Reconviction rates in Scotland: 2016-2017 offender cohort

1. Main findings: reconviction rates for court disposals

1.1 Headline figures

(Table 1)

The reconviction rate and average number of reconvictions per offender (Table 1 and Chart 1) have generally decreased over the past decade. Between 2007-08 and 2016-17, the reconviction rate has fallen by four percentage points from 31.2% to 27.2%, and the average number of reconvictions per offender has fallen by 16% from 0.57 to 0.48. These reductions are set against the context of a 38% drop in recorded crime over the same period between 2007-08 and 2016-17 (Recorded Crime in Scotland, 2017-18). The Scottish Crime and Justice Survey also reveals a similar pattern of falling incidence of crime.

There was a slight increase in both the reconviction rate and average number of reconvictions per offender in 2008-09. This is due to the Summary Justice Reform, which was designed to take less serious cases out of the justice system at an earlier stage, and to improve the efficiency of court processes, which resulted in cases being processed faster through the courts.

Over the shorter-term, there was no change in the reconviction rate or average number of reconvictions for the latest cohort of 2016-17 compared to the previous cohort of 2015-16.

The size of the cohort had been increasing every year since a low of 41,695 in 2012-13. However, in 2016-17 the cohort size decreased by 8% to 40,606 from 44,051 in 2015-16. This is the smallest cohort in the past 20 years, and reflects the decreases in the number of people convicted seen in the Criminal Proceedings Statistics since 2015-16. The slight lag in the reconviction cohort compared to the number of people convicted is likely to be because those given custodial sentences are counted at sentence date in the Criminal Proceedings Statistics, whereas they are counted later in a reconvictions cohort at the estimated time when they are released.

1.2 Age and gender

(Table 2, Table 3, Table 4 and Table 5)

Gender

Continuing a persistent long-term trend, males have higher reconviction rates and a higher average number of reconvictions per offender than females (Table 2 and Chart 3 and 4). However, the gap has narrowed in recent years. The average number of reconvictions per offender for the 2016-17 cohort was 0.49 for males, and 0.43 for females. This is the same as the figure in 2015-16 for males, but there was a slight increase for females (from 0.42) in the same period. The reconviction rates in 2016-17 were 28.0% for males and 23.3% for females, which is a slight 0.2 percentage point decrease for males (from 28.2%), but a small 0.6 percentage point increase for females (from 22.7%) from the preceding year of 2015-16.

Over the longer-term, numbers for males and females have generally decreased, but have occasionally fluctuated year to year. In the past decade, the average number of reconvictions fell by 17% for males from 0.59 in 2007-08 to 0.49 in 2016-17, and the reconviction rate fell by 4.2 percentage points from 32.2% to 28.0% in the same period. The figures for males are now similar to those for females just over a decade ago. For females, the average number of reconvictions decreased by 10% from 0.48 in 2007-08 to 0.43 in 2016-17, and the reconviction rate decreased by 3.1 percentage points from 26.4% to 23.3% in same period.

As the decrease has been greater for males over the past decade, the gap in reconvictions has narrowed between males and females. Ten years ago, the average number of reconvictions for males were 23% higher than females, compared to 14% higher in the most recent cohort.

Chart 3: Average number of reconvictions per offender for males and females, 2016-17 offender cohort

Chart 4: Reconviction rate for males and females, 2016-17 offender cohort

Age

There was a mixed picture in the changes in reconviction rates and average number of reconvictions between 2015-16 and 2016-17. The under 21 and over 40 age groups showed decreases, the 21 to 25 age group showed an increase, whereas the 26 to 30 and 31 to 40 age groups were near identical to the previous year (Table 3).

Reconviction rates and average number of reconvictions for under 21s were the highest of the age groups in 2016-17 (32.2% and 0.59, respectively), and lowest for the over 40s (19.9% and 0.34, respectively), as they have been historically. The other age groups (21 to 25, 26 to 30, and 31 to 40) were quite similar to each other in 2016-17, with reconviction rates ranging from 29.3% to 29.7%, and average number of reconvictions ranging from 0.49 to 0.54 (Table 3).

Over time, the average number of reconvictions for under 21s have decreased by more than a third (37%), from a high of 0.93 twenty years ago in 1997-98 to the latest figure of 0.59 in 2016-17. The figures for under 21s had seen a rise in 2013-14 and 2014-15, but have decreased in the past two years and the 2016-17 figures are the lowest levels recorded in the period covered by this publication (Table 3).

It should also be noted that the under 21 cohort size decreased significantly in the past 20 years by 69% from 13,796 in 1997-98 to 4,340 in 2016-17. It has decreased by 65% in the past decade alone from 12,403 in 2007-08. In terms of its effect on the national cohort size, 20 years ago the under 21 cohort represented 26% of the national cohort, compared to 11% in the most recent cohort. This large change in the under 21 cohort size, coupled with the decrease in reconvictions for this group, means that they have been a significant driver in the reduction in the overall national reconviction rate.

The reconviction rate for the 21 to 25 age group increased by 1.1 percentage points between 2015-16 and 2016-17, from 28.5% to 29.6%. The average number of reconvictions per offender increased very slightly in the same period from 0.48 to 0.49. Reconvictions for this group have fluctuated in the past 4 years, but they are still lower than they were a decade ago, with the average number of reconvictions 22% lower in 2016-17 than in 2007-08 (0.49 compared to 0.63).

Reconvictions for the 26 to 30 and 31 to 40 age groups are essentially unchanged between 2015-16 and 2016-17. For the 26 to 30 age group the reconviction rate of 29.3% in 2016-17 was slightly higher than the figure of 29.1%, in 2015-16, and the average number of reconvictions was 0.51 in both years. Similarly, for the 31 to 40 age group, the reconviction rate in 2016-17 was also slightly higher (29.7%) compared to 2015-16 (29.6%), and the average number of reconvictions was the same in both years (0.54). Over recent years the patterns have been different for these two groups, with reconviction rates for the 26 to 30 age group decreasing each year between 2011-12 and 2015-16, whereas reconviction rates for the 31 to 40 group have fluctuated (Table 3), although the average number of reconvictions for both groups has showed a downwards trend over the same period.

The reconviction rate for the over 40 age group increased from 2010-11, but has decreased in the past two years. In the past year, the reconviction rate decreased slightly by 0.4 percentage points from 20.3% in 2015-16 to 19.9% in 2016-17. The average number of reconvictions has been gradually decreasing for the past three years and decreased slightly from 0.35 to 0.34 between 2015-16 and 2016-17. Reconvictions for this group in 2016-17 were slightly higher than they were a decade ago, and also higher than they were at their lowest level in 1999-00 with a reconviction rate of 16.8% and average reconvictions of 0.29 (Table 3). The cohort size for the over 40 group is almost double the size it was 20 years ago, and as the over 40 group has the lowest reconviction rates of any age group, this has been a partial driver of the reductions seen in the overall national reconviction rate.

Age and gender

Patterns of change in reconvictions (both rates and average numbers) for males of different age groups are generally similar over time (Table 4) to those for all offenders (Table 3), as males comprise the majority of offenders in the cohort (82% in 2016-17) (Chart 5).

For male age groups between 2015-16 and 2016-17, the reconviction rate decreased for all age groups except the 21 to 25 age group. The decreases ranged from a 0.1 percentage point decrease in the 26 to 30 and 31 to 40 age groups, to a 1.4 decrease in the under 21 age group. The reconviction rate for the 21 to 25 year age group increased by 1.1 percentage points. The average number of reconvictions also saw the same directions of change as the reconviction rates over the same period, with the exception of the under 21 age group which did not change (Table 4).

Under 21s had the highest average number of reconvictions of the male age groups in 2016-17 (0.63), followed by 31 to 40 (0.54), 26 to 30 and 21 to 25 (0.51); with the over 40s having the lowest (0.35) (Chart 3). Historically the average number of reconvictions used to decrease with age, but in recent years the pattern has changed as the gap has narrowed between the age groups, and the 21 to 40 year age groups are now very similar to each other (Table 4 and Chart 5).

For females, patterns of change in reconvictions between 2015-16 and 2016-17 were mixed. The reconviction rate and average number of reconvictions decreased for the under 21 age group, with a 1 percentage point decrease in the reconviction rate and a 9% decrease in the average number of reconvictions. The over 40 age group saw a small decrease in both measures of reconvictions. The reconviction rate increased by 1 percentage point in the 21 to 25 and 31 to 40 age groups, and increased by 1.7 percentage points in the 26 to 30 age group. The average number of reconvictions increased by 9% for the 26 to 30 age group over the past year, and by 6% for the 31 to 40 age group. Although the reconviction rate increased over the past year for the 21 to 25 age group, the average number of reconvictions decreased slightly for this group (Table 5).

For females, the pattern of reconvictions across age groups is slightly different to males. In 2016-17, the age group with the highest average number of reconvictions was the 31 to 40 group (0.53), followed by 26 to 30 (0.50), under 21 (0.41), 21 to 25 (0.37); with the lowest for the over 40s (0.31) (Chart 3). The most notable difference to males is for the under 21 age group, which has always been highest of the male age groups, but this hasn’t been the case for the under 21 female age group since 2001-02 (Table 5 and Chart 6).

Chart 5: Average number of reconvictions per offender, males by age: 1997-98 to 2016-17 cohorts

Chart 6: Average number of reconvictions per offender, females by age: 1997-98 to 2016-17 cohorts

1.3 Index crime

(Table 6 and Table 7)

An “index crime” is the crime which resulted in the “index conviction”, the reference conviction which reconvictions are counted from. If a person was convicted for more than one charge in a set of proceedings, then the crime that was given the most serious disposal is counted as the index crime (see Annex A4). See Annex A5 for definitions.

In general, offenders who were convicted for lower level index crimes which tend to be committed in higher volumes, are more likely to be reoffend than those who commit more serious crimes. However, note that different disposals are given for different crimes, which may also affect the likelihood of reoffending. As has been true since 1997-98, offenders with an index crime of dishonesty, for example shoplifting (see Annex Table A2 for crime groupings), have the highest average number of reconvictions per offender and reconviction rate of any of the index crimes (Table 6 and Chart 7). For offenders convicted of crimes of dishonesty in the 2016-17 cohort, the reconviction rate was 43.5%, and there was an average of nearly 1 reconviction per offender (0.94). This compares to offenders with an index crime of a sexual crime, which had the lowest reconviction rate (10.6%) and lowest average number of reconvictions per offender (0.16) of any index crime (Chart 7 and Table 6).

Offenders from the 2016-17 cohort who had index crimes, other than sexual crimes or crimes of dishonesty, had an average number of reconvictions per offender ranging between 0.34 for violent crime and 0.50 for other crimes and offences. The reconviction rates ranged between 21.5% for violent crime, and 30.4% for other crimes and offences (Table 6 and Chart 7).

As with other trends in this publication, reconvictions for different crimes in 2016-17 were very similar to 2015-16 across all crime types (Table 6 and Chart 7). Reconviction rates and the average number of reconvictions for sexual crimes, criminal damage, breach of the peace, and other crimes and offences decreased slightly between 2015-16 and 2016-17. The largest decrease in the reconviction rate was a 0.8 percentage point decrease for breach of the peace from 26.0% to 25.2%, and the largest decrease in the average number of reconvictions was a 7% decrease for the Other crimes and offences category from 0.54 to 0.50.

The largest increase in reconvictions between 2015-16 and 2016-17 was for violent crime, which had a small 0.6 percentage increase from 20.9% to 21.5%, and a 6% increase in the average number of reconvictions from 0.32 to 0.34. However, reconvictions for violent crime in 2016-17 are still relatively low, with the second lowest recorded figures in this publication after the 2015-16 figures.

For drug offences and crimes of dishonesty, the reconviction rate slightly increased by 0.2 and 0.3 percentage points, respectively, between 2015-16 and 2016-17. The average number of reconvictions remained the same for the two types of crimes/offences in the same period.

Over the past decade, trends in reconvictions across crimes have been mixed. Violent crime, breach of peace, and drug offences have shown a general downwards trend, whereas other groups have tended to fluctuate from year to year with no clear trend (Table 6 and Chart 7).

Chart 7: Average number of reconvictions per offender, by index crime: 1997-98 to 2016-17 cohorts

Reconviction crime by index crime

Table 7 shows the types of crimes that offenders in the 2016-17 cohort were reconvicted for, by each type of index crime. The majority of offenders in the cohort (almost three quarters, 72.8%) were not reconvicted for any crime. For those that were reconvicted, more were reconvicted for breach of the peace than any other type of crime (10.2% of all offenders) and fewer offenders were reconvicted for a sexual crime (0.3% of all offenders).

Table 7 also highlights the degree to which offenders specialise on particular types of crime. Offenders convicted of crimes of dishonesty, drug offences, and breach of the peace were reconvicted for the same type of crime more than other types. Even in the cases where the majority of offenders were reconvicted for the same crime as their index crime, there were still other offenders who were reconvicted for different crimes to their index crimes. This suggests that offenders do not necessarily specialise on a particular type of crime.

1.4 Domestic abuse index crimes and offences

(Table 8a and Table 8b)

Table 8a and Table 8b show the reconviction rate and average number of reconvictions per offender, respectively, for offenders with domestic abuse and non-domestic abuse index crimes and offences from 2009-10 to 2016-17. The tables also show the percentage of offenders reconvicted for a domestic crime or offence, and the average number of reconvictions for a domestic abuse crime or offence per offender. Note that the crimes and offences are mostly referred to as “crimes” below for brevity.

The crimes counted here as domestic abuse index convictions or reconvictions are not standalone domestic abuse crimes. Those counted are crimes that are domestically aggravated, and are marked with a non-statutory domestic abuse identifier on the CHS. For example, a common assault offence committed against a partner would be given a domestic abuse identifier. Crimes are marked with the identifier by the police or COPFS and are taken into account during sentencing. A new standalone offence of domestic abuse was created by the Domestic Abuse (Scotland) Act 2018. This offence covers a course of behaviour which is abusive of a person’s partner or ex-partner. This offence was introduced on 1^st April 2019 which is later than the date covered by this bulletin so they are not counted here. This will be included in the future once they feed into the SOI. See Annex B for information on data quality.

Table 8a shows that offenders with a domestic abuse index crime were more likely to be reconvicted for a non-domestic abuse crime than a domestic abuse crime. In 2016-17, 20.4% of offenders with a domestic abuse index crime were reconvicted for any crime, compared to 9.2% convicted for domestic abuse crimes. Note that the percentage reconvicted for any crime includes those reconvicted for domestic abuse crimes, so the percent reconvicted for non-domestic abuse crimes can be calculated by subtraction, which gives a figure of 11.2% of offenders with a domestic abuse index crime reconvicted for a non-domestic abuse offence.

Table 8a shows that the percentage of offenders with a domestic abuse index crime who were reconvicted for a further domestic crime increased slightly by 0.4 percentage points from 8.8% in 2015-16 to 9.2% in 2016-17. Table 8b shows that there was a very small increase in the average number of reconvictions for a further domestic abuse crime for domestic abuse offenders from 0.10 in 2015-16 to 0.11 in 2016-17. Over the last six years, for domestic offenders the percentage reconvicted for a further domestic abuse crime and the average number of reconvictions for a further domestic abuse crime has been similar, with small year to year fluctuations.

Additional tables showing reconvictions for domestic abuse offenders by age, gender, crime, and disposal are published alongside this publication. The patterns for offenders with a domestic abuse index conviction that were reconvicted for a further domestic abuse crime follow similar patterns to reconvictions for all offenders across all types of crimes that are presented in the other sections of this bulletin. For example, a higher percentage of men are reconvicted for another domestic abuse crime than women (9.8% of males and 5.6% of females in 2016-17), and reconvictions for those given custodial sentences were higher than those given community sentences. There was no relationship between age and reconvictions.

1.5 Index disposal

(Table 9)

The index disposal is the sentence received for an index conviction (see Annex Table A1 and Annex A5 for definitions). If a person is convicted for more than one charge in a set of proceedings, then the charge that receives the most serious disposal is counted as the index disposal (see Annex A4).

A disposal may reduce the likelihood of reoffending as offenders are rehabilitated. However, different disposals are given for different types of crime and differing offending histories, and as seen elsewhere in this bulletin, these factors are also predictors of whether an offender is likely to reoffend or not. These factors should be considered when comparing the effectiveness of different types of sentences. Table 11 and the Reconviction rates by disposal, age, sex and previous convictions table in the additional tables published with this publication give reconviction rates for different offender characteristics for the disposals.

Custodial sentences

Reconviction rates for extended sentences (ES) and supervised release orders (SROs) are presented separately from other custodial sentences (Table 9). The custodial sentence category only includes those that were sent to prison or young offenders institutions, plus a small number of orders for lifelong restriction. Note that the SROs and ES are included in the custodial sentence length table along with the other custodial sentences (Table 10a).

Offenders released from a custodial sentence in 2016-17 had a higher reconviction rate and average number of reconvictions per offender than offenders given any other disposal except a DTTO. The reconviction rate for offenders released from custody in 2016-17 cohort was 42.3%, a 1.4 percentage point decrease on the 2015-16 rate of 43.7%. The average number of reconvictions per offender decreased by 5% from 0.86 to 0.82 in the same period (Table 9 and Chart 8). Note that reconviction rates for different lengths of custodial sentences vary considerably which is discussed in the next section of this bulletin.

Over the last decade, for those released from custodial sentences, there has been a 20% decrease from just over one reconviction per offender on average (1.02) in 2007-08, to below one on average (0.82) in 2016-17 (Table 9, Chart 8).

Extended Sentences (ES) are custodial sentences given for sexual crimes, or violent crimes that attract a custodial sentence of 4 years or more. ES have a period of supervision of up to 10 years in the community after the custodial sentence. If offenders breach their licence during the extended part of the sentence they can be recalled to prison. Reconviction rates for ES, like other custodial sentences, are based on the estimated release date from the custodial part of the sentence. We do not have information on the length of the supervision period on our dataset, just the length of the custodial part of the sentence. The reconviction rates are low compared to other disposals, in part, because they are generally given to sex offenders, who are typically reconvicted less than other offenders (Table 6 shows that offenders convicted of a sexual crime had the lowest reconviction rates for any index crime). Reconviction rates for ES decreased between 2015-16 and 2016-17 by 1.1 percentage points from 12.6% to 11.5%. The average number of reconvictions in 2016-17 was 0.13, which was slightly lower than 0.15 in 2015-16. Reconvictions have fluctuated from year to year for ES, which in part is probably due to the small cohort sizes and low numbers of reconvictions, where a small change would have a greater effect on percentages compared to larger cohorts (Table 9 and Chart 8).

SROs are given for crimes other than sexual crimes and consist of a custodial sentence of 1 to 4 years, followed by a period of supervision of up to a year by a social worker. If the offender breaches the order they can be returned to prison. As with ES, we do not have information on the length of the supervision period on our dataset. Reconvictions for SROs are lower than non-supervised custodial sentences. The reconviction rate of 36.6% for SROs in 2016-17 was slightly lower (0.2 percentage points) than the figure of 36.8% in 2015-16, and the average number of reconvictions decreased by 6% from 0.62 to 0.58 in the same period. In the last decade, reconvictions for SROs have been fluctuating, but were broadly similar in the last five years (Table 9 and Chart 8).

Community sentences: CPOs, DTTOs, RLOs

Community Payback Orders (CPOs) are a community sentence and consist of one or more of nine requirements imposed by the courts, including: offender supervision, compensation, unpaid work or other activity, mental health treatment, drug treatment and alcohol treatment. Every order must contain either an unpaid work or other activity requirement, or an offender supervision requirement (or both). If an offender fails to comply with a requirement in the order, the court can impose a number of sanctions, including a restricted movement requirement.

CPOs replaced the legacy community orders in 2011. There were still a very small number of offenders given a legacy community order in 2016-17 as they were given for offences committed prior to February 2011. During the transition from legacy orders to CPOs from 2010-11 to 2013-14, there were changes in the characteristics of offenders that were given these disposal types. Annex D gives a brief overview of the trends during the transitional period.

CPOs are the mostly widely used community sentence, with a cohort size of 10,016 in 2016-17. Reconvictions for CPOs were almost the same in 2016-17 as they were in 2015-16. The reconviction rate in 2016-17 was 31.2% which was 0.1 percentage points higher than 2015-16 and the average number of reconvictions per offender was 0.55 in both years. Reconvictions for CPOs have slightly fluctuated year to year after the transitional period from 2013-14 onwards, with no upwards or downwards trend (Table 9 and Chart 8).

A Drug Treatment and Testing Order (DTTO) is a high tariff disposal for people with serious drug use problems, and includes the requirement for regular reviews by the court and that the person consents to frequent random drug tests throughout the lifetime of the order. Offenders given a DTTO have the highest average number of reconvictions per offender and the highest reconviction rate compared to the other disposals (Table 9 and Chart 8). The average number of reconvictions per offender was 1.47 in 2016-17, which is a decrease of 16% from 1.76 in 2015-16. The reconviction rate decreased by 2.9 percentage points over the same period from 67.2% to 64.3%.

Over the longer term, reconvictions for DTTOs are lower than they were a decade ago, with the average number of convictions falling by 24% over the past decade from 1.94 in 2007-08 to 1.47 in 2016-17, although they fluctuated between 2008-09 and 2015-16. Reconvictions rates are also lower than they were 10 years ago, and have also been somewhat similar since then, with year to year fluctuations.

Restriction of Liberty Orders (RLOs) are imposed for periods of up to one year, and involve restricting an individual to a specified place for up to 12 hours per day and/or from a specified place for up to 24 hours. The reconviction rate for RLOs in 2016-17 of 33.6% was slightly lower (0.7 percentage points) than the figure of 34.3% in 2015-16. The average number of reconvictions was 0.59 in 2016-17 which was the same as the previous year.

Over the longer term, reconvictions are now much lower for RLOs than they used to be, with the average number of reconvictions decreasing by 43% over the past decade from 1.04 in 2007-08 to 0.59 in 2016-17. RLOs have been more widely used over the past decade, with the size of the cohort more than doubling from 556 in 2007-08 to 1,391 in 2016-17 (Table 9 and Chart 8).

Monetary and other disposals

The reconviction rate and the average number of reconvictions for monetary

disposals in 2016-17 were very similar to the figures in 2015-16. The reconviction rate decreased slightly by 0.2 percentage points from 21.3% in 2015-16, to 21.1% in 2016-17. The average number of reconvictions also decreased slightly over the same period, from 0.34 to 0.33.

The numbers of offenders with a monetary index conviction has more than halved in the past decade from 27,495 offenders in 2007-08 to 13,305 in 2016-17. This may, in part, reflect the impact of Summary Justice Reform which was designed to take less serious cases out of the court system, and deal with them using non-court disposals (see Section 2). Reconvictions have also fallen for monetary disposals in the past decade, with the average number of reconvictions decreasing by 28% from 0.46 in 2007-08 to 0.33 in 2016-17 (Table 9 and Chart 8).

For Other disposals, which includes admonishments, cautions, and absolute discharges; figures were also very similar between 2016-17 and 2015-16. The reconviction rate of 20.2% in 2016-17 was slightly lower (by 0.2 percentage points) than the figure of 20.4% in 2015-16. The average number of reconvictions per offender was slightly higher in 2016-17 with a figure of 0.35 compared to 0.34 in 2015-16.

Over time, reconvictions for Other disposals have fluctuated with no clear trend. In the past decade the reconviction rates have ranged from 20.2% to 23.5% and the average number of reconvictions per offender ranged from 0.34 to 0.44 (Table 9 and Chart 8).

Chart 8: Average number of reconvictions per offender by index disposal: 1997-98 to 2016-17 cohorts

1.6 Sentence length of custodial index conviction

(Table 10a and Table 10b)

Note that Table 10a which presents reconviction rates by custodial sentence length, includes ES and SROs, whereas they are presented separately from other custodial sentences in Table 9. This is because the numbers of ES and SROs are small in each sentence length category and it is difficult to interpret reconviction rates calculated on small groups.

Short custodial sentences have high reconviction rates (Table 10a and Chart 9). Offenders who commit relatively less serious crimes but in high volumes are more likely to be reconvicted (see Section 1.3), and these offenders are more likely to get short custodial sentences. In contrast, longer custodial sentences are given to offenders who commit more serious crimes, but these offenders tend to commit these crimes in low volumes, and hence fewer are reconvicted. For example, the reconviction rate for custodial sentences of 3 months or less in 2016-17 was 58.3%, compared to 11.1% for sentences over 4 years.

Chart 9: Reconviction rates for index disposals and sentence lengths for the 2016-17 cohort¹

1. Chart 9 shows reconviction rates for each disposal type. The category “disposal from custody” shows reconviction rates for all offenders discharged from a prison or young offender institutions in 2016-17. SROs and ES are presented separately. Custodial sentence length includes all custodial sentences (prison, young offender institutions, ES, and SROs).

There is a mixed picture for changes in reconvictions over the past year for different lengths of custodial sentences (Table 10a). Reconviction rates decreased for all sentence lengths of 1 year or less, but increased for sentences over 1 year. Average numbers of reconvictions also followed a similar trend, except for sentences of over 2 years and less than 4 years that saw an increase in the reconviction rate of 1.5 percentage points (the largest increase of any sentence length), but a small decrease in the average number of reconvictions.

The decreases in reconvictions for sentences between 3 months to 1 year, over the past year were notable. There was a 6% decrease in the average reconvictions for sentences of 3 to 6 months, from 1.08 in 2015-16, to 1.02 in 2016-17. Sentences of 6 months to 1 year decreased by 10% from 0.79 in 2015-16 to 0.71 in 2016-17. However, the 2016-17 figures for shorter sentences remain relatively high compared to other disposals (compare Table 10a to Table 9, and Chart 9 shows a comparison of reconviction rates).

The picture is somewhat complicated over the longer term for all sentence lengths (Table 10a). Over the past 20 years, reconvictions have fluctuated for all sentence lengths.

Table 10b shows reconviction rates and average number of reconvictions for all custodial sentences of 1 year or less from 1997-98 to 2016-17. These combine all the 1 year or less custodial sentence length reconvictions that are presented in Table 10a. This table has been included in response to user need for these figures.

1.7 Conviction history prior to index conviction

(Table 11)

Conviction history is a strong predictor for the likelihood of reconviction, as reconviction rates increase with increasing numbers of previous reconvictions. Offenders with more than 10 previous convictions have the highest reconviction rates, whereas offenders with no previous convictions in the past ten years have the lowest reconviction rates. This pattern holds true even when age, sex, or disposal (all of which have an association with the likelihood of reconviction) are taken into account (Table 11).

1.8 Two year rates

(Table 14)

Reconviction rates in Scotland were reported with a two-year follow-up period before the 2009-10 cohort bulletin. After this point, the focus has been on a follow-up period of one year rather than two years as, in general, the one-year rate tracks the two year rate, and has the benefit of being more timely.

Using the two-year follow up period there has been a decrease in the reconviction rate and in the average number of reconvictions per offender every year from 2009-10 onwards (Table 14). Between 2014-15 and 2015-16, the reconviction rate decreased by 1.7 percentage points from 39.0% to 37.3%, and the average number of reconvictions fell by 7% from 0.92 to 0.86. Over 10 years from 2006-07 to 2015-16, the average number of reconvictions per offender has fallen by 21% from 1.09 to 0.86, and the reconviction rate saw a 6.8 percentage point reduction from 44.1% to 37.3%.

These declining trends mirror those seen for the one-year follow up period (Table 1) but as there is a longer follow-up period, the associated values are typically greater, for instance:

• In 2015-16, the two-year reconviction rate was 10.1 percentage points higher than the rate for the one-year reconviction rate (37.3% for the two year compared to 27.2% for the one year rate).
• The average reconvictions per offender are over a third of a conviction (0.38) higher for the two-year rate (0.86 reconvictions per offender on average over two years, compared to 0.48 over one year).