7 Proportionality
Introduction
195. In Chapter 3 we noted that the Court, in assessing
the extent of the "margin of appreciation" enjoyed by
member states in interpreting Convention rights, inevitably goes
on to consider what it describes as the "proportionality"
of the measure in question. This assessment of proportionality
was at the heart of the Court's finding in Hirst, in which
the Court accepted the Government's argument that section 3 of
the Representation of the People Act 1983 pursued the "legitimate
aim" of "sanctioning the conduct of convicted prisoners
and also of enhancing civic responsibility and respect for the
rule of law", but held that it was "arbitrary in its
effects". In reaching this conclusion the Court also held
that there was "no evidence that Parliament has ever sought
to weigh the competing interests or to assess the proportionality
of a blanket ban on the right of a convicted prisoner to vote."[204]
196. It follows therefore that if Parliament were
to decide to change the law with a view to complying with the
Hirst judgment, and in the very likely event that further
litigation were to follow, the ECtHR would not only assess whether
the new law was, in its view, "proportionate", but as
part of that assessment would consider the extent to which Parliament,
and this Committee, had conducted a similar assessment.
197. In this chapter we accordingly consider which
of the options for compliance proposed in the course of this inquiry,
either by the Government or by witnesses, would be regarded as
"proportionate". In so doing, we do not pre-judge Parliament's
decision on whether or not to comply with the judgment, which
is a matter for Parliament. Instead we seek to identify a single
compliant option, which we believe to be proportionate, for Parliament
to consider alongside the option of non-compliance.
The meaning of "proportionality"
198. The term "proportionality", like so
many we have encountered in this inquiry, has a specific meaning
in the context of ECtHR case law. As Lord Faulks QC observed,
"rationality and proportionality are rather in the eye of
the beholder … We are speaking a great deal with the language
of the European court, which is very evolved—it has terms
of art and nuances—but not all constituencies are familiar
with the nuances of the European court."[205]
199. What does proportionality mean? In the context
of ECtHR jurisprudence, that Court has regard to a series of factors
when considering whether an interference can be justified. These
will usually include whether the interference with the right is
in accordance with law and is necessary in a democratic society
to pursue a legitimate aim. In such cases, the Court will consider
whether the extent of the encroachment is in reasonable proportion
both to the seriousness of the interest being protected and the
importance of the legitimate aim pursued.
200. Another way to describe the test of proportionality,
in the context of sanctions imposed on offenders, is to ask whether
the punishment (in this case, disenfranchisement) fits the crime.
Such a test has ancient origins, as the Archbishops' Council of
the Church of England reminded us: "The Bible contains insistence
on proportionality in punishment—the limits of the lex
talionis are set at 'an eye for an eye'".[206]
We have adopted this simpler test, whether the punishment fits
the crime, in considering the issue of proportionality in the
remainder of this chapter.
Does proportionality require
individual consideration?
201. The difficulty is to apply even a basic conception
of proportionality to the multitude of unique individual cases
that come before the courts. As we have noted, the ECtHR held
in Hirst that "the principle of proportionality requires
a discernible and sufficient link between the sanction and the
conduct and circumstances of the individual concerned". Such
individual assessments are of course already made in the context
of the primary sanction imposed by the sentencing judge, whether
custodial or non-custodial. In the words of Professor Julian Roberts:
"A sentencing court in this or any common-law
jurisdiction will consider all the factors relevant to the case
in terms of culpability and harm, decisions from the Court of
Appeal, the sentencing guidelines and so on. Those factors will
be taken into account. They will be explicitly identified in the
guidelines and in appellate judgments and they are part of the
case law of common law. To assist judges in this task detailed
sentencing guidelines are published by the Sentencing Council
for England and Wales and analogous bodies in Scotland and Northern
Ireland."[207]
202. Thus every sentencing judge is required to link
the harm done by a convicted criminal to a particular point on
a complex scale of punishment, including fines, community orders,
suspended sentences and custodial sentences ranging from five
days to life. Martin Howe QC argued that "prisoners are sent
to prison for offences which vary greatly in nature and seriousness,
[and] these factors are reflected in the sentence imposed. The
voting disqualification lasts for the length of the sentence,
and so in that respect the period of voting disqualification is
proportionate to the offence."[208]
203. The difficulty in this argument lies not in
the proportionality of the judge's decision to impose a custodial
sentence per se, but in the proportionality (which has been rejected
by the ECtHR) of adding loss of voting rights as an automatic
consequence. Not only is there no reference within the sentencing
guidelines to loss of voting rights as a relevant factor in deciding
between a custodial and non-custodial sentence, but, as Aidan
O'Neill QC noted, "it is not the length of the sentence that
determines the right to vote, it is the length of detention. That
means that somebody sentenced to life will regain their right
to vote once they are out of prison on licence."[209]
Since sentencing judges cannot anticipate the facts that will
ultimately determine an individual's release date (good or bad
behaviour, psychological assessment, expressions of remorse, and
so on), nor can they, at the time of sentencing, predict the effect
of that sentence in respect of the franchise.
204. These factors led some witnesses to conclude
that the test of proportionality required individual case-by-case
consideration, and an explicit decision by a judge distinct from
that on sentencing—thereby adopting the approach of the
ECtHR in the case of Frodl v Austria, an outcome which
was significantly modified in the later case of Scoppola v
Italy (No. 3). The AIRE Centre cited "the crucial importance
of retaining some element of judicial discretion on the decision
to disenfranchise a prisoner",[210]
and Lord Ramsbotham said that "I would leave everything to
do with length firmly in the hands of the judge."[211]
205. The Government, on the other hand, noted the
"additional burden"[212]
such a requirement would place upon sentencing judges, and invited
us to consider whether decisions should be appealed. Professor
Roberts opposed the introduction of a new element of judicial
discretion relating to the seriousness of the offence: "the
seriousness threshold … is problematic. Presumably, the courts
would devise guidance as to the nature of the crimes … that
were so egregiously wrongful as to justify this, but it would
require the exercise of judicial discretion, and it would be tricky
for a court to make those distinctions."[213]
Linking loss of voting rights
to specific offences
206. There are some offences for which disenfranchisement,
with or without imprisonment, could be regarded as obviously "fitting
the crime"—notably those offences which might be seen
as undermining the fabric of democratic social order. It is in
fact already the case that individuals convicted of illegal or
corrupt practices in the conduct of elections are subject to a
disqualification from registering to vote of up to five years,
irrespective of whether or not they have been sentenced to a term
of imprisonment.[214]
A similar approach, but extended to include, for instance, terrorism
offences, has been adopted in Germany.[215]
207. Dr Dirk van Zyl Smit urged us to adopt the same
approach in the UK: "If you were to recommend, say, that
someone convicted of treason or a crime against the state, or
of an electoral offence were to lose the right to vote …
I would be very surprised if that did not stand up [in the ECtHR]
because you could say that this is an appropriate punishment for
that sort of offence." He did not envisage such punishment
extending to other serious offences such as rape: "I would
say that we are talking not about the moral blameworthiness of
the offender but the appropriateness of removing the right to
vote for a particular offence."[216]
208. The implication of Dr van Zyl Smit's approach
is, as the Government put it, that "offenders who might potentially
be serving long sentences [would] retain the vote on the basis
that they were not convicted of a particular offence."[217]
In other words, an offender convicted of a relatively minor offence
relating to the conduct of elections would lose the right to vote,
whereas someone convicted of murder would not.
209. We have noted, in chapter 5, the argument of
Professor Waldron that the vote should be regarded as a "power",
and that there is a legitimate discussion as to "whether
people who have shown disregard for the rights and interests of
others should, for the time being, be entrusted by their fellow
citizens with that minuscule power over them."[218]
This consideration is, we believe, a factor underlying the strong
public opposition to giving convicted prisoners the vote, and
we therefore do not consider that an outcome which would result
in those convicted of the most heinous crimes being able to vote
would be acceptable to either Parliament or the electorate. This
in turn lends support to an approach involving an element of automaticity,
whereby those convicted of crimes of a certain level of seriousness,
who have shown utter disregard for the rights of others, would
be temporarily disenfranchised without the need for separate judicial
decision-making.
Linking disenfranchisement to
sentence length
210. The Government's proposal, as set out in the
draft Bill, is to link loss of voting rights to sentence length,
with the three options setting the threshold at 4 years, 6 months,
or the imposition of any custodial sentence, respectively. In
each case this means, as Professor Roberts noted, linking loss
of voting rights to the seriousness of the crime, and using "sentence
length as a proxy for crime seriousness."[219]
211. The evidence cited in Chapter 6 suggests that
a cut-off which takes effect at the point an offender is given
an immediate custodial sentence may be somewhat arbitrary in its
effects. But the same will be true of any other cut-off set by
reference to sentence length—sentence length will always
be, as Dr Susan Easton put it, a "rudimentary measure with
which to distinguish offenders and offending."[220]
Aidan O'Neill QC noted that "there are going to be people
who fall on the wrong side or the right side of that … rule."[221]
The Government acknowledged accordingly that any ban based on
sentence length risked being regarded by the ECtHR as "general,
automatic and indiscriminate."[222]
212. Wherever the threshold is set, there will be
hard cases just above and just below. Nevertheless, we heard persuasive
arguments in support of the Government's position that sentence
length was the best indicator of the seriousness of the offence.
Professor Roberts acknowledged that courts "think very carefully
about the length of sentence; there is a custodial threshold prior
to the imposition of a term of custody. It is a good index."[223]
Asked about the use of sentence length as a measure of crime seriousness,
Paul McDowell of NACRO said that "I do not think it is ideal,
but I do not know of a better measure."[224]
Lord Faulks agreed: "The particular facts of the case are
much better reflected in the sentence than by the description
of the offence. Although it is arbitrary … if you are going
to have a cut-off point, it is better to do it by length of sentence."[225]
We agree.
Setting the threshold
213. The Draft Bill proposes two possible thresholds:
6 months and 4 years. We have received much evidence on the advantages
and disadvantages of these thresholds, but no conclusive logical
case been made for either of them. In the words of Professor Mike
Hough and Professor Julian Roberts, "there is no significance
[in either]; both are arbitrary thresholds."[226]
214. A threshold of 6 months equates to the sentencing
powers of magistrates' courts in respect of a single offence,
though sentences of up to 12 months may be imposed by a magistrate
in respect of multiple offences—indeed, Professor Hough
and Professor Roberts confirmed that 810 individuals sentenced
in magistrates' courts in 2012 were given custodial sentences
of more than 6 months.[227]
Moreover, as the Government pointed out, "it is possible
for an offender sentenced in a Crown Court to be given a shorter
sentence."[228]
In reality, of 90,386 people convicted in the Crown Court in 2012
only 50,683 received custodial sentences at all, of whom 41,460
(just 46 percent of those sentenced) received sentences of over
6 months.[229] The
link between the 6-month threshold and the sentencing powers of
magistrates' courts is thus weak.
215. The arguments in favour of a 4-year threshold
are little stronger. As the option in the draft Bill which would
enfranchise the largest number of prisoners, it was preferred
by many witnesses. For instance, Angela Patrick, while stating
her preference for a system of judicial discretion, described
4 years as the "most likely to be considered proportionate"
of the options in the draft Bill.[230]
However, the choice of 4 years appears to be, in itself, arbitrary.
The Government noted that both the Criminal Justice Act 1991 and
the Legal Aid, Sentencing and Punishment of Offenders Act 2012
used 4 years as "an indicator of seriousness."[231]
But this is a circular argument, against which must be set the
relatively large number of serious offenders who, if such a threshold
were adopted, would be enfranchised. We see no likelihood that
either Parliament or the wider population will accept a 4-year
threshold.
216. An alternative was proposed to us by Robert
Walter MP, namely that the threshold should be set at 12 months.
In defence of this proposal Mr Walter noted that, as we have shown
in Chapter 2 of this Report, the Forfeiture Act 1870 tied the
loss of voting rights to imprisonment for a felony leading to
a sentence of more than 12 months. The result of this Act, and
subsequent legislation, was that at the time the UK signed the
ECHR prisoners sentenced to terms of imprisonment of 12 months
or less were able to and did vote in elections. Mr Walter accordingly
proposed that "we should go for a return to the situation
that persisted at the time we signed the Convention."[232]
Mr Walter also noted that 12 months was the maximum sentence length
that can be imposed by a magistrates' court, in respect of multiple
offences. This linkage would, in Mr Walter's view, be "easily
understandable, both by our colleagues and the British public."[233]
It would mean that no prisoner would lose the right to vote unless
sentenced by a judge sitting in the Crown Court.
217. Another argument may be advanced in support
of a 12-month threshold: at present, under section 1 of the Representation
of the People Act 1981, any person sentenced to a term of imprisonment
of more than one year shall be disqualified, while detained, shall
be disqualified from sitting in or being elected or nominated
to the House of Commons.[234]
It could be argued that, if a prisoner who has been sentenced
to a term of less than 1 year is entitled to sit as a Member of
Parliament, or to stand for election, that prisoner should also
be able to vote in a parliamentary election (including for himself,
if standing as a candidate). Thus a 12-month threshold would help
in bringing consistency across relevant statutory provisions.
Regaining the right to vote
218. The establishment of a threshold for disenfranchisement
would not rule out the possibility that prisoners serving longer
sentences could regain the vote as they approach the end of their
detention. As we have noted in Chapter 5, such reinstatement of
voting rights could help re-engage prisoners with their responsibilities
to the wider community, thereby contributing to prisoner rehabilitation.
219. We see two possible approaches to prisoner re-enfranchisement.
The first would require prisoner to "earn back" the
right to vote—in other words, a mechanism could be established
whereby prisoners would be able to apply for reinstatement of
the franchise, for instance at any point after becoming eligible
to apply for parole. As part of the application, they would be
required to demonstrate their suitability, either on the basis
of general good conduct, or by completing particular educational
programmes, such as a citizenship course.
220. An argument in favour of this approach is that,
by introducing an element of individual consideration of each
prisoner's needs (comparable in nature to the process whereby
those stripped of voting rights in Italy are able to apply for
their reinstatement), it would help meet the test of proportionality,
as set out in ECtHR jurisprudence. It could also deliver benefits
in terms of prisoner rehabilitation, by providing an incentive
to prisoners to engage with their wider social responsibilities.
Against this approach, on the other hand, is the cost and the
drain on resources inherent in adding a new layer of judicial
or other individual consideration. We also heard forceful objections
from some prisoners to the suggestion that they should be required
to "earn back" the right to vote.[235]
221. An alternative would be automatic reinstatement
of the franchise as each prisoner approaches release—for
example, at a point six months before the scheduled release date.
This approach would be simple and cost-effective: the onus would
be on the prisoner, having reached the scheduled point in their
sentence, and on provision of a certificate to this effect by
the prison authorities, to apply to be added to the electoral
roll in the constituency into which they are scheduled to be released,
on the basis of a declaration of local connection. There would
be opportunities to use this reinstatement of voting rights as
a means of promoting rehabilitation and civic responsibility in
the months leading up to release.
222. The main disadvantage of this proposal is its
unpredictability—scheduled release dates are by definition
prospective, and can be pushed back if a prisoner is guilty of
misconduct or some further offence. It would also require the
use of a prospective address as the basis for establishing a local
connection. As the Electoral Management Board for Scotland noted,
"allowing 'local connection' to a future address will …
involve a change to the legislation defining local connection."[236]
Conclusions of the Committee
223. The test of proportionality requires, in
our view, that the punishment (disenfranchisement) should be a
proportionate means of achieving the aim of sanctioning the conduct
of convicted prisoners and also of enhancing civic responsibility
and respect for the rule of law. This requires that the punishment
should "fit the crime".
224. We do not believe that it is feasible to
require judicial consideration of the possible loss of voting
rights in each individual case. Judges already take full account
of the particular facts of each case in passing sentence, and
we see no case for duplicating this complex and laborious task
in respect of voting rights.
225. We do not consider that linking loss of voting
rights to specific offences seen as undermining the fabric of
democratic social order, while allowing those convicted of the
most heinous crimes, including murder, to retain the vote, would
be acceptable to public opinion or to Parliament.
226. Sentence length, though arguably a rudimentary
measure, remains the best indicator of the seriousness of each
individual offence, and thus the best means of determining the
point at which the prohibition on prisoner voting should take
effect.
227. On balance, if some convicted prisoners are
to be enfranchised, we find the arguments for a 12-month threshold,
reinstating the pre-1967 position and bringing prisoner voting
rights into line with their existing right to sit in or stand
for election to the House of Commons, more persuasive than those
advanced for either a 6-month or a 4-year threshold.
228. We believe that allowing prisoners to regain
the right to vote as they approach the end of their sentence,
by showing that the loss of the right to vote is supplemented
by measures promoting prisoners' reintegration into society, would
assist in demonstrating the proportionality of whatever approach
is adopted to prisoner disenfranchisement.
204 Hirst, paragraph 79 Back
205
Q 41 Back
206
Written evidence from the Archbishops' Council, Church of England Back
207
Q 86 Back
208
Written evidence from Martin Howe QC Back
209
Q 34 Back
210
Written evidence from the AIRE Centre Back
211
Q 23 Back
212
Written evidence from Her Majesty's Government Back
213
Q 91 Back
214
Sections 160 and 173 of the Representation of the People Act 1983 Back
215
Written evidence from Her Majesty's Government; Q 51 Back
216
Q 48 Back
217
Written evidence from Her Majesty's Government Back
218
Q 119 Back
219
Q 84 Back
220
Written evidence from Dr Susan Easton Back
221
Q 41 Back
222
Written evidence from Her Majesty's Government Back
223
Q 85 Back
224
Q 129 Back
225
Q 33 Back
226
Written evidence from Professor Mike Hough and Professor Julian
Roberts Back
227
Ibid Back
228
Written evidence from Her Majesty's Government Back
229
Written evidence from Professor Mike Hough and Professor Julian
Roberts Back
230
Q 43 Back
231
Written evidence from Her Majesty's Government Back
232
Q 138 Back
233
Q 138 Back
234
Erskine May, 24th Edition (2011), p 35 Back
235
See Appendix 5 Back
236
Written evidence from the Electoral Management Board for Scotland Back
| 
