Ipsos UK – written evidence (VID0010)

Submitted by Dr Glenn Gottfried, Associate Director in Public Affairs, Ipsos UK

House of Lords Constitution Committee inquiry into voter ID

Ipsos written submission for follow-up questions for the House of Lords Constitutional Committee review on requirement to present photo ID when voting

As to the 8% of voting age adults who said voter ID made them less likely to vote in the local elections, are you able to isolate the voter ID factor from other factors (such as political apathy or being likely to be away on polling day)? This is so we have a more accurate figure on the effect of voter ID on people not wanting to vote.

As discussed, it is not possible from this research to give a purely causal figure for the impact of voter ID on non-voting.  Probably the most relevant figure from the research is the 4% of non-voters in local election areas who said that voter ID was one of the reasons why they were not able to vote in the May elections (the 8% refers to all those in local election areas who said the requirement made them less likely to want to vote). 

This 4% (of non-voters) splits further into 2% who said they did not have photo ID, 1% who said they did not agree/felt uncomfortable with the requirement, 1% who said they could not find their photo ID, and a small number of mentions of other reasons. Overall, this 4% was higher among the following groups of non-voters: social renters (8%), the West Midlands (8%), those living in the most deprived quintile areas (7%), those with a disability that has an impact on them voting in person (8%), those without photo ID (36%), and those dissatisfied with the voting process (11%).

This 4% of non-voters is a small number (base size only of 73), so further analysis of their views should be taken as indicative only given the wide margin of error.  With that caveat in mind, looking at this group in more detail we find:

• They tend to be less likely to vote in elections (only 5% say they always vote in local elections vs 47% of all those in local election areas)
• They are less satisfied generally with the process of voting at elections (30% vs 70% of all those in local election areas) and less confident that the recent local elections were well run (53% vs 79%)
• Most of this group (57%) also gave other reasons for not voting, such as not wanting to vote for any of the parties/candidates running (16%), feeling there was no point in voting (16%), not trusting politicians/councillors (16%), not knowing which way to vote (11%), etc.
• Were more likely to feel that the voter ID requirement made it more difficult to vote in person in the local elections (76%, vs 12% overall) and made them less likely to want to vote (79%, vs 8% overall).
• Are less likely to have accepted forms of ID such as a passport (38%, vs 84% overall) or a driving licence (33%, vs 82% overall).  Overall, 37% of this group said they had none of the accepted forms of ID, vs 2% overall in local election areas.

Looking at the slightly wider group of individuals who say that the requirement to provide ID when voting at the polling station made them less likely to want to vote in the May local elections (8% of those in local election areas), they are more likely to say voting being easy or convenient as the most important to them when voting compared to the overall mean. Three in five (60%) of this group mention voting being easy or convenient (compared with 49% overall in local election areas) while 25% say their vote being safe from fraud (compared with 48% overall).

Just under half of those who say that the requirement to provide ID when voting at the polling station made them less likely to want to vote in the May voted in the election (47%) while half did not vote (50%). Something related to the need to show photo identification is the most cited reason for those within this group who did not vote (mentioned by 25%) followed by reasons of apathy – not wanting to vote for any of the parties/candidates (23%), no point in voting because it makes no difference (18%) and not being interested in local politics (16%). One in eleven (9% - again, of this overall 8% group) mentioned something related to the need to show photo identification as the only reason for not voting in the election.

One in ten (10%) of those who say that the requirement to provide ID when voting at the polling station made them less likely to want to vote in the May local elections do not hold a valid form of identification for voting (compared with 2% overall).  There’s little evidence that they are any less interested in politics with 67% saying they are interested (compared with 65% overall in election areas) and 33% are not very or at all interested (compared with 35% overall).  However, they are less likely to be regular voters in local elections – 32% say they always vote in local elections vs 47% overall (there is little difference in their views when it comes to general elections).

Could you provide a full breakdown for each demographic group as to what percentage said voter ID made them more likely to want to vote and what percentage said voter ID made them less likely to want to vote (including age, people from deprived areas, renters, people from ethnic minority backgrounds and people with disabilities that that prevent them from voting in person without assistance or on their own)? And what the net figure is for each demographic group?

The computer tables with breakdowns have been published, so there is more detail available there.  However for ease the below table presents data on various demographics of interest on the question Did having to present photo identification make you more or less likely to want to vote at the recent local government elections in your area?. Figures highlighted in red are those that differ from the overall figure AND are statistically significant. Therefore, we can consider these figures to be different from the overall figure. Those that are not highlighted in red are considered to be similar to the overall figures. 

Base: All adults aged 18+ in England with local elections in May 2023

Para 4.17 of your report includes information on a greater number of demographic groups compared to paras 4.11-4.13. If you have data regarding the impact of voter ID on likelihood to vote for those additional groups included in para 4.17 please could you share it with us?

Please see our previous answers for more details on different demographic groups for the question on whether voter ID was a reason for not voting in the local elections or whether it made people less likely to want to vote in the local elections. 

Are there any significant differences between your figures and those of the Electoral Commission?

Unfortunately, the timeframe for this written submission did not allow us to conduct a full in-depth analysis of comparing our results against those from the Electoral Commission’s report. However, we understand that DLUHC are happy to do a more comprehensive comparison of the findings from the different surveys which will likely include collaboration with the Electoral Commission and would be happy to share this analysis with you.

Nonetheless we’re happy to point out a few key comparisons that we believe may be beneficial for the Committee. 

• First of all, it’s important to point out that there are meaningful differences between the different surveys which makes it difficult to make direct comparisons, and means we would not expect results to be exactly the same in any case.  In particular, there are differences in the exact question wording, scales used, and timing of fieldwork (as well as the normal margins of error there would be in any survey).  The fieldwork for the Ipsos survey was conducted between the 18th-24th May while the Electoral Commission data comes from three different surveys conducted by YouGov on 26 April – 2 May, 5 – 22 May 2023, and 5 – 24 May. Furthermore, the Ipsos data is conducted with respondents in the whole of England (with breakdowns for those in local election areas) while we understand that at least the post-poll by the Electoral Commission was conducted in just local election areas. Fieldwork dates should be kept in mind as one of the Electoral Commission surveys was conducted in the midst of an awareness campaign period before the poll while the Ipsos survey was two weeks after the local elections in England to allow for the Northern Ireland elections to be held too.  The Ipsos survey also included some individuals who had been interviewed in the previous wave, and some who were “fresh” sample – the awareness figures reported were only from the fresh sample.
• We have been informed by DLUHC, who have been in contact with the Electoral Commission, that the wording for the Electoral Commission’s awareness questions were very different from those we used. Both surveys also use very different response scales, where the Electoral Commission uses a binary Yes/No response and the Ipsos survey uses four response options “a great deal” or “a fair amount” to “not very much” or “nothing at all”.
• The question on awareness of VAC comes from the EC’s campaign tracker, which we understand was carried out before not after the elections, which is a clear point of difference.  The wording was also different and read “Voters who do not currently own an accepted form of voter ID can get a free voter ID, known as a ’Voter Authority Certificate’, from their local authority. Before taking this survey, were you aware of this? (yes, no)”. 57% said yes. The Ipsos question asked a longer preamble, below, and with a four-point scale (plus don’t know/prefer not to say).  In the Ipsos survey, 21% said they had heard a great deal or fair amount about VACs, and 46% a great deal/fair amount/not very much (but not nothing at all). When observing just individuals in local election areas the figures are 21% said they had heard a great deal or fair amount, and 47% a great deal/fair amount/not very much.

Ipsos wording: “Anyone who is registered, or has applied to register to vote, is able to apply for a Voter Authority Certificate from their local council.  The Voter Authority Certificate is a form of photographic identification containing a voter’s name and photograph. It is designed for people who do not already have identification that can be used to vote, and is available free of charge.

A Voter Authority Certificate will be valid for as long as you are still recognisable in the photograph provided. It is recommended that a Voter Authority Certificate is renewed within ten years after being issued.

Before today, how much if anything had you heard about the introduction of Voter Authority Certificates?”

• On the question of wider awareness of voter ID, there is more similarity.  The EC data on this was taken from their post-poll research, and asked “From now on, all voters in England have to show photo identification at the polling station before they receive a ballot paper. Voters who do not currently own an accepted form of voter ID can get a free voter ID, known as a ’Voter Authority Certificate’, from their local authority. Before taking this survey, were you aware of this? (yes, no).”  87% said yes.  The Ipsos survey asked “Before today, how much if anything had you heard about the fact you will be required to show a form of photo identification in order to be able to vote in person at a polling station?”.  76% said a great deal/fair amount and 90% said a great deal/fair amount/not very much. When observing just individuals in local election areas the figures are 78% said they had heard a great deal or fair amount, and 91% a great deal/fair amount/not very much.
• Both surveys have similar proportions of individuals who did not vote in the May local elections saying they did not vote because of the photo ID requirement (3% in the Electoral Commission survey said they didn’t vote because they did not have a valid ID and 1% because they did not agree with it while 4% in the Ipsos survey said they didn’t vote because of any reason related to the need to show a photo ID).  The top reasons for not voting in the Electoral Commission study were lack of time/too busy (15%), not interested in/fed up by politics (13%), medical/health reasons (8%) and being away (8%); in the Ipsos study they were being away/not having time (22%), not being interested/not knowing enough (17%), not wanting to vote for any parties/candidates running (16%) and thinking there is no point in voting (also 16%).
• Levels of confidence that the local elections were well run differed slightly between the two surveys, although were relatively high in both. Four in five (79%) in the Ipsos survey said they were confident the elections were run well compared with 68% in the Electoral Commission survey. Those saying “Don’t know”, however, were higher in the Electoral Commission survey. Those with negative views were similar between the two surveys (13% in the Ipsos survey said the local elections were not run well compared with 14% in the Electoral Commission survey).
• Views on safety of voting from fraud are also not directly comparable due to different scales being used (4 point scale in the EC study and 5 point scale in the Ipsos research). There is though a similar pattern with around 8 in 10 thinking voting in general (81% EC, 79% Ipsos) and in person (83% EC, 86% Ipsos) is safe from fraud, around 6 in 10 thinking the same for postal voting (63% EC, 62% Ipsos) and around half for proxy voting (46% EC, 50% Ipsos).

Lastly, we would like to share with the committee that since our session providing evidence on 22 November the data tables from the Ipsos survey have been made public by the DLUHC and are freely available at https://www.gov.uk/government/collections/evaluation-of-the-electoral-integrity-programme

7 December 2023