HoC 85mm(Green).tif

Science and Technology Committee

Oral evidence: The Evidence Base of Parole Board Decisions, HC 712

Wednesday 7 September 2022

Ordered by the House of Commons to be published on 7 September 2022.

Watch the meeting

Members present: Carol Monaghan (Chair); Dawn Butler; Chris Clarkson; Tracey Crouch; Katherine Fletcher; Rebecca Long Bailey; Graham Stringer.

Questions 1-83


I: Dr Jo Farrar, Second Permanent Secretary and Chief Executive Officer, Ministry of Justice and Her Majesty’s Prison and Probation Service; Dr Jo Bailey, Head of Psychology Services Group and Deputy Director, HMPPS; Martin Jones, CEO, Parole Board; and Sonia Flynn, Executive Director Chief Probation Officer and Women.

II: Professor John Podmore, Professor of Sociology, Durham University, and Professor Stephen Shute, Professor of Criminal Law and Criminal Justice, University of Sussex.

III: Professor Seena Fazel, Wellcome Trust Senior Research Fellow, Department of Psychiatry, University of Oxford, and Dr Ruth Tully, Consultant Forensic Psychologist and Clinical Lead, Tully Forensic Psychology.

Written evidence from witnesses:

– [Add names of witnesses and hyperlink to submissions]

Examination of witnesses

Witnesses: Dr Jo Bailey, Martin Jones, Dr Jo Farrar and Sonia Flynn.

Q1                Chair: This evidence session of the Science and Technology Committee focuses on the scientific basis of policy and how the Parole Board uses evidence to assess the risk that a prisoner will reoffend, and the frameworks and processes employed to ensure that this is done consistently and effectively. Just checking if anyone would like to declare any interests? No. I welcome our first panel of the morning. Thank you for joining us today. We have Martin Jones, the CEO of the Parole Board; Dr Jo Farrar, second permanent secretary for the Ministry of Justice; Dr Jo Bailey, head of psychology services group and deputy director of Her Majestys Prison and Probation Service; and Sonia Flynn, the chief probation officer of Her Majesty’s Prison and Probation Service. Thank you all for joining us this morning.

I will start by asking Martin Jones: to what extent is there a threshold that prisoners must pass before being released on licence or moved to an open prison?

Martin Jones: The threshold you have to pass to be released by the Parole Board is set by law—it is a statutory threshold. A person can be released only where it is no longer necessary for the protection of the public that that person be detained in custody, so that is quite a high threshold that the prisoner has to pass. There is a different threshold for us to recommend somebody’s suitability for a move to open conditions. That is only a recommendation; the final decision rests with the Secretary of State in those cases. What you are looking at there are the potential benefits from somebody being tested in open conditions in relation to rehabilitation and preparing them for release back into the community versus the risk to the public. You cannot recommend that somebody is moved to open conditions if you think that risk is not manageable in conditions of lesser security.

Q2                Chair: Is it usually the case that somebody will almost definitely pass these conditions unless particular concerns have been raised?

Martin Jones: It is important to put it in the context of the numbers that we deal with. The Parole Board’s job in law is to ensure that we do not release people who are a significant risk to the public. In the past seven years that I have been in charge at the Parole Board, we have reviewed just under 100,000 prisoners. Our release rate across those cases has been about 23%, so 77% of the people reviewed in the time that I have been chief executive have been directed to remain in custody for the protection of the public. That involves reviewing all the evidence we receive in relation to the offences they have committed, the assessment of what they have done in custody and their plans for the future. I think that suggests the bar is set quite high, but of course what we are looking for is the progress that somebody has made in the time they have been in custody. How have they changed from the person who committed that offence in the first place? If we are satisfied in law, of course we have to direct the release of the prisoner.

Q3                Chair: Would you also take into account the circumstances, such as the community and the support group that might be there for them when they are released?

Martin Jones: That is absolutely vital for us. If you look at the empirical evidence, we know the factors that are likely to lead to an offender reoffending or not reoffending. There are some pretty clear things in relation to, for example, accommodation. A lot of work has been done with the prison and probation service on accommodation when people leave prison, particularly the high-end risk people we are dealing with. Employment is also a significant factor. If you are releasing somebody and they will be jobless, that is obviously a huge issue for them in relation to the likelihood of them reoffending. The other question relates to support structures in the community. Do you have a family to go back to? Do you have relationships to go back to? That is incredibly important as part of that.

For every parole case, we get a risk management plan developed by the probation service, which explains how they would manage that risk in the community—of course, nobody we see is risk-free. It is about reducing risk to the point where it can be managed in the community, and setting licence conditions to ensure that risk can be mitigated through measures such as an exclusion zone, drug testing and alcohol testing—measures to ensure that if somebody starts to sway away from the right path, that can be identified and tackled, and where necessary that person can be recalled back into custody. The risk management plan is crucial for the decisions that we make.

Q4                Chair: Looking across a number of parole boards, how do you ensure consistency?

Martin Jones: It is really important to make the point that we get evidence from prison and probation officers on whether a person is safe to be released or not, and work by the Ministry of Justice some years ago suggested that 90% of our decisions are in line with the evidence provided by report writers. That provides some evidence of consistency.

Every board member is appointed to ensure that the board has the abilities and the skills to make the right decisions in these cases. When people join the Parole Board, they get an extensive set of training, and we have a published, structured, decision-making framework that every parole board has to have regard to when looking at a case. Ultimately, however, these are judicial decisions; the board is sitting as a court—independent of Government—to debate a decision. That is what the courts and the law require: that an independent body makes these decisions.

We cannot go behind that, but there are some checks we can do in the background. We have quality assurance processes to ensure that decisions are in line with the framework. A welcome change in 2019 was the introduction of a reconsideration mechanism where if either of the parties—the prisoner or the Secretary of State—believes the decision was irrational or made in a way that was unfair, they can challenge the board’s decision, which is then reviewed, ordinarily by a judge on the Parole Board, to ensure that it was made properly, in essence applying a judicial review type of test to ensure that the decision was fair. Again, that provides a way of checking that our decisions are sound.

Q5                Chair: You mentioned training. Does refresher training take place, or does training take place only on appointment to the Parole Board?

Martin Jones: Everybody who joins the Parole Board has mandatory training that they have to do—ordinarily about a week long—to explain the intricacies of risk assessment and risk management, and the way the work is processed. There is also mandatory training that is provided every year for Parole Board members in relation to different parts of practice. That is something we keep under review, in relation to what the different changes are.

About two years ago now, following the London Bridge attack, the Parole Board introduced a specialist cohort of members specifically to deal with terrorist offenders. That specialist cohort get mandatory training that is very intense, and they get annual training on the latest developments. We bring in experts from the probation service, psychiatry and sometimes counter-terrorism police, for example, to ensure that we have the very best evidence on which to make our decisions—but of course the decisions are individual ones, based on the circumstances in each case.

Q6                Chair: I understand that a decision-making framework was introduced in 2019. Have you seen any impact as a result?

Martin Jones: Yes, we have. Again, it is important to be clear that the principles set in that framework are based on looking internationally at work in other parole jurisdictions, such as Canada, New Zealand and Australia, to ensure that we follow the best in the world.

A lot of the factors are, unsurprisingly, things that we would expect any reasonable parole panel to look at, so of course part of the framework is looking at the original offence that led to the person coming into custody. We ensure that that is made very clear to Parole Board members and panels—here are the factors that you must have regard to, who was the individual who committed the crime, what were their previous convictions and so on. We make that very clear in the framework. We have certainly seen benefits in the clarity and understanding of our decisions.

That is also important for victims of crime, who of course want to know the reasons for our decisions. It is much easier to understand those decisions if we can provide them with a summary of our reasons, alongside the framework, which explains the factors that we take into account.

Chair: Thank you. I will now turn to some of my colleagues, starting with Chris Clarkson.

Q7                Chris Clarkson: Martin, you mentioned behaviour in open conditions. To what extent and how is the behaviour of offenders in open prisons and their engagement with rehabilitation programmes factored into the decision-making process when considering whether they are suitable for early release?

Martin Jones: It is absolutely crucial for us. It is important to be clear that most people who go to prison are released automatically, without seeing the Parole Board; we only look at those who have committed the most serious sexual or violent offences, or who have been recalled to custody because they are a danger to the public. If a prisoner goes to prison for committing a serious offence, what we are looking for is how they have changed from the person who committed that index offence to someone who is now safe enough to be released back into the community. Engagement in courses is therefore important. There is some good science around desistance theory. What are the signs that someone has changed, or is willing to change, and has perhaps matured during their time in custody? Willingness to engage is really important. We take the evidence from, for example, an intervention that someone has done.

It is important to be clear, of course, that it is not a matter of going on a course and it fixes you. There is nothing that fixes somebody who perhaps has a really deep-set problem, but it will give you an idea of understanding what their issues and problems may be—what the causes of their offending may be. It may give some helpful techniques in relation to understanding how to avoid risky situations. However, that is crucial for then understanding Parole Board decisions.

Open conditions, I would say, is a critical part of our decision making, and sometimes I think it is misunderstood. If you have spent 20 years in closed conditions, serving a mandatory life sentence for murder, that is a pretty controlled environment in terms of what you can do in that space. Rather than release that person directly from closed conditions, it seems wise to test them in lower security conditions to understand how they would reintegrate back into society. Research in 2019 showed that, provided it is done carefully, with the right risk assessments, release on temporary licence from open conditions is much more likely to lead to that person being released from prison and not committing serious offences thereafter. I think that is a really important safeguard.

In essence, we say that open conditions is testing, so there would have been an apparent reduction in your risk over a period of time, but before we are prepared to release you, it is important that we test whether that apparent change is proven as we start to then give you the opportunity to go in to the local town centre and open a bank account, go to the library, or start seeking employment as. Of course, we take account of that evidence. If somebody spends two or three years in open conditions and, actually, their behaviour is demonstrating that their show of change in prison is then coming through in the community, there is a very high chance that they are going to get a good outcome from their next parole hearing. The release rate for people who have had good success in their time in open conditions is probably 80% to 90%. That is unsurprising because the answer is almost that it is confirmation that they are safe to be released.

It is about every individual case and, of course, going to open conditions also gives the opportunity to test whether that is not quite right. If you go to open conditions and you start taking alcohol and you miss coming back—you abscond—then of course that is evidence that perhaps your compliance is not what we would expect, and generally speaking that will result in a significantly longer time spent in custody. Any breach in open conditions is likely to result in that person being sent back to a closed prison, where you can look again at their behaviour.

Q8                Chris Clarkson: You said every case is different. What baseline standard are you using to metric whether somebody has changed, improved or matured? Presumably there has to be some sort of standardised indicator.

Martin Jones: For every parole dossier that we get, we get evidence—normally, it is a paper dossier; it is not electronic dossier evidence about that prisoner—for most of the complex individuals, which will run into many hundreds of pages. In the most complex cases, there will be 1,000 pages of information about that prisoner. The key part of that will be the risk assessments undertaken by the professionals in that case, whether those are psychologists, probation officers or others.

The key point is that the parole hearing that takes place—it is a court hearing, where you are hearing evidence—will maybe last three hours in which you take evidence from the prisoner and the professionals to make that judgment. The probation officer, the prison officers and the psychologists will have spent much more time than that with the prisoner, looking at their behaviour, understanding and talking to them, so that assessment of whether their risk has changed is vital to us. It is a key indicator of whether somebody is likely to progress from prison to release at that point.

A huge number of algorithms and risk assessment tools are provided as part of the dossier, depending on the nature of the offence. Of course, if that person is a sex offender, there is a particular set of assessment tools, while terrorist offenders are subject to different risk-assessment tools. Those risk assessments are really key indicators. The job of the Parole Board is then almost to add the judgment on top of that, saying, “Okay then, how do we test that?” It is not a matter of simply saying, “Okay then. Well, apparently your risk has reduced.” We are looking for evidence of that in their behaviour and we will be testing that significantly at the hearing itself.

Q9                Chris Clarkson: Thank you. Sonia, to what extent are victim statements taken into account? Presumably, in some of these cases, the victim of the crime will still be very much involved and will have a view on whether or not this individual should be released into society. To what extent is that included in the metric?

Sonia Flynn: It is in statute that we must consult victims of serious crime on their view of release, and for them to also give our victim liaison officers a view regarding the protections that we need to put in place to reduce their concerns about that individual if the Parole Board does choose to release—particularly the obvious concern that they could bump into them in the street.

On the victim contact scheme, the probation officer will consult the victim liaison officer, who in turn will turn will consult the victim or the deceased’s family—because of the type of crimes we are dealing with—and that information is then relayed back through from the probation officer to the Parole Board.

I think that one of the things we have to manage when we are working with victims is the expectation that they can somehow prevent release. Ultimately, the I see our job very much as giving an assurance to the victim that, if that decision is made, we and other agencies will ensure that they are fully aware of why the decision was made and of the parameters we can put in place to manage their concerns. But as Martin has indicated, the Parole Board has a good track record in terms of the judgments it makes about release, and very few people go on to commit further serious crime. No serious crime is to be tolerated, but we are also giving the victims an assurance that this is a system that makes good judgments.

Q10            Chris Clarkson: Thank you. Dr Bailey, can I turn to you, before I let somebody else have a go? To what extent do psychologists monitor risk? Obviously, these individuals will have to be very carefully scrutinised by a professional such as yourself. To what extent is risk factored into any decision to release, and what sort of standardisation is there? Obviously, different psychologists will have different approaches to psychotherapy, treatment and so on. Could you talk us through that?

Dr Bailey: Psychology reports produced for the Parole Board are very much based on a risk assessment of that individual. As Martin has already suggested, there is a whole plethora of risk assessments. In HMPPS, we tend to use standardised tools that have published validity and reliability. If you wanted to have a look, the RMA in Scotland has a lovely directory that talks you through all of the reliability and validity, if that is of any interest. In HMPPS, we tend to use fairly standardised tools. If new tools come on the market, we will get people trained up, and will have a look at them. We might even use them as sort of a shadow tool, but we would not use something that we did not think was fairly standardised.

Having said that, a lot of tools have been validated on adult white male samples. When we are working with women, young people or people with learning difficulties—whatever the change is—we have to ensure that we are using a proper tool. If there is not a better tool, we have to be very clear about the limitations of what we are doing—that it is more of a framework than a risk assessment. I think I am probably going off your question, though.

Q11            Chris Clarkson: No, it is very interesting. It is something that I was trying to get to, and I think you have done the work for me, because I could not think of a direct route in. I am concerned about the standardisation piece. As you say, some of the diagnostic tools were designed for one particular cohort. What latitude do you have to disapply bits or modify bits of them?

Dr Bailey: It depends on whether we are talking about actuarial tools, which are pretty standard, because they are based on historic, static factors—they are just tools that you apply, and there is no discretion in there—or structured professional judgments, which tend to be more of a psychological risk assessment: there is an element of clinical decision making, because you look at the evidence and decide how to draw out the evidence to score within the structured professional judgment.

On standardisation processes, within HMPPS we have a variety of QA processes. All of our trainee psychologists are fully supervised by registered staff, who are responsible for the output that they produce. For all of the reports of our registered staff, we have a QA process, particularly, for example, on terrorist offenders. That is quite a developing area, and we have quite a firm QA process on that, and then are slightly lighter on other things. We also use a tender framework, as we call it, to procure reports where demand outweighs supply, or when the Parole Board wants a particular specialist assessment that we might not have the skill base to do. We also have a QA system around that.

Q12            Graham Stringer: Some of the statistics and processes you have told us about, and I have read about, are impressive. In your annual report, the reoffending rates—about one in 200—are over a three-year period. Why?

Martin Jones: It is about trying to get a link between the decision that we make and the fullness of time. Generally speaking, the evidence is that, if an offence will happen, it will happen quite quickly thereafter. There is really strong evidence for that from the MOJ’s published serious further offence figures. Each year, we probably get one or two cases notified to us that are outside that period, so it is not a huge number if you measure it over a longer period in that way, but it is a pretty good representation of the fact that we are making a decision. Of course, with some of the people we see, particularly if they are serving an extended determinate sentence where they perhaps have six years to serve in custody and then three years on licence, it is not an indefinite period of risk that you are looking at. Thereafter, it is only a limited period within that.

Q13            Graham Stringer: My notes say that when you look at sexual offences against children outside the family, the reoffending rate is about 25%. Is that right? Over a six-year period.

Martin Jones: Over a six-year period? If you look at published Ministry of Justice figures on reoffending, sexual offenders have the lowest proven reoffending rate of any category of offender, but it is important to take those in relation to the context of a whole group of people. When the Parole Board is making its decision on whether to release somebody, we are looking at the risk that that person presents in relation to the question, “Is it manageable in the community?” If we thought somebody was likely to present a significant risk to children over the long term, we would not be releasing them; we would be looking to ensure that that was fed into the decision making. I have never seen a figure suggesting 25%. That seems extremely high.

Q14            Graham Stringer: That is specifically for offences against children outside the family. You don’t recognise that figure?

Martin Jones: I do not recognise that figure at all. I have seen much lower figures for sexual offenders.

Q15            Graham Stringer: Thank you. Percentages cover a lot of sins. Do you have actual figures for the number of people murdered, seriously assaulted or sexually assaulted in relation to people who have been released? Do you have an absolute number?

Martin Jones: An absolute number? In relation to the last year for which figures are available, published last year, there were no murders, as I understand it, committed by people released by the Parole Board. If you looked over a longer period, you would probably be looking at maybe 12 cases, or something like that. I can confirm the figures to the Committee if that would be helpful. Generally speaking, last year, for example, the Parole Board released somewhere in the region of 4,000 people, and the number of serious further offences hangs between 20 and 30 cases each year. Of course, we examine each case in detail to understand what has gone wrong—was this an expected offence?

What you quite often see, and one of the difficulties around parole, is that most serious offences committed by people released from custody don’t come from the people you expect; it is extremely unusual for somebody who has committed a murder offence to be released and go on to commit a murder. It is much more common for somebody who has perhaps been recalled to custody off a burglary offence—they are assessed, and you are thinking about the risk of acquisitive crime, for example—to go on, out of the blue, to commit a serious offence thereafter. That is part of the difficulty of risk assessment. What you know about is the past—what is the person’s background and what are their previous convictions? What you don’t know about is the future. If we had a crystal ball, that would be incredibly helpful in making our decision. We have to rely on the evidence that we know in relation to what works and what affects people to change.

Q16            Graham Stringer: Again, the process you have put forward is impressive, but there was a public outcry about the release of Worboys and there was a judicial review, and when you look at the papers that were presented to the Parole Board it is pretty appalling—the evidence that wasn’t put before the Parole Board. The chair resigned. How typical was, and is, that poor quality of evidence presented to the Parole Board?

Martin Jones: It is a major challenge. Each year, the Secretary of State presents the evidence that he wants the Parole Board to consider. There is a set of mandatory documents, under the rules that were set by Parliament in relation to what has to be disclosed to a panel. If that information does not come to us, we reject the dossier, but the quality of the information is paramount. One of the things on which we have been working hard with the prison and probation service is ensuring that we get that information, because clearly it is difficult for us to take account of information that is not presented as evidence. You can’t ask Parole Board members to read the newspapers to get their evidence. It has to be presented to us in the way of evidence that is presented as part of the file, in the same way that any court would.

Q17            Graham Stringer: But the point made by the court was that serious evidence was not considered that was available from the court decision that convicted Worboys. That is extraordinary.

Martin Jones: Obviously, it would not be appropriate for me to talk about individual cases, so if I can talk about the generality and the points that were made, I think the really difficult—

Q18            Graham Stringer: But that is public information, isn’t it?

Martin Jones: It is public information, but I—

Q19            Graham Stringer: It changed the law.

Martin Jones: The difficult point about that case was the extent to which the Parole Board should take account of alleged but unproven offending. What you are convicted of by the court is perhaps six, 10, 12 offences, but the suspected pattern of offending is much wider than that. What the court found was that the Parole Board was obliged to then consider wider but alleged offending, but doing that in a way that is fair is obviously a challenge on the basis that we cannot simply say, “The newspapers think this.” You have to adduce that evidence as part of that, and the Parole Board has now developed guidance in relation to what is a fair way of doing that.

It would be wrong, for example, if we said that this person was taken to court, charged with four offences and acquitted of two, but, “We are going to keep you in prison because of the two offences for which the jury found you not guilty.” That would be quite difficult law if we said, “We are going to hold to these things even though the jury has found you innocent of those charges.” How you do that fairly is a challenge for us, but we have developed some very detailed guidance on how to approach this. We have worked with the prison and probation service on how they then present the evidence in relation to what impact that might have on risk assessment.

Q20            Graham Stringer: In answer to questions from Carol earlier, you said that you took into account support from family and all sorts of other things that you believe are factors in not reoffending. Mr Forde did a study that said there was no relationship whatever between those factors and reoffending.

Martin Jones: I don’t think that is backed up by the international evidence on this. If you look at the probation service and the inspectorate of probation work and research behind the supervision of offenders, support, education, employment and accommodation is absolutely critical to understanding why some people reoffend and others don’t. That is my understanding of the evidence.

Q21            Graham Stringer: Can you cite evidence that shows that that is the case?

Martin Jones: I am very happy to share with the Committee in writing the evidence that I have seen and the references and studies that sit behind that.

Dr Farrar: I could add that we currently have a reducing reoffending programme. It’s wider than Parole Board cases, but there has been research to show that employment, education and accommodation on release are the three things that make a real difference. We are evaluating that at the moment. The Committee will probably want to have a look at that evaluation. It’s early days, but the early evidence seems to suggest that particularly accommodation is significant. As we complete that evaluation, we will be happy to share it with the Committee.

Q22            Graham Stringer: The difference between criminal courts and civil courts is “no reasonable doubt” and “balance of probabilities”. When you assess these cases and look at a risk analysis, is it a balance of probability judgment or no reasonable doubt?

Martin Jones: Certainly it is not clearcut. It is not set out in law. It is more the balance of probabilities that we look at in parole proceedings. Of course, the more challenging bit of that is when you look at, for example, unproven offending on the basis that the bar to that has to be high. You need to be pretty sure that there is something behind the event, but, generally speaking, it is the balance of probability that we tend towards.

For example, one of the things that a Parole Board can do is make a finding of fact. They see evidence, it is contested, and we say we think more likely than not this happened. This may be particularly true around domestic violence allegations. Somebody is recalled to custody. There has been an allegation of domestic violence that is not taken to court, perhaps because the victim does not want to take the case in, but the parole panel can adduce the evidence and say, “Despite the fact that this is not proven, we think it is probably more likely than not, if we look at the evidence behind that and the statements made to the police at the time, that there was an incident of some kind—some sort of altercation and disturbance that could potentially go to that person’s risk—and we apply the balance of probabilities when we make that sort of finding.

Q23            Graham Stringer: In terms of the basis of your function, do you believe the biggest driver of what you do is the sheer size of the prison population, or is it humanitarian concern or concern about costs? What is the main driver behind the Parole Board’s system?

Martin Jones: I can be very clear on this. Our statutory purpose is to ensure that we do not release back into the community people that are dangerous. If we decided that 100% of people were too dangerous, that would be what the Parole Board had decided, but of course we look at each individual case and decide whether the test for release is met and if somebody has changed his partner. But I don’t see it as part of my role, sadly, to give the Ministry of Justice back more prison places. Our job is to ensure we make the right assessment in those individual cases. But of course, there is a different way of describing this. Last year, we decided that 12,000 people needed to remain in prison for the protection of the public. It has to be said that that has a huge consequence for the public purse, given that it takes about £50,000 a year to keep somebody in prison. But, as a consequence of our decisions—

Q24            Graham Stringer: I suppose, though, the question I was asking was not about the statutory basis, which I understand—you have articulated it perfectly—but why the statute is like that.

Martin Jones: I think the statute is there on the basis that of course, if you go back into the mists of time to when the Parole Board was first created in 1967, it was a political decision as to whether people should be released or not. Of course, the law changed over time to make this a decision of a court, to ensure that you have a fair decision made as to whether the person is to be released. I think what it acknowledges is that of course somebody can do an absolutely dreadful offence but actually, after the passage of time, they may be safe to be released—after 20, 30 or 40 years. It’s a matter for the courts and Parliament to decide how long people should spend in prison for punishment. It is not at all uncommon now for people to be sentenced at the Old Bailey and receive 30 years in custody which they must serve in full before they get a chance of a Parole Board hearing. But of course we are looking at somebody who perhaps has changed during the course of that journey. That is what this acknowledges. But of course there’s no assumption. The figures show that actually our release rate sits at about 25% and has sat pretty consistently at that rate. It suggests that we approach that with some caution, given the fact that obviously we do not want to put dangerous people back on to the streets. I would say public protection is our absolute, No. 1 priority and I would strongly rebut any suggestion that it’s not.

Q25            Chair: Martin, could I go back to something you have just said to Graham? You talked about the evidence that you would use in order to make a decision and you said that if there were not convictions, that couldn’t be used as part of the evidence. Certainly in the Worboys case, we saw that an awful lot of potential cases had built up. It wasn’t one or two; it wasn’t hearsay. There was a body of evidence. Is there a threshold or criteria that that evidence has to reach, in terms of numbers, severity or convictions, before it is taken into account?

Martin Jones: The guidance, which I am very happy to share with the Committee if it would be helpful, makes it very clear that it is not limited to one thing. So actually, you might find that a single allegation, if you can evidence that this is more likely than not to have happened, can influence a Parole Board decision. But you have to do it in a way that is fair, so you have to adduce the evidence. It’s pretty common, for example, for the Parole Board to direct a report from the Crown Prosecution Service or the police to ensure that we have enough information to make a judgment as to whether it is likely to happen or not. Of course, if you have a single allegation, that carries much less weight than if you get, for example, a dozen allegations. If, for example, you have a man and maybe six or seven allegations that they have assaulted different partners but that has never been proven in a court of law, that is likely to carry much more significant weight than a single allegation, perhaps. You can start to see how you can work across that. But of course, the Parole Board can’t simply say, “Well, there have been a number of allegations—no smoke without fire.” We have to then draw in the evidence and normally direct that evidence to ensure that we can take that into account, and then ask the professionals, “Okay, what do you think about that?”

Sometimes you find that you are dealing with somebody and their index offence may be an armed robbery, but actually the risk that you are now dealing with—following recall, for example—is not that they are going to rob a bank, but a risk of serious violence to a future partner. That’s the elegance of the parole process: you can take into account new, developing information about what you know about that prisoner. But the crucial point, because we’re a court, is doing it in a way that is fair, to ensure that the prisoner has a chance to see that information and, of course, comment and ask questions about the evidence that has been presented.

Q26            Rebecca Long Bailey: This is a question to Martin and Dr Farrar. Why does the mix of panel board members need changing, and what extra skills and experience will law enforcement members bring? We’ll start with Martin.

Martin Jones: I think it’s really important to be clear. We have 346 members on the Parole Board. They come from a huge, diverse range of backgrounds. We have about 50 to 60 judges sitting on the Parole Board. That’s either retired or serving judges bringing their wisdom to our decision making. We have about 100 specialist members, being either psychologists or psychiatrists, looking at, for example, people with mental health problems. And probably about 60% of our members will then come from what we describe as an independent background, so they are not a qualified professional in the way a judge or a psychologist or psychiatrist is, but they come with a professional background. They may come from a probation background—quite a few former probation officers serve as part of the Parole Board. It is actually about the assessment of risk, and obviously that expertise is incredibly important.

Bringing together the right expertise on panels is absolutely crucial for us. We have about 25 former police officers on the Parole Board, who are incredibly valuable. We have been using them extensively on terrorism cases, where there is a real value to understanding those things.

One thing I am really clear about when looking at the data is that the composition of a panel does not affect the decision making. Most decisions are made by a panel of three who come together to make those decisions. It would not be true to say that if you stuck loads of policemen on panels, you would never release anybody. There is a benefit from the approach to questioning that people can bring. In serious and organised crime there might be some benefit in those cases, while you might use judges on cases that give rise to complex questions of law.

I have no problem appointing more policemen to the Parole Board. The question I would ask is: why are we doing it? If it is to seek to influence Parole Board decision making, that is the wrong approach. It is about ensuring that we have the expertise we need to make the right decisions on the very difficult and complex cases that we are asked to decide on.

Dr Farrar: To add to what Martin said, first, the Parole Board has a huge and diverse range of experience, and we really support that at the Ministry of Justice and recognise the value of it. We are definitely not setting out to ensure that there is undue influence in the Parole Board. We really value the independence of that board and the range of diverse experience and the amount of professional expertise.

The idea behind ensuring that there is a law enforcement background comes from learning that we have had from serious further offences. Every time we have a serious further offence, we are very thorough in looking at what happened and how we learn the lessons. Although we cannot talk about individual cases, we make sure that there is thorough scrutiny.

One of the things that we feel that the Parole Board benefits from and already has is expertise from people who have first-hand experience of that offender behaviour and tackling crime. As someone who has come into the prison office quite recently as chief executive, and now as second permanent secretary, seeing the experience of frontline law enforcement officers, prison officers and probation officers, you see the value of people who have worked with individuals. The idea behind that is to make sure that we have that frontline knowledge of individuals when decisions are made, to make sure that we have really high-quality, good decisions.

As was said earlier, the aim of us all is to protect the public. Our absolute aim is to make sure that people who are released are ready to be released.

Q27            Rebecca Long Bailey: This is a question for everyone on the panel, and relates to the evidence available to the Parole Board when making their decisions. I understand that probation unions wrote to the Justice Secretary in July, raising concerns about changes. I will read out quite a worrying paragraph, which says, “The decision you have taken to prohibit Probation staff from offering recommendations on prisoners progression and release severely endangers the ability of the probation service to protect victims of the most serious offences, and indeed the wider public, from the risk of serious harm posed by many individuals involved in the parole system.” How would you respond to those concerns? I will start with Dr Farrar.

Dr Farrar: I am sure that others will want to add to this, because we have thought carefully about it. The first thing to say is that we are absolutely not stopping probation officers and psychologists giving really valuable evidence to the committee. We are operating now in a similar way as we would in court. As Martin said, this is a judicial process. We put forward all the evidence but we no longer put forward individual recommendations. Sometimes, we will bring recommendations together, and the Secretary of State will give a view, but in most cases, we will not put forward a recommendation.

We will put forward the high-quality evidence that we have always done and are, as Martin said, improving constantly to make sure that a balanced view is given to the Parole Board. As I said, the prisoner and the Secretary of State are party to parole proceedings, similar to in the judicial process, whereas probation officers are not. We want to make sure that we act in a very similar way to how we would act in other judicial processes. I am sure that Sonia will want to pick up some more detail.

Sonia Flynn: It is absolutely correct that we have made changes, both to the guidance that we have given to probation officers and to the report template that they have to complete; we have removed the section where there was a recommendation. As Jo has outlined, that has not in any way taken away from the core body of work that they are there to do, which is to present the risk and, if the Parole Board is minded, to set out the requirements of a licence, so that the individual can be safely managed in the community. That has resolved an issue that we have seen over time, whereby Jo’s staff could provide a recommendation to the Parole Board that is different from that of the probation officer. That seems quite confusing, given that we are all one HMPPS. This kind of tidies that up. It is for the Parole Board, based on the evidence they have been given by the psychologist and the probation officer, to reach their judgment and make their recommendation. I believe that this has clarified the distinction between the roles of the Parole Board, my staff and Jo’s staff, in terms of giving evidence.

Q28            Rebecca Long Bailey: Martin?

Martin Jones: Obviously, how the Secretary of State and the Ministry change the rules is a matter for them; the officials are their officials—they can instruct them on how they operate.

I always say that Parole Board decisions must be based on evidence, not recommendations. It is not a matter of saying, “We released this person because somebody recommended it”; the evidence behind that judgment has to support it. Of course, part of the job of the Parole Board and the training we give is to ensure that we understand the evidence that is presented.

In a way, I think there is a danger of a recommendation being a bit of a distraction. In reality, if the evidence of the prison and probation service, and the psychologists, is that a person is safe to be released, the question of whether they recommend that might be a bit of a distraction—other than on one point: the research that I mentioned in my opening remarks shows that 90% of our decisions are in line with the recommendations and evidence of report writers. If there is more ambiguity, particularly in cases where we are sitting on the fence, that may present a challenge for the clarity of our decision making. That is something that we will work on closely.

The job of the Parole Board members, as the judicial decision makers, is to scrutinise the people who have put forward that evidence, and they are quite skilled at understanding what the right decision is at the end of the process. It will probably take some questions to tease out those people’s view on whether the risk is manageable in the community.

it is really important that the parts of the system work together. We will ensure that we do not have a lack of clarity about whether a person remains a significant risk to the public. I want to ensure that we get the right evidence to make the right decisions.

Q29            Rebecca Long Bailey: Just so that I am clear, you mentioned that a probation officer can state in their evidence that there is a risk. Surely that would be a recommendation.

Martin Jones: Yes. When you are looking at parole decisions, it is not so much about there being no risk; it is about the risk reducing over a period. You are quite right, in a way: if somebody says to a parole board, “There’s a really high risk that this person will commit a serious offence after their release,” it would be pretty obvious to the panel that had a recommendation been made, it would have been, “Do not release this individual.” Then you have people who have been in prison for 50 years, and the probation officer, the prison officer and the psychologist may say, “This person is now very low risk”—it is never zero risk—and then you can see which way the recommendation would be.

The difficulty is the blend in between. The art is in trying to get underneath to test the risk. The really crucial part in how that comes through in hearings is in looking at the risk management plan developed by the probation officer, and asking, “Is this plan going to manage the risk in the community?”. To take a pretty obvious example, we might know that a person’s risk is around drinking alcohol. Over the last year or so, we have started to release people on the condition that they wear a tag to monitor their use of alcohol. That is a pretty useful tool for us in relation to people who may fall off the wagon. In the past we may have taken some assurances by asking, “Okay, what do we know about them?”, but of course things can go wrong pretty quickly in the community. Some of the new technologies can be helpful ways of understanding the risk.

Q30            Rebecca Long Bailey: Dr Bailey?

Dr Bailey: I do not have terribly much to add. From a psychology point of view, in order to give the clarity that Martin is talking about, we would tend to spend a bit more time talking through the scenario planning—"If this man went into the community, we think this about his risk, and this is how the risk could be managed.” We would probably just spend a bit more time spelling that out. There are green shoots; we will see how things go.

Sometimes, the recommendation was a bit reductionist; you would have a 10-page report with lots of nuances about risk and behaviour, and it would come down to “I recommend that…”, and that is all anyone would look at, particularly the prisoner. That can be quite challenging, in terms of future engagement, motivation, and working with someone. There are some benefits of taking that bit out. As I say, there are green shoots—we are waiting to see what happens. It is similar to how we work in courts, tribunals and other settings. The key thing for me was ensuring that all these changes stayed in line with guidance from the British Psychological Society and the Health and Care Professions Council, so that we make sure that we meet standards.

Q31            Rebecca Long Bailey: I still feel that there is a bit of ambiguity here so—just to be clear—if the probation officer thought that an offender posed a significant risk and did not feel that they should be released back into the community, they wouldn’t be able to say in their evidence, “I do not recommend that this offender be released.”

Martin Jones: That is my understanding of the way the recommendations work.

Dr Farrar: But they would be able to say, “Based on evidence, this is the risk that the offender poses.” That would be very clear.

Q32            Rebecca Long Bailey: What consultation has there been with the probation unions about the changes that you have made to Parole Board decision-making processes?

Sonia Flynn: There has been extensive consultation following the decision’s being made. We have agreed to set up a joint working group with the probation trade union, so that we can monitor these changes over time. If we feel that there need to be any changes to the guidance that we have issued to the staff, we will work jointly with them.

Q33            Rebecca Long Bailey: When was the last meeting that you had with the probation unions about this issue?

Sonia Flynn: I was with the JNC last week, or the week before. Plans for the membership of the joint working group have commenced.

Q34            Rebecca Long Bailey: There are a number of probation unions, as you are aware, and they haven’t indicated any lengthy or extensive consultation at all.

Sonia Flynn: Well, I know that there have been meetings with them to talk through the decision made by the then Deputy Prime Minister and how we intended to respond to that. We have also agreed to set up a joint working group with them to monitor the changes over time.

Rebecca Long Bailey: Thank you, Chair.

Q35            Tracey Crouch: Helpfully, from a time perspective, my questions have been answered, but I have two follow-up questions. One for you, Dr Farrar: what is the process for evaluating these reforms? Is there a timeframe for when you will look back to see whether they are working?

Dr Farrar: Yes. Sonia might want to pick this up in a bit more detail, but the process that she just mentioned with the trade unions is a very important part of that evaluation. We have implemented only a few of the public recommendations; a number of other recommendations will form part of legislation, so there will be extensive consultation on that.

Q36            Tracey Crouch: Are you expecting that legislation soon?

Dr Farrar: No, that still needs to be negotiated. There are one or two things in the recommendations that we are also piloting. If they prove positive, we will roll them out. All that leads to more evaluation and time for engagement and consultation as we move forward.

Q37            Tracey Crouch: Sorry to be a boring parliamentarian, but is that legislation likely to be primary or secondary? Does it need to be primary? Dr Farrar: I think it requires primary legislation, yes. We are obviously in discussions about potential for a slot.

Q38            Tracey Crouch: Finally, Dr Bailey, I am very pro the role of psychologists in this process. Presumably you would like more, not less, interaction in some of these decision-making processes for your profession.

Dr Bailey: We are very involved. The demand from the Parole Board for our input is high—so high, sometimes, that Mr Jones and I have a conversation about the level of demand. We are involved in a lot of Parole Board work. Also, as Martin said, on the panel, there are a lot of psychologists—I can’t remember exactly how many you said, Martin. The entire process is quite psychologically informed.

Q39            Tracey Crouch: We MPs obviously deal with acute mental health issues in our own constituencies, and one of the things that we often hear about is the lack of skilled professionals within the mental health system. Are you confident that there is a pipeline of people coming into the system to help with criminal forensic psychology, in the way that you deal with it in the Prison Service?

Dr Bailey: There is a pipeline through. It is fair to say that, nationally—not just in the Prison Service—there is a dearth of forensic psychologists. We also see that in clinical, and in some of the other psychology domains. I am talking about HMPPS, which is largely forensic psychology, although we do employ people in a few other domains. We also have mental health in-reach teams and a whole plethora of different psychologists doing different work in prisons. Am I confident that there is a pipeline? I am managing a pipeline. Psychologists are in a lot of demand, so I work very hard to keep their jobs enriched and fulfilling in order to retain staff.

Q40            Tracey Crouch: One final question. Is this something that you can do in the community as well? Is it something that you come into and go out of, or are you a full-time psychologist in the Prison Service?

Dr Bailey: In psychology services, we have a mainly prison-based team. There are some probation-based psychologists in HMPPS, but we tend to work with partners in the community. Some of our staff are part time in HMPPS and have another part-time role, which might be in a mental health or community setting. We have a variety of work roles in our team.

Q41            Tracey Crouch: But you are confident that the Ministry and those involved get your importance and understand your value.

Dr Farrar: Oh yes. I am a psychologist myself, and we absolutely do get that.

Q42            Katherine Fletcher: I want to pick up the point that you made, Sonia, about the change to the form and probation officers not being asked to give a specific recommendation. It occurs to me that that can perhaps improve the conversations between the probation officers and the customers, in that they do not feel that a judgment is being made on them in any situation, or it could, vice versa, slightly change behaviour. Very briefly, because I am conscious of time, I wonder if you could open that out. Does the probation service give better information in that because people know that they are not having a specific judgment?

Sonia Flynn: Yes. I think Jo referred to that, because she has observed this in psychology as well. The change focuses the prisoner’s attention on the content of the report, rather than the recommendation, and as I say, it resets the responsibility, so that the Parole Board is again the overseer and custodian of the decision. These are really early changes, but we will clearly monitor the impact. Certainly, the thinking behind the change was that it would improve engagement and the thoroughness with which all the evidence is presented to the board. As Jo said, focusing on the recommendation can be quite a reductionist approach, but yes, we undoubtedly will keep these changes under review.

Chair: I thank the first panel for their evidence, and for managing to get here despite transport difficulties this morning. Thank you very much.

Examination of Witnesses

Witnesses: Professor John Podmore and Professor Stephen Shute.

Q43            Chair: Good morning, and thank you for joining us. I welcome Professor Stephen Shute, who is a professor of criminal law and criminal justice at the University of Sussex, and John Podmore, who is an honorary professor at the School of Applied Social Sciences at Durham University. Professor Shute, can you tell us the extent to which Parole Board decisions are based on evidence?

  Professor Shute: Clearly, good evidence and information is an essential prerequisite to good risk assessment, and that is what the board is about. It is about risk assessment, and public protection is its top priority—it has said that for quite some time. The evidence that it has, as Martin Jones was indicating, is drawn from the written dossier that is submitted to the Parole Board by the Ministry of Justice, and the content of that dossier is governed by the schedules to the Parole Board rules, which set out the things that need to be in that dossier. Then, in an oral hearing, the board also has the opportunity to gather oral evidence from the prisoner and from witnesses. In assessing all that evidence, the board must do it with a critical eye, and in oral hearings it must press, it must probe, and it must test the oral evidence that it receives. What follows that assessment, of course, is a judgment as to whether the offender does or does not meet the statutory test for release.

Q44            Chair: Are you confident that they have the correct evidence when they are making these decisions?

Professor Shute: The Parole Board rules set out what needs to be in a dossier, and they identify the right things that the board needs to have. Of course, then they need to overlay that with a judgment as to what they need in an individual case, because that is a generic piece of guidance, and in individual cases, they may need different pieces of evidence. In the triaging process, a member will look at that dossier and decide if anything else is needed, and indeed, if you are going to an oral hearing,  the chair will then look at the dossier again and decide if anything else is needed. That process doesn’t always work; John Worboys’ case has been mentioned this morning. It went wrong in that case, quite badly. There were a whole series of points in the board’s decision-making process when the board could have got it right but didn’t, and there was also an error on the part of the Ministry of Justice, in that there was an administrative oversight, so the sentencing remarks were not included in the dossier.

Q45            Chair: How do we improve that, given that there is a whole series of issues?

Professor Shute: What you need is a robust review of the dossier, and you need the Parole Board members to do the job that they are charged to do; in Worboys they didn’t do it. How often those errors occur, of course I can’t say. It is only when something happens that brings it to light that you lift the stone and see the problem. What you don’t know is how often it occurs. Martin, I know, would say—quite a lot of things were said in the previous session that I would not fully concur with—that the board does now have a more robust process in the secretariat to check that the dossier that comes from the Ministry of Justice has the things in it that need to be in it. Without that, the process will go awry.

Q46            Chair: So what did you hear in the previous session that you were not in agreement with?

Professor Shute: Quite a number of things. Let me start with open conditions. Martin highlighted the importance of open conditions for providing a period of testing. It is in open conditions that you can get released on temporary licence—ROTL, as it is known. You can’t get that in closed conditions. At the same time, there has been quite a change in the release patterns of prisoners from closed conditions. If you go back over the history of the board, for a long time, an indeterminate-sentenced prisoner was expected to progress through the security categories, to move to category D—to open conditions—and only then be eligible for release. That meant that release from a category C prison was very exceptional.

Indeed, in the ’90s, when the Home Secretary had a ministerial veto over releases, my colleague Professor Roger Hood and I were observing decisions taken by paper-based panels about mandatory lifers—murderers. There were occasions when the chair of the panel floated the idea of releasing a prisoner directly from category C, only for the panel secretary to turn around and say, “That won’t be accepted by Ministers.” The chair then changed their mind. Certainly, release from a category A prison was unheard of—it was unthinkable; that is the word I am using in an article I am writing about this.

John Worboys was in a category A prison, HMP Wakefield, when the board decided to direct his release. Only a few years earlier, in 2011, Lord Justice Gross in the Court of Appeal on the Mackay case said that, because of the nature of the categorisation process, for a prisoner in category A the chances of being released on parole were “nil”—his word. That was not of course the case with Worboys. If we look back over the past few years, relative to the period that preceded them, we start to see releases from B and C, not just from D. Open prisons are no longer a necessary condition.

Q47            Chair: Perhaps you could describe what an open prison may be.

Professor Shute: One of the other features of open prisons, as compared to the closed prison estate, is that open prisons do not provide those rehabilitative programmes and other rehabilitative interventions that you have referred to; those are available in closed conditions. So, we have a tension now in the system: we are finding prisoners being released from closed conditions and, in very extreme cases—which in my view is problematic—category A prisoners being released. Worboys was not alone in that period. It is interesting to ask why, in the 2016-17 and 2017-18 financial years, we got to the point where a small number of category A prisoners were released. Remember that category A classification comes because the Secretary of State regards there to be a serious danger to the public should the prisoner escape.

Q48            Chair: Thank you. Apologies, we are slightly pressed for time, but do you have any particular points you want to make?

Professor Shute: Mr Stringer mentioned the point about the three-year period across which serious further offences are measured—indeed, not just serious further offences. The Parole Board’s review committee, which was established in 2003-04 and reinforced with independent members after the Hanson and White case and the Anthony Rice case—some of you may remember those—also operates on a serious further offence being committed within that three-year period.

My own view is that that is a problem. Why would we not want to take the learning that might be available to us if a prisoner released on licence commits a serious further offence four or five years after their release on licence? Most of the work of the Parole Board now concentrates on indeterminate sentenced prisoners—that is, prisoners who were considered dangerous by the sentencing court at the moment at which they were sentenced. That is a different group of prisoners from those who were considered before the 2003 reforms. Most of them are on licence for life—for the entirety of their natural life—and yet we are taking the smaller snapshot of a three-year period.

For those reasons, personally, I think that is a mistake. If we look at the reconviction rates for certain categories of offender, most notably sexual offenders, what we find—all the research has found this, including the research that I did with my colleague in the early 2000s following a parole sample—is that there is an initially low level of reconviction, then a slow accumulation over quite a long period. There are all sorts of reasons why that happens, but it means that you need to take a longer perspective in relation to sexual offenders and some other types of offenders than you would take—

Q49            Chair: We will need to move on. Apologies. We are pressed for time and we have another panel. I am sure there will be opportunities to explore some of these issues with some of my colleagues. Professor Podmore, do you have anything to add? How well is evidence used and can we improve on it?

Professor Podmore: If I may be blunt, they probably do not have enough evidence, they do not have the right evidence and a lot of it is not scientific. It is a big problem. We can draw a distinction. There is an important distinction between the prison and probation system and the parole system. We should cherish the parole system—it is the mark of a civilised society. We should challenge it, as well as cherish it. We should not politicise it and we should not blame it for wider ills. But the role of the prison and probation service is to take someone at the point of sentence—and yes, they deal with the more serious end—and throughout their sentence, to manage their sentence and, from the day they are sentenced, to prepare them, if appropriate, for the parole process, so that when they get before the Parole Board there is sufficient evidence, it is scientific, tested and interrogated, and the Parole Board can properly assess it. I don’t think that is happening at the moment, for a whole variety of reasons that we might want to talk about.

We talk about evidence. One particular thing I think we need to talk about is offending behaviour programmes. This is the Science and Technology Committee. The science around offending behaviour programmes needs—I and some of my colleagues have challenged it; we do not get very far. It is a bit of a closed door.

To illustrate a point, there was an inspection of HMP Onley, the report of which came out this morning. I would read you just one sentence from that: “Sentence progression should be at the heart of a category C training prison, but the offender management unit was, in most cases, providing little more than piecemeal support and prisoners said they rarely heard from their prison offender managers.” That is not untypical. Also in that report, which is also typical, is prisoners not getting out of their cells, not getting to work, not getting to education and training. Also in that report, and typical of many others, is a turnover of staff. In terms of assessment, there are complex tools and I commend this book to people, in terms of risk assessments—

Q50            Chair: What is the book?

Professor Podmore: The book is “Risk Assessment in Forensic Practice” by a colleague of mine, David Crighton. It goes into some of the quite complicated risk assessment tools that are out there. There are tools, and they are important, and they can be used, but they should be used in conjunction with—we are in the people business. We are talking about people—people in prison and potential victims outside, so we should not forget that. But what is important is that we know the prisoner.

There was an interesting question about whether we should have more policemen on the Parole Board. I think that is slightly intellectually unsound, if I may put it in those terms. We need the skills and competencies on the Parole Board, but we also need the skills and competencies within the prison and probation system, whereby people get to know their prisoners.

If you listen to prisoners, if you read tweets or blogs or whatever from prisoners and their families, one of the biggest problems is, as indicated here, that they do not see their prison offender managers, they do not see their community offender managers, or if a report is presented to the Parole Board, it is presented by someone who has only met them once over a 20-year sentence, or they have had three or four offender managers in prison or in the community over the past three or four years.

There is something in prison called OMIC, Offender Management in Custody. Nostalgia is not what it used to be—it is something I was doing in prisons in the ’80s. It is basically prison officers, social work in prisons, call it what you will, getting to know their prisoners. Because of the turnover in prison staff and the inexperience of prison staff, there are fewer and fewer people in prison who are really getting the opportunity and the training to get to know prisoners. Whether they give a recommendation is a side issue in a sense, because in my view the prison service is not having the processes and the people in place to make those assessments.

Q51            Chris Clarkson: Can I pick up on some things you have been saying, Professor Podmore? I want to talk about the balance of expertise and experience in offender management. How well do you think that balance has been struck already, and where do you think the deficiencies lie?

Professor Podmore: Do you mean in terms of offender management or in terms of the Parole Board?

Chris Clarkson: The Parole Board.

Professor Podmore: Yes, I think the Parole Board can be improved. Yes, there are judges, there are psychologists. The important thing with the Parole Board, and why I revere it so much, is to remember that it is a function of a civilised society, so it represents the community. Who is in prison? By and large, members of the community passing through. What does the prison system do? It takes people from communities, keeps them in a prison community, then puts them back into the community. So, fundamental to a parole system is that it has members of the community within it. Now, they need the right skills and competences, and that is about being able to read 1,000 pages—there are some technological issues there that I could talk about—and being able to debate, argue and interrogate.

I recently did a risk management review for the Scottish Prison Service, and I made some recommendations. There was some good practice there, not least that the Scottish Prison Service holds risk management boards within the prison. They evaluate the prisoner before the matter goes to the Parole Board for Scotland.

What I saw in there—there is a lesson here for the Parole Board—was some skilled and capable community-based and prison-based social workers. I saw some skilled addiction specialists. If I were looking at constituent Parole Board members, I would want those skills. I would want judges, psychologists and psychiatrists with skills, and I would want general members of the community who might have some of these people coming to live next door to them, but if you look at the morbidity of prisoners, it is mental health and it is drugs and alcohol.

Those specialisms have been put to one side, particularly addiction specialists. We know addiction is a problem, and it is not just about heroin and cocaine; it is about substances whose chemical composition is changing yesterday and tomorrow, and about which we know little in terms of how it affects people. Who else might we have on there? No, not more police officers. Let’s put in some addiction specialists. Let’s put in some skilled and competent social workers.

Q52            Chris Clarkson: As you say, it is a changing ecosystem; it is developing constantly. How well resourced do you feel the Parole Board is to get the evidence it needs in a timely fashion, especially if it is having to commission it outwith its own ranks?

Professor Podmore: The number of cases the Parole Board is dealing with, the number of members, the turnover of cases—I will be careful here, but I think an issue that illustrates the point is that about 40% of hearings are adjourned. You might like to check that with Martin Jones. That worries me. Why are hearings being adjourned? Is it about resources, the panel or the information that is coming to the Parole Board from the prison? In a sense, that is a wasted resource. I think Martin said there are 350 Parole Board members. That is not very many, and I have yet to come across an organisation that is sufficiently trained.

The Parole Board is a bit of a one size fits all; I think that was alluded to at the end of the previous session. We have within the prisoner population some seriously identifiable groups, and TACT prisoners—terrorist prisoners—have been identified. There are now TACT panels and people who are trained in TACT offending. There are problems with the interventions. We are only at the very beginning of understanding a large number of people sentenced under TACT who are being released after relatively short sentences. I am not sure we as society, or the Parole Board, are ready for that.

We have looked at sex offenders. There is a scandal that has been overlooked in terms of sex offender treatment. It was found that sex offender treatment programmes were making people worse. The data was sat on for five years. We need to talk about that. We need to talk about what it means for other offending programmes.

One of the things I identified in my Scottish review in terms of a cohort, which was alluded to, was serious organised crime. We know that there is disguised compliance within sex offending. Sex offenders can groom, condition and manipulate everything from their victims to Parole Board members, the police and prison staff. There is also what I would call corrupt compliance, which we have never talked about. If you are a serious organised criminal and you are coming before the Parole Board, okay, there is the very basic, “Here’s £1,000. Let me out on parole.” Put that to one side, there is what we call accelerated promotion.

If you are a serious organised criminal with a lot of influence, a lot of resources on the inside, skilled and competent at conditioning and a manipulator of the system, you can quickly go through sentence. You do all the programmes, you pass the programmes, do all the work, education and training, and go before the Board, probably in your Armani suit, being very convincing in terms of whether you are going to be released. There are skills and training we need in that, in terms of how the Parole Board works.

Q53            Chris Clarkson: A very blunt question: do you think the reports, as they stand, are fit for purpose, based on the resources available and how they are being discharged?

Professor Podmore: I think the Parole Board is doing a very good job. It is doing the best job it can with the information and resources it has got.

Q54            Chris Clarkson: Could it do more with more resources?

Professor Podmore: Everyone could do more with more resources. This is the issue for the Parole Board. If I wanted to change things for the Parole Board, I would not look at the Parole Board and beat it over the head—Worboys distracts us immensely. I would go back to what are the fundamental principles of the prison and probation system. It is to prepare people for parole. Are they doing that at the moment, with enough evidence, with the right evidence, with the scientific evidence? I would say, no.

The Parole Board can better do their job if they are presented with prisoners who have been, from day one, prepared for that process. We have talked about ROTL and open prisons. Yes, if someone is before the Parole Board and has spent a year or two years in an open prison, the first thing the Parole Board will say is, “Have they been out on temporary licence? Have they been out working in the community?”

If they haven’t, the Parole Board will think maybe that should happen. If that has happened in the year prior to that prisoner coming before the Parole Board, and if that person has worked out in the community, Monday to Friday, 9 to 5, and has had home visits, there is the evidence for the Parole Board to say, “Risk of serious harm is relatively low. Because the prison and probation system has done its duty.”

We should be building more open prisons, not so-called resettlement prisons in the middle of north Yorkshire that are nowhere near any communities. We have talked about people needing to be tested. You cannot be tested from a category A prison. You cannot really be tested from a closed prison. It is about taking risks. That is not the parole system taking the risks, but the prison and probation system taking risks. I do worry.

Usman Khan, who murdered Jack Merritt and Saskia Jones at Fishmongers’ Hall, was not a parole case, but he illustrated the point. He was a young man sentenced to nine years at the age of 18. Clearly a very dangerous man. He spent every single day of his sentence, from the age of 18 for nine years, when we knew he was going to come out, in highly secure conditions, surrounded by other terrorists. If we were going to make him a successful terrorist, we put every system in place in order to do that, and what happened? Okay, you have got to take some balanced risks.

My worry is the prison and probation system. My worry is that it is still the prison system, where it is security, security, security. Some 60,000 prisoners a year are released from prison. Locking up is easy. Releasing people safely into the community, so that they do not create victims outside, that is the difficult bit. The parole system is part of that, but we need to look at the prison and probation service before we start kicking the parole service.

Chris Clarkson: I will ask my last question of Professor Shute.

Chair: Sorry to interrupt, Chris. I am aware that time is passing and that we are getting behind. I would also ask the witnesses to be aware of that and keep answers as brief as you can.

Chris Clarkson: I am happy to concede and let the next person go.

Chair: Thanks. I will move now to Graham Stringer.

Q55            Graham Stringer: In the previous session, I focused on some obvious failures. If somebody is released on parole and they murder somebody, it is a pretty dramatic and awful failure. What reassessment of the process of releasing that person is carried out when you have one of those failures? Does the Parole Board look at what it could have done better? Or do they just say, “Made a mistake there. Very sorry.”?

Professor Shute: As I indicated earlier, the Parole Board established what it calls a review committee. It was set up in 2003-04 and its job is to look at the decision-making process by the board in those cases where a released prisoner commits a serious further offence within a three-year period.

Q56            Graham Stringer: On what Professor Podmore said, do you think two years is appropriate?

Professor Shute: Well, it is three years and I do not think that that is appropriate. I think it should be looking over a much longer period, as I indicated a moment ago. If a released prisoner commits a serious further offence four, five or six years after their release on licence, and they are on a life licence, why wouldn’t the Parole Board want to look at those decisions as well? My view is that it should have a much longer remit.

The workings of the review committee, despite the board’s commitment to openness, have always been shrouded with a degree of secrecy, but there is more secrecy now than there was in the very early days of the workings of the review committee. You cannot see the criteria that the review committee uses to make its judgments on whether the original release decision was appropriate on the board’s website, for example. As Martin Jones knows, I have suggested that it ought to be much more open about the way that committee works, because what more important committee could there be within the Parole Board than one that looks at those issues and tries to learn the lessons from the past?

Q57            Graham Stringer: That is very helpful. We are looking at how the Parole Board uses evidence when it comes to its conclusions, and we have also found some flaws in how it does it. One point in the briefing is that the external assessor, the probation officer from outside who writes a report before somebody is considered for parole, seems to be the largest factor in deciding what happens to the prisoner—whether they are given parole or not. Is that a fair assessment to you?

Professor Shute: If you look back at the influence of recommendations made by report writers, their correlation with parole outcomes is very, very strong. It has been very strong throughout the history of the board. If offender managers—formerly home probation officers—did not recommend parole, the chances of the prisoner being released by the board were very slim. There has been a strong correlation between recommendations and outcomes. That is one of the arguments that might be brought to bear when thinking about whether recommendations should be made by Ministry of Justice report writers, because you definitely do not want to have a situation where the Parole Board shortcuts its decision making—turning to the recommendations instead of critically assessing the evidence that underlies it.

Q58            Graham Stringer: That is good. One final question. You are both experts. If you could write one recommendation, having said what you have just said, that this Select Committee would make to the Government about the Parole Board, what would that recommendation be?

Professor Shute: I would make a number. One would relate to the membership. John said that 350 is not very many; I think it is quite a lot, and it is going to be getting up to nearly 500 soon. The problem with corralling a membership of that scale is ensuring there is a consistency of decision making, a common approach, and that the different risk appetites that different board members have are not producing aberrant decisions.

I would reduce the size of the board. I would have more full-time, employed members of the board, rather than relying on fee-receiving board members, which is how board members are paid. You have a small cohort of board members—13 or 14 in the last three years—all receiving fees in excess of £100,000. The highest fee is £170,000, and it is in their published report. I would start employing more board members full time, rather than, in effect, having full-time members through fees.

I disagree with Martin that the composition of panels does not affect outcomes. The composition of panels is crucial, and I think a single person within the board needs to take responsibility for the composition of panels and have an overview role. That should be the chair or the deputy chair.

Graham Stringer: Professor Podmore?

Professor Podmore: Transparency is what I would recommend. It has been said that Parole Board hearings should be public. I think that would open things up. I have a view of how they operate, but the wider public should see how they operate and be able to make an assessment as to how well the board functions and how well people are prepared for the board. By “transparency”, I mean having parole hearings in a setting like this—a court setting—not in a remote setting, where victims or families can just Zoom in. But that costs. There is a lot of talk about it. I don’t think much will change, because it would cost too much money. You cannot have transparent hearings in prisons—it is logistically too difficult—but transparency is essential. As well as my having a view, the rest of the community, on whose behalf the Parole Board acts, can then have a better-informed view of whether it is operating properly or not.

Q59            Tracey Crouch: You heard the conversations in the previous evidence session about the reforms that have taken place or are in the process of being implemented. I wonder what your views are on the advantages and disadvantages of being more prescriptive for the test, to ensure the protection of the public when assessing the risk posed by offenders.

Professor Podmore: What do you mean by “more prescriptive”?

Tracey Crouch: The prescriptive test that has been set out as part of the reform.

Professor Shute: This is in the root-and-branch review of parole. The former Secretary of State said that he wanted to make the test more prescriptive. Obviously, that will need primary legislation, because it is sitting in primary legislation at the moment, so to change it would require further primary legislation. One of the issues is that there are different interpretations of that statutory test, and even of a key term such as “public protection”, which is in the statutory test. What does it mean? There are different views about burden of proof issues. Indeed, you heard a view from Martin about burden of proof. I do not agree with what Martin said about burden of proof, if you meant to connect it through to the statutory test for release. But those ambiguities sit there, and there is the possibility that different board members will interpret a statutory test in different ways. Indeed, different judges in legal cases will interpret it in different ways. Clarification would be helpful.

As to setting the bar higher, one of the things one needs to look at is the release patterns over time. It is definitely the case that the board is now releasing a higher proportion of unreleased post-tariff indeterminate sentence prisoners—not the recall cases, which are different—than it has done historically. If you go back to the ’90s and the early 2000s, the release rate for murderers and mandatory lifers was between 10% and 20%. Now, it is up into the high 30s. Indeed, for IPPs, it is 40%. If you have a more prescriptive test, you will get a lower release rate. What should the release rate be? There is no easy answer to that. It depends on how much risk you want to take, because the higher the release rate, the greater the risk that the public is being exposed to. There is no a priori answer, which is why it needs to be a political decision. In my view, the statutory test needs to go through the parliamentary process.

Q60            Tracey Crouch: Sounds like you could already write your evidence to the Bill Committee.

Professor Podmore: I think the tests are sufficient and we are detracting from the real issues. I would leave them as they are.

Q61            Tracey Crouch: My final question is about the recommendations to remove the roles from the offender managers and prison psychologists. You heard what the previous witnesses were saying as justification for that, and I wonder if you have any views on that process.

Professor Shute: It was in answer to Graham’s question that I said I don’t think it is a good idea if board members are, on occasions, shortcutting the process and relying excessively on recommendations in the reports from report writers. They should always be looking at the underlying evidence. Clearly, removing recommendations will mean that they just can’t do that, and they will have to look at the evidence base.

On rehabilitative programmes, I think there ought to be intelligible rating scales for rehabilitative programmes, and individual judgments made by those who run those programmes on what difference going through that rehabilitative programme made to the prisoner. Those should be available to the board. I have long thought that. As far as I am aware, that information still isn’t going forward to the board.

Chair: That concludes our questions to this panel. I thank both of you for joining us this morning.

Examination of witnesses

Witnesses: Professor Seena Fazel and Dr Ruth Tully.

Q62            Chair: Welcome to our third panel. We have two witnesses joining us, one virtually. We have Professor Seena Fazel, a senior research fellow in the Department of Psychiatry at the University of Oxford. Joining us virtually, we have Dr Ruth Tully, a consultant forensic psychologist and the clinical lead at Tully Forensic Psychology. Welcome to both of you this morning.

I will start with Professor Fazel. Could you tell us what role psychologists are playing in assessing the risk that prisoners pose when they are considering parole?

Professor Fazel: Do you mean psychiatrists, because I am a psychiatrist—Dr Tully is a psychologist? I can speak to both.

Q63            Chair: Why don’t you tell us about psychiatrists, then?

Professor Fazel: The specific role that psychiatrists have is that they have quite a lot of training and skills in diagnosis, because diagnosis is often quite complicated. What psychiatrists can do is to deal with comorbidities quite well. They can diagnose people who have multiple disorders, often with physical health problems as well as mental health problems. That is one thing.

Another thing is of course that most psychiatrists work externally. They work in other spaces—hospitals, community settings—and they bring with them some knowledge of what community services are available and what is realistic in terms of community services. They have information about the sort of links between the prison and the community.

The third thing would probably be a little bit of scepticism. Most forensic psychiatrists in particular have quite a long training and they work with often very difficult offenders—exactly the type of people who can sometimes be difficult to assess. Forensic psychiatrists can bring an element of scepticism because they are working with these people day in, day out in high secure hospitals and medium secure hospitals like Broadmoor and those sorts of places.

Those are the three things: knowledge about diagnosis, a little bit about what community services exist and an element of scepticism to the process, which is I think quite helpful—it is a corrective, sometimes.

Q64            Chair: Thank you. Dr Tully, what about psychologists, then?

Dr Tully: Psychologists are involved in the parole process in a range of different ways. They are most commonly involved where there is serious violence and sexual violence and so on, which is many parole cases, of course. In the Prison Service, they will often be involved in terms of providing a risk assessment. That is often a structured professional judgment approach to risk assessment. Previously, as you have discussed quite a lot this morning, they would make a recommendation as to whether someone is safe to manage in the community in their opinion, on the basis of their risk assessment.

They might also be involved in other ways. For example, there may have been a diagnostic assessment of some sort—a personality assessment, for example.

My experience is within the Prison Service, historically, and the probation service, and also within secure psychiatric hospitals. I have provided evidence to hundreds of boards, both as part of the Prison Service, and also as an independent psychologist that is external to the service, brought in to give an opinion on a case. Primarily, the focus of that is risk, based on my experience, risk-assessment tools and the evidence. We offer an opinion to help inform the panel’s decision making.

Q65            Chair: What sort of challenges do you face when you are trying to gather evidence to present?

Dr Tully: Quite often, it is down to information. Again, I am aware that you have discussed that this morning. Any risk assessment is only as good as the information on which it is based. One of the key skills of a psychologist is recognising what type of information could be missing and what further assessments might be needed, whether they are psychological, psychiatric or other types of assessment, to inform the risk assessment. It is important that the individual doing the assessment identifies the limitations in the information provided.

The parole dossier is often standard; it is produced. Some are better informed than others. For example, determinate sentence recall dossiers in my experience are much more sparse than typical life sentence or indeterminate sentence prisoners’ dossiers, so I am often asking for more information. It is really down to the assessor to identify those limitations.

Other skills include forming a rapport with difficult-to-reach, difficult-to-assess people. Again, that is down to the skill of the assessor. Sometimes operational issues can understandably get in the way of doing an assessment—for example, if you are due to see a prisoner and the prison is on lockdown for some reason. There are challenges, but largely any limitations of a risk assessment should be identified by the assessor, and they usually are in my experience.

Q66            Chair: If evidence is sparse, as you have just mentioned, what can be done to improve it?

Dr Tully: I think a lot has been done to get evidence, particularly in the light of the Worboys review and other similar issues. Since then, I have had quite a few cases where there have been allegations and the Parole Board have had to make lots of quite complex directions involving multiple agencies to try to get reports that were done around the time of allegations, even if they were 20 years ago—for instance, by trying to get multiple police forces to work together to provide information. I had one recently where the Parole Board directed that two particular officers, who were now retired, be spoken to so that information could be gathered.

There is a lot going on to get the relevant information so that risk assessors and the Parole Board can consider that in their decision making. Again, that is down to the humans involved in the process—the board and the professional and expert witnesses—identifying what additional evidence they need. It is possible to get the evidence you need, but there are those challenges, and sometimes it is essential to say, “Actually my risk assessment is limited for these reasons.” When it is not possible to access information, it is often due to the passage of time and documents not being available anymore.

Chair: Thank you.

Q67            Katherine Fletcher: Thank you both. Your time is appreciated. This has been a really interesting session. As I have been sat listening, I have been thinking, “How do we help people start to quantify the risks?” because that is ultimately what we are doing. Professor Shute said in the previous session, “What should the release rate of murderers be?” I want to use the expertise of both of you to understand what is possible from a scientific perspective in quantification. Although I appreciate that this would not cover everybody who goes through parole boards, I just want to home in on psychopaths. Quite famously, Hare’s psychopathy checklist is an attempt to quantify the likelihood of psychopathy and actions that the broader community would consider deleterious due to biological and environmental factors. Do you have anything like that? I will start with Professor Fazel, and then I will bring you in, Dr Tully.

Professor Fazel: Quite a lot of my work is around evidence synthesis, so I spend a lot of time trying to bring together evidence and make sense of it. What is clear is that the evidence around risk assessment tools—things like the psychopathy checklist—is actually quite poor to moderate. The tools that are currently used are not very good, despite what people say. If you use the standards of the rest of, let’s say, science or medicine, they perform moderately at best, and most of the time their performance is actually quite poor. They have been taken up wholescale partly because of skilful marketing, and partly because there was a lack of anything else, so it was understandable that people just latched on to whatever was there.

Katherine Fletcher: Oh, it gives me a number—hooray!

Professor Fazel: Yes. In the cold light of day, if you look at good-quality studies—these are large studies and it is helpful if they are done by researchers with high-quality standards—there is little out there that is any good. They are being used. They are very time consuming. There is an element that they can inform decision making, but in our view they should not be a key determinant. We did a review just this year and that was our conclusion. These tools may have a role, but if anything, it would be to identify low risk rather than high risk. That is what they are better at.

Also, people need to be aware of the limitations of these tools. They sometimes provide some sort of framework, but there are also some harms and costs associated with them, and relying on them too much is a problem. Hence, I agree with what the previous panel said about having experts on these panels—people who can actually appraise all the evidence and not just get a bunch of checklist tools and come up with conclusions on that basis. They need to be critically aware of the evidence and the limitations of these tools.

Q68            Katherine Fletcher: Fair enough. Do you share that view, Dr Tully?

Dr Tully: Not quite in those words, no. There is an evidence base for the use of these tools. Research consistently shows that—certainly above unstructured clinical judgment—a structured tool has a better predictive value in terms of the research. On an individual level, what they help identify are risk-raising scenarios. I don’t want to give the impression that these tools are just about labelling and giving a quantification, because that is one thing they do. I will perhaps come away from the psychopathy tool because it is more of a personality assessment than a risk assessment, although the construct of psychopathy is linked to increased risk.

Mainly, the tools used are, for example, the HCR-20 and the RSVP. They are violence and sexual violence tools, and they are used to inform a judgment rather than to dictate it. For example, I may assess someone whose risk according to my application of that tool, which is based on the evidence—it is not an algorithm, it is a structured clinical judgment with guidance—and they may come out, in my opinion, as moderate.

Q69            Katherine Fletcher: I just want to make sure I am following you because you’re dead clever. For example, is it like, “Has the person beaten up a woman in the past? Has he also beaten the dog up?”? Is it that you have a fact that you slot into this thing?

Dr Tully: No. It might help if I add a bit of background. Before we had these tools, professionals relied on unstructured clinical judgments. They would weigh up the evidence and give an opinion, maybe on manageability or level of risk. The next development in violent risk assessment was the actuarial movement and actuarial scales. They are a bit like your car insurance, so like what you described there. “Has this ever happened?” Tick. “Has this ever happened?” Tick. It is slotting in the facts, looking at the research and then saying, for example, “Okay, this person is similar to a group of people who offended at a 10% rate over two years.”

Things have now moved on, so we tend to use a structured professional judgment, which involves less tick-boxing. There are identified risks that you look at whether they are there and that the research on these individual items supports as being linked to risks, such as a history of serious violence, problems with intimate relationships or problems with supervision. It also looks at the clinical aspects of risk. That would include somebody’s insight into their risk, their understanding of why they committed their offence, how much they have addressed that, and their current mental health. It is not just whether they have ever had a mental health problem, for example. It takes into account things like progress in treatment.

One of the earlier witnesses said that they are not aware of reports of an individual prisoner’s progress in offending behaviour programmes being provided to the panel. I can say that they routinely are. Updates from those working with these people are provided to the panel. They are also provided in the dossier, and I would consider that as part of my risk assessment. I would bring all that clinical data, as well as historical information—that is, “Have you ever done x, y or z?”—together to form an opinion on risk.

What I was going to say earlier was that somebody could present a moderate risk on that tool, but overall it is not the tool that dictates my opinion; it is my experience combined with that. Someone else could come out as moderate, but one of those people may be safely manageable in the community at that point and one of them may not be, because it also depends on the imminence of that risk. There are things that might increase the risk that are likely to happen in the community, or things that they have not yet addressed in terms of core intervention work. So these tools are used in an informed way.

Q70            Katherine Fletcher: Can you give us a brief “for instance”, so that we can follow you? If somebody has a problem with red tricycles, what is the likelihood of seeing red tricycles in the community—is it that kind of thing?

Dr Tully: It is not as straightforward as that. For example, you might have somebody who has a history of violence within intimate relationships when combined with alcohol use and unstable accommodation. As a psychologist, I would look at how these factors came together, whether the person has addressed the core, underpinning factors, which are often, and in the example I have just given might be, about attitudes towards women and the reasons underpinning alcohol use as well as the actual alcohol use. I would look at those things coming together and the likelihood that the person may form a relationship in the future and that they have more access to potential victims in the community, to help me weigh up a decision about how imminent risk is. That would include their understanding of their offending, which is often developed through engaging in offending behaviour programmes, but they are not the only route to reducing risk.

Q71            Katherine Fletcher: I am conscious of time, so I would be grateful if each of you gave a brief answer to this question, which was raised with the previous panel. The process of engagement and improvement and addressing behaviours in some instances has a deleterious effect. I just wondered: in the risk assessment and the roles that psychiatrists and psychologists have to play, how are you stopping us, frankly, teaching sex offenders how to pass these tests and avoid getting caught? Briefly, because I am conscious of time. Professor Fazel?

Professor Fazel: The first thing to say is that the evidence base behind a lot of these treatments is very poor. We heard about the sex offender treatment programme, but actually, this also exists across lots of other programmes—the evidence base is poor. You hear a lot of people talking about “before and after” studies. What you need to look at is trials, and if you look at the trial evidence, it’s really poor. That’s the first thing to consider.

The second thing to consider is exactly what I said: there is a level of scepticism. People can learn how to game the system—particularly if they have a history of gaming systems. So you need to bring expertise to these panels, those who can interview people and see through checklists—there is a bit of a culture of this.

Q72            Katherine Fletcher: So by its nature it is about long-form interviews with experienced experts?

Professor Fazel: It doesn’t have to be long form, but it is about clinical expertise/ That is where psychiatrists and psychologists can add to it. I will just add that these tools—the ones that Dr Tully has mentioned—provide broad categories, like low, medium and high, and the problem with those categories is that they are also quite subjective. If you look at what “high” means, over a one-year period it can actually mean anything between 9% and 90% if you look at the evidence. So I don’t think that these tools are really the answer to these problems. They may help a little bit with informing decision making, but actually, you need better-quality tools that are based on newer evidence and developed in the right way—

Q73            Katherine Fletcher: So there is a call to arms for the academic community to continue to develop tools—

Professor Fazel: Well, no, they are—this is partly what I have been doing in the last 10 years and other people have been doing as well. It’s about having transparent, high-quality tools. A lot of the tools that people are using were developed 20 years ago and are based on very dated methods.

Katherine Fletcher: Yes, Hare’s came out in the ’70s.

Professor Fazel: And there are also the other ones that people have mentioned. They are 20 years out of date. The methods have dramatically improved since then. We have learned a lot from the rest of science and medicine about how to do these things properly, and this is only just appearing in this area—criminal justice is quite slow on the uptake.

Q74            Katherine Fletcher: Dr Tully, would you add anything to that? I could see you nodding at one bit.

Dr Tully: I would say the skill is in the application of any tool, to inform any opinion, that experienced psychiatrists and psychologists bring to the table. If you think about a prisoner who may still be high risk and may be motivated to offend in the future, of course they are going to tell me, if they want to get released, how much they have reduced their risk, how they have learned and all the things they think that I might want to hear. So it is about, importantly, the public knowing that we don’t take people at face value on everything they say. We look to the evidence; we look to behavioural evidence. We look to the past and we look to the future.

That is the importance of providing an opinion. In terms of professionals within the prison service and psychologists within the prison service not being able to provide an opinion, I disagree that panels go with opinions on recommendations for release rather than the evidence. In my experience of hundreds of hearings, they absolutely address the evidence for those opinions and they challenge those opinions. My professional view would be that adding ambiguity by removing the ability to provide an opinion is a mistake.

Katherine Fletcher: Thank you.

Q75            Graham Stringer: Professor Fazel, it is refreshing to hear a scientist say that we should be sceptical about the evidence from a particular branch of science. Is that another way of saying that psychiatry, unlike astronomy or other sciences, is not predictive?

Professor Fazel: That is broadly right. We are in the business of what is called prognosis. We do a lot of prognoses, and what a lot of science and medicine does is help us improve prognoses. If you look at your weather app on your phone, it is now pretty good compared with 20 years ago. There is a lot of prognosis that is trying to predict what happens to, let us say, a disease. In this area we go a little bit beyond that and try to predict what happens to people in terms of offending. The criminal justice system has been very slow to pick up on how to do this properly.

Over the last 20 years, there have been dramatic improvements in the methods and standards that you apply to say something is reasonably good, and that has not really happened in criminal justice. We still rely on tools 20 years out of date. People talk about reliability and predictability, but they use these words in ways that are not very precise. That is where evidence synthesis is key. You need a degree of scepticism not only in your clinical work, but also in the way you appraise evidence. Psychiatry and psychology can add some value to this process by that approach and razor-like appraisal of evidence.

Q76            Graham Stringer: I am not—and I am sure you are not—knocking the scientific method as it applies to human behaviour. I was asking a question about how the scientific method does not seem to have found a way of predicting reoffending.

Professor Fazel: Oh, I see what you are saying. I think it has, so I disagree, but it has in the sense of giving you probabilities—in the same way the weather app does. We have now gone beyond the high-low-medium business, which was dominant in this field, and we are now thinking about probabilistic interpretations of future behaviour. They work quite well over a two-year period, a five-year period, a 10-year period. They will not work well necessarily for a month or two months. As long as you see the limitations with that, that can help provide some sort of framework going forward. If I tell you there is a half a chance of it raining today or tomorrow, that might help with a decision about whether to carry an umbrella or not. In the same way, predictions in science are using a probabilistic framework, but at the same time seeing the limitations with that, because that is only going to tell you a probability. It will not tell you anything more certain than that.

Graham Stringer: That is exactly where I was hoping you would lead. Accepting that we are dealing with statistics and probability, it is very difficult when it comes to individual behaviour, isn’t it?

Professor Fazel: Yes.

Q77            Graham Stringer: Very difficult indeed when it comes to writing reports. How do you then interpret reports about an individual and the probabilistic behaviour? How do you get over that difficulty?

Professor Fazel: It is thinking about what you can do to reduce risk. That is key. You can talk about all the things that can be done. That is where I think psychiatrists have a role because we understand the role of medication, for instance, in mental illness—people with mental health problems—and the linkage to community services, the GPs and also community mental health services. I suppose we are advocates for the importance of having well-funded community services. Ultimately, if people are going to be safe in the community, they need well-funded mental health services that can support individuals if they have mental health problems. We know that about one in seven have a serious mental health problem.

Q78            Tracey Crouch: That is a really nice segue into the next series of questions, specifically on mental health and the risk of offending. Martin Jones has said that it is no secret that a significant number of prisoners have a mental health problem, but one of the fundamental problems is that the term “mental health” is often ambiguous. You, Professor Fazel, have written a number of papers about the significant relationship between particular mental health conditions and reoffending, and also types of reoffending. My question is primarily for you, but I also seek your views, Dr Tully: how should mental health factors be assessed and weighed when considering the likelihood of reoffending?

Professor Fazel: They are important, partly because they are modifiable, so they can be dealt with. You cannot deal with the fact that someone is very young—we know that age is a predictor—but you can deal with the fact that someone has a treatable mental health problem. When we talk about mental health, we are talking about diagnosable mental health conditions; the reason we diagnose them is because we can treat them, so there is a pragmatic element to this.

Broadly speaking, when you look at the evidence around prisons, about one in seven prisoners has a serious mental health problem. By that, I mean a mental health problem that can be treated, usually with medication that is effective—we know that it is effective from trials over 50 years or so. It is important in that sense, because it is something that you can do something about, but there is a lot we cannot do anything about. People turn up in prison with lifelong histories of problems, and you cannot realistically address that in short periods of time, but you can deal with mental health problems.

The other problem, alluded to previously, was drugs and alcohol. That is very important, because alcohol often gets overlooked. There was a lot of emphasis on drugs, particularly drugs that you can inject—for obvious reasons—about 20 or 30 years ago, because of HIV problems. In a way, alcohol got left behind, and that is another area where we know there are effective treatments; there are often psychological treatments, such as talking therapies and group therapies. It is a very important area; about one in five people turns up in prison with significant alcohol and drug problems. Those are the two issues that should be weighted. As for their weights, I would say they are as not as large as, say, previous offending, but they are substantial, in the sense that they make a difference, and it is a difference that you can change.

Q79            Tracey Crouch: I want to pick up on something you said, Professor Fazel. You mentioned substance misuse. You heard the evidence from the previous witness, John Podmore, about potentially including addiction specialists on parole boards. Is that something you think would be useful, in terms of assessment?

Professor Fazel: Not necessarily. It should be core to the treatment services available, not just in prison—where, as I said, disorders such as alcohol disorders sometimes get overlooked—but in the community. We have challenges with treating people effectively for substance problems in the community, and if you want to maintain those lower levels of reoffending beyond two years, it is important to fund those treatment facilities in the community, and to fund them well.

Q80            Tracey Crouch: Thank you. Dr Tully?

Dr Tully: I emphasise the importance of individual assessment, particularly where mental health is concerned. It is not just about someone having a mental health problem; it is about understanding the nexus between that and other risk factors that might come together to contribute to the risk of violence. Not everyone with a mental health problem is violent, and not everyone who is violent has a mental health problem.

It is important that we understand the impact of mental health on why someone committed a serious offence, so that we can provide a hypothesis on how to adequately weight that in risk assessment. That comes into psychiatry and psychology all the time—understanding an individual and how something impacts on them, rather than relying only on probabilistic estimates, for example.

Q81            Tracey Crouch: Do you think parole boards have the nuanced understanding to make those kinds of assessment?

Dr Tully: I do, but that is also why they look to professional and expert witnesses for their opinions on these matters. For example, two professionals might weight things slightly differently; that may be because of their professional background, or because of their experience. Sometimes views on progression and safe management in the community differ between professionals in the same case, but that promotes a healthy level of discussion; the evidence base is quite heavily addressed at parole hearings. Panels do have the experience and expertise to figure that out, to conduct their own risk assessment, and to make their decision. They take on board and listen to the evidence of experts, and they weight that accordingly, having checked whether the expert has had access to information, has considered it adequately, and has the right experience to advise them. In short, yes—but a combination of factors come together to inform their decision making.

Tracey Crouch: Brilliant. I am conscious of time, so thank you.

Chair: Thanks, Tracey. Our final question today is from Chris Clarkson.

Q82            Chris Clarkson: This is to both panellists. Because this is the Science and Technology Committee, we have to have a science and tech question. What, if any, role do you see for artificial intelligence in helping to assess the risk posed by offenders when they are being looked at by the parole board, and do you see any advantages or drawbacks to the use of that technology?

Professor Fazel: I have done some work on this. If, by “artificial intelligence”, you mean using big data and high-quality methods, I would say that it definitely has an important role. If you mean creating an algorithm that no one really understands, that is marketed, and that people make a lot of money selling to parole boards, I would say it has very little role. Anything that is developed using high-quality methods and using big data has to be transparent. The protocols have to be published. The algorithm has to be published. Everyone needs to be able to see how it works, because that is the only way that you can gain the confidence of the people who use it, and of the public. That is often the missing link in this. Yes, if by “artificial intelligence” you mean high-quality methods that are transparent in how they present their information and tools.

Once it gets complicated, I do not think that it has a role, because often the simple approaches work very well. Once you add complexity to the system, it costs a lot of money with very little additional benefit. I would go for scalable, simple tools that do not take a lot of time. Be transparent. Publish everything about them, so that people can see how they work, and that will inspire confidence, I think. That is also scientifically the best way of doing it, in my opinion.

Dr Tully: I think it could add some limited information to actuarial data on group behaviour—"People with these features may reoffend.” It comes back to something similar to what I was describing as actuarial tools, although this is machine learning, and algorithms will develop over time. On the role of that in making a decision about an individual’s safe release into the community, I would be horrified if artificial intelligence or machine learning had a huge role in that, both as a psychologist and as a member of the public who deserves to be kept safe.

An analogy that I might use is the “unexpected item in bagging area” when you get information that is incomplete. I recognise it when I have been given the wrong data—if I have been given someone else’s list of previous convictions, who had the same name and a similar date of birth—so I can recognise flaws in the data. I can also look at how to accommodate the flaws in the data. The “unexpected item in bagging area” analogy works also if I am at a supermarket and I weigh a lime on the scale and put it in as a lemon. I can recognise that, as a human, and I can take action to correct it.

AI might provide almost a false sense of security. We think that things like AI and machine learning are progressive because they sound like they are, and it is the way of the future. We can access big data to look at helping with decision making, but when we make decisions about human beings, we need to recognise the human factors—professionals being able to hypothesise about what might increase risk, and what about a person might tell us that their risk is going up. We can ask, “What do I need to look out for? What does probation need to look out for?”. I am yet to be convinced that AI should have a huge role in decision making in the future, but I am open to looking at how the science develops over time.

Professor Fazel: It is worth adding that artificial intelligence also adds a corrective, because human beings are quite biased, so they often make the mistake with the lemon and the lime. There is a corrective element to it, which is important if the clinical algorithm is transparent. There are often straw man arguments against artificial intelligence. I do think that it has a role, as long as it is done properly.

Chris Clarkson: Excellent. Thank you very much.

Chair: I lied and said that that was the last question. Tracey Crouch would like to ask a final question.

Q83            Tracey Crouch: It just follows on rather nicely. I don’t know whether you heard the first evidence session, but I asked a question about the pipeline for psychologists and psychiatrists. It feels to me that forensic psychology and psychiatry is incredibly important in getting these decisions right. Do you think that enough people are coming through that pipeline for us to feel confident that these decisions will continue to be made in the right way going forward? Dr Tully, why don’t I come to you, as a much wanted and needed forensic psychologist?

Dr Tully: My training was in the Prison Service. I became a trainee psychologist. I qualified in the Prison Service, doing my doctorate while working across prisons and the NHS. Prisons are doing a lot, as was mentioned earlier, to help to maintain people and the staff group in prisons. I still think that there is a lot to be done. In terms of people coming on board as trainees, life experience is really important, as is experience in other settings. I was hugely enriched by working in the NHS and working in the community when I did my training. Not everybody has that experience, but it equipped me with more experience in the field of mental health and personality than I would have had if I stayed in prisons. It is important to get a wide range of experience, to get the right people in these roles, and to keep them in these roles. That is more of a workforce issue, which perhaps Dr Bailey can advise on a little more.

I did not stay in the Prison Service; that was my professional choice. I now have a more varied role, so it might also be worth seeking the views of those who are no longer in prisons to see how to maintain the staff in them. A lot of effort is going into making sure that there is a pipeline of psychologists. Prison services have outsourced some of the work to external psychologists, which also helps to provide a breadth of experience and people with different kinds of training and skills. There is a lot going on with that, but it will be an ongoing issue to make sure that people have the experience to be able to assess very serious offenders.

Professor Fazel: Some people come originally from medical school. As long as increasing numbers of people go to medical school, and more and more people choose psychiatry—in psychiatry, forensic psychiatry is very popular—the pipeline issue will be addressed much more upstream.

Chair: I thank our witnesses on the final panel for joining us this morning. That concludes the evidence for today’s session of the Science and Technology Committee.