Reform UK: the jigsaw piece that doesn’t seem to fit
There’s been a lot of attention given to the rise of Reform UK in the polls. There’s been surprisingly little effort given to considering where they might most usefully direct their efforts.
Despite a name change, Reform UK is not a new party. In 2019 it stood as the Brexit party. It tallied 2% of the UK-wide vote. However, it only stood in 275 seats, so this considerably understates the actual vote share that it would have received.
So where did the Brexit party actually stand and how did it do? It decided not to stand in any Conservative-held seat. It stood in most (but not all) Labour-held seats, most (but not all) Lib-Dem held seats and some SNP-held seats.
This produces a very lop-sided map. I’ve created a heat map which I’ve put at the top of the page. If you want to look at this interactively, you can do so here.
I’ll look at the hot spots in a minute but for now I want you to concentrate just on the geography. Vast swathes of territory that you would have thought would be promising for the Brexit party saw no candidate stand. No seat in Kent, no seat in Essex, no seat anywhere further southwest than Bristol South other than Exeter and Plymouth Sutton & Devonport. The only seat they stood in in Lincolnshire was Lincoln, surely the least promising prospect for them in that county. The shires, stuffed with Conservative-held seats but also with Brexity types in ready supply, were untapped.
On the other hand, the Brexit party stood in most London constituencies, an area that has always been difficult territory for them. They held their deposit in just three of them. In more than half the London seats they stood in, they got under 2% of the vote.
If we pro-rate the Brexit party’s vote share up to the 632 seats in Great Britain, we would credit them with a national vote share of 4.6%. But for the reasons I’ve just given, I believe that substantially underplays their strength of support. We can’t know with certainty how people would have voted if they had a Brexit party option in front of them at the time. I find it hard to credit them with a support base of less than 5% nationally at that time and 6% feels more reasonable. You might even go higher.
Now this came as a bit of a surprise to me. What also came as a surprise to me is where the actual vote strength was concentrated. They kept their deposits in 111 seats. They broke 10% in 37 seats. In descending order of vote share, these are:
Again, have a look at the map. These are clustered around South Yorkshire, the Welsh valleys, the North East and to a lesser extent Greater Manchester.
Now for the reasons I’ve already given, we don’t know whether the Brexit party would have had other areas of strength. From the performance of UKIP in the past, you have to suspect the Thames estuary might have been fruitful for them.
But that doesn’t mean we can draw no conclusions. Perhaps surprisingly, the Brexit party didn’t perform particularly well in the West Midlands. Nor did it do quite as well in the North West. If you’d asked me to guess without checking, I’d have imagined Blackpool South would have been on this hotlist — a forlorn coastal town constituency with a strong Brexit vote. But in fact the Brexit party managed just 6.1% in that constituency — only the 85th best constituency for them. Bear that in mind when the by-election results come out.
Anyway, that was then. What now for Reform UK? In the Sky News poll tracker, they’re currently tallying 12% in the averages. That’s roughly double the latent support that the Brexit party had in 2019. I’ve expressed my doubts about this in the past but let’s take this at face value for now. At the very least, it sets Reform UK’s reasonable aspirations.
We don’t have a clear idea on where Reform UK’s support is distributed and the problem is accentuated by the fact that we don’t have all that clear idea how their support was distributed at the last election. We have information even if it’s imperfect. We can use that.
Compounding the problem, we have new boundaries for the next election. The Rallings & Thrasher notional results can be perused here.
Polls tell us that roughly two-thirds of their current support comes from 2019 Conservative voters. Polls also tell us that roughly 20% or more of 2019 Conservative voters are now professing to support Reform UK.
OK, but what might this mean? We’ve had a series of MRP polls, all of which have predicted that Reform UK will get no seats. Very possibly, but the moment you look at the detail of the MRP polls, you realise that they look dubious when it comes to handling Reform UK vote share.
Let’s take a glaring pair of examples. YouGov’s latest MRP poll predicts (on a 12% national vote share) that Reform UK will get 27% of the vote in Barnsley North and 25% of the vote in Barnsley South. Survation’s latest MRP poll predicts (on an 8% national vote share) that Reform UK will get 17% of the vote in Barnsley North and 23% of the vote in Barnsley South.
But the Rallings & Thrasher notionals for 2019 on my estimated 6% national vote share gave them 29.5% in Barnsley North and 29.2% in Barnsley South. Is it really credible that Reform UK will go backwards in one of its hotspots at a time when its national vote share looks set to take either a small step forward (if you believe Survation) or a large step forward (if you believe YouGov)?
I’m not going to try to create a shadow MRP poll. First, I don’t have the data. Secondly, I don’t have the expertise. I am going to make the simple observation that if the MRP polls are struggling to cope with this (I am sympathetic, it seems pretty hard to model given the imperfect information that we have), we should emphatically not place much weight on their other conclusions either.
As for Reform UK, it seems to me that they need to adopt a very regional approach. And with their support base drawn strongly from former Conservative voters, they should be looking closely at seats that could be made into three way marginals. If the cards fall right and their vote share turns out as the polls suggest, they might just nick a few.