Skip to main content
EVIDENCE

[Recorded by Electronic Apparatus]

Thursday, February 13, 1997

.1109

[English]

The Chair: Welcome to you all.

We are today considering Bill C-55 as well as Bill C-254, Val Meredith's private member's bill.

We have with us Professor Robert Hare from the University of British Columbia and Professor Stephen Hart from Simon Fraser University.

.1110

Professor Hare, will we hear from you first?

Professor Robert Hare (Faculty of Psychology, University of British Columbia): I think so, yes.

The Chair: After you we'll hear from Professor Hart, and then we'll have a lot of questions.

Welcome to you both.

Prof. Hare: We're appearing before this committee not to argue for or against the bill but simply to provide information that we think might be useful in your deliberations. Both bills are relevant to the types of research we and our colleagues have been doing for quite a few years now, in my case for some 25 years.

We're talking about the ability to assess dangerousness in the context of the criminal justice system. In the past it has been almost a given that you cannot assess dangerousness. You cannot determine who is at risk to commit a violent offence. This old conventional wisdom has long since been replaced by information and techniques that allow us to make reasonably good predictions about who is likely to commit a violent offence. In other words, we're looking at risk for violence. We're trying to develop risk procedures for determining the potential for violence.

Much of this research revolves around the clinical construct of psychopathy, the psychopathic personality. Now, a lot of people don't like labels or attempts to pigeon-hole people. I only would like to point out that we're defining this particular disorder in terms of a set of personality traits or characteristics and behaviours. We have a group of these traits and behaviours. When we have enough of them, we refer to the individual as a psychopath.

Of course, you don't have to use that terminology. It would be possible for anybody interested in this particular area to use the instrument that has now become the standard measure for the disorder. It's referred to as the ``psychopathy check-list'', or PCL-R. The ``R'' stands for ``revised''. This is a 20-item scale completed on the basis of detailed information, semi-structured interview with the individual and so forth.

On could always, instead of using the term ``psychopath'', say that we're dealing with somebody who has a high score on this particular scale, and we know the high score has implications for the criminal justice system. It's to this that I now turn.

There is a great deal of research...most of it in Canada, by the way. Perhaps because of the way our system is set up it's much easier for Canadians to do this type of research than it is for the Americans, which is always a source of surprise to them. The evidence from right across Canada - Ontario, Quebec and the prairie and Pacific regions - is that it is possible to use instruments of this type, plus other risk-assessment devices, to make reasonably good predictions about outcome. Outcome is defined as, say, reoffending following release from an institution.

The evidence is quite clear. If we take only this construct of psychopathy as measured by the psychopathy check-list, the PCL-R, it is possible to estimate that an individual with a very high score on this particular instrument is very likely to reoffend within four or five years following release from prison. The likelihood can be as high as 75% or 80%, compared with a much lower value for other offenders - that is, those who have lower scores. So their outcome might be 25% or 30%.

Much of the research indicates that the odds that one of these individuals we have labelled as a psychopath will commit another offence following release from prison, including a violent offence, would be four or five times the odds for people who do not meet these criteria.

.1115

This particular way of looking at personality disorder is the central point of most of the more recent risk equations that have been developed for this particular purpose.

This information becomes particularly important if we look at sex offenders. Research conducted in Ontario and research conducted in British Columbia indicate a very clear point. A psychopath - somebody high on this instrument - who is sexually turned on or sexually aroused by violence is particularly dangerous. So the research with adults in Ontario indicates that people with a high score on psychopathy who show evidence of being sexually aroused by depictions of violent events, including rape and violent interaction between male and female - a very deadly combination - are at very high risk for committing another sexual offence. By high risk I mean that within the first few years following the release from an institution and following fairly intensive treatment, these individuals would probably reoffend sexually at a rate of 65%, 70%, or 75% within a relatively short period of time.

Evidence from adolescent sex offenders, which was accumulated in British Columbia at youth court services, indicates precisely the same thing. We can identify these psychopathic individuals very well from age 13 or 14, without any difficulty at all. Having done so, the diagnosis or assessment has extremely important implications for what's going to happen over the next couple of years.

In particular, one study of adolescent sex offenders - there are well over 200 of them now - indicates that the psychopathic adolescents who are sexually aroused by violence reoffend at a very high rate within the first year. In fact the estimates are that within 30 months, almost 90% of these individuals - the psychopaths who are sexually aroused by violence - are going to reoffend and be charged with and/or convicted of another offence, a very conservative estimate of criminality. In this particular case, the offences are not confined to sexual offences; they can include all sorts of offences.

We could summarize that part of it by saying a psychopathic adolescent who is sexually excited by violence is generally dangerous; he's going to commit a wide variety of different crimes. By comparison, the other adolescent offenders will reoffend generally at a rate of 20%, 25%, or 30%, so there could be a three or four times difference between these two particular groups.

I have one other point. Even among adolescents who have been sent for evaluation by the courts, we have a 10-year follow-up on these particular individuals, and the reoffence rate is extremely high for the psychopaths. Within 10 years after release from the institution, 80% to 85% have been re-convicted or charged with another offence.

To summarize, what I'm trying to say is we now have at our disposal instruments that can be used reliably by different clinicians and that have major implications for outcome. That is, they predict not only general reoffending but most particularly violent reoffending. And it turns out that this particular construct we're talking about - that is, psychopathy - is the single most potent predictor of violent reoffending that we have currently at our disposal.

Thank you.

The Chair: Thank you, Professor Hare.

Professor Hart.

Professor Stephen Hart (Faculty of Psychology, Simon Fraser University): Good morning. I will take less time than Dr. Hare has taken. Many of my views were gathered from working with him in the past.

I work in a slightly different area now, which is really to build upon the research that's been done by people like Professor Hare and to work more closely with Corrections and forensic psychiatric agencies to try to apply this research in real-world settings.

.1120

One of the things that's very important for me is to make clear some basic conclusions from the research that are relevant to this bill. I would echo Dr. Hare's conclusion that the most important thing here is that violent and sexually violent recidivism is predictable. The scientific research conducted in Canada and abroad is quite clear in suggesting that it is now possible to identify offenders who are at high risk for violence.

Scientists cannot tell you what you should do with this knowledge, but we can certainly tell you it is possible to identify, reliably and accurately, offenders who are at high risk for violence.

A second important conclusion here is that the factors that reliably and accurately predict violence are relatively few. There are a small number of risk factors that are known to be important. Because of this, any type of risk assessment procedure should include those risk factors, at a minimum. So although scientists can identify people who are at high risk for violence, it is sometimes difficult to ensure that people working in the field are actually assessing the correct variables or risk factors.

I'm not sure that the Criminal Code and the Corrections and Conditional Release Act are the proper places to actually try to develop a full set of policies and procedures concerning risk assessment, but one recommendation I have is it's important for the House of Commons to try to encourage the Correctional Service of Canada or the National Parole Board to develop some very specific and detailed procedures for assessing risk for violence, in order to identify offenders who would be captured under these proposed bills - that is, long-term offenders and dangerous offenders.

I can also talk about three issues it would be helpful right now to try to clarify. One of them is to try to identify the goals of the bills, the goals of these pieces of legislation. It is very difficult for researchers to try to evaluate the effectiveness of legislation or proposed legislation without an extremely clear statement of the goals of that legislation.

So, for example, in the preamble to the bill or in some other kind of public statement, it would be very nice if we could have it made clear for us once and for all the fact that this legislation is designed to protect public safety; that's the primary goal and in fact may be the only goal. If that's true, then we should make sure any procedures used to identify offenders are actually identifying those at highest risk for violence - not just people who are reasonably high-risk, but the highest risk for violence.

A second recommendation - and I noted this earlier - is if we can try to specify some risk assessment procedures, this will help to ensure public safety and quality of assessments. It will also help this proposed legislation to withstand legal criticisms by pointing out that it is in fact non-arbitrary, that we are identifying the most dangerous people and saving our most severe sanctions for those people.

The third point is we may easily develop risk assessment procedures and we may identify ways to identify high-risk offenders and detain them, but there must be additional risk management procedures or offender management procedures put into place, or at least an attempt to develop new procedures.

.1125

Another way for these types of proposed legislation to withstand legal criticism is to try to make sure offenders have a way of getting out of the system better - that is, to prove themselves not dangerous over time. Right now we're doing a very good job of trying to identify people and get them into the system for extended periods of time and not devoting much in the way of resources to getting them out. Or if we are devoting resources, it's for the same things we've done for the last 15 or 20 years. There have been relatively few attempts to develop something new or original, and I think we should put a higher priority on trying to develop management procedures.

That ends my comments.

The Chair: Thank you very much.

We'll go to ten-minute rounds. For the Bloc, Mr. St-Laurent, ten minutes.

[Translation]

Mr. St-Laurent (Manicouagan): Your data is both interesting and troubling. This will be a meaningful discussion, because you are providing rather remarkable answers to questions on issues that concern us, particularly the risk of recidivism amongst teenage sex offenders.

Mr. Robert Hare was the first person to say that we could identify the risk in 75% of the cases. How did you come up with this figure? Today we can all do this. At one point, an individual goes before the parole board and this is quite difficult. This is the first leg of the screening process and, although it is quite strict, it is not always successful.

Even your statistics indicate that the failure rate at this level is greater than 30%. However, you claim that in 75% of the cases, you can identify individuals who are very likely to reoffend.

[English]

Prof. Hare: Part of the answer has to do with what we mean by ``psychopathic personality''. We're identifying individuals who I think most of you would know; you would be able to point to examples in your own life and your own experience of people of this sort. These are individuals whose predominant characteristic is a stunning lack of interest in you as an individual. These are people who lack what people refer to as empathy. They're not socially connected to the rest of society. They don't believe that other people have rights in the same way they have rights.

So we have individuals who are not only lacking in concern or empathy for the welfare and well-being of other people. We have individuals who tend to be impulsive, who have a grandiose sense of their own self-importance. These are egocentric individuals, people who can be extremely irresponsible. They do not have long-term realistic goals, and they are unwilling to accept responsibility for their own actions.

As a matter of fact, in the criminal justice system one of the questions that's always asked, I presume, in parole board hearings is ``Do you have remorse for your actions?'' This is a standard question. These individuals don't experience this. They seem incapable of experiencing remorse the way the rest of us do for our actions.

When you look at all of these characteristics as a package, as a syndrome or as a cluster, it's not surprising to anybody that these individuals are going to continue to violate the rules and regulations that society has set up. These are not people who believe the rules we impose apply to them. So given that we can identify these individuals, it's not at all surprising that these are the people who are going to go on to continue to violate social conventions and social norms, including the Criminal Code.

The problem is, as Dr. Hart has pointed out, that while we can identify these individuals with a high degree of accuracy, and we can do so at a fairly early age, as early as 13 or 14 - which surprises a lot of people, but not people working in the criminal justice system - the difficulty is what we can do about it once we've identified them.

.1130

So far the research has simply followed these individuals once they've been released into the community. But we don't have in place, anywhere in the world as far as I know, adequate intervention or management strategies that would be designed to reduce the risk of these very high-risk offenders. They tend to be treated like any other offender.

And when I referred to adolescent sex offenders, a point I should have emphasized is that the adolescent psychopathic sex offender is not only a sex offender, he is a generalized offender. He commits all sorts of different crimes. He may have been caught for a sexual offence but in fact he typically is committing all sorts of other offences. And the same thing would apply to the adult population.

[Translation]

Mr. St-Laurent: Mr. Hart, your recipe is somewhat of a correctional service secret. At least, this is how I see it. You can assess the risk of repeat offenses as well, since you said that you use the data provided by Mr. Hare and the members of his team in order to assess, with some degree of accuracy, the likelihood of an inmate committing a repeat offense.

The way that things work right now, an individual must go through many required steps before being paroled. I will not list them all since I am sure that everyone here as well as yourself are familiar with them.

During the process, the parole board members are called upon to make decisions, and these are not always easy to make and which are often wrong because sometimes an individual commits a repeat offense as soon as he is released. Fortunately, these people are a minority, but it is always catastrophic when such an incident occurs.

In your opinion, we have the means enabling us to go a bit farther in analyzing with greater accuracy the likelihood of recidivism.

My question is two-fold. First of all, what degree of accuracy can we obtain if we use your methods? Secondly, could you outline the case of an inmate who goes through all of the required screening. If we were to use your method, how would the accuracy of the assessment be different, bearing in mind all of the standards that must be met, of course?

[English]

Prof. Hart: I will start with your second question first, talking a little bit about the procedures that are currently used and how they could be changed in order to improve the accuracy of risk assessments.

Currently most risk assessment is conducted by Correctional Service of Canada employees. The information is condensed, or collated, and then passed on to National Parole Board members, who then will sit down, look at this material, interview an offender and make a decision.

The primary problem with this is that making a decision on the basis of complex information is very difficult for human beings in general to do, especially when much of that information may actually be irrelevant. Predicting performance on parole is much the same as trying to predict who will be a good employee. And one of the things we know from research on industrial and organizational psychology leads us to the same conclusions as we get in the area of forensic psychology, which is that it's much better for people to have a few pieces of good information rather than a mass of irrelevant information. The more information decision-makers are presented with, the more difficult it becomes for them to make a good decision. So in some cases we say less information leads to better decisions. That's the first thing.

.1135

In fact, much of the research that's been done here in British Columbia, in Ontario, and in Quebec shows that we can predict recidivism accurately, even among people who are deemed fit for parole and have received parole from the National Parole Board. We can predict who will be a recidivist while on parole, even when they've convinced the parole board they're actually a good risk. This is because we're focusing on a small number of very important risk factors. Much of the research that Dr. Hare referred to and presented you with in fact suggests that you can use one risk factor and make very good predictions. That's just thinking of one here.

If we add other risk factors, there's a possibility we may be able to go beyond that, but that one risk factor, psychopathy, is much more important and a much better way of summarizing risk than are our own other variables, such as whether a person has a family member in the community who is willing to support him upon his release.

This may sound contrary to common sense or conventional wisdom, but it is true that certain kinds of personality factors indicative of somebody's long-term functioning are much more important than a single thing in the community, such as the presence of a job, a place to live, and whether or not the person is still abusing substances. Yet a parole board may consider these other kinds of factors equally important, and they're simply not.

To jump back to the first part of your question - and I believe we were talking about characterizing the accuracy of decisions - if my recollection of your question is correct, you're asking about how accurate the decisions we're making are.

Prof. Hare: How we would improve upon them.

Mr. St-Laurent: Exactly.

Prof. Hart: I think Correctional Service Canada is using many of the correct procedures, those that research identifies as being the most important. Indeed, Correctional Service Canada is among the best corrections agencies in the world in this area. Right now, at this very time, Dr. Hare is working with the parole board and corrections services in Australia and New Zealand. We have a visitor here from Sweden. We have been to the United Kingdom and across the United States to do this with various state and federal governments. They are trying to adopt the procedures currently in place in Correctional Service Canada.

The problem is that when the information is passed on to the parole board, the information gets used in a way that is not structured or systematized in any way.

To rephrase this, what happens right now is that Correctional Service Canada collects all the right information, but it may not be communicated to the parole board in a way that is maximally effective and it may not be used by the parole board in a way that is maximally effective. They're getting the right information, but they're not making the best use of it.

The Chair: Thank you, Mr. St-Laurent. Ms Meredith.

Ms Meredith (Surrey - White Rock - South Langley): Thank you, Madam Chair.

Welcome. It's been a long time, and I appreciate your taking the time out of your busy schedules to appear before the committee.

I want to expand a little bit on the psychopathy test. If I'm correct, Professor Hare, you are responsible for the test that corrections is presently using. Didn't they contract with you many years ago to prepare a formula for their risk assessment?

Prof. Hare: Yes and no. Certainly the instrument was developed in my laboratory as far back as 1978. Dr. Hart and two or three other of my graduate students were instrumental in helping to develop this particular instrument. It was designed originally only to make it easier to do research, to publish articles in scientific journals. We were working in laboratories in the ivory tower. In the process, other people around the world began to hear about this particular instrument, because it provided researchers and clinicians with very specific criteria for making this particular diagnosis. It turns out that after a number of years of gathering data, we began to look at the implications for the criminal justice system. We came out of the ivory tower and went into the real world.

.1140

The results were consistent, repeated time and time again in different countries, different jurisdictions. We now know that the instrument is a highly reliable method for identifying these individuals and it's a valid instrument in that it does important things such as predicting outcome. It predicts response to treatment within the prison system. In fact, some jurisdictions will not take into the expensive treatment programs those who are deemed to be psychopaths.

Correctional Service Canada has for a number of years been interested in this particular procedure. As Dr. Hart said, CSC is at the forefront in the world in doing innovative work, research...and in their mission statement, which makes an awful lot of practical and theoretical sense. They've been very active in this area in trying to improve their programs.

One of the things they have done is to adopt this particular instrument, because, as somebody in a review referred to it, it is state of the art for this particular procedure. That is its purpose. They are using it routinely now. I think every inmate who goes into Correctional Service Canada is administered the instrument, and it is used as part of the package.

I should have said at the outset that this is a formally published instrument, so I'm in somewhat of a position of conflict in talking about its merit. Clearly, I get royalties from its sale and I'd like to declare that.

Beyond that, it works. That's the bottom line, I suppose.

Ms Meredith: In conversations about psychopaths it was brought to my attention that psychopaths are not necessarily criminals or dangerous offenders. Would that be an accurate statement?

Prof. Hare: Yes, it's perfectly true. I've been quoted several times as saying that if I weren't able to do research in a prison context, where the base rate for this disorder is high, I would probably go into some other organizations. At the risk of committing liable or slander, I won't say which organizations. But yes, these individuals exist out in the real world. They could be your neighbour, your boss, or a member of your family. These are people who live for themselves; other people don't exist in any important sense, except to provide them with something. Think of them as social predators who have no concern about your rights. They don't all end up in prison. It's quite possible for some of these individuals to have by early training, breeding, luck, and experience the social skills to pull things off, as it were, without actually committing an overt crime. But they're typically going to bend the rules.

Ms Meredith: Then it would be fair to say that in your risk assessment of dangerous offenders, it's not just that you deem them to be psychopaths; there are other factors that bring the risk element to the surface. It's not that we're capturing all psychopaths. You're able to specifically pull out those who are violent in nature.

Prof. Hare: We can certainly identify the individuals we think meet the criteria for psychopathy. This would be a high score in this instrument. There are other risk factors, as Dr. Hart has pointed out, but these other risk factors simply amplify the predictability of this particular instrument.

As I explained before, if you have an individual who is cold, callous, egocentric, ruthless, and predatory, it's not surprising that he or she is going to commit a lot of criminal acts or other nasty things. A study done in the United States by the FBI found that 44% of the killers of law enforcement officers were psychopaths, according to the criteria we list. This is not a theoretical, mythical concept we're taking about. It has real implications for the real world.

.1145

Ms Meredith: Thank you.

Dr. Hart, in a conversation we had probably a couple of years ago, we were talking about the record-keeping of law enforcement and how whenever a case was brought to court it was in some cases minimized. In other words, somebody who did a violent sexual assault could end up in court with an assault charge. You mentioned that whenever you gathered more information during your research, it always seemed to worsen the scenario. Is that a good summation of the conversations we had?

Prof. Hart: Yes, I think so. Right now it's very clear that a criminal record captures the tip of the iceberg in terms of overall criminal activity and that almost always it a very inaccurate reflection, a gross underestimate of the nature, the extent, and the severity of criminal history.

You're talking about cases in which, for example, somebody charged with a sexual assault initially may be convicted of a break and enter with attempt to commit an indictable offence. It would be somebody who at first looked as if he was going to be charged with a violent crime ending up being convicted of a non-violent crime. This is still a concern for us.

There are other concerns about criminal records, one of them being that it is much more common now for offenders to request a pardon and to have their criminal record expunged. That includes some of the worst offenders we've seen over the years. People who may be able to keep themselves straight for a number of years can have their criminal records removed from the system. Now we get requests from the RCMP asking us to please destroy the criminal record we requested for research purposes since the man is being pardoned. This is very dangerous, in my view, when you have the system being so much focused on somebody's criminal history.

A mitigating circumstance is that Correctional Service Canada actively goes beyond criminal records and over the last five or six years has spent much more time trying to find out what offenders have done rather than what they got caught for or what they've been convicted of.

Correctional Service Canada is well aware of this problem and is trying to overcome it, but we still face a major problem, because of inadequate record-keeping, in trying to track our offenders over time and see what they're doing. It sometimes takes up to a year or a year and a half for a conviction to be entered into a man's criminal record, for example. So it looks as though he's safe in the community, but actually he may have already committed another offence and is in prison. We won't know it, however, for another year and a half.

Ms Meredith: In relation to the two bills before us, some of the comments from previous witnesses have been that we are trying in these pieces of legislation to predict violence, to predict somebody's actions in the future, and that just can't be. From your testimony I take it you feel very confident in saying that you can predict with great reliability whether an individual is going to violently re-offend. Would that be fair?

Prof. Hare: Not with great reliability, but certainly well above chance. It's simply not true to say that we cannot predict reasonably well who is going to re-offend generally or violently. This is not an opinion I have. The empirical literature supports this. There is no doubt about this any more. We do make mistakes, obviously.

Prof. Hart: The way I have tried to summarize it in the past is that when we look at the accuracy of predictions of violence, we predict violence about as accurately as we predict that cardiac by-pass surgery will improve somebody's angina pain. I don't know anybody who's ever turned down cardiac by-pass surgery when it was recommended by a physician; yet we predict that with about the same accuracy as we predict violence. For people to say that we cannot predict recidivism or violence is simply inaccurate. It's based upon a profound misunderstanding and misreading of the scientific literature.

.1150

What's clearly true is that we don't predict violence anywhere near perfectly - I think everybody agrees - but we certainly do much better than nothing, much better than chance, and we do about as well at predicting violence as we do at predicting all sorts of important outcomes.

Ms Meredith: Thank you.

The Chair: Mr. Telegdi.

Mr. Telegdi (Waterloo): Thank you, Madam Chair.

I was curious when you said that conditions such as release plans aren't as good predictors for success as your psychopathy test scale. There are certain conditions you have to meet. If you don't meet them, you're not going to succeed in the community. You have to have housing, you have to have some income from some place - hopefully, from a job - and you have to have some support within the community.

Let me try to get you to answer something that maybe I can visualize more easily. You have 100 people coming up for parole. Break down those numbers for me and tell me, of that 100, how many of them you would identify as psychopaths who are prone to violence. How would you break it down? What kind of success would you have with your test? Let's take the 100.

Prof. Hart: If I can summarize, out of the 100 people typically in Correctional Service Canada who receive conditional release, after about five years on the street the recidivism rate will be about 40%; that is, about 40% of those people will be reincarcerated for breaches of the conditions or for new charges. Out of those 100 people, about 20%, or somewhere between 15% and 20%, will be seen as high risk, using a one-variable test such as the psychopathy checklist. Using that one test, you will be able to identify with 75% to 80% accuracy whether somebody was violent or non-violent during that period. That's the simplest way I can put it to you: about 75% to 80% accuracy overall.

Mr. Telegdi: Now, when you say recidivists who have their parole revoked, what percentage of that 40% commit a new offence as opposed to not abiding by some conditions set out in their release conditions?

Prof. Hart: To be sure, the types of questions you're asking depend very much upon what type of a cohort of offenders you're looking at. It depends on what security they're released from, it depends on what their charges are, and it depends on what kinds of measures of recidivism you're looking at. We could try to give you a much more specific answer, but first you would have to give me a more specific question.

Mr. Telegdi: You say that in doing your research you basically look at past cases when making your predictions. In some ways people accuse you of 20-20 hindsight in doing that.

Can you go through the methodology? How did you do it?

Prof. Hare: It is not hindsight. One of the problems of this particular instrument, from the perspective of, say, criminologists and sociologists, is that it does include several items having to do with past criminal history. You could say that all we're doing is predicting what they're going to do by what they've done before, but you can remove these items from the instrument and do a true prediction study.

The first study was done by Dr. Hart and Randy Kropp a few years ago. They had 231 offenders released from Mission Institution, in the valley at Mission, and they followed up. This is a true prediction study.

.1155

The parole board knew nothing about the assessments that were done for psychopathy, and the people who were doing the psychopathy assessments knew nothing about what was going to happen later on. They were followed up for a period - I think on average of four years.

When we first obtained the results, they were very, very good. We submitted them to a journal, one of the top journals in North America, an American psychological association journal.

The reviewer asked whether what we had done improved upon standard prediction instruments, such as number of past crimes and criminal history. To publish the paper we had to show, to demonstrate statistically, that once you take into account all of these other variables, this instrument adds something important to whatever they do.

We have done that. People who do this research now in Canada and in other parts of the world routinely make sure there is no contamination of past criminal history with the prediction made with this particular instrument. Does that make sense?

Prof. Hart: Yes.

Also, to go back to one of your earlier questions, this disorder is identified on the basis of somebody's life history. This is a personality disorder. It refers to how somebody acts in most situations, across time, etc. It's a reflection of somebody's personality.

Other types of variables that people look at may be very much dependent upon chance factors. For example, you mentioned release plans, housing. One of the problems with paying attention to something like housing plans and release plans with offenders - and this is sometimes a shock to people - is offenders lie, and they lie quite a bit.

The other thing about offenders is that they are actually quite unstable in terms of their lifestyle. Some people are predictably unstable. Housing plans can change from one week to the next, yet personality, if what you're assessing is how somebody acts over a period of 15 or 20 years, is liable to be quite characteristic of how they usually get by in the community.

If I can use a weather analogy, here in Vancouver today it's raining. I can try to predict the weather next week. I can do two things. I can predict the weather next week based on the weather today, or I can predict the weather next week based upon the usual climate in Vancouver in the month of February. My prediction based upon climate is always going to be more accurate than my prediction based upon what's going on today. Or course in Vancouver it's easy, because it's always going to rain. I think you understand the point to that question.

Mr. Telegdi: Actually, not necessarily. I'll trade your weather for ours anytime.

Let me get back to the other point. I worked with a population of people who came out of jail and were placed under supervision. Many times we had people who did not have a residence, and sometimes we would go through a number of residences with the individual, knowing that we had to deal with the issue of income or drugs and alcohol, if they were involved.

To the extent that we were able to deal with those issues, the experience of the person was, for the most part, satisfactory. I agree with you that the housing situation can be unstable and, yes, people do lie, particularly when the alternative is incarceration.

To the extent that you can stabilize those critical factors, which I consider to be quite critical - if you've got no place to stay or any money, you're going to steal and you're going to do something else and chances are you're also going to flee the jurisdiction - your chances for success are much, much higher than if you don't.

Prof. Hart: Is this true? We would say that's especially true for people who are not psychopaths. One of the problems with psychopathic offenders is that lying is characteristic of their usual activities. They look on the surface as if they are well adjusted. Yet if you go on the basis of their history, you realize this is misleading.

.1200

What we're worried about are the people who look as if they have a stable residence and a good job. It looks as though everything is going fine. Those are the ones, however, who are actually going to violently recidivate in the community. Those are the people about whom we're most concerned.

Prof. Hare: The issues you raise are extremely important for the majority of offenders. Nobody would deny that. They are less important for the psychopathic offender because this is an individual for whom planned acts mean very little. If you put them into a halfway house, for example, these are the ones who are going to violate the rules and regulations of the house and get the rest of the inmates in trouble, and so on.

We're not saying that we can identify the high-risk offender with absolute reliability. Nobody would ever claim that. On the other hand, we can identify a sub-group that is at reasonably high risk. We can also identify groups that are at low risk. That's an important point. Sometimes it's lost in discussions.

We have a continuum of risk here that goes from very low to very high. It's easy to identify the high ones and the low ones. The ones in between are much more problematic for everybody, including the National Parole Board.

Mr. Telegdi: Thank you.

The Chair: Thanks, Mr. Telegdi. We'll go back to Mr. St. Laurent.

[Translation]

Mr. St-Laurent: Earlier you talked about predictions. Could you tell us when the next election will take place? This would help our side.

On a more serious note, let's talk about statistics. You described your role, Mr. Hare. Earlier, you told Ms Meredith that you were, to some extent, in a situation of conflict of interest. I didn't understand this. How are you in a conflict of interest?

On the one hand, I would imagine that you advise the Correctional Service about certain data, particularly data pertaining to the way that parole is reviewed. You said that you can predict 75% of all repeat offenders, whereas Mr. Hart said earlier that, right now, 40% of all parolees commit repeat offenses. How can you, as a witness, as someone who is advising the government, say that we can reduce recidivism? Is it because your method doesn't work? Explain this to me. I have misunderstood.

[English]

Prof. Hare: We all strive for perfection, but few of us ever actually reach it.

The procedures we developed were designed primarily for research purposes. They have been adopted by various criminal justice jurisdictions. They've been adopted simply because they do a reasonably good job.

I don't consult or provide advice to the criminal justice system. I do research within the system.

I can tell you that one of our current projects is to look at people we have studied since 1978 for whom we have follow-up information on their subsequent criminal histories. We are trying to develop a battery of risk factors that will predict outcome - that's recidivism and violence - among various types of offenders. That's in progress. I'm doing that with the Correctional Service of Canada and with some consultation with the National Parole Board, because they are interested in this sort of information.

Remember that this is all empirical. We try it, see if it works, and if it doesn't work, then we will change it. So far, things work very well, but they're far from perfect.

You ask whether we can reduce that 40% recidivism rate. Marginally, I suppose, we could have some impact upon it. I think what we could do with these procedures and with common sense is to reduce the risk of violent recidivism, which is the really serious reoffending that does occur. We may not have much impact upon less serious reoffending.

As a matter of fact, say we let low-risk offenders out. What do we mean by low risk? Is that a low risk for violence? If it's a low risk for violence, then these individuals might be at a fairly high risk for breaking and entering, theft and other crimes that don't necessarily involve a victim.

.1205

So we have an interesting balance here that we have to try to handle. We'll let a lot of people out who are going to commit crimes that are very bothersome to a lot of people, but they may not include violence.

I brought up the thing about the conflict because I've given quite a few addresses to professional organizations. One of the addresses I gave was to a large psychiatric organization a few years ago. Every speaker was required to sign a sheet indicating whether or not any of the things they talked about was of personal interest to the person, such as whether they had a conflict or a vested interest in a particular instrument.

I signed this. When I got up to give my talk, I said that I was the developer of the psychopathy checklist. I said it's published commercially in Canada and I receive royalties from it. Therefore, I said, you can pay attention to what I say or you can dismiss it, because I could be seen as somebody who's commercially oriented.

Having said that to the audience, they all thought it was rather humorous, because every other speaker there had developed instruments that were being used by people, and they didn't declare their interest.

So I guess in a sense there is a conflict position. I would like to put that right up front. If this instrument is used.... It is being used. I don't promote it, by the way - it has been adopted by all sorts of people because it is useful - but I do receive royalties from its sale. It's perfectly true.

The Chair: Thank you, Mr. St-Laurent. Mr. Discepola, five minutes.

Mr. Discepola (Vaudreuil): Thank you, Madam Chair.

Dr. Hare or Dr. Hart, I have two areas of questioning. You seemed to indicate to this committee that it's quite a scientific feat or process whereby you can actually predict behaviour. You can actually predict, as you said, if someone is going to recidivate or not.

Here is one of my concerns. You also mentioned that the National Parole Board seems to be doing a half-decent job in certain areas in which they're capable. But we hear far too often of the exceptional cases, the 20% of cases, in which they're not so successful.

Is there a weakness in the system? You mentioned the lack of interpreting the information. Your statistics and your research data collected over a period of an inmate's life while in incarceration seems to be a very good predictor of future behaviour. However, you indicated that the National Parole Board, first, may not be qualified to interpret that, and second, may not have the proper training or qualifications.

So I'm just wondering whether we can release fewer criminals into the community by enhancing either the qualifications for the National Parole Board or by introducing some training for them that would be tied in, obviously, with your research, which you seem to indicate works well.

Prof. Hart: I think the parole board people are not selected to be scientists; they're selected to represent the community. So in some respects we're choosing people deliberately for their lack of expertise. That's a fact of life.

I suppose our job in working in the criminal justice system is to try to provide those people who are representing the community with information they can use and digest in an easy way. I think right now we're providing the parole board with too much information and not enough simple information, if that clarifies it.

I don't expect the National Parole Board to become an expert in this area. If it did, then we would have a redundant organization. We might as well just let the Correctional Service of Canada make the decisions itself.

So if we want to keep the National Parole Board to try to reflect community values and community opinions, then I think that it's incumbent upon the Correctional Service of Canada and the National Parole Board to develop guidelines for communicating information about the risk for violence and other kinds of criminal risks.

Mr. Discepola: Here's my other question. It's fine and dandy to do all this research. It seems to be quite scientific with your predictions. However, you used the words ``marginal'' or ``marginally improved'' or ``reduce the recidivism rate''. What's the value of all this research if we can't put it into practice in order to actually reduce recidivism?

.1210

If we can predict it, then is it our communities that are failing us? Is it the system that's failing us? Is it the support structures that all of a sudden stop the minute the inmate is out into the community?

Very often during incarceration they're receiving all kinds of treatment programs for their behaviours. I guess we've been criticized, the Correctional Service, that quite often the minute they're no longer under our care, they're left to fend for themselves. That's when they fall. They don't have the support structures to essentially - I guess this is the best word I could use - wean them off slowly.

Should we be doing more in the communities? One of my colleagues alluded before to social housing, for example. Should we be doing more or can we be doing more in order to prevent and improve that 40% rate to maybe 30% or 20%?

Prof. Hart: I will make one quick comment before Dr. Hare speaks.

I think first of all that you should not take the recidivism rate simply as a measure of success. Some people hope to see recidivism rates go down. I am sometimes happy when they go up, not because it means that people are more criminal, but that we're better at catching them. I think this is a critical thing.

We could all have a very low recidivism rate if we just asked the police to take some time off. It would drive the recidivism rate down a lot, but the same amount of crime would still be committed. So there are occasions when an increased recidivism rate is good because it means you're catching people before they do serious things.

That's one of the things we'd be looking at here. The overall recidivism rate may not change, but hopefully we're catching serious offenders before they commit violent offences and not worrying so much about catching low-risk offenders committing technical breaches of conditions, like getting drunk when they're not supposed to. We want to catch people who are trying to hang around school yards to molest kids.

Does that make sense to you? It's not simply the recidivism rate that's a measure of success; you have to interpret that in a quite sophisticated manner.

Now I'll turn it over.

Prof. Hare: As I understand Bill C-55, the concern really is with sexual reoffending and violent reoffending. I would argue that it would be possible - it certainly is possible - to reduce the recidivism rate for those types of crimes with procedures of this sort.

Now whether we can reduce the rate for other less serious crimes is another matter, because we're not targeting these particular individuals.

You raised an interesting point about what happens in the Correctional Service of Canada. Under this long-term offender proposal, it's possible that people are going to be in for a very long period of time, but eventually they will get out unless this other provision is invoked saying that they'll be declared dangerous offenders at the end.

The problem with all of this is that we don't have procedures in place in the system to actually treat - maybe that's not a good word - manage intervention procedures for the type of individual we're talking about.

The Correctional Service of Canada is probably a world leader in standard treatment programs for the average offender. These programs work reasonably well, as you've probably heard from a number of speakers in the past.

They don't work very well with the type of individual we're describing. These individuals quite often will put on a very good show in group and individual therapy. They will impress everybody with their new-found insight. They say they understand what impact their behaviour has upon victims. Quite often this revelation is only word-deep for these individuals, but they will end up getting a very good review. Someone will write a report and say that this person has turned his life around; he's not the person he was before; he's at a low risk to reoffend. But in fact he's conned the system. The program that was designed to develop a capacity for empathy or to build up his conscience is not going to work for these people.

The Correctional Service of Canada understands this problem. I gather that plans are afoot to try to develop programs specifically targeted to these individuals. You cannot keep them in forever without actually trying to develop innovative programs that are likely to reduce their risk for violence both within the institution and outside it.

Mr. Discepola: Have you done any studies to measure the success rate of these programs?

Prof. Hare: Some people have, yes. The Correctional Service of Canada has a couple of studies run by Dr. Stephen Wong at the Regional Psychiatric Centre in Saskatoon in which they've looked at the effectiveness of these programs for the average offender and the psychopathic offender. They're not very effective for the latter, not only in the view of the people who run the program, but in terms of what happens once they're released into the community. These people have all gone through a very expensive, therapeutic-based treatment program. When followed up, it turns out that the recidivism rate of the psychopaths who went through that program is far, far higher than it is for the other offenders.

.1215

Similar research is being done in Ontario at Penetanguishene that indicates that in fact the traditional treatment programs are not only ineffective for psychopaths, but they might be counterproductive. They learn new ways of finding buttons to push on other people. When they come out, they're worse than they were before.

To finish up that point, several years ago, I was commissioned by the Correctional Service of Canada to put together a program for offenders at a high risk for violence that would include psychopaths. I did that. The director of research and statistics then was Dr. Frank Porporino. I think the commissioner was Ole Ingstrup at the time.

That program was submitted to the government several years ago, but nothing has been done with that for reasons of which I'm not aware. There was a program designed to reduce the risk that these particular people posed to the community, but it was never implemented.

The Chair: Thanks, Mr. Discepola.

I have two other people. I'm going to give Ms Meredith the last word because she found these witnesses for us.

Mr. Regan, you had a brief question.

Mr. Regan (Halifax West): Dr. Hart, you mentioned that the criminal record was often just generally the tip of the iceberg of a person's criminal activity. While I would, without any empirical evidence, tend to accept that concept, I'm wondering what empirical evidence you would have on that. It would seem to me that normally the evidence you have to prove a person's criminal past conduct is in fact their convictions. What can you tell us about the basis of your conclusions then?

Prof. Hare: You could actually ask offenders. Although we say that offenders lie, the assumption is that, in most cases, they lie to actually underestimate their criminal activity.

In a variety of studies in which people have given offenders the chance to anonymously report crimes, report them in an interview setting, or check the veracity of their reporting with a polygraph or lie detector, in all these cases you find that criminal convictions generally account for about one-tenth of the total criminal activity of those individuals.

It's horrifying, actually. It means that we're catching people for about one out of every ten times. In some cases, for sex offenders, we suggest that's about one out of every 75 offences they commit. But in general, a conviction is a good indication that somebody's criminality is serious. They're committing enough serious crimes to get caught. We only use official convictions as a gross index of the level of criminality, not an indicator of the absolute level of criminality.

In the original parole study that Dr. Hart was involved in, they did go to talk to the parole board supervisors to find out the extent to which the parolees actually adhered to the release plan. Did they stay where they were supposed to stay? Did they do what they were supposed to do?

The evidence was very strong that the psychopathic offenders in fact did not follow the pre-release plans. They did not follow the rules and regulations of the program. They were violating them all the time.

The Chair: Ms Meredith, you have the last word.

Ms Meredith: Thank you, Madam chair. I just want to follow up on what Mr. Regan brought up.

You also, as I understand it, went into police reports. You found that where it indicated what actually did happen, when it reached the court system, it was downplayed in many cases, whether it was through plea bargaining or just the charges they felt they could successfully get through the court system. The actual event was very often downplayed by the time it came up as a charge or a conviction. Is that a right assessment?

Prof. Hart: Yes, that's correct. We actually pay very little attention to the Criminal Code section through which somebody has been charged or convicted.

.1220

We spend most of our research actually reading through police reports to try to see what kinds of violence an offender has committed. When you do that, you find that not only do psychopaths commit significantly more violence overall, but they actually commit very different kinds of violent offences as well. They are, in words we've come to use to describe them, impulsively instrumental. That means that psychopaths tend to make decisions quite quickly to commit violent crimes that are instrumental for economic gain or other reasons. It's not simply a result of strong emotional arousal or things like that.

If we look at the actual behaviour of psychopaths as described by the offenders themselves, victims of the offenders, witnesses and police, you find they are actually much more violent and violent in a different way.

Prof. Hare: Their violence is predatory, planned, premeditated, dispassionate and cold-blooded when compared with the intense emotional arousal that often leads to a violent act for the rest of the offenders.

Prof. Hart: You will never find this type of information in something that says ``assault'' on somebody's PF sheet; it just says ``assault''. It doesn't tell you who the victim was, what the circumstances were that led up to it, or the nature of the violence that was committed.

Ms Meredith: Thank you.

I want to end with just a question on young offenders. You're saying that the results are much the same when you run these tests by them. If you get them at a younger age, like 13 or 14 years old, do you feel there could be a treatment for the psychopathic violent offenders? Do you feel there is some chance of any success with early intervention at a younger age?

Prof. Hare: I would have to say yes. That should be so.

The people who have looked at the development of anti-social behaviour in general terms throughout the world will all point out that by the time an individual is 14 or 15 and engaged in a life of crime, it's very difficult to change that pathway in general. Of course it must be much more difficult to change the path of an adolescent psychopath. But unless we do that, we're doomed to spend the rest of our lives dealing with adult offenders who have been through a system for many years who have learned all sorts of other ways of expressing themselves that are not very good for society.

I think that unless we try to develop innovative programs targeted to these individuals, then the problem is always going to remain with us. The only way to start is with young people. I think eventually we're certainly going to have to consider moving farther back in the age span.

Prof. Hart: I think it's also very important to point out that if we're talking about identifying 15% to 20% of offenders who are psychopathic, at a high risk for violence, and unresponsive to the usual correctional treatment programs, then this means that the flip side of the coin is that we can also identify very accurately about 80% of offenders who in fact respond quite well to standard correctional programs, who are at a low risk, and on whom we are in some ways squandering resources.

We are devoting too much in the way of control and social sanctions to those offenders and not enough to the people who need it the most. We don't necessarily have to spend a whole lot more money on the criminal justice system, but we can certainly reorganize the services to deliver the most to the people who need it the most.

Ms Meredith: Thank you very much. Thank you for attending.

The Chair: Thank you, Ms Meredith.

Gentlemen, thank you very much for your assistance. It was very helpful to have you here. We enjoyed speaking with you.

I'll just remind you that our next meeting is at 9 a.m. in this room on Tuesday, February 18.

The meeting is adjourned.

Return to Committee Home Page

;