2. Impact on voters: experience

In line with this data, we found that most of those who voted in the pilots knew that they would need to show ID to be able to vote. Our public opinion research after the election found that 85% of people who voted at a polling station said they were aware beforehand that they would have to show ID.

However, some groups of polling station voters were less likely to say that they knew they would need to show ID. Awareness was lower among first time voters (73%) compared to people who had voted before (87%), and among 35-54 year olds (80%) compared to those aged 55+ (89%).

Awareness was consistent across the photo and mixed ID pilots and lower in the poll card pilots. However, as with the 2018 pilot scheme, we think this is not due to poor public awareness in the poll card pilots but because of respondents’ interpretation of the question we asked – some people may not see needing to take a poll card as the same as needing to ‘show some identification in order to vote’.

In addition to those in the photo ID pilots, nearly nine in ten voters in the mixed ID pilots also chose to use a type of photo ID – most commonly their photo driving licence.

Nearly six out of ten voters across both the photo and mixed ID pilots showed their photo driving licence, which was similar to the proportion in the three photo ID only and mixed model pilots in 2018.

The vast majority of voters in the poll card pilots chose to show their poll card. A small proportion of people showed their photo driving licence (5%) or passport instead (1%).

Roughly 10% of voters in the mixed ID pilots used non-photo ID. Of these, the majority of people used a combination of their poll card and a bank card.

A small number of people did apply for these locally issued IDs in the two photo pilots and in two of the mixed ID pilots:

• 70 people in Pendle (63 voters showed it in the polling station)
• 24 people in Woking (37 voters showed it in the polling station, which included some issued in the 2018 pilot)
• Two people in Broxtowe (one shown in the polling station)
• One person in Braintree (none shown in the polling station)

All of the pilots allowed voters to show their identification in private (for example, because they were registered to vote anonymously or if they normally cover their face for religious reasons). Electoral Commission staff who were observing across the pilot scheme areas saw that this arrangement was in place in almost all of the polling stations they visited.

The arrangements in place varied between polling stations with some planning to use a separate room while others had a screen, curtain or purposely designed polling booth that voters could go behind. Almost all of the polling stations that Commission staff observed had at least one female member of staff who would be able to view ID of female voters in private.

From our observations on polling day, we are only aware of this facility being used by a small number of electors in Derby and Pendle. We therefore cannot draw any firm conclusions about how this process would work in practice across Great Britain.

However, if the requirement to show ID to vote is introduced, it is important that the needs of these electors are considered alongside the administrative implications for Returning Officers of providing appropriate facilities and staffing arrangements to allow voters to show their ID in private in polling stations.

Year-on-year turnout comparisons are difficult owing to the local government electoral cycle. For example five of the pilots have not held local elections since 2015, when they were held in combination with a UK general election. This means we cannot usefully look at differences by pilot model.

For the pilots with comparable figures, turnout was down by 2-6 percentage points compared to the 2018 elections.

For Watford and Woking, both in the second year of piloting voter ID, turnout was lower than in their 2018 pilot and the previous elections in 2016. In two of the other pilots (Craven and Pendle) turnout was also down on 2018 and 2016. However, in Derby, while 2019 turnout was lower than in 2018 it was higher compared to 2016.

Overall, across England, local election turnout in 2019 was around two percentage points lower compared to 2018.

As in our evaluation of the 2018 pilots, it is not possible to draw a clear connection between the pilot scheme and any changes in turnout. Limited data is available and where it is available, the pattern is not consistent. We also know that turnout is volatile and dependent on a number of factors.

As well as assessing the impact on those that tried to vote on 2 May, we have also looked for evidence of whether a voter ID requirement could deter electors from voting.

We used our public opinion survey to ask non-voters why they did not vote. The main reasons given were in line with previous surveys (too busy - 30%, away on holiday - 9%, forgot - 6%). However, 1% of people who didn’t vote in the pilots said it was because they didn’t have the right ID and less than 1% said it was because they disagreed with the requirement.

The limitations of sample-based surveying mean that we do not have enough responses from specific groups of people or within specific pilot models to be able to report experiences or views across those groups. This is because the samples are designed to provide data that is representative of areas as a whole.

This means that for some demographic groups, that are small relative to the overall population, the number of people surveyed can be too few to analyse.

We also asked respondents whether the requirement to show ID made them more or less likely to vote. Most people in the pilot scheme said it made no difference or made them more likely to vote (90%). However, a notable minority said it made them less likely to vote (3%), that they didn’t have ID (1%) or that they didn’t know (6%). Non-voters on 2 May were more likely than voters to say that they would be negatively affected or that they were unsure.

Our evaluation has also considered whether the identification requirements might have disproportionately affected particular groups of people.

We asked 165 national and local organisations, including those representing people with protected characteristics under the Equality Act 2010, to provide us with evidence of how the pilots affected the specific groups they represent.

A total of 29 organisations responded, with 16 providing more detailed feedback.

None of the organisations that contacted us provided any direct evidence about people who were unable to vote in the pilot scheme because they couldn’t show ID. However, almost all of them gave us more general views and raised concerns about the possible impact of an ID requirement for the people they represent.

Building on our evaluation of the 2018 pilot scheme, these have increased our understanding about the specific challenges that some groups of people might face.

Key themes from the responses were:

• Charities representing people with learning disabilities, the BAME, LGBT+, gypsy and traveller communities and people without a fixed address raised general concerns that some of the people they represent are already less likely to register and vote, and they are also less likely to have ID

• Many of the responses highlighted existing difficulties their users face in accessing services requiring proof of identity, including barriers faced by people who don’t have easy access to the internet

• People without a fixed address would be less likely to have forms of non-photo ID such as utility bills

• A transgender or non-binary person’s ID may not reflect their gender expression or identity

• People with learning disabilities would need accessible information to ensure they knew about an ID requirement and were able to access ID

• The cost, inconvenience and time required to obtain ID may be prohibitive for some people

Returning Officers in the pilot schemes ran public information campaigns before the elections and we know these raised overall levels of awareness; people knew more about the need to show ID by May 2019, compared with earlier in the year. In January 2019, 17% of people said they had seen or heard something about the requirement, increasing to 45% in May 2019.

Our research did find variations in levels of awareness, however, particularly among those who did not vote. For example, 57% of voters told us they had heard something compared to 31% of non-voters. This pattern is then also reflected by demographic groups where one is more likely to have voted, for example:

• Those aged 55+ were more likely to have heard something – 54% compared to 31% of 18-34s

• White respondents were more likely to have heard something than BAME respondents – 46% compared to 27%

These findings tell us something about the potential for a negative impact at other polls with higher turnout. If there were to be a disproportionate impact on particular groups of voters this could also have a negative impact on public confidence; we know that problems at elections can affect voters’ and non-voters’ overall perceptions of the poll.

Of course, it is possible that groups reporting lower levels of awareness in relation to the 2 May elections, where they were less likely to vote, would have higher levels of awareness around an election in which they intended to vote, such as a UK general election. However, we cannot assume that would happen automatically. It is likely to depend, to some extent, on the nature of the public awareness activities used.

These findings on awareness therefore emphasise the importance of ensuring that public awareness activities around any national rollout are genuinely effective across all communities and groups, particularly more infrequent voters.

Year-on-year turnout comparisons are difficult owing to the local government electoral cycle. For example five of the pilots have not held local elections since 2015, when they were held in combination with a UK general election. This means we cannot usefully look at differences by pilot model.

When we ask about how easy people think it would be to show ID, if it was required at future polls, we see similar results. Most people (91%) say that they would find it easy to show ID if they had to; however:

• Voters were more likely to say it would be easy (95%), compared to 88% for non-voters

• Those aged below 34 were slightly less likely to say they would find it easy than their older counterparts (84% compared to 93% for 35-44s and 94% for those aged 55+)

• BAME respondents were less likely to say it would be easy compared to white electors (87% compared to 92%)