Scottish Crime and Justice Survey 2017/18: main findings

3. Focus on violent crime

What was the extent and prevalence of violent crime in Scotland in 2017/18?

There were an estimated 172,000 violent crimes in 2017/18, representing around three out of every ten crimes experienced by adults during the year.

The Scottish Crime and Justice Survey estimates that 172,000 violent crimes were experienced by adults in Scotland in 2017/18^[13][14]. This figure accounts for 29% of all crime measured by the SCJS in 2017/18, with the remainder being property-related.

As a sample survey of the general public, SCJS results on the extent of violence are estimated values with relatively wide margins of error around them, rather than exact counts. Further information on the process used to calculate estimates is contained within the Technical Report. Taking into account these confidence intervals, the SCJS estimates that there were between 125,000 and 219,000 incidents of violent crime in Scotland in 2017/18. The following analysis is focused on the best estimates for each sweep of the survey.

The level of violent crime in Scotland is estimated to have almost halved since 2008/09.

Figure 3.1 displays the number of violent incidents estimated to have taken place by each sweep of the SCJS since 2008/09, and shows a large fall over the past decade^[15].

Figure 3.1: Estimated number of violent incidents, 2008/09 - 2017/18

Base: SCJS 2008/09 (16,000); 2009/10 (16,040); 2010/11 (13,010); 2012/13 (12,050); 2014/15 (11,470); 2016/17 (5,570); 2017/18 (5,480) Variable: INCVIOLENT

Table 3.1 examines results from key comparator years^[16] and shows that the estimated amount of violent crime experienced by adults:

• has fallen by 46% since the 2008/09 baseline, from 317,000 to 172,000 incidents in 2017/18;
• has shown no change since the last SCJS in 2016/17 – the apparent decrease from 231,000 violent incidents is not statistically significant.

Table 3.1: Estimated of number of violent crimes (2008/09, 2016/17, 2017/18)

Number of violent crimes	2008/09	2016/17	2017/18	Change since 2008/09	Change since 2016/17
Best estimate	317,000	231,000	172,000	↓ by 46%	No change
Lower estimate	275,000	172,000	125,000
Upper estimate	358,000	290,000	219,000
Number of respondents	16,000	5,570	5,480

Base: SCJS 2008/09 (16,000); 2016/17 (5,570); 2017/18 (5,480). Variable: INCVIOLENT

The fall in violent crime over the last decade has been mostly driven by decreases between 2008/09 and 2010/11, with some fluctuations but broad stability seen since then.

Violent crime estimates derived from the SCJS are based on a relatively small number of respondents who disclose experiences of such issues in the survey in any given year^[17]. As a result, analysis of findings between adjacent surveys are often less likely to identify statistically significant changes. For example, all of the apparent fluctuations shown from year-to-year since 2010/11 in Figure 3.1 are not statistically significant.

However, where they exist, the SCJS can often identify significant changes and trends over the longer-term (such as since 2008/09 as discussed above). Taking this into account, a more detailed examination of changes in the level of violent crime over the last decade finds that the estimated number of violent incidents:

• fell markedly between 2008/09 and 2010/11, and has remained below the 2008/09 baseline since then; but
• has been more stable in recent years - for instance, the apparent fall in violence from 2010/11 to 2017/18 is not statistically significant.

When looking at intermediate years, although the decrease from 2012/13 to 2017/18 is statistically significant, the wider trend and lack of significant change seen in recent years suggests this may represent fluctuation in the data. Therefore, overall the SCJS suggests that the level of violence experienced by adults in Scotland has been relatively stable since 2010/11.

The likelihood of experiencing violent crime is relatively small and has fallen since 2008/09.

Whilst the SCJS produces crime estimates which make it possible to examine trends in the volume of crime experienced over time, a particular strength of the survey is its ability to provide findings on the proportion of adults experiencing crime in any one year with a relatively good level of precision.

Looking at the victimisation rate, as in previous years the vast majority of adults in Scotland did not experience violent crime in 2017/18. The SCJS estimates that 2.3% of adults were victims of at least one violent crime in 2017/18. In comparison, an estimated 10.8% of adults experienced property crime over the same period.

However, like incident numbers, it is worth noting crime prevalence rates are also estimates with associated margins of error around them as they are derived from a sample survey of the population. Taking into account these confidence intervals, between 1.8% and 2.8% of the adult population were estimated to have experienced violent crime in 2017/18, with 2.3% representing the best estimate^[18]. Again, as with incident counts, analysis from this point onwards will focus on the best estimates for results across the survey for each sweep^[19].

Looking at trends over time, the proportion of adults experiencing violent crime has fallen from 4.1% in 2008/09 to 2.3% in 2017/18. Overall, this suggests that violent crime victimisation in Scotland has been relatively uncommon since 2008/09 and has become a rarer experience still over the last decade.

Since the last SCJS in 2016/17 there has been no change detected in the proportion of adults experiencing violent crime. The apparent decrease from 2.9% in 2016/17 to 2.3% in 2017/18 is not statistically significant, as shown in Figure 3.2.

Figure 3.2: Proportion of adults experiencing violent crime by year

Base: SCJS 2008/09 (16,000); 2016/17 (5,570); 2017/18 (5,480). Variable: PREVVIOLENT

However, as noted earlier, the SCJS is often better able to identify trends and changes, where they exist, over longer time periods. For example, looking further back reveals that the proportion of adults experiencing violence in 2017/18 (2.3%) was lower than the 3.1% in 2012/13 (and preceding years).

What types of violent crime were most commonly experienced in 2017/18?

As in previous years, the majority of violent crimes in 2017/18 were incidents of minor assault resulting in no or negligible injury.

Just over three-fifths of violent incidents in 2017/18 (62%) were cases of minor assault with no or negligible injury to the victim, as shown in Figure 3.3.

By comparison, one-in-twenty violent incidents (5%) in 2017/18 were serious assaults, whilst robberies accounted for 6% of all violence. Taken together, all categories of assault accounted for 94% of violent crime.

Figure 3.3: Categories of crime as proportions of violent crime overall in 2017/18

Base: 2017/18 (5,480). Variable: INCSERASSAULT, INCMINORASSINJURY, INCMINORASSNOINJURY, INCATTEMPTASSAULT INCROB.

Between 2008/09 and 2017/18, the SCJS has consistently estimated that assaults (including attempted, minor and serious assaults) have accounted for around 95% of violent crime experienced by adults. As such, trends over time in the number of assaults very closely mirror wider trends in violence, having fallen by 46% since 2008/09 (from an estimated 297,000 incidents to 161,000).

The strength of the SCJS lies in looking at the prevalence of robbery and serious assault, rather than estimating the number of incidents.

SCJS reports have previously provided estimates of the number of robberies (the other main category of violent crime captured by the survey) and sub-categories of assault, such as serious assault. Whilst being significant events for victims, these crimes represent small proportions of violence overall and are experienced by small proportions of the population (and therefore the SCJS sample), so have relatively large degrees of error around them^[20].

As such, for lower volume (although often higher harm) crime categories like serious assault, the strength of the SCJS is in examining how prevalent such experiences are in the population (i.e. demonstrating that a relatively small proportion of the population are affected), rather than estimating the number of incidents of these types of crime that occur in a single year or over time.

Therefore, with effect from 2017/18, the main body of the SCJS report will examine only the prevalence of such crimes. Estimates of the extent of these crimes will continue to be provided in Annex tables (see Annex Table A1.2) with advice on where such findings should be used with caution to help users interpret such results.

As (minor) assaults account for the vast majority of violence crime overall, it is worth noting that the later sections looking at the characteristics of violent crime in general are also mainly driven by the nature of these (higher volume, often lower harm) incidents. To enhance the wider evidence base on robbery and serious assault, Scottish Government statisticians and Police Scotland have carried out two studies into the characteristics of police recorded ‘Robbery’ and ‘Attempted murder and serious assault’. These studies have involved examining a sample of police records (rather than all records) to provide a broad indication of the characteristics of the crime type.

The report on robbery was published in September 2018 and is based on a review of police recorded robberies from both 2008-09 and 2017-18 to explore the extent to which robbery committed more recently in Scotland may differ in its character to robbery committed 10 years ago. The second study, on the characteristics of ‘Attempted murder and serious assault’, will be published in Spring 2019, and will follow a similar format.

The likelihood of experiencing assault and robbery have both fallen since 2008/09.

Returning to SCJS findings, as in previous years the prevalence rate for different categories of violent crime varied, with the SCJS estimating that 2.2% of adults were victims of assault in 2017/18, whilst 0.2% experienced robbery. In other words, despite still being a relatively rare experience, adults were more than ten times as likely to have been victims of assault in 2017/18 than robbery.

Examining trends over time, the SCJS finds that the prevalence of assault has fallen from 3.8% in 2008/09, again demonstrating a similar trend to violence overall. The proportion of adults experiencing robbery halved from 0.4% to 0.2%^[21] over the same period, as shown in Figure 3.4.

Figure 3.4: Proportion of adults experiencing types of violent crime, 2008/09 – 2017/18

Base: SCJS 2008/09 (16,000); 2016/17 (5,570); 2017/18 (5,480). Variables: PREVASSAULT; PREVROB.

Whilst a small proportion of adults were victims of any sort of assault in 2017/18, experiences of more serious forms of violence were rarer still. This is in line with the finding that the vast majority of violent crime was accounted for by minor assaults. For instance, a greater proportion of adults experienced minor assault with no or negligible injury (1.5%) than minor assault resulting in injury (0.4%) or serious assault (0.1%).

How did experiences of violent crime vary across the population?

Those in deprived areas were more likely to be victims of violence in 2017/18, whilst such experiences were less likely for adults aged 60 and over.

The SCJS enables us to examine how experiences of violent crime varied across the population by demographic and geographic characteristics. As shown in Figure 3.5:

• The likelihood of experiencing violence decreased with age - for example those aged 60+ were least likely to be victims of violent crime in 2017/18, with fewer than 1 in 200 experiencing violence in this age group, compared to just over 1 in 20 of those aged 16-24;
• Adults living in the 15% most deprived areas in Scotland were almost twice as likely to have been victims of violence than people living elsewhere.

Figure 3.5: Proportion of adults experiencing violent crime, by demographic and area characteristics

Base: 2017/18 (5,480). Variable: INCSURVEYCRIME, QDAGE, SIMD_TOP.

The proportion of younger adults experiencing violent crime has more than halved since 2008/09, whilst the prevalence rate for other age groups has been stable.

Looking at trends over time, the proportion of 16-24 year olds who were victims of violence has more than halved since 2008/09, falling from 12.0% to 5.8%, as shown in Figure 3.6. In contrast the prevalence rates for all other age groups showed no change over the same period – the apparent decreases shown are not statistically significant.

Figure 3.6: Proportion of adults experiencing violent crime by age over time

Base: SCJS 2008/09 (16,000); 2009/10 (16,040); 2010/11 (13,010); 2012/13 (12,050); 2014/15 (11,470); 2016/17 (5,570); 2017/18 (5,480) Variable: PREVVIOLENT, QDAGE.

In contrast, the prevalence rate for adults living in the 15% most deprived areas (3.8% in 2017/18) has not shown any change since 2008/09, whereas the rate has reduced for those living in the rest of Scotland (from 3.8% to 2.1%).

Following decreases in victimisation over the last decade, there was no difference in the likelihood of being a victim of violence by gender or rurality in 2017/18.

Since 2008/09, the proportion of males and those in urban areas experiencing violence have fallen, whilst the prevalence rates for females and those in rural areas have shown no change. Whilst females and those in rural areas have previously been less likely to have been victims of violence than direct comparator groups, the differing trends in victimisation over time mean that:

• unlike results in previous years, the 2017/18 SCJS did not detect a statistically significant difference in the likelihood of having experienced violence between those in urban and rural areas.
• the survey detected no difference in the proportion of males and females who experienced violent crime in 2017/18 (2.5% and 2.1% respectively) – continuing the finding from 2016/17 which also found no statistically signifcant difference in victimisation by gender^[22].

Figure 3.7 below demonstrates how prevalence rates between these comparator groups have narrowed over the last decade – the apparent differences shown between groups in 2017/18 are not statistically significant, nor are the apparent changes in the prevalence rates for females or those in rural areas since 2008/09.

Figure 3.7: Violent crime victimisation rate by gender and rurality over time

Base: SCJS 2008/09 (16,000); 2009/10 (16,040); 2010/11 (13,010); 2012/13 (12,050); 2014/15 (11,470); 2016/17 (5,570); 2017/18 (5,480) Variable: PREVVIOLENT, TABQDGEN; TABURBRUR.

However, when considering findings by gender, it is important to note that victims of partner abuse may not report such experiences through the face-to-face element of the SCJS which produces the main survey prevalence rates – including on violence. As such, questions on experiences of partner abuse (covering both physical and psychological abuse) are answered in a self-completion element of the survey – with key findings on this topic from 2016/17-2017/18 presented in Section 9.3. Nonetheless, the SCJS main survey estimates have been captured on a consistent basis since 2008/09, meaning trends in victimisation are useful measures of trends experienced by the population.

Looking at more recent changes in victimisation since 2016/17, in line with the national average, there have been no statistically significant changes in the violent crime victimisation rate across any of the sub-population groups discussed above (including age and deprivation).

What can the SCJS tell us about repeat victimisation?

The SCJS estimates that most adults did not experience violent crime in 2017/18, whilst 2.3% of the population were victims of at least one violent crime. However, the survey also enables us to further explore how experiences varied amongst victims and examine the concentration of crime, including what proportion of victims experienced a particular type of crime more than once during the year^[23]. This is known as ‘repeat victimisation’.

Further information about the approach taken to process and derive SCJS results, including on repeat victimisation, is provided in the Technical Report.

Fewer than 1 in every 100 adults were victims of repeated incidents of violence, but their experiences accounted for around three-fifths of violent crime in 2017/18.

Table 3.3 explores the volume of crime experienced by victims in more detail to outline the extent of repeat victimisation and further unpack the concentration of violent crime amonst the adult population. It shows that 1.6% of adults were victims of a single violent incident over the year, with a smaller proportion of the population (0.7%) experiencing repeat victimisation (two or more violent crimes). These repeat victims are estimated to have experienced on average around 3 violent crimes each during 2017/18, whilst together this group of adults are estimated to have experienced almost three-fifths (59%) of all violent crime committed against adults over this period. The table also highlights that an even smaller proportion of the population (0.1%) were high frequency repeat victims who experienced five or more incidents each.

Table 3.3: Proportion of violent crime experienced by repeat victims, by number of crimes experienced (2017/18)

Base: SCJS 2017/18 (5,480). Variable: PREVVIOLENT, INCVIOLENT.

All levels of violent crime victimisation were lower in 2017/18 than in 2008/09, but recent fluctuations mean these findings should be monitored into the future.

Figure 3.8 shows trends in single and repeat violent crime victimisation over time. It highlights that the proportion of adults experiencing only one incident of violence, two or more incidents, and five or more incidents were all lower in 2017/18 than in 2008/09.

Whilst this means repeat violent victimisation (for both 2+ and 5+ incidents) was less prevalent in 2017/18 than a decade ago, findings comparing single years should be interpreted with caution and be considered in context of the broader trend over that period.

Looking more closely, the proportion experiencing two or more violent crimes has been below the 2008/09 baseline (1.6%) since 2010/11 with the exception of 2016/17 when the 1.1% estimate was not significantly different to the 2008/09 figure. The return to a significant decrease comparing 2008/09 and the latest figure suggests that the 2016/17 figure may have been an outlier in an otherwise declining trend^[24].

On the other hand, although such victimisation is consistently very rare, the proportion of adults experiencing five or more violent crimes has shown a less consistent trend and has only been below the 2008/09 baseline figure (0.3%) in 2014/15 (0.2%) and 2017/18 (0.1%). As such it will be important to monitor these findings into the future to see whether the lower victimisation rate seen in 2017/18 is maintained.

All levels of violent crime victimisation have shown no change since 2016/17 – the apparent differences shown in Figure 3.8 are not statistically significant.

Figure 3.8: Proportion of adults experiencing number of violent crimes

Base: SCJS 2008/09 (16,000); 2009/10 (16,040); 2010/11 (13,010); 2012/13 (12,050); 2014/15 (11,470); 2016/17 (5,570); 2017/18 (5,480). Variable: PREVVIOLENT, INCVIOLENT.

Expanding the evidence on repeat violent victimisation

A rapid evidence review of the research on repeat violent victimisation is due to be published by the Scottish Government in Spring 2019. This review will assess the evidence on the extent and prevalence of repeat violent victimisation, the context and circumstances around this, and highlight the evidence on preventing repeat violent victimisation.

What were the characteristics of violent crime?

The majority of violent crime took place in public settings in 2017/18.

When locations are combined into broader categories, the SCJS estimates that just over three-fifths of violent incidents in 2017/18 (62%) occurred in a public setting, with the remainder taking place in a private space^[25]. This proportion is similar to the figures in 2008/09 (67%) and 2016/17 (72%) – the apparent differences are not significant.

Figure 3.8 looks at particular locations more closely and demonstrates that violent crime was experienced in a variety of settings in 2017/18, with the respondent’s place of work the most commonly cited specific location – accounting for almost three in every ten violent crimes (28%).

Figure 3.9: Proportion of violent crime incidents occuring in different locations

Base: Violent crime incidents (120); Variable: QWH1 / QWH3 / QWH5 / QWH7

The proportion of violent incidents estimated to have taken place in or around a pub or bar in 2017/18 was 15%, similar to the results in recent years.

Although a higher proportion of violent crimes took place during the week, the incidence per day was greater at weekends.

Where respondents provided details about when an incident occurred^[26], around three-fifths of violent crimes (59%) happened during the week, with 41% taking place at weekends^[27]. However, taking into account the number of days within each category means that the incidence of violent crime per day was higher at weekends.

What do we know about perpetrators of violent crime?

In 98% of violent incidents reported in 2017/18, respondents were able to provide some information about the offender^[28]. The section below presents headline results on the details provided^[29]. All findings are proportions of cases where respondents were able to say something about the person or people who carried out the offence, unless otherwise stated.

Almost four-in-five violent crimes were committed by male offenders.

The SCJS results highlight that the vast majority of violent crimes in 2017/18 (78%) were carried out by male offenders only – a consistent finding over the years. 15% of incidents involved female offenders only, whilst in 7% of cases both men and women were responsible.

Violent incidents most commonly involved offenders under the age of 40.

Figure 3.10 shows that violent crimes involved people from a range of age groups, but only around one-in-four incidents involved any offenders over the age of 40, suggesting that perpetrators tend to be from younger cohorts^[30].

Figure 3.10: Percentage of violent crime incidents involving offenders of each age group

Base: Violent crime incidents where respondent could say something about offender (120); Variable: QAGE

Violent crimes often involved offenders who victims knew or had seen before.

Most violent incidents (74%) in 2017/18 were committed by people who the victims knew or had seen before. Where offenders were known by the victim, more than two-thirds of incidents (68%) were said to have involved people ‘known well’.

What do we know about the role of alcohol and weapons in violent crime?

The proportion of violent crimes involving offenders under the influence of alcohol is estimated to have fallen from just over three-in-five in 2008/09 to fewer than half in 2017/18.

Offenders were believed to be under the influence of alcohol in 46% of violent incidents where victims were able to say something about the offender in 2017/18. This figure has fallen from 63% in 2008/09 but is not significantly different from the estimate in 2014/15 (56%) or 2016/17 (42%).

This suggests that alcohol has played a less prominent role in violent crime overall in recent years compared to a decade ago – although it remains a factor in a sizeable proportion of incidents.

Figure 3.11^[31]: Proportion of violent crime offenders under the influence of alcohol

Base: Violent crime incidents where respondent could say something about offender (2008/09: 570; 2017/18: 120); Variable: QAL

For wider context on the role of alcohol in violent crime, victims reported having consumed alcohol immediately before the incident in 25% of cases of all violent crime in 2017/18.

Relatedly, victims reported that just over one-in-three violent crimes (36%) involved offenders who were thought to be under the influence of drugs in 2017/18, up from 22% in 2016/17 but unchanged from the 2008/09 figure (29%).

Violent crime in 2017/18 did not commonly involve the presence of weapons.

Victims who said that someone saw or heard what was going on (98% of violent incidents) were asked additional questions about their experience, including the presence of weapons. 12% of such incidents^[32] in 2017/18 were said to have involved perpetrators with weapons, down from 25% in 2010/11 (when the wording of this question was updated to its current format). A knife was reported as being present in 7% of violent incidents where someone saw or heard what was happening.

What was the impact of violent crime?

Two-thirds of violent incidents resulted in injury, although serious injuries were relatively rare.

Where violent crime resulted in some sort of injury (66% of incidents), the most common injuries sustained were minor bruising or a black eye (60%), severe bruising (35%) and scratches or minor cuts (35%). More serious injuries like head injuries and broken bones occurred much less frequently, as shown in Figure 3.12^[33]

Figure 3.12: Type of injuries sustained as a proportion of violent incidents resulting in injury

Base: Violent crime incidents where respondent was injured (70); Variable: QINW

Anger and annoyance were the most common emotional reactions to violent crime.

Consistent with previous years, the emotional impacts most commonly reported by victims of violent crime in 2017/18 were anger (reported in 45% of violent incidents) and annoyance (44%). Victims in just over a quarter of incidents (27%) said they experienced none of the listed emotional impacts, up from 10% in 2008/09 and 8% in 2016/17. In contrast, victims of property crime reported no emotional impact in only 3% of incidents in 2017/18.

However, victims of violent crime were more likely than victims of property crime to report experiencing fear (19% compared to 9%) and anxiety or panic attacks (23% compared to 10%) as a result of their experience.

What proportion of violent crime was reported to the police?

The majority of violent incidents in 2017/18 were not reported to the police, although the reporting rate is no different to previous years.

The 2017/18 SCJS estimates that just under two-fifths of violent incidents (39%) were brought to the attention of the police. The reporting rate in 2017/18 was not significantly different from the rate in any SCJS sweep since 2008/09 (for example, 43% in both 2008/09 and 2016/17) and not significantly different from the reporting rate for property crime in 2017/18 (34%).

There can be a range of factors which influence whether or not an individual reports a crime to the police, not least how the victim views their own experience. For instance, the SCJS finds that in around half of violent incidents in 2017/18 (51%) victims thought their experience should be described as ‘a crime’ as shown in Figure 3.13 below. Just over three-fifths (62%) of incidents which victims considered to be a crime were brought to the attention of the police in 2017/18.

Figure 3.13: Victim's description of violent crime incidents experienced

Base: Violent crime incidents (120); Variable: QCRNO

When asked directly why they did not report their experience, victims cited a range of reasons. The most common explanations provided by victims were that:

• they reported the matter to other authorities (35% of unreported violent crime);
• they dealt with the matter themselves (19%);
• the issue was considered a private, personal or family matter (13%);
• the experience was too trivial or not worth reporting (11%).

Additionally, ‘other’ reasons were cited in 24% of cases.

Where incidents did come to the attention of the police, victims received information or assistance about the investigation and the case (where relevant) from the police in almost three-quarters of instances (72%). Respondents received information or assistance from the Witness Service/Victim Support Scotland in relation to 16% of incidents, from the Crown Office and Procurator Fiscal Service/Victim Information and Advice service in 9%, and from Scottish Courts and Tribunals Service in 7%. Victims said they had not received information or assistance from any organisation in only 6% of cases, whilst a further 15% of incidents were not investigated according to the victim.

What consequences did victims believe offenders should have faced?

Most victims of violent crime did not think the offender should have gone to court, but views varied on what action would have been appropriate.

Regardless of whether the incident was reported to the police, victims in just under two-fifths (39%) of violent crime thought the offender should have been prosecuted in court. This proportion has fallen from 52% in 2008/09, and is lower than the proportion of property crime victims in 2017/18 who thought offenders should have been prosecuted in court (60%).

Just over a fifth (22%) of those who did not think court was appropriate said nothing should have happened to the offender, down from 42% in 2016/17 but similar to results in previous years. A quarter (25%) thought offenders should have been given help to stop them offending, whilst 23% thought they should have been made to apologise for their actions.