Research and analysis

Electoral Integrity Programme evaluation: Year 1 - executive summary

Published 30 November 2023

Introduction

Policy background

The Elections Act 2022 contained measures designed to strengthen the integrity of the electoral system and ensure that elections remain secure, fair, modern and transparent. These measures are being delivered through the Electoral Integrity Programme (EIP) as part of the UK Government’s electoral reform agenda. The measures are likely to be implemented by the end of the current Parliament with two having already been introduced for the May 2023 local elections. These were to:

• introduce photographic identification requirements for voting in polling stations; and
• require Returning Officers to consider the needs of disabled electors and make specialist equipment available to support them to vote if needed and reasonable to do so; and allow anyone over the age of 18 to accompany disabled electors in the polling station.

To evaluate the implementation and impact of these measures, in May 2023, the Department for Levelling Up, Housing and Communities (DLUHC) appointed IFF Research to synthesise the findings from a wide range of research, data collection and analyses that have been conducted as part of DLUHC’s monitoring and evaluation activities for the EIP as well as conducting new qualitative research.

The evaluation approach

The evaluation included both process and impact components. The process elements gathered evidence on how electoral registration officers, returning officers, electoral services teams and polling station staff had implemented the measures. The impact evaluation elements assessed the effectiveness of the measures and their impact on both electoral staff and electors.

The evaluation adopted a theory-based approach using theory of change models to set out how each of the voter identification and accessibility measures might affect change in the short, medium and long term. IFF Research then used contribution analysis developing a series of contribution claims to articulate how each measure within the theory of change models could lead to change, while recognising the importance of other influencing factors. The extent to which claims had been met was then assessed using evidence from a broad range of data sources including implementation surveys and monitoring data gathered by DLUHC, public opinion surveys conducted by Ipsos, research published by the Electoral Commission and new primary qualitative research conducted by IFF Research among electoral services teams, polling station staff and selected groups of electors. In the compressed timeframe, not all claims could be assessed conclusively and the report highlights where further investigation is needed to reach a more definitive view.

Key findings: voter identification

Eight contribution claims were developed from the voter identification theory of change to examine how the measures were implemented for the May 2023 English local elections. They covered the preparedness and capabilities of electoral services teams within local authorities and polling station staff to deliver the measures, awareness and response among voters, and impact on personation.

The evaluation found that electoral services teams and polling station staff were well prepared for the implementation of the measures for the May 2023 elections and checks on voter identification were conducted efficiently with very few voters initially turned away because they did not have any, or any accepted, photographic identification.

Communications designed to raise awareness of the voter identification requirement by both the Electoral Commission and local authorities were generally effective, but for future elections, consideration could be given to targeting audiences found to be less aware, such as those who rarely vote, members of ethnic minorities and young people. Local authorities had sufficient resources to process Voter Authority Certificates (VACs) for those without accepted photographic identification, but more could be done to raise awareness among those who do not have photo identification. For those wishing to have their identification checked privately, polling station staff felt they were able to make sufficient provision using privacy screens or private rooms. However, given the very small number of voters taking up this option, the evaluation was not able to conclude on the voter experience of this measure.

The Electoral Commission reports that allegations of personation reported to the police from the May 2023 English local elections were very low (two) and consistent with previous elections. It is not possible to conclude in this evaluation whether any of the small number of voters turned away on polling day were attempting to use fraudulent identification or if the measure has had a deterrent effect on those who might attempt personation.

Overall, the introduction of the photographic identification requirement did not impact on the likelihood or experience of voting in the May 2023 English local elections among voting age adults with accepted photographic identification and trust in the in-person voting system remains high. However, those without accepted photographic identification were much more likely to say that the voter identification requirement made them less likely to vote.

Claim 1: Local authority staff and the electoral services teams meet the requirements of law

Electoral administrators made effective use of the Electoral Commission’s guidance to deliver appropriate levels of staffing and training to process Voter Authority Certificate (VAC) applications and conduct the required identification document checks in polling stations. Nearly four in five (79%) Returning Officers deployed more Poll Clerks and over half (56%) more Polling Station Inspectors, whilst most managed VAC applications using either electoral services teams or by transferring staff internally from other duties as required. Only 3% of electoral administrators felt they had inadequate resourcing to process VAC applications and only 4% felt they had too few Poll Clerks (May Reflections Survey of Electoral Administrators). Some electoral administrators had been concerned about the impact of the new measures on the recruitment and retention of polling station staff. Although around 10% of polling station staff dropped out between recruitment and polling day, dropout was mainly attributable to personal circumstances, including holidays around the Bank Holiday, with only one fifth (18%) of those who dropped out citing the new measures.

Reflecting on the May 2023 experience, around a quarter (26%) of local authorities surveyed felt they would make little to no changes to their preparations for any future elections (May Reflections Survey of Electoral Administrators). Although the claim was strongly evidenced for the May 2023 election, some electoral administrators anticipated greater resource needs to process a higher number of VAC applications and recruit more polling station staff for a UK Parliamentary General Election and had concerns about sufficient lead in time to do so.

In total, the review of the evaluation evidence found this claim has been met.

Claim 2: Returning Officers and polling station staff meet the requirements of law

Nearly all (95%) polling station staff surveyed recalled being trained in checking voter identification (Electoral Commission’s Polling Staff Survey). In the qualitative research, polling station staff said that they found guidance from the Electoral Commission, as well as information provided in polling stations to remind voters of accepted identification, useful to refer to on polling day. Although the majority of polling station staff surveyed felt that the process of voting took longer than previous elections (62% said a little longer, 3% said a lot longer), the difference was not substantial and nearly all (99%) staff felt the polling day still ran smoothly (Electoral Commission’s Polling Staff Survey).

The qualitative interviews suggest that Returning Officers and polling station staff did not deviate from the requirements and instructions they received in training, even when faced with challenging situations in the preparation for and during polling day. This suggests a high level of confidence amongst polling station staff, and good understanding of their role. That said, polling station staff were less confident about explaining why some types of photographic identification are accepted and others are not and how to identify voters if they had changed significantly in appearance since their photographic identification was issued. Further training in these areas would be valued and could improve the voter experience.

Overall, the review of the evaluation evidence found this claim has been met.

Claim 3: Communication campaigns raise electors’ awareness of voter identification

Nearly all (99%) electoral services teams, often drawing on material provided by the Electoral Commission, issued communications to the electorate about the photographic voter identification requirements, whether directly to households or via on street billboards, social media and local radio. By May 2023, 84% of voting age adults living in areas holding local elections were aware of, that is had heard a great deal or fair amount about, the voter identification measures (Public Opinion Survey Wave 2), partly as a result of these communications but also due to national and local news and other media coverage. But awareness was not consistent across the electorate. Those most likely to have heard nothing at all about the requirement were people who rarely or never vote at local elections (18%), ethnic minorities (12%) and younger adults (11% of those aged 18-34) (Public Opinion Survey Wave 2). The evaluation did not find examples of local authorities delivering targeted communications campaigns to reach these specific audiences on voter identification measures. Some electoral services staff attributed this to the limited timeframe to prepare for the elections.

The proportion of voters who were initially turned away – usually because they had not brought an accepted photographic identification document – was small (0.7%) (Voter Identification Monitoring Data). However, this cannot be entirely attributed to the communications campaigns given many voters were carrying accepted photographic identification documents as a matter of course. Future campaigns will be able to take the learning from the Public Opinion Surveys together with feedback from local advocacy organisations given subsequent to the May 2023 local elections to develop more targeted communications to raise awareness about the types of photographic identification that are accepted.

In conclusion, the evidence reviewed indicates that this claim was partially met.

Claim 4: Communication campaigns raises electors’ awareness of Voter Authority Certificates (VAC)

In May 2023, 96% of voting age adults in England owned an accepted form of photographic identification, 2% did not, 1% preferred not to say and 1% did not know (Public Opinion Survey Wave 2). Those who did not have an accepted form of identification could apply for a free VAC to present as accepted photographic identification in order to vote at a polling station in the local elections. In response to communications from both the Electoral Commission and local electoral services teams, awareness of VACs did increase from 9% of voting age adults in England in January 2023 to 21% in May 2023, suggesting that the campaign raised awareness to some extent (Public Opinion Survey Wave 2). But awareness was not much higher among voting age adults with no accepted photographic identification of whom only a quarter (26%) were aware of the VAC by May 2023 (Public Opinion Survey Wave 2; note that this figure includes Wave 1 re-contacted sample in addition to Wave 2 ‘fresh’ sample, and low incidence and resulting sample size for voting age adults with no accepted photographic identification).

The evaluation did not find evidence of VAC awareness communications being targeted at those with no accepted photographic identification. Future evaluative work could monitor the design, reach and impact of VAC messaging for a UK Parliamentary General Election to assess effectiveness more thoroughly. Continuing to work to increase awareness among the general population will also be of benefit as those who need a VAC may hear about them via friends and family.

The evidence for this claim was therefore sufficiently consistent and strong to conclude that the claim has not yet been met.

Claim 5: Electors who want a VAC, and are eligible, are able to obtain one

Electoral administrators reported that VAC application and issuing processes went smoothly. They felt all the necessary resource and training was available to deliver against this requirement at the scale required for the local elections.

A total of 70,017 applications were submitted in the run up to the May 2023 local elections in local authorities that held elections (usually using the online service; 95% of applications) with half (50%) submitted in the month prior to the VAC application deadline on April 25th (Voter Identification Monitoring Data). Overall, 92% of VAC applications (equating to 64,240) resulted in a VAC being issued before the closure of the poll. Although some initial teething issues were flagged as challenges, the online Electoral Registration Officer Portal (EROP) was generally viewed as an enabler and effective tool in processing VAC applications. While providing suggestions for improvements to the Portal, electoral services staff were confident that it will be effective for a UK Parliamentary General Election.

Electoral administrators felt they had sufficient staff to process VAC applications for May 2023, but largely attributed this to the low numbers of VAC applications, and expressed concerns about the potential increase of applications for a UK Parliamentary General Election. Although awareness of DLUHC support for VAC processing was fairly high, nearly one in five electoral administrators (17%) remained unaware of it by May 2023 (May Reflections Survey of Electoral Administrators). Electoral services teams would benefit from a reiteration of this message from DLUHC in preparation for future elections.

The most common issue with applications was non-compliant photographs (33% of rejected applications had a non-accepted photograph) (Voter Identification Monitoring Data), and time spent chasing electors who submitted applications with issues was sometimes resource intensive. Steps to improve electors’ understanding of the type of photograph required could reduce this need.

Raising awareness of the availability of in-person applications could challenge some negative elector perceptions about the accessibility of the VAC application process for electors who may not have access to a computer or struggle with an online application.

The evaluation evidence is consistent and strong in supporting a conclusion that this claim has been met. However, it is important to note that there was uncertainty on whether electoral services teams could manage the greater demands of a UK Parliamentary General Election, especially if many VAC applications were received close to polling day.  

Claim 6: Electors who want to have their identification checked in private, are able to do so

In preparation for the May 2023 local elections, nearly all (95%) of polling station staff recalled receiving training on how to help voters show their photographic identification in private (Electoral Commission’s Polling Staff Survey). All electoral services teams and polling station staff felt they were able to provide appropriate privacy arrangements, whether that was with privacy screens or alternatives using old polling booths, screens or separate rooms. The discretionary nature worked well to allow electoral staff to be proportionate to likely demand in their local area.

In total, 2,250 voters had their identity checked in private across all local authorities holding English local elections in May 2023 (Voter Identification Monitoring Data). Around half (51%) of local authorities did not receive any requests for a private identification document check on polling day, whilst 46% received between one and 50 requests, and seven (3% of) local authorities received more than 50 requests. Given the very small numbers of electors using private identification checks it was not possible to gather evidence on their experience of the process or any perceived barriers to requesting private checks.

Although the evidence indicates that provision was made effectively, without more detailed evidence from voters who used, or might have wanted to use, this provision, it is difficult to firmly conclude that this claim has been met.

Claim 7: Personation is identified more easily and reduces in the longer-term

In preparation for the local elections, nearly all polling station staff surveyed (95%) recalled receiving training on how to review photographic identification and generally felt well prepared to carry out their duties on polling day (Electoral Commission’s Polling Staff Survey). The Electoral Commission’s Handbook for polling station staff (PDF, 5.5 MB), (particularly Annex 12) consolidated this training effectively and provided guidance for polling station staff on the day. On the day of the May local elections, around 14,000 (or 0.25%) people who tried to vote at a polling station were turned away and did not return later in the day with accepted photographic identification (Voter Identification Monitoring Data). Similarly, across the UK parliamentary by-elections held in July and October, around 350 (0.25%) people who tried to vote at a polling station were turned away and did not return (Voter Identification Monitoring Data).

The Electoral Commission reports that allegations of personation reported to the police from the May 2023 local elections were very low (two) and consistent with previous elections (Electoral Commission Report on the May 2023 local elections in England). Similarly, confidence in the protections from fraud for in person voting remain consistently high among polling station staff, electoral administrators and electors alike. When it comes to getting away with electoral fraud in polling stations, half (51%) of voting age adults think it is difficult, a small increase of 3 percentage points since January. Amongst voting age adults in areas that held English local elections in May 2023, there was evidence of higher public confidence that it would be difficult to get away with fraud in a polling station (52%) than those areas that did not hold elections (48%) (Public Opinion Survey Wave 2). In the qualitative research, some polling station staff did, however, raise concerns about their ability to detect a counterfeit identification document, but generally all groups held the perception that fraud was more likely to be conducted via postal or proxy voting.

It is not possible from the evidence available to the evaluation to assess the impact of the voter identification measures as a deterrent to those who might have otherwise attempted personation at a polling station. But, contextually, there was no associated fall in in person voting (with turnout, at 32.0%, comparable to the last similar set of local elections in 2019, when it was 32.5%) or shift to postal or proxy voting, with the proportion of voters opting for a postal vote falling slightly from 19.9% to 19.0% and proxy votes remaining constant at 0.1% between 2021 (the first local elections post-pandemic) and 2023 (Electoral Commission Report on the May 2023 local elections in England).

In total, the evaluation evidence reviewed is inconclusive in determining whether the claim has been met.

Claim 8: Electors remain satisfied with the process and ease of in-person voting

In terms of voter experience, the evidence reviewed consistently suggests that the introduction of the photographic identification requirements did not impact on the local election experience of voting age adults with accepted photographic identification. However, of the 2% of voting age adults in England without accepted identification, over four in ten (41%) of those in areas conducting local elections in May 2023 reported that the photographic identification requirements made them less likely to vote compared to fewer than one in ten (8%) of all voting age adults in areas holding English local elections in May 2023 (Public Opinion Survey Wave 2).

Evaluation of voter satisfaction concluded that, although satisfaction with the voting process fell slightly among voting age adults in England from 74% in January 2023 to 69% in May 2023 (Public Opinion Survey Wave 2), many other factors could impact voter satisfaction and so the change cannot be attributed to the voter identification measures.

Trust in the voting system remains high with 82% of voting age adults in England in May 2023 feeling confident in the way elections are run (Public Opinion Survey Wave 2) and findings from the qualitative research indicate that electors did not perceive election fraud to be an issue prior to the introduction of the measures. Shifts in trust may emerge over the longer term but it is too early at this stage to assess whether this element of the claim has been met.

Given the weakness and inconsistencies in some of the evidence for elements of this claim, it is therefore not possible to draw a conclusion on whether the claim has been met. For the claim to be met in future elections there would need to be more sustained and consistent intention to vote and satisfaction with the process across different types of voters.

Key findings: accessibility measures

Four claims were developed from DLUHC’s original theory of change to examine how the accessibility measures were implemented for the May 2023 local elections. These covered the preparedness and capabilities of electoral services teams within local authorities and polling station staff to deliver the measures, plus awareness and experience of disabled voters.

Electoral services teams and polling station staff were well prepared to implement the new accessibility measures including providing equipment to support in-person voting for electors with disabilities and delivered provision effectively on the day. Training and provision could potentially be improved through more engagement with disability advocacy groups in preparation for future elections. However, there was low awareness of the measures among disabled voters in areas conducting elections in May 2023 and some local authorities did have difficulties engaging with disability advocacy groups in the time available before polling day. But when told about the measures, more than three-quarters of voting age adults in England with a disability anticipate that the accessibility measures will make it easier for them to vote in person.

Claim 1: Local authority staff in electoral services teams with Returning Officers meet the requirements of law

By March 2023, nearly all electoral administrators in England holding local elections in May 2023 (99%) felt they were prepared to implement the accessibility measures of providing equipment to support in-person voting for electors with disabilities and allowing a wider range of people to act in the role of companion to assist disabled people to vote in polling stations (March Readiness Survey of Electoral Administrators). Confidence was similarly high in Northern Ireland. Most (80%) electoral administrators had engaged with voters with disabilities by this stage also, either through direct communications or engagement with local advocacy groups. They were confident (98%) that they had sufficient staffing to meet the requirement (March Readiness Survey of Electoral Administrators).

Views on the levels of discretion in the guidance were inconsistent, with some local authorities’ electoral services teams interviewed appreciating the levels of discretion while others were looking for a firmer steer on minimum levels of provision. Most electoral services teams offered the provisions that the Electoral Commission’s guidance said they ‘should provide’ (May Reflections Survey of Electoral Administrators). In preparation for the UK Parliamentary General Election, some electoral administrators indicated that they would benefit from greater clarity on how to apply the guidance.

Overall, the evidence supports a conclusion that the claim has been met.

Claim 2: Returning Officers with polling station staff meet the requirements of law

Nearly all (94%) polling station staff recalled being trained in supporting disabled electors to vote and polling station staff’s confidence in their ability to support electors with disabilities increased and was very high by May 2023, ranging from 88% confident they could assist a voter to use a tactile voting device to 97% confident they could assist a voter with mobility issues (Electoral Commission’s Polling Staff Survey). But some polling station staff interviewed in the qualitative research felt accessibility requirements were not given sufficient attention, with a greater focus on training for the voter identification checks. That said, polling station staff confidence in supporting electors with different disabilities to vote in-person has increased since 2022 (Electoral Commission’s Polling Staff Survey).

One in ten (10%) of polling station staff reported voters having difficulties with accessibility at polling stations on the day, but around two-fifths of these (43%) were difficulties accessing the building which is part of the Electoral Commission’s guidance on venue accessibility rather than the new legislation (Electoral Commission’s Polling Staff Survey).

In planning for a UK Parliamentary General Election, the evidence suggests that it would be beneficial to allow more time to engage with disability advocacy groups to help in delivering the training and identifying ways of improving support for disabled voters on the day. Evidence for this claim was limited by the very low incidence of disabled electors who wanted to vote in person and were either unable to or did not attempt to. Future evaluations should also engage with disability advocacy groups to engage these electors in the research.

The current evaluation evidence indicates that the claim that Returning Officers and polling station staff are delivering the requirements of the new accessibility measures is being met.

Claim 3: Communications effectively raise awareness of reforms amongst electors with disabilities and their carers

Although many local authorities attempted to engage disability advocacy and support groups to increase awareness on the new accessibility measures, by May 2023, only a minority of disabled adults of voting age in England knew a great deal or fair amount about the new accessibility measures (12% in England and 10% in Northern Ireland for companion changes, and 12% in England and 10% in Northern Ireland for equipment measures) (Public Opinion Survey Wave 2). Some local authorities (11% of those surveyed) did identify engaging with advocacy and support groups in the time available as a challenge in meeting accessibility measures more broadly, including their role in raising awareness, however 62% of local authorities surveyed stated there were little to no problems in this area (May Reflections Survey of Electoral Administrators). There was also limited evidence on the outcomes of that engagement. This evaluation did not collect evidence directly from advocacy groups and this could be considered for future investigation to better understand the barriers to engagement with local authorities on the disability measures.

The evaluation found little evidence of direct communications with disabled electors. If communications via advocacy and support groups are proving challenging, more direct or alternative approaches could be considered for a UK Parliamentary General Election. Furthermore, these efforts do not necessarily need to be targeted; raising awareness generally will increase the likelihood of awareness of these new measures reaching disabled electors (for example, by reaching friends or family of disabled electors). These approaches could be further supported with more communications guidance from the Electoral Commission.

Given the low awareness levels and difficulties with engagement, the evidence suggests that this claim has not been met.

Claim 4: Disabled electors are enabled to vote in-person as a result of the new measures

There is consistent evidence that attitudes towards in-person voting among disabled electors are largely positive and unchanged (Public Opinion Survey Wave 2 and qualitative interviews) with over eight out of ten (83%) voting age adults in England with a disability limiting their activity a lot found the process of voting in person at the May local elections easy. However, satisfaction with voting in general was significantly lower among voting age adults with a disability that limits their activities a lot (60%) compared to among all voting age adults in England (69%) (Public Opinion Survey Wave 2).

Some electoral services teams feel they are still in the early stages of implementing the accessibility measures. There was consistent evidence that electoral services teams intend to draw on engagement with local advocacy groups that occurred close to, or since, the May 2023 local elections to prepare for a UK Parliamentary General Election. (May Reflections Survey and qualitative research).

A strong majority of voting age adults in England with a disability and those with an adult with a disability in their household continue to feel the changes will make it easier for people with disabilities to vote in person at a polling station (77% for providing a wide range of equipment; 77% for allowing more people to act as a companion) (Public Opinion Survey Wave 2) and a similar sentiment was shared in qualitative interviews. Similarly in Northern Ireland, strong majorities of voting age adults feel that both policy changes will make it easier to vote in person for those with disabilities (78% for providing a wide range of equipment and 75% for allowing more people to act as a companion). The focus therefore needs to be on ensuring that those that would benefit from these changes are aware and able to make use of them.

In total, this evaluation cannot conclude on whether the claim that disabled electors are enabled to vote in-person as a result of the new accessibility measures has been met yet. Future evaluation of a UK General Parliamentary Election will need to assess whether engagement with local disability advocacy groups has been more effective and examine the in-person voting experience of disabled voters in more depth.