Levelling Up, Housing and Communities Committee
Oral evidence: Electoral registration, HC 841
Monday 3 July 2023
Ordered by the House of Commons to be published on 3 July 2023.
Members present: Mr Clive Betts (Chair); Ian Byrne; Ben Everitt; Andrew Lewer; Mary Robinson; Mohammad Yasin.
Questions 155 – 231
Witnesses
I: Peter Stanyon, Chief Executive, Association of Electoral Administrators; Graham Farrant, Chief Executive and Returning Officer, Bournemouth, Christchurch and Poole Council; Miranda Cannon, Returning Officer, Leicester City Council.
II: John Pullinger CB, Chair, Electoral Commission; Ailsa Irvine, Director of Electoral Administration and Guidance, Electoral Commission; Craig Westwood, Director of Communications, Electoral Commission.
Witnesses: Peter Stanyon, Graham Farrant and Miranda Cannon.
Chair: Welcome, everyone, to this afternoon’s session of the Levelling Up, Housing and Communities Select Committee. We are going to have a further evidence session this afternoon on electoral registration. We have two panels this afternoon. We are also going to explore the question of the introduction of voter ID at the recent local elections, which was the first time that voter ID had been required for people to vote at elections in this country. We are going to explore with our witnesses what impact that had on voting at those elections.
Before I come over to our first panel of witnesses, I am going to ask members of the Committee to put on record any interests that they may have that may be directly relevant to this inquiry. I am a vice-president of the Local Government Association.
Mohammad Yasin: I am a member of Bedford Town Deal Board and employ a councillor in my constituency office.
Q155 Chair: Thank you very much to our witnesses for coming to give us first-hand experience of what happened at the recent local elections, particularly with the new voter ID requirements. Could I ask you to introduce yourselves, saying who you are and the organisation that you are representing today?
Graham Farrant: I am Graham Farrant. I am chief executive of Bournemouth, Christchurch and Poole Council, or BPC Council as it is often known.
Peter Stanyon: I am Peter Stanyon, chief executive of the Association of Electoral Administrators.
Miranda Cannon: I am Miranda Cannon. I am director of delivery, communications and political governance and the returning officer for Leicester City Council.
Q156 Chair: I will begin by asking Peter Stanyon first, followed by Graham and Miranda, whether the implementation of voter ID at the recent elections was a success. Did it go according to plan? Did it go well?
Peter Stanyon: Yes. The feedback we received is that, administratively, it went as smoothly as could be expected, so it was a success. There will be pockets of instances where there was a degree of confusion at individual polling stations, but generally, across the board, the way that the staff took to the training that they received and then the delivery in polling stations meant that, in the main and to our knowledge, no major issues were experienced at any of the polling stations across the 230 local authorities that delivered the elections.
Graham Farrant: Overall, there were no problems. Some presiding officers did not take up their role because they felt it was going to be too complicated, so it had a minor impact as it was coming in, but in terms of the implementation, it was smooth. It was well advertised, people seemed to know about it and the process seemed to be fine. We can talk about some numbers later on in terms of what the impact was and how that appeared, but generally it was well implemented.
Miranda Cannon: Similarly, it was perhaps better than expected. It went very smoothly. It was challenging for the staff, in the context that we were running two polls—both mayoral and local elections—but generally I was pleased with the level of understanding and awareness of voters who were turning up at the polling stations.
Q157 Chair: There were, apparently, 14,000 people who turned up and did not have the right ID or any ID at all and were turned away. Is that an accurate reflection of what happened?
Miranda Cannon: As you may be aware, at my authority we had greeters at every polling station. We chose to do that from a positive customer experience point of view. Clearly, some people, as we explained the requirement at the door, went away at that point. The data that we captured inside was not necessarily entirely the picture overall, but I would still say that it was a comparatively low percentage who were turned away. The recording of the data was a challenge on top of all the other requirements to achieve, and I am sure that others will comment on that too.
Q158 Chair: If you had meeters and greeters, you really do not know the actual figures for any polling station.
Miranda Cannon: No. We have a rough estimate, but no, we do not have an accurate one. We recorded only those who came to the desk in the polling station where they did not have ID at that point. There were quite a number who were explained to at the door and then who chose to go. Quite often, they came back, but the numbers will differ.
Graham Farrant: We have greeters only where we have a joint polling station, so where there are two polling stations in the same building. We did not have greeters at the majority of ours, so our figures are reasonably accurate. Some 0.35% of voters were turned away in the first instance, and two thirds of those came back with suitable voter ID. Our figures are probably more accurate, particularly for those polling stations.
I also visited somewhere between 45 and 50 polling stations on the day of our election, because that is what I do, and got good first-hand knowledge of it. Most people, in my experience, were arriving at the polling station with a form of ID in their hand, and you could tell that they were going to vote, because they were carrying a passport or a driving licence, or sometimes a passport and a driving licence and a bus pass, because they wanted to be certain. We had some really good experiences.
My sense is that there was a high level of awareness among those who were going to vote. In casual conversation with people who do not vote, I am not sure that the awareness was there, primarily because they are not interested in elections and, therefore, just did not take an interest.
Peter Stanyon: I can only agree with what my two colleagues have said. As is shown in the Electoral Commission’s interim report, statistically there are challenges with the figures. The other thing that is not necessarily reflected is that there were greeters in every station, such as in Leicester, or in some polling stations, as Graham as described, in BCP. In polling stations where there were no greeters, there was clear signage to say, “If you do not have your ID, go back and get it before you arrive in the station,” so there is another factor in there as well in terms of the customer service element grinding against the statistical analysis of the effect of the voter ID. To the point that Miranda initially made, it was there to assist people through the process, but that then had a negative effect in terms of the analysis that would then flow from that.
Q159 Chair: It did not go too badly, then. Clearly, some people were turned away or may not have arrived in the first place. Is it going to be a bigger challenge when the general election comes round?
Peter Stanyon: Yes, absolutely. Drilling down into the interim analysis that has been done already, in terms of the areas where people do not regularly vote at local elections, there will inevitably be a lag in terms of the need to bring that ID with them, because they have not been through the process this time around. Over the years, voters will get used to having to bring ID, but it needs that cycle for everybody to have been through a process. There will be those who are very engaged and will vote at every single type of election, and others who will vote only at general elections, for example, so we would expect there to be a degree of explanation needed on a higher level than it was at the locals, where there are lower turnouts.
Q160 Chair: Graham Farrant, what is your view? Are you looking forward to that challenge? Or might it be a little bit more overwhelming than the recent one?
Graham Farrant: As people hear about a general election, they will get more interested. As they switch on, they will see the adverts. One of the examples that I gave was in March, when I said, “I am an interested party in the election and I have not noticed a single advert about voter ID.” I remember saying that distinctly in March, and then I saw the advertising campaign and began to see that come through. There is an issue about whether people are genuinely interested and, leading up to a general election, there is a much greater awareness about elections.
To be honest, local elections do not catch the eye. We have just over 30% turnout. It might be worth saying that we had a resignation of a councillor two days after the local election, so we have just held a by-election in one of our wards six weeks after the previous election, and we did not do a great big campaign about voter ID. We had a slightly higher percentage of people turning up without voter ID this time. It was marginal. We only had 2,000 people coming to vote anyway, so they are not exactly reliable statistics, but it is worth mentioning.
Some people came in and made a point that, “Last time you turned me away. This time, I have my photo ID,” but there was a slightly higher percentage. Twenty-one people out of about 2,000 were turned away, of whom 13 came back, so eight out of 2,000 is a slightly higher percentage than we had first time. They are still small numbers, and it is worth making the point that it is still a much lower number than those who spoiled their ballot paper. In our first election, we had 75 who were turned away and did not come back, but we had over 300 who spoiled their ballot paper. It is worth putting into that perspective.
Q161 Chair: Miranda, will the bigger challenge at a general election be that you have a wider group of people who are probably not normally that interested in the elections and do not normally turn up, but may just turn up for the general election? There are people like that, because there is a higher turnout. Do you think it might be harder to get to them and get the information to them?
Miranda Cannon: I think so, generally. As Peter said, at a local election, with lower turnout, you are typically reaching those who are very much engaged. I expected them, in some senses, to be more likely to have some form of relevant ID. I worry that at a general election, with much higher turnout, there will be those people who vote only sporadically and periodically, and perhaps, in that sense, some of those in that group are more likely to not have the requisite ID and to be less aware of the voter authority certificates.
My worry is not about the general awareness of voter ID. In some senses, it is about getting across the process about the free certificates and how people can apply for those, and so on. As the Electoral Commission research has shown, there was a much lower level of awareness of that, and I would say that my gut feel from talking to people locally in Leicester is that that is very much the case on the ground. We need to do more to get awareness across of that particular element.
Q162 Ian Byrne: Did voter ID impact the voter turnout of any specific groups? If so, what can be done to alleviate any negative effects?
Miranda Cannon: There was not an apparent overall impact on turnout. In fact, our turnout at the local elections in May was higher than the last time we had local elections, and there was no evident drop-off in terms of any particular groups or wards. The patterns generally replicated which wards we would normally expect to have higher levels of turnout and, as I say, the overall turnout was about 3% higher overall than our last local elections. I cannot discern any particular negative impact on any particular groups at this point.
Q163 Ian Byrne: So you have had no feedback from any of the disabled groups or anyone like that.
Miranda Cannon: No, we have not. Sometimes after an election, you expect a number of complaints or issues to come to the fore. Interestingly enough, there fewer on this election than we might have had previously. As I say, I felt that overall it went smoothly and did not have a major impact on people’s engagement, but I do worry that that will be slightly different when we get to higher numbers at a general election.
Peter Stanyon: Anecdotally, turnout was about where it was four years ago, or just slightly below.
Q164 Ian Byrne: There was no chilling effect.
Peter Stanyon: No, nothing major. There was a general feeling that it was slightly down by a percentage point or otherwise, but again, that is only anecdotal. We have to wait to see the stats from the commission in terms of what the effect was in that respect.
Graham Farrant: It is a similar story and probably slightly down in our area. There were no complaints from any specific groups. I have heard that those with a disability are sometimes less likely to have voter ID. That is one of the things that has been said, but I do not have any empirical evidence for that. Talking to people from the polling stations, their feeling was that older people who were absolutely going to vote were coming along with, in some cases, passport, driving licence and bus pass, because they wanted to make sure—they wanted to vote. One or two people tested the system, but there did not seem to be any particular group that we turned away, such as students or older people.
Q165 Ian Byrne: In terms of the 14,000 who were turned away nationally, are we going to get a breakdown of who was affected? Was there any pattern within that 14,000?
Graham Farrant: We did not see that.
Peter Stanyon: I do not believe that the commission has the ability to break that down in terms of individual groups, as they were not recording that in that format at polling stations.
Q166 Ian Byrne: That may be something that we should look into, if there are 14,000 people who were turned away. I want to touch on the postal element of it. Do you have any concerns over Government plans to implement voter ID for postal votes in the next election?
Graham Farrant: I only have concerns about the complexity of introducing it and getting everybody to understand it. I have met lots of people who use postal votes and who struggle with the postal vote system as it exists. There is a habit about voting in elections, is there not? It is sometimes a reassuring habit, because you go in and it feels like quite an old-fashioned environment when you are doing it in person. People have got used to the postal votes. Any change is likely to cause some concerns.
Q167 Ian Byrne: Did you see any increase in the postal votes that were applied for with regard to getting around the voter ID element of it?
Graham Farrant: We wondered if that might be the case, but we did not see huge increases. There is always a steady increase in the number of postal votes as more people register, but we did not see any real evidence of that.
Q168 Ian Byrne: The Government are going to start implementing voter ID for postal votes. What would that look like?
Peter Stanyon: It will mirror the registration process, in that, at the point of application, the national insurance number will be required, and that will then be checked against the DWP databases to ensure that there is an individual of that identity. Leading on from Graham’s point, the complexity will be at the point of application. It is a three-year reapplication process rather than it being there almost in perpetuity.
The issue with that is that, first, the national insurance number will need to be produced. Secondly, there could be chase-ups, because someone has not produced the necessary information, and there is a chase cycle. As it gets closer to a major electoral event, that is where the complexity will come, because inevitably elections drive both registrations and now, with the online absent voting element that is being proposed, applications for postal votes quite late into the process.
Q169 Ian Byrne: Miranda, do you have any worries about ID being introduced for postal votes?
Miranda Cannon: Just building on what Peter said there, it is about the impact close to any poll, and particularly larger general elections. We already see that surge in registrations and in demand for postal votes. If you have more checking and more processes to do at a time when you are preparing for what is a major event, it is a huge burden at a time when a lot of electoral services teams are already under the cosh—and they are not big teams. Trying to recruit and hold on to that technical expertise is a big challenge, so it can only start to impact. You will see perhaps more staff leave the sector. It is a very specialist area. You do not get technical electoral services staff in any other walk of life, so it is that worry. It is about having the capacity to deal with it.
Ian Byrne: It is about the capacity to carry it out.
Q170 Mary Robinson: It just occurred to me that we often hear about GDPR difficulties between data held by central Government and data held locally; would voter ID, in terms of whether any other information was required from DWP and so on, be impacted by that?
Miranda Cannon: I do not think that it will be massively, because we already have pretty strong protocols introduced with individual electoral registration and a lot of data-matching processes that happen already. In some senses, there is a very established way of working now, which builds on that. It is more about the checks and balances, as Peter referenced, and that chasing process, where perhaps the information has not been accurately provided or people did not have that information. That is the worry, not particularly the data-matching. That would be something that we are used to and we have quite an established approach to.
Graham Farrant: It is about the complexity. When you are getting somebody to write down a nine-digit number, and to get it in the right order, with people transposing numbers and so on, that is where the extra work comes that we never allow for. You assume that if someone is going to give their national insurance number, they are going to get it right. It is all that extra work, and that is the administrative burden that just adds to the complexity of it and makes staff doing that just feel less valued. We should not underestimate the complexity of not getting it wrong.
Q171 Mohammad Yasin: We have data that 14,000 people were turned away from polling stations because they did not have valid voter ID, but I believe that thousands did not turn up to polling stations. Is this data misleading?
Graham Farrant: It is as accurate as we could get. We will never know whether some people did not vote because of it, but it was pretty much in line with the trends on voting. We did not see a higher number. For us, the total number who were turned away initially was 223. That is across 33 wards, so it is less than 10 per ward. That would not have impacted on the results of any of our wards in terms of the closeness of the election, but there are times when it would.
We do not know about people who did not turn up because they knew that they did not have voter ID. In conversations, I have not heard of anybody who has said that or anybody who said they knew people who did not turn up. There are bound to be some, but I cannot put a number on it.
Q172 Mohammad Yasin: People said this to us when we were door-knocking, or even after the local elections. What is your experience, Miranda?
Miranda Cannon: It is probably similar to Graham’s. We have not had anyone come back afterwards and say that this was a problem for them and a barrier to them. Like Graham did, I went out on polling day and there was a general positive level of awareness, and I was not sensing from people that this was a great barrier. We just will not know whether some people were sat at home thinking, “I am not going to go out today because I do not have that ID.” That would be very hard to discern. In our case, turnout was a bit above what we have seen before, which implied to me that it was not proving a major problem.
Q173 Mohammad Yasin: Moving on to impact on local authorities, was new burdens funding adequate for local authorities to cover the costs associated with implementing voter ID?
Graham Farrant: From my perspective, it was adequate for the one-off this time. Will it be enough long-term? For example, you buy privacy screens and mirrors; will there be enough in the budget to continue to replace those when they get damaged or lost in between elections? That is the key concern for me. It has added to the total running cost of the election.
What it does not address are the issues about recruitment of polling station staff, some of whom just said, “That seems like it is going to be too complicated. I am not going to be a presiding officer any more,” so we lost some.
Q174 Mohammad Yasin: Was it difficult?
Graham Farrant: It was tighter than we have ever had it before. BCP Council would not have been able to run our local election if Dorset Council had had local elections at the same time. Four years ago, they were both at the same time, and we coped; this time, we got some staff in who would normally work for Dorset Council. In our geography, there are not many other places to go, because there is a big forest to the east and the sea to the south, so we are reliant on each other.
When it comes to a general election, we will struggle. We will have to change the terms and conditions. We are looking at our offer to people who work in the polling stations, and we might have to improve that, which will be a cost to the council. I do not think that has been thought about or taken into account anywhere in the calculations that I have seen.
Peter Stanyon: I totally support what Graham says. The number one concern that came out of our survey of AEA members was polling station staffing. This was not a national poll. It was 230 local authorities out of the 360-odd in England, and that does not include places like London or Birmingham and the big metropolises. Scotland and Wales have the challenge coming ahead again. That is the huge backdrop to how the resources are given.
They were used differently in different local authorities. Certainly, some spent an awful lot of that resource on things such as privacy screens and access for those with a disability, for example. Others pushed it towards staffing and things like that. Until it has been through the whole process on the justification-led bids, which is the second half of the funding element, we will not know how effective the funding was, other than the fact that no one has come out and said, “We were bereft of funding.”
It is more about how that now affects, as we move forward towards a higher turnout election, with more focus on the elections themselves and the issues that Graham talked about, with staffing being the No. 1 concern. On the basis that it has been a concern for a long time, this is adding another layer to that concern as we go forward.
Miranda Cannon: We are still, in effect, trying to calculate that overall cost, as it were. The points that colleagues have made are absolutely about what the one-offs were this time round versus what some of those additional ongoing costs are going to be. We will be submitting a bid for some of the extra funding, hopefully, that will be available. The initial allocation was enough, but we did choose, for example, to invest in having a privacy screen in every polling station, and one that we could use again and that would be sufficient and robust enough for that purpose.
We chose to write to every household in January, setting out who was on the register, as well as details of voter ID, because we felt that it was important enough to do that. That was a £60,000 cost, which was a very significant cost but an important one to do in terms of getting the message across. We chose to do some things because, to us, they were really critical and important, but we will need to challenge what we will do on an ongoing basis as things start to settle down around ID. Inevitably, there will be ongoing costs around things like voter authority certificates, where, again, additional staff time and costs are going to be needed on an ongoing basis.
Q175 Mohammad Yasin: How challenging was it to recruit staff at the polling stations?
Miranda Cannon: We started our recruitment many months before. We already have quite a strong database of staff. We were not in the position of needing to borrow staff but, equally, the district councils around Leicester all had elections too, so that would not have been an option for us. We were all competing, in some senses, for a pool of staff.
We had enough people initially. We built up a contingency pot of 30%. We used all of those contingency staff, because we started to see staff dropping out as we got closer to the poll. Some staff dropped out after they had done the training. They felt a little bit concerned and uneasy about the additional work that was involved, and so came back, sometimes at quite short notice, to say that they were not prepared to do it.
We were lucky that we had that big contingency, but we used it all, and that is my worry. Suddenly we are asking people who have done it time and time again to do more on polling day, and it is just something that they do every so often. There may come a point where they say, “Enough is enough. I am not willing to take on all of that responsibility any more.”
Peter Stanyon: If could just add to that, one other factor that has been fed back to us on this occasion was that it was the King’s coronation, so there were two bank holidays. There would be people taking the opportunity and who would normally have worked, so that was like the icing on the cake of the additional challenge coming along, but also the fact that people were taking the opportunity to take a longer break over that period.
Q176 Mohammad Yasin: Just before the local elections, some people did not realise until the last minute that they needed some ID to vote and applied for voter authority certificates, and there was a surge in some local authority areas. Did local authorities meet that demand effectively, or was there a lot of struggle?
Miranda Cannon: From our point of view, we met the demand. Overall, we had planned on the basis that we would see a bigger surge than materialised in practice. Without doubt, on the final deadline day for voter authority certificates, and the day before that, we got the same amount as we had done over the course of the whole preceding week. Even so, we are talking about only 40 or 50 voter authority certificate applications. We had geared up to manage significantly more.
Again, it was great this time around in the sense that we could manage it, but I will worry when we get to a general election next time that we will see a much bigger surge, which will be much more difficult to manage. That will be in my mind next year, certainly.
Graham Farrant: I had a very similar experience. The problem is—and I come back to this point—when you start thinking about the local election if you are the elector. Some will think about it on the Tuesday or Wednesday of the Thursday election and at that point it is too late for a voter authentication certification.
We could be a little more lenient on the identity that we allow. We had a lot of people who turned up with their council passes. It is a perfectly good council pass—it has my photo on—but that is not allowed. A lot of NHS staff turned up with an NHS staff pass, but that was not allowed. We could allow employer or public sector passes. We could be more lenient around the definition, which might get over some of the hurdles.
Because of the early deadline, there will have been people who thought too late about the voter authentication certificate and could not apply. It is about thinking about human behaviour and when the election becomes real to them.
Q177 Mohammad Yasin: Peter, what needs to happen to ensure that the voter authority certificate system works effectively for electoral registration officers prior to the next general election? As Miranda said, there is going to be a bigger surge in a general election than there was in local elections.
Peter Stanyon: I have said before the Committee previously that the front end of the portal was fantastic for the elector in terms of uploading the information, but it took a long time to get the administrative side working. It needs to be smooth and simple to allow that processing to take place. I believe that the key element to limiting the number of applications will be good communication in areas to say, “If you have your driver’s licence or your passport, you do not need a VAC, but if you do, these are the deadlines.”
It needs to be brought into the psyche, as Graham says, as soon as you know an election is taking place, in order to make sure that people know what the new system is. Over time, that will become slightly easier, because people will become used to having to go through that process.
The challenge is in that window, and it does fill me with a degree of trepidation that, if all of the parts of the Elections Act are rolled out in time for the next general election, we are moving into an area where you have minus 12 and minus 11 away from the action. You will have the registration to vote, where we know already that there are spikes in registrations. Online absent voting will inevitably create spikes just before the election. As people engage with voter authority certificates, there will be that spike as it gets closer, six days out from the election.
We also have a degree of concern about the statistics, and the completeness and accuracy of registry reports that have been done by the Electoral Commission recently. Are the registers as full and complete as they were in 2019? Again, that might exacerbate that challenge as we get closer to the election. The reality of that election timetable at the moment is that everything happens in that short window, while the same staff are preparing to deliver the actual election.
That can be resourced up. We are aware of it and would expect returning officers and their teams to be looking to mitigate that very early on, but the brutal reality is that, as Miranda made reference to, local authorities had to over-resource for voter authority certificates, because there was no benchmark to put that against. At the next general election, we will be in a similar position, where we do not know where the benchmark will be for the amount of resource and complexity that will come in.
The Government have responsibility to make sure that the systems are smooth, efficient and accurate coming into local authorities, and then it becomes the responsibility of the local authorities to assist returning officers as best they can in making resource available in that period and seeing where it goes.
Q178 Mohammad Yasin: How confident are you, Miranda and Graham, as returning officers, that you will be able to meet the need?
Miranda Cannon: We always deliver. I work on the basis of the worst-case scenario—I expect Graham will say similar—and you plan up to that. Elections are not just about your electoral services team. It is about drawing in the wider organisation. I am lucky in that I do not work in a small district council. I work in a unitary authority with 6,000 staff, so I have a pool of others, if necessary, and my colleagues at senior level understand the importance and, if necessary, will assist, but that becomes more and more difficult as the financial pressure on local authorities continues to squeeze.
This time round, I was drawing on finance and revenues and benefits staff, for example, to train them up on the portal to be able to step in if we needed them. I will probably anticipate doing similar and bringing in some of those extra resources. I have no doubt that we will meet the challenge, but it will be a big ask and one that concerns me somewhat at this point.
Graham Farrant: My sense is that everything is getting more complex and tighter in terms of the availability of staff. Going back to 20 years ago, for the count, we used to bring in bank staff who were really good at counting notes and counting paper. They do not exist any more, because it is all done by machine, so that pool of resource has gone and we have different people in doing the count.
What we will find is that as we make these requirements more complex, we will need more people. Those people will be less highly skilled. They have to be. I will give you an example. We had a by-election last week, because one of our newly elected councillors resigned after two days. We did the count using our election services team, and I am sure that they will not mind me praising them. They did the count very professionally. Every single box that was counted matched to the number first time. The count was spot on first time. You could tell the difference between having a professional team who does it for a living compared with the normal election team that we have, who do make some mistakes.
The danger with these requirements is that we are requiring more people to be involved, and they have less expertise, so the risk of something going wrong gets marginally greater every time you add another layer of complexity. When you add a layer of complexity with the speed with which we are doing it now, you make it more and more likely that something will go wrong.
On voter ID, we could have gone for a model that said, “For the first four years, we will ask whether you have voter ID. We will tell you that you should have voter ID, but we will still let you vote.” We did not. We went for a model first time out with that: “Bang—it is compulsory.” The more we do that, the more we layer risk on, and that is the challenge.
As a returning officer, you are constantly assessing the risk. Do you print ballot papers for everybody? A few years ago, people did not, and then we had a peak. Therefore, that is a risk that we now get over by printing ballot papers for everybody. That is the problem. We are just layering risk on, and that risk might be marginally greater, but it is getting greater every time we put in a new requirement and the skills base gets weakened.
Q179 Andrew Lewer: I would just like to take your views on whether you thought that the Electoral Commission’s information campaign on voter ID was successful, particularly in terms of targeting the groups who are less likely to have ID. We have touched on it a bit, but I am interested in the specifics of the campaign.
Graham Farrant: I was concerned in March. As a returning officer, I did not feel that I had seen anywhere near enough publicity. By the time of the election, I had seen and recognised some publicity—some adverts and some campaigns, on television, on buses and at bus stops, and on leaflets. When I speak to friends and neighbours, those who were interested in elections were aware of the requirement. Those who were not really interested but might vote or might not did not really know. They did not have a clue about the new requirement, but all had a photo ID in their pocket or in their wallet that would have been suitable, interestingly. There is this issue about how good the publicity was, and it did get much better. In the last four weeks, it was a strong campaign and it made an impact.
In those groups who do not have voter ID, that might have been a bit late. I keep coming back to when you trigger your thoughts about the election and when you apply for your VAC. You may not do that until the week, and then it is too late. We are going to need to campaign on this issue for a good number of years to get people into the habit of it. It is not good enough just to say, “We will do it for this election and not the next one.”
We saw a marginal drop-off in people bringing ID for a by-election six weeks later, so that message needs to get through until people get into the habit: “To vote, I need to bring photo ID.” Some populations—and particularly those from eastern Europe—said, “Of course you bring photo ID. I’m amazed that we used to be able to vote without it.” They were used to it. They have seen it. They expect it. That is where we need to be and we are not there yet, so we need to keep the campaigning going probably for the next three or four years.
Q180 Andrew Lewer: When they brought this in in Northern Ireland, did they do it for years and years, or did they just do a local and a general and then it was accepted?
Graham Farrant: I do not know.
Andrew Lewer: I am not asking you to catch you out. I do not know. I am genuinely interested in how long they decided to go along with it.
Peter Stanyon: I can only echo everything that Graham said. It is a long-term strategy. The figures that the EC report were that there was 57% awareness at the start, which rose to 87% on polling day and 92% after polling day. In terms of the direction of movement, it was a very positive response in that respect. Equally, when you look at the statistics, there are pockets within that of lower take-up. I am not a comms expert; we need to look to those to get that message across. How do we get it into every single stream in terms of raising that awareness over a period of time, which means that, ultimately, it becomes second nature to bring that along?
I saw the adverts when they were on the television and in the paper, for example, but I am engaged enough to be looking out for them. Experience says to me that every time there is an election, you become more aware of it the closer you get to. If you go back to the voter registration stats, it will show that, when there has been a leaders’ debate on the TV, the next half-hour sees a whopping great rise in applications, because it is in the face—if that is the best way of saying it—of the engaged electorate. It was a positive campaign, but a lot of work needs to be done as well to continue that going forward.
Miranda Cannon: It was complex, because you are trying to get across a number of messages for the first time—first, ID is needed; secondly, these are the forms of ID; and thirdly, if you do not have these forms of ID, this is what you need to do. In terms of the first of those three—you need ID—it was strong and the awareness grew quickly, which was really positive. It hit most of the buttons in terms of people understanding what forms of ID. The challenge is the next level about, if you do not have ID, what you need to do, and getting that message across about the voter authority certificates, of which, as we all understand, there was a much lower uptake than might have been anticipated.
It inevitably has to be supplemented by local comms, particularly in a place like, for example, my own in Leicester, where it is extremely diverse and language is a particular challenge. There were quite a lot of the national materials that were not translated, for example, into the sorts of languages that are important in our city, Gujarati being the top one. Nothing like that was provided, so it was very important that we supplemented it, but again, there was a cost to that.
It was a really good start, but as colleagues have said there is more to do, and there is more to do to understand some of the more detailed nuances of the system as opposed to the overall awareness, which got out quite quickly.
Q181 Mary Robinson: The impact or lack of impact of photo ID is something that is important to discuss, but it does highlight another issue, which is around the number of people who do not vote anyway. Linking those two criteria would be quite difficult. However, we have a valuable piece of material, which is the marked register, so we know which people vote every year or vote regularly. Is there an opportunity to examine that more closely and link that to a voter ID campaign in order to find out why people are not voting in first place and to get them to vote?
Peter Stanyon: That is a good question. In terms of a returning officer or an electoral registration officer using that data, they are not able to look at and use it for a campaign, ironically. They are the guardians of that data. That is not to say that data is not available for someone to undertake that research. In many respects, that is probably the better way: for someone independent of the entire process to try to identify and then drill down into why people do not vote and what the reasons are for why turnout was lower at a local election than it would be at a parliamentary election.
I totally support the idea that the awareness needs to be properly presented in different local authority areas dependent on the demographic or the challenges. At the moment, the returning officer or the electoral registration officer is not in a position to be able to do that piece of work other than to say, “There is the data. You go off and do that for us.”
Graham Farrant: But we should be investigating why people do not vote. Our figures are that average turnout is about 32%. It was much higher—probably about 40%—in terms of postal votes, but only 25% in terms of those who were eligible to vote in person. In the by-election that we have just had, we had a total turnout of 18.5%. That by-election could have been won by somebody getting one-twentieth of the eligible votes. As a returning officer and as a representative of local government, my sense is that we need more people to take part in the democratic process. They do not, and we should be researching why. We get 30%-plus for a local and about 60% for a national election, but that is still only two-thirds of people voting.
Q182 Mary Robinson: We have discussed photo ID and have touched along the way on the impact on the other things that will be implemented when the Act comes into force. Has the implementation of voter ID changed your assessment of EROs’ capacity to implement the remaining provisions of the Elections Act, such as overseas, proxy or postal voters?
Peter Stanyon: One of the difficulties that we have is that we have yet to see the secondary legislation and the effect that it will have on administrative processes. I speak with what is a best guess. That is the best way of describing it. Inevitably, despite whatever work is done with comms, the general election will drive overseas electors. The new residency requirement rather than registration at a station, together with the 15-year rule disappearing, has the potential to increase the number of applications, and then the complexity comes on the back office to work out where that individual should be registered.
I have already referred to online absent voting, in that it will be closer to the election, alongside VAC applications and the applications to register. The challenge about the handling of postal ballot papers is beginning to fill me with a degree of concern. In the polling station, there is another process to be levelled on to the staff or volunteers to ensure that that individual completes a form. If they do not do so, they are all rejected. In fact, if they come with seven of them, six will be accepted and they have to pick which one is rejected. There is a lot of complexity. Ironically, in some ways, voter ID was the easier element to introduce, because it is more upfront and it is a straight, “You have the ID or you do not,” so it is a yea or a nay when it comes to that.
The secondary measures that are coming in now on the phased process have the potential to stretch the experts in terms of some of the back office procedures, but also making sure that staff in the stations continue to do what they have to do, which sometimes is to say no to a voter, and that goes against the grain of most individuals in a polling station. There is a hugely testing period coming as we go forward, certainly for those who have not had elections in May, because they will be having the first voter ID with the other elements at the same time.
Q183 Mary Robinson: You mentioned stretched expertise; how critically stretched will it be? Is it doable?
Peter Stanyon: As Miranda said, it will be delivered. All I can say is that in our membership organisation, we have advertised more electoral services and electoral services managers’ jobs since May than I can recall in my time at the association. There are some highly intelligent people coming to electoral services, but I would always say that until you have been through that five-year election cycle, you need to know the nuances of where it can go wrong and to nip it in the bud very quickly. That is the bleed in terms of the experience. The bodies will be there; it is the experience of those individuals in post that is the worry.
Miranda Cannon: On our ability to deliver, the level of risk goes up dependent on a number of other factors, those being the timescales that we are expected to deliver within, voter ID being a classic challenge of getting the secondary legislation very late, and my staff getting guidance and training on the portal that they need to use to administer it only at a very late stage, just days before it was going live. If we are put under those conditions again, that elevates the risk. Having sufficient lead-in time in terms of the legislation, the detailed guidance, the communications campaign, the IT systems and the training are things that will make a bigger difference in terms of my confidence in our ability to deliver it well. I have pretty much no doubt that we all deliver it well under whatever circumstances, but to do it well and do it properly, those things need to be properly done in a considered timescale.
Graham Farrant: I would agree with that entirely. It just increases the probability of, at some point somewhere, a significant error being made that undermines the sanctity of the election process. Your original question was, “Would it change your evaluation of it?” It absolutely will change the evaluation. We will do it again. We will layer on some more risks that we probably were not aware of when we first had voter ID. We will do everything that we can to get the right people in the right place, in the right quantities, to do what we need to do, but it just increases that risk of something going wrong that is material and that materially affects the result of an election.
Q184 Mary Robinson: Peter, given that timing is critical on this, what is the latest that you want to have the secondary legislation in place in order for these measures to be implemented successfully?
Peter Stanyon: Tomorrow. Joking apart, it is the Gould principle that has been constantly quoted: a minimum of six months in advance. That did not happen with voter ID, which was delivered four months in advance and was delivered well in the circumstances.
One of the big challenges is not so much what the legislation says, but the certainty about what it will say and then the ability for the commission to produce the guidance and for the training providers to produce the training. Often, it is not just the electrical services core team who need that awareness raising; it is those who will deliver it in polling stations, and those are different messages at different levels for different people. Certainly, six months is the bare minimum that we would want to see.
Linking it to the words that Graham used in terms of a failure in an election, one of the beauties of our system is that 650 separate elections take place to the House of Commons, so failure in five seats is not a failure of the system; it is failure in those five seats. In some respects, that will hide what could be an endemic issue for each of those that others have been able to cope with slightly better. It is just layering change upon change, so the sooner, the better, but certainly six months at the absolute bare minimum.
Q185 Mary Robinson: In terms of the impact on EROs, the changes to voter ID were important and impactful, but the challenge was risen to. With these other measures that will have to be introduced, is this going to be far greater?
Peter Stanyon: Yes, for the core teams. We are stepping into the unknown. I cannot quote you the number of potential overseas electors coming, but I think it is potentially an additional 250,000 arriving on to the electoral register. It is the process to follow, but it is that worst-case scenario planning. If you are worst-case scenario planning on that, you are looking at 1,000 or 2,000. If you get 50, in some ways that becomes as great a risk to the rest of the election, because you have thrown resource at a certain area. A lot of it is the uncertainty and, once we have all been through this in the real-time scenario of every single registration and returning officer delivering an election, rather than the two-thirds that we had this time, that is where the real stress test will come on the ability of the sector to manage.
The other point that I would flag on that is that it is not just about local authorities; it is about the electoral management software suppliers, the Royal Mail, the printers and the knock-on chain that comes from that, which is a learning exercise for everybody involved in the process.
Q186 Mary Robinson: This is going to be a difficult one, or it may not be: how would you rate the impact of the changes that are going to be brought in, such as overseas voters and so on? How would you rate them in terms of challenge?
Peter Stanyon: I would rate them all coming together as a bigger challenge than voter ID.
Graham Farrant: That is sensible.
Miranda Cannon: Yes.
Chair: Thank you all very much for coming this afternoon to give evidence to us about the challenges that you have already faced and dealt with in terms of voter ID, and the further challenges to come, which may be even more significant and difficult. Thank you for sharing that with the Committee this afternoon; it is appreciated.
Witnesses: John Pullinger, Ailsa Irvine and Craig Westwood.
Q187 Chair: Good afternoon, and thank you very much for coming. We have witnesses from the Electoral Commission. Would you just like to introduce yourselves and say what your position at the commission is?
John Pullinger: I am John Pullinger. I am chair of the commission.
Craig Westwood: I am Craig Westwood. I am the director of communications, policy and research at the commission.
Ailsa Irvine: I am Ailsa Irvine. I am the director of electoral administration and guidance at the Electoral Commission.
Q188 Chair: While the inquiry is about electoral registration, we do want to ask about voter ID and the experience of the local elections. Hearing what we have done from the registration officers, we ought to probably put on record our thanks to the electoral registration officers and their staff for everything that they did in the recent elections to help make things go as smoothly as they did with all the new challenges that they faced. It would be appropriate for us to do that.
You have oversight of this—or at least you had an initial look. You said 14,000 people were turned away, who you are aware of and have been counted, who did not have the right voter ID when they came to a polling station. Is that a reasonable estimate? Or is it an underestimate of the number who got somewhere near a polling station but were turned away, and certainly of the people who did not even attempt to go to a polling station because they did not have the ID?
John Pullinger: First of all, I would echo your thanks to electoral authority administrators. They did an amazing job. They all started off by saying it went smoothly, and that is our experience, from what we have seen from local authorities around the country.
We wanted to get data out quickly. We got it out last week, but it is still work in progress. These were initial findings, so everything that we are about to say is, inevitably, provisional. First, in terms of the number of 14,000 that you have mentioned and that we mention in our report, we are saying “at least”, because those are the people who got as far as the polling station. They may have arrived and turned back, or been turned back by greeters, but they got to the polling station and ultimately did not vote, so those people definitely were deterred, but that is only part of the story.
There are two other pieces of evidence that we have put in our report, and there is much more to come. The first is relating to the voter authority certificate. Local authority colleagues have spoken to you about that. Before the elections, we had estimated that between 250,000 and 300,000 people who would have been likely to vote but did not have ID were out there, yet the data that we have so far suggests that only 25,000 voter authority certificates were used, so what happened to the rest of those is a question that we need to try to find answers to.
We think we have some of the answers in our final bit of research that we have published so far, which is a reasonably large-scale public opinion survey that we did immediately after the elections, where we asked people whether they voted. Of people who did not vote, 4% gave us an unprompted reason, which was voter ID. We are trying to unpack exactly what that tells us, but we want to understand in the further data that we collect—particularly data from returning officers but also from candidates and parties, and from charities and civil society organisations that we are working with—to understand which communities were most challenged and most likely to be deterred from voting.
From the evidence that we have at the moment, for example, disabled voters seem to have been more affected than some others. We are talking to disability organisations and gathering more evidence, so we will be able to give you more precision on that in September. Unemployed people were another group that we have some concerns about, as well as some ethnic and minority communities—we heard about language problems from our colleague in Leicester just now—so we want to get under the skin of that. We are giving you a first sight of it. You asked about 14,000, but that is only part of the story.
Q189 Chair: They are the ones who got to the polling desk and were told no.
John Pullinger: Yes.
Q190 Chair: And they did not come back.
John Pullinger: They did not come back.
Q191 Chair: That is helpful to understand. In terms of your timetable now for the results of your further surveys and investigations, will that give us a different number?
John Pullinger: That will give us the extra pieces of evidence. I would be concerned about going on any one piece of evidence. Sometimes you have to build a picture from different perspectives. As I said, I want to get the perspective of candidates, because, certainly at the count, I spoke to many candidates and they had a view on what was happening to their supporters. We do a poll of all candidates and agents to get their perspective, which we have not heard at all yet. We want to get a much deeper understanding from returning officers and their staff about their experience and what they saw.
We also want to test some of these bald numbers. We have heard already that it was quite a challenge to collect the data effectively. We still need to test that and improve it.
The other group that I am especially keen to hear from is civil society organisations, which are working with different communities who we knew during the campaign were struggling. From the testimony that you have just heard, it was definitely harder to reach some of those communities who are less likely to have ID already and to make sure that they knew what to do and came along prepared to do it.
Q192 Chair: Would they be communities that might generally have a low propensity to vote, or even to register, anyway?
John Pullinger: We need to work with that. Craig might want to add something, because he leads on the campaigns in terms of how we reach those voters, but when we were looking at how many people we thought would be likely to vote but did not have ID, we had a wealth of evidence that helped us to work out broadly the size of that number. We estimated that it was between 250,000 and 300,000. That takes into account that some groups are less likely to vote anyway, but it does need more unpicking. This is our first cut. There is more to come. Would you like Craig to add a little?
Q193 Chair: We will come to the information campaign in just a second, if that is okay, and bring Craig in then. I have a couple of follow-ups. In terms of voter ID, there were very specific forms of ID that could be used. Do you think it ought to be widened? If so, what other forms of ID would you be happy to see introduced that would still be as valid as the forms of ID that were acceptable at these elections?
John Pullinger: From the experiences that we have heard about already—you have had some of it mentioned today—some people came along and produced their work ID, which had their photograph on it, and were surprised to be turned away. Examples included police officers with warrant cards and NHS staff with their cards. When the Bill was going through Parliament, we worked with parliamentarians to think about the list.
I would particularly refer to the amendment that Lord Willetts put down in the other place, which had a longer list, which was not agreed to by this House in the end. That had looked at other alternative forms of ID that would meet the criteria of something that would be pretty strong and that would identify that person’s identity with the form of ID they were presenting.
Q194 Chair: Did all of these new arrangements then deal with the problems that you had identified with previous elections?
John Pullinger: When the policy came forward, we had identified three criteria that would make it a success. The first is whether it would improve the security of the vote. That is what voter ID is intended to do. It means you have that link between the person turning up at the polling station and them demonstrating their identity. For it to be a success, it also needed to meet this test of workability. Could it be something that could be safely delivered by local authorities?
Colleagues have already talked about how they made it work, but my assessment is that they made it work by the skin of their teeth. We heard risk mentioned several times, in terms of that building up, so we need to worry about that in terms of judging the success. It is only successful if it is accessible and people are not disenfranchised as a result. One person turned away is one too many. We still need to work at those two aspects of it.
Q195 Chair: That was at a fairly low-turnout local election, where the people who vote may have been the ones who were most aware of electoral process and procedures. Do you think it can work and that the skin of the teeth may not fall away and be laid bare at a general election?
John Pullinger: It will be harder. It will be harder because local authorities will all be doing it at the same time; they will not be able to help each other out, as they did this time. It will be harder because you have a different group of people who are likely to turn out, and people who only vote for general elections are a different group to those who vote in council elections. There will be extra work to reach those people. It will be harder because we are trying to get into areas that have not yet had to do this. We have only been in England. We have not had all of England. We have not had London. We have not had Scotland or Wales yet.
It will be harder for the reasons that came up in the end of the last evidence session: that we are layering this on top of other changes in the Elections Act that are due to be coming into force around the same time. I concur with local authority colleagues that, of those changes, this is one of the easier ones, in the sense that we had enough advanced notice of it.
The other thing with the general election that makes it harder is you cannot plan because you do not know when it is, and you have to do all your work in a very short timescale, including finding premises, recruiting staff—all the very practical things we have heard from our colleagues from local government.
Q196 Chair: What are you saying to Ministers officially about all of these other changes? Are they deliverable?
John Pullinger: We are collecting the evidence. We are giving you a first sense that it has gone smoothly, but we have concerns in the area of deliverability and those sections of the community who struggled to meet the requirements. We will come up with a report in September that has gathered all of the evidence. We will be able to give you a clearer perspective then.
I am confident in saying to you now that the risks are increasing, as they have been for some time. I do not know whether you would like Ailsa Irvine to add a little, because she has been working on this for many years and has been seeing this stretching of local government resources around election time growing over many years.
Q197 Chair: When you produce that report, are you going to look at the changes coming up and how they might be added on to the existing challenges, successfully or not successfully?
John Pullinger: The report will come forward with recommendations as to what needs to happen. That will obviously bring into play the context of the further changes and what we know about them by that point. As Peter Stanyon said just now, there is still some uncertainty about precisely what they are going to be and when they are there. We will say as much as we can in September, which, in terms of parliamentary time, is not all that far away from now.
Q198 Mary Robinson: I am sure it will be covered in more detail later, but just to clarify, you were speaking about people who are more likely to vote. In order to ascertain who those people are, do you look at the crucial figure, which is the people who do not vote regularly or do not vote at all? For instance, anecdotally, 75% of my constituency turned out to vote at the last general election. That is high, but it is still 25%—a quarter—who did not. Where is that relevant in this discussion around voter ID?
John Pullinger: In evidence we have at the moment, we have just been looking at what happened on 4 May in those local authorities in England that voted. We are focused on that. In building the bigger picture, we are drawing on the evidence we have collected from general elections and elsewhere. Craig is probably the better person to give you exactly how we triangulate that evidence so we understand exactly the differences that you are trying to make sense of.
Craig Westwood: At a national level, for the communications work that we are going to do, the most important thing for us to understand is the demographic breakdown, because that is how we are often able to target communications and make sure that we are using the right channels, whether that is through paid advertising, through using the media and press coverage, and also increasingly through the civil society organisations that John has been talking about, working in partnership with them and using them as a trusted intermediary to reach other audiences, sometimes those who are harder to reach, who are more sceptical about politics or about establishment organisations, including the commission. We use third-party organisations to reach them. By doing survey work to understand who has been voting at a particular set of elections, who has not, the reasons why and the demographic data behind those, that enables us to see who we need to target next time.
Q199 Mary Robinson: I can look at the marked register and I can see exactly who has not voted in previous years. It is there. Do you look at that?
Craig Westwood: We do not at an individual local authority level. It would be too much of a challenge for us to do, particularly because the electoral register does not have any of that demographic data on it. You would need to look beyond that to understand what the characteristics of these people were. What we are looking at is how you do an advertising campaign nationally that is going to reach broad audiences.
Q200 Mary Robinson: I am being very simplistic about this, but if I had 75% of people and I had the ability to get in touch with them directly, I would be getting in touch with them and asking them or sending them information. Is that not something that you could actually do: target people directly?
Craig Westwood: One of the main partnership networks that we have is local authorities. They have talked about the resources that we give them to support their communications activity. They then have it in their gift to engage their local communities and to really understand what is going on in the local area. It is not really something that we could do at a national level.
Q201 Mary Robinson: I am sorry, but I am going to push this point a little bit. I touched on it previously, but the answer I got back was that that data was not shared or able to be used on an individual basis, if I am correct in interpreting it. Can a local authority get in touch with Mr Smith and say, “We would like to know more about the reason why you vote or do not vote,” and ask for that information and follow up on it?
Craig Westwood: I am not aware of the specific restrictions that are placed on them in how they use the register.
Q202 Mary Robinson: It just seems to me to be an obvious one—to ask people.
Ailsa Irvine: There are limitations around how the marked register can be used, who can access it and for what purpose. I would have to look to see whether that is a purpose that would be permissible or not, but the legislation, as with many aspects of electoral law, is incredibly prescriptive and does not give a lot of flexibility. It may well be that that the complexity of our electoral law framework is another barrier to being able to engage in the way people might want to.
Mary Robinson: I am sure there are all sorts of reasons for it. It just seems like an obvious thing to do. Thank you very much for your answers.
Q203 Ian Byrne: You might have covered a little bit of this, Craig, but data from polling stations showed that 0.7% of voters were turned away because they did not bring the required ID in the first instance. Why do we think that the information campaign did not reach that 0.7%?
Craig Westwood: What we know, which Peter Stanyon referred to it in his earlier evidence, is that in the four months that was available to us we managed to get public awareness up from 22% to 87%, which for us is a massive achievement. It obviously leaves 13% that we were unable to reach. What we do not know is the extent to which that 13% overlapped with people who were not interested in voting, were never going to vote anyway or people who actually needed to get the message because they had the intention to vote.
Q204 Ian Byrne: These people were turned away. They had an intent to vote.
Craig Westwood: Yes. What we need to do is look at that data and more fully understand it. As well as the people who were specifically turned away, we are concerned about that 4% that John Pullinger talked about earlier, who were people who were saying that the reason that they did not vote was because of the policy, because they may have not left the house or been turned away earlier, decided not to go to the polling station or would pick up one of the other forms of voting. That is the audience that we are particularly interested to understand, to see what the demographic breakdown of that is and how we can deliver more to reach those people.
We are very comfortable that we did a really good job at reaching the people in the time available. It was only four months. We would have liked much longer to engage, particularly some of those hard-to-reach audiences, but in no way are we complacent about that.
We want to double-down on the communications. We are continuing to work, particularly with those civil society partners and charities, to really drill into those organisations. In our work with those charities, from one of the partners, we did hear the feedback from them quite early in that process that actually, for them, the local authority elections were just too close, and that their focus would be on the general election, because they understood that for their community it needed much more time to bed in, for the messages to stick.
Q205 Ian Byrne: That is a very good point you make. I congratulate you on getting the 87%, which is really good, but is there a fear? We are all talking about local elections here, with low turnout and the ability to manage the situation. We listened to Peter before and undoubtedly there is a real worry about the general election. Are you worried about the impact in terms of a chilling effect on democracy and voter turnout in a general election? Is that something you are concerned about?
Craig Westwood: We do not have the turnout figures yet. Obviously, we have heard from some returning officer colleagues about what the experience has been at their local level. We have some anecdotal evidence from individual authorities; we do not have a sense of the overall picture.
Q206 Ian Byrne: By September, would that potentially be looked at and you could give us the answer? That is a real concern. Obviously, we do not know how many people were just turned off by the whole process of voter ID. It is a concern for us all, isn’t it? How did you target your information campaign to reach those who are least likely to have the necessary ID to vote? It was a success to a degree, wasn’t it?
Craig Westwood: Yes. What we did was base it on evidence, which is how we do all of our work. We did detailed work of our own, and also the Government had done studies to look at who demographically was likely to possess the eligible forms of ID. Then, by looking at that, you could see which groups we knew were going to need much more targeting. Then we assessed which of those groups we could effectively reach through mass market advertising. For example, the unemployed, people from lower social grades and lower educational attainment were all groups that we assessed we could still reach through the paid advertising campaign.
Then we looked beyond that to other demographics who we knew particularly faced multiple and compounding barriers to engagement with the democratic process as a whole, who would need extra support. Some examples of those would be the over-85s, who are, to your earlier question, more likely to be registered and to vote, but much less likely to possess ID in order to overcome this barrier. The trans and non-binary community, and people with certain disabilities, particularly visual impairments and learning disabilities, are a few examples of the groups that we were looking at.
Some of them are relatively large. Some of them are very small demographics, but actually very important for us to reach to make sure that the process is not having a punitive impact. We therefore established charity partnerships to co-create resources. We really worked in ways that we have never worked with partners before, to really reach deep into those audiences, to make sure that we were speaking in the right way to them in messages and languages that they could understand and that would reach them. It was just a limitation of time: we only had four months to do it when we would have liked much longer.
Q207 Ian Byrne: What would you need for a general election to get that level of effectiveness time-wise?
Craig Westwood: For England, much of that has started now, because those partnerships are in place and, obviously, we have already started bedding in with some of the audiences. We know that there were no local elections this time in London, but that awareness of the policy was almost at the same level as in the areas where elections were. We have already gained a certain amount of distance for London.
The key question for me, and the challenge we have, is about Scotland and Wales, where voters will only be having to show voter ID for certain elections—for those that are reserved—and where there is not that advanced work. We are obviously getting in ahead of time, making sure that we are building up those charity relationships. As John mentioned earlier, we will not know the date until seven or eight weeks out from it.
Q208 Ian Byrne: What work are you doing around voter ID for postal votes?
Craig Westwood: In terms of our understanding, it is the requirement, as was mentioned earlier, to show your national insurance number at the point of the application. People will have already had to have done that to register. Apart from the administrative questions that were raised earlier, we do not see that as a particularly new barrier for the voter.
Q209 Ian Byrne: Lastly, on voter authority certificates, your analysis shows that over 40% of people turned away from polling stations were completely unaware of the potential of getting a certificate. Is this a significant problem? How are we going to overcome that?
Craig Westwood: Yes, that is the main challenge that we need to look at. As Miranda mentioned in the last evidence session, there were some really complex messages that needed to be got across there. The message to bring ID to the polling station is a relatively simple one. We saw that work for the broad range of people across the election. For those audiences that needed to get the message about the voter authority certificate, it is a very complex message for some people who are more likely, in a number of categories, to be less engaged with politics and less trustworthy of authority figures. That message does need that additional time to penetrate. We are doubling down on that activity to make sure for the next elections.
Q210 Ian Byrne: I suppose that is the work you have been doing with different partners, with different ways of doing things.
Craig Westwood: Yes, exactly, and learning from the experiences that local authorities have had for this set of elections in engaging their communities. They obviously know their local communities better than we ever could. They will be able to share that best practice with their other local authority colleagues for the next set of elections.
Q211 Ben Everitt: Peter is getting quoted a lot today. I can see he is still at the back; thank you, Peter, for staying. He has previously said that the electoral registration system was creaking under the pressure and being held together with gaffer tape. Do you share that assessment, John?
John Pullinger: Simply, yes. The current electoral registration system is neither effective nor efficient. It does not work for voters. It does not work for parties. It does not work for electoral administrators. We have a situation where millions of people are left off, and that is not good enough. We need to do something quite radical to improve that and to automate the process in quite a significant way, so that people are not left off in the way that they have been increasingly over the last quite a few years now.
Q212 Ben Everitt: Do any options spring to mind?
John Pullinger: I will ask Ailsa to add to what I say, but we have done a number of feasibility studies and pilots that certainly demonstrate to me quite clearly a much more automated system is entirely possible. It is the norm in many countries that achieve both much higher levels of accuracy and completeness in their registers. There is lots of good practice to work on. A step-by-step approach, with the ambition of making sure everybody who could be registered is registered, seems to be entirely within our grasp. If you would like, I am sure Ailsa can say quite a bit more on what we have done already that has given us that kind of confidence and the kind of people we have already worked with to test out the feasibility of it.
Ailsa Irvine: In looking at the feasibility studies, we looked at a range of different options, really focused on how we can use public data to keep registers up to date throughout the year, to not be relying on event-led registrations in the same way as we are currently. There is a range within there that we have looked at. This is not necessarily an either/or. They could potentially even be steps on a journey towards a greater system of automation and automaticity.
We looked at integrating it into other public service transactions. For example, we know each year the DVLA deals with 60 million customer transactions. Actually, if you could look at sharing some of that information with EROs, to use that to contact electors, you could use it to populate a registration application. You could integrate the experience for the elector, so they could simply say, “Yes, I would like that information passed to the electoral registration officer to update my registration details.” There are ways that you could use that type of integration to target one of the groups who we know are traditionally under-registered, which is home-movers.
You could also look at a form of automatic registration. For example, one of the areas that we know is a significant challenge with getting registered is attainers—those who are just about to reach voting age. We know that has continued to drop off since the introduction of individual electoral registration. Our last estimate in 2018 was that only 25% of that age group are registered. You could look at a form of automatic registrations whereby, with the data that the HMRC holds when it issues the national insurance number, you could look at some form of automatic registration there.
These are all things that we explored the practicability of. We spoke to the different data-holding organisations. We have identified that these are reforms that could be implemented without a radical overhaul of the fundamental structure of how electoral registrations run in the UK. We have the individual registration and digital service, which already provides a hub and a way of data coming in from voters, being checked with the DWP and going back to electoral registration officers. There is a platform there that could be built on without fundamentally reforming the system.
Q213 Ben Everitt: It sounds like it would be a hell of a back end on the IT system linking all of these Government data points together. That is not to say it would not be worth doing, but it does raise the issue around data security, which leads me to a question about the open register. Again, we have a quote from Peter. He says that not having restrictions on what data can be used for “flies in the face of everything else around data”, which is pretty snappy. Essentially, we are in a position now where businesses or potentially agents of foreign Governments can buy the open register, which is details of people’s registration to vote. If we make it even bigger and more comprehensive, using the data that you have described, that is probably expending an existing risk to privacy rights. What are your views on having an open register, John?
John Pullinger: We should abolish it. It is not serving a useful purpose. I will stop there.
Ben Everitt: Brevity is a virtue.
Ailsa Irvine: We have consistently recommended against the open register since it was first introduced in 2002.
Q214 Chair: You have talked about automatic registration or partial automatic registration, where you can use certain bits of information, such as with attainers from DWP and so on, to get automatic registration for a group. What about completely automatic registration? Have you looked at that as well? Is that achievable or feasible?
John Pullinger: It is achievable and feasible, but you would not do it in a big bang way. You would build up to it and each step would be a step in the right direction. I would set that as an ambition, because there are a lot of practical things that need to be addressed. Although I gave a very simple answer to Mr Everitt just now, his point about cyber-security is well made. The more you automate these systems, the more you have to think very hard about that. We are in a situation where we are under threat from different people for different reasons. This is a data set that could be of interest to people. We need to work very hard with relevant experts on that to make sure we are building cyber-security in at every stage.
As an ambition, we have to make sure that this is something that is straightforward in the vast majority of cases. It is not going to be totally automated. Ailsa’s example was a good one: you have an automated system that has a checkpoint where someone can say, “Do you want us to use this information to be added to the register?”—for example, when you are having your driving licence updated when you move house. The system is automatic, but it has points at which the electoral registration officer or the individual can come in and say, “Yes, this is correct,” or, “This is not correct,” because the issue of accuracy is very important.
Some of the concerns about an automated process are well made. If you just have no human intervention, accuracy could be put at risk. I just maintain that many other countries have done this very successfully, and we can learn from that experience from a technological point of view so that we build a system that will then work for people and administrators as well.
Q215 Chair: These would be systems that would flag up the information to the electoral registration officer, who then checks with the individual potential voter whether it is accurate, and then puts them on the list.
John Pullinger: Yes, potentially. It is into the workflow. We have solved this problem for passports, for example. We have solved it each year when we do our car insurance, where they check we have an MOT. That would be the kind of system you could envisage for electoral registration, because it is not inherently complicated in the sense of the number of criteria that you have. You do have to check for eligibility and you have to check other aspects of accuracy, such as the addresses, particularly for people like students or people in private rented accommodation who are moving around. There has to be a human intervention potential. The amount of resource you could save by automating the simple cases would be significant.
Q216 Chair: Why has the Electoral Commission not then set out the framework you would like to see, where you would like to end up and how you gradually move towards that process? Is that not what we ought to have from you?
John Pullinger: Ailsa, can you explain what we have set out through the feasibility work that we have done?
Ailsa Irvine: Feasibility studies have drawn out some of these angles of complexity around trying to find the data sets that would actually have all the information that you would need to be able to register someone to vote. There is not really a single data set that would enable you to do mass automatic registration on its own. It would require the complexity of bringing a number of data sets together. We have drawn on examples like Australia, where it is exactly as John sets out, where there is an automatic element to the system, but the voter has to confirm before they are then added to the register. We have set out a lot of that in the feasibility studies. We completed those and published that in 2019.
What we are now doing is refreshing those feasibility studies. We are looking at that now, at the same time as we are carrying out and completing a piece of research and updating our accuracy and completeness research. By later this year, we will be in a stronger position to understand what the latest picture on the registers are, whether the patterns of under-registration remain as we have understood them to this point, and how the different solutions could come together. There is an opportunity as we move forward over this year to pull a lot more of that together.
Q217 Chair: You know the sorts of data that you can collect fairly easily to achieve some significant improvement in registration. Have you set out how you might want to gradually incorporate that into the system, if you do not want a big bang?
Ailsa Irvine: No, probably not as explicitly as you have just outlined it. What we have really set out is that you do not need to go straight to automatic registration. There are different staging points that you could go to on the way, but we have not set out a preferred road map to this stage.
Q218 Chair: Would that not be something useful for you to do?
John Pullinger: It is certainly something where we could lead the process, but we cannot do this in isolation. In terms of some of the challenges of delivery, we would want to do this jointly with our local authority colleagues. On the challenges with policy, we need to work with the Government, the Houses of Parliament and the devolved Administrations as well. We have not talked about the issues of challenges for electoral administration because of diversion processes. A solution would not be an Electoral Commission road map: it would be an electoral system road map.
Q219 Chair: Who would lead on that, because currently it is not happening?
John Pullinger: Currently it is not happening. We could convene such a discussion to make it happen.
Chair: With the electoral administrators and the key people involved.
John Pullinger: Of course, yes.
Q220 Chair: Together you could actually sit down and do something.
John Pullinger: We could, yes. I would not exclude the political parties from this as well, because it is central to the whole way campaigning is organised. We could create that framework to move forward, yes.
Chair: We will move on to the engagement with under-registered groups, which is a big issue around the current system.
Q221 Ian Byrne: It is a fascinating conversation. Do local authorities have sufficient resources to ensure that all individuals from different backgrounds receive adequate support when registering to vote, John?
John Pullinger: The simple answer has to be no. Ailsa has already mentioned that the 16 to 17-year-olds are the group least likely to be on the register. That has got markedly worse since we have had individual registration. There is a challenge to get that community to understand the value of being registered and the process of being registered. Our campaigns are deliberately targeted at them, but there is a bigger question around education, working with schools and other colleges.
Craig is the expert on this. He has led on this and has been trying to get very innovative in the way that we think about this. He has also been working very closely with the devolved Administrations, where of course we have voting at 16. We have learned a lot from the experience they have had in Scotland and Wales with that younger age group.
Craig Westwood: One of the main things we do when we are planning our public awareness campaigns ahead of elections is to think about how we can best support and back up the work that the local authorities are doing at a local level. It is about really thinking about the resources that we can provide them, so that we are creating them once and they can then effectively get out and spend as much of their activity and time as they can on the frontline, engaging with communities, using the right resources that we have provided them. That is always a factor, to make sure that we are making best use of their time.
In terms of the education work that we are doing, we stepped back into doing democratic engagement resources about four years ago, because we were hearing from a lot of citizenship teachers that they were crying out for something from a dependable, independent source that they could trust as being utterly impartial in terms of the materials, because they are obviously concerned about engaging on politics in the classroom.
We have worked with the Association for Citizenship Teaching to make sure that we are producing the right resources that work in each part of the UK on the topics that are most of interest. That started with the nuts and bolts of the electoral process—so really helping young people to understand what it is they are voting for, why it matters and how the process works. We will then continue to expand the range of subjects.
Q222 Ian Byrne: Do you have any outcomes from that work you have done over the last four years?
Craig Westwood: We are undergoing an evaluation at the moment. What we do know is that we are having a significant rate of pick-up of schools that are using the activities, particularly around things like National Democracy Week, where we have had hundreds of schools across the country that have engaged with us, used our resources and particularly used it to drive registration of young people within the classes.
Q223 Ian Byrne: Do you have any concern that the registration process for attainers was linked to individuals’ national insurance numbers?
Ailsa Irvine: That was one of the recommendations that we have made: that you should be exploring automatic registration at the same time as someone is issued with a national insurance number. It would address that key area of under-registration with attainers.
Q224 Chair: At one time the annual canvass was the great event, was it not, in the registration year? Local authorities employed people to go and knock on doors. I always wondered how effective it was, and it probably got increasingly less effective as you knocked on the doors of the people who did not tend to register themselves—they were the ones, in HMOs, where you would not find them in. There were probably people in the BAME community who did not have English as a first language. Increasingly, there were probably some concerns about safety as well. People did the job in the evening when people were likely to be in, but it was dark. Is it now a defunct process?
John Pullinger: I am sure Ailsa will add something. Improvements have been made recently, which have helped, certainly from an administration point of view. Frankly, it is tinkering compared with the challenges that we have talked about and the levels of under-registration that we have. It is not shifting the dial on that in any way. Ailsa, do you want to add a little bit about what we have seen?
Ailsa Irvine: In the absence of these more fundamental reforms to the system that we would like to see, the annual canvass still plays an important role in helping to audit the register and helping to limit the event-led registration and the pressure that puts on the system by people applying close to elections. It does still play an important role. As John said, it was reformed back in 2020. We have had positive feedback from administrators about the lesser burden it puts on running the process, because it does enable them to target the resources at those properties where they understand, from the data available to them, that there are more likely to have been changes in household composition.
However, what we are seeing is that it does not seem to be keeping pace with population movement. We will not know for sure the impact on the accuracy and completeness of the registers until we complete our research, which we will be reporting on in the autumn of this year. We can see from the numbers and the data that we collect from local authorities each year that it is not picking up all the volumes of changes. For example, we know that of those households where they have identified that there is likely to be a change, a third of those households are not responding to the annual canvass. Therefore, those are changes that we think need to be made, but people are not responding to and falling through. It is unlikely, then, that the registers are as accurate and complete as we want them to be.
Q225 Chair: They are getting worse, effectively.
Ailsa Irvine: We will know that when we see our accuracy and completeness research in the autumn, but there is a concern at the moment that the system is not keeping pace with the movement.
Q226 Chair: The concern is that they are getting worse.
Ailsa Irvine: We shall see shortly. That is why we are doing the research. We really want to understand the picture.
Q227 Chair: You are doing the research because you have a concern that it may be getting worse. We will get all that information in your further report in the autumn, which will look at those issues. You will also cover, when you do your report on the implementation of voter ID, your analysis of the future changes that are coming. Will you be indicating to the Government the necessary timeframe for them to bring in secondary legislation for the registration officers to be able to implement those changes in time for the general election in particular?
John Pullinger: I echo the comments that Peter Stanyon made earlier. The Gould principle of having six months between knowing what you need to do and actually doing it is something we really need to keep to. We did not for voter ID. I have used the words, “We got through it by the skin of our teeth,” and that is true. In terms of the changes that are coming, we really do need to be quite strong in maintaining enough time to train people and make sure we have a consistent approach across all parts of the UK in implementing whatever changes are coming forward. We need to be quite tough on that.
Q228 Chair: This will be in the report you are doing following on from your analysis of the voter ID implementation.
John Pullinger: We will be and are saying this now. In terms of any further changes coming down the track, including any changes that might come forward relating to voter ID—you have asked us about forms of ID—that is a discussion that Parliament needs to have. Whatever changes we choose to make, there needs to be not just secondary legislation coming out but the guidance coming out and everything being ready, so that local authorities can plan with confidence. It is uncertainty that is the real enemy of this. If you know what you can do, you can plan for it.
Q229 Chair: It is still going to be a massive challenge, is it not?
John Pullinger: Absolutely, yes. Even with that it is going to be a massive challenge.
Q230 Chair: You said the voter ID changes are probably the easiest of some of them to bring in.
John Pullinger: Yes, and they are more certain than some of them. With overseas voters, the scale of it and the nature of the problem is going to be unknown until we are in it.
Q231 Chair: Are there any other issues you would like to raise with us while you are here today?
John Pullinger: There is only one, if I may. All of the changes we are talking about are adding complexity to a system that is vastly overcomplicated already. As many parliamentary reports have indicated, the fundamental problem is that we have yet to consolidate electoral law on to a platform that makes sense. It is really hard for administrators and parties to understand what the rules are by which they need to operate. The Law Commissions of England, Wales and Scotland have come up with a very clear prospectus that would achieve this, which we believe is something that could be implemented relatively rapidly. Once it is, then it is possible to build future changes on with much more confidence that the system will not be subject to the kind of risks that you have heard today.
Chair: On that note—I do not know if that is an encouraging note or a concerning note; it is probably both—thank you very much indeed, all of you, for coming to give evidence to us about what are obviously some really important issues. It is about our democracy and how we elect our Governments at national and local levels. It is an important issue. Thank you for your advice, information and comments about how things might be taken forward and improved in the future.