PROJECTING THE IMPACT OF ‘REDUCE AND EQUALISE’

Stuart Wilks-Heeg and Stephen Crone, Democratic Audit.

Background

In August 2010, Democratic Audit was asked by BBC Newsnight to produce estimates of the
likely impact of the coalition’s ‘reduce and equalise’ proposals for Westminster constituencies.
These proposals involve reducing the number of MPs from 650 to 600 and seeking to equalise
the number of electors in each constituency. With the exception of 2, or possibly 3, remote
Scottish constituencies, this will mean that all constituencies should fall within +/- 5 per cent of
the average – approximately 76,000 electors per seat. The process of drawing up the new
constituencies will be led by the fully independent Boundary Commissions for England,
Scotland, Wales and Northern Ireland.

Our estimates are principally focussed on the likely geographical impact of the reforms, i.e. the
areas of the country which are most likely to lose seats. We do not yet have the electorate
figures which the Boundary Commissions will use to inform their recommendations (these will
only become available in December 2010). Neither do we have full information about how the
Boundary Commissions will approach the task. For instance, it is not yet clear whether
constituencies will be permitted to cross regional boundaries. Nonetheless, it is possible to
predict with a high degree of accuracy the changes in the number of seats at a regional level.

Newsnight also asked us to consider the possible political impact of the changes. While it is
possible to draw conclusions about how the proposals could impact on party representation,
these findings must be regarded as purely indicative. The effect of boundary changes can only
be assessed once the geography of the new boundaries is actually known. However, based on
current evidence, we can make some reasonable assumptions about how the regional changes
would translate in terms of political representation. Stark regional contrasts in political
representation, together with the geographical clusters of constituencies with smaller
electorates in metropolitan areas outside of London, provide some obvious pointers.

Which parts of the country will lose seats?

The current distribution of seats by country/region is shown in table 1. There will be a separate
boundary review process for each constituent country of the UK (England, Scotland, Wales and
Northern Ireland), each of which will have a specific seat quota. Using the 2009 Electoral
Statistics, we first calculated how many seats each country would be entitled to in a 600 seats
House of Commons, assuming an average of 76,000 electors per seat. It is then possible to
calculate how many seats each country is likely to lose under the proposals.

We assume that the reforms mean there will be 503 seats in England (-30), 30 in Wales (-10), 52
in Scotland (-7) and 15 in Northern Ireland (-3).

We then calculated, on the same basis, how many of the 503 seats in England each of the English
regions would be entitled to. The results are shown in table 1. The North West is predicted to
lose the most seats (-7), followed by the West Midlands (-5) and Yorkshire and the Humber (-4).
The fewest seat reductions are likely in the East of England (-1) the East Midlands (-2) and the
South West (-2).

1
Table 1: Number of current Parliamentary constituencies by region/country, and
projected distribution of constituencies in 2015

Before After Change

Greater London 73 70 -3
South West 55 53 -2
West Midlands 59 54 -5
North West 75 68 -7
North East 29 26 -3
Yorkshire and the Humber 54 50 -4
East Midlands 46 44 -2
East of England 58 57 -1
South East 84 81 -3
ENGLAND TOTAL 533 503 -30

Scotland 59 52 -7
Wales 40 30 -10
GB TOTAL 632 585 -47

Northern Ireland 18 15 -3

UK TOTAL 650 600 -50

It is important to note a number of caveats about these estimates:

1. The figure produced for the ‘quota’ for each region or country is rarely an integer,
meaning that numbers need to be rounded up or down. Within England, we have made
some attempt to take into account sub-regional variations in the size of constituencies,
as well as clusters of small constituencies, but a more detailed analysis would be needed
to confirm the estimates for the English regions.
2. The estimates are based on the December 2009 electoral registers – whereas the
boundary review will be based on the December 2010 registers. Assuming late voter
registrations added in the run-up to the 2010 General Election are mostly retained
(‘carried forward’) on the new registers, the average number of electors per
constituency will be moderately higher (although this is unlikely to make much
difference to estimates at a regional scale).
3. While constituencies will not cross country borders, some of the new constituencies
within England may need to cross regional boundaries.

Despite these caveats, we would expect these figures to be accurate to the degree of +/- 1 seat
for Scotland, Wales, Northern Ireland and each of the English regions.

What would the impact be on party representation at Westminster?

It is very difficult to produce precise estimates of the likely partisan impact of these changes
without knowing precisely how the new constituency boundaries would be drawn.

However, we can make a number of assumptions to produce a rough estimate of the likely
impact on representation in the House of Commons.

2
 First, there are clear regional variations in party representation, which give us some
idea of how the parties are likely to be affected. For instance, since almost two-thirds of
North West MPs are Labour, the seat reductions in the North West will hit Labour
harder than anyone else. Conversely, as there are 0 Liberal Democrat MPs in the East
Midlands and only 2 Labour MPs in the East of England, we know that the changes in
these regions will have little or no impact on these parties’ representation.
Second, current variations in constituency sizes highlight that those with smaller
electorates are disproportionately located in metropolitan areas. For instance, it is clear
that all of the former Metropolitan County Council areas of England (Greater
Manchester, Merseyside, South Yorkshire, Tyne and Wear, West Midlands, West
Yorkshire) will lose at least one seat. Merseyside is likely to lose two seats and ‘Greater
Birmingham’ may lose as many as three. As most of these metropolitan seats have
Labour MPs, we have assumed that seat losses in regions such as Yorkshire and the
Humber will disproportionately impact on Labour.1 Likewise, seat reductions in
Scotland and Wales are most likely in more densely-populated, Labour-voting areas.
Third, it is possible to identify specific constituencies within regions which are
effectively ‘islands’ of representation for a particular party and have very small
electorates – thus making them highly vulnerable as a result of boundary changes. For
example, Berwick-upon-Tweed in the North East (one of only 2 Liberal Democrat seats
in the region) has only 57,000 electors. On balance, we have assumed that boundary
changes will cause the Liberal Democrats to lose this seat. Likewise, the neighbouring
Plaid Cymru seats of Dwyfor Meirionnydd and Arfon in North Wales have 46,000 and
41,000 seats respectively – it is difficult to imagine how the party could retain two seats
in this part of Wales given the likelihood of the country as a whole losing 10 seats.

Using these assumptions, we produced the regional estimates shown in table 2. The aggregate
estimates for the overall impact on party representation are shown in table 3.

Table 2: Possible regional outcome of 2010 General Election in Great Britain if it had
been fought in 600 (rather than 650) constituencies

Change Labour Lib Dem Cons SNP PC

Greater London -3 -2 0 -1 --- ---
South West -2 -0 -1 -1 --- ---
West Midlands -5 -3 0 -2 --- ---
North West -7 -4 -1 -2 --- ---
North East -3 -2 -1 0 --- ---
Yorkshire and the Humber -4 -3 -1 0 --- ---
East Midlands -2 0 0 -2 --- ---
East of England -1 0 0 -1 --- ---
South East -3 0 -1 -2 --- ---
ENGLAND TOTAL -30 -14 -5 -11 --- ---

Scotland -7 -5 -1 0 -1 0
Wales -10 -6 -1 -2 0
GB TOTAL -47 -25 -7 -13 -1

1
It should be noted that several metropolitan areas outside of London have 1 or 2 Liberal Democrat MPs – often
representing constituencies surrounded by safe Labour seats. Depending on how boundaries are re-drawn in
these areas, it is therefore possible that additional Liberal Democrat seats could be vulnerable.

3
Table 3: Possible outcome of 2010 General Election if it had been fought in 600 (rather
than 650) constituencies

Seats won at Impact of Est. MPs in

2010 General seat 600 seat
Election reductions Commons
Conservatives 307 -13 294
Labour 258 -25 233
Lib Dem 57 -7 50
Others 28 -5 23
TOTAL 650 -50 600

Note: these estimates are based on specific assumptions and should be interpreted principally as a guide
to the likely impact. The number of seats each party would actually lose is probably in the range of +/- 4.
Thus, Conservative losses could range from 9-17, Labour from 21-29 and the Liberal Democrats from 3-
11.

These estimates suggest that, if the 2010 election had been fought on the basis of 600 seats with
equalised electorates, Labour would have lost ground relative to the two other main parties. As
many as half of the 50 seats due to disappear from the Commons could be notionally Labour
seats. The boundary changes and seat reductions are therefore likely to provide a slight benefit
to the Conservatives. However, it is not enough to give the Tories a majority – they still fall 7
short under these projections. This outcome might have been just enough for the Conservatives
to govern with the DUP, who would probably have been reduced from 8 to 7 MPs as a result of
the changes – although such an arrangement would almost certainly have been difficult to
broker. Given the political circumstances, hindsight suggests that a Conservative-Liberal
Democrat coalition would still have been the most likely outcome.

About Democratic Audit

Democratic Audit is an independent research organisation, which researches issues concerning

democracy and human rights in the United Kingdom. We are grant funded by the Joseph
Rowntree Charitable Trust, governed by an independent Board of Trustees, and based at the
University of Liverpool. The Audit’s Director is Dr. Stuart Wilks-Heeg, who is also Senior
Lecturer in Social Policy at the University of Liverpool.

About the authors

Stuart Wilks-Heeg is Executive Director of Democratic Audit and Senior Lecturer in Social Policy
at the University of Liverpool. He is the author of Purity of Elections in the UK: Causes for
Concern (Joseph Rowntree Reform Trust, 2008) and was the lead author of the Electoral
Commission’s (2010) report on electoral registration.

Stephen Crone is undertaking a graduate research placement with Democratic Audit. He

graduated from the University of Liverpool with a BA in Politics (First class honours) in 2009.

13 August 2010