Good afternoon and welcome. Starting from my left, could you all introduce yourselves for the record?

I am Detective Chief Inspector Kevin Ives. I am the national co-ordinator for the election crime portfolio, which focuses mainly on fraud offences and Representation of the People Act offences. Nik Adams is the chief officer lead on that, and he apologises that he cannot be here; I am here in his stead.

Thank you for coming.

I am Peter Stanyon, chief executive of the Association of Electoral Administrators.

I am Vijay Rangarajan, chief executive of the Electoral Commission.

I am Niki Nixon, director of communications at the Electoral Commission.

That’s great. Let’s start straight off. The first question is from Jessica.

I have a very general question to begin. What main lessons do you all think came out of the 2024 general election? How do you think it compares with other elections that you have been through?

We published a report after the 2024 election taking account of as much data as we could lay our hands on about what had happened. That included surveys of members of the public, electoral administrators, and of course candidates. That was a core part of the information; we also worked a lot with the academic community to bring all that information together. What were the main lessons? Abuse and intimidation were at high, unacceptable levels. The core data we have is that 55% of candidates felt they had a problem with harassment; women were twice as likely to experience harassment, and ethnic minority candidates three times as likely. That is definitely affecting candidate behaviour; a number of people changed the way they campaigned. Fundamentally, that is one of our core concerns, because it is impoverishing some of the interactions between voters and candidates, and making candidates unable to interact in the way that they want to. It is also clearly beginning to put people off standing for elected office, which is another real long-term threat to our democracy. In addition, it is incredibly personally upsetting and disruptive for the candidates themselves. It is just awful. In the evidence we gave to you, Mr Speaker, and the Committee we published some of the quotes we received in our qualitative work. You can see the sort of effect it has had on people and, even worse, on their families, colleagues, volunteers and so on. I have two final comments. It has the potential to get worse. On the front page of the evidence we submitted to you, we put a graph showing that young people found it more acceptable than older people to abuse and intimidate politicians. The differential grows with age. We think there is a number of things that we need to do, from candidate addresses through to a really serious programme of education, as well as online screening. There is a very interesting online-real life resonance going on: sometimes the harassment is happening to be filmed and then put online, and sometimes the online abuse is then generating offline behaviour. That is difficult. Niki is going to add a little more, if that is okay.

We see, from the experiences that people have had, that the abuse is most definitely crossing a line. We have seen masked men interrupting meetings, hate mail sent to people’s children, and people’s offices being attacked. There is no doubt that it has crossed a line, and as Vijay said, it is starting to affect how people are campaigning: for instance, people no longer want to campaign alone. They are censoring themselves, they do not want to share their views, because they are concerned about what the repercussions will be.

I would widen it out slightly. I support everything that Vijay and Niki have said, but I highlight the fact that, because it was a truly snap election, everyone in the administrative community was slightly on the back foot from the outset. As we went through, it became very clear that there was a higher level of mistrust in the system and how we deliver an election. By widening it, I mean that there is a lot of, not necessarily intimidation, but a lack of understanding about there being deadlines for deadline reasons—you have to have them for this—and about the lack of power of returning officers to address issues that arise. There is no power of investigation for returning officers; they simply pass these sorts of thing on to colleagues to deal with. It was a strange situation; it was a successfully delivered election in the circumstances, but the things that were going on around the outside highlighted the fact that there has been a change in public perception of access to the system. I also highlight that the fact that returning officers are not quite in the place that many people think they are, because they have no power of investigation.

For policing, yes, it was a snap election, but we had our networks in place; we had trained people in each force. What we definitely could do is drive consistency more. There is no doubt about that. With the UK policing model of the 43 forces, if you have an incident at a hustings or you have canvassers being approached in the street, a lot of the time the officers turning up to deal with these incidents are frontline officers and not the trained experts. We certainly come out of this looking for ways to drive more consistency across the country. Some areas are getting a gold-star service; others do not feel the same way. We have some good ideas for that. We certainly take that away from this.

You all clearly have a role, as police, electoral administrators and the Electoral Commission, in dealing with intimidation and other electoral crimes. How do you all collaborate?

Personally, I have routine meetings with the Electoral Commission. We are on first name terms. We meet several times a month. We have discussions about issues and routinely discuss individual incidents. As national co-ordinator, I collect information on the things that happen around the country and get the right advice to the right forces. We are speaking on a weekly basis, generally speaking.

Can you explain how the police nationally work with the Operation Bridger contacts in each of the local forces?

That is no doubt confusing from the outside, and that is something that we can talk about as well. I believe the Met police are coming as well. We have the electoral crime portfolio. That is Representation of the People Act offences, electoral crime. There is clear crossover there, because things like the way candidates behave and interact with each other and some of the posters that go up cross over into harassment and intimidation. We have our own single points of contact; they are usually economic crime officers and there are at least two in each force in the country. They are trained by City of London Police crime academy. They are in place to advise and make sure those crimes are dealt with properly. Operation Bridger, as you all know so I won’t go over that, is there for MPs. With Operation Ford, the new force elected-official advisers are there for every elected official. We speak routinely, more than once a month. We are all in the same meeting, in the same room.

How do you get a more consistent approach? From speaking to colleagues, I know there have been instances of different approaches in the same force. I remember one where the police were very good about what was happening at a meeting, and others where they just sat in the van and let intimidation and whatever go on, and did absolutely nothing. Another issue is police sometimes treat things as individual incidents rather than as a pattern of behaviour. Individual incidents might not cross a particular line, but if you look at it as repeated behaviour as a whole, it is clearly being done to intimidate.

I can tell you some of the things that we did without trying to give the impression that it worked perfectly, because that is not what we are coming away with. We have an authorised professional protocol. It is on the College of Policing website, and all officers should refer to that. It was updated—we finished updating it just about a month before the election. We did that alongside the College of Policing. So the APP was in place. We filmed a video that we sent out to every police force to put on the front page of their intranets telling officers why policing the election was important and that they should show zero tolerance, setting out what the offences are and where they should look for guidance, and saying what Op Bridger and the FEOAs would be. We sent out ChiefsNet letters—that is the online system where chief constables can send each other messages. Those went out to make sure that there was proper response in place for polling stations, because there was some intelligence beforehand for that, and also to make sure that election calls were prioritised. Those went out. We also have the SPOCs in place and they should drive consistency. As I said, the gap here is that they are not there on the streets when the canvassers have called the police because they have just been chased off. We need to do more to get that into the minds and hands of the frontline officers who are turning up first. We ran a debrief after this election just for policing. We have come away with recommendations. We need to do more of these, because we need all the SPOCs talking to each other to make sure that we are not missing those linked offences.

The SPOCs were brought in as part of Operation Bridger to look after MPs.

My apologies. These are the electoral crime SPOCs.

Right. Not the SPOCs—

I agree that it is confusing. We need to explain better.

That is good. Just to back up Mark's point, we have got a big constabulary area where one end is doing something different from the other. It is about trying to ensure that we can get consistency. It is good that that has been taken on board.

We have obviously had changes in that you can have your home address not published on the forms, but many candidates and agents are still publishing their home addresses. Can you shed some light on why addresses are still being published at elections? Would you support removing the option to publish them?

It is the choice of the candidate as things stand. That change was brought in reasonably recently, and I would say the option is taken up more often than not now, although there is always a debate about a candidate showing that tie to an area, which may well be seen as a positive. You would know more about that than I do. One of the areas where there is a huge gap is when an independent candidate acts as their own agent. On the ballot paper they can withhold their address, but on the election agent notice their home address is shown. There are some anomalies at this stage. Whether or not the choice should be given is really for Parliament to decide. It is a straightforward admin process for returning officers solely to withhold those details. They may be withheld, and those addresses are destroyed 21 days after the election, so it is very much a constrained piece of information. It works, but there are still holes in the system at this stage.

Our recommendation is to make a change on this, partly so that no candidate—whether they are their own agent or whatever—is forced to publish their home address. The Welsh Government have made that change already, and we would recommend following suit. You want to prove a local link, but that does not mean publishing your actual home address in any form. We have provided quite a lot of guidance to candidates telling them they do not need to do that, but it can put them in an invidious position of being accused of not having a local link by not publishing their address. We do not want to force them to choose between winning an election and their own security; it can be a dangerous position to put them in.

There is a consequence to that, though, isn’t there? Who is going to do the checking? At the moment, you are not allowed to check that the address is correct, or even that the name given by the candidate is correct. The problem you then have is that the other candidates can say, “Hang on a minute—that’s not right,” because it is not published. What safeguards are you going to put in? I want to protect candidates, but at the same time I want to know that they are the person they say they are and that the address is correct. What are you going to do about that? What is the consequence of doing one without challenging the other?

Can I add a little bit to that, Mr Speaker? It is also about whether it is their actual home address.

That is what I am saying. It is proof of identity and proof of address, and that becomes more distant and harder if nobody else knows. I do not want to know, but I want to know what you are going to do about that. It is a ridiculous situation, because you have to prove who you are to vote, but to stand as a candidate, you do not. That is what we are trying to get across to you.

One of the key elements to it is that there is case law that requires returning officers to take at face value what is presented to them.

Actually, I agree with that.

I gave returning officer training yesterday where we said this. Even if you know it is a false address, there is nothing you can do. It is about power. In terms of the checking mechanisms that might be needed, it will open up a whole new stream. How do you do that checking? How do you check the individual’s name or address that they provide? How can that be done within the election timetable so that it does not delay that side of the elections? One of the things talked about is potentially providing photo ID at the point of nomination. That is straightforward for a parliamentary election, where we are dealing with five or six candidates in a constituency; but if we are trying to keep consistency across elections, I point you to Wiltshire, where 2,100 parish council seats are up for election in May this year. To have consistency all the way through, unless we are going to make a break between parliamentary and local government rules, lots of questions need to be asked about what is the right level to go at. It is an absolutely fair point to say that the checks ought to be there.

To come back a little, the problem is that if you admit you know it is wrong, there is nothing you can do about it; surely the starting point ought to be that if you know the address is wrong, that should become an offence.

The issue comes down to the basis of the case law and the terms we are working to. If local knowledge suggests an individual’s address is not correct, that is the point at which the returning officer will pass it to the SPOC, but they will still run the election; that would be dealt with in a post-election situation.

They do not look into it. What we are saying is that we need something in statute to require proof of identity and proof of address, even in local elections. I do not think that you should be able to claim to be someone you are not. That will be a fair recommendation. We will look to introduce something.

You mentioned destroying the evidence for people’s addresses after 21 days. In a constituency near mine, for example, one of the independent candidates acted as his own agent, so while I appreciate that while you get rid of his address after 21 days, that is scant consolation after he has put tens of thousands of leaflets through people’s doors with his address on them. What steps could we take to reassure people who are standing as independent candidates and probably acting as their own agent that they may do so without having to divulge that information? I agree that it is a loophole, but it is something that might dissuade people from standing even in the local elections coming. It sounds like something that we need to deal with fairly swiftly.

The advice that returning officers and I would give is ultimately to provide an office address, not your home address. That may be more difficult than it actually is because of the rules on election agents’ addresses, but the difficulty is where the home address is the only option available to the candidate. The standard process is for the agent’s notice to go up on the web and come down immediately after the election has finished, but that does not stop the fact that it is already out there in that situation. Leafleting is exactly the same. There is very little that can be done while that loophole allows for the address to be published. Historically, it may be there for a long time, given the nature of data out there.

Can we only close that with legislation?

Only, I think, by saying that a home address does not need to be published on an agent’s notice. It is to do with where the notice for an election goes. The old-fashioned way is to say, “You go to that office, knock on the door and talk about the candidate.” Taking that out of the equation will simplify things a lot more easily, but there still needs to be some route to the candidate through the agent, in a contact way.

Among the suggestions we have seen was one about allowing PO boxes for that purpose. At the moment, that is not allowed; you cannot include a PO box. Is that a suitable alternative, given the issue that has been highlighted?

I do not think that we have a clear view of that. What we would like to see is what I think Mr Speaker was suggesting, which is returning officers being able to check that a candidate is the right person. We are not looking to make this massively burdensome, but we have had allegations, which damage trust in the system, so it is about being able to check that someone has an address there, but they do not have to divulge it. Clearly, if they want to use a PO box for communications, they can do that as well, but trying to check that they are the right person is the core of the issue.

Regarding agents, it is about the communication address that you get hold of the agent through. I would have to rely on the PO boxes as a suitable route to do that, or it might be a web link in modern-day communications. A number of issues that are not directly related to the candidate flow from that, as a result of an independent candidate standing.

Many in this room and previous candidates will have stood multiple times. They might well have divulged their personal address in the past, but even if they choose to make the change, the information is often still available after elections, if you go looking on the internet. Are you aware of that? Are there ways for it to be removed? Would you want to see something to make that a lot easier?

It has changed in my 30 or 40 years on elections. We used to state the person nominated, with the notice of poll put on the website by the council. That raises the issue—I have personal experience of this with a celebrity who stood for election—that it can be tracked through from four or five years previously. The advice or good practice would be that all that information need not be provided. All the public need to be aware of at that stage is who is standing in a particular area and, potentially, what the party moniker is, because those notices are still published at the returning officer’s office, so there is a paper copy that is actually there. That is the best practice way of doing it—limiting the personal detail that goes out as things stand, without actually preventing access to who it is standing in those areas. But once it is out there in the online domain, I have no expertise as to how we would get it taken off.

We all want to see elections run fairly and properly, but then we come on to issues of intimidation, harassment and abuse. We have three organisations—the Electoral Commission, electoral administrators and the police—that can oversee that process to hopefully get it right. How would each of you, individually, define what you think harassment, abuse and intimidation amount to? Who do you think should be dealing with it?

There are some very clear examples of harassment and intimidation, such as threats and clear abuse. There is absolutely no argument or grey area there; they fall totally to the police to deal with. Naturally, during the course of the election, there were other things said and printed that some sides would claim were harassment and should be taken down, while other sides claimed that they were factual statements that should be allowed to remain. It should be legislation that allows the police to make those decisions clearly. One of the issues that we have right now for policing is that the police officer will turn up and be desperate to be independent, to be seen to be independent and not to come down on either side of the argument. The clear issue at this election was the Gaza issue. There were statements that were very clearly crossing the threshold, and other statements where there was an argument. We do need clear guidance for frontline police officers around what falls on each side of that argument. That is difficult, because at the next election there may be a totally different issue and therefore the statements made will be very different. But, as a clear answer to your question, harassment and intimidation is a police matter. It should be a police matter, and it should be dealt with consistently. There should be zero tolerance for that; there is no ambiguity there. That is why the new force elected-official advisers will be in place for all candidates. Hopefully we can learn from this election and give that consistency.

Can there be harassment that is not a criminal offence but still adversely affects the operation of an election, so the police cannot deal with it and somebody else has to?

“Harassment” is the legal term, so if there is harassment, that is breaching the Act and should be dealt with by the police. Are there matters that come to the police where decisions are taken that they are not harassment? Yes, there are, but they can be dealt with in other ways, such as simply talking to and warning other candidates or members of the public. But if it is an offence of harassment—if it meets the Act and meets the definition—it is a police matter.

Let us move on to Peter Stanyon. Do the staff in elections offices—returning officers up and down the country—not face situations in which the police are not going to act, because it is not really criminal, but it still starts to affect how an election operates or the candidates start to feel really concerned or worried about the situation? What do staff do in that situation? Is the guidance for them clear? Is the law clear?

The guidance is as clear as it can be, I think, based on the hard definition of what is harassment and what is intimidation. Again, it very much comes down to building those relationships with the individual agents for the parties. What we say to returning officers is that it is about being seen to be independent regardless of where someone is on the political scale, so that everybody is getting the same information at the same time and access to the system, in terms of that side of things. But the powers available to a returning officer are absolutely limited, and it would be a simple case of passing to colleagues in the police forces to investigate.

You are talking about political parties in a traditional sense. I think we have all heard evidence of campaigns that are not run by traditional political parties, or even political parties at all. Effectively, they have got groups of followers who treat themselves like a football-supporting group and go around behaving like that. They do not follow the normal rules; there is no agent to go to who has any control over them. How do you deal with that situation?

In terms of campaigning, there is very little that returning officers can do other than reporting allegations made to the police. In terms of polling stations and the count, there is the ability, but it will solely be to remove those individuals from the count itself. We had our conference last week. Jess Phillips gave a presentation on her experience of the Birmingham count, of which I am sure many of you will be aware. One of the difficulties was the clickbait nature of the intimidation that was taking place. It is about trying to get that happy medium: if you take someone out of a count, are you actually giving more fuel to the fire? In polling stations, you can only take someone to the edge of the polling place, the school playground or the church hall field. Beyond that, there is nothing a returning officer can do, other than reporting it.

I will come on to that point in a second. Vijay, what is the Electoral Commission’s view on whether your guidance is clear enough for electoral officers at a local level to use it to stop unfairness in elections?

We produced joint guidance with the police before the general election and sent it to all candidates and to all returning officers and/or political parties. That was broadly welcomed and, I think, used. I think we can go further. We would like to get the guidance as clear as we can, so that everyone is operating on the same basis as to what is and is not acceptable. But there will be a grey area somewhere in between. Some statements will clearly give offence, but there is nothing to stop someone taking offence at any statements. From our point of view, we want to maintain the tradition of extremely robust but respectful political debate. What we saw at the election was so far outside that, though, that we can tackle some of the harder edges. Going up to someone’s family when they emerge from their family home and shouting at them is not in any way defensible. Some of the posters we saw were not defensible. We would say, if it is possible, that we should concentrate on the ring of things that we definitely saw happen at that election. I very much take your point about how sometimes, for some candidates, it was very hard to find someone to complain to or someone to tackle what was going on. It comes back to policing on the ground, or what can be done on the ground. There is room for debate, and maybe this conference is going to help us find the right balance as to where to put the line so that it does not constrain free speech, because after all we do not legislate. Nobody controls and nobody regulates political free speech—it is our most protected area—but we saw some things that went well beyond that.

Sometimes some of this campaigning can be an attempt to restrict free speech.

Absolutely.

Let me come back to Peter’s point about polling day and polling stations. It is a really difficult area. I have had some very good conversations with the chief executive and returning officer in Sheffield, talking through the issues they face and whether they feel certain that they have had absolutely clear advice about what constitutes unacceptable behaviour, to use a non-technical term. It is clearly unacceptable for people to go into a polling station and start lobbying and campaigning. That would not be allowed. Outside, political parties have traditionally had someone on the door taking numbers. We know why they do that; it is part of an old political process. That has now grown into having several people giving out leaflets at the door—a scrum of people through which a voter has to walk, all of whom are shouting on behalf of one candidate. If that is in the grounds of the polling station, you probably have the right to do it. But what happens if it is just outside the polling station at the school gate, or just down the road? It is still pretty much intimidation for someone to have to walk through that. Is there a need for clearer guidance? The returning officer said to me that the police cannot do anything about it, because there is no specific offence.

One thing we have done in Northern Ireland is work with parties on a code of conduct that specifically addresses behaviour outside polling stations. It is, of course, a voluntary code of conduct.

Exactly: it is a code of conduct. Quite frankly, some of the organisations we are dealing with now are local movements rather than political parties. Have they all read the code of practice? Good luck. Will they abide by it? It is not going to happen, is it, so what are we going to do?

We have section 97 of the Representation of the People Act, which concerns disturbances at election meetings. The behaviour deemed illegal is just described as that which is done in a “disorderly manner”. There is quite a wide variance between one police officer or person and the next on what they believe to constitute a “disorderly manner”. What “disorderly manner” means at an election meeting could be better defined.

In the guidance that we sent to all candidates, we tried to define this as far as we could. It had a clear section that was just two pages long on “When should I contact the police?”. It states that they should call the police for anything that involves “Threat of imminent violence...Communications, on or offline, which contain abusive or threatening language…Repeated unwanted contact that may constitute harassment or stalking…Racial, homophobic or other discriminatory abuse or threats”. The final behaviour listed is “Fixation on you or an issue”, because we saw some fixated campaigners. That is definitely the sort of thing that, when raised with the police, they were looking at. It was about trying to make this as real as we could for campaigners and setting out what the barriers were and when they should contact the police.

What about vexatious actions? What if there are 20 people with placards and a loudspeaker shouting outside the polling station and people have to walk past them?

It depends on where the individuals are. According to the definition, you have the polling station, which is the room in which voting takes place, and the polling place in which it is sited. That could be a school hall; equally, it could open straight out on to the highway. There have been many examples in the past of political party offices emblazoned with posters directly opposite a polling station. That is outside the jurisdiction of the returning officer, because it is not within that polling place. If it is within the place, the returning officer has a degree of—perhaps not control, but persuasion, because there is not a specific—

We have a void here, haven’t we?

Yes. Absolutely.

Yes.

One of the things that we would very much be interested in looking at is whether returning officers should have the power, maybe with the police, to set a perimeter around their polling station more broadly. That is what happens in Australia, in some areas, and other jurisdictions have it. Small polling stations in alleyways, for example, are hard to get to, and such a perimeter can prevent the kind of pile-on effects that you are describing.

But in fairness, there are rules that are a bit greater, Peter, that say that posters cannot be within a certain proximity of a polling station. At the last election, they were literally put up on the rails as you walked in. We asked for them to be removed, but neither the polling clerk nor the police would do anything about it. In the end, somebody took them down—it just happened to be me—but it should not get to that, because that could cause an incident. It is about people understanding the rules and being more effective. The rules are there, but they are not being interpreted in the way that they were meant to be. The other thing I would say to the police—I have to say this—is that at one time you used to have at least one policeman regularly visiting on a circuit around polling stations. We are lucky if we even see one person coming out during a polling day. All that has gone, because we have lost that structure that used to be there of asking, “Are there any issues, problems or anything?” The fact is that they have to phone somebody to come out. That is where we have lost the authority that we should have on a regular basis. Maybe, once every hour, somebody should be checking to make sure that those fire points are not coming rather than waiting, especially in an area like mine. Chorley is an 80 square mile constituency—some are bigger, some are smaller, but it is about having that consistency of police being effective and visible.

On the guidance that was sent out, do you think that it would be better, perhaps through the police, if every individual police officer that had to go out to a call from a candidate had actually read that guidance, knew what the guidance was and had an understanding of what our understanding of the guidance was, rather than making decisions on law that does not necessarily apply during elections?

I totally agree. The authorised professional practice is there for them to do that. The idea of having the video on every police officer’s computer screen was to try to get them to look at that. It has not been enough. One of the recommendations we have taken from our debrief is that every police officer will be issued with, hopefully, a paper and a digital Z-card to give them some guidance. It does come from getting those police forces and all those community policing teams to buy into that, and that is a big bit of work that we are going to take forward.

I want to go back to something you said, Peter, about the count for Jess Phillips, because a lot of us saw that. I was a bit confused, because I thought there were restrictions on the number of people or supporters who could accompany any candidate—certainly in my experience this year there were, and there were when I was a local election candidate. Is that not standardised across the country?

Legally, to be a scrutineer at the count, there is a formula based on dividing by the number of candidates, for example. I am not fully aware, but I think that one was within an area where all results were being announced. It was more general, for the 10 constituencies that Birmingham had, but the control was to remove those individuals. They did not actually leave the room—they turned their back on the stage, as I understand it from what was said at the conference—but that is the limited control. But you are absolutely right: the actual scrutineers of the election are governed by a statutory formula.

It is fascinating to hear all this. I don’t recall seeing the guidance, but would it have gone to the named agent?

Yes, it should have gone to the named agent. It went via the bulletin that we send to all returning officers to make sure they can send it on to all candidates.

Got you. Maybe I wasn’t paying attention. I don’t know—it’s not impossible.

But you may have been, because one of the things that we definitely found difficult at times—this was a lesson that I think a lot of us took from the general election—was direct communication with candidates. If I can speak for the police and us, having more direct information that we can send to candidates—whether that is in local elections, where there are thousands and thousands, or at the general election—information on what is going on, particularly if there is a fast-rising security threat or for whatever reason, would be very helpful.

We were not sure whether they existed, because we could not find any trace of them anywhere and they did not come to any hustings. We are talking about things like polling day and the count. In between elections, does the Electoral Commission operate, or is your remit within that short and long period of the election?

It is very much in between. Obviously, a lot of the abuse and intimidation actually happened in between elections, and some of it built up very gradually, particularly with some of the local issues. Local Facebook and WhatsApp groups, which were remarkably closed and hard to see into, were where things really built up locally. Then there were sparks that happened during the election campaign itself. Certainly, from the work that we are doing on abuse and intimidation, we are keen on looking across the whole period, but also at all candidates, not just MPs and councils. It has an effect. If someone is standing even for a parish council and they are abused, they might not stand further, so we think we have to draw the scope wide.

Sometimes town halls are used for international elections. For example, the Polish election was done at the POSK community centre up the road from us. They have their own arrangements, do they? You guys would not cover that? Vijay Rangarajan indicated assent.

Perhaps I can start on that question. You are absolutely right about the speed at which some of these rumours can spread online, and of course the barrier to entry to producing very realistic videos is falling and the cost of that is now incredibly low. That is another area in which it is possible things will get worse. First, we work with the social media companies. There is a spectrum of how willing and how quick they are to respond, but clearly, if anything is threatening life, is illegal, or is clear misinformation or disinformation—for example, is designed to put people off voting—we will act on it as fast as we can. I cannot say that we have had complete success, but we are looking at and working very closely with Ofcom and other regulators on the Online Safety Act and the various other pieces of legislation coming in, and we are asking, “Can we use that?” Secondly, there is a part that is about voter education. We are keen to build up the work we do in schools, discussing questions like: why is it great to be a candidate? Why are politicians necessary, important representatives of you? What do they do that is important for you? How should we treat them and humanise more? We did notice a big difference between people who had had contact with their elected representatives, at whatever level, and those who had never had any, and that helped a bit on the abuse coming back from a video. It also helped digital literacy—just knowing that you shouldn’t believe everything you see online. Finally, in relation to the very top end of the produced videos that some candidates, particularly women, are getting—we are talking about the high-end pornographic rape threats and so on—there are some very interesting examples of legislatures putting in place systems to tackle that. One, which we drew attention to in our evidence, is the Scottish Parliament. It has a system running whereby they can collate together “public source” for their candidates who opt in—it is a pure opt-in system. They look at all of the horrendous videos coming in and help them to filter, which means that some poor young person sitting in one of your offices doesn’t have to do that, but they also run a legal and a police rule over it to ask, “Is that legal? Should we actually be prosecuting that? Is it coming from the same person?” Then they try to track back. That takes some resources, but we think that it is a model definitely worthy of exploring in order to protect candidates from online abuse, which otherwise can just explode.

They definitely need to be checking more of what they do, what they allow and what they then disseminate. I very much agree. As for how they do it, I think the technology changes so fast, it is up to them. It is probably not just them though; parts of Government, parties and other organisations can also look to see what kinds of techniques are being used—because again, they evolve very fast.

It is about the duty of care and responsibility of companies. As I always say, I believe it is time for Governments to actually legislate if companies will not take action responsibly.

I want to chat to you a bit about the support available for candidates. Several witnesses have stated that candidates aren’t clear on what can constitute harassment and what doesn’t constitute harassment. When I was a candidate, I only reported something to the police because a Labour party staffer was there and said, “This has crossed a line. You need to report it.” I think that is quite common among candidates, but independent candidates in particular have raised that concern. Have you considered the additional guidance that they might need?

We provide a regulatory support function year round, and we ramp it up before elections. One of the things we do as part of that is run webinars, where we see a lot of attendance from independent candidates because they don’t have the party machinery around them to support them. We are able to provide that advice there. We run webinars specifically on security, and on how to handle abuse and intimidation. In our joint guidance, which we do with the police, we have a section on how to recognise harassment—when what you are experiencing has crossed that line—and to report it. The guidance itself also contains the necessary contact details so that people can take action straight away. We run a helpline as well, so people can call us for advice and support, too.

I would be interested to know if you were told about it, but there was also available a one-day or few hours input on staying safe online and how to protect yourself online, put out by the police. I do not know if that got through. It was fairly well attended, but one of the issues we had was making sure that every single candidate or MP saw the advert.

I do vaguely remember that. The other point as well is that the election period is six weeks, but harassment can start a long time before that—I know some of our candidates were in place nearly two years before the election. How early do you think it is possible to engage with potential candidates, as well as those who are selected straight after the election has been called? As you say, it was a snap election, so how do you make sure you are able to get in with them as well?

We are happy to engage with people whenever they become a candidate—indeed, we try to engage people before they become candidates, because quite often they come asking, “What does this mean? How do I do this? How do I fill in the forms?” We are about to publish some more materials and videos on how exactly you fill in the candidature forms. We get in as early as possible.

So when a candidate is being given advice, if they are subject to harassment, the police can be involved and evidence is gathered. But if they are elected, presumably Operation Bridger kicks in at that point. Is there a clear pathway for the evidence to go through, or do you just start all over again?

I do not know of a specific case where that has happened, but knowing the police, that would be fairly seamless. It would not be a difficult ask for any investigation.

Is guidance issued to the 43 forces about the transition from candidate to MP?

The Operation Bridger SPOC is in place in each force, so they will be there for that purpose. It is also key to remember that the Operation Bridger SPOCs—the same is true with all portfolio SPOCs—are not necessarily the ones carrying out that investigation. They are not going to be the lead officer in that investigation; they are there for advice and to try and make sure that things are happening. The actual investigation will potentially be carried by the local CID. That would not change through that transition.

We have run out of time for this panel, so thank you for speaking with us. If there is anything you did not have the opportunity to say, or further thoughts you wish to share, please do send them in writing to the Secretary. Thank you, all four of you, for coming along. Witnesses: Will Fletcher, Dr Sofia Collignon and Professor Helen Margetts.

Good afternoon, and welcome. Please introduce yourselves.

I am Will Fletcher, Interim CEO of the Jo Cox Foundation

I am Sofia Collignon, an academic at Queen Mary University of London and the Director of the Mile End Institute.

I am Helen Margetts, Professor of Society and the Internet at the University of Oxford and Director for Public Policy at the Alan Turing Institute.

We would say that there is a group of different factors at play causing these things. We definitely agree with what was said by the previous witnesses about the need for more political literacy, because it is really important that our population understand what our politicians do and learn how to engage with them in productive ways. I think we learned from the previous witnesses too that, ultimately, there is no silver bullet. There is a range of factors behind this growing problem of abuse, and we will need properly co-ordinated action from a range of different institutions if we want to solve it. That will include organisations like security and the police, but it will also involve doing some of the hard work to change the culture that we see in this country and right across the world, because this is not a problem restricted to us. For us, it has to begin with education work to raise political literacy, and quite possibly some public campaigning work to raise awareness among the population of what is and is not acceptable.

One thing that I think is relevant in this case is that in interviews some police officers mention that sometimes they are called once it has happened. I cannot comment specifically about that mob, but the police were asking whether there is a way to encourage people who are going to organise a political event to engage in risk assessment beforehand. The police are operating with constrained and limited resources now, but if they have time to prepare, it is much more likely that they can prepare a positive and much more strategic way to tackle potential mobs. This is of course the case when we know that an event will be taking place that could potentially attract a focus, but it is difficult to foresee that in every situation. To the second part of the question about what is driving the perception that politicians might be fair game, I will say that among the population there is a general feeling on frustration about issues that are long standing and perhaps have not been solved as quickly as they want them to be. When we ask candidates who is behind actions of harassment, abuse and intimidation, most of them say that it is angry members of the public. We need to think very carefully about how we can address this underlying anger and improve the reputation of not only politicians but public life and what it is to be in public life.

I want to ask about identity-based abuse. We know that the survey data comes up with misogyny, Islamophobia, antisemitism—in fact every type of racism—ableism and all those factors. Do you think specific measures are needed? What would your advice be for people from these categories, so they feel safer engaging in politics?

I was going to make a comment about the online world in answer to the previous question. We know there is an association between online violence and offline real-world violence. We must think about whether part of the perception of MPs being fair game comes from the online world, where people may become accustomed to thinking that MPs are fair game. In relation to your question, it is certainly the case that women and people of colour receive much more abuse than white men. It is difficult to measure, and you will see studies suggesting that men receive just as much abuse as women—I have been part of one of those studies myself. The challenge is that there is an association between activity online and the abuse that you receive online. Quite often, men are more active online. There is also a relationship between seniority in a party and the extent to which someone is a public figure. There are all sorts of challenges to measuring it, but we definitely know that female MPs, along with all women, are much more likely to do safety work in terms of keeping themselves out of online spaces, not expressing opinions in online spaces, and using user controls such as blocking, which, when you think about it, is a worrying issue for democracy in general.

Do you think that there should be better legal definitions of misogyny, or that we should rethink these characteristics?

Listening to previous witnesses, one of the things that came across was about where we define abuse and where we draw the line between abuse and robust disagreement or a very heated discussion. One of the important things to recognise is that abuse comes when it is personalised. We can all disagree on policies, when one insults the topics, but when the content becomes about the person, it can become abusive. It is true that if you are a member of an under-represented minority, an ethnic minority, or a woman—because the perpetrator will always try to use what is unique or special about you as a person—you are an easy target. They take your identity and make abusive comments based on that. I am torn, because I think that a lot of the measures that we have right now put a lot of stress on the potential victim, which increases their mental load, such as the discussion that if you are vulnerable on the street and more likely to be attacked, why are you willing to put yourself at risk? I think that we need to change this and be very clear that any identity-based abuse is not just a comment; it becomes abusive quickly, or at a higher speed, so it should not be tolerated. In political discussions, we should be aware of that. I also think there should be a clear definition of misogyny and identity-based abuse.

I agree with a lot of what I have heard. Vijay from the Electoral Commission spoke about some of the evidence that they have gathered from the general election. One finding they had was that, of those candidates who had experienced abuse, 11% said they had experienced abuse from other candidates, so one issue we need to talk about is the behaviour of those in politics and the role that political parties can play in ensuring that there is good behaviour from their campaigners and candidates. We all know that all political parties have codes of conduct, and I am sure we have all heard stories of times when those codes of conduct have not necessarily been applied in the way we think they should be, so we would advocate that the political parties think about that, particularly on an equalities basis. We would also recommend that political parties think about civility and equalities training for their candidates to solve the issue of candidate-on-candidate abuse. Finally, regarding misogyny, we clearly know that violence against women and girls is a growing issue that is separate from but overlaps with this issue, so I think that monitoring how efforts to tackle violence against women and girls have positive impacts on abuse towards women in politics will be very important. For instance, we are looking very closely at the new guidance that has come out from Ofcom—their new code on violence against women and girls—to see how that can be used to ensure that abuse towards women in politics online is being effectively tackled and addressed.

My question is for Professor Margetts and is with regard to your experience of social media generally. In terms of racist abuse, and to your point about high-profile people, we have seen the abuse that Kemi Badenoch, Rishi Sunak, David Lammy and Diane Abbott all receive, but it is not just people at the higher level. I got a huge amount of abuse when I was just a candidate and effectively a nobody—very few people outside my immediate area would even have known that I was standing for election. I get that abuse from the left and the right and from black people and white people. What is it about the correlation between race and opinion that means that I can say fairly uncontroversial things—that a white politician could say and not receive any abuse for—that then make me a target, related directly to my skin colour? How has that evolved, and why is it such a feature of politics now?

In the online environment, it has been a feature for a very long time—since social media platforms have existed. I agree that it is terrible. I cannot give you an explanation, other than that those views exist in society and are being reflected online. I should say that it is not a normal distribution; there are no normal distributions here. Within any category like race or gender, there tend to be a few figures who get a phenomenal amount of abuse—all MPs get abuse, but most people do not get those levels of abuse. In terms of tackling it, I think you sometimes have to think about it differently; you have to think about how you would deal with the problem that some very high-profile figures have, but which does not apply to others. But I am not solving the problem that you raise. It is illegal. It is covered also under the Online Safety Act. There are things that were not illegal before but that under the Online Safety Act are now illegal. Implementing that may be one route to tackling this.

Will, you mentioned civility training. Is there anything else that you think political parties and politicians could do themselves to disagree better and improve the tone?

Absolutely. One of the many drivers of this issue is distrust in politics. Ensuring that we have good standards in politics is an important part of achieving trust. We absolutely would call for all politicians to role-model good behaviour and to find ways of disagreeing agreeably, in a world that is becoming increasingly polarised.

May I add something? Parties need to expand the scope not only to candidates and politicians, but to their party members in general. It should be a condition, once you become a party member, that you get training on what is not acceptable behaviour.

This is very interesting. I have had, for 10 years, probably about one person locked up every year; the 2024 one was pre election. We now have the Elections Act 2022, which should disqualify people if they are found guilty of that behaviour. Do we know how many people that has applied to? Having policies is great, but are they being executed?

I don’t know, but I can add something on that. There is a lack of monitoring on this issue. There is now a multiplicity of actors or stakeholders, but we do not really have reliable data or any way to systematically compile all the data and make it available. In a way, we have the problem that it is a very complex issue with a multiplicity of solutions and initiatives—sometimes complementary and sometimes overlapping—but because we are not measuring things, it is becoming very difficult to evaluate what has been working and what is not working.

Theoretically, the disqualification order means that the penalty is that someone cannot stand if they have intimidated a person. You have talked about this happening internally, between candidates, but we do not know.

No.

But we have heard, and you heard from previous witnesses, about how police interventions can successfully act as a deterrent to perpetrators going on to do this again. As a broader point, there have been some very high-profile cases of perpetrators facing criminal action—they have been reported in the press—but that does not seem to be acting as a deterrent to others going on to do it. So there are good deterrents to stop perpetrators reoffending, but there is still the question of how we change the culture more broadly.

I have always thought that it is best not to advertise and encourage more people, so I do not go public about these things. Brilliant—thank you.

We have heard from you, and the other panel, about how stuff that starts online is moving into the physical, outside world, and vice versa. I am not going to dig into that in too much detail, because we have established that it is a thing, but how do you think that regulation of social media could be strengthened here in the UK to try to help address that whole point about online activity then driving what is going on outside?

I do think there are things in the Online Safety Act, but we have not seen them in practice yet, because at the moment the companies are still supposed to be getting their act in order, so as to comply with the Act, until 16 March. There are things in there that could improve the situation, but that is never just going to be down to legislation. There are all sorts of other issues that could make things better. I do not think this is specifically covered in the Online Safety Act, but we will have to see how it plays out. It relates to Ben’s question about race. Automatic methods of detecting racial abuse are so much better than they were before, so there is no excuse for the platforms allowing unchecked racial abuse. There is a lot more that could be done in that respect, particularly now that methods are so much more automated than they were before. But it also depends on lots of things. The other witnesses have been talking about education and questions of digital and political literacy. There is also the point—of course, I do not want to put the work on to MPs—that all these other things are quite clumsy tools for individual people suffering abuse. User controls and ways of stopping people commenting on posts and things like that will need to be used, and people need to get really literate at using those. User controls are mentioned in the Online Safety Act, but we do not know very much about how aware people are of them and how much they are willing to use them.

A lot of the stuff around getting social media platforms to take action themselves focuses on the big ones, but there are obviously these smaller, more niche ones that sometimes can be more problematic. Do you have any thoughts on the risks that those pose and how we might try to address them?

The online world is quite complex. A lot of very serious abusive activity starts in places like Telegram and then moves on to the mainstream platforms. Most of the platforms used by a lot of people are covered by the Act, but I think it has to apply to all platforms.

Regarding the social media world, do you have any thoughts on statements and comments made underneath posts that are potentially libelous or defamatory, which I think many of my colleagues will have experienced? I know that there may be some coverage under the Online Safety Act, but I have a deep concern that if the onus is on the MP or the candidate to try and take legal action, they may not have the resources or the wherewithal. However, the actual defamatory statement does intimidate them, because the effect in the real world is that people following the election or the MP see the defamatory act and might believe it is true, and the MP or candidate therefore thinks that lots of people out there think that that terrible allegation about them is true. That puts them at greater risk, and it makes them feel at greater risk, so I think that is something that needs to be addressed. What do you think?

I totally agree with you. That is one of the things that the Act is supposed to do something about, and I think that that should happen. The question is whether it can, of course. But we know that Brazil pushed back—in this case, it was the Supreme Court—and eventually got X to turn around and moderate content better. I think that that is what we hope for from the Online Safety Act—pushing back at the companies and implementing the Act.

This is another area where there is a role for political parties where candidates and Members are part of political parties. In the previous session, we heard Lee Anderson’s example of how his political party at the time provided legal support, and we would encourage political parties to provide support when civil action is more appropriate than criminal action.

I completely agree that we need more civility in our politics, and I commend the work that you have done at the Jo Cox Foundation, but I also think that the public have a responsibility. That responsibility includes not libelling people and thinking you can get away with it, because the effect on whether people want to stand and on the way they feel is hugely toxic. If they think members of the public think they are a terrible thing—I will not mention the things that other Members have been called, and that I specifically have been called—they must conclude that lots of people think they are worthy of insult, or worse. That is very dangerous, and it is not just about political parties.

One of the operational things that should be changed is the threshold for the police to act. Currently, it is on the victim to collect all the evidence, which means that a lot of people are re-victimised over and over again when they have to take the screenshot and save it. It is their responsibility to see these things, and there is no clear threshold for when the police can act or when any legal action can be taken. A very useful thing would be to change the threshold or clarify where it is.

I am interested in what you think about the fundamental design questions around social media, and in how much we can really counteract them. I agree with everything that has been said, but there is something specific, is there not, about the way that political debate unfolds on social media, as opposed to many other fora? One thing is the fact—we might come to this—that people can remain anonymous if they so choose, in order to make the comments they want to make. The other is that they are linked to a much wider group of other speakers than they otherwise might be. Finding the 200 other people who believe the crazy thing you believe used to be very difficult; it is now much easier. Do you think that that is, to some extent, just a consequence of the way social media is designed, and we can do nothing about it, or is there something that we can do about those more structural challenges in social media?

Social media is not all the same, but we tend to talk about it a bit as if it was. That is particularly true in academia, actually, because for a long while the only place we could get data was Twitter, so a lot of talk about social media was really about that one platform. We have to be careful, because all platforms are not the same; some are more civil than others. A while back, Twitter became a bit of a public sphere, in that every single MP and every public institution and organisation tended to be on there. It was even used as a place for complaining about trains and things like that. I think we have moved away from that now. X has lost 30% of its users, and a lot of people and institutions have moved away from it. In many ways, that is a shame, in democratic terms, but it also means that there is more choice. As a political culture, we have to decide where people are going to be. We cannot all make that decision; some of it will be about individual decisions. It will be about all the kinds of controls that we have talked about—legal controls, and social and cultural norms. I do not think it is all down to the basic design of social media. You already see that some representatives are just not using X anything like the way they were before. They have moved to other places or have used user controls. Members of Parliament will have to think about why they want to be on social media—what they are there for. Are they advertising or broadcasting what they are doing? Do they want people to be able to communicate with them and respond to them? It may be that there are different channels for different things. If you want your constituents to be able to get in touch with you, maybe the channels on your website or parliamentary channels are how you want to do it, and then the broadcasting happens in a different way. I think that is how we will have to think about it. There is no obvious successor to that kind of public sphere concept.

We are obviously talking about social media, but where do you see the role of mainstream media, particularly newspapers? I think that quite often the firework for some of the things you see on social media is actually lit by certain newspapers, and then people run off with it and let it explode all over social media. In terms of them acting more responsibly, how do you see that?

Based on some of the interviews that I have done with candidates, one of the most harmful things that can happen is a story explodes on traditional media, it gets covered but they never get the right to reply, or the story is not fully covered so in the end their voice is not heard. Some of them say that it is very empowering, once the story has been solved, if they have the opportunity to respond. I think the media has a responsibility to also provide a right to reply to people affected by the news, and especially to answer to requests, once the whole issue has been cleared, and to give them the same space to respond that they gave to the original news.

Picking up Dr Collignon’s point about the threshold at which police can act, I have reported a lot of abuse to the police, and I have seen huge inconsistency in the way different forces handle issues. I don’t think it is necessarily the threshold at which the police can act, but the determining of what constitutes a crime and therefore meets that threshold. Only this week I spoke to a chief inspector around an issue that he said police wanted to push forwards, but it was actually a crime registrar—someone I had never heard of who would determine whether it —met the threshold. That crime registrar is answerable only to chief constable of that force. The Metropolitan police, for example, do not use crime registrars; a sergeant can take a decision on whether they want to push something forward. That vast inconsistency between forces in the way things are handled means that you will get completely different outcomes for identical issues depending on where you are in the country. I think that is probably something we want to look at in terms of ensuring consistency. Most people assume that the police are the police no matter where they are in the country, but there are 43 different forces that operate under completely different rules and statutes. That is something that needs to be aligned so that we can better handle this.

Can I add to that? That also corresponds to a lot of experiences that different candidates have. It has been mentioned that in certain communities or areas they have very close contact with the police during the campaign and after the campaign, they have a nominated contact person and so on. In other places the police are not on board, and they basically do not act, or it is a much weaker relationship. We have been working hard to encourage a stronger relationship with the police, but there is huge variation and one of my recommendations would be to have mandatory training at the College of Policing on this issue, so that all officers receive at least very similar baseline knowledge.

On a general point, we have talked about social media moderation and civic literacy; is there anyone else out there in the world looking at those things?

Looking at what things?

Looking generally at how we tackle this issue of threats and harassment, are there any other countries where work is being done that you think might work with some of this social media regulation?

Australia was an early innovator in safety legislation, and they have done some impressive—

So they immediately put corrections up at the side when things are not factual—there are things that are being done. I think the problem is that with the media—I know Mark mentioned newspapers—there is IPSO, the press complaints commission, and there are different codes that they have to adhere to. But it is the wild west when it comes to social media. That is the thing that is missing. I would have thought perhaps your emphasis might have been, “There should be a code of conduct that they have to adhere to, the same way that the media does.” Is that fair to say? Professor Margetts indicated assent.

I don’t think we are in a moment where social media will self-regulate or will be very likely to become a nicer place. We have examples of how it is becoming even more the wild west, and that has opened the door to international interference in political issues in the UK. Now, they have to moderate and their codes and their responses to harassment and intimidation are individualised. Another way would be to provide them with what they have to do, and the very basics that they all have to abide by if they want to keep on operating in the UK.

But would it be fair to say the press, quite rightly, get handcuffs when they cross the line—though maybe not as much as we would want them to—but here there is nothing, so we need to begin to ensure there are handcuffs we can put on?

Can we talk about artificial intelligence? We have not mentioned it so far. How do you think artificial intelligence will change the picture that we have been describing, either positively or negatively?

We at the Jo Cox Foundation are already concerned about misinformation and disinformation and how they drive abuse and intimidation. There is clearly a concern that the increased availability of AI will make it much easier for many more people to produce very convincing misinformation and disinformation, so that is absolutely something that we need to think very carefully about. I am also particularly concerned about deepfake images and how that interrelates with the issue we have already discussed of misogyny. When it comes to deepfake pornographic images, that is a very concerning trend that we are already starting to see. We think it is important that we have better media literacy as a country. Part of the solution has to be helping people to understand and question what they see online, so they can make those judgments for themselves. As to how we effectively tackle it from a technological point of view, that is not our area of expertise, but it is clearly something that needs to be explored in great detail.

In terms of giving people the tools to exercise their scepticism about what they might be seeing, do you think that there is more scope for watermarking, or something similar, that enables people immediately to understand that something is an AI-generated image, text, video—whatever it may be?

Absolutely. I know that some of the social media platforms I have seen have watermarking capabilities and things; it is about getting them to be used consistently. I would assume that those who use AI to deliberately generate misinformation or disinformation would not use watermarking tools if they were purely voluntary.

The main thing that AI will do is turbocharge the online harms we already know about, so we need to use whatever expertise we have developed to deal with those when we think about AI. Last year was supposed to be the year of the AI election. People were talking about that from January onwards. We did not really have, and certainly not in the UK, an AI election. That is something to be glad about, but of course AI is developing all the time. You are absolutely right that deepfakes and in particular misogynistic abuse and targeting are concerns. Some of the concerns we had at the beginning of last year look not quite as concerning as we had thought. For example, with a PhD student and a research team, we have done some work on microtargeting and large language models. There were a lot of predictions about that. It looks like microtargeting is not actually an effective form of political persuasion, not any more so than, say, the best message of GPT. In terms of people looking at it, I have at least six PhD students trawling over all this, including watermarking and all the strategies, such as prebunking, warning labels and so on. But obviously, that is not going to solve the problem. There are lots of researchers looking at this and working out ways of dealing with it. One important thing to note is that AI will be needed to tackle AI. We have seen that with other online harms—I mentioned the automatic identification and removal of abuse. We are going to need to innovate to tackle this question; there is absolutely no doubt about that. If I could make one last point, we have seen quite a lot of abuse and violence on social media that has actually come just from people, without any AI, and it is important to remember that. We should not blame AI for everything.

I will be happy with a one-word response. Are we missing out on great potential candidates because people are afraid of the intimidation and abuse they might receive?

Yes.

There is evidence to suggest that around twice as many women as men are now scared to even express their political opinions online—very worrying.

There is also evidence that shows that it is having an impact at a very young age. Girlguiding UK did some research that found that more than a third of girls and young women are put off entering certain professions because of the level of online abuse, including politics.

There is also differentiation between the stages at which their political interests are activated. In some interviews with party activists, they say they are considering their safety before they decide whether to canvass, support a candidate or engage in any electoral arguments.

Thank you for coming along today and for the way you have presented your answers to the questions you have been given. If you have any further thoughts, please would you pass that information on in writing and be in touch with the Clerks? Thank you once again.