Electoral Registration (Hansard, 2 June 1997)

§ Mrs. Liddell

The information requested falls within the responsibility of the Chief Executive of the Office for National Statistics. I have asked him to arrange a reply to be given.

Letter from Tim Holt to Mr. Keith Vaz, dated 2 June 1997: The Chancellor of the Exchequer has asked me to reply as the Director of the Office for National Statistics (ONS) to your recent question asking how many of those eligible to vote he estimates were not registered to vote at the time of (a) the 1987 General Election, (b) the 1992 General Election and (c) the 1997 General Election. 91W There is not enough information to give a reliable figure for those missing from the Electoral Register. It is not possible to compare registration figures with estimates of the resident population directly because these two sources are not compiled on the same basis. The differences between the coverage of the Electoral Registers and the mid year population estimates have been published in Appendix B to the ONS publication '1996 Electoral Statistics' series EL no. 23, a copy of which is attached. An approximate figure has been calculated for the UK resident population who were eligible to vote in 1987 and 1992 and, relative to the number of electors registered in 1987 and 1992, this suggests that 2.6% and 4.5% of those eligible to vote were not registered in these two years respectively. An estimate cannot be made for those registered at the time of the 1997 General Election because mid year estimates for the resident population at the preceding mid year are not yet available. This will be possible to do when the ONS publish the 1996 mid year population estimates which is planned for August 1997.

The table below shows the numbers of parliamentary electors in Great Britain between 1976 and 1996 collated by ONS and the General Register Office for Scotland from returns made by local Electoral Registration Officers. These are compared with roughly corresponding estimates of the resident population.
	Parliamentary electors (Thousands) (a)	Approximate corresponding resident population at preceding mid-year¹ (Thousands) (b)	(a) as percentage of (b)
1976	39,739	40,870	97.2
1981	41,069	42,108	97.5
1982	41,405	42,369	97.7
1983	41,642	42,584	97.8
1984	41,906	42,859	97.8
1985	42,048	43,167	97.4
1986	42,305	43,470	97.3
1987	42,563	43,722	97.4
1988	42,596	43,965	96.9
1989	42,492	44,164	96.2
1990	42,533	44,360	95.9
1991	42,424	44,502	95.3
1992	42,583	44,628	95.4
1993	42,565	44,713	95.2
1994	42,624	44,781	95.2
1995	42,727	44,861	95.2
1996	42,808	45,012	95.1
¹ Persons aged 17 and over, together with 63 per cent of those aged 16, who became 18 during the period when each register was in use. No account has been taken of adults who died between the mid-year reference date for population estimates and the 10 October qualifying date for the following year's electoral register.

The ratios in the final column suggest that coverage of the electoral register has been steady in recent years, after a decline in the period between 1986 and 1991. However, these ratios need to be treated with care and do not represent the proportion of eligible persons who were on the register at their qualifying address for a number of reasons: the electoral register (the numerator) contains some redundant entries for people who had already moved elsewhere or died at the time when it was compiled but whose names were nonetheless carried forward to a new register; the electoral register contains some legitimate dual registrations, for example by students and owners of second homes; the electoral register, but not the population estimates (the denominator) includes overseas voters and service voters who are living abroad; the population estimates include residents who were ineligible to be on the electoral register because they were neither British Commonwealth nor Republic of Ireland citizens. 92W The effects of these various factors, which operate upon the ratios in both directions, are not known precisely. A small sample survey (Electoral registration in 1991—S. Smith, ONS (1993)) carried out by ONS following the 1991 Census of Population gave a central estimate of 92.9 per cent. (standard error 0.36 per cent.) for the proportion of eligible residents in Great Britain who were registered as electors at their usual addresses. This figure needs qualification, too. It relates to people who lived in private households, and it does not cover people who were not enumerated in the Census. On the implausibly extreme assumption that all people missed by the Census were also missing from the electoral register where they lived the proportion of eligible people covered by the latter could have been as low as 91.0 per cent. Those living in London, attainers and young adults, people who had changed their residence within the last year, New Commonwealth citizens in the non-white ethnic groups, and those living in rented accommodation were all found more likely than other people to be missing from the electoral registers.