3 Commentary on two year reconviction rates for 2005-06 cohort
(Tables 1 to 8)
3.1 The reconviction rates presented here apply to all offenders who were sentenced to the index disposal as their main penalty, irrespective of whether the offender was compliant (for non-custodial sentences). Information is not available via the Scottish Offenders Index on completion rates of orders.
3.2 Overall reconviction rates
After a drop in the number of offenders between 1996-97 and 2000-01 (from 54,192 to 41,569), the number has been increasing steadily at a rate of approximately 2,000 per year to 53,244 for 2006-07 (Table 1 and Table 10). While the reconviction rate has remained fairly constant over this period, the number of offenders has increased by 11,675.
3.3 Similarly, the number of offenders discharged from custody has also been increasing steadily since 2000-01, although the scale of increase is perhaps not as marked (Table 1).
3.4 Of the 50,242 offenders discharged from custody or given a non-custodial sentence in 2005-06, Table 2 shows that 22 per cent were reconvicted within 6 months, 32 per cent within one year and 45 per cent within two years. The proportion of offenders who received a custodial reconviction was 6 per cent within 6 months rising to 15 per cent after two years. There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.5 Reconviction rates by sex and age
Male offenders were more likely to be reconvicted than females, 46 per cent of males were reconvicted within two years compared to 36 per cent of females. This pattern applied across all age groups. However, for those persons who were reconvicted within two years, there was little difference between the sexes in the average number of reconvictions acquired. There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.6 For male offenders, the overall reconviction rate decreased with age. Fifty-eight per cent of those males aged under 21 were reconvicted within two years compared with 36 per cent of those aged over 30. Females aged 21-30 were more likely to be reconvicted than any other female age group, with 42 per cent being reconvicted within two years. Females over 30 were the group least likely overall to be reconvicted within two years (31 per cent). For custodial reconvictions, offenders aged in their twenties were more likely to receive a custodial reconviction compared with younger and older offender age groups. There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.7 Reconviction rates by number and type of previous convictions
Table 2 shows that the likelihood of reconviction increased markedly with the number of previous convictions held by the offender at the time of their index conviction. Around three quarters (74 per cent) of those offenders with over 10 previous convictions were reconvicted within two years compared to around a quarter (27 per cent) of offenders with no previous convictions. Similarly, 61 per cent of those offenders with over 10 previous custodial convictions were given a further custodial conviction within two years compared to 7 per cent of those with no previous custodial convictions. There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.8 The length of time to reconviction was also shorter for those offenders with higher numbers of previous convictions. Fifty-eight per cent of offenders with over 10 previous convictions were reconvicted within one year compared to 38 per cent for offenders with 3 to 10 previous convictions, 27 per cent with 1 or 2 previous convictions and 18 per cent of those with no previous convictions (Table 2). There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.9 For those offenders who had at least one reconviction within two years, the average number of reconvictions rose from 2.0 for those with no previous convictions to 3.2 for those with over 10 previous convictions (Table 3). There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.10 Reconviction rates by index disposal and index crime
Offenders given a Drug Treatment and Testing Order (81 per cent) or a Restriction of Liberty Order (68 per cent) were on average more likely to be reconvicted within two years than those given a Community Service Order (42 per cent) or a monetary penalty (41 per cent). Offenders given a Drug Treatment and Testing Order were most likely to acquire a custodial reconviction (Table 2).
3.11 Table 3 provides the average number of reconvictions for those offenders with at least one reconviction within two years. It shows that average number of reconvictions for those who were given a Drug Treatment and Testing Order or discharged from custody was 3.7 and 3.2 respectively. Whereas the average number of reconvictions for those who were given a Community Service Order, a monetary penalty or another form of disposal was 2.3, 2.2 and 2.3 respectively. There has been some change from the 2004-05 cohort for offenders given a Drug Treatment and Testing Order or a Restriction of Liberty Order. This appears to be of practical significance (see 2008 bulletin) but there are low volumes of these orders and the year on year differences of reconviction rates will be volatile.
3.12 For offenders discharged from custody, reconviction rates varied with sentence length, those with shorter sentences being more likely to be reconvicted. 74 per cent of those with a sentence of 6 months or less had a subsequent reconviction within two years, compared with 27 per cent for those with a sentence of four years or more. 54 per cent of those who had a sentence of 6 months or less had a subsequent custodial reconviction within two years, compared to 17 per cent for offenders discharged from a sentence of four years or more (Table 4).
3.13 The comparison of reconviction rates across disposals is complicated by a number of issues. Pseudo reconviction rates can have a differential effect across different types of disposal and this is discussed further in Section 5.1. In addition, the characteristics of offenders receiving different disposals are likely to be different. Indeed, the risk of re-offending is often taken into account at the time of sentence.
3.14 Once the age, sex and particularly the number of previous convictions of offenders are taken into account, the differences observed in average reconviction rates for different types of index disposal are found to be less significant. These factors are all strongly associated with the likelihood of reconviction. Table 5 shows that for all types of disposal, and for all sex and age groups, the groups of offenders with the highest reconviction rates were those with over 10 previous convictions. In other words, the number of previous convictions held by an offender appeared to be the dominant factor in terms of the likelihood of reconviction.
3.15 Reconviction rates also vary according to the index crime. Table 2 shows that those offenders with a sexual crime (excluding prostitution offences) as their index crime had the lowest reconviction rate on average (17 per cent within two years), although this is based on a particularly small number of offenders (520). Those with a crime of dishonesty as their index crime had the highest reconviction rate (58 per cent reconvicted within two years). There has been little change from the 2004-05 cohort (see 2008 bulletin).
3.16 Chart 2 and Chart 3 illustrate the two year reconviction and custodial reconviction rates for more detailed crime groups. Chart 2 and Table 6 show that the crimes within the dishonesty category which had the highest two year reconviction rates were shoplifting (70 per cent); housebreaking (70 per cent) and theft by opening a lock-fast place (64 per cent). Chart 3 shows that the violent crime of robbery (34 per cent) and housebreaking (40 per cent) were the two index crimes with the highest two year custodial reconviction rates. There has been some change from the 2004-05 cohort (see 2008 bulletin) concerning the crime of prostitution which had a reconviction rate of 84 per cent in 2004-05 and a reconviction rate of 77 per cent in 2005-06. However, there are low volumes of offenders committing these crimes and the year on year differences of reconviction rates will be volatile.
Chart 2 Reconviction rates by index crime: 2005-06 cohort
3.17 Many offenders acquire convictions for a variety of crimes. For example, Table 6 shows that, with the exception of those convicted of fraud, between 16 and 28 per cent of offenders with an index conviction in 2005-06 in the dishonesty crime categories were reconvicted for a violent crime within two years. Conversely, offenders convicted for certain categories of crime are more likely to be reconvicted for the same type of crime: the percentage reconvicted within two years for the same type of crime as the index crime was highest for prostitution offences (58 per cent); shoplifting (46 per cent); breach of the peace (29 per cent) and housebreaking (27 per cent).
Chart 3 Custodial reconviction rates by index crime: 2005-06 cohort
3.18 Reconviction rates over time
The analysis in this section compares two-year reconviction rates for the 2005-06 cohort of offenders with those for the corresponding cohorts from 1996-97 onwards. When considering trends over time in reconviction rates, it should be borne in mind that these will be affected by changes in criminal justice policy and practice. For example, the increased use of alternatives to prosecution such as fiscal fines in the last 10-15 years will, other things being equal, tend to reduce reconviction rates; while increases in police clear up rates may act to increase reconviction rates. In addition, this analysis is based on unadjusted reconviction rates. It is important to note that reconviction rates may also be affected by the characteristics of each cohort group ( e.g. age, sex, index crime, index disposal etc).
3.19 Table 1 shows that the overall reconviction rate was relatively stable between 1996-97 and 2005-06 with the exception of a slight fall between 1997-98 and 1999-00. From 1996-97 to 2005-06, the rates ranged between 42.5 per cent and 45.2 per cent.
3.20 A slight decrease in the reconviction rate for offenders aged under 21 was generally offset by slight increases in the rates for older offenders. The reconviction rate for males in previous cohorts were broadly similar to the 46 per cent recorded for the 2005-06 cohort. The reconviction rate for females decreased from the 38 per cent recorded for the 2004-05 cohort to 36 per cent for the 2005-06 cohort, mainly reflecting decreases in the rates for offenders ages between 21 and 25.
3.21 For those sentenced to a Community Service Order, the reconviction rate fell from 47 per cent for the 1996-97 cohort to 42 per cent for the 2005-06 cohort. The corresponding rates for custodial index disposals dropped slightly from 64 to 62 per cent as did the rate for Probation Orders from 60 to 58 per cent.
3.22 Table 1 also shows that the two year reconviction rate for those offenders with an index crime of violence remained largely similar with a rate of 38 per cent in the 1996-97 cohort and 37 per cent for the 2005-06 cohort. There was a dip in the series of reconviction rates for violent offenders between these two cohorts where the rate was 34 per cent at its lowest points in 1998-99 and 1999-00. The corresponding rates for breach of the peace fell by 1 percentage point from 45 per cent in 1996-97 to 44 per cent in 2005-06, whilst the rate for drug offences increased slightly from 39 to 43 per cent across the period. The rate for those offenders with an index crime of dishonesty increased from 51 per cent in the 1996-97 cohort to 58 per cent in the 2005-06 cohort.
3.23 Reconviction rates by local geographic breakdown
The analysis in this section compares two year reconviction rates for the 2005-06 cohort of offenders across approximate local authority areas, Community Justice Authorities and police forces, based on the location of the court where the index conviction was heard. Whilst this approach provides a useful estimate for these areas, it is important to note that an offender may not always be supervised in the area in which they are convicted and, additionally, that subsequent reconvictions may have occurred in different areas. The characteristics of offenders are also likely to vary across these areas.
3.24 The 2005-06 cohort can be grouped into approximate local authority areas based on the court of their index conviction, as in Table 7. This shows that the highest reconviction rate was for offenders whose index conviction was at a court in the Clackmannanshire area: 55 per cent of such offenders were reconvicted within two years, however there was a low number of offenders (577) and the rates will be subject to fluctuations from year to year. Falkirk had the second highest reconviction rate within two years (50 per cent) and there were a sufficient number of offenders (1,310) to limit year to year fluctuations. The lowest reconviction rate was for offenders whose index conviction was received at courts in Argyll and Bute, East Lothian and Orkney, 39 per cent of such offenders were reconvicted within two years.
3.25 These approximate local authority areas can also be grouped into Community Justice Authority areas. Table 7 shows that the highest two year reconviction rate is for offenders whose index conviction was at a court in Fife and Forth Valley (49 per cent). The area with the lowest reconviction rate for offenders convicted in court in this area was Lothian & Borders (41 per cent).
3.26 Table 8 illustrates the variation in reconviction rates by police force area. These rates are also based on the court where the index conviction occurred. The two year reconviction rates for Central (50 per cent) and Tayside (47 per cent) were the highest when presented by force area. In contrast, the rates for Lothian & Borders (42 per cent) were the lowest.