Boundary changes: How the electoral impact of the new political battlegrounds are calculated
The electoral geography of the UK is changing.
Following the recommendations of independent Boundary Commissions for England, Scotland, Wales and Northern Ireland, the next UK general election will be fought on new constituency boundaries, replacing those in operation since 2010.
This is the sixth periodic review to be implemented since the war. The next review is not scheduled until October 2031.
Exploring how this movement of voters affects the political makeup of the House of Commons is a task that Colin Rallings and myself have been doing over the past thirty years following previous boundary adjustments.
Winners and losers
There are winners and losers in different parts of the UK. With the total size of the House of Commons fixed at 650 seats, the 2023 review used a national ‘electoral quota’ to determine the number of constituencies available to the different countries.
The quota, 73,393 electors, was based on the 47.1 million registered electorate on March 2, 2020. With few exceptions, particularly affecting island seats, the Commissions had to work within plus or minus 5% of the quota. The electorates for new seats could be no smaller than 69,724 and no larger than 77,062 electors.
England’s tally of seats rises from 533 to 543, a net gain of 10 seats. Scotland loses two seats, from 59 to 57. Wales is the biggest loser. It had 40 constituencies before but now has only 32, a reduction of a fifth of its seats.
The situation in Northern Ireland is unchanged on 18 seats.
Despite the overall increase in English seats, there are some regions that benefit, others lose.
The biggest increase is across the South East where 84 constituencies will be replaced with 91 when the next election is fought. Both the East of England and the South West show a net gain of three seats.
The North East, North West and West Midlands each has a net reduction of two constituencies.
What does it mean for a general election?
Beginning with the changes introduced prior to the 1997 general election, we have constructed vote estimates for the previous election assuming the new constituencies had then been in place.
These figures are used by the principal broadcasters, Sky News, BBC and ITV, together with the Press Association, all of whom commission the work, to set the baseline numbers as we go into the next election.
How much each party’s vote has changed within a constituency, whether a seat has changed hands or been held by the incumbent, are measured against our estimated results.
See your your constituency has been affected by entering your postcode in our lookup.
Why and how do we do it?
The ‘why’ is because general election votes are only available for whole constituencies.
Boundary commissions use local electoral wards as the building blocks for constructing new constituencies but we don’t know how general election votes are cast at that level. Hence, the need for estimates.
Local election results are our means for constructing these. We begin by building a set of local results in the wards that form the existing constituency. Since the patterns of party competition at parliamentary and council elections can vary, some adjustments are necessary.
This might mean taking an average of local votes when multiple councillors are being elected. Sometimes, historic ward results are used to fill in any missing gaps when one of the major parties has either been unopposed or failed to field a local election candidate.
Many parts of the country see local voters support non-party or independent candidates. Where such candidates seem to have been given a free run by one of the major parties, their vote is given to that party as if they were its surrogate.
Where these candidates stand, in addition to those from the major parties, their vote is ignored for these purposes.
Making these adjustments eventually gives a local elections total for each constituency.
We then make the assumption that the proportion of a party’s total local vote received in any given set of wards is equal to the proportion of its total general election vote in the same wards.
Each party’s strongest and weakest wards will remain the same regardless of any gross differences in performance between the general and local elections. This enables us to translate local votes into notional general ones through a formula which takes account of the relationship between the two.
In most cases, adjustments to parliamentary boundaries involve the transfers of small numbers of wards into or out of neighbouring constituencies. Leaving the core of each current constituency intact and only moving relatively small numbers of ward-level votes around means that we reduce the risk of arriving at inaccurate estimates.
Accuracy can be gauged from previous general elections where the baseline figures have been estimated. Changes in constituency vote shares are indistinguishable between seats with boundary changes, and therefore estimated votes, and those using the actual results from the previous election because there have been no boundary adjustments.
The extent to which a constituency has been transformed is captured by what we call the Index of Change. This is zero when no or very few electors have been moved. The index climbs when an existing constituency loses electors because it exceeds the quota. It climbs further still when electors from a neighbouring constituency are transferred in.
Making a similar number of gains at the next general election would give Labour a reasonable-sized majority. But to do that, Starmer needs to match Blair’s performance and hope that Sunak’s Conservatives fare even worse than their predecessors did under John Major.
Click here to open a document written by Professors Colin Rallings and Michael Thrasher, which explains the methodology in even more detail.
And click here to DOWNLOAD a spreadsheet of all of the new vote estimates for each constituency.