There is a further exchange between Rob Hayward and Lewis Baston on Conservative Home about what the results of the boundary review may or may not look like. Meanwhile Democratic Audit have put up a lot more detail on their proposals on their website here.
As I’ve mentioned before, I’ve been playing with possible boundaries myself (in my case it’s mainly as a distraction to doing the preparatory work for calculating notional figures when proper proposed boundaries appear) and came up with very different overall figures to Lewis – I haven’t done every region yet, but so far have Labour losing about 18 seats more than the Conservatives (and haven’t done Scotland yet!)*. That’s not to say that my proposals are in any way more likely to reflect what the boundary commissions come up with than Lewis’s are – only that it is possible for different people to come up with what they think are plausible and non-partisan boundary changes…but have significantly different results.
A few further observations though – when I started looking at the detailed differences between my guesses and Lewis’s, they were oten quite minor, counties where we had both come up with the same basic layout in seats, but had put just a couple of wards in different places leading to different notional winners for seats. Those very marginal differences aren’t really important (indeed, as Lewis has said, despite the positive headline figures for Labour in their projection, underlying that there are fewer winnable marginals for Labour making the boundaries worse for them). Of course, there are also some areas were my own ideas of what is likely are very different to Lewis’s.
Rob Hayward makes a couple of criticisms of Lewis’ projections in the article. In terms of split wards, Democratic Audit have used a reasonable amount of them (so have I), Rob suggests there are too many and that the boundary commissions really will avoid them at all costs. He suggests there may end up being virtally none of them. We know from the Commission’s guidance that they will only be splitting wards in exceptional and compelling circumstances, but have no idea how high the Commissions will set that bar in practice.
As far as I can tell splitting wards probably can be avoided in most cases if you work hard enough at it…but that doing so sometimes requires sacrifices in terms of making seats that look rational, that don’t split small towns, that don’t cross too many boundaries and so on. Personally I think it better to split wards sometimes, but the Boundary Commissions don’t make their decisions based on my preferences! Whether these issues count as exceptional and compelling in the eyes of the Commissions remains to be seen – there is some logic in what Rob says… whether or not such arguments are compelling, they are likely to be so common that it’s going to be a push to say they are exceptional.
Rob then picks up a couple of odd-looking constituencies on Lewis’s model – I think that an almost wholly rural North Lancashire isn’t wholly incredible, but agree with Rob that linking Powys in Wales westwards across the mountains rather than northwards looks unlikely (though I’ve not seen a plan for central Wales that doesn’t have to resort to something clumsy). I’ve already said that I think linking Suffolk to Essex is unlikely when you can divide Essex up reasonably enough without the link, leaving minimal change in Suffolk.
The point that Rob ends on is also the thing I think is most unlikely in Lewis’s recommendations – there are some instances where Lewis’s proposals produce seats that cover wards from four or more local authorities, sometimes in different counties. The Bolsover & Ollerton constituency that Rob highlights is indeed a shocker! I think Lewis’s Mid-Kent (the M20 seat!) is pretty horrid too, but then, I’m a Kentishman and I’m sure any plan I came up with for Lewis’s back-yard in Camden would equally ring all sorts of alarm bells for him.
All that aside, the original purpose of Democratic Audit’s paper was to make a plausible projection of what the Boundary Commissions may suggest. There are some odd bits in there, but I expect the Boundary Commission’s actual proposals will have a few odd bits in them too. I’d be very surprised if the Boundary Commissions proposed something like Lewis’s Bolsover & Ollerton, but it wouldn’t surprise me if they came up with something just as odd somewhere else in the country that none of us could possibly have guessed (as Lewis mentions – last time round they suggested a Wallasey & Kirkdale seat that crossed the Mersey Estuary, so they have form for somewhat surprising proposals!).
(*Someone will ask soon if I’m going to release them. I don’t know – I may set up some specific threads for discussing county or regional boundary changes in the constituency guide part of UKPR and stick them up there at some point. I don’t intend to make a big fuss though – if nothing else, I haven’t even started collating data for Scotland and by the time I get round to that we’ll probably be on the final straight to the real provisional recommendations!)