Sunday, 1 November 2015

What are the boundaries and limits of Labour's 2020 performance?

You know what we say, when everyone says that this blog is overly gloomy? We say: it's not us, it's the numbers. It's not what we think. It's the reality. And our recent scepticism about the UK Labour Party's electoral prospects is just the same. Take a look at the two charts above (click on them to expand your view): such figures can and should be the first port of call for every historically-informed and data-capable commentator. We've assembled these from Mark Pack's invaluable opinion polls database of all the figures he's been able to get his hands on for the years since 1945. And what they show is not pleasant reading if you think that pluralist, competitive political systems are vital for good governance and a healthy democracy.

Let's start with the first chart. This shows the polling distance that separated Labour from the Conservatives each time the former party found itself in Opposition since 1970, and after six months had in each case passed since the previous election (as it has now). This doesn't actually look all that bad for Labour right now. Sure, they're an average of 6.2 per cent behind when we average out all of October's polls, but they were about the same distance behind late in 1983 too, after Mrs Thatcher's post-Falklands War landslide, and they were an enormous 13.4 per cent back late in 1987, when Thatcher's Conservatives were in the purple of their pomp - enjoying a huge economic boom, and basking in the glory of a third successive victory.

So Labour has been here before. That shouldn't, however, be entirely comforting if you are a Labour supporter, for right now they're registering polling numbers for this stage of Opposition that are worse than all but two periods in our entire recent history. Also, and of course, the party went on to lose the two elections that followed those low scores. Only where they already led by now have Labour, in modern times, ever emerged from a Parliament to win, and they even managed to lose once out of the three occasions on which they were ahead (in 1979, after which they went on to be buried while led by Michael Foot in 1983). Yes, we'll let that sink in for a moment. The fact that Labour is behind is already a red signal that they are probably about to lose again.

It gets worse. Now move on to the second chart. This examines the fall in Labour's rating from this point onwards, on every occasion they've been in Opposition. This picture isn't very pretty either. Only once, in every example we've got to look at, have they moved forward at all from the numbers they were posting six months after a General Election. And that was when they were led by Tony Blair, at the height of his remarkable popularity, and faced by an obviously flagging (and very badly divided) Conservative government that just seemed to be running out of time.

The average fall, from this point to an election four or five years hence? About 6.1 per cent - which would put Labour in 2020 right back on about 25.7 per cent, or 190 seats on the current boundaries of our 650-seat House of Commons, and winning only 170 out of 600 if the Conservatives' boundary and voting reforms go through (you can play around with election simulators such as that built by YouGov's Anthony Wells here). Think we're being too gloomy? Well, we could have included every Parliament since 1970, including those during which Labour were in government, and then the average fall from this equivalent point in time is 8.4 per cent, pushing Labour right down to a 23.4 per cent share of the vote. Then they'd win only about 170 seats on the current boundaries - or an eye-watering, epochal, shattering 145 seats on the new lines. Throughout we've held the Conservatives in our projections at a fairly realistic 40 per cent (a very, well, conservative assumption if Labour does plumb these depths). We could have made some even more fearful assumptions. You get the picture.

Now, we do know that there are all sorts of problems with these numbers. We're comparing apples with pears. Polls have undergone a number of transformations and methodological changes throughout the period under review, not least after the 1992 polling debacle which saw pollsters begin to adjust for a 'spiral of silence' in which 'shy Tories' refuse to admit their real loyalties. A lot of what is apparently a 'fall' in Labour ratings between year one and the final year of each Parliament may well have been polling error, especially in the late 1980s and early 1990s. The polls overstated Labour in 1992, 1997, 2001, 2005 and 2015 (though not in 2010). Labour has been historically overestimated, between final polls and General Election results, by about three per cent. So the 'decline' to each result was probably less than it appears. If pollsters have now ceased to overstate Labour (if, for instance, ComRes' new and very tight voter screen is producing the 'right' results), then there'll be less decline running up to 2020. Maybe Labour will hold the line about three points above our historically-based projections here - and secure between 26.4 and 28.7 per cent of the vote. Actually, that looks a bit more realistic, doesn't it? The major problem with that less catastrophic guess, though, is as follows: it still involves wining only something like 170 to 190 seats on the new boundaries.

There are 232 Labour MPs right now. If these numbers are anything like right, somewhere between 42 and 87 of them could be losing their jobs in 2020.

Well, that's it. These are all the indicators we've got. And they say this: if history and data are any guide at all, Labour can hope, at best, only to escape from the next General Election having merely been badly defeated. But if the party is unlucky, or things go very badly for them, at one extreme of established precedent they could be facing a historic rout that will reduce them to being the party only of English and Welsh inner cities and radical university towns. At one end of the data's limits, Labour will have ceased to be a truly national party.

Don't blame us. That's what recent history suggests. This is what the numbers say: cold, clear, inescapable, and there for all to see.