Figure 2.1 below shows the estimates of accuracy and completeness that have been produced between 1966 and 2018. National estimates of completeness have been produced irregularly but have been produced with a degree of uniformity that allows for making valid comparisons over time. However, comparable accuracy estimates have been produced less frequently.
Although different methods have been used to calculate accuracy, the estimates produced in 1981 and subsequently from 2011 to 2018 demonstrate that the accuracy of the electoral registers has remained relatively consistent over time. 123
There have been greater fluctuations in levels of completeness. The highest levels recorded were in 1950 and 1966 (96% by the end of the canvass in both 1950 and 1966) but had declined slightly by 1981 (94%) and again in 1991 (91-93%) and 2001 (91-92%).4
This gradual decline from 1966- 2000 was influenced by an increase in population mobility and the decline in the 1980s and early 1990s has been linked to the Community Charge (commonly referred to as the ‘poll tax’) where an estimated 350,000 people removed themselves from the electoral registers during this time in an attempt to avoid paying the new tax.5
Levels of completeness declined more dramatically after 2001, falling to 85% in 2011. This decline has been associated with population change, population mobility due to changing housing circumstances, a decrease in public engagement with electoral democracy and changing registration practices combined with declining form response rate in the early 2000s.
In the analysis of the 2018 results we consider these factors to place our findings into context of the evolving challenges associated with maintaining accurate and complete registers.
- 1. Todd and Butcher, Electoral registration in 1981 (1981).This study set out estimates for the accuracy of the registers for April 1981 and used these estimates to calculate what the accuracy of the registers may have been at the time of the qualifying date for the canvass (October 1980). The study found that between 10.4% and 13.5% of the names on the April 1981 registers belong to people who by that time were not living at the address listed in the registers. The Office of Population Censuses and Surveys (OPCS) also estimated the proportion of names on the registers that were inaccurate in October 1981 was between 6.1% and 9.4%.↩Back to content
- 2. The 1991 OPCS study also considered redundant entries on the registers in Great Britain at the time of the 1991 census. This found that between 6.0% and 7.9% of the names listed at addresses which were occupied at the time of the census did not correspond with people living there at the time, However unlike the 1981 study, an estimate for the percentage of names on the registers that were listed at addresses which were unoccupied at the time of the census was not added to this percentage. Consequently, the figures do not represent a complete picture of the possible scale of inaccuracies on the registers.↩Back to content
- 3. The method used to calculate the accuracy of the 2014 registers was different from the one used by the Office for National Statistics in 2011.↩Back to content
- 4. P.G Gray and A. Gee, Electoral registration for parliamentary elections: an enquiry made for the Home Office (HMSO: London 1967)↩Back to content
- 5. McLean, I. and Smith, J. The UK poll tax and the electoral register: unintended consequences? (1991), Warwick University, Economic Research Papers, N. 398. McLean and Smith subsequently upgraded this estimate to 600,000 individuals. The Electoral Commission, The completeness and accuracy of electoral registers in Great Britain (March 2010).↩Back to content
In December 2018, there were 45,775,758 entries on the parliamentary electoral registers and 47,785,498 entries on the local government registers in the UK.
The December 2018 registers represented a decline in the number of register entries from December 2017 where the number of entries on the electoral registers in the UK was at its highest following the 2016 EU referendum and June 2017 UK general election.
Overall, there was a 0.8% decrease in the number of entries on the parliamentary registers and a decrease of 0.3% on the local government registers between December 2017 and December 2018.
This follows two years where the registers have increased in size - albeit only a small increase between December 2016 and 2017 when the parliamentary registers increased by 0.8% and the local government registers by 1.2%.1
This study looks at the quality of these December 2018 registers for Great Britain. A separate assessment is published for Northern Ireland.
It is important to note that while the registers may contain more entries than our last assessment in December 2015 that this does not necessarily lead to them being more complete.
This is because no judgement on the accuracy of the entries is made in administrative statistics so all entries are counted equal regardless of their veracity.
This assessment allows us to make a judgement on the quality of the registers and, therefore, determine how complete the registers are.
- 1. https://www.ons.gov.uk/peoplepopulationandcommunity/elections/electoralregistration/bulletins/electoralstatisticsforuk/2018↩Back to content
The electoral registers are a property-based database and, in order to maintain accurate and complete registers, EROs have to add and delete records for home-movers as well as to remove entries for those who have died and add entries for newly eligible electors and those not previously registered.
The figure below presents additions and deletions as a percentage of December local government register entries from 2015 to 2018. This figure shows that there was greater activity in 2015 following the transition to IER and this suggests that the December 2015 registers, the registers on which our last assessment of accuracy and completeness was based, were of a high quality (as our 2015 assessment proved).
Notably, in each of 2016 and 2017, activity around additions was greater in the rolling registration period than during the autumn canvass whereas deletions were still primarily recorded in the autumn canvass period.
However, in 2018 more additions were recorded during the autumn canvass than during the rolling registration period. This difference between these years is likely to be the absence of a major national poll in 2018.
In each of 2016 and 2017 there were high profile electoral events in the shape of the referendum of the UK’s membership of the EU and a UK Parliamentary general election.
There was a high volume of registration activity in the lead up to these events. In 2018 there were local government elections in 150 local authority areas in England and no elections in Scotland or Wales.
This data demonstrates the value of the autumn canvass, particularly in years where there is less election activity, in maintaining good quality electoral registers.
An alternative way to present this information, which emphasises the effect of elections on when additions and deletions occur throughout the year, is to look at the proportion of additions and deletions during the canvass and rolling registration.
This chart shows this for 2017 and 2018. It further demonstrates the above point that in a high profile election year we tend to see additions weighted towards the rolling registration period (61% of additions in 2017), whereas in years with less elections the canvass continues to act as the primary vehicle for additions to and deletions from the register (68% for both).
The electoral registers are a record of the names and addresses of people eligible to vote in elections and referendums. The registers are the foundation of the voting process: they provide the list of those who are eligible to vote, and those not included on the registers cannot take part in elections and referendums.
The registers are also used for other public purposes such as conducting boundary reviews, selecting people to undertake jury service in England and Wales and assisting law enforcement, as well as for credit ratings, and an edited version of the registers, containing only details of those who have not opted out of appearing on it, is available to anyone who wants to buy a copy.
There is no single electoral register in Great Britain but one for each local authority area in England, Scotland and Wales. Each local authority’s Electoral Registration Officer (ERO) is responsible for compiling and maintaining the local list of electors.
Not every resident in the United Kingdom can register to vote and eligibility to register is directly related to the eligibility to vote which differs depending on the type of election. This means that EROs are required to keep:
- a parliamentary register: this is the list of electors used for UK Parliamentary elections
- a local government register: used for elections to the Scottish Parliament and National Assembly for Wales as well as local government elections, mayoral elections and Police and Crime Commissioner elections
EU citizens have been eligible to vote in local, devolved and European Parliamentary elections since 1999. Therefore, EU citizens are entitled to be on the local government register. In order to vote in European Parliamentary elections in the UK, EU citizens must also complete an extra registration form (commonly referred to as a ‘UC1’) to declare that they intend to vote in the UK rather than in their country of origin.
The table below sets out entitlement to vote by citizenship. Those not listed on this table are not eligible to be on either of the two registers.1
Franchise for citizenship and elections
|Citizenship||Parliamentary Register||Local Government register|
Notes: Commonwealth Citizens resident in the UK who either have leave to remain or do not require such leave
British citizens living abroad who have appeared on a UK electoral register in the last 15 years (or were resident in the UK and too young to be registered at the time of residence) are also entitled to be on the electoral register as an overseas elector and can vote at European Parliamentary and UK Parliamentary elections.
They are registered in the same parliamentary constituency as before they went abroad or, if they were too young to register, in the constituency where their parent or guardian was registered.
In England and Wales, the age at which citizens become entitled to vote is currently 18, but the electoral registers also include records of ‘attainers’ - 16 and 17 year olds who will turn 18 within the twelve month period starting on the 1 December after they make their application.
In Scotland, the voting age for Scottish Parliamentary and local council elections is 16. This means that all 16 and 17 year olds in Scotland are entitled to be on the local government register and that attainers are 14 and 15 year olds who will turn 16 during the twelve month period starting on the 1 December after they make their application.
However, entitlement for the parliamentary register in Scotland has not been affected by the change to the franchise so 16/17 year olds are attainers on the parliamentary register.
Proposed franchise changes in Wales will mean that those aged 16 and 17 in Wales will have an equivalent entitlement as in Scotland.
- 1. There are some specific exceptions to these rules. For example, convicted prisoners lose their right to vote while they are imprisoned regardless of citizenship.↩Back to content
The electoral register is a property based database, with register entries linked to a property. This means that the quality of its information is affected by ongoing population change and EROs must add and delete records for home-movers, remove entries for electors who have died and add entries for newly eligible electors.
In Great Britain a new register is prepared annually and reviewed most months. A key element of preparing the annual register is through house-to-house canvassing. The majority of canvass activity takes place between July and November.
Each ERO is required by law to conduct an annual canvass of all properties in their area to confirm their electoral register entries and to identify electors who have moved or were not previously registered.
During the annual canvass, EROs are required to send a Household Enquiry Form (HEF) to all households (and follow-up with non-responders). This form requires the respondent to reply to the ERO with the names of any individuals no longer resident and of those who are now resident and eligible.
Since the introduction of IER it is no longer possible for a person to be added to the register as a result of their name being added to a returned HEF. Where returned HEFs include new (non-registered) individuals, a second communication – an Invitation to Register (ITR) – is sent out with a registration application for such individuals to complete. It is this application that asks for the person’s National Insurance Number (NiNo) and date of birth which allows them to become registered. Individuals can apply to register online.
An ERO also cannot delete an entry based solely on a name being crossed off on a returned HEF (as they could under household registration) – so they will need either a second source of information or to carry out a review of the individual’s entitlement to remain registered.
The 2020 annual canvass commencing from July 2020 is expected to be the first under a reformed canvass model. The new canvass model aims to allow EROs in Great Britain to better focus their resources in the areas of greatest need, and better meet the objective of the canvass than under the current system.
The idea is that the canvass process will be streamlined for those properties which are likely to have not changed household composition, allowing the ERO to target their resources to where responses and updates to the electoral register are required. There is also an exemptions process for certain types of properties less suited to traditional canvassing methods – for example, care homes, Houses in Multiple Occupation and student halls of residence – which allows the ERO to seek information on residents in a different way.
We see these reforms as an important first step in modernising electoral registration. Read our full response to the UK Government’s proposals for reform of the annual canvass process
Since 2004, the Electoral Commission has been the principal body in the UK undertaking research into the electoral registers with the long-term aim to:
- provide an overview of the accuracy and completeness of the registers in Great Britain and Northern Ireland
- inform the Commission’s guidance and support for Electoral Registration Officers (EROs)
- provide up-to-date information on those groups which are more likely to be under-registered and thereby inform our approach to public awareness activity around elections and voter registration
- provide ongoing tracking of how electoral registers change in response to legislative developments and administrative or population changes and use this tracking to inform our role in scrutinising proposals and policies to revise the registration system
This study of the 1 December 2018 registers establishes a baseline as we seek to measure the impact of upcoming changes to the annual canvass.
In order to measure the impact of reforms to the annual canvass we anticipate conducting two studies on the accuracy and completeness of the registers:
- ‘Before measure’: Presented in this report measuring the 2018 registers ahead of reform to the annual canvass.
- ‘After measure’: in order to assess the impact on the reforms to the annual canvass which we expect to be introduced in 2020.
We will also be collecting a variety of other data and working alongside EROs to understand the impact of the reformed canvass processes and how they are working in practice