Oral evidence: Hate Crime and its Violent Consequences, HC 609
Tuesday 15 November 2016
Ordered by the House of Commons to be published on 15 November 2016.
Members present: Yvette Cooper (Chair); James Berry; Nusrat Ghani; Mr Ranil Jayawardena; Tim Loughton; Stuart C. McDonald; Naz Shah; Mr Chuka Umunna; Mr David Winnick.
Questions 1-126
Witnesses
I: Nick Antjoule, Hate Crime Manager, Galop, Shane Gorman, Disability Hate Crime Adviser, Leonard Cheshire Disability, and Nick Lowles, Director, Hope Not Hate.
II: Carl Miller, Research Director, Centre for the Analysis of Social Media at Demos, and Dr Pete Burnap, Director, Social Data Science Lab, Cardiff University.
Written evidence from witnesses:
– Galop
Witnesses: Nick Antjoule, Shane Gorman and Nick Lowles.
Chair: Welcome to this afternoon’s session. We are doing the extremely unusual thing of starting a session early, rather than late. I thank our witnesses for coming today. Could you begin by introducing yourselves and telling us which organisation you are from?
Nick Antjoule: I am Nick Antjoule. I work for Galop. We are an LGBT anti-violence charity that covers hate crime, domestic abuse and sexual violence against LGBT people.
Shane Gorman: My name is Shane Gorman. I am a disability hate crime advocate in Northern Ireland with Leonard Cheshire Disability.
Nick Lowles: I am Nick Lowles from Hope Not Hate.
Q1 Chair: Thank you very much for coming to give evidence. As you know, this is an inquiry into rising hate crime in Britain. Could you start by giving us a sense of the trends that you have each observed over the past three years and also give us an example of the experience of hate crime and the impact it might have on someone?
Nick Antjoule: Just thinking in the short term, post-Brexit we did see a sharp rise in the number of LGBT people who were coming to talk to us about experiences of hate crime. That was quite a large rise, but I know that a good portion of that was made up of people who were saying, “Usually I would be putting up with this. I would shake it off and say, ‘That’s my experience of going on the bus or the train or wherever else and experiencing transphobia’, but because of the context, I thought it was important to get it recorded.” I think a lot of the rise was made up of that.
The context is that there is no good data or research on the levels of confidence among LGBT people in the criminal justice agencies. Contextualising the rises in the amounts of homophobia and transphobia that get recorded by the police is tough to do. My position tends to be that there has been a rise over the past several years. In the absence of any evidence of that being a massive rise in the number of hate crimes that people are facing, I guess I tend to think that so few get recorded that a rise is, on balance, probably a good thing.
Q2 Chair: The police statistics show a 29% increase in sexual orientation hate crimes and a 41% increase in transgender hate crimes, just in the space of 12 months. How much do you think that is about a growing problem and how much is it about increased reporting?
Nick Antjoule: My general feeling is that it is probably both. People are probably facing an increased amount of hate crime and reporting more, whether that is because they are more confident, because they feel it is more important to get it recorded or because it is more likely to be recorded by police once it is reported to them.
An important bigger picture that sometimes gets missed out is that roughly a quarter of hate crime against LGBT people ends up getting reported, but actually the amount that gets recorded by police is slightly smaller than that. There is a drop-off there that I think is important to factor in as well.
Q3 Chair: What do you identify as the possible causes of this?
Nick Antjoule: During the debate about leaving the EU, there was a lot of rhetoric around opening up the door to voicing prejudicial opinions that perhaps might not otherwise have been voiced. I cannot help feeling that perhaps that might have played some small part as well.
Q4 Chair: For example?
Nick Antjoule: Obviously a big part of this was race and faith. Lots of negative opinions were voiced as part of the debates in relation to race and faith. I cannot help feeling that that opened up the door to prejudicial opinions about other communities being voiced as well. Just as a quick example, several people came to us and said that they had been verbally abused by people saying “Now we can get these people out of the country and you are going to be next”—meaning people who were visibly LGBT. There is a very visible link there.
Shane Gorman: Over the last three years we have been providing independent advocacy support to victims of disability hate crimes in Northern Ireland. We have been working on a one-to-one basis with victims to ensure that they can access the PSNI as well as the wider criminal justice system. We believe that that support has increased the reporting of disability hate crime, because we have removed potential barriers in the disabled victim’s life that might have prevented them reporting it. Generally, these can be communication difficulties, access issues or mobility issues, so if we work in a person-centred way we can support the individual in reporting to the police but also through the wider criminal justice system.
An example is an individual who I have supported personally. She had a learning disability and she was afraid to report a serious assault to the police because, due to her disability, she was afraid of police officers in uniform. Through my independent advocacy support, I could arrange for an officer to meet her in her own home without wearing a police uniform, which made that part of the criminal justice process accessible. The individual went on to fully engage with the criminal justice process because of those small changes.
Q5 Chair: Again, the figures for recorded crime in England and Wales show a 44% increase in disability hate crime in 12 months. How much of that do you think is increased reporting and how much is increased prevalence?
Shane Gorman: Again, from a Northern Ireland context, 1.8 million people live in Northern Ireland and an estimated one in five have a disability. Out of the estimated 100,000 crimes reported to the police in 2012, there were only 15 recorded disability hate crimes and 33 recorded incidents. In the last number of years, we have seen that increase to 74 recorded crimes and 134 recorded incidents. We welcome that, because we believe that, given the population of disabled people in Northern Ireland and the amount of crimes being recorded, there is a significant gap.
We are working in a person-centred way to remove those barriers in order to ensure that people can access the criminal justice system, but there is a lot more work to do. We have been doing a lot of awareness-raising throughout the community. Anybody can report a disability hate crime—or any hate crime, in fact—so it is important that people in society recognise that if they recognise something, they should report it to the police as a hate crime.
Q6 Chair: If you put aside the reported crime, do you have any evidence about what is happening to prevalence—whether there are increases, decreases or changes in the level of abuse, violence or hate crimes towards disabled people?
Shane Gorman: From our point of view, our service has been in operation for three years and we have seen an increase in reporting. That is the basis we are coming from. We welcome an increase in the reporting of hate crime; we believe that for hate crime to be addressed we first have to see an increase in reporting, so that we can work towards prevention.
Q7 Chair: Do you have any sense of what the wider prevalence is or of what people say their experience is? If you talk to people, do they say that they are experiencing more abuse or less abuse?
Shane Gorman: Speaking to victims from the disabled community, they would generally say that they have been a victim, and that they have been a victim of a variety of offences over a number of years, but that they never recognised what a hate crime was and never knew that it could be reported to the police. The fact that there are now alternative methods for them to report it to the police makes it easier for them to come forward.
Q8 Chair: Would they say that this problem has increased, decreased or not changed?
Shane Gorman: Well, for us, in the last three years it is very hard to say in relation to change. We welcome the increase in reporting. As an organisation, we believe that there hasn’t been an increase as such in the number of offences that have occurred, but reporting has increased in Northern Ireland.
Q9 Chair: Do you have any sense of an increase in prevalence, or any change, after the referendum?
Shane Gorman: As I mentioned, our service has only been in operation for three years, and we welcome the increase in disability hate crime reporting.
Chair: We’ve got that point. Mr Lowles?
Nick Lowles: I would like to make three wider points that I think are positive. Attitudes in society over the last three years have improved, and there is greater tolerance. Hope not Hate has done some surveys over the past five years called “Fear and Hope”—we did one in 2011 and one in February 2016—and there was a big increase in more positive attitudes towards inclusion and multicultural society. The number of people who have very hostile views, and particularly one group who have views that mean they are prepared to think about violence as a means to an end, has shrunk from 13% in 2011 to 8% in February. That is partly a change in demographics, but I think a lot of it is due to an increase in economic security; the more economically insecure you are, the more pessimistic and fearful you are of the other.
The general trend up until this summer has been positive. There have been big improvements in awareness of hate crime and reporting of hate crime, and the Government and the police have put a number of things in place that have improved that. There are still lots of holes, but there have been some good advances.
One of the other positive trends, particularly over the last 12 months, is that people will take action themselves. We have seen a whole load of social media campaigns where people have filmed incidents, and those campaigns have gone viral. There is often a sense of isolation and victimhood, particularly with online hate, and there is an online community that will rally around. That has become a trend now.
The other thing we have seen over the past three years is spikes. There have been external and internal incidents—the Charlie Hebdo killings in Paris, for example—after which we have seen a big rise in hate crime. Obviously people get motivated by that to express a prejudice.
The final thing I will say is that we did another “Fear and Hope” report the week after the referendum to see how society’s attitudes had changed, and we asked many of the same questions we asked in February. What was quite interesting was that there was no discernible increase in prejudice or intolerant ideas. What has happened is that a small group of people felt emboldened by the result to express a kind of anger and hatred. The referendum debate obviously gave a green light to people who wanted to express their prejudice, but it didn’t necessarily change society as a whole.
The trend going forward could be different—I am happy to spend a couple of minutes talking about that. Our pollsters are Populus, and we have always found a correlation between economic pessimism and fear and hate. Unsurprisingly, in February those people who had the most negative attitudes about immigration and multiculturalism were the most pessimistic about their own future—their own personal future, the future of their family and the future of the country. After the referendum, this group were now the most optimistic people; they thought things were going to change for them.
Straightaway our pollsters said, “The biggest finding of our new poll is that, sooner or later, if their expectations aren’t met, these people are going to be very, very angry.” The possible increase in insecurity around Brexit and whatever agreements are made could increase the pool of people who are in that fearful group. Certainly for us, it is about not only looking at what’s happened, but making sure we are prepared for the potential spikes over the next few years.
Q10 Chair: The police report that there was an increase in reported hate crime immediately after the referendum. It then fell after that, but not to the levels of the previous year—it was still at a higher level than previous years. How far do you think that is about increased prevalence and how far do you think it is about increased reporting?
Nick Lowles: Probably both. I would probably say more the latter. Some of the gaps in the reporting of hate crime are, for example, in eastern European communities. It is interesting that even the Government’s “Hate crime action plan” talked about the groups that are potentially victims of hate crime, but it didn’t include eastern European communities. There is no organisation such as CST or Tell MAMA representing or monitoring the eastern European communities, so there is massive under-reporting.
I was at an event in Manchester with Baroness Williams when she was still at the Department for Communities and Local Government straight after the referendum. There was a woman from the Polish community, and one of the things she pointed out was that many people in her community wouldn’t go to the police if there was a problem. Many of them are living on the edge of society anyway. What is most depressing about it is that they consider it part of the deal of being here. They are here, they are taking jobs, and the backlash is just one of the things that they have to accept. If that attitude is prevalent in the Polish or wider eastern European communities, there is obviously going to be a reluctance to come forward, so I think there is massive under-reporting.
The other thing we have noticed anecdotally is an increase in incidents involving young people and school-age kids. I suppose we all have faith. All the polling shows that young people are much more tolerant and much more mixed, partly because of the demographics, but obviously what happens in wider society has an impact on young people. There have been more and more incidents in school playgrounds, added to the insecurity around their status. I think these are very difficult times for many people in this country.
Q11 Chair: You said that you thought a small group of people felt emboldened. What was it that gave them that sense?
Nick Lowles: We commissioned two academics to do some work on hate on Twitter—because of the Jo Cox murder trial, we can’t publish it until after the trial—and they studied 53,000 tweets. We can provide it as evidence. There were three strands of hate, certainly online—we are talking about thousands and thousands of tweets. There was one that was directly targeted at eastern Europeans—“Get out. We voted out. Go home”. A second group was wider prejudice, be it aimed at Muslim communities or non-whites generally. The third was connected with the Jo Cox murder itself. An unbelievable number of people welcomed that murder. It was truly horrific. Obviously, that was the more political people.
I think what happened was that a small number of people saw the online hate or offline hate, and they saw the referendum result as a green light to express their prejudice in a way they might not have done before. They had the confidence to do it. We have obviously seen exactly the same thing happening in the US in the last week as well.
Q12 Chair: Clearly, huge numbers of people who voted to leave the EU will be absolutely appalled by any form of violence, abuse, racism, extremism or hatred. What is it that you think then allowed a minority of people to think that they had a licence? Would you see the way campaigners behaved as significant? Would you see the way events were reported as significant? What it is it that allowed this minority to respond in that way?
Nick Lowles: It was probably a combination of factors. I guess people who had deeply entrenched, prejudiced views decided that they had the green light, so they acted themselves and found a justification for doing so. Obviously, the context and the nature of the debate and particularly how it is portrayed in the media. I personally believe that posters like “breaking point” had a role. For most people who voted for Brexit, immigration was not the No.1 reason, but obviously for some people it was, and in a way the toxic nature of the debate polarised the whole thing. Maybe because of the type of people they are, or whatever, they then acted upon it. So I think it is a combination of the toxic nature and also themselves looking for a reason to do it.
Q13 Chair: Mr Gorman and Mr Antjoule, do you want to add any further comment on whether the “breaking point” poster or other things inflamed the situation?
Nick Antjoule: Nick has made his case very well. I’d agree.
Q14 Nusrat Ghani: Mr Antjoule, when you gave evidence earlier you mentioned that there had been a rise in hate crime against the LGBT community over several years. Is that correct?
Nick Antjoule: Yes.
Q15 Nusrat Ghani: And you also stated that you were not sure whether the recent rise was because of the EU referendum, or whether people just felt a little bit more confident in reporting LGBT hate crime, where previously they felt it would not be taken seriously. Is that correct?
Nick Antjoule: Yes.
Q16 Nusrat Ghani: So if you cannot pinpoint that the rise is definitely down to the EU referendum, why do think there might have been an increase in people reporting? What has led them to be confident enough to come forward and say, “These incidents are taking place”?
Nick Antjoule: I will admit that it is a difficult question, partly because of the lack of data. It is very difficult to see. I know that we saw a rise of roughly 147% in the people we were seeing in the three months after the referendum. A bit of context there is that we are not a monitoring organisation in the same way that Community Security Trust or Tell MAMA are. We are not funded to do that. We are primarily focused on policy change and directly supporting individuals, so we are not about creating data, but I think that is probably quite telling that there was something going on there.
The missing piece of information around LGBT people is that with the Crime Survey for England and Wales you have very good sets of data around confidence in authorities for disability, faith and race but, because the survey does not ask those questions around gender identity and sexual orientation, literally, I just do not know. So, without that, I am slightly hesitant to hazard a guess, if I am completely honest, just because crime statistics are so difficult to interpret. I guess, because so few actually end up getting reported, generally, my default is: if more people are feeling able to report, that is probably quite a good thing. Like I say, it is probably a little of both.
Q17 Nusrat Ghani: That is fine. Sometimes when we are dealing with hate crime we can underestimate the power of words. Just to help us understand the kind of abuse that the LGBT community faces, can you just give one or two quick examples of the hate crimes that they face? One or two incidents in the seven years over which you said there has been a rise. It does not have to be the most recent.
Nick Antjoule: Sure. It is very easy to think of all hate crime as violent crime. A lot of it is and has huge impact on people but, actually, a lot of it is people being targeted by people that they know—neighbours, family members, colleagues. To give you an example from both ends of that spectrum, we assisted someone who had been attacked on his way home from work late at night. He had an arm broken. He reported it immediately, which was great, but he was not referred anywhere for specialist victim support services. He was fired from that job—he had a zero-hours contract. He was not entitled—
Q18 Nusrat Ghani: Why was it an LGBT crime?
Nick Antjoule: I am sorry; he was attacked in a homophobic manner on his way back home from work. Because he was fired, he was not entitled to sick pay. He basically went through the whole criminal justice system and from the point of view of criminal justice, everything worked out fine. They identified the suspect. They got him to court. But a month or two down the road, this guy had not been out of his house for a month because he was so terrified. His income had gone. He was on the point of being thrown out on the street. We worked with him to guilt his employer into giving him a job back that he could do with a broken arm and worked with his landlord to make sure that he had a home. We managed to help him to stay and assisted him through that process with things that do not necessarily seem to be criminal justice-related, but actually without them, he would not have been able to give evidence.
In terms of the other end of things, we assisted a trans woman who was being targeted by a neighbour. Her gender history had been outed to neighbours by someone who worked for her housing provider. She ended up being threatened with being shot by a neighbour. She was really terrified. She had some pre-existing disabilities, mental health problems, from previously being targeted, so this being targeted grew into something that was incredibly difficult for her. It was completely life changing. That may have seemed on the surface like a piece of verbal abuse that perhaps other people might have been able to throw off, but the context of her facing transphobia and abuse and violence meant that it caused a huge problem for her. We assisted her by working with the housing provider to make her home safe and to get her moved.
Q19 Nusrat Ghani: I have some quick questions for Mr Lowles. You mentioned that some individuals have become emboldened to engage in race hate crime, because people do not become racist overnight. You also mentioned the difference between people identifying more that they have been victims of race hate crime and coming forward to have that recorded, and the spike having continued and people being victims of race hate crime. We know that, where incidents take place, such as the terrorist attacks in Paris, particular communities find a spike in race hate crime. You also mentioned that the eastern European community was not one that was acknowledged in Government reports as one that could become victims of race hate crime. Do you think, because the communities that have faced race hate crime in this country were easily visible and had been facing race hate crime for generations, they probably do not come forward and have that registered, and that the new community, because they were new to it, were much more emboldened to come forward and say that they had been victims, because it was such a new thing for them to face?
Nick Lowles: No; if you look at some of the other groups listed in the “Hate Crime Action Plan”, many of them are white—many of them are, you know, indistinguishable on the street sort of thing. I think it is partly maybe that they are new communities. I think maybe there is a perception among many that they are transitory communities—that they are here for a short time and then they will go. There is not the same community structures within some of the new communities because people are here to work and in a sense, they have not got the settled community organisations; also, because these communities are not represented, be it in politics, the media, the police or whatever, people do not think about it so much. It is white communities. Certainly I have not spoken to a single eastern European who has not experienced something in the immediate aftermath. As I said earlier, I think it is compounded by this kind of insecurity about their status. I think it is a combination of those things.
Q20 Chair: I am sorry; that is quite a big statement you just said: “I haven’t spoken to a single eastern European who hasn’t experienced something in the aftermath.”
Nick Lowles: Yes. Absolutely.
Q21 Nusrat Ghani: But I assume that the Hope Not Hate organisation probably has not come across a person of colour who has not experienced an incident of hate crime at some point in their lives either.
Nick Lowles: I would not say that that is untrue. I think it is the concentration of cases—obviously, in the course of a lifetime, but we are talking about a very concentrated period when people felt or experienced animosity against them. Obviously, if you are a Muslim, there is a lot of anti-Muslim hatred out there, but I think it was the concentration in the period.
Q22 Nusrat Ghani: You produced a report recently about the collapse of the BNP and the EDL—so the decline of far-right extremism, but the rise of neo-Nazi groups, especially groups such as National Action, which you described in a recent report as “the most ideological Nazi group”. Do you think that because the activities of these groups are so much more horrific—even though the BNP and EDL did terrible things, these groups are so extreme—that somehow the abuse they perpetrate is so much worse? Even though all race hate crime is abusive, I am just trying to find out why these groups must be doing even more damage than the previous groups—because they are more organised, the hatred is more vitriolic and they are just more organised because of social media as well.
Nick Lowles: That last point is key. Their use of social media is quite different. They are younger and they use social media in a different way. If you look at the abuse that Luciana Berger got, we are talking about 2,000 online hate tweets, but it was actually from quite a small number of people. I think that is their way of operating—they operate in packs, wind each other up, all go for a particular target and all just pile in, and often they set up false Twitter accounts, so they are anonymous. So use of social media certainly. If you look at Britain First, for example, here is an organisation that has a massive social media footprint—over 1 million Likes on Facebook—but really it is an organisation of a handful of people, and they use technology and their videoing of going into a mosque or provocative little stunts. For example, I know that with a lot of councils their concern about Britain First is far bigger than the actual threat. That is an issue.
Q23 Nusrat Ghani: May I make a last point about social media? We often talk about the rise of Islamic extremism, but we do not look at the rise of radicalisation of white young men or the role of the media. For this report, the inquiry is going to look at hate crime across many spectrums. I wondered if your organisation looked at hate crime against women.
Nick Lowles: We have not specifically, but what is quite clear, if you look particularly at online hate, is that often there is a connection between race hate and anti-women language. We have seen that in some of the stuff aimed at politicians and celebrities—that is an extra component that is really prevalent.
Q24 Mr Umunna: May I go back to the point you made, Mr Lowles, about people feeling emboldened to express hate? As opposed to there being an increase in the numbers of people expressing hate, it is those who have hate in them expressing it more frequently. Reflecting back to the EU referendum campaign and, frankly, the disgusting and vile videos often put out by Leave.eu, or some of the statements put out by leading Leave politicians, to what extent do you think that the Leave groups and some of the leading Leave politician campaigners bear responsibility for emboldening the groups or people whom you talk about to express their views in ways that they had not before?
Nick Lowles: I think I would answer that by saying that, if you look at it, there was a deliberate strategy by Arron Banks’s group to use emotion—deliberately to highlight immigration, sovereignty, nationality—as a tool to wind people up. Obviously, for the vast majority of people, that was just about voting in the referendum, but I think there was a deliberate attempt to toxify and polarise the debate. I think the use of companies that were brought in by that campaign to deliberately look at how people’s emotions can move—I think that that made what was going to be a difficult and contentious debate about identity even worse. I think there was a deliberate strategy here. It was not to encourage attacks on the street, but that was one of the consequences.
Q25 Mr Umunna: Did it have the effect of radicalising people?
Nick Lowles: Did it radicalise people who were not radicalised in the first place? I do not think there is any evidence of that. What it did was wind people up who already had prejudice or strong views and were okay about the use of violence as a means to an end. I have no evidence to say that new people were radicalised.
Q26 Mr Umunna: You mentioned Mr Banks and his group Leave.EU. What of the other public figures and politicians who made different statements during that campaign?
Nick Lowles: I think that across the board there was a deliberate attempt to raise issues around such things as the Turkish passport. There were Boris Johnson’s comments on Obama. Those things all contributed to an atmosphere where fact and reality in a way did not matter. It was all about emotion. It was all about trying to polarise the public identity. There was also the press, as well. Some of the press stuff was fairly awful. It all contributed.
Q27 Mr Umunna: Can I ask you, Mr Antjoule, whether you agree with what Mr Lowles has just said about the different people campaigning for us to leave the European Union—particularly Mr Banks, his group Leave.EU and the Foreign Secretary?
Nick Antjoule: It sounded reasonable to me, in that it seemed as though the campaign opened the way for people to express prejudice that they already held. If I may slightly change the angle of your question to look at the result of that referendum—obviously we are where we are, but one of the positive benefits of being part of the EU was that it provided a good platform for solidarity and collaboration among organisations that are combating hate crime across Europe. For us, working with European partners is very useful. As part of the negotiations around our leaving, having a platform to collaborate is a useful thing for us.
Q28 Mr Umunna: Can I just ask all of you one final question? Given what we have seen happen since the referendum—obviously this has impacted not only on people of eastern European extraction who I represent, but also on people of colour—do you think that the Government have done enough since the referendum to provide reassurance and to protect those at heightened risk of attack both online and offline in the wake of the EU referendum? Let’s start with Mr Lowles and go all the way down.
Nick Lowles: If we are looking specifically at the eastern European communities—I don’t, but if you look at reporting and awareness, then a whole load has to be done, whether that is about schools or the police. Then there is the wider issue about a lack of security about status, which has contributed. Just going back to your previous question, Nigel Farage and Arron Banks have form on this, if you look at what they have said over several years. They deliberately lie about statistics to wind people up.
Shane Gorman: In relation to disability hate crime in Northern Ireland, since Brexit there has been an increase in disability hate crime being referred to our service, but that coincides with a lot of awareness raising we have been doing with the Department of Justice in Northern Ireland and the Police Service of Northern Ireland, so it would be impossible to say that Brexit is the main reason for the increased referrals to our service. However, we believe that the Department of Justice, the police and a number of key organisations in Northern Ireland are working very hard to address disability hate crime.
Nick Antjoule: In terms of the Government response, there has been a very good response from Government. I know that civil servants do not often get a lot of praise but they have really done lots of amazing work. Debbie Goodier in the Home Office and Paul Giannasi have done amazing work on this issue, driving the work within Government. They have both done great work.
Q29 Mr Umunna: Could I just put this to you, Mr Antjoule? Mr Lowles has just singled out a few politicians who played leading roles in the EU referendum campaign. I just asked whether you think the Government’s response has been sufficient. You agreed with Mr Lowles that the current Foreign Secretary has helped produce an environment in which people felt emboldened to express themselves in ways they did not before. It would seem that part of the Government response has been to appoint as Foreign Secretary one of the perpetrators of the problem.
Nick Antjoule: Obviously, that is up to Government who is appointed to a role. I guess I come at this from a criminal justice-focused place, which is thinking about how to improve the response to people who report hate crime and also how to prevent it. There is the macro-political environment, which is your specialism rather than mine.
In terms of a criminal justice response working with communities, there has been some positive work since. The Government are actually planning lots of great stuff. There were some very big holes in there, which I can briefly mention, if you like.
No. 1 is offender work. There is very little work to prevent hate crime happening. For people who are found guilty of a hate crime there are no offender programmes that work with all different strands of hate crime right now. There are one or two who work with offenders of race hate and, occasionally, around faith. But actually that is the big gap on this issue.
No. 2 is law. There are big gaps around hate crime law. With race and faith there is one system. With LGBT and disability there is a different one. There is a much lower maximum sentence for LGBT and disability hate crime. In some cases it is just a quarter of the maximum sentence. There are other problems with that. So I guess that would be another bit missing from the equation.
No. 3 is satisfaction. I know that we talk about reporting, which is a big and important part of this debate, but actually we tend to forget those people who do report. Once we persuade them to report, what kind of service are they getting? Is it something that leaves them feeling respected and that it was something useful to report? Is it something that they would do again or advise their friends to do? Or is it something that leaves them feeling marginalised?
Sitting suspended for Divisions in the House.
On resuming—
Chair: We will carry on as everybody comes back. Tim Loughton.
Q30 Tim Loughton: Mr Lowles, so far all the examples of hate crime seem to have been attributed to people who were on the Leave side or minded to support the Leave side. How would you regard people who used social media to suggest that anybody who voted Leave is akin to the Nazis and is xenophobic and racist? Is that an example of a hate crime?
Nick Lowles: Do I think all people who voted leave—
Q31 Tim Loughton: No, I said: what do you think about somebody who has used social media to suggest, either directly to people, MPs or others, that those who voted Leave or supported the Leave campaign are akin to the Nazis and are racist or xenophobic? Is that an example of a hate crime?
Nick Lowles: No, but I don’t think it’s helpful at all, and I think certainly—
Q32 Tim Loughton: Why is it not a hate crime to do that?
Nick Lowles: Well, sorry; I was thinking in terms of race hate.
Q33 Tim Loughton: It is a hate crime we are talking about, whether it is racist, xenophobic, sexist or whatever.
Nick Lowles: Is calling someone a Nazi a hate crime? I wouldn’t say it is. I wouldn’t agree with it—
Q34 Tim Loughton: If somebody has tweeted to an MP who has supported the Leave campaign that they are indirectly responsible for the death of Jo Cox, is that a hate crime?
Nick Lowles: I think things like that could well be, because that is very targeted at a person. I am not a lawyer and I don’t know where the definitions are. I think it is deplorable. Whether it would be counted as a hate crime, I don’t know.
Q35 Tim Loughton: Do you think it should be?
Nick Lowles: I think there is a distinction with abusing a group and—I don’t know, to be honest. I don’t know.
Q36 Tim Loughton: Have you investigated examples of tweets and social media messages, such as those which some people might suggest, as I would, are akin to some of the messages that people have been emboldened to use against eastern Europeans or whatever it is, supposedly on the back of the referendum result? Have you investigated that as well?
Nick Lowles: We have certainly looked at—one of the things that our Fear and HOPE survey in the week after the referendum found was an increasing polarisation of society. The key change that happened from February was that there was a sense that some liberal-minded people had become quite illiberal in the immediate aftermath of the referendum.
Tim Loughton: That wasn’t my question.
Nick Lowles: No—we haven’t studied the tweets, and I think—
Q37 Tim Loughton: So you have only studied complaints or social media posts which could be attributed to people who supposedly are in support of the Leave campaign against others, rather than people who might be in favour of the Remain campaign against Leavers.
Nick Lowles: No, we—
Tim Loughton: It hasn’t been one-sided, or has it been one-sided?
Nick Lowles: No, no, I think you have misunderstood the research we did. We were looking at race hate. We are an anti-racist organisation and we were looking at race hate. Whether they were carried out by Leavers or non-Leavers, I’ve got no idea. We do not monitor all hate on the internet. We did a particular study that was looking at race hate, so that is all I am here to talk about.
Q38 Tim Loughton: Okay. What concerns me is that you have used terms such as people being emboldened or wound up. You also cautioned that people who voted Leave who could be emboldened to be racist or xenophobic or issuers of hate-crime messages might be triggered into action if things don’t turn out as they might have hoped after the Brexit vote. Have you considered that people who voted Remain, if things turn out very differently to what they had hoped, might also be triggered to issue more of the sorts of messages of which I have given examples just now, which were all real messages, some of which came to me and, I know, many other Conservative Leave MPs?
Nick Lowles: Our study was based on research by Populus, who I think do the Conservative party’s polling, which looked at people’s attitudes—
Q39 Tim Loughton: Why is that material? Why is that relevant?
Nick Lowles: No, I was just explaining that it was a serious and robust polling organisation. Over a number of years we have been looking at attitudes to immigration and multiculturalism and everything like that, and that isn’t talking about anxieties or concerns, which everyone is very entitled to have. It is looking particularly at those who would see violence as a means to an end. Some of the questions then allude to that. We were looking at this particular group and so we were not certain—
Q40 Tim Loughton: So you do not think—
Nick Lowles: This group voted over 95% one way.
Q41 Tim Loughton: You are not answering my questions, because so far this has been exceedingly one-sided. I think you have a point, to an extent, but you are now suggesting that it is only people who, for whatever reason and some of whom may be complete morons, supported Leave are the only ones whose action could end in violence. Do you not understand?
Nick Lowles: I did not say that at all. I would not say that at all.
Q42 Tim Loughton: Well, that is very much the gist of where you are going.
Can you just qualify a point you made in answer to Mr Umunna just now? You very strongly suggested that the now Foreign Secretary was accessory to effectively emboldening people to be racist, because of some actions or things he said during the campaign. Can you justify and give evidence of that?
Nick Lowles: No. I was saying that it contributed to a toxic atmosphere of the campaign. I have not said that he—I said it contributed to a toxic atmosphere. I am not trying to distinguish between—I am not seeking to try to demonise Leave voters or not. I am looking at where the evidence is leading and if you look at race hate—
Q43 Tim Loughton: But you have only looked at evidence on one side of the argument.
Nick Lowles: I am talking about race hate. There is no evidence that race hate— Race hate is being perpetrated by racists. I am studying this particular group. That is the authority—
Q44 Tim Loughton: Were the Nazis racist?
Nick Lowles: Yes.
Q45 Tim Loughton: So if somebody on the Leave side is accused of being a Nazi or akin to a Nazi because of the stance they took in the referendum vote supporting Leave, that is racist, then, isn’t it? There is evidence for racism on both sides of this argument, but your studies and research seems to have been limited entirely to apportioning all the blame to people who might have supported or suggested they supported the Leave side, which slightly undermines the credibility of what I think may well be a useful piece of research that you have done.
Nick Lowles: I think you are trying to twist words and twist definitions. To be honest, if you are looking at an attitude of a racist on social media—telling someone they should eff this and eff that, they should go home, they should die, they should this, that and the other—I am not excusing or saying it is acceptable, someone calling somebody else a Nazi or whatever; it is not, but I do not think it is the same. I don’t think for a victim of racist abuse it is appropriate for you to try to say it is the same thing, because it actually isn’t. And I think you are trying to score a political point here and that is what I am not trying to do. I am not trying to do.
Q46 Chair: Mr Lowles, can you confirm that hate crime is defined as a “criminal offence which is…motivated by hostility or prejudice towards someone based on a personal characteristic” where there are five strands— race or ethnicity, religion or beliefs, sexual orientation, disability or transgender identity—and where those are the characteristics of the person who is targeted or the victim of the hate crime?
Nick Lowles indicated assent.
Q47 Chair: And can you confirm that your organisation looks at race hate crime and victims who are targeted because of their race?
Nick Lowles: Yes.
Q48 Mr Jayawardena: Mr Lowles, have you seen the John Lewis Christmas advert?
Nick Lowles: I have not.
Q49 Mr Jayawardena: Have you seen the Marks and Spencer Christmas advert?
Nick Lowles: I haven’t, actually.
Q50 Mr Jayawardena: Well, you are missing out on both counts, and I have to say I think this year Marks and Spencer just pipped John Lewis to the post. Both adverts involve emotion to sell their products. The Labour party, as it happens, used emotion in its 2015 campaign. It said, “Next time, they’ll cut to the bone” when talking about the Conservatives with the NHS. I am struggling to see why it is such a surprise to you that some people may have used emotion—your word—in the EU referendum campaign.
Nick Lowles: I think that there were people who used emotion to talk about sovereignty, about who controls laws, where decisions should be made. Totally fine, that’s what politics is about. There are those people who deliberately try to exaggerate, to lie, to use racist stereotypes, to abuse a whole community by definitions around—whether it be criminality or whatever. That is a deliberate attempt to whip up hatred towards particular communities. The people I talked about—and that wasn’t Boris Johnson; I wasn’t saying he is in this category—but the Nigel Farages and Arron Banks, particularly Nigel Farage, have got form for doing this. The whole thing about the 29 million Bulgarians and Romanians that were going to flood into the country in 2014—these are deliberate things.
Q51 Mr Jayawardena: But that is not emotion, is it, Mr Lowles? That is simply someone’s own view of what might or might not happen. It is for the people to judge whether that’s true.
Nick Lowles: And then if you look at the South Thanet campaign, where he was trying to associate eastern Europeans with criminals in some of his stuff, I think there was a deliberate policy of some people to wind up and create a fear and an antagonism towards outsiders.
Q52 Mr Jayawardena: Well, much as you said around the John Lewis and Marks and Spencer adverts, I’m afraid I wasn’t directly involved in the South Thanet campaign, so I can’t comment on that. But you talked about eastern Europeans earlier and you said that people had contacted you because they faced “animosity”—your word, again. But isn’t it true that only people who experienced animosity against themselves would actually contact you in the first place?
Nick Lowles: No, no, no. I didn’t say that they contacted me; I said people I have spoken to.
Q53 Mr Jayawardena: Fine—people you have spoken to. But your organisation deals with people who have faced animosity.
Nick Lowles: l was talking about me on a personal level. People I have spoken to in the street, in my local community, or whatever.
Q54 Mr Jayawardena: Well, I would implore you to visit North East Hampshire, then, because we have very happy community relations. Moving to some standard questions, are we the victims of our own success in the way that reporting has gone up, because people feel more confident, but does that not also lead to a risk that some people abuse that success and feel that they can actually report crimes as hate crimes, even if they are not, devaluing those who have experienced real hate crime?
Nick Lowles: I think that’s a very dangerous line that you’re going down.
Mr Jayawardena: Well, I asked the question.
Nick Lowles: I would say no. I think for someone to contact the police or contact a reporting agency to report a hate crime—I would be very surprised if there was a lot of exaggeration going on. I think the effort to go down that road—what is far more prevalent is that people don’t contact the authorities. When I began I couldn’t say whether I thought that hate crimes were going up or down in the last three years, because there has been so much improvement in how they’re measured, new categories and definitions and awareness, it is impossible to work out. But what I wouldn’t say is exactly what you’ve said there. I don’t think that’s leading to people deliberately trying to abuse the system.
Q55 Mr Jayawardena: I then have this confusion in my mind around why the hate crime prosecutions fell by 10% which, to my mind, in an independent criminal justice system shows that there is lack of evidence proving that these crimes actually were committed. So do you have any light to shed on that?
Nick Lowles: I kind of don’t. I think that they were probably quite difficult crimes to solve or to bring to court.
Q56 Mr Jayawardena: Would any of the other witnesses like to comment on the decline in prosecutions and a reason why that might be the case, or is it, as I might suggest—I might not suggest, but for the sake of argument I do—that people are defining crimes as hate crimes even if they weren’t hate crimes?
Nick Antjoule: I guess the definition of a hate crime is a criminal offence that is motivated by hostility against someone because of their perceived membership of a group. Within those strands but also, a court is empowered to treat other things outside hate crimes as hate crimes too. In terms of people who are targeted, the target of crimes where they feel it was motivated by prejudice against them, then that, too, fits within the hate crime framework. So—a bigger picture. A tiny, tiny fragment of hate crimes that are committed end up being recorded. Within the LGBT community it is roughly a fifth of the estimated level. That is estimated by very good data from the crime survey for England and Wales. So I think you are right to raise it, but actually the massive scale and reporting mean that that’s the big picture, here.
Q57 Mr Jayawardena: The Government have not yet responded to the Law Commission’s recommendations for new guidance from the Sentencing Council on recording enhanced sentencing, which is increased sentencing due to hate crime hostility. In the view of the witnesses, should those recommendations be adopted?
Nick Antjoule: I am sorry; could you repeat the question?
Mr Jayawardena: Essentially, enhanced sentencing means that there would be increased sentences due to hate crime, alongside the crimes that had been committed. That is the Law Commission’s recommendation. The Government have yet to respond to those recommendations. Would you agree with the Law Commission’s view?
Nick Antjoule: I think what the Law Commission report recommended was that there be a review of this disparity between the different strands of hate crime, and that either a plan be put in place to remedy the disparity between them or, in the absence of that, action be taken to remedy it. That was a couple of years ago now. I think it still needs a response, in terms of the gap between maximum sentences. The fact that homophobic, transphobic and disability hate crimes tend not to be recorded on somebody’s criminal record, because there is no offence that can be named there, means that courts cannot see if someone is a serial hate crime offender. Probation services cannot see the prejudice driving someone’s action so they cannot take proper action.
I guess in answer to your question, right now there are rules there within sexual orientation, gender identity and disability hate crime, to enable courts to give an uplift in sentencing. They are already there. Within race and faith hate crimes, it is kind of baked into the criminal offences that they are available. I guess what is behind your question is the disparity. I think that yes, that does need remedying.
Q58 Mr Jayawardena: And as you say, recording on the PNC as well.
Nick Antjoule: Yes.
Q59 Mr Jayawardena: And I suspect the other two witnesses would agree.
Shane Gorman: In Northern Ireland, enhanced sentencing is available for disability.
Q60 Mr Jayawardena: Because of the different criminal justice system.
Shane Gorman: Yes. As you touched on, it is very hard when it gets to evidential value in court to prove that and to apply it for enhanced sentencing. However, I think that is in relation to the perception test, which the Police Service of Northern Ireland work with in relation to disability hate crime. If something is perceived as a disability hate crime, it is reported and recorded as such. It is then up to the officer to obtain the appropriate evidence and then to pass it to the PPS to decide if it can be applied for enhanced sentencing.
Q61 Mr Winnick: Mr Lowles, on hate crime, is it not the position that it has increased in the last few years? The police figures that I have are that in 2011-12, there were just under 36,000 incidents. In 2015-16, that figure was just over 49,000. Unfortunately there has been an upward trend—that is the situation. You wouldn’t wish to dispute those figures?
Nick Lowles: No; but I suppose what I was trying to say is that over the last few years there has been much greater awareness and police forces around the country have put in totally new processes. I don’t know how much that increase is due to the fact that there has actually been an increase or other factors, such as increased reporting or better police methods of community engagement. I can’t say.
Q62 Mr Winnick: If the argument is that the decision taken by the majority in the referendum on 23 June has resulted in an increase, and there may well have been an increase, would I nevertheless not be right in saying that the figures I quoted, over five years, demonstrate that the upward increase has very little if anything to do with the referendum decision?
Nick Lowles: If you look at the spike and the month-on-month comparison between June and July 2016 and June and July 2015, there was a massive jump, which suggests something happened in that month to cause that spike.
Q63 Mr Winnick: But that hasn’t continued, has it?
Nick Lowles: No, it has dropped down again. Again, it is about the incident. Something happened in that period that caused the spike. If you look at the day-by-day data, it all happened in the immediate days after the referendum result. That was the spike. It has gone down, but it is still above the year-on-year average of 12 months ago.
Q64 Mr Winnick: Over five years there has been an upward trend.
Nick Lowles: Yes.
Q65 Mr Winnick: You said that you had not met an eastern European since the referendum result who had not suffered some abuse.
Nick Lowles: Yes.
Q66 Mr Winnick: Could you tell me how many eastern Europeans you have actually come across?
Nick Lowles: It is certainly in the scores. That is me personally having met and interacted with them—I haven’t gone to seek them out. They are people I have met through my work or my daily life.
Q67 Mr Winnick: I am sure. You don’t make these things up as you go along, but could you give us the figures for the numbers of Poles who reside in the United Kingdom?
Nick Lowles: The numbers?
Mr Winnick: I have got the numbers, so I will not try to say that your information is incorrect. I will put it differently. Broadly speaking, there are 831,000 Poles in Britain. They are the largest group. Eastern Europeans amount to 2 million. Apart from some very unfortunate incidents and one very tragic case in which a Polish person was murdered in Harlow, is there really any evidence that the overwhelming majority of people who are Polish or are in the broad category of eastern Europeans are suffering persecution, discrimination or abuse on a daily basis as a result of the decision taken on 23 June?
Nick Lowles: First of all, I never said it was on a daily basis. Secondly, I think it is indicative that the Government and the police do not have any data at all about the numbers of hate crimes targeting eastern Europeans. One of the points I made earlier is that that is one of the gaps in the system. But I put it back on you: you can go and talk to any eastern European community organisation in the UK and you will not find one that has not got anecdotal evidence or some sort of evidence that there has been an increase. Everyone I have spoken to has sensed it, and I cannot believe that I have spoken to the only people who felt it. What the wider level is, I do not know, but one of the problems is that, while great advances have been made in monitoring certain types of hate crimes—the police and the Government have upped their game over the last couple of years—there is a black hole of data when it comes to European communities.
Chair: I am conscious that we may soon have another vote and we need to move on to our second panel. One last question, David, and then we need to move on.
Q68 Mr Winnick: Obviously, if one single person is discriminated against because they are eastern European, because of their sexuality or for any other reason, that is one too many. But I put it to you, Mr Lowles, how important it is not to exaggerate. I accept entirely the extra work done by your organisation and those of your colleagues, but would it not be right to say that all the evidence seems to point to the fact that, despite the abuse that has occurred towards eastern Europeans, which we all deplore, the overwhelming majority in Britain go about their lives in just the same way as before the referendum result.
Nick Lowles: Absolutely. I am sure that is the case. But, anecdotally, when I was in a meeting with a Government Minister we were told by a woman from the Polish community in Birmingham that unfortunately hatred and animosity against them are accepted as part of life. They accept that and get on with it.
The next few years are going to be quite fraught. Without taking a political view one way or the other, there will be increased anxiety and, potentially, increased tensions over the next few years. The Government and the police need to put in a more robust system to work with communities and faith organisations to make sure that, if abuse does happen, people have a route in and the confidence to—
Q69 Mr Winnick: I agree with that, I assure you.
Q70 Stuart C. McDonald: Mr Gorman, earlier on you spoke a little bit about the work that Leonard Cheshire are doing to overcome barriers that exist for people with disabilities to report hate crime. You also mentioned work that you are doing with the Department of Justice and the police. Could you tell us a little about the sort things that they are doing to try and overcome these barriers and what else would you encourage them to do?
Shane Gorman: The advocacy service was formed in partnership with the police service in Northern Ireland so they would actually sit on our disability hate crimes steering group, which is attended by the Department of Justice, the Northern Ireland Policing Board and a number of disability charities. That is the steering group for the direction of the disability hate crime advocacy service. We provide direct support to individuals in terms of reporting hate crime to the police and providing support throughout the criminal justice process.
In terms of indirect support, we also work to raise awareness of disability hate crime throughout the statutory, voluntary and community sectors of Northern Ireland. Every six weeks I would actually go to the police training college and deliver a community safety session on disability hate crime to all new recruits. It means the police service is very well equipped to deal with victims of disability hate crime, and they recognise the barriers that certain victims will come up against when reporting a crime and then take a part in the wider process.
Furthermore, we have embarked on a number of promotional campaigns. The Department of Justice supported our Get Support to Report campaign which was to highlight that anybody in society can report a disability hate crime. Recently, we also supported the police in their #Voices campaign, which was victims of all strands of hate speaking about their experiences. That was done through social media.
Q71 Stuart C. McDonald: You mentioned that essentially Leonard Cheshire operates as a third-party reporting organisation. What are the procedures for that? There are one or two reports that maybe there isn’t enough awareness about these organisations. Is that a challenge that you face? Consider procedures first of all, and then what you can do to overcome the lack of awareness of these organisations.
Shane Gorman: Through our service level agreement with the police service they will accept reports of disability hate crime from ourselves. Also, through their own policies and procedures, they will accept reports of hate crime from anybody as hate crime can be reported by anybody.
In terms of how we follow up with victims of disability hate crime, it depends very much on the individual and their individual circumstances. For example, if somebody who was deaf wanted to report a crime to the police, which has happened, they have contacted me either via social media, email or text message and I am able to facilitate a meeting at a police station with and officer and an interpreter so they can fully take part in the process.
Q72 Stuart C. McDonald: Mr Antjoule, you also spoke quite warmly about some Home Office action that has been undertaken to overcome barriers to reporting. Can you say a little about what that work includes?
Nick Antjoule: Yes. There is quite a lot of it. One of the things that the Home Office has funded is an online hate crime hub, which is a very positive development. Very little online hate crime is reported; very little is made visible, but there is a huge amount out there. That is one of the big challenges. It passes through lots of jurisdictions and getting some kind of result is difficult. The thing that is lacking right now is community response alongside that. I know that the Home Office are doing some counter- narrative work, working with communities to put a different point of view online, to challenge prejudicial attitudes, which is very positive.
One of things that I am keen to address is satisfaction levels. There is a big gap in levels of satisfaction between those who report hate crime and those reporting other forms of crime. Three quarters of people who report crime say it was a positive experience—it was useful. With hate crime, it is roughly half. There is this huge gap of people making that leap of saying, “I am going to make this very difficult decision to come forward and report hate crime” and then feeling it was not useful or that is was positively difficult for them. That is one of the things I would really like to see Government action on, asking how do we improve that service, how do we make it more victim centred, how to do actually meet people’s needs?
Q73 Stuart C. McDonald: May I also ask you briefly, I think Galop is also a third-party reporting mechanism. How does that work in practice and do you have problems increasing awareness about that among the LGBT community?
Nick Antjoule: Yes, in practice. Essentially, we have people coming to us saying, “I’m thinking about reporting and these are the pros and cons. Is this something that is going to open me up to prejudice? Is it something that is wasting police time?” We will talk them through the process and say, “Let’s think about what is involved.”
Usually, we would either directly make a report for them—so send something to police on their behalf that could be recorded—or, like my colleague here, have them sit down with a police officer who we know will understand some of the identity issues that are tied up with the people we work with, to help them make a report directly.
Q74 Stuart C. McDonald: Mr Lowles, in terms of race hate crime, is the Government doing enough? What more should it be doing to overcome some of the barriers that there are to reporting?
Nick Lowles: Beyond what I have said at the moment, I think there remains a discrepancy in how police forces around the country deal with the situation. Some greater standardisation still needs to happen. There could be still more done on increasing access to and the number of ways people can report hate crime into the system.
Another practical thing is schools. First, we are anecdotally hearing about an increase in hate crime and race hate in schools as well as prejudice and so on. There are particular forms of online bullying and hate that happen among young people. More needs to be done to train teachers or raise awareness at a school level, which I think the Government should do.
Q75 Naz Shah: I have two questions for Nick Lowles. Do you think the Home Secretary helped to take the heat out of issues by suggesting that companies should publish the number of their foreign workers? In the light of that, do you have confidence in her capacity to address these issues?
Nick Lowles: I think it was a political error to do that. In the wider context, what she was trying to do, I don’t know. Certainly, the way that it was interpreted and the way that it added to a toxic feel, it was completely wrong. Do I have faith? I think good things are happening at the Home Office. I would hope that she has learned from that and will not make that sort of error again.
Q76 Naz Shah: Following on from Mr Umunna’s earlier question about the role and responsibilities of politicians, we talked about social media, but what about mainstream media? I am referring in particular to the campaign for the London Mayor when there was a front page of the bus from 7/7. Where do you think the responsibility lies and what should we be doing? What is your biggest worry?
Nick Lowles: Clearly, the mainstream media have a platform to speak to millions of people. I’m sure I was not the only person shocked at the IPSO decision not to take action against Kelvin MacKenzie for the article he wrote about a Channel 4 presenter.
We are in a situation where the press continue to police themselves and that is clearly unacceptable. People should be held to account for their actions. I strongly believe in freedom of speech but with that comes responsibility. As we saw with the judges after the High Court decision, there is a culture of bullying that goes on. The way the media took up the London Mayor thing and some of the stuff around the referendum and the immigration issue generally—not just in the referendum campaign—adds to the toxic atmosphere.
Q77 Chair: We are about to move to our next panel, but I have a final point to clarify. In terms of some of the evidence and the points made by everyone, Mr Lowles, you referred to some research that you had done, but that you were not able to publish until after the trial for Jo Cox’s murder has taken place. You also referred to some analysis that has been done of—I am not sure if this online—some abuse targeted at eastern Europeans, some welcoming Jo Cox’s murder, and some not targeted. Is that part of the same evidence?
Nick Lowles: Yes.
Q78 Chair: Okay. Obviously, when that is published, we would be very interested to see it. Also, do you or anyone have any further evidence? I think you referred to incidents in school as well. We will also be taking further evidence about the impact on eastern Europeans living in Britain, because clearly that is an extremely important point that has been made.
But let me ask you finally, before we have our panel on social media, what has all your experience been of the impact of social media? First, have you evidence of increased hate crime online? Secondly, have you any evidence of online abuse having an impact offline? That is a final question for you.
Nick Antjoule: My general feeling is that online expression of prejudice and hostility is a very big problem. It is an international problem. One of the issues in our response to it is that we have police forces—I think we are very unusual in that we 30, 40 or whatever police forces, where other countries just have one—and I think that is your barrier to dealing with this issue, because it is a UK problem and an international issue, so finding a response to that is very difficult. I know you have Demos coming to speak to you this afternoon, and they have done some very valuable work. I would really like to see some investment in algorithms to find out about hate crimes, or online incidents of hate speech that look like they might be offences that originate in the UK—perfectly doable—and to eliminate this problem around reporting. We could see them and look at whether there are some offences there that we want to be targeting. Right now we are relying on reporting, and very few of them—a tiny proportion—end up with a positive result, because there are so many difficulties. Targeting, in a way, is far more effective. But alongside that, work with community organisations. Criminalising it is not going to be a good response. In the vast majority of cases, empowerment of communities to speak up and to put other points of view is the longer-term thing that needs to go alongside that.
Shane Gorman: In relation to online abuse, regarding victims of disability hate crime, it is something that has occurred. It would not be massive numbers in relation to the number of victims whom I have supported personally. However, if it does happen, we encourage it to be reported to the police and to be treated as a crime and an incident by the PSNI. However, the effect of online hate crime can obviously be greater, as someone has been targeted in their own home. It is bad enough for someone to become socially isolated, to be cut off in their own community if they are maybe being verbally abused or harassed because of their disability when they are out and about, but when they are at home and people can target them through their computer, it can have a great effect.
Nick Lowles: More work has to be done. I welcome the police setting up the hub to look at online hate. I personally experience it quite often. After we helped to stop Pamela Geller and Robert Spencer coming to the UK just after Lee Rigby’s murder, in a 24-hour period I received over 2,000 hate messages on Twitter. I can tell you, I have done this for 25 years and I have a fairly thick skin, but after a while it does grind you down—and that is as an activist. I have had offline hate as well—I had a pig’s leg left on the doorstep, and various other things.
My experience is that some of the worst perpetrators act in groups. It would be really interesting to see some work on it—the police trying to look at some of the trolls—because I think you would find that for some people it is a kind of semi-profession. More police intervention has to be made against those people.
Q79 Chair: Do you think that there is a relationship between online abuse and offline hate crime?
Nick Lowles: I don’t know. Obviously, it is far easier to do online hate—you can set up a false account, you can abuse people, you can follow people and the more that they react to it, the more you wind them up. In cases of National Action and some of the political groups, I think that that then does translate—as we saw with some of the harassment of some MPs—but for most people, it is just a far easier way of carrying it out, particularly the targeting of individuals, which is the really awful thing, because it is persistent and the more you show it is hurting you, the more they go after you. Unless you have experienced it, it is hard to understand how miserable that is.
Q80 Chair: This is the issue that we will pursue in the next session, but we started this debate with a discussion of increasing hate crime, and the different levels of increasing hate crime. Is it your perception, though, that online hate crime—in terms of things that would actually count as crimes, in addition to online abuse—has increased in the last few years?
Nick Lowles: Yes, I think online hate crime has. I really do not know about offline, but online hate crime has increased.
Chair: Thank you very much, all of you, for your time. We hugely appreciate the time you have given us today, and the immensely important work that each of your organisations does in tackling hate crime. We thank you for that important work, as well as for giving evidence.
With perfect timing, we will adjourn for 15 minutes and then come back and take evidence from our second panel. Thank you.
Sitting suspended for a Division in the House.
On resuming—
Examination of witnesses
Witnesses: Carl Miller and Dr Pete Burnap.
Chair: Thank you very much for coming to join us. Please will you quickly introduce yourselves?
Carl Miller: I am Carl Miller, research director of the Centre for the Analysis of Social Media at the think-tank Demos.
Dr Burnap: I am Pete Burnap, the director of the Social Data Science Lab at Cardiff University.
Q81 Chair: Can you start by giving us a picture of what you see as the main trends in online abuse, hatred and extremism?
Dr Burnap: Sure. Our studies have focused primarily on event-specific reactions. We have found that online hate spikes following events. For example, one of our major studies was on the killing of Lee Rigby; we saw a definite spike in online hate at that time. We have looked at the duration—how long that lasts in social media, and in particular Twitter. We have found that the production of hate drops off very quickly following the event—24 to 48 hours after the event, the extreme hate reaction drops away. We have also looked at the actors who are involved in the narrative and who “lives longest” on Twitter, so from the point where something is published to the point where it no longer is propagated and people stop retweeting and publishing it. We have found that police and far right activist groups were the longest living compared with other actors.
Q82 Chair: When you say “other actors”, what is your differentiation between them and the far right groups?
Dr Burnap: We looked at a collection of tweets and manually annotated them. We looked at media outlets, other political parties, celebrities, far right activist groups and the police, and others would then fall into an “other” category.
Q83 Chair: How do you define the category of hatred that you look at? Is it just those that would fall into the CPS’s category of potentially criminal abuse, or do you look more widely at hatred?
Dr Burnap: We look more widely. We tend to differentiate between hate crime and cyber-hate. The cyber-hate definition we use is where the language is antagonistic or inflammatory based on protected characteristics. Those cases perhaps would not always fall into the law of hate crime. The law is very oriented toward threat and a likely outcome of the threat becoming a reality. We tend to take more of a victim-oriented perspective—that something would be offensive or antagonistic to people who fall into those groups.
What we measure varies in levels. We look at extreme levels and obvious antagonism, and we also look more broadly into what we might call “othering”. We look at anti-immigration narratives and, as was alluded to earlier, the “Us and them”, “Send them home”, “Get them out” and “Told you so” in relation to immigration policies and so on, as well as, to some degree, calls for collective action as well—“Get them out” and that kind of stuff. We look at it more broadly, rather than within the specific legal definition of hate crime.
Q84 Chair: Putting aside the spikes of trigger events and things, have you noticed long-term trends?
Dr Burnap: To be honest, we have not measured long-term trends. One of the main reasons for that is the sheer volume of data that would need to be collected to do long-term trends. We are looking at this at the moment with the Community Security Trust and we have collected data for a 12-month period with specific terms. We are in the process of extracting of the order of tens of millions of tweets over that year. We need to pull that apart and look at that over time.
We specialise in the development of algorithms to detect the hateful narratives: cyber-hate at scale across millions of tweets. We know there is an issue with under-reporting. We know there is an issue with policing this in a broader sense, not necessarily within the legal system but monitoring and curtailing it. We have used machine learning—training a machine to recognise the signals of cyber-hate—so we can categorise it at scale, which humans simply would not be able to do.
Q85 Chair: What are the particular events in the last couple of years that have triggered increases?
Dr Burnap: The main one we looked at was Lee Rigby. That had a massive impact. We have seen a large spike in the data we have collected following Brexit. It is very context-dependent as well. There was the Ken Livingstone incident recently, which was a massive spike in the collection we are looking at for the CST. It is largely driven by events that are reported and the online reaction to that in the immediate aftermath. The US elections made a large spike in the volume as well.
What is important for us as researchers—our challenge now is to dig into this volume of data. We know that the volume is there—the volume of traffic, the responses to these events—but we do not yet know the volume of cyber-hate within that. We are also keen to pull apart the cyber-hate we are identifying and break that down into what the narratives are around this. Are they immigration-based? Are they formed around other opinions and beliefs and so on?
Q86 Chair: Of the events you have looked at so far, what proportion of the response you have looked at has been an increase in hatred and what proportion has been an increase in counter-speech?
Dr Burnap: The volume of hatred we typically see is between about 8% and 11% of a sample. We tend to take a sample to try to reduce the volume. It is around 8% and 10%. We do not have volumes of counter-speech. With counter-speech, we have looked at how it has a role to play in the stemming of cyber-hate. What we have found is that the more people actively engage in counter-speech, the more it reduces the continuation of that narrative. For example, if one, two, three, four, five people start getting involved in trying to call to order or suggest this is not acceptable, that tends to lead to longer drawn out interactions around cyber-hate and it tends to get more extreme as well, but when more people get involved, it tends to stem that. That is the finding at the moment.
Q87 Chair: What impact do you think the EU referendum had?
Dr Burnap: I cannot talk specifically to the EU referendum. I think Mr Miller here might have more data around that. All I can say at the moment is that it has led to a spike in the data we have collected and we are collecting terms that would be likely to be indicative of cyber-hate, but we have not yet measured the volume of cyber-hate due to the sheer volume of data and also because we are trying to look at this over time.
Q88 Chair: In the incidents that you looked at, do you detect patterns and networks behind abuse or is it very random?
Dr Burnap: We have certainly seen patterns in that there is normally the same actor type getting involved—the far right narrative is normally there, the police narrative is there and the media narrative is there. A previous panel mentioned that there may be collective action—for example, organising around political orientations. We have not measured the network effect or the social capital effect in terms of networks themselves, but we have found that where people have high social capital—on social media, that may be a high number of followers—that certainly leads to an increase in the propagation of their narrative. The more people follow individuals or actors, the more their information is likely to be propagated.
Q89 Chair: Mr Miller?
Carl Miller: All of that.
Chair: I will prompt you. Starting from the beginning, what trends have you seen?
Carl Miller: It is important to say at the outset that measuring social media in the way that we are presenting evidence of to you now is a new and young discipline. Alongside trying to find the trends and present stuff that is actually useful for policy makers, we are also trying to build a way of doing the research in the first place. We do not have as big a picture as studies into offline hate crime, because we have only been able to measure with some degree of appropriate accuracy and robustness the amount of online hate crime—and only then on certain platforms—for quite a short amount of time.
We began to measure Islamophobia on Twitter in March, and we have been running that all the way through since. The latest numbers that we have are for the end of July. From that reasonably small time window, we can say that July was the highest month on record and globally, Islamophobia tended to increase over that time period. I would say that that is not exposing a long-term trend, although I think that long-term trend is there, but that evidence very much agrees with my colleague’s point that important events actually drive the phenomenon of online hate to a huge degree in the first place. Social media is incredibly event-specific and reactive, and the vast amount of the hateful stuff that you see either coincides with a significant event that has happened in the offline world or is explicitly related to it in some way.
The bigger picture, as the previous panel alluded to, is that it can only be that online hate, much like almost any other practice that you see happening on social media, is increasing. It is increasing because social media is increasing. In the space of just a few years, we have gone from 250 million tweets a day to 500 million. We have seen the explosion of platforms like WhatsApp that were not previously being used at all. In a short time, social media use has become a prevalent practice in society. As people have jumped on to those platforms, they have brought both good and bad practices with them. The big picture has to be that online hate has increased, along with all the other things that have increased on social media.
You asked about the impact of Brexit. There are two lots of numbers I can present to you on that. First, we did look at Brexit for xenophobia and the prevalence of xenophobic attitudes expressed on Twitter over that time window and slightly more broadly. Some 16,000 tweets that we measured used a xenophobic word or hashtag—a hashtag that we and hate crime researchers had compiled as one which could be used in a xenophobic way. Of those, over 10,000 were not xenophobic when we algorithmically analysed them; they were actually people going to Twitter to take on xenophobes and bigots, declare support for victims, and so on. Around 5,000 were indeed derogatory and xenophobic.
Alongside that—both these pieces of evidence point to something that is also really important when it comes to social media—we saw 17,000 tweets that were either sharing or relaying accounts of offline experiences of hate. These were victims sharing, and presumably in some sense trying to validate, the experiences they had had or seen offline with others on Twitter. Of course, both those things point out that social media is also a really important forum for victims to come forward and find ways to publicise their own experiences. I think that brings me up to date with the questions.
Q90 Chair: You have both talked about events triggering responses. You have talked about the Ken Livingstone incident and the Brexit incident. Does that mean that the way in which campaigners, politicians or the media report an incident has an impact on the kind of hatred or vitriol that follows online?
Carl Miller: I think there are two things there. Yes, it does. When you look at the nature of online hate and the kind of things it contains, a very large amount contains links and a lot of those links are to mainstream news websites. A whole kind of media ecosystem has built up around hateful networks. Often, those are echo chambers—this is a really important thing, which perhaps we can discuss later—of likeminded individuals who all talk with each other, and continually confirm and validate each other’s views.
You often see websites that cherry-pick mainstream news coverage and puts it all in one place, which is basically a wall of confirmation for a particular viewpoint. The media certainly play an enormous role in all that. Powerful digital voices, some of which are powerful only on social media and some that have come from the offline world, also drive a lot of conversation. Social media is an incredibly unequal space. Some people are much louder, much more highly followed and much better networked than others.
Another thing is also true, though. You can say that the mainstream media, politicians, celebrities and so on importantly frame events for people, but social media also allows people to frame those events themselves and to write their own meaning on to those events. For the Brexit vote, there was a phenomenon where lots of people, or at least, reasonably large numbers of hateful people, were declaring on social media what they felt the Brexit vote meant—what it was really about and what it meant for them. At all times, that was at distinct odds with the mainstream campaigns or most of the mainstream media coverage.
Q91 Chair: Do you think that the things that were said or the way in which things were said by mainstream campaigners in the case of Brexit, in other incidents or even in the US elections had an impact? Did it give licence to those who already held prejudice to voice it online, or not?
Carl Miller: To answer that fully, I need to jump into the heads of the people who are conducting the hatred, which is really difficult to do. Empirically speaking, there is a link. People who are sharing hatred are also linking to speeches of politicians, mainstream news coverage of events, controversial columnists and so on, so from an empirical point of view, yes, there is certainly a relationship between those two things. Whether it is driving it, whether it is permissive or whether they have gone out there and wanted to find something that confirmed their beliefs in the first place is too difficult to say.
Q92 Chair: Can you say anything more about the organised nature of it? Are there particular groups that might be provoking or encouraging it? I am thinking particularly about some of the targeted abuse, as opposed to people who are almost just shouting at the world.
Carl Miller: Social media has never allowed networks to form more easily, which has meant that all kinds of networks are formed that otherwise would not exist. Some of those networks are entirely new kinds of political mobilisations and groups—Britain First was mentioned earlier. There has been a massive fracturing, in a sense, of extremist opinion into much smaller, more amorphous, more organically forming groups, which are constantly coming together and breaking apart online.
There are also trolling networks. This is not, I think, predominantly a solo pursuit or hobby. These people talk to each other a lot. They have websites where they discuss trolling. They share their latest wins, which is when they get a reaction out of the person they are going for. In the strange world of internet subcultures, trolling has long been a recognised art form, believe it or not. It was always seen as the art of trying to—I will not use the words they use—really annoy someone and get a rise out of them. Certainly by some it is still considered in that vein, which means that they will constantly try to build their reputation with that subculture.
Q93 Chair: A lot has been written recently about the alt-right movement online. What is your view on that? Does it concern you?
Carl Miller: It is hugely concerning. The earlier point that we have not spent enough time thinking about the radicalisation of young white men online is entirely true, but that is not because the alt-right, the far right and other right-wing networks have not been on social media for a long time. They were the first people to join. [Interruption.]
Chair: There is a Division. We will adjourn for 15 minutes. To give you notice of the remaining questions that I was going to ask before I turn to the rest of the Committee, I want to ask you about not just the alt-right but the Breitbart website and some of the websites and networks that have been associated in reports with increased hate crime or online abuse.
Sitting suspended for a Division in the House.
On resuming—
Q94 Chair: Sorry, votes are occupational hazard. Can you tell me your thoughts on whether the alt-right movement, and whether things such as the Breitbart website, have increased first, online racism, and secondly, targeted online racist abuse.
Dr Burnap: May I earlier return to your question from on, in particular on the role of the media and, I suppose to some degree, the use of evidence? What we have found in our studies is that, when social media posts are published, we looked at the number of media releases with a similar topic in the headline, to look at the likelihood of that information being propagated. We found that media headlines around a topic certainly correlate with an increase in the propagation of social media posts. To some degree, that would suggest that there is some shaping of what is being propagated and talked about.
I hate to keep coming back to what we are currently doing, but this is such a nascent area that we are still looking to produce evidence on this. What we are currently looking at is to what degree the topic of the story in the media has any correlation with the narrative on social media—whether there is any kind of stirring up of fear or anxiety around particular groups, whether that is then evident in the social media stream and whether that has any impact on the narrative. As I said, to add to the propagation factors of social media, when a post includes URLs it is much more likely to be retweeted than ones that don’t, which would suggest links to evidence, further information, further discussion or maybe links to media. Again, we are looking to pull that apart and to look at what types of hyperlinks. Certainly, the use of hyperlinks to extend the discussion into the broader web is significant in how much information is propagated in social media.
Q95 Chair: Okay. Do you have any specific thoughts on the impact of the alt-right movement or things like Breitbart?
Dr Burnap: I have no evidence on that, sorry.
Carl Miller: I am sorry. We are just beginning to look at the US context now.
Q96 Chair: Do you have any sense of how much things that are happening in the US have an impact in terms of affecting or being targeted at individuals in the UK?
Dr Burnap: As individuals is difficult, I think. Because of the volume of data that we get, and perhaps the lack of data on reporting on an individual basis, it is difficult to look how at broader global events have a relationship with individuals. Certainly, as relationships to groups, we studied previously the last presidential election—2012—and saw an increase of global cyber-hate towards people based on protected characteristics. Geolocation is quite limited on social media, but we saw that occur in different geolocations. That was definitely the case.
Q97 Chair: We have seen, with some of the consideration around antisemitism and some of the consideration around some of the misogynist attacks that have been targeted at particular individuals in the UK, some links to networks in the US. Have you looked at any of those kinds of issues? It’s okay if you haven’t. I don’t want to ask you about things you haven’t looked at.
Carl Miller: Yes and no. The spread of information—hate or otherwise—is of course not constrained by geography whatsoever, so making geographical distinctions is not normally the way we would go about doing that kind of research, unless we were trying to measure the impact on a particular geography. I might speculate, rather than offer hard evidence. If it is true to say that important events in general drive online hatred and, of course, counter-speech and everything else, and if online hatred is moving between countries more easily than it has before, I would be astonished if the US election has not had an impact on it. For us, the most important events we see that have driven that kind of hatred are terrorist attacks, but major political events are certainly up there as well.
Dr Burnap: One of the major points about evidence for this kind of discussion and the relationship between global events and targeted hate in other countries is the limited ability to geolocate where social media posts come from. In our studies, specifically for Twitter, we have seen that around 1% to 2% is geolocated. That is a very small amount of data to use when looking at where the hate is either coming from or being targeted at.
Q98 Mr Jayawardena: Your data, Mr Miller, showed that on 24 June UK tweets containing a word or hashtag related to xenophobia were five times more likely to be supportive or contesting racism than actually being xenophobic. Is that right?
Carl Miller: That is right.
Q99 Mr Jayawardena: Even though you identify that supportive tweets peak and fade, while derogatory tweets tend to maintain a steady volume most of the time, there were twice as many supportive tweets across your sampling period of 19 June to 1 July. Is that right also?
Carl Miller: Yes.
Q100 Mr Jayawardena: You count tweets rather than accounts, but do you have any indication about the volume of accounts? Are we seeing a very small number of very active, very xenophobic individuals making hundreds of tweets, or is it reflective of a more widespread problem?
Carl Miller: That is a really good question. For that particular sampling period you referenced, no we haven’t. Broadly speaking, you are right. The rough ratio of tweets to users tends to be, in the research we have done, about three tweets per person over a month. They are generally repeat users. But within that, typically you get something that looks a bit like a power law: a small number of people are responsible for a vastly disproportionate amount of the actual content. If you look at the most prolific 25 users within a given period, they will be responsible for hundreds, if not thousands, of these messages. That is prevalent across pretty much any social media dataset that you collect.
Q101 Mr Jayawardena: Turning to another set of hashtags you have looked at, there are 98,948 tweets using the hashtag #postrefracism or #safetypin. Only 17,000 of those, though, actually provided specific hate crime incidents. Are we blowing this issue of social media hate crime out of proportion?
Carl Miller: Those hashtags were not entirely dedicated to simply reporting offline incidents. I should have caveated before that we did not actually verify whether those reports were true; they may not have been. We were trying to work out the extent to which Twitter was being used as a forum for people to report offline hatred. “Postrefracism” was used, especially as a hashtag, for lots of things. It was used to simply raise awareness of an increase or reported increase in hate crime over that period. Of course, most of that dataset probably was not going to be about offline accounts in the first place.
If you compare those 17,000 that algorithmically were considered to be so with hate crime figures over the same period, you have a really significant forum for the reportage of offline hateful incidents. Of course they are not verified, of course they are not necessarily crossing the criminal threshold and certainly they are not prosecuted, but in circumstances where, as we have heard today and as is well recognised, the general reporting of hate crime to the police is a massive, massive issue—such a big issue that we do not really know at any given moment whether it is the reporting or the amount of hate crime that is going up—the emergence of Twitter for this kind of purpose is massively significant.
Q102 Mr Jayawardena: But although 941 hate crime incidents are still far too many, that is less than 1% of the overall discourse of the 99,000 tweets using those hashtags—
Carl Miller: Sorry, I don’t understand where we are getting that 941 from.
Mr Jayawardena: The statistic of 941 reported hate crime incidents has been published. I do not have the source in front of me, but that is less than 1% of the 99,000 tweets that were made using that hashtag. The point I am coming back to is whether the presence on social media is far too overbearing. Is it over-egging what is still a problem, but is disproportionately reflected on social media because of the echo chamber that you have talked about and the fact that people want to participate in a debate about something rather than reporting something that has really happened?
Carl Miller: All of those things are happening at once. Plenty of people are—in varying degrees of respectful and disrespectful—debating with one another online. That 90-odd thousand figure that you quote from my research is about the number of people who were simply using those hashtags over that date. Within those hashtags, 17,000 were sharing an offline incident of alleged hateful abuse in that period.
Q103 Mr Jayawardena: As you say, alleged—that may not have been reported to the police.
Carl Miller: Yes.
Q104 Mr Jayawardena: And may not have actually passed the hurdle for what is legally a hate crime.
Carl Miller: Indeed—and might be complete lies for all I know. We didn’t verify any of them.
Q105 Mr Jayawardena: One final quick question. I understand that almost one in three of the extremist Twitter accounts that have been sampled actually belongs to another extremist who has been suspended. Is that your understanding and, if so, is there a way that we can tackle that?
Carl Miller: I don’t think that is my research you are quoting, but it reflects that the logic of social media is such that it is much easier to upload content and to set up new accounts than it is ever going to be to close them down. Every time you close a Twitter account down, someone in Twitter, usually liaising with someone in a police service, has to make the decision to do so. In that time, you can send 10 tweets and set up 10 Twitter accounts. We have seen that especially prevalent in terrorist exploitation of the internet. So-called Islamic State has automated the setting up of Twitter accounts. They set up hundreds of thousands every single day, just as Twitter and others struggle to get them off. It is a general problem.
Q106 Mr Jayawardena: You have not thought of any immediate bright idea or solution to that.
Carl Miller: No.
Dr Burnap: What we have tried to do is build these automated methods for detecting unacceptable behaviour, essentially. The issue is false positives. If we are able to get that 100% accurate, then as soon as something occurs that falls into the category of unacceptable, we take it down straightaway. That can be done. The issue is that algorithmically, given the short, informal nature of social media, it is difficult to get that accurate every time, so you are going to get false positives. People are going to get fed up with the platform and they may stop using it, which is an issue.
To comment on one of your other questions, I think that, in terms of what happens online, it is very important to have the debate about the difference between what is antagonistic and offensive and what is a hate crime. You can include statistics of reported crimes. You can report statistics that have been followed through and prosecuted as hate crimes, but given that social media is so prevalent in people’s everyday lives—there is no real escaping from it; they use it for their personal interaction—if they are being targeted online and they feel it is intruding on their ability to interact with people online, that problem needs to be addressed and it cannot just be ushered away because it does not meet the high benchmark of a hate crime.
Q107 Naz Shah: I have two questions, for both of you. How effective have far-right groups been in exploiting social media?
Dr Burnap: Our empirical evidence suggest that they have been very effective, actually. They are one of the longest-lasting narratives on Twitter. They were in the events we have looked at. It is important to note that they are not always using outright racist or xenophobic language as such. It is very nuanced and an evolving “Us and them” narrative. They are using the social media platforms very well. I would just add to that: the flip side is that the police are also using it and we are up there with the far right groups in terms of the longevity of their narrative. There is a counter-narrative there, but in direct answer to your question, they are using it very well.
Carl Miller: I agree.
Q108 Naz Shah: That leads on to my next question. On counter-speech, your research shows that a common response to extremism is to engage in that. How effective is that in curtailing hate speech?
Dr Burnap: We have looked at it at an individual level. We have looked at threads of posts and the follow-up reactions to those posts. We found that if the target individual responds, they are going to get another response. If somebody else chips in and starts to defend them, they are going to get another response. The more this goes on, the longer it lasts.
There are particular types of responses as well, which lead to other types of responses. If somebody is identified—“You’re such a racist” for example— they will sometimes ease off. If you call them a keyboard warrior or a coward, the hate gets worse and more extreme: this is not the answer they were looking for. We have also found the more people get involved in a counter-narrative, the shorter those threads become. So it is a community effort and the more people are directed towards that, the better.
Q109 Chair: Just explain that?
Dr Burnap: In the statistical models we built there is a correlation, in that an increase in the number of individuals getting involved with counter-speech leads to a decrease in the length of the responses to a hateful remark. This could go on and on if there is only a small number of people involved, while it tends to slow down and stop if more people start getting involved. There are active social media accounts that support this as well. We call them sentinel accounts—things like “Yes, You’re Racist”—and they will actively seek out racist tweets and retweet them. They have a number of followers who can then see that and get involved in the response and the counter-narrative. The more people get involved in that counter-narrative, the more likely it is to shut down those hateful individuals. We have seen that across—we had a couple of thousand cases.
Q110 Naz Shah: In terms of far right groups, I had one really bad experience where Britain First came to my constituency in Bradford West and on one thread, one statement, I had over 2,000 comments. It was just shocking. I did not respond to it. I do not respond to anything on social media; it is just not what we do. I would not have a day job left if I did. Out of all the far right groups, which group would you say, if you could single some out, is most effective in using that? From my own experience, I have seen Britain First do it.
Dr Burnap: I do not have evidence on individual groups. There are so many of them and they come and go so quickly online and, to some extent, people merge between the two. The best way to identify individuals—you have got the groups in terms of their narrative, but then you also have individuals who associate with those groups. You might have some profile information that suggests they are related to Britain First or the EDL or whatever, but then they might have both.
It is very difficult to disambiguate supporters of those groups, because the supporters are as responsible for propagating the information and the success of their online narrative as the groups themselves. It is hard—you really need to pull apart the supporters and where their allegiances lie in terms of trying to allocate some success of those groups as individual groups, for propagating their message. It is very difficult to do that.
Q111 Naz Shah: A final point. In terms of post-Brexit and the rise of the far right across Europe in particular and especially with what has happened with Trump, is there or should there be any work ongoing on mapping out the links between these groups, some algorithms?
Dr Burnap: Yes, we are doing that right now, looking at relationships between groups and also relationships between online cyber-hate—and I mean that in more general terms, in terms of antagonistic, inflammatory and offensive commentary—and offline reports of hate crime. This is a study we are starting to undertake with the US, which is funded by the Department of Justice in the US and is with the RAND institution.
We will be looking at that exact relationship, so we will be able to look at the US, UK and into Europe as well. We will be looking at online, offline and networks in the types of actors. I keep apologising for the infancy of the work but it is quite new and will be relevant to you going forward.
Q112 Chair: When will that report?
Dr Burnap: The work is starting in January. It will be ongoing and we should have some work out in the first quarter to second quarter of next year. Following that will be the CST stuff that we have done, which will also look at the kind of questions that you are asking in relation to those kinds of events.
Carl Miller: May I make a quick point about counter-speech? In partner with Facebook, we have been trying to measure empirically the effect of counter-speech for quite a long time. A simple point I want to put in front of you is that it is really important to get clear who you think counter-speech should be targeted at.
There is a big difference between sending counter-speech directly at an extremist. Intuitively, frankly, as well as empirically, I am not sure there is a huge amount of evidence that that over a large scale will actually get that extremist to change their mind. I cannot believe that when anti-racist activists take on Nazis on Twitter that it is genuinely changing the Nazi’s mind about what they believe.
For us, the evidence shows that it is effective as a kind of mood music. A surge of support for the victims can encourage them to come forward. For people potentially of two minds or teetering on the edge, it can dissuade them. It can make the hateful activity seem more socially unacceptable, for sure. But that is a big difference to taking on already committed die-hard bigots. There I think actually there is a danger in doing that, which is reciprocal radicalisation. That is often when groups on two extreme ends of the political spectrum start firing abuse at each other, start cherry-picking each other’s statements and putting them into their own networks. That can actually get people very riled and incredibly angry and contribute to extremism, not the opposite.
Q113 Mr Winnick: The nub of the issue, certainly as far as Members of Parliament are concerned, is can the social media companies do more to police the sort of hate crime being used on the routine, regular scale as just mentioned? In fact, some are hardly doing anything.
Carl Miller: I think they are doing some things already. They are providing support, for example, on a hate crimes hub. Twitter released today that they are changing; basically, they are trying to empower their users better to mute content that contains certain words and phrases, that kind of thing. They are trying to improve how they handle the reports being sent to them.
The problems that social media companies face are twofold. No. 1: should they actually go out and proactively look for that content in the first place? My response is that I do not think that that is their role in this. I don’t think they should be analysing the vast amount of their own platform, if it hasn’t been reported to them. Simply because, unlike other kinds of socially problematic behaviour online, hateful stuff is ridiculously difficult to analyse at scale.
This is all a kind of grey area of linguistics and semantics. It would never be possible for Twitter, with thousands of people in a huge tower block somewhere, to be monitoring all of their content. The second area is whether they could and should be effective in responding to requests from police and from reports of their own users in flagging and ticketing poor content. There, I think they should be. They should do better and they could do better. There, too, the problem is simply the scale.
As a company, they are basically having to police, in Facebook’s case, an environment that is larger than any country. In Twitter’s case, 500 million tweets a day are flowing through their platforms. They are always going to get more ticketed content sent to them by users, some of which will be complete nonsense. It will be people trying to game the system, trying to get content down. It will be more than they are ever going to be able to handle manually. The real question is how they prioritise all of that. How do they actually send stuff to the top of the queue? How do they protect the most vulnerable people?
Q114 Mr Winnick: There is a contradiction. Racial hate in any other media outlet would be illegal and breaking the law, and yet on Twitter and Facebook outright racist, fascist hatemongers have a free run.
Dr Burnap: I have a slightly different view from my colleague here. In answer to your first question, Twitter, for example, has released a new policy. At the moment, it is mainly oriented towards muting content, so you do not have to see the content. My personal view is that social media companies generally could do more to actively seek out this content online.
Q115 Mr Winnick: They are refusing to do so.
Dr Burnap: I don’t think they are refusing to do so. There are difficulties for these guys. For a start, they are US-based, so they have much more freedom of speech than we do, and we have to respect that. That’s fine. On both sides of the Atlantic on the criminal aspect, the threshold to something becoming criminal is very high. Typically it has to suggest an active threat. Personally, I think that is too high and the bar should be set slightly lower, because this is a huge environment. You are talking tens of millions of people involved in these debates, and the ongoing narrative is dangerous. More could be done to seek out this content and block it even temporarily, or perhaps send messages back to suggest that this is perhaps bordering on offensive, or something along those lines. I understand this is difficult because the context is always difficult. The false positives may be high and social media companies do not want to block users who are legitimately using their platform from having a narrative. So I don’t believe they are ignoring it. I don’t believe they are deliberately allowing criminal behaviour.
Q116 Mr Winnick: But they could do more.
Dr Burnap: They could do more to actively seek out this content. For example, they have massive departments looking at online fake advertising.
Q117 Mr Winnick: Mr Miller, do you agree they could do more?
Carl Miller: I think they could do more to protect users. They could roll out new products, but I simply do not think that they can go out proactively, and nor do I want them to. We already worry about the power of social media platforms in general in public life. I worry that in electing them to be some kind of global policeman of their services, we are giving them another power that is only properly held by a state and held accountable to a democratic process.
Q118 Mr Winnick: It would be helping to stop the sort of hate that is so constant.
Carl Miller: Yes, but I think the right way of doing that would be for them to work with law enforcement agencies to take it down. If you are asking a company that operates pretty much in every single jurisdiction in the world to form their own view of what should and should not be allowed on their platform, you are elevating them as a private company to an area where I would not want them to be, which means that they are making their own private organisational calls.
Q119 Mr Winnick: In combating racism that presumably you and everyone else like us would agree should be combated.
Carl Miller: But if something is illegal, a law enforcement agency should represent it to them as illegal and it should be taken down. If it is not illegal, where is the grey line in terms of something being very offensive and something being a little bit offensive, and why should a social media company unaccountable to any democratic Government draw that line?
Q120 Mr Winnick: My last question. If it is illegal outside of social media, you are saying it should not be allowed on social media. Do you accept that?
Carl Miller: I know we are running thin on time, but I did want to make a point about the law, if I may, because this is a hugely important area for an inquiry to consider. I am speaking now not as a lawyer but as an interested observer.
It seems to me that the law is incredibly unclear about where the line on criminality really is between something that is actually illegal and something that is not illegal when it comes to online hate. If you look at what is down on the statute books in the Malicious Communications Act 1988 and the Communications Act 2003, in terms of the things that define offences on social media, one of the things that is currently illegal under those Acts is indecent, grossly offensive or menacing in character communications, as well as anything sent to somebody that is false or known to be false. There is incredibly draconian stuff down on the statute books. What the Director of Public Prosecutions has done is load loads of restrictive guidance on top of that, obviously to prevent the chilling effects of that kind of information. So it has now to be very, very offensive, and it is really unclear where the line in the sand is drawn.
I genuinely think that Parliament should spend some time discussing what should be illegal and what should not be illegal in the age of social media. We have not had a proper law passed on this since social media became in widespread use. If you talk to lawyers about this, most of them will say they don’t even know which Act really applies here. Some of it is the Communications Act, as I said; some of it is the Protection from Harassment Act. Some people say that it is public order legislation; others say that counter-terrorism or incitement of racial hatred legislation applies here. It all seems really fragmentary. In terms of policing this and getting a robust enforcement approach in line, as well as liaising with the social media companies, for Parliament to clearly define what is illegal would be the first step towards a robust enforcement solution to hatred online.
Dr Burnap: I will start with another example. Following the US election, Facebook has been absolutely slammed for the spread of misinformation. People are claiming that the spread of misinformation on that site has led to an outcome in a political vote. That seems to be important, and pressure is on Facebook to shut that down. However, the same pressure is not being applied for commentary that is antagonistic and offensive in nature. There are many, many millions of people who are being affected by this type of communication online. The framing of the problem is the important thing.
Personally, I think that bringing the law into it is very difficult. I understand that a line needs to be drawn for prosecution, but more could still be done to prevent antagonistic and offensive commentary before it becomes illegal. I don’t think it necessarily has to break the law before it can fall foul of terms and conditions of use of social media. The main reason is that in such a huge ecosystem for people to share beliefs and opinions, it is very easy for antagonistic and offensive commentary to spread to millions of people and for them to become involved in that mindset. If such a huge group of people were doing this offline, something would be done about it, but because it is online, it is largely seen as unpoliced.
Of course you have jurisdictional issues. A handful of companies are responsible for this, but I think more could be done across the board to bring these people together to discuss how we can frame the problem and where we draw the line between what is acceptable and what is not, without something necessarily needing to fall foul of the law before it becomes a problem.
Q121 Chair: Can you give an example of anything that any of the social media companies could do, in theory, but are not? Often this debate is quite abstract.
Dr Burnap: It is difficult. As Mr Miller said, Twitter has released a new policy today, which is good in the way that Twitter is trying to help its users to become less exposed to disinformation and suggest there are other ways to report it, but it does stop short of seeking out disinformation. For example, why can’t disinformation be sought out and muted before it has been seen? It has to be seen by people to be muted. This is a difficult problem, and I am in no way suggesting that the social media companies are not doing everything they can to bring this argument to the forefront. They are not trying to hide it or remove themselves from the discussion. They are definitely right there, doing it face-on. It is a very difficult problem, and it needs high-level framing of what is acceptable and what is not. That is the thing that has not been done yet.
Q122 Stuart C. McDonald: Mr Miller, you made arguments about whose role it should be first and foremost to police the content of Twitter, Facebook and so on. That must be a hugely expensive project, and some might say that if Twitter is not responsible for doing it, why should the taxpayer be responsible for investing significant sums of money in getting the police to do it?
Carl Miller: To my mind, the taxpayer pays for the police to enforce the law. I worry when we delegate to a private company that happens to be in control of what is basically a massive digital commons what is offensive and what is not offensive in the way suggested. Ideas of what should and should not enter the public space are always going to be incredibly contested. People are always going to have different ideas about what that should be. I think a hard legal line about what is and isn’t acceptable in the age of social media is the way to start doing that.
Q123 Stuart C. McDonald: In terms of the police, from your own observations, how well are they placed, in terms of being able to tackle this problem either through their understanding, their resourcing or even their technical capabilities?
Carl Miller: I think the police, through no fault of their own, are going through an enforcement crisis. I think that in a very short amount of time we have seen an enormous amount of either criminal or sub-criminal but problematic behaviour transferring on to social media platforms. This is profoundly different from trying to enforce the law in the offline world. They are dealing with enormous amounts of scale, as we have said.
To give you an example, 1,000 British people can have a hate crime offence conducted against them by one person at one moment. A thousand British people can have that same offence conducted against them by someone behind an encryption wall in a part of the world that has absolutely no relationship with British law enforcement of any kind. All of that is so challenging when it comes to how the police can reasonably quickly and efficiently investigate each of those complaints that I think it levels really profound new questions at them. This has all happened in policing terms in the blink of an eye.
Q124 Stuart C. McDonald: From what you say, it might now be impossible to police that without, for example, resort to perhaps use of algorithms and so on; and then you have other dangers, such as you end up taking things out of the Twittersphere and off Facebook which are actually probably not hateful.
Carl Miller: These are choppy waters, for sure. I do not think it is impossible to enforce the law on social media. I would like the law to extend as firmly on to social media as it does in the offline world. It certainly has impacts on there as it does in the offline world. I think that there is a real danger that the public will have diminishing confidence in the capacity of the police to protect them in the online world as they do in the offline one. I think that is already beginning to happen around things like online hate crime. In lieu of a robust enforcement response—and things like the online hate crime hub that we heard earlier are certainly a step in the right direction—you will see the rise of things like online vigilantism. That what happened with child abuse images online. I think increasingly that kind of stuff will happen in other areas where the police are not seen to be robustly enforcing and protecting British citizens.
Q125 Stuart C. McDonald: Dr Burnap, any further thoughts on where the police are?
Dr Burnap: I think the police are doing everything that they can. The online hate crime hub is great. I think they are actively engaging with it. I agree with Mr Miller in terms of the resourcing required to do this is huge. My personal view on it is that the most successful way to police this online will be community-based policing—communities actively engaging in the use of counter-speech to try and reduce the volume and propagation of cyber-hate and hate speech online. How that happens is just as difficult as actually policing it anyway, because you are talking about getting huge numbers of people together who can work together, collaborate to engage in the social media narrative. I just think it would be more productive coming from the communities themselves rather than from the police.
Q126 Stuart C. McDonald: Finally, you said earlier in your evidence that a lot of hate speech online is driven by particular events, and so on. Are the police more proactive after such events, or do they have the capability to be flexible and up their presence or up their action in response to such events?
Dr Burnap: They are certainly active. I mentioned earlier that their narrative lasts the second longest in terms of over time narrative, alongside the far right political groups and above the media and other types of actors and individuals, and so on. The interplay, I suppose, would be interesting to look at: to what degree does what the police are actually saying have an impact on the rest of the narrative on social media, if you like? Is it topics-based, or is it the sheer volume of police-level tweets is having an impact? That would be something that could be unpacked, I suppose.
One thing I have not mentioned, and this is a good opportunity to mention it, is actually the propagation of online cyber-hate. In our studies we have seen that actually cyber-hate is less likely to be propagated than other types of tweet around an event. Actually tweets that are supportive, which would include tweets from, I suppose, active parliamentary politicians and also the police, are more likely to spread than the negative and hateful narratives. Their actions in spreading a positive narrative are successful following events.
Mr Jayawardena: Mr Miller, I just wish to clarify the statistics I gave you, because I have the source now. For the record, 17,000 tweets provided or relayed specific hate crime incidents. That was 17% of the overall number, roughly. Only 14% of those were original messages; the others were amplifying or retweeting them. You reported that only 6,720 of those came from the UK, so if the same ratio applies, then it is 941. I wanted to put that on the record.
Carl Miller: Thanks very much.
Chair: Thank you. I am conscious that as a result of the votes, this has run rather later than we intended. We appreciate your time. If you have any further thoughts in terms of developing research you are doing about the nature of online hate crime and broader online abuse, we would be very interested in that, as we would be in any further thoughts on specific things that social media organisations or the police could do. I know that this debate can often be quite general in calling on people to do more without having precise proposals of further things that could be done, so anything like that would be welcome. Thank you both for your time; we really appreciate it.