We continue our alternatives to custody inquiry. I welcome Professor Gill McIvor, who is from the social work research centre at the University of Stirling; Fergus McNeill, who is a lecturer in social work at the University of Glasgow; and Bill Whyte, who is the director of the criminal justice social work development centre for Scotland at the University of Edinburgh. I understand that Professor McIvor will be the speaker.
I will informally chair the panel.
It was tactless of me to call you the speaker. The phrase "informally chair" is more polite.
I said in my submission that one difficulty in a small country is often that relatively few offenders are made subject to some types of disposal. That has implications for gathering outcome evidence in the longer term. However, in the past 12 to 15 years, most of the community-based disposals in Scotland have been subject to some evaluation.
Has the research been collated in such a way that the committee could get its hands on it?
Some of those data have been collated in the Scottish Consortium on Crime and Criminal Justice's report. However, a single volume that pulls the material together does not exist. Most of the research was commissioned by the Executive and has been published as research reports in the Executive's social research series.
Is that in the public domain?
Yes.
Are the consortium's findings also in the public domain?
Its report is available on its website. The report is also published in hard copy, as are the Executive research reports.
All the summaries are available in the central research unit series; the full documents are fairly lengthy. The challenge is, as Gill McIvor said, what is meant by evaluation. Everything has been subject to some kind of evaluation, but inevitably there is a debate about how well something works in comparison with something else. We have not had a tradition in Scotland of setting up very good control groups or match groups to compare one initiative with another. Such data remain somewhat limited.
Are you saying that evaluation should be a routine activity?
Yes.
You have said that it depends what we mean by evaluation, but it also depends what we mean by effectiveness. Do we measure that in the short or long term? Is a standard test not applied for different alternatives that are being used?
There will usually be a focus on reconviction rates as the ultimate aim is to reduce reoffending. What the research examines additionally will depend on what the initiative is intended to achieve. For example, the research may focus on changes in the offender's attitudes or circumstances. In some cases the outcomes that are examined will be general—such as recidivism—but in other cases they will be specific to the initiative and its objectives.
Can you give me an example of two distinct programmes? It would be difficult to say that you compare apples with apples.
Community service and probation might be a good example. The explicit objective of the supervision in probation would be to bring about changes in the offender's attitude and behaviour and probably in their social circumstances so that offending would become less likely. Community service, on the other hand, has been set up not so explicitly as a rehabilitative type of disposal; it has been set up largely as a fine on the offender's time. We would not necessarily expect to find deliberate changes in the offender's behaviour to be the focus of the supervision of the offender when they are performing community service. We may find that there are additional benefits, but the purpose is different.
You have said that to make a proper comparison it is necessary to set up a control group. How easy would that be to do? Would it be affected by the sheriff's discretion and the kind of sentences that he gives? How can you focus on a certain court and say that you will consider 15 cases of one type and 15 cases with a similar background but a different sentence? I would not have thought that it was possible to do that.
Difficult ethical questions arise when we talk about random allocations to different types of justice. That is a huge barrier to undertaking that type of controlled experiment. People would rightly be concerned if they felt that justice was being administered randomly. We can get round the issue in other ways. For example, if there is an excess demand for places over the supply of places on a new initiative, we can compare the outcomes for people who were not able to take part in the initiative because there were no places available but who otherwise would have been appropriate for that type of disposal.
A number of courts do not have access to all the disposals that are available, but most disposals have been rolled out over time, so it would be possible for some comparisons to be made between jurisdictions and between matching offender samples. I do not think that random allocation would be ethically acceptable.
The other possibility is to compare predicted rates of reconviction with actual rates of reconviction. That would be done by determining the expected rate of reconviction for a sample of a particular type of offender with a particular type of history and of a particular age, but that would rely on having information on how the wider population of offenders progress through the justice system. We are still some way behind England and Wales in that respect.
The situation is beginning to improve. We approached the justice department's statistics unit with a sample of cases of people who attended the Airborne Initiative, and we asked the unit to generate a similar sample of cases where alternative disposals were received, using a number of different variables to match the two samples. By using that approach, as Fergus McNeill mentioned, we were able to compare the reconviction rates of people who went to the Airborne Initiative with those of people who received prison sentences or other community disposals.
If it is not part of an on-going academic paper, it would be useful if that information could be provided to the committee in written form.
Restriction of liberty orders are being piloted. I do not have—
I do not expect you to have the information to hand. To some extent, we are prodding around and trying to find out which data are available. Perhaps Audit Scotland will have some for us. If you have information from your own research, it would be helpful to the committee if you could provide it.
It strikes me that there are two or three points. The first concerns the data that we already have. I talked about routine data. One of the issues that concern me is that academic institutions are often commissioned to do projects, but when the data stream finishes, whether that is 12 months on or 24 months on, nobody in the criminal justice service carries on—I am not trying to do us in the academic world out of a job. A practical illustration is that it is difficult for services to get routine access to Scottish Criminal Record Office data. Indeed, we know from some of the research that we have done that data are not even configured in a way that would allow agencies to follow up their clients and determine how they are doing over, say, two years or five years. Other jurisdictions have the required data—that is certainly the case in England, Australia and parts of New Zealand. It is a huge challenge. I gather that a project is under way in Scotland to get the data systems together, but that is a real weakness in our system.
That is a good issue to raise with the Minister for Justice when he comes before us.
In the written evidence that the committee received from Professor McIvor, it was suggested:
There are two types of evaluative information that would be of value. The first is the independent evaluation of new initiatives, which aims to establish whether they are doing what they purport to do. As Bill Whyte suggested, the other type is the on-going collection of data, by which services can routinely establish how well they are performing. They can use that information as an educative tool for the public and for sentencers. To some extent, sentencers are in the dark about the effectiveness of community-based disposals. Sentencers might be more positively disposed towards such disposals if they had available, in an accessible format, some of the data that academics know are available.
To put the matter in an historical context, since the advent of national objectives and standards, services have primarily been required to provide monitoring data that relate to meeting those standards, which, in effect, are about efficiency. The standards relate to whether a service sees somebody enough times within the prescribed time scale, whether work is done on time and whether reports are submitted timeously. To date, the focus on whether the objectives on reducing reconviction or the use of custody have been met has been less comprehensive than the focus on monitoring and evaluating services.
Professor McIvor argues at the end of her report:
I think that it would. We can also draw on evidence from other jurisdictions for areas on which no Scottish data are available. Services will become more effective if they are developed under the rubric of getting best results. We can begin to draw on the large amount of research evidence from other jurisdictions to show us which approaches are more likely to be effective. We must implement ideas from other jurisdictions on the ground in Scotland and ensure that they are evaluated and monitored properly to find out whether they are fit for purpose in the Scottish context.
Many offenders who are eligible to receive a non-custodial sentence might have a large number of problems that contribute to their offending behaviour, such as drug misuse, psychological distress or homelessness. How effective are the available alternatives to custody in dealing with offenders who have multiple problems?
An increasingly wide range of community-based disposals that can be tailored to particular offenders' needs is becoming available. For example, drug treatment and testing orders provide a real alternative for dealing in the community with offenders whose offending is linked primarily to their use of drugs. Another example is probation orders with drug treatment requirements. Probation orders are generally used with offenders who have a number of personal and social problems, whereas drug treatment and testing orders are more directly focused on drug use and offending. Probation provides a vehicle to enable offenders to access a range of services in the community, such as housing or employment services.
Can I glean from your reply that, as a rough-and-ready guide, probation is probably the best option when an offender has multiple problems?
The issue of providing a range of services is one reason why the structure in Scotland is designed in the way that it is. Local authority social workers are involved because, as well as focusing directly on the offence and the victim, they can link up and broker services from a wide range of services. That is one advantage of the structure in Scotland.
The recent Audit Scotland report entitled "Dealing with offending by young people" said that there is anecdotal evidence that some local authorities try to avoid putting people on probation because the costs fall to the local authority. What is your view on that?
There are disadvantages. If you are talking about youth justice, it might not necessarily be probation. You are referring to the example of secure accommodation, for which the local authority has to pay. If those young people were drawn into the criminal system, the state would have to pay for it. Criminalisation therefore becomes an incentive for the local authority.
Do you believe that there is some merit in that proposal?
With the kind of expenditure that secure accommodation demands, the current funding arrangements raise practical problems.
I want to go back a question or two and talk about evaluation. You seem to think that until there is proper evaluation, we will not be able to persuade sentencers to use alternative disposals. It is a chicken-and-egg situation. If we are not going to get the sheriffs to use alternative disposals, how are we going to get enough disposals to evaluate? Could we consider the information that is given to sentencers rather than waiting for evaluation?
That is probably absolutely right. There is a need for evaluation but the pilot schemes for restriction of liberty orders have already been evaluated so data exist that can be used to persuade sentencers that restriction of liberty orders are a useful addition to the range of community-based disposals.
I agree that that is important. However, there is a strand in the argument that assumes that when community disposals are good enough, judges will use them and custody will fall. We should consider examples from other jurisdictions such as Finland. However, I am afraid that the Executive has decided that custody will fall and community disposals will then get used. That is a challenge for the committee.
We have evidence that the relative share of all sentences for community disposals between 1990 and 2000 doubled. The same happened to rates of custody.
Your written submission says that.
We are stuck with a conundrum that says that we can work tirelessly to make such disposals more effective, to publicise them, to enhance their credibility and to resource them, but sentencers need to be permitted, encouraged or required to think more carefully about community disposals. I should not say "more carefully" because I think that sentencers are extremely diligent in attending to the range of options that they have. However, they need to go beyond the credibility of community penalties. There are bigger questions about the purpose of sentencing.
That is helpful. Instead of simply considering data, I wish we had time to ask you more questions about this very deep area. Maureen Macmillan will ask you about public perception, which is an issue that will also be in the minds of sentencers. After all, there are several audiences out there for whom justice must be done and must be seen to be done.
Absolutely.
We were interested in the suggestion in Bill Whyte's submission that
The quotation from Lord Bingham that I have included in my submission captures fairly well part of the problem that judges face. People feel that if offenders are not taken into custody, they have not been dealt with properly. We have to challenge the major issue of the lack of confidence in community disposals and highlight the fact that those meaningful and purposeful disposals contain sanctions and punishment and have effective outcomes.
Are there any data to show that community disposals turn round young people's offending behaviour within three years?
We have data that show that reoffending rates are highest among that particular group of young people, with 18 to 19 as the peak age for offending among young people. Usually, the majority of young people have stopped offending by the time that they have reached their early to mid-20s.
If a community disposal does not work, there is a likelihood that the young person will be given a custodial sentence without another community disposal being tried.
There is nothing in the legislation to prevent the courts from imposing another community disposal, and the person's circumstances might have changed in such a way that it would be more appropriate to use a different disposal. There is nothing to prevent the court from imposing another probation order and perhaps adding conditions to it that would focus the order more directly on some of the issues that seem to cause the young person to continue to offend.
As I say in my paper, national objectives and standards make the distinction between standard probation, probation with conditions and intensive supervision. However, it would be very hard to discover in Scotland the difference between a standard order and an intensive order, and the graduated nature of orders. The research suggests that intensive probation should be reserved for those who are likely otherwise to be in custody and that it might take up as much as 70 per cent of their week, whereas the standard order might involve seeing somebody for a few hours a week.
When social workers write a social assessment report, they examine a set of antecedents and previous convictions along with the complaint. If they see from the list of previous disposals that probation and community service have already appeared, even if they have appeared only once, they know that they are on a sticky wicket and that they will have a difficult job persuading the sentencer that such disposals can work next time. They go back to ask the person about the circumstances of their supervision—why it was not more effective, what is different now and what they can do differently to assist the person. They then have to present the different nature of the new community disposal. The disposal might have the same label, but it has to have different content before the social worker can reasonably expect to persuade the sentencer.
I do not think that I have time to ask you about the interesting information about Finland in your paper. If I do not have time to ask about that today, I might ask you to write to us to explain why certain things are happening. I have noted the information that it would be useful for us to know about.
You have mentioned the patchy nature of the availability throughout the country of various alternatives to custody and the problems that can occur as a result of the attitude of the sentencers. Would you say that there are sheriffdoms in which there are an adequate number of alternatives to custody but that they are not being used effectively, or is there a general lack of alternatives to custody across the board?
It would be difficult to identify such areas, as the picture is much more complex that you suggest. One of the reasons that distribution is patchy, as Bill Whyte said, is that in recent years there has been a substantial increase in the number of community-based disposals that are available, such as restriction of liberty orders, drug treatment and testing orders and supervised attendance orders.
I asked whether there should be a baseline. Should all sheriffdoms have a certain range of alternatives to custody available to them? I understand and am aware of the fact that a patchy approach is taken because of the Executive's piloting process, which I believe to be appropriate. The issue that must be addressed, however, is how the programme should be rolled out. If I am a sheriff who travels to various parts of the country and I turn up in a sheriffdom in which alternatives to custody are not available, I am left without much option other than a custodial sentence.
The simple answer is yes. That is why, as a nation, we went for national objectives and standards. It is clear that differentials exist between sheriffdoms in rural areas—for example, on the islands—but standards ensure that everything is exactly the same. The member used the word "minimum", and I agree that there should be a minimum standard.
We would be on delicate ground if the committee tried to tell a sheriff that, but I see where you are going.
In your submission, you mention that
Of the reduction in offending?
Of a combination of approaches.
A number of good projects have been evaluated, but I cannot tell you about them off the top of my head. We are beginning to see that the methodologies that we have discussed and which have been drawn from wider research in other jurisdictions are being translated meaningfully into practice in Scotland.
It is worth emphasising that much of what works has emerged or has been popularised in the latter half of the 1990s and subsequently, whereas the services that have been mentioned began their redevelopment after the national standards at the start of the 1990s.
I do not want to get too technical, but, on the example of aspirin, the committee might be interested to know that the effect size—the outcome measure—for aspirin is half as good as the average outcome for community disposals, yet people have followed that without any doubt. The evidence is much stronger than we think it is.
I did not quite follow that.
I think that I did.
Is community disposal a pill?
There are no pills involved.
There are specific examples of what Bill Whyte has described as multimodal approaches. A multimodal approach simply involves focusing on a range of issues, rather than just one issue, in relation to why young people offend and providing a range of inputs to try to change the circumstances or the behaviour.
You say that such services are not cheap, but do you have any evaluations of the costs relative to, say, keeping someone in prison for a year or several years? It might be better to spend money earlier to save money later.
The Scottish Executive's recent research studies into the effectiveness of various disposals have usually contained a costing of the initiative along with an estimate of the comparable cost of an alternative sentence. Some costs relating to probation, community service and imprisonment are published annually by the Executive in its statistical series and we also have individual costings for the area initiatives that I have described and for particular disposals, such as drug treatment and testing orders and supervised attendance orders. Cost data are available for the majority of the initiatives that have been introduced and evaluated in recent years.
Are such data available even for initiatives that involve multiple agencies? In addition to having costs in one direction, such initiatives might involve medical costs, for example. Is comprehensive costing available to enable us to see whether the costs are comparable to the cost of custody?
The costs involved would generally be the immediate costs of providing the service. In most instances, the alternative disposals are significantly cheaper than the prison sentence that they probably replaced.
Would it be fair to say that although projects such as Matrix or Cluaran, which work with younger people and family groups at an early stage, may initially be more costly than the provision of secure accommodation, if such younger people are put on a probation order once they get older, such programmes will work out significantly cheaper than custody? Is it more costly to intervene and work with the family at an earlier stage?
The intervention may be more costly in the short term, but if we are able to turn families around, it may be less costly in the long term because of its impact on the risk that, further down the line, the children may need to go into secure accommodation or may end up in custody. The invest-to-save initiative recognises that an initiative that is trying to have a preventive impact may have resource implications at the start but will hopefully bring about longer-term benefits, such as a reduction in offending and other problem behaviours, several years down the line.
This is not a terribly scientific measure, but the research on the costs of youth crime that was conducted in, I think, 1998—I cannot remember the exact date—came up with a figure of around £2,100 per crime. The accountants that produced the costing—I cannot remember which company it was—thought that £1,700 of that cost would be recoverable if the crime could be prevented. That figure of £1,700 per crime gives us an idea of what could be saved through investment in crime prevention activity.
The working definition is five episodes in a six-month period, but that is an arbitrary measure.
So, if somebody is committing 10 offences per year, they are costing £17,000 of recoverable costs. I am not aware of any intervention that would cost £17,000 to deliver.
There is a formula linking the shift to alternatives to custody with a reduction in offending, which is based on actuarial work that was carried out for the National Probation Service for England and Wales. We can probably learn from that. The estimate was that reducing each offender's level of offending by 5 per cent would save the economy—never mind the direct cost to victims—£700 million.
Is that figure for the whole of the UK?
No, I understand that the figure refers only to England and Wales.
Given that Scotland has 10 per cent of the UK population, we are perhaps looking at a saving of £70 million.
I do not know how the figure was worked out, but it would be possible to find out.
It would be extremely useful for the committee to get the various costings for the alternatives to custody to see how those compare with custodial costs.
Perhaps we have time for a quick visit to Finland. Bill Whyte's submission states:
Despite its radical changes in custodial policy and custodial outcomes, Finland is not much further on than us in asking questions about the effectiveness of community disposals. To some extent, the science and art of community disposals that we have discussed today is at much the same stage in Finland as it is here. Finland is a small country that did not do its own research on the effectiveness of community disposals.
Is there an issue about a lot of potential criminals or troublemakers wandering about the community, even if they are doing community service? People on the other side of the argument from us may well dwell on that.
Inevitably, there must be a level of community tolerance and acceptance of policy in Finland. That is a sociological issue, and I cannot give a simple answer on why people accept or tolerate the policy.
It is worth commenting on the notion of dangerous offenders at large. I do not know much about the Finnish experience, but most of the time when we consider the effectiveness of community disposals, we are comparing them with very short prison sentences. The vast majority of relevant cases are sheriff summary cases with a maximum penalty of six months, even for second and subsequent convictions. People spend three months in jail. The incapacitating effect, or the question of public protection, that arises from such custodial sentences is marginal.
We hope that the thrust of our evidence encourages the focus to be on the short sentence. A large number of longer sentences are being given out in Scotland. Sentences are longer than they have ever been, having almost doubled in 10 years. However, we must deal with the issue of the short-sentence prisoner. The prison does not have the chance to work with such prisoners, and the contaminating effect prompts the question whether short sentences are a wise policy.
We have been considering prisoners for three and a half years, so we are aware of the figures.
I should like to touch on the point with which Fergus McNeill concluded. I want to pursue the role of social workers and social inquiry reports.
As Bill Whyte said, we do not have studies that give us information on the amount of contact time. That is a major issue for main grade workers when demands for court reports mean that contact time with people on probation or other orders suffers.
It would be helpful if you could write to us with anything that you want to add.
I was coming to that. There are many things that we would still like to ask, but the pressures of time prevent us from doing so. I suggest that members submit to the clerks any supplementary questions on matters that they wish to pursue—I have some such questions myself. If we may, we will put our questions in the form of a letter to the three witnesses, who can then provide us with written answers that we will put in the public domain along with our letter.
Previous
Convener's Report