Boundary changes

Posted by on June 15, 2015 in Articles, Electoral Boundaries, Front Page | 0 comments

Boundary changes

A few weeks ago the Times – in the person of its excellent political correspondent Laura Pitel – asked if I could have a look at the implications of the forthcoming changes to constituency boundaries which will be argued over between 2016 and 2018 and – barring eventualities – implemented in 2020 for the new election. The political map of Britain changed abruptly in 2015, thanks to the SNP landslide in Scotland, the Liberal Democrats’ disaster and the pattern of Conservative/ Labour swing that ended up making the gradient of the mountain Labour must climb much steeper.

I produced a fairly ‘quick and dirty’ version of what the new boundaries might look like, without going into too much detail because the electorate numbers on which the allocation of seats will be based are going to be different in December when the definitive numbers are taken. One cannot simply take the results of the previous aborted review in 2011-13 forward either, for the same reason, although the detail of that review and the arguments that found favour in the revised proposals are of interest in seeing how the Boundary Commissions might approach their new task. By all accounts, the Boundary Commission for England (BCE) will be rather more willing to split wards this time around, which might well reduce the number of cross-border and bizarre seats, but creates problems of its own.

Laura Pitel’s article is here (£) and if I do say so myself is well worth reading. The underlying calculations are published here. The headline finding is that the Conservatives benefit slightly from the changes, losing 19 seats (compared to 20 for Labour and 11 for the other parties), which in a smaller 600-seat parliament would increase their majority by 12 seats.

Current number of MPs Number of MPs after boundary changes Change in number of MPs Percentage change
UK 650 600 -50 -7.7%
England (including Isle of Wight) 533 501 -32 -6.0%
Wales 40 29 -11 -27.5%
Scotland (including Islands) 59 54 -5 -8.5%
Northern Ireland 18 16 -2 -11.1%
English regions        
East of England 58 56 -2 -3.4%
Isle of Wight 1 2 +1 +100%
South East (not IoW) 83 81 -2 -2.4%
South West 55 53 -2 -3.6%
SOUTH 197 192 -5 -2.5%
LONDON 73 70 -3 -4.1%
East Midlands 46 43 -3 -6.5%
West Midlands 59 53 -6 -10.2%
MIDLANDS 105 96 -9 -8.6%
North East 29 25 -4 -13.8%
North West 75 68 -7 -9.3%
Yorkshire & The Humber 54 50 -4 -7.4%
NORTH 158 143 -15 -9.5%


The new rules (legislated in 2011 but postponed until this parliament):

  • The number of seats will be 600.
  • Four seats have special status – 2 on the Isle of Wight and 2 in the Scottish Islands (na h-Eileanan an Iar and Orkney & Shetland) and are excluded from requirements on the size of constituencies.
  • The other 596 seats are allocated to the rest of the country on the basis of the numbers on the electoral register in December 2015. The law specifies the formula to be used to distribute seats between the four nations, and the Boundary Commission for England (BCE) uses the same method to distribute constituencies between the nine English regions.
  • Constituencies shall be within 5 per cent either side of the UK electoral quota, which is established by dividing the total electorate (other than in the island exceptions) by 596. The four special cases are exempt, and there is some small print about creating constituencies with very large land area and in the case where the Northern Ireland average size is significantly different from the UK quota.


The registered electorate

For the purposes of this analysis, I have used the electorate figures as at the General Election of May 2015 as supplied to the Electoral Commission. The actual numbers in December 2015 will be different, which would affect the numbers of constituencies going to each area at the margins and perhaps have large effects on some of the local detail.


Principal effects of the changes

  1. Wales will be most badly affected because its constituencies are currently significantly smaller than the UK average. It will lose over a quarter of its representation in the House of Commons.
  2. Scotland, Northern Ireland, the North and the Midlands will lose around 10 per cent, more or less, of their MPs.
  3. The South of England and London will lose smaller proportions of their MPs.

In general, this shift in the balance of representation will benefit the Conservatives – the most Conservative region (South East) loses least. However, they may also find that their toeholds in historically weaker regions (Wales and the urban North) are ‘swamped’ by adding Labour-voting areas to existing Tory seats.


How the rules change the way constituencies relate to counties

The old system of drawing constituencies placed importance on county boundaries in England (and to some extent on London borough boundaries). For these purposes unitary authorities are often counted with their ‘parent’ county, i.e. Medway is included with Kent. The Boundary Commission for England (BCE) would first examine how many MPs a county should have, and then draw boundaries within that county to equalise the electorate to a reasonable degree in each one, with an eye to other considerations like keeping communities together and minimising disruption.

The new system is different, in that it will require constituencies that cross county boundaries – something that has been avoided where possible for centuries of parliamentary history. This is caused by the strict 5 per cent threshold of variation around the UK electoral quota (which using the May 2015 numbers would be 77,618 electors). Creating constituencies with fewer than 0.95 quotas of electors (73,738 electors) or more than 1.05 quotas of electors (81,498 electors) is not permitted (except in the special cases written into the legislation).

So a county such as Cambridgeshire, with 7.50 quotas, cannot have 7 seats because the average seat would be 1.07 quotas, i.e. 7 per cent too big, which is not allowed. Nor can it have 8 seats because the average seat would be 0.94 quotas, which is also not allowed. There must therefore be at least one cross-border constituency between Cambridgeshire and a neighbouring county.

Even if a county can technically be divided into a whole number of constituencies of allowable size, it might not be possible in practice. For instance, Bedfordshire could just about get six seats of its own (average size 0.96 of a quota, 4 per cent undersized). But in practice it is not likely to be possible to chop it up so finely and evenly as to produce six allowable seats. Counties may also be fine on their own, but disrupted by neighbouring areas being too far from a whole number of seats (e.g. Berkshire may be disrupted because Oxfordshire is big).

The Boundary Commissions work by component nation of the UK and the English Commission is guided by the legislation to work within the regions that are also used as European Parliament electoral areas. The detailed findings of my work on the boundary changes are therefore also divided up by nation and region, with a page each for Eastern England, East Midlands, London, North East, North West, South East, South West, West Midlands, Yorkshire & The Humber, Wales, Scotland and Northern Ireland. (The map is, though, of the proposed constituencies in my STV model election).STV constituencies.


Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses cookies. Find out more about this site’s cookies.