I stumbled on this Tweet the other day.
The chart referred to is a list of the top 60 Tory targets in the 2019 election, ordered by the size of swing required (which effectively means by the size of the incumbent party’s majority over the Tories, but the figures look smaller). 24 are Labour seats, in England, with majorities of less than 5%; in the event the Tories took 19 of them (step forward Battersea, Bedford, Canterbury, Portsmouth South and Warwick & Leamington – good work, lads).
Here are all the gains the Tories made from Labour in England – well, almost all; the map doesn’t extend far enough south to show Ipswich, Stroud and Kensington, so there are 45 seats here instead of 48.
Following Dan’s suggestion, the colour coding on this map based on how many successive elections each constituency had had as a Labour seat when it was taken by the Tories. Pale blue are 2017 Losses, seats that Labour took in 2017, or in two cases in 2015. (Labour made quite a lot of gains in 2017, didn’t they? Wonder if anyone’s drawn any lessons from that.) The deep blue are Ye Olde Laboure Heartelandes, seats that had been Labour since at least 1983; most of them go back to February 1974 (the first general election under the current franchise). Medium blues are Labour going back to 1997 (or in one case 2001); lastly, greys are seats whose Labour election count stood at zero, as they had gone into the 2019 election with an MP who had already left Labour – and who was, in all these cases, actively campaigning against their old party. (I’m referring to John Woodcock, John Mann, Ian Austin, Ivan Lewis and Angela Smith. Any resemblance between this list and a list of “absolute dangers expelled from the Labour Party” is for the reader.) I don’t usually set much store by the Ned Lagg Effect – people tend to vote for a party, not an individual, as individuals ranging from Jim Sillars to Ivan Lewis have discovered to their cost. But 2019 wasn’t a normal election; in a campaign one of whose dominant messages was Are You Going To Hand Britain Over To Terrorist Communist Traitors?, the discovery that your own (formerly) Labour MP was actually endorsing the whole Communist-terrorist thing must have shifted a few votes in those constituencies.
First impression: there’s a lot of pale blue. There’s also a fair bit of deep blue, but it’s scattered all over the map and consists very largely of spread-out, semi-rural constituencies. But we can do better than that. Here’s a map from the previous Red Wall post – now revised and updated, incidentally, and featuring the definitive description of the Red Wall courtesy of its original inventor (tl; dr interesting, but I’m still not impressed). This map has 51 constituencies on it: the (only) 50 Labour constituencies where the Labour vote share went down in 2017 relative to 2015, and Scunthorpe (I’ll explain why Scunthorpe in a minute).
Edit 13/2 Thanks to the reader who pointed out that I’d misidentified Scunthorpe as Hartlepool. No idea how I did that – it’s not even on the coast! Corrected.
The colour-coding here is the version used in the previous post: the deep blues are long-term Labour seats where the 2019 Tory majority was 5% or more and the Labour vote had fallen by 10% or more relative to 2017 and the Labour vote was lower than at any time since (and including) 2001; the mid-blues are other long-term Labour seats that went to the Tories in 2019, while the reds are seats that Labour held in 2019. (Scunthorpe, in purple, is a ‘deep blue’ seat that doesn’t strictly qualify to be on this map, as Labour’s vote share rose between 2015 and 2017 – by half of one percent.)
Now, let’s tidy up and simplify a bit. I said at the top of the post that there were 24 seats where Labour’s margin over the Tories was 5% or less, and that the Tories took 19 of those. Let’s say for the sake of argument that any party having a good election is likely to have successes at that kind of level: if what we want to explain is why the Tories got such a big majority, or why Labour’s seat total fell so low, the sub-2.5% swing seats aren’t the place to look. So we’ll eliminate those 19 seats from the first map, to give 29 gains instead of 48, of which the map shows 29 instead of 45 (the three southern seats omitted from the map are among the 19).
As for the “long-term Labour, vote share down in 2017” map, let’s take out the Labour holds – we’re not interested in those right now – and, again, take out the 19 sub-5%-majority seats. We’re left with a fairly sparse map showing only 20 seats.
And here are those two maps.
Spot the similarity.
As I said, there are 20 seats on the right-hand map and 29 on the left; the set of Tory gains from Labour in England overturning a majority greater than 5% isn’t identical with the set of long-term Labour seats where Labour’s vote share fell in 2017. But it’s close. The left-hand map (Tory gains against a >5% majority) includes all 20 of the seats in the right-hand map (long-term holds, relative vote share down in 2017); of the remaining nine, five are pale blue (only taken by Labour in 2017), three are mid-blue (1997 gains) and the ninth is grey (step forward Ivan Lewis).
My conclusion here is pretty much the same as the conclusion to the previous, big post (have you read the big post, by the way? recently? it’s revised and updated, you know). In five words, Red Wall: real but small.
The phenomenon people refer to as the Red Wall was the unexpected, large-scale loss of Labour votes to the Tories, apparently caused by long-term Labour voters deciding that they’d liked Labour in the old days but they couldn’t be doing with all this here political correctness, and taking place in the Norththe North-East and North-Westthe North-East, parts of the North-West and parts of Yorkshirethe North-East, parts of the North-West and parts of Yorkshire, the East Midlands and some places around Birminghamthe North-East, some of the more rural parts of the North-West and Yorkshire, the East Midlands, some places around Birmingham although not Birmingham itself, and also Stoke
a whole bunch of places which really don’t have much in common other than being south of the border and north of Luton. I’m caricaturing, but I do actually think this is a real phenomenon: look at those two maps. But it’s only one phenomenon, and it wasn’t what won the 2019 election for the Tories – arguably it was only because the Tories were already winning the 2019 election that the Red Wall effect really kicked in.
If we’re interested in the Red Wall phenomenon, we’re interested in something that (a) genuinely happened and (b) happened up and down the country, but (c) only happened in a small number of places. Labour needs to make a lot of gains next time round, but whether it needs to make precisely those gains is more debatable – and whether the kind of Labour campaign that would win back Ashfield and Great Grimsby would win the country is very dubious indeed. Apart from anything else, look at the sub-5%-majority places that Labour did hold in 2019 – Portsmouth South, Bedford, Canterbury; look what happened to Labour’s vote share in 2017 in the south-east (scroll down, and brace yourself). If you were thinking tactically for Labour, which area would you concentrate on – the one where Labour lost vote share despite intensive campaigning and national media attention, or the one where Labour gained vote share with hardly anyone even noticing?
So if we are interested in the Red Wall phenomenon, at this stage we’re interested in it partly for purely historical reasons (something unusual did happen in those seats), and partly on a secondary tactical level. Nobody should be asking “how might learning from the Red Wall be useful for Labour?” – but “what errors might the belief that the Red Wall is useful for Labour lead to?” is an interesting and potentially useful question, as is “what biases and presuppositions are likely to have led people to believe Labour should learn from the Red Wall?”. And I think the answer is going to come from a closer look at those 50 seats. (Or 51 if you count Scunthorpe.)
Workers’ Playtime 