In recent months, Labour has endlessly debated the ins and outs of “electability”. Jeremy Corbyn’s detractors say that he is “unelectable”, a throwback to the dark days of the 1980s when Labour never looked, let alone acted, like an alternative government. Corbynites, in their turn, reply that a coalition of Green and Liberal Democrat leftists, previous non-voters and socialists who have in recent years turned to the Scottish Nationalists or Plaid Cymru might now ‘return’ to the Labour fold, allowing it to win unlikely yet famous victories. But which camp has electoral experience, and the data, on their side? Is it possible, even at this early stage of the new Labour leader’s reign, to scrape together enough evidence to draw out general conclusions?
Here I will argue that there are now enough statistics out there to justify at least a tentative answer to this vexed question of new-look Labour’s electability (or otherwise). We now have a great deal of information from headline voting intention surveys, the detailed specific questions asked by pollsters, from local by-elections and one Westminster contest (in Oldham West) to draw at least an initial picture of how Labour is doing. That picture should be in general deeply discouraging for Labour people, for it reveals a party that now seems to have retreated even from the position of fragile and limited recovery it thought it was enjoying early in the last Parliament.
First, the headline voting figures. The average Labour vote share projected by looking at the last poll conducted by each of our regular pollsters, including multiple scores for companies who use more than one method, is about 31 per cent; by contrast, at the same stage of the 2010-15 Parliament, Labour was running at an average of just under 40 per cent. It is true that pollsters have amended their methods since the debacle of the last general election, and that this may be massaging down Labour’s figures from what they would otherwise be – though that is not absolutely certain, for we cannot be sure when the divergence between polling scores and voters’ ‘true’ intentions began. The British Polling Council has not completed its post-mortem on May, and many pollsters have not changed their methods as much as (for instance) ComRes, which has started to count respondents not if they say they are going to vote, but on the basis of past experience for each demographic element of the electorate. ICM have also recently concluded that differential response rates to telephone polling may still be flattering Labour, something that is very hard to adjust for whatever techniques pollsters adopt: in this situation, the figures now and in 2010 may be rather more comparable than many believe. If Labour really were running 8-9 cent below their score in late 2010, and were to carry on doing so until the next general election, they might attract a vote share of only 23 or 24 per cent: a chilling prospect for anyone who wishes Labour well.
Looking to contemporary history, there are few precedents for a Labour Opposition doing so badly. We have again to be very wary of comparisons here, for polling has changed out of all recognition since the 1980s: the “spiral of silence” adjustment, to adjust for so-called ‘shy Tories’, is one example of such differences. Even so, what figures we do have show that only once, in January 1988, has Labour been so far behind the Conservatives while in Opposition at this stage of a Parliament. Then, in the heyday of Thatcherite triumphalism, Neil Kinnock’s Party trailed the governing party’s score by around eight per cent – about the same gap that separates Corbyn’s Labour and David Cameron’s government now. Each of Labour’s many other spells in Opposition since 1970 has seen the party doing rather better by this stage: leading by one per cent in January 1971, and by six per cent in December 1979, trailing by ‘only’ six per cent in January 1984; and leading again in our next two examples, by a huge 18 per cent in November 1992 and a tiny 0.4 per cent in December 2010. The basic impression is stark, but inescapable: Labour has not in modern times been behind eight months after a general election and gone on to win the subsequent contest. This is all the more worrying given the lack of new leader ‘bounce’ that Labour should – if the last four decades’ polling are any guide – have been enjoying by now. The Party should now be five per cent or so up on their pre-Corbyn travails over the summer: instead they have gone nowhere at all.
The below-the-line details reported in these polls are little better. Corbyn’s own numbers are absolutely appalling, rivalled only by Michael Foot at his nadir. Some of the numbers are so bad they are a little hard to grasp. His net score for ‘doing well’ or ‘doing badly’ with YouGov is -41 per cent; those who take an unfavourable view of him lead those seeing him in a favourable light by 28 per cent with Opinium; Ipsos-Mori’s ‘satisfied’ or ‘dissatisfied’ rating stands at -17 per cent (Foot’s score was -21 per cent at the same stage of his leadership). What is worse, these numbers have been getting steadily worse: Corbyn’s high initial (and strong) favourability among his large following of left-leaning enthusiasts has not prevented less committed ‘don’t knows’ moving largely against him. That net YouGov score was -8 in late September, and -20 in late October.
Some of the geographical, as well as personal, details are just as bleak. When I last surveyed the scene for The Staggers in late September and early October, I spied a few signs to give Labour some hope in this respect: in particular, a handful of by-elections that pointed to the Party doing rather better than previously in Scotland, and in Wales (taken together with an encouraging YouGov poll) perhaps even moving forward a little. These signs now also appear to have fizzled out. Polling for May’s Scottish elections has seen Labour make absolutely no progress against the extraordinarily dominant Scottish National Party, registering scores for the constituency part of the vote that have stagnated in the low 20s ever since the general election. The latest YouGov Welsh Political Barometer Poll shows Labour back where it started in May in terms of Westminster voting intention, registering exactly the 37 per cent the Party attracted as it actually managed to lose Welsh seats to the Conservatives. Those numbers are not any better – indeed, they are a little worse – when it comes to polling for the Welsh Assembly contest in the spring.
Local by-election results follow the same pattern. Labour’s gross voting figures since Corbyn’s election as leader are some way down on those recorded in the last Parliament, regardless of the year in which the seats were last fought: the Party’s gross vote share in comparable contests is three or four points down on seats last contested in May 2015, while the Conservatives are only down a point or two. These figures have to be treated with extreme caution. Local by-elections usually attract only a very small turnout, swings between the major UK parties are often affected by local parties or independent candidates (or larger parties’ failure to field a candidate), and are usually held in random and extremely unrepresentative locales. All that said, the acknowledged masters of local election analysis, Michael Thrasher and Colin Rallings of Plymouth University, reckon that once we smooth out these factors the figures from post-Corbyn contests portend a one point Conservative lead, of 32 per cent to 31 per cent, in May’s local elections. Labour will therefore be down seven or eight per cent on 2012, and might stand to lose around 200 council seats – good corroboration, if and when this happens, for our opinion poll findings.
More importantly, if such an outcome does transpire, we can start to sketch out some of the implications for Labour’s vote in 2020. According to modelling by Matt Singh of Number Cruncher Politics, and on which he in part based his famous prediction of the 2015 polling failure, a weighted average of a one point government lead in local elections across a Parliament would imply on past evidence a very large government lead come the next polling day – of something like 12 or 13 per cent. It that was indeed to be our projection by 2018 and 2019, something made much more likely by such a result in May 2016, it would be entirely consistent with Labour’s poor UK polling at the moment, as well as its potential failure even to match last time’s performance in elections to the Scottish Parliament and Welsh Assembly. Such a weak performance in local government and devolved contests would be one more sign that, overall, the situation is worse even than the initial picture I painted back in September and October.
This brings us neatly to the Oldham West parliamentary by-election, where Labour defied widespread expectations of an only narrow win by romping home against a very disappointed and deflated United Kingdom Independence Party. Labour’s majority, at over 10,000, was seven per cent up on May, a result that on the face of it flies in the face of everything we see in polls, polling internals and local by-elections. But any reliance on such a superficial impression would display, once more, a failure to see these things historically. In the Oldham East by-election, held nine months into the last Parliament, Labour’s vote went up by 10 per cent; at Barnsley Central, two months later, the Party’s vote increased by 13.5 per cent; and two months later, its vote went up by more than 12 per cent at Leicester South. In this context, Oldham West once again showed us a Labour Party that is significantly under-performing even its anaemic performance in 2010-15, a Parliament which ended in a dramatic and decisive defeat. Trailing the early by-election successes of 2011 by between about three and six per cent is by no means out of line with an opinion poll performance eight or nine points shy of what it was then, especially when we take into account a slight UKIP decline in polls conducted since the general election.
The upshot of all this? Labour is likely to do pretty badly in next May’s English local elections. It is probably going to be trounced in Scotland and to be forced into retreat in Wales. In each of these cases, it will have gone backwards on Ed Miliband’s performance in 2011 and 2012, and quite markedly so: Labour can hope on this evidence to gain only about 30 Members of the Scottish Parliament, and perhaps around 27 Members of the Welsh Assembly. They won 37 and 30 respectively early in the last Parliament. Failing to reach the heights of even Miliband’s limited appeal is no-one’s idea of progress.
Only in London can Labour hope for real success. Here there is a very good chance that Sadiq Khan will defeat Zac Goldsmith for the London Mayoralty. Local by-election results in the capital look much better for Labour: since September, they have advanced by three or four per cent on the last time these seats were up for election in 2014, while Khan has a lead of six per cent (53 per cent to 47 per cent) in the latest opinion poll there. Here Labour’s renewed appeal to leftish and Green voters, and the massive activist army created by London Labour numbers surging to a fifth of all Labour members across the whole of the UK, may well pay off. It will be at least one victory for Corbynism.
Such an undoubted triumph may burnish May’s results with a very thin sheen of success. But that will probably be deceptive. Things could, of course, as ever be even worse: Labour’s poll share could have fallen further, as the most advanced doomsters were predicting back in September. Ukip might have broken through among traditional Labour voters in Oldham West, and pushed Labour really hard and close for that seat. But there are enough ominous signs to be very worried for Labour’s future. The party’s historically very poor opinion poll performance and lack of new leader ‘bounce’, Corbyn’s abysmal personal ratings as they plumb the depths of Michael Foot’s, the weakness of Labour’s local by-election showings and the failure to do any better in Oldham West than in (say) Barnsley in 2011: these are all deep, deep red warning signals of potentially resounding defeats to come.
These conclusions are only interim inferences from a limited amount of data. The deluge of numbers we can expect in early May will tell us much, much more. But there can be no doubt now: Labour is weaker than under Ed Miliband. If nothing changes soon, and previous relationships hold, the Party is dicing with a double-digit defeat at the 2020 general election, at which it might attract a vote share some way down into the mid- to high-20s. Everything we know – every last scrap of data – says that the Labour Party as we have known it is in very profound trouble indeed.