Getting it wrong, getting it right

So it may not have escaped your notice that the UK has just held a decisive General Election (above). The Conservatives triumphant; the Scottish National Party celebrating; everyone else flatlining or crushed. The age of English and Scottish Nationalism is upon us, and our next constitutional battles are likely to see those two forces fight it out for the future of the United Kingdom. Oh good.

But where does that leave our predictions, here at Public Policy and the Past? One of the main losers on 12 December was Britain’s main Opposition, the Labour Party. They got run out of town in whole areas of the country where they used to dominate – not just stereotypical ‘Northern England’, where if you read some of the papers you’d think there was a whippet and a pint of warm ale on every corner – but in parts of the Midlands and South where they used to dominate.

Let’s go to Harlow and Stevenage, shall we? Two New Towns full of blue-collar workers where Labour held the historically-marginal seats until 2010. Now it’s a sea of blue as far as the eye can see. What about Cannock Chase, or Redditch in Worcestershire? They’re now so far out of Labour’s reach that they would need arms like a Mr Man to get anywhere near. There are simply not enough urban, young or studenty seats (hello, Edinburgh South and Truro and Falmouth) to make up for Labour’s historic collapse across Deep England – North, South, East and West.

That presents us with a problem, because, er, we said Labour could win this – not as a majority (without any real presence in Scotland, that looks impossible), but as a minority governing with the say-so of other parties – particularly the SNP and the Liberal Democrats. That was, well, let’s not gloss this… wrong.

That’s okay in a way though, because as we’ve said before the job of speculating (let’s not call it forecasting, shall we?) is to learn – to see clearly where you thought the pieces would fall, and the reasons why you thought that, against how they actually broke down. So this election result is a great opportunity to test our priors against reality. Why did we think Labour could get so close to the Tories, and why didn’t they?

Here’s what we thought back in the summer: Labour was deeply unpopular, but it still had three advantages over the Conservatives. One, the Tories were imploding. Their Parliamentary Party was in the process of what looked like a historic split between Liberal Conservatives and Tory particularists (as in 1846). Two, Boris Johnson was a great leader for Labour, deeply, deeply unpopular among all those swathes of liberal and Remain England in which the red team had to get a hearing and win back Liberal Democrat and Green defectors (and those famous Don’t Knows). Three, Labour had and has a huge membership that could give them a big advantage in the ground game – flooding marginal seats with activists that might not be able to convert people to their cause, but sure could Get Out The Vote.

Turns out this was really wrong. But we’ve at least got three categories in which to ask the question: why? Setting up opinions, and setting yourselves up to get shot down or proved wrong, is a good thing for these reasons. It allows self-reflection. It permits self-audit. It gives you the colour-in boxes to fill in after the event, and maybe to ask better questions and get it wrong more narrowly next time.

So, category one. The Tories didn’t implode. Prime Minister Johnson was able to expel the dissident pro-European wing from his Parliamentary Party and lose almost no electoral support. Amidst all the talk of Labour Leavers and their desertion from Labour, there’s been nowhere near enough talk of Conservative Remainers. In the end, a big majority of them stayed with the Tories. Why? Well, they were simply afraid of Labour leader Jeremy Corbyn. On top of (and related to) that, the Liberal Democrats had a torrid campaign in which yet again almost everything went wrong for them, as in 2017 - though the experiences of the February 1974 and June 1983 elections ought to have alerted us earlier to the possibility that two unpleasant extremes might tear their voter base apart rather than glue it together. In any case, a more emollient Labour leader and a more humble, focused Liberal Democrat advance might have rumbled the Tories. It didn’t happen.

On to category two. Johnson was indeed unpopular, but there are two reasons why this didn’t matter in the end. He wasn’t all that unpopular in Leave England (or Leave Wales): and because his mission was to unite the Leave vote around himself, and not around Nigel Farage’s upstart Brexit Party, that was all fine and dandy for him. The Brexit Party crashed to almost nothing, and despite having a couple of what amounted to good by-elections amidst the din (in Barnsley and Hartlepool), the very darkly comic character known as ‘Boris’ in the end looked like the best bet for everyone who wanted to leave the European Union.

The second reason why Johnson triumphed, despite being one of the most unpopular PMs ever at this stage of his stay in No. 10, was that he was pretty popular when set against Corbyn. Johnson was the political equivalent of a McDonald’s: divisive, likely to make you pretty unhealthy in the end, but a fast and dirty meal. Corbyn was more like a Little Chef: much talked about, never visited.

What about three, Labour’s ground game? Well, that didn’t work that well. You can’t polish a turd, of course – and Labour’s manifesto was absolutely deadly in that it made people laugh, not read. But the interesting points here go deeper. Activist turnout and effects were good in Putney, Labour’s only gain of the night, where the party was able to put out hundreds and hundreds of activists. Anywhere near a train station, on the Tube, at the end of a tram line? Great. Young engaged activists could pour in and make a big difference. Anywhere else – anywhere where you needed a car, say, oh… everywhere in Deep England? Much less successful.

Labour’s targeting operation also sent those really fresh and optimistic troops into dead-cert Tory seats, thus throwing away one of their only advantages. They did that partly because election supremo Karie Murphy doesn’t really know anything about elections, to some extent because their reading of 2017 was that they weren’t aggressive enough to gain more seats, and in places because they wanted to move people away from where they might actually save Labour MPs who don’t like Corbyn. As so often, one of the most tragic elements is just how much hope and goodwill has been squandered by the Labour-haters who now occupy the Labour cockpit.

One last thing. As so often, the feeling from the gut, and the first trigger movement, were right – and all the intellectualising and post-hoc data were wrong. Here at Public Policy and the Past, our first instinct was that Labour was heading for a terrible defeat from the moment it elected Corbyn. That’s not a Left-Right point so much as a point about the people around him, the long associations and ideas with which he was associated and would by which he would become known, and the poisonous influence of the super-union Unite – which has now taken over the Labour Party in all but name.

Many Corbynite insights are right. Britain does need much better public services, better organised public transport, more lifelong education. But as we’ve said again and again, these weren’t the people, and their presumptions weren’t the ideas, by which to carry that argument. Let’s end this experiment where we began, in the autumn of 2015:

There is a difference between inspiration and the peddling of false hope. Because what will happen when Mr Corbyn is either ousted by his Parliamentary colleagues, or – even worse for Labour – is actually allowed to collide with the electorate, like a piece of space debris burning up as it smashes into the atmosphere? The eye-popping but fake sugar rush of this microwaved Tony Benn’s elevation having passed, the subsequent crash will be terrible. There will be the blankest, darkest, most painful despair you can imagine, followed by blame – of Blairites, the media, the public themselves – who were not clever or far-sighted enough to accede to the Corbyn revelation… And then what? Then what? The answer, you know in your hearts, is this: decades of unbroken Conservative dominance. And a Britain that becomes less fair, less equal, less open, less liberal, less European – and less respected – with every passing day.

So there are three interesting academic lessons to be learned from the 2019 General Election: about the nature of the Conservative vote, now quite dependent on older, more socially conservative Britons who live in towns; the importance of relative and not absolute popularity; and the limits of a load of activists carrying a message that voters just don’t like. But the most important lesson of all? Sometimes, trust your instincts.

So where are the don't knows now?

Regular readers will know that Public Policy and the Past is obsessed, absolutely obsessed, with the ‘don’t knows’ that you don’t usually read about when you scan the headline figures in voting intention polls. So - as we head towards the finishing line of yet another national election (thank the Lord), maybe it’s time to have another look at them. 

There are three reasons for going back over this territory. The first is a general point. The don’t knows form the background hum of where the voters are coming in and out of each big camp – where the parties don’t have to detach people from another tribe to rally them around their colours, but only from a kind of weak attraction or half-remembered past association.

The second reason we’re obsessed with this point is more specific, and it’s, well, once bitten, twice shy. When then-Prime Minister Theresa May called a snap General Election in 2017, Britain’s Labour Party looked completely dead and buried. Opinion polls put them twenty points and more behind. A couple even gave the Tories double Labour’s vote. But then they zoomed up and zoomed up in the polls. Eventually, they hit 40 per cent in the final vote.

Why? Well, partly because May ran the worst campaign in British political history (at least since Labour’s in 1983), but also because there was one key point most of us prognosticators had missed: the sheer number of people who had voted Labour in 2015, but who were saying ‘don’t know’ at the start of the 2017 campaign. They duly turned up and voted Labour when the chips were down. We’re not making the mistake of leaving them out of account again.

The third factor behind this latter-day voyage around the don’t knows? Well, the 2019 General Election is beginning to look a bit like the 2017 one. Not exactly, not precisely, but quite a bit. For one thing, both parties are polling below the levels they reached during that campaign, and secondly, Labour’s score doesn’t seem to be accelerating upwards as fast it did last time.

But Labour are all the same gaining ground now, cutting the Conservatives’ lead in a number of surveys to bare single figures. Not a single national Voting Intention poll has yet implied a Hung Parliament, but one might well soon (and, given normal variation, probably will) – sparking that panic in Conservative ranks that we saw in 1987 and 1992, before their polling woes abated.

So where are we with the don’t knows this time? Could they come to Labour’s aid again, cutting the Conservatives’ lead from – on average – maybe nine points, pushing it below the all-important six points which means thatthe Conservatives lose their overall majority? Well, the answer is the classic academic’s cop-out: sort of, and sort of not. If we could chuck one of those hands-in-the-air don’t know emojis at you, we would. Which is funny really. Okay, maybe you had to be there.

Let’s have a scoot around the figures. We’ve gone through the last eight pollsters to report (and put out their tables), and excluded Deltapoll and Kantar, who don’t provide a crossbreak for ‘don’t know’ now and party allegiance in 2017. That leaves us with the data from six companies – Survation, YouGov, Opinium, SavantaComRes, Panelbase and BMG. That should be quite enough to get a general impression of where the don’t knows are right now. There are lots of ways you could cut this data (by gender, for instance, which suggests that Labour probably will benefit from a late move), but for brevity's sake here we'll focus on the 'past vote' category.

The answers aren't as encouraging for Labour (and for those in search of a Hung Parliament) as they might be. There is a differential, in that there are more ex-Labour don’t knows than Conservative, but it doesn’t look like there are enough on their own to close that polling gap. Survation will give Labour people the most hope. That firm suggests that 7.6 per cent of Tory voters from 2017 are now ‘undecided’, against a much bigger 13 per cent from the red team – though on the other hand ‘refused’ amounts to 2.2 per cent of 2017 Tories and 0.6 per cent of Labour voters from the last election, so we’re probably better off saying 9.8 per cent Tory to 13.6 per cent Labour. If they all move back to their prior teams, that’s worth maybe a point off the Conservatives’ lead.

Elsewhere, the news is less rosy for the Left. The latest YouGov poll has 10 per cent of 2017 Conservative voters saying ‘don’t know’, or refusing to answer, and 13 per cent of Labour – with rounding, not much of a better result than Survation’s for Jeremy Corbyn’s party, but still worse (for reference, the split was 12 per cent to 18 per cent in the last YouGov poll before the Commons voted for an early election). Opinium does have 16 per cent of ex-Labour voters outside London saying ‘don’t know’, and only 9 per cent of ex-Conservatives – a differential that might be worth a couple of points extra to Labour – but with a pollster which shows than lagging 15 per cent behind Boris Johnson’s party.

Lastly, there are three pollsters which show only a one point difference between the don’t knows among 2017 Labour and Conservative voters: Savanta ComRes, Panelbase and BMG. Those firms are showing 6 per cent of Conservatives undecided against 7 per cent of Labour, 5 per cent and 7 per cent, and lastly 9 per cent and 10 per cent. Not much comfort there.

What does this mean? It means that Labour can’t rely on the don’t knows. It will, in all likelihood, get a bit of uplift from that source, but not the three or maybe four points extra it needs to force Boris Johnson into another round of Brexit hell – or even, perhaps, form a government themselves. Labour will need to seek votes elsewhere. This will, of course, prove a harder task.

Labour’s chances therefore now rest on squeezing the Liberal Democrats and the Greens even harder, since the real battleground seems to be across the English Midlands and North, where the Conservatives are hoping to win a string of seats that have been traditionally (and culturally) Labour. That will be hard. Not impossible given a very volatile and uncertain electorate, but more difficult than convincing the don't knows. Labour have already pulled over a lot – and we mean a lot – of those votes already (the Liberal Democrats are five or six points down from their autumn peak). And those votes are unusual and sparse in many of these areas – in Great Grimsby, for instance, which at this point looks fairly doomed as a Labour seat. Hoping to convert almost all of Grimsby's remaining Lib Dems seems like a long shot.

So we’ve got a fix on the don’t knows. There aren’t that many of them left: very likely not enough on their own to force the Tories below 322 seats and into an effective minority. But Labour have climbed two ladders. They’ve definitely relegated the Liberal Democrats into second. They’ve powered up with ex-Labour returnees. Now they’ve released those gravity-defying rockets, the third stage is the hardest: get back Labour voters going Tory in small town England. Those afterburners might fire. They might not. A lot hangs on what happens when Labour presses that button.