The Scottish Government

« Previous | Contents | Next »

Listen

RECONVICTION FOLLOWING DRUG TREATMENT AND TESTING ORDERS

CHAPTER TWO FINDINGS

2.1 Reconviction data were initially sought in respect of 271 offenders who had been given DTTOs during the first two years of the schemes: 127 in Glasgow and 144 in Fife.

CHARACTERISTICS OF THE SAMPLE

2.2 Information about the sex of the offender had not been sought from the schemes (since it was assumed that the number of women in the sample would be too low to enable any meaningful separate analysis of their data). The ages of offenders given DTTOs varied from 17 to 51 years, with a mean age of 26.9 years. As Eley et al. (2002) had previously observed, Glasgow DTTO clients were older, on average, than those in Fife (28.9 years compared with 25.2 years) ². In terms of age, the reconviction sample matches the sample used in the original study.

2.3 Across the sample as a whole, 118 DTTOs (44%) had been completed successfully (including 3% that were discharged early) and 130 (48%) had been revoked for a breach of conditions or because it was impractical for the order to continue ³. In 15 cases in Glasgow (6% of the total sample) breach proceedings had been initiated but the order expired before the breach was heard in court. Three orders were still ongoing, one was transferred to another area and two clients had died. Information about the outcome of the order was not provided in two cases.

2.4 The subsequent analysis focused upon the completions and revocations (that is, the six orders that were ongoing or that ended for other reasons were excluded). The 15 outstanding Glasgow breaches that completed were defined as 'revocations' (bringing the total in this category to 145) since it is likely that they would have been revoked if they had been brought to court before the expiry of the order. With the data re-categorised in this way, 45 per cent of orders were successfully completed and 55 per cent were revoked. However there was a marked difference in outcomes across the two DTTO schemes, with a significantly higher percentage of completions in Glasgow (59%) than in Fife (32%) ⁴.

2.5 Those who completed their orders successfully were older, on average, than those whose orders were revoked (28.3 years compared with 25.9 years) ⁵. It is therefore possible that the lower completion rate in Fife is attributable to that scheme supervising younger offenders. The data were therefore re-analysed excluding individuals under 21 years of age, resulting in completion rates of 59 per cent and 37 per cent respectively in Glasgow and Fife ⁶. It appears, therefore, that factors other than age are responsible for the differential 'success' rates across the two schemes.

2.6 Offenders who completed a DTTO were on their orders for, on average, 70.6 weeks. This compares with an average of 48.4 weeks prior to a DTTO being revoked ⁷. There was no significant difference in the mean time on an order in Glasgow and Fife (59.5 and 57.2 weeks respectively).

2.7 The 271 cases in this sample included 15 individuals who were given a second DTTO during the relevant period (usually shortly after their initial order was made). These duplicate cases were excluded from subsequent analysis. In addition, the information provided by the schemes proved insufficient to locate the SCRO records of 42 individuals (possibly as a result of inaccurate recording of URNs by the schemes or by social workers who gave this information to them). JSU were able to locate the conviction records of 214 individuals. There was no evidence of differential attrition of data across schemes ⁸ nor according to whether the order had been completed or revoked ⁹, giving no reason to suspect that this smaller sample was any less representative than the larger sample from which is was obtained. The subsequent analysis therefore focused upon these cases: 102 from Glasgow and 112 from Fife. Twelve month reconviction data were available for each of these cases. Full 24 month data were available in respect of 188 offenders. In the other 26 cases, the DTTO had been imposed during 2002 and, because there is a delay in JSU receiving updated conviction data from SCRO, it was assumed that by July 2004 JSU would not necessarily have complete 24-month data for these cases.

CRIMINAL HISTORY PRIOR TO A DTTO

2.8 The previous criminal histories of the Fife and Glasgow cases are summarised in Table 1. The findings support the conclusion by Eley et al. 2002 (based on a smaller number of cases) that DTTOs were being targeted in general upon apparently high tariff offenders. For example, the number of previous convictions among the sample varied from one to 71, with a mean of just under 18 convictions per person. The number of offences involved will be even larger, given that a single conviction often includes several charges. It is also apparent from Table 1 that the criminal histories of offenders from Fife and Glasgow were very similar. None of the small differences found between them with respect to previous convictions or custodial sentences were statistically significant.

Table 2.1: Convictions and custodial sentences prior to a DTTO

2.9 Some differences were found in the criminal histories of those who completed their DTTOs and those whose orders were revoked or breached. The latter had significantly more convictions in the two year period prior to their DTTO being imposed (5.2 compared with 3.7 ¹⁰) and had served, on average, more custodial sentences before being made subject to an order (8.7 compared with 6.3 ¹¹). Age was found to be negatively correlated with criminal history. In other words, younger offenders tended to have more convictions in the 24 months before their DTTO than did older offenders ¹².

RECONVICTION FOLLOWING A DTTO

2.10 The analysis of recidivism was able to take account of both the rate of reconviction and the frequency of convictions before and after the date of imposition of the DTTO.

2.11 Overall, 69 per cent of offenders were found to have been reconvicted within 12 months and 82 per cent within 24 months. Twenty-nine per cent of the sample had been imprisoned within 12 months and 44 per cent within 24 months. However, when the data were adjusted to exclude convictions that were known to relate to charges pre-dating the DTTO the reconviction rates after 12 and 24 months were 41 per cent and 66 per cent respectively and the imprisonment rates over the same periods were 21 and 39 per cent. These data are summarised in Table 2 by scheme.

Table 2.2: Overall and adjusted reconviction and imprisonment rate by scheme

2.12 For both the overall reconviction data and the adjusted reconviction data the rates of reconviction and imprisonment were slightly lower in Glasgow than in Fife. The 12 month adjusted reconviction rate was significantly lower in Glasgow than in Fife ¹³, while the 24 month rate was also lower but not significantly so.

Table 2.3: Overall and adjusted reconviction and imprisonment rate by how order was terminated

2.13 The rates of reconviction and imprisonment varied widely between offenders on DTTOs who completed their orders and those whose orders were breached or otherwise revoked. These differences were statistically significant for all comparisons, with completers having a lower likelihood of reconviction and imprisonment at both follow-up periods and with and without adjustments of the data to take account of pseudo-reconvictions ¹⁶. In this respect the findings echo those of Hough et al. (2003), though their 24-month reconviction rate for non-completers (91%) was higher than the adjusted rate for the present sample (79%) ¹⁷.

2.14 Although it is possible to compare these findings with the reconviction data following custodial and non-custodial sentences published by the Scottish Executive (2003) such a comparaison is problematic. It suggests that 2 year unadjusted reconviction and imprisonment rates for DTTO completers and non-completers were higher than those following custodial and non-custodial disposals. However, given the high tariff nature of the DTTO, it is likely that those given such orders have more serious criminal histories than those given disposals such as community service or probation and previous criminal history is a strong predictor of reconviction ¹⁸ (e.g. Lloyd, et al., 1994; May, 1999). The Scottish Executive statistical data includes an analysis of reconviction by number of previous convictions. The reconviction rates for those with 10 or more previous convictions ¹⁹ were 67 per cent and 80 per cent after 12 and 24 months respectively, which is similar to the unadjusted data for the total DTTO sample (69% and 82% respectively). This suggests that reconviction following a DTTO is at least no worse than following other disposals when account is taken of criminal history. Recidivism following a completed DTTO appears a little lower than that among those with similar criminal histories given other disposals.

2.15 In the absence of a comparison group of offenders matched on a number of key variables, it is possible to examine the proportions convicted before and after the imposition of a DTTO (that is, excluding the conviction for which they received this order). The relevant data are presented in Figures 1-3 for the sample as a whole and separately for completers and non-completers. These data are also summarised in Table 4 along with significance levels derived from the McNemar test (in each case the comparison is with the 'before' figure in the same row).

Table 2.4: Percentage convicted before and after imposition of a DTTO

***p<.001 **p<.01 *p<.05

2.16 For the sample as a whole, the percentage reconvicted after 12 months was not significantly different from the percentage convicted in the 12 months prior to being made subject to an order. However, the effect of pseudo-reconvictions is particularly marked with shorter follow-up periods, and with the convictions for 'old' offences removed the conviction rate in the 12 months following release (12 months adjusted) was significantly lower than in the 12 months before (Figure 1). In the case of the 24 month data, unadjusted reconviction rates were slightly lower than the rates of conviction in the 24 month period before the DTTO while adjusted rates were significantly lower (Figure 1). The differences were most marked among those offenders who completed their DTTOs (Figures 2 and 3).

Figure 2.1: Percentage convicted before and after a DTTO

Figure 2.2: Percentage convicted before and after a DTTO - those completing an order

Figure 2.3: Percentage convicted before and after a DTTO - those not completing an order

TIME TO FIRST RECONVICTION

2.17 The mean time to first conviction for the sample as a whole was 167 days, rising to 302 days when pseudo-reconvictions were excluded. In other words, those who were convicted of further offences remained free of further convictions on average for just over 43 weeks. The mean number of days to first conviction (adjusted for pseudo-reconvictions) was 312 for those who completed their orders and 299 for those who did not. It perhaps appears surprising that those who did not complete their orders were able to remain free of convictions for a similar period of time to the completers, however there are a number of possible explanations. First, further convictions per se will not necessarily result in the breach of an order so that it would be possible to commit further offences yet go on to complete a DTTO. Second, revocations for reasons other than breach are likely to have been prompted by the imposition of a custodial sentence of such as length as to render continuation of the order impractical. These offenders would not, therefore, have been at liberty to commit further offences for a significant period of time. Likewise, breach of an order for non-compliance is likely to have resulted in the imposition of a custodial sentence, reducing the time during which offenders were at liberty to re-offend.

2.18 The time to first new conviction did, however, vary significantly across the two DTTO schemes. Offenders in Fife were convicted more quickly than were those in Glasgow. The numbers of days to first conviction (unadjusted and adjusted for psuedo-reconvictions) were 224 and 353 respectively in Glasgow and 116 and 258 respectively in Fife ²⁰. In other words, Fife offenders were likely to remain free of convictions, on average, for just under 37 weeks while those in Glasgow were free of new convictions for, on average, just over 50 weeks. This would suggest that DTTOs imposed in Glasgow were effective in suppressing further re-offending for longer periods of time though it is also possible that these findings could reflect differences in the speed of processing cases in Glasgow and Fife. However, it has already been noted that completion rates were lower in Fife. Taken together these findings suggest that the DTTO regime in Fife may have been less effective in engaging offenders in treatment and reducing their need to re-offend.

2.19 It must be noted, however, that time to first conviction does not equate to time to first offence since a period of several months will often elapse between a charge being made and the case being disposed of in court. This would suggest that on average first offences were probably committed in Glasgow around 4-5 months into an order and in Fife within the first few weeks of an order being made. These data would certainly be consistent with the self report data from Glasgow DTTO participants, who suggested that they had ceased offending after being given an order but in some cases had had a relapse four or five months into their DTTO (Eley et al., 2002) ²¹. The small number of offender interviews that were conducted in Fife for that evaluation ²² prevents us similarly linking the two sets of data in this way. However, it is possible that prescribing practices across the two schemes may account for the differences. The DTTO evaluation indicated that offenders in Glasgow progressively registered fewer positive tests for heroin over time but that a similar steady reduction was not achieved in Fife, though again this may have partly been a result of the short timescales involved. However, subsequent evaluation of the Drug Courts - which developed from the pilot DTTO schemes - suggested that the prescribing regime in Fife resulted in offenders being slower to access the levels of medication they required to prevent the need to 'top up' with additional street heroin (Eley, Malloch et al., 2002; Malloch et al., 2003).

OFFENCES COMMITTED FOLLOWING A DTTO

2.20 Although JSU can not provide details of each of the offences subsumed under a conviction, the nature of the main offence was recorded. These data are summarised in Table 5. By far the most common group of offences committed by the sample involved dishonesty (for example shoplifting, housebreaking, theft from vehicles and fraud), with half of the sample being convicted of offences of this kind within two years (adjusted figures). Public order offences were also relatively common (including breaches of the peace, petty assaults and possession of offensive weapons) as were offences against justice (such as breaches of bail and obstruction of a police officer). Whilst a small proportion of offenders were convicted of relatively serious violent offences (including robbery and culpable homicide) the majority of new convictions were for 'offences' rather than 'crimes'. This is further underscored by the disposals received by those convicted of further offences committed after being given a DTTO. While 52 per cent of this group were imprisoned within 12 months and 60 per cent within two years, 38 per cent were at worst admonished or received a fine in the first 12 months while these options constituted the most severe disposals imposed upon 28 per cent of the sample over a period of two years.

Table 2.5: Nature of new offences for which reconvicted (main offence only)²³

FREQUENCY OF RECONVICTION

2.21 Another means of assessing whether DTTOs had any impact on recidivism is to compare the frequency of offending before and after the making of an order. If the DTTO has had little or no effect, it would be expected that the frequency of reconviction would be broadly similar in the two time periods. If, on the other hand, it has impacted positively upon recidivism, the frequency of reconviction would be expected to be lower following the making of an order. The relevant data are summarised in Figures 4 to 6 for the sample as a whole then separately for completers and non-completers. They are also presented in Table 6. The relevant comparisons for which significance levels are provided are between the 'before' data and adjusted 'after' data.

Table 2.6: Mean number of convictions before and after imposition of a DTTO

***p<.001 **p<.01 *p<.05

2.22 Compared to the similar periods before the DTTO was imposed, the adjusted mean numbers of offences in the 12 and 24 months after the DTTO was imposed were lower for both non-completers and completers. However the decrease in the frequency of convictions was more marked for the latter group where the mean number of convictions after 12 and 24 months was around one-third of that in the corresponding periods before the DTTO was made. By comparison, the mean number of convictions among non-completers following the imposition of a DTTO was roughly two-thirds of that prior to the order being made.

Figure 2.4: Mean number of convictions imposed before and after a DTTO - all cases

Figure 2.5: Mean number of convictions for imposed before and after a DTTO - those completing an order

Figure 2.6: Mean number of convictions for imposed before and after a DTTO - those not completing an order

2.23 There were also differences between the two schemes in the frequency of reconviction following a DTTO. Overall, offenders in Fife had more new convictions within both the 12 month and 24 month follow-up periods. However, this might be at least partly explained by the fact that a higher proportion of offenders in Fife failed to complete their orders and non-completion of orders was associated with a higher rate of recidivism. For this reason, the frequency of reconviction by scheme was compared controlling for the outcome of the order.

2.24 Among those who failed to complete their orders, offenders in Fife were convicted of more offences, on average, in the 12 months after being put on an order (3.8 compared with 1.8 in Glasgow ²⁴). A similar finding emerged with the data adjusted to take account of pseudo-reconvictions (with a mean of 2.0 convictions in Fife and 0.7 in Glasgow ²⁵). Despite being similar to their counterparts in Glasgow with respect to criminal history, offenders in Fife, therefore, were quicker to be reconvicted, were reconvicted of more offences and were more likely not to complete their DTTOs.

« Previous | Contents | Next »