Overnight the English and Welsh boundary commissions released their initial proposals for the boundary changes – the English proposals are here, Welsh proposals are here. If you missed it, I discussed the wider background to the boundary changes and how they work yesterday. The initial proposals are very much as expected – we’ve known the broad strokes of the review since the rules were set in 2011 and the electorate numbers were released in the spring. The review will cut the total number of MPs from 650 to 600, including a reduction from 533 to 501 in England and a reduction from 40 to 29 in Wales. The English losses are disproportionately in the North, where the population is falling relative to southern England, and will consequently hit Labour harder than the Tories. Today’s proposals are about the details – finding out what the specific proposals are and what the specific impact will be.

As usual, I’ve worked out notional projected figures for what the new boundaries would mean at a general election (using a similar method to that used by Rallings and Thrasher for the official notional figures they will produce for the media once the final report it out). The changes are, as ever, a case of swings and roundabouts. A net total of 43 seats are abolished across England and Wales, but the impact is more than just those abolished and created seats, other seats may notionally change hands as wards are moved from one seat to another. Just because a seat held by a party is abolished, it doesn’t mean that party necessarily looses out as parts of that dismembered seats may push neighbouring seats into their column.

Even defining which seats disappear can be debatable (especially when there are contentious selection battles to come). In England there are around 35 existing seats that don’t really have a seat which is its clear successor, 12 of these are Tory seats, 23 are Labour. Looking at the knock-on effects elsewhere, there are around 16 seats where the Conservatives would notionally have won seats they don’t currently hold, 13 seats where Labour would have won seats they don’t currently hold. In Wales the changes are more drastic, given the sharp reduction from 40 seats to only 29. By my calculation Labour lose 7 seats in Wales, the Conservatives lose 4.

There are only three seats in England and Wales that the Liberal Democrats would have won on the proposed boundaries: North Norfolk, Westmoreland & Lonsdale and Ceredigion. I should add a caveat to this though – these projections are purely an attempt to work out what the result would be if the votes cast at the last election had been counted on these new boundaries. It’s purely an accounting exercise, estimating what the general election vote was in each ward, allocating them to their new seats and totting them up. It doesn’t account for the fact that people might have voted differently if they’d been in a different seat – for example, it’s possible that people who voted Tory in Penistone East ward in 2015 when the ward was in Penistone & Stocksbridge might instead have voted tactically for Nick Clegg had their ward been in Sheffield Hallam. For that reason I think such notional figures can sometimes underestimate the Lib Dems. Note that the Liberal Democrats have suggested they might win Cambridge on the new boundaries – this is not unrealistic at all, on my notional figures Labour have a majority of under 200 there.

Aside from Nick Clegg, several other high profile figures face seeing their seats abolished or made tricky to win. Looking at the Conservatives, George Osborne’s Tatton seat is abolished (the largest part of it goes into David Rutley’s Macclesfield seat), David Davis’s Haltemprice seat also goes, mostly into the new Goole seat, the successor to Andrew Percy’s Brigg and Goole. Priti Patel’s Witham seat forms the smaller part of a new Witham & Maldon seat, mostly made up of John Whittingdale’s Maldon seat. Jeremy Wright, the Attorney General, also sees his Kenilworth and Southam seat abolished. Some media reports have suggested Justine Greening may also be in trouble: she’s not, her seat has Wimbledon Common added to it but will still be very Conservative.

For the Conservatives, MPs in danger of losing their seat poses a risk to getting the boundary changes through, and it’s likely the party will try to use retirements, peerages and so forth to ensure no current Tory MPs lose out. For Labour MPs the boundary changes take place against a different background and MPs who have opposed Jeremy Corbyn may face difficulties in selections if there are local boundary changes. High profile MPs whose seats are abolished include Tristram Hunt in Stoke Central (divided between Stoke North and South), Ian Austin in Dudley North (his seat mostly ends up in the new Dudley East & Tipton, the successor seat to Adrian Bailey’s West Brom West), Vernon Coaker (whose Gedling seat is divided between Chris Leslie’s Nottingham East and Conservative-held Sherwood). Jeremy Corbyn’s own seat sees substantial changes: Diane Abbott’s Hackney North and Stoke Newington is abolished and split between a Hackney Central seat (mostly made up of Meg Hillier’s Hackney South) and a Finsbury Park and Stoke Newington seat (mostly made up of Corbyn’s Islington North), with a much smaller part going into a new Hackney West and Bethnal Green seat.

I’ll put up full notional results later on today, but looking at England and Wales as a whole, it looks as if the Conservatives would lose 10 seats, Labour would lose 28 seats, the Liberal Democrats would lose 4 and the Greens would lose one (by my calculations the new Brighton North would be a close three-way marginal, with the Conservatives narrowly ahead of Labour). That means at the last general election the Conservatives would have won a majority of 40, rather than the majority of 12 they actually got.

I’ll release full notional figures for each seat later on today.

32 Responses to “Who wins and loses on the new boundaries”

  1. Quite strong suggestions from Laura Kuensberg on the BBC that the Boundary changes are unlikely to be approved.

  2. Afternoon everyone. I was about to comment on the old thread, but just noticed the new one, so I’ll just comment here instead :)

    I haven’t been around much lately, mainly because I decided to take the summer off from political buffoonery – I just had enough of fact-free bilge around the time of the referendum. Thanks AW for your two fascinating posts on the boundary recommendations, as well as the comments BTL on the previous thread.

    There has been a fair bit of discussion about whether it’s a good idea to reduce the HoC to 600, and whether the new boundaries constitute gerrymandering in favour of the Conservatives. However, the extent of the proposed changes really do mean that most incumbent MPs will have to attract new voters from the wee bits tacked on to their constituencies. What’s not quite clear to me yet is whether the new boundaries increase or decrease the geographical polarisation that we’ve seen since 2001 – where Labour have been losing ground in the South of England while gaining vote share in the North of England.

    I’m sure there will be some comments about how unfair FPTP is and that we need to move to a better/nicer/fairer/superer electoral system, but I think we’ll have to wait for another hung parliament for that to occur. If I’ve time later, I’ll post my tuppence-worth on electoral reform (just for fun….).

  3. Ah, I’ve a post in mod – I used the term related to the manipulation of electoral boundaries for unfair advantage which sounds like a lizard, and said that some people had accused the government of such a heinous act…..

  4. Thanks Anthony for the post.

    Merseyside loses two seats under the proposals. Walton would be partitioned – it would be interesting as the current MP likely wins the metro elections. If he steps down, then there would be an MP for two years (it doesn’t make sense to me).

    Two Wirral constituencies merged (one was just gained from the Conservatives). Both MPs are Labour, and their political orientations are quite different.

    Anthony’s point about movement of wards is quite important.

    When starting to read the redrawing Wavertree, first I thought that it could give the LibDems some chance, but then I got to the end, and I saw what else the constituency would get, and hence Labour will likely hold it easily. The same apply for West Darby.

    The redrawing of Southport (into Lancashire) could make it more marginal between LibDems and Conservatives I think.


    “As somebody who has always desired a small state sector I’m still waiting for a Government that will make us * lean and mean” have been all my life :-)”


    Well the Govt. have at times been “mean”, so that’s halfway there already innit. (“Lean” prolly won’t happen till Boomers don’t need keeping happy any more, but it’s already getting leaner for assorted peeps!!…)

  6. It would be a real shame to lose Caroline Lucas, she’s a real asset to the HoC and political life in general. I really shouldn’t hope that lots of labour supporters will vote for her even if there isn’t an electoral pact.

  7. @AW (or anyone who knows).

    I’ve a quick question about your determination of who would win certain seats (notionally). Do you just use information from local elections (where the votes are broken down by ward anyway) combined with the full parliamentary consituency results from 2015, or are the Parliamentary results available by ward. It might seem a stupid question, but ballot-box by ballot-box vote breakdowns are typically available in Ireland (via tallying), in Germany (they’re printed on the Internet), and the US (by precinct reporting). Is this information publicly available in the UK (I know you can pay to obtain the marked electoral register)? Forgive my ignorance, I’ve been away for a while :)

    [They don’t exist. Votes from multiple ballot boxes are counted before they are mixed, so no official numbers per ward exist. Party observers do carry out informal tallies at some counts, of course, but these are not easily available. When I first did this sort of exercise a decade ago I did include box counts that various party activists provided me with, but I decided the patchy availability and reliability skewed the numbers too much so I stopped – AW]

  8. CR

    I think what AW says about the Lib Dems would also apply to Caroline Lucas. Neighbouring area which didn’t vote green because that wasn’t the election in their area would now have the option and so we should expect her to outperform her notional result.

    Ultimately it’ll be up to the people of Brighton North to decide if they agree with you or not.

    [I think it does too, and I think Caroline Lucas may very well have won. People in wards moving into Brighton North would probably have been more likely to have voted Green if they’d been in Brighton Pavilion. As I say, this is purely an accounting exercise though, I’m not second guessing if people would have behaved differently – AW]


    Returning officers may choose to disclose ward level votes, but they are not required to do so. There are ward level counts, but not everywhere.

  10. LWVG

    Tallying isn’t the refined art in Britain in the same way that it is in Ireland, so reapportionment will be done based on local election results and a bit of guesswork. That is why different projections of boundary changes will produce slightly different results, as you will see in the next few days I suspect.

    The only ward-level breakdowns within larger elections that I know of, occur in the elections for the London Mayor and Assembly.

  11. Sorry, Anthony has already replied.

    In some countries the wards to the counting, they forward the data both electronically and on paper to the constituencies, these report to the national election organisation that can then release the results.

  12. The whole business of people having to go to local polling stations, make marks on bits of paper, and wait for these to be counted is so antiquated, archaic and expensive.

    A worthwhile reform would be to switch to online voting in time for the 2020 GE. There is already a secure system in place for registering to vote online, I can see no reason why this shouldn’t be extended to the actual vote itself.

    Not only would we get the results quicker, but there could be a full ward-by-ward analysis of the voting available immediately.

  13. Senior Conservative source says there a “60-40” chance the boundary changes WON’T happen.

  14. As the Lib Dem candidate for Tatton I have a can you all about what it means for George Osborne! Effectively the constituency is split in two with the north going to Altrincham (already has a sitting Tory with a solid majority) and the rest to Macclesfield (again extremely blue). It does make both constituencies rather large geographically and makes little sense on the ground to voters.

    If I was a betting man David Rutley the current MP for Macclesfield will be 60 by 2020 and offered a rather tasteful peerage so the seat can be vacated for George…

    As a candidate that’s completely losing their constituency I really am all for boundary reviews, what I really, really find deceitful is the reduction of MPs being sold to the public as ‘cost cutting exercise’ while the Tory’s appoint hundreds of new Lords at a cost of millions to the taxpayer.

  15. David Rutley has only been an MP since 2010 compared to 2001 for George Osborne, so I think it’s unlikely Rutley will stand down for Osborne. In fact the other way round might be more likely. In other words I think length of service is more important than age in this case.

  16. If, as is being suggested, enough Tory backbenchers rebel so that the changes don’t go through, there is an alternative solution often used by private firms wishing to shed staff.

    Quite simply, over say the next 10 years, whenever a sitting MP decides to retire, stand down, or dies, that constituency is broken up and areas allocated to neighbouring constituencies. Once that has happened 50 times, you’ve got your 600. It’s called natural wastage.

    The Boundaries Commission could then look at the size of the electorates in the 600 constituencies, and maybe shuffle things around a bit to make them more balanced.

  17. you are a wonder ,Anthony
    very grateful for your excellent work


    Not a Boomer myself so still waiting.

  19. BT SAYS… (FPT)

    […] the idea that [Cameron] didn’t think a ‘no’ vote was possible when he promised (and kept to his promise) a referendum, is laughable.

    I suspect that Cameron no more expected to have to keep his promise for a referendum than he bothered with the long list of other broken promises which someone quoted earlier. Like almost everyone else he believed the polls[1] that said he would have to rely on the Lib Dems again and they could be expected to veto any referendum as they had 2010-15.

    However even with a majority, polling and Cameron’s vanity would have colluded to convince him that he could win a referendum. Up to the 2015 election YouGov regularly asked a question:


    Imagine the British government under David Cameron renegotiated our relationship with Europe and said that Britain’s interests were now protected, and David Cameron recommended that Britain remain a member of the European Union on the new terms.

    How would you then vote in a referendum on the issue?

    While polling on a more general question was usually roughly evenly split (as the eventual result turned out to be) this post-negotiation question always produced a big majority for Remain (or equivalent) as Conservative voters backed their leader.

    But when YouGov started to ask more detailed questions after the election, it became clear that these switchers were expecting substantial changes in the UK’s relationship with the EU – something Cameron was never going to be able to deliver – in part because of his inept handling of the EU in the past and lack of caring how it operates. Cameron assumed that he would be able use his authority and charm to deliver these voters, but he couldn’t. The failure is all his.

    [1] We know that the private polling done for the Conservatives by Populus produced told the same story as the other polls. Of course Cameron’s majority was mainly produced by targeting Lib Dem seats, which undermined his safety net – a good example of how Cameron’s administration never seemed to think things through.

  20. Thanks again AW, interesting piece. It seems to me that a smaller Commons is desirable on cost alone. As for the changes to parties likely from the review that seems only fair to correct the current unfair advantage to one party.

    I see that Graham appears desperate for it not to happen. I would have thought there was a good chance of it happening myself but no doubt we shall see in the fulness of time.

  21. It seems that there is no mention of Carswell, whose constituency is moving a fair bit to include Harwich.

    The UKIP vote seems to be more broad and not respecting constituency boundaries and so I wouldn’t expect him to benefit as much from the effect that might save certain LD and Green MPs from the accounting exercise.

    How safe is he going to be under the new boundaries?

    It seems that Planet Thanet (I’m so sorely tempted to make a suggestion for this as a constituency name as opposed to the boring Thanet East) might well be an encouraging set of boundaries.

  22. The report makes it clear that they have focused on using existing ward boundaries as units to re-allocate depending on number of electors.

    But lots of local authorities are currently undertaking boundary reviews themselves that will change the ward boundaries in 2018 too (Birmingham, Leics, Cambs, Southwark, Bexley).

    For example, in Southwark, the Dulwich and West Norwood constituency is being give the South Camberwell ward – but the councils proposals see that ward being split up further itself, and the boundaries won’t align with the new constituency.

    Regards working out notional majorities etc., for London at least, is there any value in calculating the majorities based on the most recent 2016 GLA elections? They provide the data ward by ward, so you could add up each ward as it makes the new constituency to give a notional election outcome.

  23. I can’t see conservative MPs voting against this in sufficient numbers to stop this, very few of them are going to lose their seats as polling stands at the moment and those that do will be compensated richly. Its going to be very difficult to vote against party advantage, lots of down sides really very little upside. Rebels that keep their seats at the expense of losing a conservative majority govt could expect to be very unpopular

    The effect that this review will have on labours civil war is another reason why conservatives won’t vote this down, how perfect for them that divisons within labour are reopened just before the GE campaign.

  24. CR

    I think the fact that Corbyn, Abbot and Thornbury are going to have to come to some arrangement will be the icing on the cake.

    Another opportunity for Labour to handle things badly.

    I agree with the Tories riding high in the polls will lead some to think that even if their constituency becomes notionally marginal, they can still win comfortably.

    It’s only those whose seat is abolished entirely who might hesitate, although there seems to be a policy of finding a new seat for every MP who will need it through retirements/ennobling a few people to smooth things out. As long as the parachute drop isn’t too far from their previous constituency I doubt things will go too horribly wrong.

    If it turns into a free for all in the Labour ranks it could be interesting to say the least.

    I don’t think the process of normalising constituency sizes is as big a thing as it seems, a “normal” boundary review (which would have been due anyway) would typically move some seats from Labour to Conservative due to the current trends of internal migration. The bottom line overstates the effect from normalising which probably is about half the total effect.

    It certainly isn’t the end of Labour as a potential party of power, with the largest share of the vote they would still be likely to be the governing party.

  25. ToH

    To be fair, a lot of things are conflated in this review.

    1) unequal size of constituencies, which is obviously wrong
    2) the lack of a proper statue how to do it, when to do, on what basis (not just this one, but all)
    3) the reduction in the size of the HoC is a bit of a logical problem with the growth of population, the growth of the HoL, and the enduring growing differences. In my view, considering the arguments for the FPTP (they are possibly flawed, but I didn’t invent them), in particular the link to constituents, the argument would be for reducing the size of the constituencies. Alternatively, in absence of any logic, we could just have one constituency.
    4) there is an argument about the cut off point – well there would be always a cut off point, so I don’t agree. The real trouble is the contradiction between automatic registration and the British traditions (although not quite, considering the Act on Compulsory Registration of Marriages in the 18th century). Automatic registration would get rid of most of the problems raised, however, I do highly appreciate the lax Bristish attitude about it.
    5) it is really fascinating to read the commission’s explanations in some redrawings – there is no single principle, quite clearly different members used different considerations.

    So, in my view, it would be much nicer if the HoC set out a generic law on the frequency, basis, and methodology of redrawing the boundaries and as long as we have FPTP, setting the constituency size much smaller, perhaps 45,000 or so (it would probably also increase the homogeneity of the constituencies).

  26. David Carrod,
    “The whole business of people having to go to local polling stations, make marks on bits of paper, and wait for these to be counted is so antiquated, archaic and expensive. A worthwhile reform would be to switch to online voting in time for the 2020 GE. There is already a secure system in place for registering to vote online, I can see no reason why this shouldn’t be extended to the actual vote itself.”

    Paper voting is auditable and physically non-trivial to alter: it would be quite a logistical exercise to physically intercept millions of ballot papers without being detected at any point by observers or officials. It’s also beautifully anonymous as ballot papers can be identical and mixed together before counting, and the vote is personally placed in the box by the voter with no potential for being intercepted. In most countries you can also check the identity of the voter at the polling station by asking for a passport or driving licence.

    Online voting wouldn’t be so transparent: you either trust the result or not, as there are no recounts possible. Online wouldn’t be anonymous as internet votes are traceable to individuals’ voting IDs and/or IP addresses, and you would have to simply trust the government not to be keeping tabs on who voted for who (there would be nothing technically stopping them from doing so). Internet data doesn’t travel straight from one place to another, it goes through many hops between various connections and third parties, with lots of potential for monitoring/alteration by nefarious groups en route.

    It would be very worrying if online voting replaced the paper kind.

  27. Thanks AW (and others) for the info – however, I’m sure Anthony will be happy that his projections seem to be in line with others reported in the G. Of course, as a few people have pointed out, personal and tactical voting would of course change these projections (never mind changes in the national voting intentions!). On the actual reduction to 600, there is of course the point made by TOH above that it reduces cost (although this is hugely undermined by the continuously increasing HoL), but on the other hand, one can argue that with increasing population and severe reduction in the importance of local government, the workload of constituency MPs has greatly increased. The other argument whether boundaries should be determined on population or electoral roll has been had to death on these pages – my suggestion has always been for mandatory registration on the basis of National Insurance numbers, which would narrow the gap between the eligible electorate and the registered. Whether people show up or not, that’s their problem :)

    I promised earlier to annoy people with my idea of electoral reform – I hope the resulting post isn’t too long. I don’t accept that FPTP is a fair system – just because it’s been there for gazillion years and that it “produces strong governments” doesn’t mean it’s any good. On the other hand, a list system might be deemed to put too much power in the hands of parties. The AMS system has some attractive qualities, most notably proportionality, but it does undoubtedly create “two types” of representatives. A system which does retain a certain degree of proportionality as well as the constituency link is STV of course – however, as we saw (or at least I saw!) in Ireland this year, unless you have electronic voting, the counting can be very lengthy. In addition, with the very large constituencies that the UK would have, the sheer organisation of the count would be tricky (unless of course, you have machine counting).

    Ok, so I propose MMFPTP (multi-member first-past-the-post). You have multi-member constituencies, with, in principle, multiple candidates for each party. However, in contrast with bloc voting, each voter has *one* vote. For N seats, then the top N candidates get the seats – simples! You retain a personal link to the MP, you retain constituencies (although, they will be larger than the present ones), and you get a system that gives reasonable proportionality (if N is around 4 or 5), while reducing the influence of tiny parties (for N = 5, you would probably need around 12% to have a chance of winning a seat). You also have the glorious, but possibly unedifying, situation whereby candidates of the same party are battling against one another (vote management is very important). I’ve played around with this a bit over the last couple of weeks – of course, there’s a practically infinite number of ways of arranging constituencies.

    Here’s a quick example based on the 2015 results (apologies for any mathematical errors I may have made).

    I’ve designed a 5-seat constituency of Leicestershire East – made up of the current seats of Harborough, E Leicester, S Leicester, W Leicester and Rutland and Melton. Sorry for anyone who lives there who feels that these don’t belong together – it’s just an example. Of these five seats, Labour won 3 and the Conservatives won 2 at the last election. On total votes, the breakdown is: Labour 37.9%, Conservatives 37.2%, UKIP 12.8, Lib Dem 6.9, Greens 4.4, Others 0.7. Now, it would be crazy for Lab and Con to put forward 5 candidates each – 3 would be a more realistic number, since it’s possible with 4 candidates only to have 1 elected. On these figures, you would expect 2 Lab, 2 Con and a very close race for the fifth seat *if* the votes were evenly divided between the candidates (which is never the case in reality).

    I know this isn’t relevant to the present thread, and it has about as much chance of flying as the pig that Oldnat referred to in the previous thread, but I’d love to see even a local election run on this basis. On the other hand, the battle to become a candidate would make Labour’s present predicament seem like a cakewalk!

  28. New thread – same topic


    To make it a bit more complicated (though I like your suggestion) :-)

    the number of votes needed for a seat could be set. The non-winning votes then summed up, and reallocated according to the proportion of votes in the constituencies.

    So, in a 5 member region the required vote to get a seat is, let’s say, 35,000. Anything above it (but less than 70,000) and under it goes to a pot. If Party X got 60% in region 1 and 30% in region 2, then Party X’s candidate(s) in region 1 get twice as many residual votes than in region 2 (but would still need 35,000 votes).

    This would mean that the size of the HoC would not be set, but would depend on the number of people who bother to vote and the distribution of these votes.

  30. Re: Brexit means Brexit means ???? Etc.

    From The Times, page 2!!

    “Tens of thousands of foreign workers could get temporary visas to pick fruit and vegetables under government plans to prevent production shifting abroad because of Brexit… They need 75,000 workers a year because British people are unwilling to do the work…”

  31. Alan

    “It seems that there is no mention of Carswell, whose constituency is moving a fair bit to include Harwich.”

    I mentioned it on the last thread!

    Harwich joining Clacon will make it much more difficult for Carswell to win. Of the 20 UKIP Councillors on Tednring District Council,only one is from Harwich, the rest are from the Clacton part of the district. All other Harwich councillors are Labour or Conservative.

    The really ridiculous thing about the boundary change in Clacton is taking Jaywick out and putting it in a rural constituency!

  32. I remember Mars bars being 6d in approx 1970, so shilling for pound sounds like a good rule of thumb to me.


    “They need 75,000 workers a year because British people are unwilling to do the work…”

    A significant problem is that seasonal work is not a good option. What do you do for the other 6 months in the year. When I was a student many of my contemporaries spent their summers e.g. grape picking in France. The French didn’t want those jobs for the same reason.

    I see no issue with seasonal work being made available to workers from poorer countries, who may see the (from our PoV) low wages as being well worth it. The problem comes when they are exploited and effectively paid well below minimum wage.