2019 report: Accuracy and completeness of the 2018 electoral registers in Great Britain
Summary
We run accuracy and completeness studies to measure the quality of the electoral registers, and assess how this changes in response to legislative developments and administrative and population changes.
The results for Great Britain in December 2018 show that:
- Parliamentary registers were 85% complete and 89% accurate
- Local government registers were 83% complete and 89% accurate
The findings lead to two main estimates:
- between 8.3 and 9.4 million people in Great Britain who were eligible to be on the local government registers were not correctly registered
- there were between 4.7 and 5.6 million inaccurate entries on the local government registers
This is our first study since the 2015 assessment of the registers, following the transition to Individual Electoral Registration (IER). Since 2015 the levels of completeness have stayed at around the same level, with a not statistically significant decline of around 1 percentage point on the local government registers, while the accuracy has fallen by two percentage points.
Age and mobility
Our research confirms that age and mobility continue to be the strongest variables associated with lower levels of completeness. Plans to reform the annual canvass starting in 2020 are an important first step, enabling Electoral Registration Officers (EROs) to focus resources on households that have changed composition.
Modernising electoral registration
However, there is more that could and should be done to modernise registration processes in Great Britain to provide the best opportunity for ensuring that as many people as possible are correctly registered.
Earlier this year we published the findings from a series of feasibility studies exploring how reforms, including automatic or more automated registration, could be delivered. The studies found that these changes are feasible from a technical and operational perspective and could be implemented without radically altering the structure of the electoral registration system in the UK.
What we mean by accuracy and completeness
The quality of the electoral registers is measured in two main ways: their accuracy and their completeness.
By accuracy we mean that ‘there are no false entries on the electoral registers’. It is the measure of the percentage of entries on the registers which relate to verified and eligible voters who are resident at that address. Inaccurate register entries may relate to entries which have become redundant (for example, due to home movement), which are ineligible and have been included unintentionally, or which are fraudulent.
By completeness, we mean that ‘every person who is entitled to have an entry on an electoral register is registered’. It refers to the percentage of eligible people who are registered at their current address. The proportion of eligible people who are not included on the register at their current address constitutes the rate of under-registration.
Methodology
The accuracy and completeness estimates presented in this report are based on a house-to-house survey of 5,079 addresses in Great Britain across 127 local authority areas. The majority of information was gathered from face-to-face interviews of 4,968 households, conducted by trained interviewers with the aim of gathering information from residents which could then be checked against the details held on the electoral registers. A small number of postal questionnaires were issued in order to attempt to reach those addresses where a face-to-face interview was not achieved; of these 111 were returned. The fieldwork and data analysis from the study was carried out by Ipsos MORI.
This method is consistent with the approaches used for our 2014 and 2015 assessments of the registers. This approach has been validated through a separate study (‘Electoral registration in 2011’) using data from the 2011 Census. The results from that study were compared to those generated from previous house-to-house studies and found a high level of consistency between the results and methods. Therefore, we are confident in the estimates produced using the house-to-house survey approach.
We also compare some specific findings to our study of the 2015 registers in order to provide a sense of how the quality of the registers has changed with time.
Any estimate of accuracy and completeness represents a ‘snapshot’ at a particular moment in the lifecycle of the registers. We have generally taken this snapshot in the period immediately following the compilation and publication of the annual registers when, historically, the completeness of the registers was its highest. Our previous studies on the topic suggest that the completeness of the registers can decline by as much as up to one percentage point a month from the completion of the canvass. We would expect this effect to be less pronounced since the introduction of year-round registration and the introduction of online registration.
In line with previous studies, the findings presented in this report are for the registers published in England, Wales and Scotland on 1 December 2018. We also present national estimates alongside the estimates for the whole of Great Britain.
Using the percentage figures produced from this research, it is possible to estimate the number of people who are not correctly registered, or who have inaccuracies in their register entries. However, it is important to be aware that these can only be estimates for several reasons.
Firstly, both the accuracy and completeness estimates are subject, like any survey findings, to confidence intervals (+/-1.1% and +/- 1.0% respectively). This is because the results are from a surveyed sample of the overall population. The confidence intervals denote the range around any survey statistic where the true population value is likely to lie.
Secondly, related to completeness, it is not possible to determine with certainty the size of the population eligible to register to vote in Great Britain. Eligibility is determined by age and nationality. Mid-year estimate data from the Office for National Statistics, based on the 2011 census, offers the most accurate estimate of the size of the population. However, while annual estimates are made available which include data on age, they do not include information on nationality.
Any attempts to calculate the absolute number of people not correctly registered at their current address is therefore based on an estimate of completeness and an estimate of the total eligible population. It can, therefore, only be an approximation and should be treated as such.
The findings from this accuracy and completeness study lead to an estimate of between 8.3 and 9.4 million people in Great Britain eligible to be on the local government registers who were not correctly registered in December 2018. Our 2015 assessment estimated between 7.8 – 8.3 million.
Although completeness has remained stable, over time the estimated number of incorrectly registered people in Great Britain appears to have increased. This is mostly because the population has grown by close to two per cent between 2015 and 2018. It is also partly because of slightly wider confidence intervals for the 2018 estimates due to a smaller sample size.
A further point to note is that the range presented above does not mean that there should be an additional 8.3 – 9.4 million entries on the registers in total. Those not correctly registered may still be included on the registers but for instance at a previous address (an inaccurate entry).
The estimates also allow for the total number of inaccurate entries on the December 2018 registers to be approximated. In Great Britain it is likely that between 4.7 and 5.6 million entries on the local government registers are inaccurate. Our 2015 assessment estimated between 4.0 and 4.5 million.
This estimate suggests the number of inaccurate entries has increased although, as with completeness, factors such as population growth and different confidence intervals play a role. There are currently fewer inaccurate entries than in 2014, which was the last study carried out before the introduction of individual electoral registration (IER).