Impact on voters

Most people in the pilot scheme areas knew about the identification requirements. Our research found that nearly nine out of ten (86%) of people who voted at polling stations said they were aware beforehand that they had to show identification to vote at their polling station.

Some groups of people were less likely to say they knew about the identification requirements. Polling station voters in the C2DE social grade were less likely to say they knew beforehand (18% said they did not know about the requirement compared to 9% of ABC1s). In the population as a whole, some groups were less likely to say they had heard something about the pilot. This included people aged under 35, C2DEs, those who said they were generally less politically active, and those who said that they didn’t vote in the elections.

People in Swindon and Watford were also less likely to say they were aware that they needed to show identification. This may be because they didn’t consider the requirement to show a poll card as ‘identification’, although we cannot confirm this from the available data.

Returning Officers in the pilot scheme areas ran public awareness campaigns before the elections, and awareness about the identification requirements improved between January and May 2018. In January, 36% of people in the pilot scheme areas said they had seen or heard something about the requirement, and by May this had risen to 55%.

In Bromley, Gosport and Woking most voters were able to show either their passport or photo driving licence to vote at their polling station. Local travel passes were the third most frequently shown type of identification.

In Swindon and Watford, the vast majority of voters showed their poll card to vote at their polling station. A small proportion of people showed their photo driving licence, passport, debit or credit card instead of their poll card.

Very few people applied to use the alternative options provided for those who did not have the required identification. No one in Bromley and Gosport applied for a Certificate of Identity or Electoral Identity Letter. Feedback from the Returning Officers in both these areas suggests that most people who contacted them to ask about the alternative option found that they did actually have one of the acceptable forms of identification.

In Woking, 64 people applied for a Local Elector Card before polling day and 43 voters showed their card as identification at their polling station.

Voters in Swindon and Watford could show their poll card as part of the required identification, and they were able to apply for a replacement poll card before polling day. Returning Officers issued 66 replacements to voters in Swindon and 3 to voters in Watford. Voters in Swindon could also ask another elector at the same polling station to ‘attest’ their identity, and 107 voters used this option.

Although all of the pilot schemes allowed voters to show their identification in private (for example because they were registered to vote anonymously or if they normally covered their face for religious reasons), we are not aware that anyone used this facility. We therefore cannot draw any conclusions about how this process would work in practice.

Although nearly everyone in the pilot scheme areas who went to their polling station to vote was able to show the required identification, a small number did not have the right identification with them and were not given a ballot paper. Many of these people returned later with the right identification, but some did not.

There is little evidence that the voter identification requirements had a direct impact on turnout in the May 2018 pilot scheme areas. In three areas turnout at the local authority level was no more than 1 percentage point lower than the most recent comparable elections. Turnout in the other two areas was higher than, including in Swindon where turnout was 5 percentage points higher. Overall turnout across England in 2018 was the same, at 36%7 , as at the last comparable set of elections in 2014.

There were some larger changes in turnout at the individual ward level in the pilot scheme areas. These ranged from an increase of 12 percentage points in one ward in Swindon to a decrease of 8 percentage points in one ward in Bromley.

The data on turnout and participation cannot tell us much about people who may have decided not to go to their polling station at all because of the identification requirement. However, our research with people in the pilot scheme areas found that 2% of those who didn’t vote in May 2018 said it was because they didn’t have the right identification. Many more people said they didn’t vote because they were too busy (27%), didn’t know who to vote for (13%), were away on holiday (12%) or were not interested (10%).

Most people in the pilot areas (79%) said the requirement to show identification made no difference to whether or not they voted in May 2018, and overall 3% of people said it made them less likely to vote. However, our research did confirm that non-voters were more likely than voters to say that it had made them less likely to vote: 5% of non-voters said it made them less likely to vote, and 2% said that they wouldn’t have been able to vote because they didn’t have any identification.

Before the May 2018 pilot schemes, accessibility organisations raised concerns that some groups of people would find it harder to vote because they would not easily be able to access the right types of identification. This included concerns about disabled people and people from black and minority ethnic communities, who are also less likely to participate in elections in general.

Overall, we found no clear pattern of decreased turnout based on the different demographic profiles of specific wards in the May 2018 pilot scheme areas. The relatively small size of the May 2018 pilot schemes, the level of turnout and the limited demographic variation across wards of the pilot schemes means that it is difficult to systematically identify examples of a negative impact for particular groups of people. The limitations of sample-based surveying also mean that we did not get enough responses from specific groups of people to be able to report experiences or views across those groups.

In Watford we did find a correlation between the proportion of a ward’s population that is Asian/British Asian and the number of electors both initially turning up without identification and not returning.8 We need to be cautious in drawing conclusions from this analysis as it is based on a small number of data points (only 12 wards in one local authority area). However, this does show that it would be helpful to have more pilots with more diverse populations in order to explore this further.