The impact of Voter ID: FAQs
As frequent readers, and anybody who voted in yesterday’s elections, will be aware, photographic ID is now compulsory for voting in polling stations in England. There has been a huge amount of speculation, not least from us, about the effects that voter ID will have on turnout, trust in elections, or election results. Now the results from local elections are flooding in, and so are brand new opinions about the immediate effect of voter ID.
But how much do we now know about voter ID and its effect on the elections? In this piece I will outline a few of the big questions and how close we are now to understanding their answers. These questions should be part of any serious assessment of voter ID, and as you’ll see from the answers, we’re very far away from being able to determine the policy’s effect, or its likely impact in future elections. Watch out for the Electoral Commission and the central government’s reviews of the policy over the next few months.
Also look out for our own research on the topic funded by the JRSST- Charitable Trust. We’re launching a survey of electoral administrators in the middle of May, and we will be publishing our findings over the summer. For more information please email [email protected].
What did we learn from 4th May?
There were a few lessons we can take away from the elections.
First, there were the expected stories about voters being turned away, as in this Byline Times article or this article from the Mirror. These included stories about electors unable to vote because they didn’t have the right type of ID, concerns from poll clerks that people being turned away would not return, worries that the number of people being turned away before making it to the poll clerks would not be recorded. There were also reports of people turned away because they were wearing a mask for health reasons.
These types of concerns are to be expected – there was never any doubt that people would be turned away, but in the absence of more evidence, it is very difficult to say whether they represent broader trends in the implementation of voter ID.
Second, there were a few reports essentially saying that things went better than we might have expected. Although this is good to hear, there are three big reasons to wait before calling this a total success.
- We still don’t know the effect on turnout, especially the turnout of different social groups at risk of exclusion.
- We only know the story from the outside, from the insider perspective of electoral administrators, the state of affairs could look very different.
- This was the first test of the policy, at a set of local elections. Local elections are very important but it means that first, many people have still not had to use ID, such as voters in all of Scotland and Wales, second, the types of people who vote at local elections tend to be unusually engaged – the around 30-40% of the population who vote in general elections but not locals could have a very different experience. .
Without wanting to sound like I’m playing for time,, the main thing to do if we want to know the impact of voter ID is wait until we see the results of more scientific analysis. Even this may only give us a partial understanding. As Megan Kenyon in the Local Government Chronicle wrote, the full effect is very difficult to gauge – if it ever can be.
Were people aware they needed to bring ID?
The most up-to-date polling I can find on this question comes from the Independent who announced on the day of the election that one in five still believed they could vote without photo ID. This is backed up by a YouGov poll on May 3rd which showed 18% had not heard about the changes.
Our own polling with Ipsos from earlier in April showed that there was significant confusion about what types of ID would be accepted. 37% of people wrongly believed a student card would be accepted, and 30% that a poll card would.
Luckily, polling station staff were required to record the number of people who were refused a ballot paper because they did not show a photographic ID and the number of people who did not show an accepted form of ID. It’s difficult to say when these tallies will become available, but there is significant public interest in knowing just how many people did not know about the policy, or did not bring the right form of ID.
Did voter ID stop fraud?
The one major type of electoral fraud that voter ID can stop is called personation which is where one person pretends to be another person at the polling station. According to Electoral Commission data there were 7 allegations of personation in the polling station in 2022, and none of them received further action. Given the low levels of electoral fraud year-on-year, it would be very difficult to say if there has been a substantial decrease.
The Electoral Commission will publish a list of allegations of electoral fraud at a later date, so watch out for that. Usually accusations of wrongdoing and the resultant police action are the only official data collected about electoral fraud. If any post-election audit to determine the level of fraud takes place, we will report on it.
Another way we can assess the level of fraud is by asking polling station staff, as Professors Toby James and Alistair Clark have in previous years. If there are any studies of this type, we will report on those too, but they are likely to take significantly longer to be published considering the vast effort involved in surveying thousands of polling station staff.
Did voter ID reduce turnout?
The short answer to this question is that we don’t know. There are so many factors that affect turnout in an election, from the level of election (local, national) to its competitiveness, to whether there are any other elections held on the same day. On an individual level it is very hard to say why a person may or may not vote in one election or another.
The turnout in English local elections fluctuates but is usually around 30-35%. Unless there is a general election on the same day, and then it is much higher.
With the information we have now, it is impossible to say if voter ID changed turnout. More scientific studies, particularly over more than one election cycle, will be able to tell us more.
Importantly, the overall aggregate turnout figures only tell a small part of the story, because we are just as interested in the individuals and minority groups, who may not appear in the overall percentage figures, who have either been turned away or decided not to vote because of these new measures.
How many people were turned away?
Polling station staff were required to take note of the number of people turned away for not having acceptable photo ID. In the months ahead we should expect these numbers to become available, either from electoral services departments, or – more likely – from the Electoral Commission.
However, as several media organisations and the Electoral Commission pointed out, this would only capture the people who made it inside the polling station. The Electoral Commission said that ‘greeters,’ people who stand outside the polling station and may inform them of the ID requirements, could potentially affect the data collected on how many people were turned away. In their own words:
“However, it is clear that the presence of a greeter at a polling station – while helpful in supporting the smooth administration of polling station voting – is likely to affect the data recorded at the desk. For this reason, the legislation also requires the Electoral Registration Officer for each local authority to separate data for polling stations with and without greeters when submitting data after the polls.”
In other words, we will know if there were greeters, but not if they specifically saw people turn away.
Secondly, we will not know the number of people who were put off from voting before even leaving their homes because of the new requirements. This is not possible to say today, but as more scientific survey analysis comes in over the next few months we will definitely report on it.
Were different social groups affected at different rates by voter ID?
This is one of the most important questions about voter ID, and is also one that we simply cannot answer this soon after the election.
There will be two stages to answering this question. First, the easier question – how were large social groups affected? This includes people like the under 30s, the over 65s, ethnic minorities (in general, specific ethnic minorities are more difficult), supporters of different political parties, and electors from different regions. To figure out if these groups were affected, we will need scientific survey evidence, and very sophisticated ward-level analysis of the type that takes a lot of time. Nobody will be able to accurately tell you today. It’s worth noting too, that the official statistics collected on people being turned away will not include this type of information.
The much harder part of the question will be smaller and harder to reach social groups. A Public Administration and Constitutional Affairs Committee report noted that they had received evidence from charities representing blind and partially sighted people and people with learning disabilities who may be disproportionately affected, as well as LGBT groups who raised the difficulties trans and nonbinary people may have with the reform. Charities representing Travellers have also raised similar concerns. The participation of these groups is just as important for democracy, but they are much harder to reach with traditional survey research. Look out for research on the impact the policy had on smaller social groups within charities representing them.
This should also make us very wary of headlines along the lines of “polls show the vast majority unaffected by voter ID,” because democracy is not just about the majority, everybody deserves the equal chance to participate.
Were people put off from attending the polling station?
We do not know the answer to this question yet, but it should become clearer over the next few months. Figuring out whether voter ID put people off voting yesterday will be a very difficult task. It’s not quite as simple as just asking people if they were put off by the requirements (although that is a very important part of the story).
Any scientific study has to contend with a few major stumbling blocks: many people do not vote in local elections anyway – turnout is around 30-35% – every local election is different, and people do or do not vote for a host of different reasons.
If you were designing a study for this, you would want to find people who voted in one election without voter ID requirements, and then see if they chose not to vote in this election with mandatory ID requirements. If a study like this takes place, dealing with the problem of local elections happening in different places every year, we will certainly report on it.
In the meantime, the Electoral Commission has been conducting research asking people if voter ID will change how likely they are to vote. In 2022, 17% said it would make them less likely to vote and 12% that it would make them more likely. We will have to wait for more scientific studies to find out if the reality matches the polling.
Look out for more polling over the next few days to see if people say they chose not to vote because of the requirements.
Was the policy implemented well?
Before answering this question, it is worth noting that elections are organised and run by small groups in each local authority (usually around 4-4.5 full time staff according to Democracy Volunteers research) and that these staff have been reporting significant difficulties running elections for years, contending with short timetables, limited resources and a stressful working environment. The Association of Electoral Administrators’ chief executive Peter Stanyon has also written for us about how voter ID adds new pressures on top of the existing challenges they face.
With this in mind, we will be running research after the elections, funded by the JRSST-Charitable Trust into what the implementation of voter ID looked like behind the scenes. It is important to hear the perspectives of electoral administrators before judging the implementation of the new requirements.
As to what we can say about the implementation from an outside perspective, polling day started with reports in the Daily Mail that half of councils had given polling station staff less than an hour’s training on the new rules. This is completely unsurprising considering the DLUHC new burdens document specified that elections staff would get funding to train polling station staff for an additional 45 minutes. Whether that is adequate or not remains to be seen, but that’s how much elections staff were given.
On the day, there were a small number of reports of some people not being asked for ID, or the policy being unequally applied.
Were polling staff subjected to any complaints/abuse due to the voter ID requirements?
In the weeks before the election, we saw several stories about police preparing for potential unrest on polling day. As of the time of writing on the morning of 5th May, there are not many stories circulating about major disruption. However, the few that are are worrying.
The Guardian reported on their liveblog (08:43) that the leader of Sefton council – Ian Maher – announced that some polling station staff were ‘subjected to horrid abuse’ when implementing the new voter ID requirements.
We will have to wait until surveys of electoral administrators and polling station staff to see whether there was more hostility or abuse that has gone unreported.
It’s also the case that at a general election, especially a close general election, polling station staff could face more hostility as voters, many of whom will never have voted with ID before, are potentially turned away.
Were any attempts to use fake/fraudulent voter ID recorded?
If there were any, these will have been recorded at the polling station, and the elector asked to return with a different form of identification. When we get information on the number of people turned away over the next few months, we should also get information on the number of suspected forgeries. Our understanding is that the Electoral Commission will be collecting this information from local electoral services department.
Polling station staff were also asked to refer incidents of suspected forgeries to their Returning Officer – the person in charge of running the local elections in that council – so they can in turn refer these to the police. This may also mean there will be police records of all incidents referred to them.
Has voter ID improved confidence in the electoral process?
This is another question we won’t easily be able to answer today. Many supporters of voter ID have argued that it can improve perceptions of the process. Sir Eric Pickles led a review into electoral fraud in England. This review made it clear how important trust was as a justification for the introduction of voter ID: ‘ We need to make sure that people trust the system and that perceptions can play as big a part in undermining the system as well as actual proof of fraud.’ As the debate on voter ID has continued, supporters of the reform have increasingly referred to how it can improve confidence in the electoral process. The big question then is, has it improved trust in the process?
Survey research will seek to figure out if confidence has improved. The Electoral Commission performs regular surveys to determine the levels of public trust in elections. In their 2022 research they concluded that public confidence in elections was high. Satisfaction with the process of voting is over 80%, only 8% of people think their vote is not counted accurately (one standard measure of confidence in electoral processes across many studies). In general, this looks like a positive picture of confidence in elections.
But, there is some evidence that concerns about fraud are still prevalent, in 2022, 31% of respondents across England agreed with the statement: ‘There could be enough electoral fraud in some areas to affect the election result.’ 31% is quite high, and completely out of step with the empirical evidence, which suggests that proven cases of fraud, particularly personation, are vanishingly rare.
On the other hand, as we have written about in more detail here, voter ID could undermine trust in elections. If it has a disproportionate impact on certain minority groups their confidence may be reduced, if it is poorly implemented that may reduce confidence in the fairness of the process, and if it becomes a polarised political issue that could reduce confidence, as we have seen in the United States.
We will have to wait for more rigorous analysis, and probably for more elections to see if the introduction of mandatory ID in polling stations has changed public confidence in the electoral process, and if it has then in which direction and for which groups.
Any other questions?
If you have any other questions on voter ID, please send them to [email protected]