A new bill currently before parliament alters the rules governing the periodic redrawing of the UK’s parliamentary constituencies, most notably by replacing a requirement to limit the House of Commons to 600 MPs with a new fixed size, set at the current 650. But, as Ron Johnston, David Rossiter and Charles Pattie show, the new rules are just as likely as those they replace to result in major disruption to the constituency map at all future reviews. 

In 2011, the coalition government passed the Parliamentary Voting System and Constituencies Act, which changed the rules guiding how the UK’s parliamentary constituencies are drawn up. Boundary reviews were to take place every five years (more frequently than before). Almost all new seats (with four exceptions) were to have electorates within +/-5% of the national quota (the average electorate). And the House of Commons was to be reduced in size from 650 to 600 MPs. To date, the Boundary Commissions have conducted two redistricting exercises under the 2011 Act. Neither review has been implemented: the first was lost to infighting in the coalition, and the second was tabled in September 2018 but has not yet been approved by parliament. The proposed changes they contained would have produced the largest shake-up in Britain’s constituency map in modern times.

Now the redistricting rules look set to change again. The Parliamentary Constituencies Bill 2019-21, published on 20 May, is now moving through its Committee Stage in parliament. It retains the requirements that all constituencies (with four exceptions) should have an electorate within +/-5% of the national average, but changes the number of constituencies to 650 – the argument being that with Brexit there will be more work for MPs, and thus a need for more of them, than if we had remained a member of the EU. If the Bill is passed, the Boundary Commissions will be required to recommend a new set of 650 constituencies by 1 July 2023 – in time for the next general election, due in May 2024. Subsequent reviews will then take place on a slightly longer timetable than under the 2011 Act – every eight years.

So how disruptive would a constituency review be under the new Bill? It might be thought that the major disruptions contained in the two abandoned reviews under the 2011 Act were one-off consequences of the adoption of new rules, and that subsequent redistricting reviews under those rules would be less disruptive than in the first iteration. One might further think that the current Bill, by removing the need to cut 50 seats from the Commons would also reduce substantially the disruption to the constituency map. However, both assumptions are wrong. Major disruption of the constituency map is still very likely in the next and all subsequent reviews. 

When we gave evidence to the Political and Constitutional Reform Committee (and as we said in a two-part post on this blog) in 2014 we pointed out that the combination of a fixed parliament size and a restricted tolerance (5%) would mean that each subsequent review would witness significant change to boundaries. At that time, we estimated major change would occur in around one-third of seats and minor change in another third. And we showed that (at least while most Boundary Commissions continued to use local government wards and similarly-sized units as their main building blocks and continued to resist substantial ward-splitting) changing the size of the Commons would not substantially affect the degree of disruption. Whether parliament’s size was fixed at 600 or 650 MPs made little difference.

Six years on from that Committee’s report (which recommended a loosening of the tolerance to 10%) we can actually test those projections with real data. Although the first review under the 2011 legislation was aborted in 2013, each commission had undertaken public consultation and published revised recommendations based on 2010 electoral data (the ‘enumeration date’), so we know how disruptive that review would have been had it been implemented. When ONS published its annual volume Electoral Statistics in 2015, we already knew that changes would have been required to both national and regional entitlements – indeed that very point was made by the commissions when they commenced their still-unimplemented 2018 Review. Roll forward three more years to 2018 and we can gauge the level of change that would be required had the 600-seat constituency map proposed in that first Review under the 2011 Act been revised on an eight-yearly cycle as proposed in the current Bill.

The results are striking. At the national scale, in the first revision of the 600 constituencies proposed by the commissions in 2013 Scotland would gain a seat and Wales lose one, leaving the allocation to England and Northern Ireland unchanged. But within England (where the Boundary Commission now produces separate plans for each government region), six of the nine regions have a changed entitlement. And this is the nub of the problem. For when an entitlement changes and tolerance is small – bearing in mind that 5% is below the average English ward size of almost 6,000 – major disruption necessarily ensues. North East England, for example, would lose one of its 26 seats. The Commission would choose to either entirely redraw the map (unlikely) or select a seat to ‘abolish’. In redistributing the wards from the abolished seat it would necessarily proceed in ‘pass the parcel’ fashion, transferring wards to most or all of its former neighbours. Some of these neighbours would be below quota and hence able to accept one (but only one) new ward, but others would have a ‘legal’ electorate and simply act as transit stations, importing some wards and exporting the same number. The process would continue outwards until all seats again fell within quota. By the end of the process just a handful of seats (we estimate six or fewer in the North East example) would remain unchanged.

How many seats are significantly impacted by this process varies from region to region. The smaller the ward size, the lower the degree of disruption, but the geography of change – the necessarily unpredictable pattern of growth and decline in population, let alone electorates – also has a crucial part to play. Replaying the commissions’ initial exercise eight years on, our estimates of disruption have to be increased. We now estimate that around half of all seats would experience major change at each subsequent review, with just one in five escaping change of any sort. In part this reflects the longer time interval proposed (up to eight from five years) but it also reflects two other factors. First, the pattern of national and regional allocation has been more fluid than we projected. Second, the ongoing changes to the map of local government in England (mergers of authorities and reduced council sizes) has led to an increase in average ward size, removing a further degree of freedom from the English Commission.

What is more, (as we also showed in 2014) moving from a fixed size of 600 MPs to 650 will not prevent major (and repeated) disruption to the constituency map. The reason for this can be seen by looking at the current situation. The Office for National Statistics recently published the number of registered electors in each constituency in December 2019. Excluding the four constituencies not constrained by the +/-5% tolerance (the Isle of Wight – which gets two seats; Orkney and Shetland; and Na h-Eileanan an Iar), the UK had a registered electorate of 46,907,785, which gives a national quota for 646 seats of 72,612 with all having to have an electorate between 68,981 and 76,243. 

The Welsh electorate in December 2019 was 2,313,851 – an average of 57,846 across its 40 constituencies. Using a national quota of 72,612 it would be entitled to only 32 seats, a reduction of eight. That would require the entire map to be redrawn. Currently only four seats (three in Cardiff plus the Vale of Glamorgan – which with 76,998 is too big!) have electorates greater than 68,981; many have less than 60,000 – the smallest has just 42,664.

Excluding its two ‘protected constituencies’ Scotland currently returns 57 MPs: its 2019 electorate would entitle it to 54. But 42 of its current constituencies have electorates outside the quota – again extensive boundary changes will be needed. Northern Ireland perhaps presents fewer problems. It currently has 18 constituencies and would retain that number with the new quota: but only eight of the current 18 have electorates within the +/-5% tolerance.

Finally England would be the main gainer: it currently has 532 constituencies, excluding the Isle of Wight; with a quota of 72,612 it would be entitled to 542. But 126 have electorates less than 68,981 (twelve have less than 60,000) and 200 have more than 76,243 – 12 have more than 90,000, including both Milton Keynes seats. Again, extensive cartographic redrawing will be necessary (and, as with the commissions’ first two attempts to produce new constituencies under the 2011 legislation, changes will also be required to the boundaries of many seats with electorates currently within the quota to accommodate changes to those seats with electorates outside the quota).

So as MPs begin to debate the new Parliamentary Constituencies Bill, they should not assume that fixing the size of the Commons at the current level of 650 will reduce the degree of disruption to the constituency map in the next boundary review. Nor should they assume that major disruption will be a feature only of the first review under the revised rules. In fact, as analyses of the incomplete exercises over the last decade have shown, that disruption will be extensive – not only at the first redrawing of the map but every subsequent eight years according to the new timetable of reviews. Playing with numbers can produce substantial consequences – whose nature is known but, apparently, ignored.

To mark the 25th anniversary of the Constitution Unit we have launched a special fundraising initiative, encouraging supporters to make donations incorporating the figures 2 and 5. If you value our work, and would like to contribute to its future continuation, please consider making a one-off or regular donation. Contributions are essential to supporting our public-facing work. Find out more on our donations page. 

Sign up to our mailing lists, for news of events and publications, or to our blog, here. 

About the authors

Ron Johnston was Professor of Geography at the University of Bristol. He was working on a draft of this post in May when he sadly died, aged 79. He was a scholar of immense distinction who had contributed immeasurably to the study of political geography over many years. All of us at the Constitution Unit send our deepest condolences to his family, colleagues, and friends.

David Rossiter is a former Research Fellow who worked at the Universities of Bristol, Leeds, Oxford and Sheffield.

Charles Pattie is Professor of Politics at the University of Sheffield.

The authors have written together on the Boundary Commissions and constituency definition in the UK since the early 1980s and co-authored The Boundary Commissions: Redrawing the UK’s Map of Parliamentary Constituencies (Manchester University Press, 1999).