Plus ça change – or déjà vu all over again: the proposals for new, and fewer, parliamentary constituencies

ron johnston

Proposals for new parliamentary constituencies have now been published by three of the four UK Boundary Commissions. Ron Johnston examines the nature of those recommendations and their likely impact, on both individual members of the current House of Commons and their parties. The Conservatives are likely to gain significantly over Labour as a result of the changes, but there is much debate over the electoral data that the Commissions have to use, as laid down in the rules approved by parliament in 2011.

The Boundary Commissions for England, Northern Ireland and Wales have now published their initial recommendations for new parliamentary constituency boundaries. These are implementing the revised rules for such exercises introduced in the Parliamentary Voting System and Constituencies Act 2011. An earlier exercise deploying those rules began in 2011 but was ended prematurely by parliament in 2013. That decision delayed the procedure by five years; the Commissions now have to deliver a final set of proposals for new constituencies by October 2018, which it is anticipated parliament will approve for use at the expected next general election in 2020.

Those new rules introduced two major changes to the United Kingdom’s electoral cartography, each with a potential substantial impact on the composition of the next House of Commons. First, the number of MPs is to be reduced from 650 to 600: England will have 501 compared to its current 533; Scotland’s contingent will be reduced from 59 to 53 and Northern Ireland’s from 18 to 17; Wales will experience the greatest reduction, from 40 to 29 MPs. The second change is that with four exceptions (two for Scotland – Orkney & Shetland and the Western Isles – and two for England – for the Isle of Wight) all constituencies must have electorates deviating by no more than five percentage points from a UK average of 74,769; all must therefore have electorates between 71,031 and 78,508.

The combination of those two changes accounts for the bigger cuts in Wales than elsewhere. Currently Wales has 40 constituencies with an average electorate of 54,546, compared to an average of 70,234 for England (excluding the Isle of Wight) and 67,416 in Scotland. Only one of the current 40 Welsh constituencies has an electorate within the specified range, and so the current map has to be completely replaced.

The Scottish Boundary Commission will not announce its provisional recommendations until mid-October, at the request of the political parties there.

One issue that has caused considerable concern relates to the data on the number of registered electors that the Commissions are required to employ. They must use the electoral roll compiled in late 2015. Any individuals who successfully applied to be added to the register after December 2015 are excluded – and as many as 2 million did so in order to be able to vote in the June 2016 EU referendum. Those electors are not included in the figures. In considerable part that number of ‘missing electors’ from the Commissions’ databases is because of the government’s decision – against the Electoral Commission’s advice – to end the transition to individual electoral registration in 2015 rather than, as the legislation allowed, 2016. Many who were registered to vote in the 2015 general election – a large number of them young people – were removed from the register in December 2015 as a consequence.

That decision has had substantial consequences, because of the geography of the ‘missing millions’; they are concentrated in the large cities with substantial young and mobile populations living in rented accommodation, including university students. For example, in the unfinished exercise in 2011–­2012 the Boundary Commission for Wales’ revised recommendations after public consultation were for 30 constituencies each with an electorate between 72,810 and 80,473 (the average for the 596 constituencies then was 76,641). Of those 40, in 2016 four had an electorate below the threshold figure for the current exercise and two had electorates that were too large. The biggest changes were in Cardiff Central and Gower & Swansea West. In the Commission’s 2012 revised proposals Cardiff Central’s electorate was 72,901; by December 2015 that had fallen to only 57,730: for Gower & Swansea West, the change was from 74,515 to 65,125.

Similar reductions can be found in comparable places throughout Great Britain, with the consequence that in general urban areas will be under-represented in the 2020 House of Commons. For example, according to the data being used by the Electoral Commission, in December 2015 Cambridge Borough had 75,799 registered electors; six months later, when the EU referendum was held, it had 80,099. Oxford had an even bigger gap: 88,382 registered electors in December 2015 and 97,309 in June 2016 – but the extra 8,927 did not count in the allocation of constituencies. And across Greater London, which has the greatest concentration of students, whereas there were 5,090,487 registered electors in December 2015 there were 5,424,289 in the following June – a difference of 333,802 voters, or approximately four more seats that otherwise might go to London rather than other regions.

Reducing the number of seats and equalising their electorates means that many of the existing constituencies will be divided whereas others are unchanged. In the East Midlands, for example, the current Gainsborough constituency is unchanged and Ashfield loses just 72 voters to the proposed Broxtowe & Hucknall seat, whereas the proposed Alfreton & Clay Cross constituency draws 40 per cent of its electors from the current Amber Valley seat, 37 per cent from Bolsover and 23 per cent from what is now North East Derbyshire.

Similar variations – substantial continuity between old and new in some of the proposed constituencies but major change in others – characterise every region. In the South West, for example, in Bristol one ward is switched between two of the existing constituencies and the other two constituencies are unchanged, and in Swindon the only changes to the current two seats have been made to align their boundaries to the town’s new wards, involving just over 5,000 voters being moved. But the certain to be challenged Bideford, Bude & Launceston proposed constituency draws 43 per cent of its voters from the former North Cornwall seat and the remainder from Torridge and West Devon.

Many county boundaries, and many more borough and district boundaries, have been crossed in order to create constituencies within the narrow band of electorates required. In London, for example, of the 68 proposed constituencies less than half (31) comprise wards from just one of the 32 boroughs and six of the boroughs have no constituency that contains only wards from that borough alone. One of those six is Southwark; three constituencies combine some of its wards with others from Lambeth and another combines Southwark and Lewisham wards. Currently Birmingham has ten constituencies, each comprised of four of the city’s 40 wards. Under the proposals there will be only five constituencies made up of four Birmingham wards each; there will be three in which each combines four Birmingham wards with one from Sandwell; one combining four Birmingham wards with one from Walsall; one with three Dudley and three Birmingham wards; and one with six from Solihull and a single Birmingham ward. In most of Britain’s large cities the maps of local government and parliamentary representation will no longer largely coincide.

In the discussions during the 2011–2012 aborted exercise first implementing these rules there was much debate regarding the desirability of splitting wards to reduce the overall disruption to existing seats and communities. Many authorities, and also many MPs, have since thought this desirable and the Commissions indicated that they would certainly consider it this time. They have: the Welsh Commission has proposed to split just one ward, and the English Commission none, although in the latter case it has published comprehensive data for the country’s polling districts should those making representations want to split wards in their alternative schema. Some believe a small amount of ward splitting could significantly reduce the disruption and borough boundary crossing in the largest cities (Birmingham, Sheffield and Leeds, for example) but now we have to wait to see whether any such schemes form part of the alternatives suggested during the public consultation, and how the Commissions will respond. (The Scottish Commission is already committed to such splitting, and has done so previously without generating much opposition.)

Not surprisingly, much attention is focused on the likely impact on high profile MPs, some of whom are more kindly treated by the proposals than others. Theresa May’s Maidenhead constituency is entirely unchanged, for example, whereas Nick Clegg’s in Sheffield Hallam is split virtually in half. Jeremy Corbyn’s Islington North is split three ways; the largest part goes to a new Finsbury Park & Stoke Newington seat, which also includes five wards from the constituency now represented by his ally Diane Abbott; two of his current wards are joined with eight from the present Holborn & St Pancras seat, held by Keir Starmer; and the final ward joins with eight from the current Islington South & Finsbury (held by another Corbyn ally, Emily Thornberry) in a new Islington constituency. And George Osborne’s current Tatton constituency is split five ways: five wards go to the proposed Macclesfield seat, three to Altrincham & Tatton Park, two each to Bramhall & Poynton and Eddisbury & Northwich, and one to Weaver Vale.

What will be the electoral consequences of all this? The proposals published in September–October are only provisional and many are likely to be changed after the twelve-week statutory period of consultation that began on the day the proposals were published. But the general outline is clear:

  • In England and Wales, when the Commissions made their initial proposals for 600 new seats in 2011 the likely result would have been that a Conservative lead over Labour of 48 seats at the 2010 general election was extended to 68 – a net gain to the Conservatives of 20 seats as a result of the equalisation, despite the reduction in the overall number of MPs – and all the signs are of a similar outcome in this set of proposals: indeed the pollster Anthony Wells has already done the massive number-crunching and suggested that the Conservatives will lose 10 seats, Labour 28 and the Liberal Democrats four of their current seven, increasing the Conservatives’ lead over Labour from 88 seats to 116;
  • In Northern Ireland, where the main element of the change is a reduction in the number of Belfast seats from four to three, it is likely that one of those three will be more marginal between the province’s two main political communities (unionist and nationalist) than is the case in the current four; and
  • In Scotland, although we do not yet have its Commission’s proposals the electoral situation there means that of the six seats that are to be lost at least five and perhaps all six will be among those currently held by the SNP;

Of course, this is only the start of the exercise. There are now twelve weeks of public consultation, inviting both written and oral representations challenging some (almost certainly most) of the recommendations and suggesting alternative configurations to replace them. Each of the political parties, their MPs and local supporters will be pressing their own causes, using the other criteria that the Commissions can deploy – special geographical considerations; local government boundaries; the boundaries of existing constituencies; and the breaking of local ties – to advance their arguments for alternative sets of constituencies that meet the size rules but better reflect their electoral chances (without, of course, ever mentioning that is the main reason for any changes they want made!). The Commissions will consider all that evidence, and responses to it, and bring forward revised proposals in late 2017; these will again be subject to consultation (written but not oral) which the Commissions will then take into account in finalising the recommendations that they must send to parliament by October 2018. A majority of their initial proposals were changed, to a greater or lesser extent, during the previous exercise in 2011–2012 and that is very likely to be the case again.

The changes being implemented in this review were introduced for two reasons. The reduction in the number of MPs was to save money (£12million p.a. was suggested) and thereby improve trust in politics/politicians – and in the House of Lords a government spokesperson defended the choice of 600 as being a good round number! The equalisation of electorates – both between and within countries – was defended because it would remove a widely-recognised Conservative disadvantage because Labour seats have tended for decades to have fewer electors than the Tory-held. That disadvantage has been removed – but there is no gerrymandering, whatever some critics may claim. The changes do not create a new Conservative advantage, they simply remove a long-standing Labour one and equalise the parties on one of the factors that produces disproportional election outcomes in the UK’s first-past-the-post electoral system – but the electoral register that the Commissions have to use has introduced a new disadvantage for Labour.

We will get a final map for the reduced number of constituencies in 2018 and in some parts of the country at least it may be somewhat, if not substantially, different from what is currently before us. But almost certainly, whatever the final cartographic outcome the Conservatives will be the main beneficiaries by up to 30 seats net from a new numbers games, in which politicians have set the rules but their implementation is undertaken by strictly independent, politically neutral Boundary Commissions, their Assistant Commissioners and staff.

About the author

Ron Johnston is a professor in the School of Geographical Sciences at the University of Bristol and co-author, with David Rossiter and Charles Pattie, of The Boundary Commissions (Manchester University Press, 1999) and many subsequent analyses of parliamentary boundary redistributions.

4 thoughts on “Plus ça change – or déjà vu all over again: the proposals for new, and fewer, parliamentary constituencies

  1. Pingback: Reducing the size of the House of Lords: here’s how to do it | The Constitution Unit Blog

  2. Pingback: Should We All Be Worried About The Labour Party? – Hub Politic

  3. Do you really buy the first reason? “The reduction in the number of MPs was to save money (£12million p.a. was suggested) and thereby improve trust in politics/politicians.” The Tories’ plan to arbitrarily reduce the number of elected Members of Parliament by 50 had always been under the guise of reducing the cost of politics, even whilst they continued to cram the Lords at taxpayers’ expense. The first reason therefore does not withstand logical scrutiny.

    The “in-built advantage to Labour” that you talk of no longer exists! As Jones comments, “Labour’s in-built advantage evaporated at the last election along with 40 of the party’s seats in Scotland. While the Tories won a seat for every 34,244 voters who opted for the Conservatives, it took an average of 40,290 to win a Labour seat last year” (see further here:

  4. Your posting ignores one factor: while it is illegal to vote more than once in one’s own name at an election, it is not illegal to be registered to vote in more than one constituency. Typically, before individual voter registration the colleges organised the registration of students in their Halls of Residence, while these same students were probably still registered at their parents’ homes. This meant that the students could vote wherever was most convenient for them: at their university in term-time and at home out of term-time. However, it also meant that each student was double-counted in calculating the total number of the electorate in each constituency. What we do not know is how many of the additional electors for the Referendum were double-counted students, but since the principal examples given were for Oxford, Cambridge and London, it is likely that most of them were.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s