Written evidence submitted by Lewis Baston,
Democratic Audit (PVSCB 01)
I am currently senior research fellow with Democratic
Audit and it is under the auspices of Democratic Audit that I
offer these observations on the Parliamentary Voting Systems and
Constituencies Bill. Previously I was Director of Research at
the Electoral Reform Society (2003-2010) and I have been author
and co-author of several books on political geography, most notably
The Political Map of Britain. I am grateful to the Committee
for inviting me to submit evidence.
SUMMARY
— There are no serious problems with the
provisions on a referendum.
— The timetable for the Bill itself, and
the proposed boundary review, are both too rapid and prevent consideration
of workable alternatives.
— The purported "problem" addressed
by the Bill is not a serious one.
— The electoral register is too incomplete
and the totals too volatile to serve as a fair basis for the allocation
of parliamentary constituencies, and these problems are likely
to worsen over the next few years.
— An exception has been made for some islands
and constituencies with large land areas, but there is no acknowledgement
of other factors that impinge on the practicality of constituency
representation (population, local identities, administrative complexity).
— The banning of public inquiries is a
severe and deplorable downgrading of public participation and
transparency in the boundary process.
— A Commons size of 600 is arbitrary and
seems not to reflect any analysis of the capacity and functions
of MPs and the House in general.
1. Introductory remarks
1.1 I shall concentrate my evidence on the
provisions of this Bill relating to constituency boundaries. The
burden of my evidence is that this is a severely flawed proposal,
both technically and in terms of democratic representation.
2. The Alternative Vote referendum
2.1 The Bill yokes together two constitutionally
separate sets of provisions—it would be entirely possible
to have the referendum without the boundary changes, or vice
versa. The arguments relating to each proposal are entirely
different.
2.2 The parts establishing a referendum
on a change to the electoral system are relatively uncomplicated.
Referendums take place according to the procedure established
under PPERA (Political Parties, Elections and Referendums Act)
2000, and holding one is—although a new administrative exercise—not
a legal or constitutional novelty. Provided the administration
and the question wording are fair, the main burden of scrutiny
will be undertaken by the public and the media during a referendum
campaign, for which public funds are provided to enable the Yes
and No campaigns to put their respective cases.
2.3 The interests of maximising participation
in the referendum are best served by a combined election day.
This will also reduce administrative costs. While electing two
legislative bodies on the same day (as may happen in 2015) is
complicated, a referendum question is capable of being discussed
separately and this is indeed the usual procedure in many countries
which make extensive use of the referendum, such as the United
States.
2.4 While a more democratic process of deciding
the question for the referendum, such as the Citizens' Assembly
system developed in Canada, would be preferable to the pre-selection
of two options (particularly two options such as AV (Alternative
Vote) and FPTP (first past the post) whose basic properties are
similar), one must recognise this as a step forward.
3. The undue haste of this Bill
3.1 The provisions about the boundary review
are different in kind to those establishing a referendum. Boundary-drawing
is a complex process that involves many considerations. The interaction
of any particular set of rules with the physical and administrative
geography of the UK may produce surprising results.
3.2 The essentials of the current system
were established as a result of the Speaker's Conference of 1944
which was able to discuss issues of detail in a considered way
and reach a consensus about an acceptable way forward. The rules
are being replaced without any attempt to form a consensus, and
without adequate parliamentary scrutiny of the Bill and its underlying
assumptions about representation.
3.3 The undue haste with which this proposal
is being legislated has closed off the possibilities that a more
considered process could have entertained. Among these are:
— Public attitudes towards the general
principles involved in boundary determination. How much importance
do the public really attach to the government's definition of
equality of size? Would people, in their own constituency, prefer
an equal sized seat that does not correspond to the boundaries
of their perceived community and daily lives, or one that was
perhaps a bit large but made sense on the ground? How do people
feel about not having the same parliamentary boundaries from one
election to the next? The government appears not to have attempted
to discover what people want from representation.
— The data from the 2011 Census. This would
assist immensely with possible alternative approaches like estimating
and projecting eligible population.
— The capabilities of mapping and other
information technology in improving the process.
A further Speaker's Conference or an Inquiry,
with expert witnesses and an opportunity for different arguments
about technical matters to be tested, would have been preferable
to this Bill.
3.4 The haste extends to the timetable for
implementation. The customary public consultation through inquiries
on boundary changes is dispensed with, and there will not be the
same opportunities as in the past to allow harmonisation with
local government boundary changes (which caused some of the delay
in the recommendations for the metropolitan boroughs in 2004).
3.5 It must be asked whether this expensive
and disruptive process is justified, so soon after the last Periodic
Review was implemented. It would be a better contribution to what
is an important aspect of representation if a more considered
process could take place and the results implemented for the election
after the one projected for 2015.
4. Is there a serious "size problem"?
4.1 The evidence that differing sizes of
constituency is a significant problem is limited. The most quoted
anomalies—the Isle of Wight with 110,000 electors and na
h-Eileanan an Iar with 22,000—are there for specific reasons
and do not reflect a general pattern of widely scattered electorate
totals per constituency. Several of the constituencies at the
low end have already been given special treatment by the Bill.
4.2 More important than a few exceptional
cases is the broad pattern. Constituencies are more or
less equal already—with a few systematic exceptions namely
Wales, the islands and the Scottish Highlands (plus one or two
London boroughs where a convenient pairing was not available).
Leaving aside the appropriate distribution of seats between component
parts of the UK, the Commissions generally do a good job in achieving
numerical equality in their areas. In successive reviews the principle
of equality of numbers has gradually become more central in Commission
policy, for instance in its decision in the 1990s to cross London
borough boundaries.
4.3 A general principle of toleration of
10% variation allows for county boundaries, community identity
and practicality of representation to be taken into account, while
a rigid 5% rule cannot.
4.4 Of the 533 English constituencies in
the last review, 474 (88.9%) were within 10% of the English quota[1]
and according to the Boundary Commission for England's latest
figures available[2]
there were still 429 within this range (80.5%). Only 10 English
seats outside a range of 15% were proposed (one over, nine under)
and on 2010 electorates there were 30 such seats (18 over, 12
under).
4.5 One has to ask whether it is worth imposing
the disruption of an immediate boundary review on constituency
representation across the entire country when the bulk of them
are within 10% of what they "should" be anyway, and
most of the exceptions can be dealt with under existing legislation
with a change in policy by the Commission.
5. How complete and satisfactory is the electoral
register as a basis for precise allocations of seats?
5.1 Broad social change has made it more
difficult to maintain accurate electoral registers than in the
past. It is easy to keep track of relatively static populations,
or those that move according to planned development, as was the
case in decades past and still is the case in many rural and suburban
areas.
5.2 However, the extent of population turnover
in the large cities—especially London—is on a scale
that is unfamiliar. The proportion of people for whom family and
residential circumstances are complicated, or for whom English
is a second language, has increased (again unevenly between areas).
Public attitudes have also shifted. People are less willing to
comply with official demands to fill in forms and less willing
to answer their door or telephone. The Electoral Commission's
research noted that the register was around 98% in the mid-1980s
but has drifted downwards since then, with sharp falls around
1990 and in the early 2000s, to around 90% now.[3]
Fluctuations caused by demographic or administrative factors can
easily lead to electorate numbers varying by more than the government's
chosen factor of 5%.
5.3 The nature of electoral registration
has changed considerably since the last boundary review enumeration
date in February 2000. The first change was the introduction of
rolling registration, which means more fluctuation than before
around the number of electors located in the annual canvass. In
2002 Individual Electoral Registration (IER) was introduced in
Northern Ireland. The Electoral Administration Act 2006 (section
9) imposed a duty on ERO's to maximise valid electoral registration.
Section 11 of the same Act allowed later electoral registration
than had previously been allowed (11 days before poll), so that
people who had been alerted to the election by the campaign and
became aware of their non-registration could still participate.
The register grew by 700,000 between December 2009 and the election.[4]
5.4 Electoral registration will continue
to change. The biggest change will be the introduction of IER
in the rest of the UK. The Political Parties and Elections Act
2009 set out a timetable, agreed by consensus, for introduction
in 2015 with a period of voluntary dual running with the existing
system and the need for periodic measurements of the effect IER
is having on the completeness of the register. In the coalition
agreement, the government parties pledge to "accelerate"
IER. It is not clear what this will involve, and whether safeguards
will be weakened as a result. IER is in itself an extremely complex
public sector IT programme, and if done thoroughly will require
extensive data sharing in government and an assertive outreach
programme to find voters.
5.5 Many countries with similar methods
of electoral registration have either compulsory identity cards
or a system of population registration; neither of these will
be available in Britain. IER done properly will be expensive.
IER done on the cheap will be disastrous for levels of electoral
registration.
5.6 The introduction of IER in Northern
Ireland has had two consequences which render registered electorate
a dubious basis for strict equalisation. One is that there was
a sharp initial drop in the registered electorate. While some
of this reflected fraudulent or dead names dropping out, Electoral
Commission research showed that the proportion of the eligible
population registered dropped from 94.4% to 85.1%in 2002. Under-registration
was worst among young people, with only 71% of people aged 18-24
on the register.[5]
IER has also made the size of the registered electorate more volatile,
as the chart below shows, with the decline and subsequent volatility
being most pronounced in the large city, Belfast (although the
upward spike in 2010 reflects boundary changes as well as actual
registration).
AVERAGE REGISTERED ELECTORATE IN NORTHERN
IRELAND CONSTITUENCIES BY CATEGORY
Table based on statistics from the Electoral
Office of Northern Ireland, accessed May 2010. Categories: Belfast
city—4 Belfast seats. East of Bann suburban—Lagan
Valley, South Antrim, East Antrim, North Down. East of Bann rural—Strangford,
South Down, North Antrim, Newry & Armagh, Upper Bann.
5.7 It is beyond doubt that differences
in the levels of electoral registration already cause some distortion
in the representation of different regions, particularly the under-representation
of London. Applying a rigid standard (except for the islands and
sparsely populated areas) can only worsen this.
5.8 Compiling electoral registers is a local
government responsibility. Funds for electoral registration are
not ring-fenced and because it is not a "front-line"
service in the same way as schools, roads, etc. in a climate of
retrenchment it may be hard for councillors to avoid cutting its
budgets—particularly if there is acute social need in their
areas and attaining a near-complete register is difficult. The
large additional costs and administrative uncertainties of implementing
IER will add to the strain and conceivably lead to years of inadequate,
systematically uneven and widely fluctuating electoral registration
totals.
5.10 Neither the Electoral Commission nor
central government has the power to do more than "name and
shame" councils that are providing ineffective electoral
registration. It is worth noting that ineffective and incomplete
are not the same measure. A highly professional service in an
inner London borough may be doing a good job against overwhelming
odds but still have very incomplete registers, while an ineffective
electoral registration department in an "easier" area
may have a superior rate of registration.
5.11 In summary, the number of registered
electors is an approximation at best, and the stability and accuracy
of this number have deteriorated over recent years and are likely
to deteriorate further in the years to come. Social and administrative
change have both destabilised the idea of a definite "size"
to a constituency based on electoral registration. It is perverse
to insist on closer arithmetic perfection than ever before. The
Bill is analogous to a cookbook demanding absolutely no more than
105g, and no less than 95g, of flour for a recipe when one is
using a standard pair of kitchen scales with a thick needle calibrated
in 10g bands (it may be worse—at least flour does not trickle
around from bowl to bowl of its own volition).
6. The land area constraint
6.1 The Bill proposes (Rule 4) to prohibit
constituencies of more than 13,000km2, and to exempt constituencies
of more than 12,000km2 from the electorate equality rule. It is
reasonable to take physical geography into account in boundary
determination (although it is contradictory to the broad philosophy
outlined by the Bill's authors), but in doing so in several special
cases the Bill raises questions. Upon what criteria did the government
base their thresholds for land area and was this decision evidence
based?
7. Other factors affecting constituency representation
7.1 If the land area constraint and island
exceptions recognise the impracticality of an MP representing
a constituency that covers a large or very divided or heterogeneous
area, the Bill is lacking recognition of other factors that make
constituencies problematic to represent and limit the ability
of constituents to gain access to their MP.
7.2 An MP represents all the local population,
not just registered electors (or even eligible electors) and the
ratio between population and registered electorate varies widely
and systematically. A constituency with 76,000 registered electors
in a rural or suburban area may have a total population of 99,000,
while a central city constituency of 75,000 registered electors
has 122,000.[6]
The practical constraints of representing the interests of 22%
more people than another MP are absent from consideration in the
Bill.
7.3 In addition, the very same factors that
lead to a wide gap between registered electors and population
(ie transient populations, immigrants and asylum applicants, a
large proportion of children, a prison|) will tend to generate
much more than proportionate levels of casework. The hugely higher
workload in urban seats caused by the greater population will
mean unequal levels of representation for each citizen (or non-citizen
entitled to constituency service)—or an MP so burdened by
constituency duties that he or she may be less able to perform
the other functions of an MP. The Bill considers one phenomenon
that makes some seats difficult to represent—a dispersed
population—but not others (population size).
7.4 An overcomplicated administrative geography
may also make constituency representation either spread too thinly
between constituents or too burdensome to the MP. Interaction
with local authorities, and community leadership, are both important
aspects of the duty of the MP, and local government issues generate
casework. This work is obviously easier the fewer local authorities
with which the constituency MP needs to work. The Bill, as its
authors recognise, will involve more constituencies crossing administrative
boundaries—even county boundaries. As well as creating units
that will have a weak sense of collective identity this will mean
that the mechanical process of representation will be more difficult
in some places than others.
7.5 "Equalisation" of registered
electorate at a particular enumeration date does not mean equalisation
of representation—a constituency which has a much larger
population and multiple local authorities will be much more difficult
to represent than one in an area where these factors do not apply.
8. Public involvement
8.1 The outright prohibition on the Boundary
Commissions holding public inquiries is extraordinary. While there
are arguably too many under the current rules, and proceedings
are sometimes political theatre, public inquiries are a valuable
part of the process, and may be even more vital under the new
proposals.
8.2 Broadly, there are two levels of objections
that are raised to provisional boundary recommendations. One is
local and reflects a feeling in a particular community that it
belongs in the same constituency as town X rather than the Commission's
proposed town Y. These are valid views, but they are usually put
forward in isolation of the implications of the change for the
wider pattern of constituency boundaries across the county (or,
now, possibly region or nation).
8.3 The other sort of objection often comes
from political parties or local authorities, and suggests an alternative
scheme for the area, involving several constituencies in a coherent
framework. There is usually more than one set of proposals that
conform to the Rules, and comparing them against the Rules and
against public opinion is vital.
8.4 The public inquiry, at its best, can
be a forum for testing the strength of arguments for the provisional
recommendations and alternative schemes under the Rules, and how
they correspond with other (possibly less self-interested) representations
from the public. Assistant Commissioners often take pains to discount
self-interested pleading and ascertain which plan best fits the
constraints and the realities on the ground, and their work may
or may not be upheld by the Boundary Commission itself. The proposals
in the Bill are much less transparent and more centralising and
top-down. In terms of gaining consent and a sense of ownership
of the proposals in the locality, the level of scrutiny of the
broad pattern and local detail gained from a public inquiry is
sometimes indispensable.
9. The number of MPs
9.1 The rule specifying 600 MPs seems arbitrary.
It is arguable that increasing population, the increasing casework
demands put upon MPs, and the greater demands of scrutiny and
committee work, mean that the 650 MPs of 2010 are doing much more
work than their 650 counterparts in 1983 or the 670 a century
ago. The average number of constituents has risen from 55,000
in 1950 to 70,000 now, and population has also increased steeply.
9.2 Most comparisons with other countries
with smaller lower houses and larger population miss the points
that the US and Germany, for instance, have federal and state
tiers of government, and the legislature in some countries like
the US and France does not supply the ministerial bench.
9.3 A decision about the number of MPs should
proceed from an analysis of the functional needs of Parliament,
and the representative role with constituents, rather than being
arbitrarily imposed.
24 August 2010
1 Rallings, C & Thrasher, M The Media Guide
to the New Parliamentary Constituencies, 2007. Back
2
http://www.boundarycommissionforengland.org.uk/electoral-figures/electoral-figures.htm
as accessed 28 July 2010. These figures differ in detail from
those published by Returning Officers for the 6 May 2010 election,
another indication that "size" is hardly a fixed or
simple matter. Back
3
http://www.electoralcommission.org.uk/__data/assets/pdf_file/0018/87111/The-completeness-and-accuracy-of-electoral-registers-in-Great-Britain.pdf
page 25. Back
4
http://www.electoralcommission.org.uk/__data/assets/pdf_file/0010/100702/Report-on-the-administration-of-the-2010-UK-general-election.pdf Back
5
http://www.nio.gov.uk/electoral_registration_in_northern_ireland_-_consultation_paper.pdf Back
6
Figures are real examples contrasting Colne Valley and Birmingham
Sparkbrook & Small Heath in 2007. Back
| 
