In search of the Red Wall (8)

THE STORY SO FAR: There is no “Red Wall”.

To be more precise, it’s not possible to identify any group of constituencies that fit all the criteria for the “Red Wall” as it’s usually described (large numbers of previously long-term and solid Labour seats which went Tory in 2019, clustered together, somewhere in the North of England). The Tories did win in some surprising places in 2019, with constituencies which had been Labour for twenty years or more suddenly showing a 10-20% drop in the Labour vote. Long-term patterns of voting in these constituencies suggest that a disaffected, “none of the above” voting bloc has played a significant part for some time, and that the Tories’ success in 2019 was largely due to the capture of this anti-political vote by the Brexit Party and (in a bizarre irony) by the Conservative Party itself.

Here (again) are the voting patterns for one of these constituencies, Don Valley in south Yorkshire.

It’s not hard to see what’s going on. A sizeable Liberal Democrat protest vote in 2005 is joined by far-Right protest voting in 2010, eating into Labour’s normally substantial majority. In 2015 the Lib Dems had been discredited by their participation in the Coalition. The benefits go in part to Labour; there is also a new repository for antipolitical voting in the form of UKIP, which in addition gains support from right-wing Tory voters. Polarisation returns in 2017, with the Tories capturing most – although not all – of the UKIP vote. This means that their position relative to Labour improves again, putting them in a good position to capitalise when the next wave of antipolitical protest voting – spearheaded by the rebadged Brexit Party – peels votes away from Labour in 2019.

Disaffected voters express their opposition to both the “old parties” by voting for the Lib Dems in 2005; for the Lib Dems, UKIP and the far Right in 2010; for UKIP in 2015; and for both the (Brexiteering) Tories and the (anti-system) Labour Party in 2017 (but mainly for the Tories). Consequently, Labour’s position relative to the Tories goes Down between 2005 and 2010, goes Up between 2010 and 2015 and finally goes Down again in 2017.

For completeness’ sake we can also consider what happened in Bishop Auckland:

Which is… pretty much the same, except that Labour didn’t noticeably benefit from the Lib Dem collapse in 2015 – either because the Lib Dem votes they gained were mostly matched by losses to UKIP, or because antipolitical Lib Dem votes transferred to UKIP direct – with the result that Labour lost ground with the rise of UKIP as well as with their decline. Labour’s position relative to the Tories went Down in 2010, Down in 2015 and Down in 2017.

We see these patterns – particularly the first one – in a lot of the seats Labour lost. It’s a plausible, coherent story, too – the two motors of the whole process are the alienation of a substantial body of voters from both main parties and the failure to completely delegitimate the far Right, and both of those conditions seem likely to have applied in any number of places (particularly places with longstanding Labour councils).

The question then is: do these conditions obtain more widely? They may be scattered around the country and they may not have decided the election in themselves, but are the “red wall” seats just the most visible part of a larger problem? Bluntly, are the seats Labour actually lost in 2019 the tip of the iceberg?

There isn’t much in the way of good news in this story – sitting as we are amid the wreckage of the 2019 election – but the answer to this question does at least qualify as interesting news. Here are the details for the 418 English constituencies which have existed with more or less the same boundaries since 1979, and which were held by the main two parties in 2017.

Reading from the top, the dark red bloc are ‘UU’ seats, those where Labour’s position relative to the Tories went Up in both 2015 and 2017; the paler red are ‘DU’ (Down in 2015, Up in 2017); the pale blues are UD (Up in 2015, Down in 2017); and the dark blue, DD (Down in both 2015 and 2017). I haven’t distinguished between UUU/DUU, UDU/DDU, UUD/DUD and UDD/DDD, partly for simplicity and partly because including that level of detail would make very little difference. (Labour’s position went Down in 2010 in 387 constituencies out of the 418; the UUU subset accounts for 29 of the remaining 31 and UDD and UDU for one each of the last two.)

What do we see? We see that, in 2017, Labour’s position improved in over 70% of Labour constituencies and nearly 80% of Tory seats. Labour seats were considerably more likely to show improvements under both Corbyn and Ed Miliband than Corbyn alone; in Tory seats, by contrast, over 40% – the largest single category – saw Labour’s position deteriorate under Miliband and then improve under Corbyn. If we’re measuring the popularity of party leaders on the basis of the ability to improve the party’s vote share, particularly outside its existing heartlands – an eccentric idea, perhaps, but let’s go with it – then Jeremy Corbyn, as of 2017, was far and away the most popular leader Labour had had since 1997. I don’t recall this point being made very often at the time.

More importantly for the current discussion, where are those DUDs and DDDs? They are there, but – as it turns out – they’re not all that numerous. 39 Labour seats, around 20% of the total, fitted the DUD template that we saw so many times in the previous post; the DDD model only fits another 11. Nor was the pattern any more representative of trends in Conservative seats. 320 out of 418 seats – more than three-quarters – saw Labour’s position improve in 2017, under Corbyn; in 203 of them – nearly half – there were improvements in 2015 as well. Pitching to the DUD seats – and the disaffected 15-20% of Right-leaning antipolitical voters who made them that way – is no way to either gain votes in most Tory seats or hang on to them in most Labour seats.

Time for another map or two, or four. Here’s the north and centre of England, showing Labour and Tory seats where Labour’s relative vote share went down in 2017 (red for Labour, blue for Tory, keep up). There are fifty Labour seats in this category, about 45 of which you can see here.

And here are the Labour and Tory seats where Labour’s position improved in 2017 (slightly different red for Labour… you get the idea). There are 136 Labour seats in this category, of which you see about 65 here.

(Small squiggly constituencies, remember. There are twelve constituencies in the block running north-south from Doncaster to Bolsover, 40 in the east-west block from Birkenhead to Leeds.)

And here’s the south-east. Again, these are Labour and Tory seats where Labour’s relative vote share fell in 2017.

And here are the Labour and Tory seats in the south-east where Labour’s position improved in 2017. (You may want to sit down.)

Small squiggly red constituencies, again – forty of them in London, another ten in Birmingham. As for the blue ones… well, that’s a bit striking, isn’t it? I’m not saying that Corbyn was building a platform for power in the heart of the Tory beast – in a lot of these cases what happened in 2017 was that Labour came third with 10% instead of 8%. But, as I said, if we measure success for a party leader in terms of putting on vote share for the party…

NEXT: So, what did happen in 2019 (and where did it happen)?

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: