Select Committee on Home Affairs Fourth Report


SECTION C: IMPROVING THE ACCURACY OF THE REGISTER

The present accuracy of the register

30. There are a number of ways in which the present register at the time of an election fails to record the actual total of persons otherwise entitled to vote at that time.[54] The principal ones are:

(i)  Excess inclusions: some names will appear on the register but will not in fact represent an eligible voter; such mistakes could include:

    (a)  children who will not reach the age of 18 during the lifetime of the register

    (b)  foreign citizens (though Commonwealth or Irish citizens are eligible, as are EU citizens at local or European elections)

    (c)  people who have died or moved away, but who are re-registered because a returning officer has not been informed and judges that the failure to return the form is likely to be an error

    (d)  persons who have mistakenly been included more than once on a register (though double registration is permitted for those who can genuinely claim to be resident at more than one address).[55]

(ii)  Omissions: a large number of other eligible electors may not in practice appear on the register; the principal groups include:

    (a)  any homeless persons who have failed to establish the necessary residence qualifications

    (b)  any British voters overseas who fail to register

    (c)  some vulnerable persons (i.e. people who choose not to register because they do not wish their name and address to become publicly available)

    (d)  other deliberate avoiders (i.e. those who fail to register because they wish to avoid the obligations or perceived obligations which can attend registration, such as jury service or liability to local taxes)

    (e)  other non-deliberate avoiders (a large number of other people—often from particular groups, such as those under 25, or certain ethnic minority groups—who, research has shown, fail to be picked up by the registration process)

(iii) Lack of currency: the register is a list of persons resident at the address specified on a single night (October 10th), some 4 months before it comes into force (February 16th the next year) and some 16 months before it ceases to be in force (February 16th the following year); it will therefore

    (a)  include a number of names of people who have died since October 10th

    (b)  include an even larger number of names of people who have moved (and who therefore remain entitled to vote but who are no longer registered at their actual address) and

    (c)  exclude any persons who have attained the age of 18 since October 10th but who failed to have their names included on the form as potential voters.

31. Clearly some of these inaccuracies are more significant than others, and some may be more avoidable than others. Any measures of the inaccuracies can only be very approximate. For example, no figure is known for the number of names which appear twice, or for the numbers of homeless persons who are unregistered. It is also difficult to know how much adjustment should be made for people who have moved: such people will remain on the register at their previous address and will be entitled to an absent vote there, but in practice many will be much less likely to vote than if they were registered at their new address. Additionally, it can only be a matter of opinion as to how far all British citizens resident overseas for up to 20 years—who are entitled to register for an absent vote—should be regarded as part of the theoretical correct total of possible voters.

32. If non-registered overseas voters are excluded from the figure for the total entitled electorate, broad estimates can nevertheless be made by comparing information on actual registration with the known adult population as revealed by the census (and associated quality checks).[56] The most recent such study, based on the 1991 census, suggested an under-registration rate at that time of between 7.4% and 9.0%.[57] Applied to the population estimate at the time the 1997 register was prepared, this means that there may have been around 4 million people who were totally omitted from the register and therefore unable to vote at the 1997 general election.

33. This total is magnified further if a figure is added in for those who changed their residence between the October qualifying date for the register and the time of the election. It is estimated that around 4 million adults move in the course of a year, though a quarter of these move within the same electoral ward.[58] The number of potential voters who, although on the register, could not vote at their current address at the May 1997 election may thus have been over 2 million,[59] of whom over 1½ million could not vote in their own ward. The total number of potential voters who could not vote at all or could only vote if they took the necessary steps to vote in a different ward at the 1997 election may therefore have been around 5½ million.[60]

Current methods of preparing the register

34. In preparing the register, registration officers are under a statutory duty to "have a house to house or other sufficient inquiry made as to the persons entitled to be registered",[61] but beyond this they have a discretion as to how to carry out the duty. In practice, the canvass everywhere is based on distribution to, and return from, each household of a form every year on which householders list the names of those coming within the grounds of entitlement to vote. Different registration officers have different practices in respect of how to follow up the non-return by households of their form. There is widespread agreement that personal calling by canvassers to recover the forms is among the most effective ways.[62] Useful techniques to maximise the response include the pre-printing of the previous year's information on the form, drawing people's attention to the possible disadvantages of non-registration in respect of obtaining credit references, and paying canvassers by results.[63] Our attention was also drawn to the scope for making greater use, where legally possible, of records held by other public authorities or utilities as a means of verifying information held. A number of witnesses[64] made this point, and we recommend that consideration be given to giving registration officers further legal rights to access to specified relevant records held by such bodies.

35. It is clear that many registration officers are open to new ideas about how to prepare a register which is as accurate as possible, and Mr Monks, of SOLACE, thought that the quality of registration work was improving.[65] We are concerned however about whether best practices are being followed throughout the country. The political parties suggested to us that there was more that could be done and that there was room for greater consistency in methods.[66] We do not challenge the principle that a measure of discretion should rest with the registration officer, since the circumstances and problems faced will be different in different kinds of constituency or authority, but we do consider from the evidence we have received that more steps can be taken by the less successful authorities to bring themselves up to the level of the more successful. We note there are already mechanisms in place for bringing good practice to the attention of all areas through guidance issued by the Home Office, through the AEA's own work,[67] and through the annual ONS publication on register compilation methods.[68] We conclude however that the time has come whereby consideration should be given to the devising and imposition of national minimum standards for the work of registration officers. We note that the Minister recognised the case for the introduction of common standards.[69]

36. Such standards would help to raise levels of registration. But there are two further benefits which they might generate. First, there is some evidence that, because currently there are no minimum required standards, local authorities find it difficult to give financial priority to registration work;[70] if there were standards which they were obliged to achieve this position could change. Equally, agreed standards could prevent individual local authorities devoting too many resources in pursuit of unnecessary accuracy.[71] Secondly, minimum standards would contribute towards greater consistency in the methods and equipment used by different registration officers, which would help to prepare the way for some of the more extensive possible reforms, such as a rolling register.

Young people

37. We have been told that some sectors of the population are disproportionately under-registered. For example, there is a particular shortfall in young people who are registered,[72] with under 24s being consistently the age group with the lowest levels of registration.[73] The Labour Party reported anecdotal evidence of numbers of young people within registered households not being included in the register.[74] Some have expressed fears that there is a danger of the development of a whole generation of people uncommitted to the democratic process. We have already recorded above our support for measures to instil a sense of civic responsibility into the young through the curriculum. The Minister told us that in his opinion registration of young people was an "area of work ... we have got to develop much more sharply and much more effectively".[75] We recognise the force of Lord Parkinson's observation that low levels of voting amongst young people should not be too great a cause for concern since they will be more likely to vote as they get older[76] but, nevertheless, we feel there are no grounds for complacency. We conclude that particular efforts must be made to improve registration rates amongst young people. We note and welcome the fact that 16-24 year olds are one of the main target groups—the other being ethnic minorities—for the Home Office's (limited) publicity campaign for the 1998 registration exercise.[77] Part of the problem is to ensure that 'attainers'—i.e. those who are under 18 at the time of the qualifying date for the register but who will become 18 during the lifetime of the register—are fully registered; for this group at least it should be possible to improve registration rates by work in schools and by use of school records.

A 'rolling register'

38. Improvement in register preparation techniques will help to address the problem of inaccuracies in the register in terms of the extent to which it reflects the actual situation on 10 October. But no amount of such improvement will address the fact that under the present rules the register will be out of date by the time it comes into force the following February,[78] and will be yet further out of date by the time it expires one year later. It was argued that the present system failed to reflect changes in society over the century, in particular in the degree of mobility in the population.

39. The solution to this problem, in principle, is the introduction of a 'rolling' register. This involves abandoning the concept of basing the register on a single annual qualifying date, and introducing a register which can be continuously amended and updated. This would allow the quick correction of any errors, the incorporation of information on deaths among registered electors, and the accommodation of those changing residence. There would be some extra costs involved, though the Association of Electoral Administrators and SOLACE did not think that these would necessarily be substantial.[79]

40. Introduction of a rolling register, at least on an experimental basis, was supported by academic witnesses,[80] by the local authority and electoral officials,[81] and in principle by the representatives of political parties,[82] subject to two points we discuss below. There is no doubt that a rolling register will bring major advantages in improving the currency of the electoral register, thereby removing a major source of current dissatisfaction. The Home Office Working Party, following consultation with local authorities, has recommended the introduction of a system of rolling registration.[83] We support the introduction of a rolling register, and we therefore welcome the decision of the Home Secretary's Working Party to recommend its introduction. There will be a range of practical issues which will have to be addressed—such as the effect of a rolling register on the existing annual canvass, the mechanism for notifying moves, the process for publishing changes and allowing challenges—and which are currently being examined by the Working Party, but the evidence we have received suggests that these will not present insuperable obstacles.

41. There are however two points on which we should comment. The first concerns the technical readiness of local authorities to move in this direction. Although we were told that most local authorities were already equipped to allow introduction of a rolling register,[84] it was noted in the evidence that such a register would involve closer cooperation between local authorities than has hitherto been necessary, in particular greater compatibility between the IT systems used to support the registration process. Such compatibility was not yet sufficiently widespread because the equipment and software used by electoral registration officers depended on what was used by the rest of their particular local authority rather than on what was used by other authorities' electoral registration officers.[85] The point at issue is not just one of administrative efficiency—it is also a matter of the integrity of the register. If an elector moving house is to be entered on the register at the new address almost immediately, then it is important that that person is also deleted from the register at the old address at the same time. The Conservative Party submission placed some emphasis on this, noting that "there will need to be clear safeguards built in to maintain the integrity of the electoral register, which will become more prone to illicit interference or allegations of such interference"; they reported concerns from other countries "that the introduction of a rolling register has led to abuses of the system, particularly with regard to multiple registration".[86] We agree that the integrity of the electoral process must be defended, and conclude that moves towards a rolling register must only go forward at a rate which can be properly supported by the technology in use by local authorities and which commands the confidence and the consent of all the main political parties. The aim should be to reach a position as soon as possible where new residents can move their registration from their old to their new constituency in a single transaction at any time of year within one to two working days.

42. The second issue concerns the cut-off point beyond which changes to the register would not be allowed. It is important that there is no danger of people seeking to manipulate the electoral process by finding improper ways of claiming a registration in an area after an election has been called. The election administrators suggested that it would be possible under a rolling register to continue to make changes right up to the closing date for the receipt of nominations (i.e. 11 days before the poll). It is envisaged that polling cards could be sent out to everyone on the register on the day an election is called, thereby alerting those who did not receive one to the need to register if they wished to have a vote. It is not clear whether changes made after an election is called would be limited to the correction of errors on the register as it stood on that day, or would accommodate also further changes in residence arising after that date. We would be unhappy about allowing electors to be placed on a register because of an entitlement which arose only after an election has been called[87] because of the danger—albeit small at a General Election, but perhaps slightly less so at a by-election—of people deliberately establishing a temporary residence in a particular constituency in order to obtain a vote.

The residence qualification

43. The requirement that, in order to be registered, a person must establish a residence qualification in the relevant electoral area means that certain sectors of the population find it more difficult than others to get on to the register. This applies particularly to the homeless, but can affect other groups also, such as some patients in mental hospitals and remand prisoners.[88]

44. The problem of the homeless was a source of concern to several of our witnesses. The National Homeless Alliance drew our attention to their 1995 report on this issue which called for amendment of the law to ensure that homeless people had a right to vote;[89] they argued that "the franchise should not be dispensed according to housing tenure, but citizenship".[90] The representatives of all three major parties suggested that ways should be found to get round the problem,[91] though it was emphasised that it must be done in a way which did not give rise to opportunities for fraud.

45. Local authority representatives drew our attention to the fact that in some areas registration officers were already applying the existing rules flexibly, through using temporary hostel addresses and so forth, though they were not certain that these methods would all stand up in court.[92] The Home Office indicated that case law already established that "so long as a person can be said to be resident at a particular place, then the quality of accommodation is not relevant", thereby potentially allowing even some rough sleepers to be registered, though they accepted that in practice registration was difficult for the homeless.[93] The National Homeless Alliance suggested that electoral registration officers around the country held differing interpretations of what was and what was not permitted and that Home Office guidance failed to reflect fully the flexibility which was open to registration officers to register homeless people.[94] We consider that registration officers should not be put in a position where they have to risk breaking the rules as a consequence of being more flexible in defining a 'residence', since it is important that homeless people are not effectively disenfranchised; we recommend that, while incorporating appropriate safeguards, the Representation of the People Acts be amended accordingly. Mr Harry Barnes MP has proposed that one solution might be to introduce a system whereby an elector could make a duly attested "declaration of locality".[95] We note that if there were a rolling register then the problem would anyway be reduced since it would be easier to regard temporary addresses as a residence.[96]

46. Psychiatric patients resident for long periods in psychiatric hospitals face even greater difficulties in being registered.[97] Such persons who are compulsorily detained may not give their hospital as an address for registration purposes, even if they are away from their former home for so long that they would have difficulty in gaining a registration there.[98] There is however a special statutory procedure available to those who are voluntarily resident, enabling them to be registered at what would otherwise be their home address so long as they can make a 'declaration', involving such details as their name and their former address, without 'assistance'.[99] MIND have argued that this requirement in practice places a condition for registration for one sector of the population to which other sectors are not subjected.[100] Although the motivation for these rules may be understandable, the restrictions may have reflected in part a perceived need to avoid distortion to the electorate in particular areas at a time when there were more large psychiatric hospitals with substantial numbers of residents than is currently the case.[101] It is certainly not true that all psychiatric patients are suffering from conditions which would prevent them from exercising serious political choices. We recommend that the restrictions on using psychiatric hospitals as residences for the purpose of electoral registration be removed; alternative kinds of restriction, not based on residence grounds, should be devised if it were thought necessary to bar from voting certain patients whose condition made it inappropriate for them to have a vote.

Registration of vulnerable persons

47. There are a number of categories of person who, for various reasons, may be reluctant to register because they do not wish their name and—more specifically—their address to appear in a publicly available document. The most frequent causes for this are victims of domestic violence wishing to remain untraced by their former partners, women living alone, and persons in occupations which may make them vulnerable to attack such as police or prison officers. Such people may wish to retain and exercise their full electoral rights but may be reluctant to do so under the present arrangements and accordingly do not register at all. Such persons must be distinguished from those not registering for other reasons, such as those seeking to avoid tax or jury service.

48. The Local Government Association told us that a significant number of their members favoured finding some way of anonymising the registration process in appropriate cases, or at least a method of registration without revealing addresses.[102] SOLACE noted again that some authorities already tried to find ways of addressing this problem,[103] but were not confident that they were always acting within the rules.[104] The Home Office observed that, although there was relatively little hard evidence suggesting that the problem was yet a major one, there was some concern that the dangers were increasing as new technology made register information more easily available in electronic forms; they noted also that anonymous registration was permitted in some countries.[105] In his oral evidence, the Minister indicated that he was aware of the concern about this issue but that the Working Party had not yet addressed the matter.[106]

49. We recognise the strength of the calls for a means of registering without having to reveal information which could make the elector vulnerable, and consideration should be given to developing such a system. However, there is a point of principle involved in that it is important that the electoral process should be as open as possible and the DETR consultation paper, in discussing this possibility, rightly in our view noted that "it would be necessary to ensure that safeguards against electoral fraud remained adequate". For this reason we consider that any such system should allow for anonymous registration only in exceptional circumstances. The Association of Council Secretaries and Solicitors suggested, for example, that it might be made available only on the recommendation of the Chief Constable for the area.[107]

50. An alternative—or additional—approach would be to restrict the availability of the register to the public. The Working Party is consulting on this point, inviting comments on how far public concerns about vulnerability were justified, and what the implications would be for commercial and other users of the register if it were withdrawn from sale, or if some names were excluded.[108] However, again because of the importance of openness in the electoral process, and so that political parties are able properly to check the lists and to canvas voters, we see no alternative to the general principle that copies of the register should be generally available, at least for electoral purposes.

Penalties and incentives

51. Although voting is not compulsory, registration is. However, very few prosecutions are ever taken out against citizens for non-registration.[109] Local authority witnesses explained that the principal reason for this is that the costs of bringing a prosecution are too high when compared to the level of penalties imposed by the courts.[110] The maximum fine imposable by the courts is a level 3 fine, currently £1,000.

52. We detected no serious calls among our witnesses for offences of non-registration to be pursued more vigorously by registration officers or for courts to impose heavier fines.[111] The Association of Council Secretaries and Solicitors observed that a reluctance to prosecute reflected also a "tradition of non-compulsion and concerns for civil liberties" in this country.[112] We broadly concur, in that we do not think it is necessary or desirable for the state to be heavy handed in pursuing non-registration through the courts. We note the Government's view, expressed in the DETR's Modernising Local Government consultation paper, that even if the power to prosecute is rarely used it is nevertheless useful as an effective threat.[113] Nevertheless, it is difficult in the absence of any up to date information to assess whether present practice is reasonable, and we therefore suggest that the Home Office should gather reliable statistics from local authorities about the numbers of prosecutions and levels of fines.

53. An alternative to penalising someone for not registering which has been floated might be to provide some form of incentive for those who do. Such an approach could apply also to voting, as opposed to registration. Theoretically, for example, it might be possible to give those who register, or those who vote, a small reduction in council tax.[114] There would however be numerous difficulties in the way of such a scheme, such as allowing for the fact that most individuals (as opposed to most households) do not pay council tax, and the need to avoid any appearance that a government or council introducing such a measure might be doing so for party advantage. We are not aware of any such scheme which has been sufficiently thought through to merit serious consideration at this stage.


54  See inter alia paper submitted by Mr Harry Barnes MP (Appendix 15). Back

55  Such as some students or second home owners; such voters can vote at both addresses at local elections, but only once at national elections. Back

56  The Home Office/ONS data can be used to produce an annual indicator of the accuracy of the register based on estimated population figures. These suggest under-registration rates of around 5% (See Appendix 1 Table 6). These are less accurate than the comparisons based on the census and do not make allowance for dual registration, but can be useful in showing trends; they suggest that under-registration may have doubled since the early 1980s. Back

57  Electoral registration in 1991 (S. Smith, OPCS). Back

58  OPCS 1991 Census Report (Part I) Table 15. Back

59  Seven twelfths of 4 million, rounded down. Back

60  i.e. roughly 4 m. not on the register at all plus 1½m. registered in a different ward. Back

61  Registration of the People Act 1983, s. 10. Back

62  90% of all canvass returns in N. Ireland are collected in person (see Memorandum from N. Ireland Office); see also Appendix 17 (Professor Denver), Q 255 (Mr Bambrook, for the AEA). Back

63  For fuller discussion of the details see Appendix 1, annex C (Home Office memorandum); Appendix 10 (Local Government Association); QQ 251-267 (SOLACE and AEA); Enhancing local electoral turnout by Rallings, Thrasher and Downe (York Publishing Services Ltd., for the Joseph Rowntree Foundation, 1996) Chapter 1; Compiling the electoral register (Office for National Statistics, annual publication); and Appendix F to Agenda for Change: the Report of the Hansard Society on election campaigns (Sep 1991). Back

64  See for example Appendix 12, para 5 (Association of Council Secretaries and Solicitors). Back

65  Q 256. Back

66  Appendix 8 (Conservative Party); Appendix 7 and QQ 404-409 (Labour Party); Appendix 9 and Q 404 (Liberal Democrats). Back

67  See Appendix 9 para 3.5; see also Q41 (Dr Butler). Back

68  See for example Compiling the electoral register 1996, Office for National Statistics (TSO, 1997). Back

69  QQ 504 & 506 Back

70  See Appendix 7 (Labour Party). Back

71  Q 411 (Lord Parkinson). Back

72  Other groups include ethnic minorities, and inner city residents; see for example Appendix 1 para 3.31 (Home Office), Q 44 (Dr. Butler), Appendix 12 para 5 (Association of Council Secretaries and Solicitors). Back

73  In a MORI poll of April 1997 6% of people aged 18-24 admitted they were not registered to vote, more than for any other of the five age groups into which the population had been divided; in a similar survey in April 1998 this had risen to 36% of those aged 18-24 (paper submitted by MORI, see List of unprinted memoranda). Back

74  Appendix 7. Back

75  Q 515. Back

76  Q 384. Back

77  Home Office circular RPA 420 (31 July 1998). Back

78  Leeds City Council in their memorandum noted that since "some local authorities start collecting information for the new register as early as August, the register can be at least six months out of date by the time it comes into force" (see List of unprinted memoranda). Back

79  Appendix 5 para 4.15ff. and QQ 268-271 (though see Association of Council Secretaries and Solicitors, Appendix 12, section 7). Back

80  See Appendix 2 (Professor Blackburn); Appendix 17 (Professor Denver). Back

81  Appendix 6 (SOLACE); Appendix 5 (AEA); Appendix 10 (Local Government Association); Appendix 12 (ACSS). Back

82  QQ 403-406; see also Appendix 7 (Labour Party). Back

83  Home Office Working Party interim report, August 1988. Back

84  Memorandum from Leeds City Council (see List of unprinted memoranda). Back

85  Q 503 (Home Office). Back

86  Appendix 8, and APS 12(A). Back

87  Other than for someone attaining the age of 18 after that date; special consideration may be needed to address the situation where a by-election is known to be pending but has not yet been formally called. Back

88  For the difficulties faced by remand prisoners, see Appendix 14 (Prison Reform Trust) and QQ 519-520 (Mr George Howarth MP and Mr Limpkin). Back

89  No Home, No Vote, No Voice: Challenging the disenfranchisement of homeless people, CHAR (since renamed National Homeless Alliance) 1995. Back

90  Appendix 13. Back

91  QQ 419-421. Back

92  Appendix 6 (SOLACE); Appendix 5 para 4.22 (AEA); Appendix 10 para 4.11 (Local Government Association); see also Appendix 15, section 6 (evidence from Mr Harry Barnes MP). Back

93  Appendix 1, para 3.24. The National Homeless Alliance referred to the case of Lippiatt v Electoral Registration Officer, Penwith District Council (March 1996). Back

94  Appendix 13. Back

95  Appendix 15, section 6 Back

96  See Q 421. Back

97  We were told that there were 27,500 compulsory admissions and over 300,000 voluntary admissions in 1994/5, though the latter figure would presumably include some very short stay patients and repeat admissions; see QQ 165 and 168. Back

98  Representation of the People Act 1983 s. 7(1). Back

99  Other than physical assistance, RPA 1983 s.7 (2). Back

100  1996 MIND briefing paper The Right to Vote (see List of unprinted memoranda). Back

101  Q 438, Q 518. Back

102  Appendix 10 para 4.12. Back

103  One possible technique was to register women under a maiden name. Back

104  Appendix 6. Back

105  Appendix 1, para 3.22. Back

106  Q 510. Back

107  Appendix 12, section 9. Back

108  Details on the current rules governing the supply and sale of the registers are at Appendix 1 (Home Office memorandum, annex D). Back

109  There were 3 prosecutions in 1988 and 13 in 1989, though records are no longer kept centrally (Written Answer Official Report 30 June 1998 col 140); see also Q 45 (Dr Butler). Back

110  Appendix 6 (SOLACE); Appendix 12, section 8 (Association of Council Secretaries and Solicitors). Back

111  Although the Labour Party submission did suggest that penalties on property owners who deliberately withheld information about the numbers of young people or second families within their rented out properties should be made sufficiently severe to act as a disincentive: Appendix 7. Back

112  Appendix 12, section 8; likewise, SOLACE described prosecution for non-registration as "heavy-handed". Back

113  Para 3.17. Back

114  To avoid a substantial public expenditure cost, any such reduction would have to be offset by higher Council Tax payments by others. Back


 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries

© Parliamentary copyright 1998
Prepared 1 October 1998