Board notes: 25 September 2019 informal presentation/discussion for Commissioners on accuracy and completeness survey

The Head of Research introduced the presentation, and explained that the purpose of the research was to provide an overview of the accuracy and completeness of electoral registers in Great Britain and Northern Ireland, inform our guidance for Electoral Registration Officers, and inform our approach to public awareness campaigns.

The presentation covered the latest report on the quality of the 372 electoral registers in the UK. Care was taken to ensure that the sampled households represented geographically spread local authorities across England, Scotland and Wales and wards across Northern Ireland. This was the first time where the sample size in Wales and Scotland allowed us to make robust estimates for those devolved areas. While our response rate of 72% was quite high, the Board understood there was a risk of systematic bias from which households were likely to respond, which was difficult to address.

The survey found that around 17% of eligible voters in Great Britain were not correctly registered at their current address. In Northern Ireland, 25% of eligible voters were not correctly registered. In addition, more than 10% of entries on the registers in Great Britain and 20% in Northern Ireland were inaccurate. The Head of Research acknowledged that the study could not accurately measure voters who were registered in more than one constituency.

The Board heard that young people, recent home movers, and private renters were much less likely to be correctly registered. People who had lived at their address for under two years were significantly less likely to be correctly registered than those who had lived at their address for longer. Private renters were much less likely to have up-to-date register entries than those who own their homes outright. These findings were similar to results from previous surveys.

The Board discussed the registration rates of voters with different characteristics. The study found lower rates of registration among voters aged under 35 as compared to voters aged over 35 in both Great Britain and Northern Ireland. There was concern that the data indicated that lower rates of registration among young people in Northern Ireland was resulting in lower rates of registration in older cohorts; voters who were not accurately registered when younger appeared less likely to be correctly registered as they aged.

The Board considered the limitations of existing registration campaigns, and considered new methods to improve accuracy and completeness. The Board noted the proposed reforms to the annual canvass that were already underway.

The Board considered options that could improve the accuracy and completeness of the register. Some options included allowing Electoral Registration Officers greater access to existing national data which could support more automated or automatic registration. There was also discussion of integrating electoral registration into other public service transactions, for example applying for or renewing driving licences, to make it easier for voters to keep their registration details accurate. A move towards more automated forms of registration for specific groups, for example registering young voters when they were allocated their National Insurance number ahead of their sixteenth birthday, could also help to improve levels of registration.

The Head of Research noted the difficulties of reliably comparing accuracy and completeness figures over time and across different studies. He explained that other factors had to be taken into account, such as the number of European Union nationals living in the UK at the time of comparison. There had also been changes to the franchise over time which would affect the ability to compare figures. The Board asked the Head of Research to consider providing a comparison with other countries.

The Board discussed how best to present the findings and the recommendations of the study. The Board considered it appropriate to provide granular information on the characteristics of voters who were less likely to be registered. The Board also suggested recommending changes to registration to reduce the number of duplications in the system.

The meeting ended at 2.35pm.