The Scottish Government

« Previous | Contents | Next »

Listen

CHAPTER SIX ECONOMIC EVALUATION OF ELECTRONIC MONITORING AS A CONDITION OF BAIL

INTRODUCTION

6.1 Electronic monitoring has the potential to add to the range of bail and community-based sentencing options available, but its effectiveness and costs need to be compared with those of alternatives. The focus of this chapter is on the costs of EM bail.

6.2 The primary aim of this part of the evaluation was to compare electronic monitoring as a condition of bail with remand in custody in terms of costs and impacts. The measurement of the latter is not straightforward in this field. Ideally, the outcomes of different interventions or conditions would be expressed in terms of offending behaviour, public perceptions of safety and so on. It was not possible to employ such an approach here because of time and funding constraints. Instead we have employed more of an 'administrative' approach, couching effectiveness in terms of the processing of remand following the approach employed earlier in chapters. We conducted a cost analysis which examined the costs of achieving a given level of outcome, making comparisons between alternatives.

METHODS

Economic decision model

6.3 The approach taken was to build a decision model, the structure of which was finalised following extensive discussions with the Research Advisory Group. The analysis is based on an approach that maps the paths taken by individuals through the criminal justice system during their remand periods, the probabilities of subsequent events occurring and the economic consequences of alternatives. The main benefit of using a decision model is that it enables complex processes and activities to be broken down into component parts, each of which can then be further broken down and analysed in detail before being recombined in a logical, quantitative and transparent way to estimate - in this case - the economic consequences of different courses of action and different experiences. The ultimate purpose is to identify which of the options provides better value for money.

6.4 The decision model shows the paths that individuals charged with summary offences take through the criminal justice system in Scotland after the court rejects an application for standard bail. This model does not include solemn cases such as rape or murder charges. It also does not include appeals against refusal of standard bail, because the focus of the analysis is on the comparison between EM bail and custodial remand.

Model structure

6.5 The following provides a brief explanation of the structure of this standard decision model which was used to identify costs and probability parameters. Each triangle in the illustrated model is referred to as an 'end node' and indicates the point at which the flow of events ends in the analysis. Each circle is a 'chance node' from which emanate further options/paths that the accused can take with specific (estimated) probabilities of those events occurring.

EM bail standard model

6.6 For didactic purposes, the decision model is shown in two parts, although the process is a continuous one. The first part (illustrated in Figure 6.1) starts from the point at which a court rejects standard bail. At this juncture the individual can either make a request for EM bail or not make a request. If no request for EM bail is made following refusal of standard bail, the individual will be remanded in custody until trial, unless an appeal is lodged. If an application for EM bail is made, the court will use information provided by the suitability report when considering the application for EM bail.

6.7 If the application for EM bail is not considered, the individual may decide to appeal to the High Court, in which case they will be remanded in custody until the appeal is heard. If the appeal to the High Court is successful, then EM bail will be granted. If the appeal is unsuccessful, EM bail will obviously be refused and the individual will be remanded in custody until trial for the primary offence. A similar sequence of events will occur should the court be minded to consider the application for EM bail after the request for EM bail is made. In such cases, the individual will then be remanded in custody for what was found in this study to be an average of 5.7 working days until the suitability report is produced. An additional diet is needed in this case to consider the suitability report. There is the possibility that the EM bail application may not be successful, in which case the individual will tend to be remanded in custody until trial. If, on the other hand, the EM bail application is successful, and if the individual complies with the EM bail conditions (see below), then they will remain in the community on EM bail until their trial is heard.

6.8 The second part of the model (illustrated in Figure 6.2) identifies the paths associated with a failure to comply. Individuals granted EM bail may or may not comply with the conditions set. Failures to comply can include the following:

1. the individual absconds;
2. the individual commits a new offence; and
3. the individual voluntarily or involuntarily fails to comply.

Figure 6.1. The EM Bail process

Figure 6.2 Failure to comply

6.9 Individuals who abscond may elude capture for the duration of the pilot or they may be re-arrested and remanded in custody until the next court appearance when the court can decide whether to grant EM bail or remand the individual in custody until trial. If an individual commits a further offence, the police might record the offence and take no further action, in which case the individual goes back on EM bail until the trial is heard for the original offence. Other failures to comply can be either voluntary or involuntary. Voluntary failures to comply include cases where an individual is re-arrested on an outstanding warrant for an alleged offence other than that for which they were originally charged. The police might record the alleged offence and take no further action, in which case the individual continues on EM bail until trial for the original offence. There may also be a court appearance to assess the individual's guilt of breach. If found guilty of breach the individual can either be remanded in custody until the trial for the primary offence, continue on EM bail (with or without extra conditions) or be put on standard bail until trial. If the individual is found not guilty of breach, EM bail could continue until the trial diet. Involuntary failures to comply can occur if someone is in hospital for an extended period or a householder withdraws consent to cooperate. EM bail may be reviewed and/or continued in these cases.

COST AND PROBABILITY

6.10 The above paths will have probabilities of occurrence and costs associated with them. These costs were estimated in this study following established principles and methods. The primary study perspective was the public sector, which determined the range of costs to be measured. Consequently, only resources paid for by the government and their agencies were included in the base analysis. The public agencies included in the analysis are the Scottish Court Service, sheriffs, the Crown Office, defence agents, Scottish Prison Service, the Scottish Executive, social work departments and the police. A second perspective was also examined in which the cost consequences of crime (during the bail period) are included alongside the public agency costs.

6.11 Costs are measured at 2004/2005 price levels. If appropriate costs for this period were not already available they were calculated specifically for the study, and costs from previous years were inflated using an appropriate index.

Sources of information

6.12 Various sources of information were used to develop the model, and to estimate the comparative cost of EM bail versus remand in custody. The Procedure Manual on Electronically Monitored Movement Restriction Conditions in Bail Orders (Scottish Executive, 2005c) was one of these sources of information and was the starting point for the development of the model structure. Consultations were held between officials at the Scottish Executive, members of the Research Advisory Group and the research team to identify possible events (for a typical case) from the point at which an application for standard bail was refused, and the application for EM bail was made, to the point where the applicant's trial was heard.

Cost parameters for analysis

6.13 The costs associated with EM bail can be divided into a number of activities, although the process is a continuous one. The main activities identified are:

• Application
• Appeals
• Failure to comply
• Monitoring

6.14 The costs associated with custodial remand and bail were estimated using data from publicly available sources. Costs incurred by each agency have been included under each of the main activities identified. However, not every agency will play a role in every event associated with EM bail. The unit costs calculated in this way and used in the analysis are shown in Table 6.1 and explained below.

Application

6.15 Those agencies most likely to be involved in the application stage of the process are the Crown Office, Scottish Court Service, social work departments, defence agent, sheriff, and Scottish Prison Service.

6.16 The procurator fiscal or a depute will consider the report from the police and decide whether it is appropriate to initiate proceedings, and if so in what form. If proceedings are to be undertaken against the individual, support staff prepare case papers for court. The procurator fiscal or depute would then appear in court to present the case. If a continued bail hearing is fixed, support staff will prepare the case papers for the procurator fiscal who will again have to appear in court and make an oral presentation. The unit cost estimate for the procurator fiscal is £175 per case for the application stage.

6.17 EM bail cases would have their first motions in either the summary or solemn court. If a suitability report is called for, then the cases are continued for an average of 5 days. The solemn court keeps their own cases if reports are called for. For the purposes of the following analyses it was assumed that the cost for the extra hearing in the solemn court is the same as that in the summary court. The extra hearing to consider the EM bail application usually lasts between 5 and 10 minutes. The estimated cost per case for the extra court hearing associated with a decision for an EM bail application in the summary court is £142 per case and in the solemn court £189 per case, with an average of £166 between both courts.

Table 6.1 Summary of costs per case incurred by various agencies

6.18 There was some degree of variability in who produces the suitability reports across the pilot sites. In Glasgow and Stirling, reports are written by social work department bail officers, while in Kilmarnock SACRO has been contracted to produce them. In Stirling, the social work department employed one full-time bail officer and one part-time clerical officer for 20 hours per week. These are discussed separately under sensitivity analysis (see para 6.51 below).

6.19 Interviews with the accused can often be conducted in the holding cells at the court and tend to last between 20 and 30 minutes. It takes on average 20 minutes to write the report at an estimated cost per case of £13.49. However, if it is not possible to conduct the interview in the holding cells at the court, the interview is conducted while the accused is on remand. Depending on the prison to which the accused is remanded, travel and interview time can be approximately 40 minutes if within the area, but if outside the area it can take up to 3 hours for the bail officer to travel to the accused and back. A conservative estimate of the cost to the social work department including travel is £27 per case for interviews conducted in prison; but the figure can be as high as £74 per case if the maximum travel time is considered.

6.20 It was suggested in Chapter 3 that between 70 and 100 per cent of interviews with accused are conducted in the holding cells. We have therefore taken the above costs and weighted them to account for the likelihood of the interview being conducted in holding cells as opposed to in prison, taking the mid point of the range (85 per cent) for our base case analysis. The weighted cost per suitability report is estimated at £16.

6.21 While the suitability report is being produced the accused is remanded in custody. As we have shown in Chapter 3 in relation to the referral process, this takes on average 5.7 days, with a maximum observed time on remand of 27 days. For this average of 5.7 days, the estimated cost is £519 per case.

6.22 A sheriff would preside over the EM bail application at second hearing. The average time taken for the second hearing and the reading of the report by the sheriff is estimated to be 10 minutes. The estimated cost per hour based on the number of sitting days per annum is estimated at £164.

6.23 The defence agent representing the accused under legal aid will receive a minimum of £50 per case for all work done in connection with an EM bail application. The recent amended legislation which allows an additional fee for the second hearing took effect from 12 ^th June 2006; however, sources in Glasgow suggest no claims had been submitted up until November 2006. Nevertheless these costs are included in the current analysis to capture the likely costs if EM bail is rolled out nationally.

Appeals

6.24 The main parties involved in appeals are: the Scottish Court Service, defence agent, procurator fiscal, sheriff and the Scottish Prison Service. Appeals against refusal of EM bail are made to the Appeal Court in Edinburgh, and such an appeal hearing usually lasts about 5 minutes. For those granted EM bail after an appeal there is an additional hearing at the High Court where the accused is brought from custody the next day to have the procedures explained in court. This usually lasts for about 10 minutes. The cost per hour of the sheriff's time used in the analysis is £218. The average cost per case for the Scottish Court Service going to appeal is £331. During the bail appeal process the procurator fiscal or depute appearing in court submits a bail appeal report to the Crown Office. An advocate depute and a Crown Office Trainee attend the high court hearing. During the appeal court hearing, the Advocate makes oral presentations and the judge considers the issue of bail. After the hearing, the trainee telephones the results of the bail appeal to the relevant procurator fiscal office that submitted the bail appeal report. We have used the average cost per case in the preparation of each custody case as a proxy for handling bail appeals. However there are caveats around the use of this estimate given that information on handling bail appeal work is not usually recorded in a manner that permits meaningful unit cost estimation. Actual case costs may be higher or lower depending on the circumstances of the case. The unit cost estimate of the input by the procurator fiscal used in the analysis is £175 per case.

Failure to comply

6.25 When a failure to comply with EM bail is recorded, the accused is only at this stage potentially in breach of the order. However, Reliance/Serco is required to notify the police, and the administrative arrangements by the police have been assumed to take between 20 and 30 minutes following notification of the failure to comply. Usually, two police constables are sent to investigate and that investigation can last 2-5 hours. If the accused cannot be traced within 72 hours, the police will submit a report to the procurator fiscal asking for a warrant to be raised for the arrest of the accused. Where a warrant has been raised the case would then be allocated to a police officer to deal with. It is assumed that the investigation by the police can take between 7 and 10 days. The cost to the police per case for failure to comply is thus estimated at £1,207. This estimate does not include any office overheads, allocation for police vehicles and equipment.

6.26 The hearing for failure to comply is presided over by a sheriff; the cost per hour of this sitting is £164.

6.27 The procurator fiscal can instigate breach proceedings against the accused where it is believed that this is justified. For breach proceedings following a failure to comply the cost to the Crown Office depends on the court in which the proceedings take place. We have assumed that the sheriff court is where the majority of cases are conducted and have therefore used this in deriving our estimate. The cost per case in the sheriff court is £333 per case.

6.28 A defence agent would usually be present at the court hearing for breach of bail. Assuming this work is carried out in the sheriff court, the cost per case to the Defence is estimated at £250. This figure represents all work done in connection with ordinary breaches of bail where professional services are provided in relation to proceedings in the Sheriff Court (Scottish Statutory Instruments 2004/23 Criminal and Legal Aid (Fixed Payments) (Scotland) Amendment (No. 3) Regulations 2004).

Monitoring

6.29 Monitoring and installation charges to the Scottish Executive by the monitoring company over the pilot were obtained and used to derive an average cost per case. These are net amounts invoiced and include any discounts given to the Scottish Executive. The total charge over 16 months for 101 people was £153,158, equivalent to an average charge of £1,516. Total charge data for those on EM bail was extracted from invoices sent to the Scottish Executive.

Probability parameters

6.30 Using data obtained from the agencies involved in the pilots, probabilities were derived, and these are given in Table 6.2. The data are obtained from the sample of 306 applications for EM bail. Difficulties already noted in earlier chapters in obtaining quantitative data which required cross-referencing of information for triangulation purposes meant that there is necessarily some uncertainty about the probabilities used.

6.31 Seventy five per cent of the cases had their application for EM bail considered by the Sheriff, and 52 per cent were not granted EM bail. Of those granted EM bail, as is to be expected, the majority (98%) had a positive suitability report. In very few cases (2%) the procurator fiscal or depute appealed the decision to grant EM bail. Of the 26 cases examined where the defence appealed there was a 77 per cent chance the appeal would be refused.

6.32 In the majority of cases (67%), there were no failures to comply where the Crown appealed the granting of EM bail. However, this probability should be treated with caution due to the small numbers on which data were available. In the cases where there were no appeals to the granting of EM bail, in 67 per cent of cases there was one or more failure to comply: 32 per cent had one failure to comply, 22 per cent had 2-4 failures to comply and 13 per cent had 5-7 failures to comply. In 31 per cent of the cases there was a new offence leading to a court hearing. In 14 per cent of the cases the accused on EM bail absconded.

Remand

6.33 The elements for costing remand were data on the average number of days on remand for those who were refused EM bail and remanded in custody and an estimate of the cost per day on remand. The mean number of days remanded was 56 (median 43). The cost per day for remand was derived using data on the cost per prisoner place for the financial year 2003/2004 and inflated. The cost per day per case was estimated

Table 6.2 Summary of probabilities used in the model

* Eight per cent were not recorded as either suitable or unsuitable

** The sample size on which probabilities were derived is 2 people.

*** The sample size on which probabilities were derived is 3 people.

at £91 ⁷. These data were used to derive an average cost per accused for those on remand over the duration of the pilot of £5,096.

6.34 There is asymmetry between EM bail and custodial remand at the sentencing stage. If someone is remanded into custody and subsequently convicted and given a custodial sentence, then under section 210(1) of the CP(S)A 1995 the court must take into account the time spent on remand in custody awaiting trial or sentence. This usually means that the sentence is 'backdated' to the start of the remand period. On the other hand, there is apparently no legal requirement to consider time spent in custody when sentencing someone who was on EM bail. If the custodial sentence for someone previously remanded in custody is longer than the period spent on remand, then the cost of the latter effectively becomes zero when making comparison with EM bail. We will take this into consideration later in the chapter.

Analyses

6.35 TreeAge Pro software (TreeAge Software - Data 4, Inc., Williamstown, MA) was used to conduct base analyses of the model. The expected costs of EM bail and remand in custody were derived by combining the data on cost to each agency with the probabilities of events occurring.

6.36 In addition to the costs to the various agencies estimated here, some costs of crime are borne by victims and society. Some victims have personal and property losses, others suffer physically and emotionally and may require the use of health services. There can also be lost income (to the individual) and lost productivity (to society) where victims have to take time off work.

6.37 Victim costs are difficult to quantify, but we were able to draw on a Home Office study that estimated the economic and social cost of crime against individuals and households (Dubourg et al., 2005). We use average estimates of the cost consequence of crimes that occurred during the pilot (vandalism, theft and common assault). Using these data, we estimated the cost consequence of crime to be £843 per accused.

6.38 We present the base case analyses with and without these cost consequences of crime.

Sensitivity Analysis

6.39 There was some uncertainty and inherent variability around some of the parameters used in the base case analysis and so we conducted sensitivity analyses. Due to the variability in the parameters used in the analysis, sensitivity analyses were conducted on:

6.40 For the base case analyses we used an average of 85 per cent of interviews being conducted in the cells, the remainder assumed to be conducted in prisons. It was suggested in Chapter 3 that between 70 and 100 per cent of interviews are conducted in the cells, and so we re-ran the analyses using 70 per cent and 100 per cent respectively.

6.41 It was assumed for the base case that the average time on remand while awaiting the suitability report is 5.7 days. However, reports can be produced in a minimum of the same day (in 4 cases) to a maximum of 27 days (in one case), during which time the individual is remanded. To assess the sensitivity of the results, we looked at this full range of experience, producing values between £91 for one day and £2,457 per case for a maximum of 27 days.

6.42 For the base case, we used a weighted average cost per case for the activities involved in preparing the suitability report. In Kilmarnock, this service is contracted out to SACRO. We assess the results using cost per case where the report is prepared by SACRO. The cost per report is based on the cost to the agency of providing this service in relation to the number of reports prepared over this period.

COMPARATIVE COSTS

6.43 We were interested in the expected cost associated with electronic monitoring relative to the expected cost of remand. This was investigated using a pruned version of the standard model populated by data from the pilot for costs, probabilities and length of time on remand for those refused EM bail. We report the results for two analyses excluding and including the cost consequences of crime.

Base case results: Comparative costs excluding cost consequences of crime

6.44 The base case results of the model after the 16 months of the pilots are shown in Figure 6.3 and Figure 6.4. On average, EM bail (£4,123) looks cheaper than remand (£5,096), although we have yet to take into account any backdating of subsequent custodial sentences for those on custodial remand. An accused making an application for EM bail through a defence agent is estimated to generate a cost to all agencies of £4,123. This reflects all of the extra costs incurred by the various agencies involved.

6.45 During the course of interviews conducted during the pilot, it was evident that the activities involved in piloting electronic monitoring do not have significant workload implications for agencies. Many agencies perform activities related to electronic monitoring as part of their regular duties. In Stirling and Glasgow, bail officers were employed to undertake activities associated with EM bail. This, however, is not the case in Kilmarnock where the production of reports was contracted out; we will see the effects of this on the overall results later in this chapter. Generally, however, we were told that there was a minimal impact on the workload of most agencies, and therefore the contribution to costs is not high.

Figure 6.3 Comparison of costs over 16-month pilot - EM bail process

Figure 6.4 Comparison of costs over 16-month pilot - failure to comply

6.46 The cost differential between EM bail and custody also reflects the lower average cost of monitoring and installation negotiated by the Scottish Executive in 2006. If there were to be a substantial change in the uptake of electronic monitoring resulting in a marked increase (or indeed decrease) in the number of installations and the need for monitoring, this could alter the average cost to the monitoring company, and in turn could alter the charge to the Scottish Executive. More people on EM bail could make the processing of them more efficient, for all agencies potentially, which would reduce the average cost per case. On the other hand, if EM bail was extended to a wider number of people, there is also the potential that it could draw in more serious offenders and in so doing increase the risk of failure to comply, which would push up the average cost.

Base case results: Comparative costs including cost consequences of crime

6.47 In the model we included the cost of crime (during the bail period) resulting from a new offence. These are shown in Figures 6.5 and 6.6. The expected cost of EM bail when the impacts of crime were included was £4,230 compared to a remand cost of £5,096. In fact, even if we were to increase the weighted average cost of crime by a factor of 10, EM bail would still be cheaper than remand in custody, largely as a result of the small numbers in the sample who went on to commit a new offence while on EM bail. For those on remand, it is possible that while in custody they will engage in activities for which they will be reprimanded at the prison. It is assumed that no extra resources are needed to deal with this latter scenario, and hence there are no additional costs.

Sensitivity analysis on base analysis excluding cost consequence of crime

6.48 In the first sensitivity analysis we changed the percentage of interviews conducted in the cells from 85 per cent to either 70 per cent or 100 per cent. This made very little difference to the cost of producing suitability reports or the overall cost of EM bail. Under the 70 per cent assumption, the expected cost of EM bail was £4,124, and under the 100 per cent assumption it was £4,121.

6.49 We examined the sensitivity of the results to a change in the days spent on remand pending suitability reports for people who eventually got EM bail. EM bail becomes more costly (at £5,673) than remand when this number of days is set equal to the maximum observed length of 27 days. If the accused is held on remand for just one day, the expected cost of EM bail becomes £3,780. If someone were to be held for 19 days on remand pending a suitability reports, then the total cost of EM bail would be almost exactly equal to the mean cost for the custodial remand group (based on 56 days, the observed average) after a failed application for EM bail. Figure 6.7 illustrates the pattern of cost differences under different assumptions. It should be remembered, however, that no adjustment has yet been made for backdating of sentence length for the custodial remand group.

Figure 6.5 Comparison of costs over the 16-month pilot: EM bail process, including cost consequences of crime

Figure 6.6 Comparison of costs over the 16-month pilot: failure to comply, including cost consequences of crime

Figure 6.7 Sensitivity analysis for days on remand pending suitability report

6.50 The third sensitivity analysis used the cost per suitability report where this is prepared by SACRO. This results in a narrowing of the cost difference between the 2 options, although there are still cost savings to be accrued from EM bail. At a cost per case of £1,366, the cost of EM bail is now £5,203. The cost per case is high due to the low number of cases dealt with over the period of the fieldwork in Kilmarnock (52 cases). However, if the numbers of cases increased there are potentially economies of scale which might work through to a reduction in the price negotiated with SACRO.

6.51 We also examined the sensitivity of the results to a change in the costs of preparing the social enquiry report by bail officers at Stirling. This has a marginal impact on the costs to be accrued from EM bail over those if the social work department conducts the social enquiry reports at Glasgow. If bail officers in Stirling conduct 70 per cent of the interviews in cells the cost of EM bail is £4,121. However, if 100 per cent of the interviews are conducted in cells, the cost of EM bail is £4,119.

Sensitivity analysis including cost consequence of crime

6.52 There was no significant change in the results under these 4 sensitivity analyses when the cost consequences of crime are included. Under the third sensitivity analyses, if a person was remanded in custody for 17.5 days (during the application process) the cost of remand would be £1,602. The expected cost of EM bail would then be equal to the cost of custodial remand. Periods in custody pending the suitability report below 18 days will result in a cost saving with EM bail.

Final sentence

6.53 In this study it was not possible to measure the outcomes of EM bail and remand in terms of impacts on the individuals concerned, such as subsequent criminal behaviour, but we do need to be aware of the 'destinational outcomes' in terms of the sentence received at the final trial diet. The reason is the asymmetry mentioned earlier between custodial remand cases and individuals on EM bail. A custodial sentence is backdated to the start of the remand period for the former, but there is no legal requirement to backdate the sentence for someone on EM bail who is given a custodial sentence. When comparing the costs of the 2 options, the backdating of the sentence for one group effectively means that the period in custodial remand has zero cost (when looking at the wider picture) for someone subsequently sentenced to custody for a period greater than the time spent on remand.

6.54 We explored 3 sources to get information on the number or proportion of people who are remanded in custody and subsequently receive a custodial sentence. The first source was the data collected during the pilot: 170 of the individuals refused EM bail were remanded in custody, 18 got standard bail and 2 were released prior to the second EM bail hearing. We are only concerned in this study with those who were remanded. Sentence for the original offence is known for 61 of these people, 35 of whom (57 per cent) received a custodial sentence. Length of sentence is not known for 2 cases, but 13 were sentenced (before backdating) to 30-90 days, 17 to 120-180 days, and 3 to 230 or more days. For most of these custodial remand cases, the length of custodial sentence is greater than the average duration of custodial remand (56 days). It should be noted, however, that this group of custodial remand cases does not necessarily give us the ideal comparison, as these are people who applied for but were refused EM bail, and they might not be completely representative of everyone given custodial remand. (For information, mean sentence length was 118 days for cases in the pilot who were previously on custodial remand, compared to 130 days for the 12 cases who were previously on EM bail who got a custodial sentence.)

6.55 A second source of information was the Scottish Executive who advised us that approximately 50 per cent of people on custodial remand group get a custodial sentence (Scottish Executive, 2000: p. 11, p. 92). We do not have further information on this calculation.

6.56 Third, we looked at evidence reported on the Scottish Parliament website ( http://www.scottish.parliament.uk/business/committees/historic/justice1/2003.htm) on the percentage of all remands who received a custodial disposal (Justice 1 Committee. 3 ^rd Report 2003. Inquiry into Alternatives to Custody Volume: Evidence SP paper 826). The estimated custodial disposal for all remands in 1997 was 46.3 per cent. This estimate is close to those suggested by each of the other two sources.

6.57 We therefore have three different estimates for the proportion of custodial remand cases that subsequently receive a custodial sentence that effectively reduces their remand cost to zero: 57 per cent, 50 per cent and 46.3 per cent. We do not know whether the sentence would be backdated in every case, nor do we know exactly how the length of sentence compares with the time spent in custody while on remand. However, from the information we have been given it seems reasonable to assume that a backdating is applied in every case, and that sentence length exceeds remand duration. Consequently, we are assuming that the calculated cost of custodial remand (£5,096) applies to only 43, 50 and 53.7 per cent of custodial remand cases, respectively. Using the first source of information to illustrate the calculation, we would then estimate - for the purposes of the comparison with EM bail - that custodial remand would cost on average £2,191 (= 0.43 x £5,096). Similar calculations can be done for the other two estimates of the percentage. Table 6.3 sets out the expected costs from these three alternative sources of information.

Table 6.3: Comparative costs between EM bail and custodial remand taking into account the backdating of custodial sentences

6.58 The last estimate in Table 6.3 can probably be discounted, based as it is on national data on all people charged and convicted across Scotland. Nevertheless, the conclusion from this estimate is in line with that from the other two, namely that custodial remand is much less expensive than EM bail (mean £4,123; and £4,230 if cost consequences of crime are included) when taking into account the backdating of custodial sentences. Indeed, so long as the percentage of custodial remand cases who receive a custodial sentence at final trial diet (of a length at least as great as the period spent on remand) is greater than 19.1 per cent, then custodial remand is less costly than EM bail (excluding cost consequences of crime). If cost consequences of crime are included, this 'threshold percentage' becomes 17 per cent.

Limitations

6.59 The analyses and estimates presented in this chapter have some limitations, particularly because they are based in part on assumptions that could not come from direct observational data, but which had to be made to produce the complete representation of the EM bail and custodial remand processes. The model is necessarily a simplification of the real world, but it provides a helpful representation that could be developed further if needed.

6.60 Another limitation is that, although the model uses data gathered from various sources during the fieldwork period, there were some difficulties with data collection which may affect certain parameters, such as the probabilities associated with Crown appeals and the seriousness and frequency of failures to comply. The data available were not always ideal for model-building purposes.

CONCLUSIONS

6.61 Given the increasing pressure placed by the remand population on the prison system in Scotland and the costs associated with housing prisoners who are untried and unconvicted at the time of imprisonment (Sentencing Commission, 2005; Scottish Executive, 2004), it is natural that increasing attention should focus on costs, particularly of alternatives to remand in custody.

6.62 In 2002, 19 per cent of the average daily prison population comprised prisoners detained on remand (Scottish Executive, 2004). This represents an increase of 3 per cent over 1994 estimates. This increase, coupled with the rising average period spent on remand, puts considerable strain on the prison system (The Sentencing Commission, 2005; HM Inspectorate of Prisons for Scotland, 1999). Increases in throughput contribute to increased remand costs. An annual estimate of £27.2 million for 1999 and an annual cost of £36 million for 2002 have been suggested for the remand population in custody (Safeguarding Communities Reducing Offending ( SACRO), 2004; HM Inspectorate of Prisons for Scotland, 1999).

6.63 The model built in this study allowed us to represent the routes that an accused

on EM bail would take through a number of processes until they reach the final trial diet. From this model and our collection of data to estimate probabilities and costs, we calculated that the average cost of EM bail is £4,123. If cost consequences of crime are included, relating to offences committed during the bail period, the mean cost becomes £4,230. The estimated mean cost for a custodial remand case is £5,096. Sensitivity analyses, making allowance for different values for some of the key parameters in the model, generally did not change this initial conclusion that the process of EM bail looks less costly than the process of remand.

6.64 However, this apparent cost difference does not take into account the differential treatment of time spent in custody during the pre-trial period: for people in custodial remand a subsequent custodial sentence is likely to be backdated to the start of the remand period. There is no equivalent backdating for EM bail cases. When comparing costs between EM bail and custodial remand, such backdating becomes pivotal because, under all suggested scenarios for the proportion of custodial remand cases that get custodial sentences at final trial diet, the cost ranking is reversed. For example, if 50 per cent of custodial remand cases are subsequently sentenced to custody, EM bail would be £1,575 more expensive per case. Overall, therefore, we conclude that EM bail is more expensive than custodial remand.

6.65 The pivotal element here is the percentage of custodial remand cases who receive a custodial sentence at final trial diet (of a length at least as great as the period spent on remand). If this is greater than 19.1 per cent, then custodial remand is less costly than EM bail (excluding cost consequences of crime). Were the courts to change their sentencing patterns considerably, and/or if custodial sentences took into account time spent on EM bail or in custody pending suitability reports for EM bail, then this pattern of relative costs could be different.

6.66 It is difficult to be sure about the economic consequences of a national roll-out of EM bail. With expansion, it would be reasonable to expect economies of scale in some of the elements of the processing of cases, and particularly in the monitoring of individuals. This could reduce average costs per case (which might be translated into a lower price per case where services - such as monitoring - are undertaken by agencies under contract from a public body). Another reason for expecting costs to fall could be that processing arrangements would get more efficient in the use of staff time as experience grows. On the other hand, rolling out EM bail to all parts of Scotland could push up costs if some activities are more costly in rural areas (because of travelling time). A second reason for expecting costs to rise would be if EM bail was extended to a wider group of accused individuals, to include people with a higher risk of failing to comply with conditions.

« Previous | Contents | Next »