In search of the Red Wall (4)

The story so far: the Red Wall was originally defined (in August 2019) as a contiguous group of Labour seats, mostly in the Northwest of England, whose demographics suggest that they ‘should’ be Conservative. Nobody really went for this definition. It was redefined (in October) to include Labour seats in the Northwest and Yorkshire where the Conservative vote had increased and/or where there was a large majority for Brexit; nobody really went for this one either. A journalist writing in the Spectator then used the term to refer, more impressionistically, to a “block” of solid, long-term Labour seats in the Midlands and the North (some a hundred miles apart), which had somehow turned into winnable Conservative targets by virtue of the popularity of Brexit and the unpopularity of Jeremy Corbyn. After that, things got a bit silly.

“One measures a circle, beginning anywhere” – Charles Fort

In November the Red Wall appeared in the FT. This article suggested that even the Spectator definition had been too restrictive: it refers to the “red wall” of Labour seats in working-class areas of the Midlands, Wales and the North of England. Red because Labour, a wall because… there are a lot of them, in some parts of the UK which aren’t on the Underground? (Cheap shot, I know, but when people talk about “the North” as if it were a single area (“Bolsover, Bishop Auckland and Ashfield”) it does look like metro-provincialism.)

Then we got – as promised in the previous post – the proof. I once saw a review of a book on ley lines which noted that the book included several different maps, all with alignments of ancient sites duly marked, and concluded that the authors had provided definitive proof that it was possible to draw straight lines on maps. In a similar spirit, here – courtesy of Sebastian Payne and John Burn-Murdoch, the FT‘s indefatigable data guy – is The Red Wall: The Proof.

They said it couldn’t be done… hang on, they said it shouldn’t be done.

I can’t argue with that: if one measures a large enough ellipse on a map of Britain, it will contain a lot of different Labour constituencies. And – what actually is potentially interesting – we can see a few contiguous belts of Labour seats. Using the excellent (and free!) MapChart tool, I’ve reverse-engineered Burn-Murdoch’s graphic as follows:

Not shown: Wales. I don’t know what went on in the Vale of Clwyd, but it’s highly unlikely to be the same thing that went on in Derbyshire, particularly if flags were involved in any way. Also not shown: Blackpool South, which was a Tory target (and is on this map) but didn’t fall within the Burn-Murdoch Ellipse.

This doesn’t look a lot like James Kanagasooriam’s original Red Wall, for the simple reason that that was defined in terms of the underlying demographics of the constituencies involved; this is defined in terms of the seats being (a) held by Labour (b) quite close together and (c) er, that’s it. At best it’s a map of Conservative targets in the Northwest, Yorkshire and Lincolnshire. I want to stress this point before we go any further. Looking at the original Kanagasooriam map and then looking at the seats which actually went to the Tories, we know something we didn’t know before: those six seats had certain demographic characteristics, and it’s possible that in 2019 those characteristics outweighed the cultural and historical factors which had been keeping them Labour. (Although in the other 34 seats on the map – which stayed Labour – that plainly didn’t happen.)

By contrast, if we look at this map, then look at the seats which went to the Tories, we know… that the Tories won some, perhaps many, of their target seats in the Northwest, Yorkshire and Lincolnshire. That’s all it really tells us.

But let’s do it anyway.

Red Wall crumbling, am I right or am I right? Andy Burnham, Dennis Skinner, Mary Creagh, your boys took a hell of a beating!

Well, perhaps. (I’m not saying we didn’t lose all those seats – look at the government’s majority.) But let’s remind ourselves, again, of what we know about the Red Wall, in this iteration of it (which, sadly, is the one that’s stuck) – and what we don’t know. Do we know that these are all Labour seats with “Conservative” demographics? No, we don’t know that. Do we know that they are all socially conservative areas with big majorities for Brexit? No, we don’t. Alternatively, do we know that these are all seats that had been Labour “for generations” (K. Balls) before the Labour vote collapsed in 2019? No, we don’t. Do we know that the Labour vote hit an all-time low in all of these seats in 2019, or that the Tories won them all with a substantial majority? Also no.

All we know about the purple seats is that they were Labour going into the election; the Tories won them; and they form a belt running across the country – most of it, nearly – at approximately the latitude of Southport. And that third fact, in itself, tells us nothing – let’s face it, everything has to happen somewhere.

What if we were to distinguish between Labour seats that actually had been Labour for a while – since 1997, say – and those that had been Labour in 1997 but changed hands (twice) since then? Out of the remainder, we could also distinguish between big wins for the Tories and narrow wins, the kind of seat gains that happen in any election when the tide is running strongly in one party’s favour (as, I think we can all agree, it was for the Tories in 2019). Having looked at the data, I’m defining a “big win” as one where the Labour vote in 2019 fell 10% or more relative to 2017 and was lower than at any time since (and including) 2001, and where the eventual Tory majority was 5% or more. Comparing these factors against the (losing) Labour share of the vote, I found that the lowest Labour vote of any constituency in the “narrow win” group (where at most two of these criteria are met) was 39.8% – all the rest had a Labour vote of 40% or above. Conversely, every constituency in the “big win” group (ticking all three boxes) had a Labour vote below 37%. (Some other “big win” seats, outside the Burn-Murdoch ellipse, had higher Labour votes than that, with one – Wolverhampton North East – hitting 39.8%; none reached 40%, though.) This suggested to me that I was on the right track: intuitively – and thinking back to the first post in this series – a seat where 40%+ of people still vote Labour is not one that’s “gone Conservative”.

So here’s a revised map. The pale lilac constituencies are those Tory gains which had already changed hands twice in the previous twenty years; the mauve ones are the “narrow wins”; and the white one, as before, is the new Speaker. Lastly, the constituencies in dark grey are those Tory gains whose previous MPs, by the time of the election, had left the Labour Party and were campaigning against it, in one case actually calling for a Conservative vote. I don’t know how much weight to give to this – it isn’t a factor one usually has to take into account – but it can’t have done Labour’s chances of holding those seats any good.

This map looks a bit different from the one above. West of the Pennines – in fact, West of the Derbyshire Dales – Labour’s losses are almost exclusively seats that had been lost and regained by Labour within the last twenty years (Bury North, Burnley) or narrow wins by the Tories (Ashton-under-Lyne, Hyndburn; Bury South, in grey, is also in this group). Again, I’m not saying that Labour didn’t lose all of those seats, or that we haven’t got a mountain to climb next time round. I am saying that a world in which thirty Labour strongholds have suddenly turned into Tory strongholds (say) is a different world, and calls for different strategies, than one where this has only happened in ten Labour strongholds (but the Tories have squeaked a win in another ten, as well as re-taking ten Labour marginals).

The idea of the Red Wall – as we’ve come to know it – makes a great story: Labour heartland voters, left stranded by the decline in heavy industry and the rise of social liberalism, abandon their decades-long loyalties to vote for the party of Britain and Brexit! But we need to deal with the problem we’ve got, not with a problem that would make more dramatic sense. I’m not saying there were no “Red Wall” seats (in the sense of that label that’s now accepted) in the 2019 election; on the contrary, some seats fit the profile very well. The question is how many there were, and how typical they were – both of the seats Labour lost and of the seats where Labour needs to campaign. The question is, to put it bluntly, whether the true “Red Wall” seats were Labour’s heartland at all – and, if not, whether basing party strategy and policy on the imperative of regaining those seats might do more harm than good.

I’ll go into this in the next post. For now, here’s another map. Rather than being based on Tory targeting, this one shows all Labour seats in the region going into the 2017 election, including the ones we held. (Plus two – Birkenhead and Nottingham East – that we took back from squatters. Frank Field and Chris Leslie both stood in 2019, under their new colours; relative to 2017, when they held those seats for Labour, their votes dropped by 59.7% and 67.9% respectively. Cheers, guys, nice knowing you.)

Yes, that Red Wall of Labour constituencies is certainly… well, it’s got some gaps in it now, so there’s that…

Funny how it’s mostly the big square constituencies that go to the Tories while the small squiggly ones stay Labour, isn’t it?

One last map: here (below) is London. One loss (courtesy of Tory defector Sam Gyimah and a really disgraceful campaign to portray a divisive third-place candidate as a unifying winner); one seat reclaimed from a squatter (Mike Gapes, whose vote dropped by a magnificent 68.5% compared to 2017); and one gain. (Since the gain is Putney, which hadn’t been Labour since 2005, this offers a partial and belated confirmation for James Kanagasooriam’s original analysis: he’d cited Putney as an example of the kind of seat that the Tories held despite its demographics suggesting it “ought to” be Labour.)

There are also, of course, an awful lot of small squiggly red holds. Perhaps it’s not so surprising that Conservative strategists, and Conservative-leaning commentators like Balls and Kanagasooriam, preferred to concentrate on the wide open spaces of the Northsouth Yorkshire and the East Midlands, and make out that the Tories were winning the seats that really matter. There are – or were – gains to be made up there, of course; there was a Tory majority for the taking, and by God they took it. But if you’re looking for a real Red Wall – if you’re looking for “traditional Labour seats” in “working-class areas”, and plenty of them – look at how, even now, you can go from the Fylde Coast to Saddleworth Moor without setting foot in a Tory constituency, or cross the whole of Greater London (either way, Heathrow to Dagenham or Barnet to Croydon).

Despite what had happened in the true (and increasingly ironically named) “Red Wall” areas, and despite how disastrous the results were overall, Labour’s starting point after 2019 should have been (in the immortal words of Granny Weatherwax) “WE ATEN’T DEAD”. The big “Red Wall” losses – the purple constituencies – were serious, but they weren’t typical; by treating them as if they were, we ran the risk (at best) of losing momentum and having to run to catch up when people began to mobilise again, at worst of facing in the wrong direction and becoming irrelevant.

PS On a side-note, I’ve referred a few times to former Labour MPs who stood against Labour in 2019. Here are all of them.

Name2017 %2019 %
As TIGAs LDAs Ind
Luciana Berger79.531.9
Frank Field76.917.2
Mike Gapes75.87.3
Roger Godsiff77.68.1
Chris Leslie71.53.6
Gavin Shuker62.49.2
Angela Smith45.811
Chuka Umunna68.530.7
Chris Williamson48.51.4

Chris Williamson’s performance must be some sort of record for former MPs standing in their own former constituency. Then again, it’s only a drop of 46.9%. In that respect, Roger Godsiff – the only MP actually to be deselected under the Corbyn leadership, trivia fans – leads the pack with a 69.5% plummet in his personal vote; the last two TIGers aren’t far behind, though. Which is worse, Mike Gapes dropping 68.5% of his vote, or Chris Leslie losing 67.9% and his deposit? Either way, those are quite the performances. Even in 1983 – the big shakeup which the ChUKers were surely hoping to emulate – the biggest comparable fall I could find was from 61.5% to 11% (Arthur Lewis, Newham North West). (Labour retook the seat comfortably; nice to see that some things don’t change.)

On a side-note to the side-note, it’s rather striking that only one of the eight ex-Labour members of Change UK – Angela Smith – had received less than 50% of the vote at the previous election; indeed, only one other (Joan Ryan) had had less than 60%. The role of safe-seat complacency in decisions to defect – not to mention vanity and general Ned Lagg-ery – shouldn’t be understated.

But we should probably get back to that Red Wall.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: