Response to the report on electoral registration by the Levelling Up, Housing and Communities Committee

The Electoral Commission should undertake further research to identify effective ways to make registration easier and more accessible for people with disabilities. (Paragraph 33) 

There should be no barriers to registering to vote and voting for disabled people and we carry out a range of research to ensure we identify those that do exist. After each election, we gather evidence on people’s experiences of registering to vote and voting. In addition to our large-scale public opinion survey, we invite disability charities and the people they support to provide feedback on the election.  

Our assessment of the accuracy and completeness of the 2022 electoral registers found that people with a physical disability were more likely to be registered correctly than the population as a whole (92% correctly registered compared to 86% overall). Although people with a mental health condition had similar levels of registration to the average (84% compared to 86%).  

Similarly, our annual public opinion tracker, which asks about attitudes towards registration, found no difference in people’s confidence that they know how to register depending on whether they have a disability or not.

However, our survey found a difference in satisfaction with the process of registering. Of the people who said they are ‘limited a lot’ by a disability, 75% were satisfied with the process of registering – that compares to 80% of people without a disability. This suggests that aspects of the system are more of a challenge to some disabled people.  

A key difficulty with research into a diverse group of people is that it can mask the much greater issues faced by some, including those with more severe conditions or with other factors, for example young disabled people with no prior experience of registering or voting, as well as those with poorer access to either online resources or someone to help them.  

In order to address this, we are working with trusted organisations that support disabled people in order to explore these issues in more detail and to identify how we can help to remove barriers through our communications and education work.  

This work has included a project in partnership with Mencap, which has highlighted the importance of fully accessible registration forms including the provision of easy read versions as well as more targeted support, for example using Makaton. We have also worked with the Centre for Mental Health on research which suggested that people with a mental health condition shared some similar concerns about registration to those we find among the general public. This included the difficulty in checking if you are registered and the perception that the process takes too long.  

This initial research has been helpful in identifying barriers, but we need to do further work. We are currently developing our corporate plan for the next Parliament, which will include qualitative and quantitative work on the barriers experienced by key demographics. This includes disabled people as well as other under registered groups such as younger people and some Black and minority ethnic communities.

Alongside this research we will continue to work with partners to implement their recommendations wherever possible and to create and disseminate information on voter registration and other aspects of our electoral system. This includes information on registration contained alongside information on the voter ID requirement which we have published in Braille, synthetic audio and large print (supported by the Royal National Institute for Blind People), Easy Read (supported by Mencap) and BSL (supported by the British Deaf Association). 

The Electoral Commission should, with local authorities and academics, undertake more research on effective ways to encourage people who do not vote, and make recommendations to central and local government as to how people can be encouraged to take part in the democratic process. (Paragraph 39)

From our existing research, we know some of the main reasons that people choose not to vote. In our public opinion surveys after elections we ask people whether they voted or not, and why. We consistently find a large proportion of non-voters who say that circumstances prevented them from voting – for example, after the 2019 UK parliamentary general election 16% of non-voters said they were busy or away at the time of the election. However, it is also clear that perceptions about the effectiveness of their vote is a factor (7% of non-voters said they had not voted because it would make no difference to the result). Finally, negative views of politics and politicians are clearly also important – in 2019 we found 14% of non-voters said they had not voted because they did not like any of the parties/candidates or because they could not trust any of them.

We are also working increasingly closely with local authorities to capitalise on their understanding of the barriers people face and to identify how we can support authorities to engage them. This collaboration includes working groups of local authority communications staff in different parts of the UK which, for example, has helped inform our approach to supporting groups whose first language is not English or Welsh to meet the voter ID requirement. In our next corporate plan period, we want to work with electoral administrators to make the whole system more inclusive and participative, for example through identifying and sharing good practice and learning.  

However, we agree that further work is needed. Existing research suggests voting is habitual, at least in part, and that people are more likely to vote if they have started to do so as soon as they are eligible. New research later this year will deepen our understanding of the issues faced by younger people when participating in politics. We expect the findings to be particularly valuable in informing the work we already do to provide resources for young people and educators on voting and democracy.  

We also know that low socio-economic status is a common denominator for many of our under-registered groups, including some Pakistani, Bangladeshi and African communities, social housing renters, and those without a permanent address. We plan to commission a set of participatory action research projects targeted at the C2DE socio-economic group. This work will include projects run at a national scale to maximise reach, and smaller community-based projects. We will engage a wide range of local and national civil society and voluntary organisations to deliver this work.

As in other areas, our next corporate plan also provides an opportunity to identify new pieces of work and we expect further work on encouraging registration and participation to be included in that.

The Electoral Commission should write to us in four weeks with a detailed plan including milestones and dates for moving towards implementing automated voter registration with a clear outline of how people can protect their data privacy. (Paragraph 66) 

We welcome the Committee’s focus on the importance of detailed planning for moving towards automated voter registration. In 2019 we published the findings of feasibility studies on what would be needed to deliver these changes. More recently our report on the accuracy and completeness of the 2022 electoral registers sets out case studies of how new approaches to registration could work in practice using different data sets. It also includes information about how other countries and territories have implemented similar reforms.

Overall responsibility for developing a comprehensive implementation plan to deliver reform must ultimately rest with the UK Government, given its responsibility for policy, legislation and funding for electoral registration. We have, however, set out below some of the key challenges in delivering some form of automated registration – whether assisted, integrated or automatic, or some combination of these – and the steps and stages that would need to be completed. We have also developed an indicative implementation timeline, which forms the annex to this paper.  

Cross-government and cross-sector working, including with civil society groups, would be needed to ensure successful delivery of registration reform. In particular, strong leadership and a high level of political support would be required to ensure senior political agreement across relevant parts of government – including key agencies – and a shared vision to underpin delivery. This would include the availability and use of data sources and infrastructures from across the public sector.

In addition, early engagement with other partners involved in operational delivery would be crucial – for example, technical partners, such as Electoral Management Software (EMS) suppliers and the Individual Electoral Registration Digital Service (IERDS); Electoral Registration Officers (EROs); and the devolved governments. This would be an important step in developing a shared understanding of roles and responsibilities, governance arrangements and delivery expectations.

A detailed implementation plan would need to factor in the time needed to develop policy options for delivering automated registration. Policy development would be led by UK Government officials, followed by subsequent stages including consultation, piloting and testing, and the legislative process.  

We note that the Senedd is currently considering the Elections and Elected Bodies (Wales) Bill, which would allow the Welsh Government to introduce pilot schemes focusing on automatic registration. If this legislation is approved, we will support the Welsh Government and local authorities to design and evaluate pilot schemes that will provide robust evidence about the potential impact of this policy approach, and potential lessons for automated registration in other parts of the UK.

There are also some broader points relating to implementation. The Committee suggests that some reforms (for example, signposting registration) could be delivered without delay, principally because they would not require legal changes to implement, nor would they be dependent on significant technical or operational changes. We agree with the Committee on this point. Other reforms, such as assisted, integrated or automatic registration are more ambitious in scope. For these, it would be better to consider how they could be taken forward as part of a wider modernisation strategy, rather than as separate initiatives.

A holistic approach would include consideration of:

• How any changes will impact on eligible voters – for example, in terms of providing additional opportunities to register to vote and on under-registered groups.
• How the different reforms could complement one another in boosting accuracy and completeness, and addressing specific electoral registration challenges.
• How any changes will impact on EROs and electoral administrators and how they operate, and how any disruption and additional pressure can be minimised.
• The relative impact of each proposed reform on existing registration processes (for example, the annual canvass) and the system more widely.
• The potential for the reforms to deliver efficiencies, streamline processes and enhance resilience in the sector.
• How the necessary technical infrastructures might be aligned and integrated.
• The phasing of implementation.

The Committee has further asked the Commission to outline how people can protect their data privacy. Automated voter registration would involve the transmission of personal data from data source organisations to EROs. This raises questions around the potential loss of privacy and undermining of public trust, which would need to be addressed. It would be important for government departments and public bodies to consult with data protection professionals to advise and assist on key aspects of data protection and to ensure that data may be shared safely, responsibly and legally. More broadly, consideration of these issues should be an integral part of any government-led registration reform programme.

GDPR-compliant data sharing agreements between data source organisations and EROs would need to be in place before any transfer of personal data could take place. Data transfers would also need to comply with data protection legislation and Information Commissioner’s Office guidance.

In addition, systems of integrated registration would be based on the active and informed consent of the individual voter. For example, at the end of another public service transaction, the individual would be asked whether they also want to register to vote via a check box. If the individual proactively confirms that they do, relevant data would be transferred to an electoral registration application, with the individual providing any missing data, such as National Insurance number and nationality.

The sharing of personal data between data source organisations and EROs to support models such as assisted or automatic registration would need to be based on a clear legal gateway permitting the sharing of such information for electoral registration purposes.  

Automatic registration in particular would require additional safeguards – for example, including the option for people to apply for anonymous registration if their safety is at risk, or the ability to ‘opt out‘ of automatic registration – for example, if they are not eligible to register. People would also need to be given the option to opt out of the open register before being added to the electoral register automatically, as they can do already through the register to vote service.

In the meantime, the Electoral Commission should research the possibility of moving to single unique identifiers for citizens registered to vote as a first step towards a single national register for England and to address the twin problems of duplicate registration and registration being lost when moving house. There should be a consideration of whether National Insurance numbers could be used for this purpose. (Paragraph 67) 

As the Committee recognises, unique identifiers for each registered voter could enable EROs to identify and manage potential duplicate registration applications more easily. They could also help EROs to identify and remove out of date register entries where voters have moved home and re-registered at their new address. In the context of the response above in relation to data sharing, using a single unique identifier could also significantly reduce the volume of information that would need to be transferred between EROs and compared against registers.

Unique identifiers could be created at a number of different points in the electoral registration process. In Northern Ireland, a unique number - known as a Digital Registration Number (DRN) - is issued to people when they first register to vote successfully online. The DRN is needed if a voter then decides to apply for a postal or proxy vote.  

Our report on the 2023 Northern Ireland local government elections found that 5,118 (73%) absent vote applications were rejected because there was no DRN. This was a significant increase from the 2022 Assembly election in which 3,636 applications were rejected because there was no DRN, which equated to 60% of all rejected applications. The Electoral Office for Northern Ireland received 8,700 requests for a DRN in the run up to the election, with 53% of those requests made in the last week before the absent vote deadline on 26 April 2023. This put a considerable administrative burden on the Electoral Office and created a significant challenge for staff to respond to requests for DRN in a timely manner.

In Great Britain, EROs themselves could create and allocate unique identifiers to each of the entries on their registers. This would not guarantee that identifiers are unique across different registers, however. Alternatively, the IERDS could allocate identifiers when it is used to process applications to register to vote. This would reduce the risk that the same identifiers are used for entries on different registers, or that the same individual could be issued with two (or more) different ‘unique’ identifiers, defeating their overall purpose.  

Any system involving the issuing of unique identifiers would also need to be compatible with specific circumstances that could arise in relation to a person’s registration status – for example, where a person is lawfully registered in more than one place yet must be issued with a unique identifier that would be consistent across both registers.

Using National Insurance numbers (NINo) would be more likely to provide unique identifiers across all registers (although not all eligible voters will have been issued with a NINo, and it is possible that some people have more than one NINo). The Department for Work and Pensions would need to consent to NINos being used in this way. The identifier would only be used by EROs to manage the register, and it would not need to be made available publicly to those entitled to access the register.

It is clear from the rejection rates and low awareness levels that the DRN is acting as a barrier to voters in Northern Ireland. We have already called on the UK Government to review the DRN to ensure that the barriers it has created are removed while also maintaining the integrity of the absent voting process.    

We will continue to keep under review the potential of single unique identifiers, including drawing on the evidence of the use of the DRN in Northern Ireland, alongside whether connectivity between registers would offer any additional benefits for electoral registration or electoral reform more broadly.

Our current assessment indicates that neither single unique identifiers nor a single national electoral register for England (or for the UK as a whole) would be essential to support the introduction of automated electoral registration. The feasibility studies that we published in 2019 and updated as part of our accuracy and completeness research last year show how the existing technical infrastructure could be adapted and enhanced to support automated forms of registration in future.

The Electoral Commission should ensure that it carries out a proper assessment of the impact of the voter ID requirement before the next general election, including a review of data collected locally. (Paragraph 91) 

Through our post-poll evaluation and reporting work, we  assess changes to elections and how they are being implemented, and inform the wider debate and make recommendations for change. Throughout the development and implementation of the voter ID policy, we consistently recommended that the impact of this new requirement on voters and electoral administrators should be robustly assessed and evaluated. Drawing on our existing research and reporting duties, we are committed to ensuring that we provide an independent, evidence-based analysis of how voter ID is being implemented in practice.  

Following the May 2023 polls in England, the first for which voter ID was required, we published an interim analysis of the new ID requirement in June 2023, to ensure that our independent analysis of the implementation of the new requirement was available as soon as practicably possible after polling day.  

Our full report on the May 2023 elections drew on a wider range of information, including feedback from candidates, Returning Officers, polling station staff, election observers, charities and civil society organisations, and the police. We also conducted additional analysis of socio-demographic data in 18 local authorities to understand whether some people were more likely to struggle to meet the ID requirement. This analysis confirmed our interim findings that some people found it harder than others to show accepted voter ID, including disabled people and unemployed people. We found that this stemmed from two overlapping issues – the variations in ownership of accepted photo ID and in awareness of the need to show ID when voting in person.

For the scheduled May 2024 polls and the next UK parliamentary general election, we developed a comprehensive package of communications, research and reporting activities, and assessing the impact of voter ID comprises a key part of this. We are currently carrying out further analysis and assessment of data at the May 2024 elections to understand the impact of the voter ID requirement for voters and electoral administrators. This includes:

• Analysing data that Returning Officers can collect from every polling station about the number of voters turned away because they were not able to show accepted ID, and the numbers who returned with ID and were able to vote.
• Research with the public using a large sample post-election survey to estimate the number and proportion of eligible voters who did not attempt to vote at a polling station because of the ID requirement, and to understand awareness of and attitudes to the requirement.  
• Seeking evidence from candidates, Returning Officers, polling station staff, election observers, the police, charities and civil society organisations, about the practical implementation and impact of the voter ID requirement.
• We will publish our analysis of data about the impact of the voter ID requirement at the May 2024 elections in early July and use these findings to inform planning for the UK parliamentary general election. We will also carry out and publish an equivalent analysis of data about the impact of the voter ID requirement at the general election and share our findings with the Committee.

Central and local government and the Electoral Commission need to work together to raise awareness of the need for Voter Authority Certificates prior to the next general election, to ensure that those who do not have ID are not shut out of the electoral process. (Paragraph 94)

The Commission has a statutory role to raise public awareness of the democratic process, and conducts a range of work to support this aim. We have run specific voter ID awareness campaigns in 2023 and 2024. These campaigns use a mixture of paid-for advertising, partnership work across local government and civil society, public relations and our own social channels and website.  

The campaign’s primary focus is to raise awareness of the overall policy change for the whole electorate, but it also highlights the existence of free ID for those without approved ID. The prominence of this message varies depending on the channel of communication and the target audience – material designed for audiences with a greater likelihood of not having approved ID includes a greater level or prominence of information about the free ID. This work seeks to balance the need to reach the estimated 4% of the electorate without approved ID, with the need to not drive unnecessary applications for the free ID from those who do not need it. A significant surge in applications for a Voter Authority Certificate (VAC) could risk causing capacity issues for electoral administrators during the peak pre-election period, and displacing other important activity.

In 2023, pre-campaign research showed awareness of the voter ID policy was at 22% of the electorate in England. Post-poll awareness had reached 91%. Awareness of the VAC specifically was 57%. This informed our May 2024 campaign, including engagement with partners, such as local authorities, to make sure they had all the resources required to raise awareness of the VAC with voters who might face additional barriers to having ID.  

We will run this campaign again ahead of the UK parliamentary general election. This will include updated voter information resources that local authorities, EROs and civil society partners can use. Each of these will reference the availability of free voter ID, and several – including social media graphics and posters – will explicitly promote the VAC application deadline. These will be available in a range of accessible formats and community languages.