each seat according to our election model, which is based purely on Ashcroft and national polls.

Here’s that graphic, as it appears in this week’s New Statesman. [1] If a seat is in the top right quadrant, we and Election Forecast agree it is likely to vote Labour.


As you can see, 33 seats are in the top right quadrant. In other words, if Miliband wins all of these, he would only need a pair of the many green and light blue seats in the graphic, which we and EF think are very much in play; we either disagree on them (the green seats) or agree that the Tories will just prevail in them (the light blue ones).

But have things now changed? Are 33 seats still in the top right? What’s happened in the green seats, the election’s truly close marginals? And have the numbers changed in other seats - does Miliband still need to win 35 from the Tories?

How have things changed? Does Miliband still need 35 Tory seats?

We’ve re-run the graphic after more than 20 new national polls and about as many new seat polls. And the numbers still look best for Labour.

First, Miliband no longer seems to need as many as 35 of these seats. New Ashcroft polls have given Labour and the SNP double-digit leads in two seats we had assumed Cameron could count on – the Lib Dem seat of Bristol West and the Tories’ sole Scottish seat of Dumfriesshire. (We optimistically assume the Lib Dems, and every individual Lib Dem MP, will back Cameron.)

These Ashcroft polls make sense given national polls, so we have now moved these two seats into Labour’s bloc (we assume SNP MPs will vote for Miliband in a Queen’s Speech, whether they continue to is not our concern here). And that means Miliband now only needs 33 Tory-held seats.


So are 33 seats still in the top right of our graphic? Yes. 34 now are. (Click on the graphic for a bigger, zoomable version.)

There hasn’t been much movement overall. National polls have swung very slightly towards the Tories, moving most of the seats leftwards, but the probability of a Labour victory according to EF has, on average, actually risen by nearly 1 per cent.

Screen Shot 2015-05-04 at 03.56.50

34 seats are now in the top right, but that’s too binary a conclusion. More accurately, 27 seats look particularly likely to turn Labour. Can you see that there is an upmost upper right quadrant, marked by two thin dotted red lines? That’s an arbitrary box we created for all seats where Labour lead by at least 4 per cent and have at least a two-in-three chance of winning according to EF. There are 27 seats either in or on the edge of this box.

Miliband has to win six of these 16 truly close Tory-Labour seats.

If Labour can win all of these seats, Miliband would be a half a dozen seats away from ‘locking Cameron out’. Can he win another six? The odds seem to be in his favour - there are a dozen seats where Labour appear to be very competitive. They are the highlighted seats in the graphic.

We think these dozen seats will decide the election. If we want to cast the net slightly wider, we can add four seats – Rossendale & Darwen, Pudsey, Stevenage and Crewe & Nantwich – where Labour seem to be slight underdogs but could prevail. Miliband needs six of these 16 seats. Can he do it? Election Forecast are convinced Labour will win a pair of these 16 - Amber Valley and Lincoln. And the polls suggest Labour are a good few points ahead in another four seats where EF also think they are favourites - Harrow East, Warrington South, Ipswich and Stockton South.

Screen Shot 2015-05-04 at 03.58.02

If Labour wins all six of these, Cameron won’t be able to survive even if he wins all the other closest marginals - Finchley & Golders Green; Halesowen & Rowley Regis; Northampton North; Norwich North; Peterborough; and Ealing Central & Acton, as well as Stevenage; Rossendale & Darwen; Pudsey; and Crewe & Nantwich.

It’s clear who should be favoured if we trust the data. It’s also worth noting that Election Forecast assume the Tories out-poll Labour on Thursday by nearly 2 points; that’s why their numbers are slightly less favourable to Labour than ours, which point to a 0.5 point Tory win. Our numbers are based purely on the polls, whereas EF discount the polls in a way that slightly favours the Tories (as history - albeit a fairly limited history - suggests they should).


We think Miliband remains the most likely post-election PM. But Cameron could certainly win. We think there are four ways he keeps power.

First, the seats could simply break Cameron’s way. Our graphic could be very accurate, but Cameron could just pip Labour in 11 of the 16 key seats we've identified, as he must; as well as winning all the seats to the bottom left of the graphic, as we’ve assumed. The numbers could just go the Tories’ way on election night. We are talking about a few extremely close seats that are little more than coin tosses. (Coin tosses that are slightly rigged in Labour’s favour.)

Or second, the polls - and therefore our graphic - could be wrong. In the past fortnight a concerning divide has opened up between the two different types of polling: phone and online. Until a pair of recent ties, six consecutive phone polls had suggested the Tories now lead by more than 3 points. That is a greater lead than EF are predicting, and would shift the seats in our graphic towards the bottom left.

We think there are four ways David Cameron can keep power.

But online polls haven’t picked up such a divide. They still show a tied race, and most polls are online polls, so we and EF still predict quite a close race. If that’s wrong, and the phone polls of the past fortnight are right, the Tories’ odds in these marginal seats are better than we think. To put it another way, Ashcroft’s seat polls could all be systematically slightly wrong, and be showing vote shares that are too low for the Tories.

This is the same dilemma that faced forecasters like Nate Silver during the 2012 US election. Less than a week before polling day in November 2012, national and state polls were telling a slightly different story, in the same way that recent UK phone polls are telling a different story to online and seat polls.

Could “state polls systematically overrate Mr. Obama’s standing?”, Silver wondered five days before the election. “It’s certainly possible. (It keeps me up late at night.) If the polls in states like Ohio and Wisconsin are wrong, then FiveThirtyEight — and all of our competitors that build projections based on state polls — will not have a happy Nov. 6.”

The same will be true for us on Thursday night if polls are wrong, but, as one pollster put it to us last week, “a lot of polls have to be wrong”; Ashcroft’s seat polls fit with the picture of a tied national race shown by dozens of online polls over the past six weeks.

This is the same dilemma that faced forecasters like Nate Silver.

Instead – and this is Cameron’s third way of keeping power – the polls could be right at the moment, but swing the Tories’ way by Thursday. It may seem late for things to change, but Benjamin Netanyahu’s recent re-election was arguably evidence of how a centre-right government can squeeze the vote of a right-wing party just before polling day by stoking fears over statehood.

Fourth, if Cameron’s ‘Netanyahu strategy’ fails; and the phone polls were just an aberration; and things break in Labour’s favour as the odds suggest they should; then Cameron still has one way of keeping power. If he is seen to have won decisively on election night, perhaps some Labour MPs will balk at the chance to unseat him.

Screen Shot 2015-05-04 at 11.41.29

The Tories could win nearly a million more votes than Labour, and 20 more seats, and yet still face being voted down by an ‘anti-Tory’ majority. The Times is reporting today that some Labour MPs are concerned by whether they could then legitimately vote Cameron down.

Only 22 per cent of eligible voters voted for Tony Blair in 2005. Few recent Prime Ministers have had much legitimacy. But perhaps a few Labour MPs will choose to abstain when Cameron brings forward a Queen’s Speech. That’s not something we can model. Mathematically, Miliband is the favourite. In practice, perhaps Cameron can find a way to survive.

A flurry of final polls are due over the next 48 hours. May2015’s final election prediction will run shortly after their release.

[1] By the time our graphic ran in the NS, Miliband's magic number was down to 34, as we go on to explain.


Polls suggest Ed Miliband is likely to become Prime Minister (May2015, 19 April 2015)

This is how Ed Miliband gets to 323 seats and becomes Prime Minister (May2015, 24 April 2015)

Despite a few good polls, the chances of Cameron’s survival have not improved (May2015, 27 April 2015)

Harry Lambert was the editor of May2015, the New Statesman's election website.

Photo: Getty Images
Show Hide image

I'm far from convinced by Cameron's plans for Syria

The Prime Minister has a plan for when the bombs drop. But what about after?

In the House of Commons today, the Prime Minister set out a powerful case for Britain to join air strikes against Isil in Syria.  Isil, he argued, poses a direct threat to Britain and its people, and Britain should not be in the business of “outsourcing our security to our allies”. And while he conceded that further airstrikes alone would not be sufficient to beat Isil, he made the case for an “Isil first” strategy – attacking Isil now, while continuing to do what we can diplomatically to help secure a lasting settlement for Syria in which Assad (eventually) plays no part.

I agreed with much of David Cameron’s analysis. And no-one should doubt either the murderous barbarism of Isil in the region, or the barbarism they foment and inspire in others across the world.  But at the end of his lengthy Q&A session with MPs, I remained unconvinced that UK involvement in airstrikes in Syria was the right option. Because the case for action has to be a case for action that has a chance of succeeding.  And David Cameron’s case contained neither a plan for winning the war, nor a plan for winning the peace.

The Prime Minister, along with military experts and analysts across the world, concedes that air strikes alone will not defeat Isil, and that (as in Iraq) ground forces are essential if we want to rid Syria of Isil. But what is the plan to assemble these ground forces so necessary for a successful mission?  David Cameron’s answer today was more a hope than a plan. He referred to “70,000 Syrian opposition fighters - principally the Free Syrian Army (FSA) – with whom we can co-ordinate attacks on Isil”.

But it is an illusion to think that these fighters can provide the ground forces needed to complement aerial bombardment of Isil.  Many commentators have begun to doubt whether the FSA continues to exist as a coherent operational entity over the past few months. Coralling the myriad rebel groups into a disciplined force capable of fighting and occupying Isil territory is a heroic ambition, not a plan. And previous efforts to mobilize the rebels against Isil have been utter failures. Last month the Americans abandoned a $500m programme to train and turn 5,400 rebel fighters into a disciplined force to fight Isil. They succeeded in training just 60 fighters. And there have been incidents of American-trained fighters giving some of their US-provided equipment to the Nusra Front, an affiliate of Al Qaeda.

Why has it proven so hard to co-opt rebel forces in the fight against Isil? Because most of the various rebel groups are fighting a war against Assad, not against Isil.  Syria’s civil war is gruesome and complex, but it is fundamentally a Civil War between Assad’s forces and a variety of opponents of Assad’s regime. It would be a mistake for Britain to base a case for military action against Isil on the hope that thousands of disparate rebel forces can be persuaded to change their enemy – especially when the evidence so far is that they won’t.

This is a plan for military action that, at present, looks highly unlikely to succeed.  But what of the plan for peace? David Cameron today argued for the separation of the immediate task at hand - to strike against Isil in Syria – from the longer-term ambition of achieving a settlement in Syria and removing Assad.  But for Isil to be beaten, the two cannot be separated. Because it is only by making progress in developing a credible and internationally-backed plan for a post-Assad Syria that we will persuade Syrian Sunnis that fighting Isil will not end up helping Assad win the Civil War.  If we want not only to rely on rebel Sunnis to provide ground troops against Isil, but also provide stable governance in Isil-occupied areas when the bombing stops, progress on a settlement to Syria’s Civil War is more not less urgent.  Without it, the reluctance of Syrian Sunnis to think that our fight is their fight will undermine the chances of military efforts to beat Isil and bring basic order to the regions they control. 

This points us towards doubling down on the progress that has already been made in Vienna: working with the USA, France, Syria’s neighbours and the Gulf states, as well as Russia and Iran. We need not just a combined approach to ending the conflict, but the prospect of a post-war Syria that offers a place for those whose cooperation we seek to defeat Isil. No doubt this will strike some as insufficient in the face of the horrors perpetrated by Isil. But I fear that if we want not just to take action against Isil but to defeat them and prevent their return, it offers a better chance of succeeding than David Cameron’s proposal today. 

Stewart Wood is a former Shadow Cabinet minister and adviser to Ed Miliband. He tweets as @StewartWood.