Research and analysis

Understanding ethnic disparities in involvement in crime – a limited scope rapid evidence review, by Professor Clifford Stott et al

Updated 28 April 2021

Understanding ethnic disparities in involvement in crime: a limited scope rapid evidence review

Professor Clifford Stott, Dr Matthew Radburn, Dr Arabella Kyprianides and Dr Matthew Muscat

Scope and purpose

In October 2020, we developed a scoping paper designed to assist the Race Disparity Unit (RDU) in focusing its research questions as these relate to the current Commission on racial disparities. On the basis of that paper, the RDU commissioned us to undertake a systematic rapid evidence review of a limited range of published governmental and academic studies of crime and ethnic disparities.

Executive summary

We were asked to address 4 interrelated issues.

Patterns of ethnic disparity in crime based on a review of existing governmental studies

It is evident that disparities in recorded offending in relationship to these categories of behaviours begin with inequalities in relation to police contact and in particular the police use of stop and search powers. It is therefore evident that the relative overrepresentation of BAME people in arrest, prosecution and conviction statistics, particularly as this relates to drug offences, cannot be separated out, or understood independently from, police targeting of areas with high proportions of BAME communities. It is reasonable to conclude that this interrelationship between policing and recorded offending exaggerates the extent to which the ethnic categories are then disproportionately understood to be involved in crime more generally (see Bowling and Phillips, 2007).^{[footnote 1]}

Factors which may be correlated (though not necessarily causative) with people who commit these crimes^{[footnote 2]}

Risk factors are understood to be variables which can usefully predict an increased likelihood of violent crime, drug use, gang involvement, property offences and antisocial behaviour. The extensive body of data and analysis suggests very little if any relationship between ethnic category and involvement in these categories of crime. Therefore, BAME categorisation is not itself a risk factor. Moreover, the research highlights how risk factor-based approaches generally are unhelpful because crime is the outcome of a complex interaction between environmental and personal influences.^{[footnote 3]}

Gaps in existing research and evidence to address the above questions and known data collection and quality issues

The sample of studies and reports are extremely limited in their capacity to examine actual levels of BAME involvement in crime as opposed merely to recorded offending. This
limitation relates to the methodology and data employed, the lack of detailed specificity in the existing datasets, a detailed and contextualised exploration of the victim offender relationship. Data is also largely cross-sectional and correlational, so cannot actually address the underlying causes of crime or explore offending over the life-course, particularly as this relates to the diverse BAME communities of the UK. Datasets in academic studies also tend to lack cross cultural relevance to the UK, particularly as this relates to ethnicity.

What could be developed by government and police force areas to help us better analyse and understand the patterns and drivers of crime among different ethnic groups?

These limitations point to the need for and utility of a relatively large-scale, UK-based, co-produced mixed method study, designed to gather both primary (new) and secondary (existing) data. This might begin with government and police working in partnership with universities to secure funding from United Kingdom Research and Innovation (UKRI) in order to undertake a nationwide comparative study. By using quantitative data, it would be possible to identify a range of representative geographical ‘hotspots’ pertaining to the crimes of interest across a sample of several towns and cities in the UK. Having identified these locations, agreements could be developed with relevant local stakeholders (for example, the police force, relevant local authorities, and NHS) to enable the gathering of primary quantitative and qualitative data in a consistent and comparable way.

Patterns of ethnic disparity in crime based on a review of existing governmental studies

Our analysis of the sample of literature shows that there are demonstrable, quantifiable and robust disparities in Criminal Justice System (CJS) pathways outcomes according to ethnicity. We focus specifically on patterns of ethnic disparity in relation to a) drug offences, b) organised crime groups and county lines, c) violent crime, d) burglary, robbery and theft, e) anti-social behaviour. We then explore how these patterns may be explained in relation to the interrelated stages of a person’s contact with, and journey through, the CJS in terms of policing, courts and sentencing.

Patterns of ethnic disparity in drugs offences (including cannabis and other substances)

The available data suggests that ethnicity is associated with significant disparities within the CJS that are particularly acute for BAME men above 18 years old in relation to drug offences.^{[footnote 4]} The patterns suggest that these emerge primarily at point of arrest, where rates for BAME men are disproportionately high relative to White men (see also below). These disparities ranged from Black men being 5.4 times more likely than White men to be arrested for drug offences, to Asian men being approximately 1.4 times more likely. The disparities in police contact then flow into distinctively different ‘ethnic pathways’ through the CJS.

MOJ analysis^{[footnote 5]} explored the extent of the association between ethnicity and custodial sentencing within specific ‘higher-order offences’, one category of these being drug related.

Their data indicates that in 2015 there were approximately 4,300 offenders convicted for drug-related offences. While 73% of these offenders were White, only 45% of White offenders subsequently went on to be imprisoned, compared with 66% of BAME offenders in the same year. Correspondingly, the BAME imprisonment ratio in this year for these offences was 2.4 – more than double than that for White offenders. Therefore, there is strong evidence of an ‘ethnicity effect’ related not just to arrest but also to imprisonment in relation to drug offences, with BAME offenders more likely to be given custodial sentences than White offenders.

While all BAME men were more likely than White men to be committed to Crown Court for trial, conviction rates for this category of offences were then actually marginally lower than, or proportionate to, White men. However, further analysis by the MOJ^{[footnote 6]} of drug-related offences also demonstrated distinctive disproportionality in sentencing. It is generally the case that custodial sentencing is associated with a variety of factors, such as offender age, ethnicity, offence type and court where the case was heard.^{[footnote 7]} Their analysis found custodial sentencing for all BAME men and Black women committing drug offences was particularly disproportionate, but only at Crown Court. Indeed, this was the only offence category where custodial sentencing was consistently more likely for all BAME men relative to the White group^{[footnote 8]} but also for Black women, who were 2.3 times more likely to receive a custodial sentence for drugs relative to White women.

While the patterns of disparity are relatively clear, the higher-order category ‘drugs offences’ cover a wide range of underlying crimes, in terms of class of drug and type of offence (for example, from possession of cannabis through to wholesale importation, production and supply of class A drugs). It is uncertain whether or not the disparity in rates of imprisonment comes from patterns in different types of underlying offending. To explore this issue, we analysed the literature further. The MOJ reported that approximately a third of prosecutions and convictions of Black people in 2018 were drug related. However, the data also indicated that these figures can largely be attributed to possession of Class B drugs offences (including cannabis), which accounted for nearly half of all drug prosecutions (47%) and drug-related convictions (48%) for Black defendants. Conversely, White defendants made up the largest proportion of people prosecuted and convicted for possession of Class A drugs in 2018 (23% and 24% respectively) compared with Black defendants (17% and 18% respectively). While there are patterns in the types of underlying types of crime, it would appear that inversely White people are more likely to commit more serious drug offences than BAME people.

Accordingly, longer-term trends in the data suggest that the proportion of drug prosecutions where the defendant is White have decreased from 71% in 2014, to 63% in 2018, while there was an increase in the percentage of Black defendants, from 15% to 21% over the same period.

Patterns of ethnic disparity in organised crime groups (OCGs) as this relates to ‘county lines’

There are powerful limitations in the available data and existing analysis of ‘county lines’ offending. This is confirmed by a report from the National Crime Agency (NCA, 2017) which argues that the assessment of this OCG activity across the UK is marred by limitations of police data capture. The report estimates that there are approximately 720 county lines across England and Wales. Their analysis revealed geographic differences in the exporting hubs of county lines. London is identified as the primary exporting hub, with 65% of the UK’s police forces reporting lines into their jurisdiction originating in the capital. Merseyside is identified as the second highest exporter, affecting 42% of other UK police force areas. Due to data limitations the information supplied by the police in relation to the ethnicity of county lines, ‘nominals’ should be treated with caution.^{[footnote 9]} The NCA (2017) report that of those police forces who supplied them with information, ethnicity of suspected ‘nominals’ varied according to geographical location. For example, London ‘nominals’ were reported to be mainly Black. In Liverpool and Manchester, ‘nominals’ were mostly White, and in Birmingham ‘nominals’ were mostly Asian. Somali nationals were referenced by 33% of police forces (with lines predominantly originating in London or Manchester), and Western Balkan Organised Crime Groups were referenced by 9% of police forces.

Patterns of ethnic disparity in violent crime

To address this issue, we first examined the analysis and data relating to possession of Weapons Offences. The number of prosecutions for possession of weapons offences in England and Wales has increased by 5% since 2014, with 13,100 defendants prosecuted in 2018. When compared to 2014, an increase in prosecutions was seen across all ethnic groups, apart from those categorised as White, which saw a decrease of 2% in prosecutions.

In 2018, ethnic minority groups were overrepresented for prosecutions of possession of weapons offences, accounting for 30% of all prosecutions in this category. Of all prosecutions for possession of weapons offences, “possession of an article with a blade or point” made up 59% of prosecutions. The Metropolitan police force (London) area accounted for 66% of all Black defendants prosecuted for this offence, compared with 14% for White defendants.

To this end, the disproportionate prosecutions for this offence nationally can in part be explained by the greater ethnic diversity of London.

Of those sentenced at court, the most common sentence type for possession of weapons offences for all ethnic groups (except offenders of Mixed ethnicity) was immediate custody. In 2018, Black defendants had the highest custody rate at 42%, while the custody rate for all other ethnic groups varied between 31% and 37% Since 2014, Mixed ethnicity offenders consistently had the highest percentage of offenders receiving a sentencing outcome of a community sentence (37% in 2018). Since 2016, Asian offenders had the longest Average Custody Sentence Length (ACSL) for possession of weapons offences. In 2018, the ACSL for possession of weapons offences was highest for Asian offenders at 17.1 months and lowest for Chinese or Other offenders at 8.8 months. The overall ACSL for possession of weapons offences in 2018 was 12.8 months.

We then moved on to address this category of offending with reference to ‘acquisitive violence’. According to the sample of reports, ethnicity is not understood to be associated disproportionately with imprisonment for this category of offending. The MOJ^{[footnote 10]} explored the extent of the association between ethnicity and custodial sentencing within specific higher order offence groups, one of them being acquisitive violence. Around 1,400 offenders convicted for acquisitive violence were examined. The majority of 73% of offenders were White, of which 85% were imprisoned. These percentages were not statistically significantly different for BAME offenders. There is no clear evidence of ethnic differences between White and BAME offenders arrested or convicted of acquisitive violence. However, once again, given the fact that the offence group ‘acquisitive violence’ covers such a wide range of specific offences, that lack of variations in the imprisonment rate could actually be masking underlying variations in the patterns of specific offending. For example, in 2018 to 2019, higher percentages of White and Asian suspects (40%) were arrested for ‘violence against the person’ offences, compared with 35% of Chinese or Other ethnicity suspects, 34% of Mixed ethnicity suspects, and 32% of Black suspects.

In relation to knife crime, a 2018 report entitled ‘Justice Matters: Disproportionality’^{[footnote 11]} references data collected by the Metropolitan Police Service. This work showed that in London in 2017, 50% of knife crime offenders were BAME (up from 44% in 2008). In this total, 50% were under the age of 25 and the majority (90%) were male. 50% of knife crime victims were BAME. A similar pattern emerged when examining knife crime with injury. In 2017, 83% of offenders were male, 35% were aged between 17 to 24, and 69% were BAME. Victims of knife injuries shared a similar profile with offenders. 78% of victims were male, 32% were aged between 17 to 24, and 55% were BAME. Ethnic disparities were also evidenced when looking at knife possession. In 2017, 53% of possession of knife suspects were Black, and 37% of all suspects were Black men under the age of 25. This resonates with the arrest data on stop and search which showed that 56% of all people arrested for offensive weapons following a stop and search were Black.

Patterns of ethnic disparities in robbery and theft

Theft

Theft offences accounted for 19% of total arrests (where ethnicity was known) in 2018 to 2019. There was variation by ethnicity with 20% of Whites, 17% Mixed and Other (including Chinese), 13% Black and 11% of Asian being arrested for theft. These arrests translated into higher percentages of theft convictions that varied in a similar pattern, accounting for 38% of convictions for White offenders, and 28% for Other (including Chinese) offenders, 18% of Black offenders, and 19% of Asian offenders.

Robbery

Disparity in relationship to robbery offences were particularly salient. Young Black men were 10.5 times more likely than young White men to be arrested. However, following arrest, young Black men were significantly less likely to be committed to the Crown Court for trial compared with young White men, and were no more likely to be convicted or receive a custodial sentence. However, they were marginally more likely than young White men to be proceeded against and convicted at a magistrates’ court.^{[footnote 12]} A slightly different pattern was evident for young Mixed ethnicity men, who were 4.2 times more likely than young White men to be arrested for robbery. They were marginally more likely to be proceeded against at a magistrates’ court but no more likely to be convicted or sentenced to custody there compared with young White men. Young Mixed ethnicity men were proportionately likely to be committed to the Crown Court for trial when compared with young White men, but significantly less likely to be convicted.^{[footnote 13]}

Among adults, Black men were about 8.4 times more likely to be arrested for robbery compared with White men.^{[footnote 14]} However, they were less likely than White men to be proceeded against at a magistrates’ court. At Crown Court, not guilty pleas were significantly more likely but custodial remand actually lower for Black men relative to White men. Both conviction rates and custodial sentencing was lower than for White men. Among Mixed ethnicity men, arrest rates were about 5.5 times higher than for the White group.^{[footnote 15]} However, the likelihood of proceeding at a magistrates’ court and of being committed to Crown Court for trial were either less or equal when compared with White men. Conviction rates in the Crown Court were marginally lower for Mixed ethnicity men appearing for robbery, while custodial sentencing was not significantly different to White men. A comparable picture emerged for young Black women, who were 5.1 times more likely to be arrested for robbery compared with young White women.^{[footnote 16]} Young Black women were more likely to be proceeded against at a magistrates’ court but equally as likely as young White women to be convicted.

Patterns of ethnic disparity in anti-social behaviour (ASB)

ASB concerns acts which causes “nuisance or annoyance” in the housing context, or “harassment, alarm, or distress” in public spaces.^{[footnote 17]} ASB encompasses behaviours such as noisy neighbours, vandalism, fly-tipping, littering, street drug dealing, vandalism, graffiti, and public drunkenness.^{[footnote 18]}

There are 2 main ways of measuring the extent of anti-social behaviour in the UK. The Crime Survey of England and Wales (CSEW) provides information about people’s experiences and perceptions of anti-social behaviour. Police data provides information about the incidents they record as such. Both datasets have data quality issues which make it difficult to estimate the actual scale of anti-social behaviour in England and Wales, which is likely to be much higher.

There do appear to be some patterns of ethnic disparity in anti-social behaviour (ASB) in the sample of reports and studies that we studied. Smith’s 2004 academic review of ethnic variations in crime and ASB in England considered whether distinct patterns among ethnic groups have tended to persist from one generation to another.^{[footnote 19]} Their study argues that ethnic disparities in ASB do exist. According to their data, crime and ASB increased among certain categories over time (for example, the African Caribbean ethnic group), but not at all among certain others, most clearly for the Indian ethnic group. Certain other groups (the Bangladeshi group, especially) showed some evidence for an increase in crime and ASB over time. However, it is likely that the precise pattern of local ethnic disparity will vary across location and relate to the demographic makeup of the local population as this relates to age as much as to ethnicity.

Relatedly, a report by an independent educational charity showed data on the percentage of young adults prosecuted for breaching dispersal powers by ethnicity in London.^{[footnote 20]} The study found that Black African offenders aged 18 to 25 were more likely to breach dispersal powers than offenders in the same age group from different ethnic groups (White British, White Other, Black Other, Asian and Asian British).

Disparities in policing

The section above demonstrates consistent patterns of disparity where BAME people tend to be more likely to be arrested, charged and convicted relative to White people for the range of specific crimes focused on in this paper. However, an important issue that needs to be taken into account when seeking an explanation for the evident disparities relates to the ‘street-based’ nature of the crimes under consideration and the prevailing policing practices used in an attempt to control them (for example, stop and search is used heavily to try to prevent, deter and disrupt violent crime, robbery and drugs).

Over the last 11 years there has been a national decline in the overall levels of police stop and search. Between 2009 to 2010, and 2018 to 2019 the annual stop and search rate in England and Wales reduced from 25 to 7 per 1,000 people. However, even within this general pattern there was considerable disparity in relation to ethnicity. For every year in this period, the stop and search rate per 1,000 people was consistently lower for White people compared with the national average. The rates for Asian, Black and Mixed ethnic groups were invariantly higher than the national average across the same time period. Within these BAME categories, people from Black African, Black Caribbean and Other Black groups consistently experienced the highest rates. This long-term trend is concordant with the latest data. For instance, in 2018 to 2019 Black people had the highest stop and search rates in every police force area recorded. Figures suggest that in 2019 to 2020, BAME people were stopped at a rate 4.1 times higher than White people.

Importantly, this data is indicative of disparities in police contact in the form of stop and search, which are then associated with ‘downstream’ differences in patterns of arrest. This data is heavily skewed by patterns in London. For example, in 2018 to 2019, the Metropolitan Police Service made 48% of all stops and searches in England and Wales. In the same time period, 52% of people arrested were BAME, which is an over-representation primarily because of the high proportion of BAME communities in London. Stark patterns of disparity do exist outside London, such as in the Dorset Police area where Black people were 25 times more likely to get stopped and searched compared with White people, and 14 times as likely to be arrested.

Factors which may be correlated (though not necessarily causative) with people who commit these crimes^{[footnote 21]}

Patterns of offending

Although crime has gone down sharply over the last 20 years, some types of violent crime (homicide, knife crime, gun crime and robbery) have gone up since 2014, and across almost all police force areas in England and Wales. While approximately half the increase in robbery, knife crime and gun crime can be attributed to improvements in police data collection, the rest can be largely attributed to drugs and county lines activities.^{[footnote 22]} Between 2014 and 2017, homicides in which the suspect or the victim was known to be dealing or using illicit drugs increased by 7%.

Data has also shown that crack cocaine use is increasing in England and Wales. Crack cocaine markets have a robust connection with serious violence because of its links with county lines, gangs and organised crime groups.^{[footnote 23]} Although recorded serious violence has increased in England and Wales, the trends are mixed in relation to antisocial behaviour. While the Crime Survey for England and Wales (CSEW) showed an increase in antisocial behaviour between 2018 and 2019, police data showed a decrease in antisocial behaviour over the last 10 years.^{[footnote 24]} Unlike the mixed results concerning antisocial behaviour data which shows that property offences constitute the majority of crimes in the CSEW, there has also been a downward trend.^{[footnote 25]}

Bearing in mind these general patterns, we address what the sample of literature tells us about factors that tend to be associated with these crimes.

Definitions

Risk factors are variables which can usefully predict an increased risk or likelihood of violent crime, drug use, gang involvement, property offences and antisocial behaviour.^{[footnote 26]} Protective factors are variables that reduce such likelihoods.^{[footnote 27]} It is important to note that these predictors or correlations are not causal factors, but merely have a tendency in crime and offending records to be associated with the category of offences in question.^{[footnote 28]}

Risk factors for violent crime

The UK government’s Serious Violence Strategy of 2018 defines serious violence as “specific types of crime, such as homicide, knife crime, and gun crime, and areas of criminality where serious violence or its threat is inherent, such as in gangs and county lines drug dealing.”^{[footnote 29]}

This strategy looks at 8 studies^{[footnote 30]} and proposes 5 broad factors of risk as can be seen in Table 1.

Table 1: Risk factors for serious violence according to the UK government’s serious violence strategy 2018

Setting	Risk factors
Individual	Childhood abuse and neglect, impulsivity (low self-control), aggression, low intelligence, substance use, positive attitude towards offending, involved in anti-social behaviour, previously committed offences, low self esteem, gang membership, head injury
Family	Family socioeconomic status, anti-social parents (including substance abuse), poor supervision, parental criminality
School	Low school performance, bullying others, truancy and school exclusion
Community	Urban areas, high crime, local deprivation
Peer group	Delinquent peers

Although these risk factors are based predominantly on US data (and only supplemented by UK data), there is strong evidence supported by several studies of the ‘generalisability’ of these types of risk factors to the UK. The relevant aspects of these reports are summarised in Table 2.

Table 2: Summary of key findings from key reports and studies of risk factors in violence related crimes

Source	Population	Violent crime types	Risk factors identified
Home Office[footnote 31]	UK youths aged 17 to 24	Serious types of ‘violence linked behaviour’ such as weapons carrying or use and gang conflict	Gender, number of siblings in the household, a lack of self-control, early puberty, experience of victimisation, frequency of truanting, bullying, self-harm, risk taking or gambling, feeling isolated, and having previously committed minor violence, theft, public disorder and or cybercrime
College of Policing report[footnote 32]	This report drew on several studies	Violence and or weapon carrying	Gender (being male), age (peaks at the age of 15), adverse childhood experience (including abuse, neglect, parental criminality, substance abuse, being taken into care), educational attainment (school exclusion and low attainment)
Haylock et al[footnote 33]	Youth (10 to 24)	Weapon related violence and crime	Adverse childhood experiences, poor mental health
Sutherland et al[footnote 34]	n/a	Violent crime	Areas of deprivation, presence of transport hubs or major shopping centres or night-time economies

In contrast to the Serious Violence Strategy, where the evidence of a relationship between ethnicity and violence was at best mixed, the Home Office report of 2019 found no association between ethnicity and serious violence related behaviours (for example, carrying of weapons). A 2019 College of Policing report shows that no relationship exists between ethnicity and weapon carrying, but that age and gender (for example, young men, age peaking at 15) along with adverse childhood experiences and low educational attainment, are predictive of weapon carrying and involvement in violent crime. A further review by Haylock et al in 2020 of risk factors associated with weapon-related crime for young people aged 10 to 24 within the UK strengthens both of these reports. They found that adverse childhood experiences and poor mental health were positively correlated with youth and gang violence.

Risk factors for gang involvement

Although we consider risk factors of gang involvement in this paper, it is important to first note that membership of a ‘gang’ itself is not necessarily a crime, and that the data and analysis on these issues is generally taken from associations with other types of offending, such as violent crime and drug use.^{[footnote 35]} Also, the concepts of ‘gangs’ and ‘gang membership’ are problematic. For example, the Metropolitan Police’s ‘gangs’ matrix’ was criticised by Amnesty International for being racially discriminatory, with young Black men being over-represented, and 38% of people on the matrix being judged to pose no risk of committing violence.^{[footnote 36]} Also, the data and analysis is skewed by research from the US, where criminal ‘gang’ cultures are much more salient and deeply-embedded. To address this issue we look at the report prepared by the Home Office and the Early Foundation Initiative.^{[footnote 37]} This report provides an extensive review of several US and UK qualitative and quantitative cross-sectional and longitudinal studies on youth violence and gang involvement (see Table 3). Their analysis found that a range of individual, school and community factors were all associated with gang involvement, but the influence of these different factors varied with age.

Table 3: Risk factors for gang involvement by age group identified by Home Office and the Early Foundation Initiative^{[footnote 38]}

Setting	Risk factor
Individual	Cannabis use, displaced aggression traits and anger traits
School	Low academic achievement in primary school and learning disability
Community	Cannabis use, availability and neighbourhood

Their analysis also identified several protective factors that work against gang involvement (see Table 4).

Table 4: Protective factors against gang involvement identified by Home Office and the Early Foundation Initiative^{[footnote 39]}

Setting	Protective factors
Individual	Belief in the moral order, positive and prosocial attitudes, low impulsivity, intolerant attitude towards deviance, perceived sanctions for transgressions, low ADHD symptoms, low emotional distress and high self-esteem
Family	Good family management, stable family structure, infrequent parent child conflict, supportive relationship with parents or other adults, parents’ positive evaluation of peers
School	Academic achievement, commitment to school, school recognition for involvement in conventional activities, high educational aspirations and bonding to school.
Peers	Friends who participate in conventional behaviour, low peer delinquency, and prosocial bonding.
Community	Low economic deprivation, neighbourhood interaction, and neighbour support

It should be noted that some factors identified for predicting gang involvement are often offences in and of themselves (for example, illegal drug use). The relationship between gang membership and drugs is evidently complex. While several studies have found an association between gang involvement, drug use, sales and violence, these findings are actually based on data which put into serious question the capacity to make any direct causal links.^{[footnote 40]} The literature shows, perhaps unsurprisingly, that gang membership can be considered as a risk factor for increased involvement in violent crimes and illegal drugs. A meta-analysis of 179 empirical studies and 107 independent datasets found a strong relationship between gang membership and various types of offending.^{[footnote 41]}

Risk factors for drug use

Stone et al. (2012) conducted a comprehensive review of the literature that identified several risk factors for, and protective factors of, illegal drug use in young adulthood (aged between 18 to 26) (see Table 5 below).^{[footnote 42]} This research suggests that drug use leads to involvement in criminal behaviour due to:

• the psychopharmacological properties of drugs
• the economic motivations to obtain drugs
• the systemic violence associated with the illegal drug market.^{[footnote 43]}

Perhaps unsurprisingly, the risk and protective factors for drug use overlap with those for violent crime and gang involvement outlined above.

Table 5: Risk factors for drug use

Type	Risk factors
Fixed factors	Gender (male), race and ethnicity, prenatal alcohol abuse, parental substance abuse history, parental depression, neighbourhood instability
Contextual factors	Social norms and alcohol availability
Individual factors	History of abuse or neglect, poor family relationships, family management, internalizing or externalizing behaviour, favourable attitudes towards drug use, living situation, job status, college attendance, peer relations, belief in conformity, religious involvement, level of education, becoming pregnant, marriage or committed relationship

Risk factors for property crime (excluding robbery)^{[footnote 44]}

It is widely understood that in the UK and elsewhere, the majority of burglaries are committed by drug users engaging in property crime to support their addictions. Several studies have shown that the drugs mostly associated with acquisitive crime include heroin, crack cocaine and methamphetamine.^{[footnote 45]} Evidence also suggests that some offenders use illegal drugs in order to facilitate their involvement in property crime. For example, one study^{[footnote 46]} found that professional burglars used illegal drugs when committing offences to reduce their anxiety and remain vigilant. Studies have shown that those who engage in property crime make rational decisions to commit the offence. For instance, Bennett and Wright’s^{[footnote 47]} 1984 study of imprisoned professional burglars in southern England showed that most of their burglaries were planned. Burglars invariably make rational decisions based on target suitability, and various situational risk factors have been identified in the literature. For example, 2 studies^{[footnote 48]} showed that burglars select the most vulnerable targets based on aspects such as occupancy, wealth, layout, and security (see Table 6 below).

Table 6: The risk factors predictive of targeted properties

Type	Risk factors
Occupancy cues	Cars in driveway, lights, presence of mail, burglar alarms, dogs (irrespective of size) but not cats
Wealth cues	Appearance of residence and neighbourhood, landscaping quality and type of car driven
Layout or surveillance cues	Amount of cover or openness, neighbouring houses and rear access
Security cues	Alarms, window locks and deadbolt locks

Research that has focused on property crime prevention is based primarily on addressing situational factors. For example, Welsh and Farrington’s meta-analysis from 2009^{[footnote 49]} showed that CCTV cameras have a modest yet significant impact on crime reduction when compared with control areas.^{[footnote 50][footnote 51]} Evidence exists to show that the installation of electronic immobilisers and improvements in window and door locks contributed to declines in vehicle and residential theft, respectively.^{[footnote 52]} It is widely known that offenders tend to commit crime near to where they live, and areas with higher levels of car theft are those where vehicles tend to be older and less secure. In this sense, regarding property crime, apart from the key issue of drug addiction, the main risk factors arising from research relate more to situational opportunities and affordances than they do to factors relating the characteristics of the offenders involved.

Risk factors for anti-social behaviour (ASB)

Perhaps the best source of existing evidence and analysis on this issue is the extensive literature review of conduct disorder^{[footnote 53]} by Farrington (2005) that identified several early risk factors for ASB (see Table 7).^{[footnote 54]} These were corroborated by a literature review conducted by Fitch (2009) and by a meta-analysis conducted by Murray and colleagues (2012).^{[footnote 55][footnote 56]} On the basis of these reviews, it appears that a range of different factors are associated with the likelihood of engaging in ASB. While these studies appear to focus on experiences at school, relationships with family and peers, and substance use, it should be noted that these variables are also clearly framed by factors of economic deprivation.

Table 7: A summary of the key risk factors the available research identifies as associated with antisocial behaviour

Researchers	Risk factors identified
Farrington (2005)	Impulsiveness, low intelligence and low school achievement, poor parental supervision, child physical abuse, punitive or erratic parental discipline, cold parental attitude, parental conflict, disrupted families, antisocial parents, large family size, low family income, antisocial peers, high delinquency-rate schools, and high-crime neighbourhoods
Fitch (2009)	Physical abuse, school exclusion, poverty, lack of positive-role models, family criminality, and drug or alcohol abuse
Murray et al (2009)	Parental imprisonment (suggestive of antisocial parents and a lack of positive role models)

Summary

What is perhaps most powerfully relevant about the research on risk factors is that this extensive body of data and analysis suggests very little, if any, relationship between ethnic group and involvement in these types of crime. From this brief review it is possible to argue that a significant overlap exists between the identified risk factors. We would argue there are at least 2 important reasons for this.

Firstly, all these studies essentially use the same datasets and other studies which are then based on each other. For example, the reports focused on risk factors for violent crime referenced other studies that set out risk factors for youth violence and gang membership.

Secondly, this problem is exacerbated by the fact that the bulk of the UK reports are all ultimately based on the same interrelated datasets provided by the government, and obtained from stakeholders largely through statutory reporting requirements.

Also, offenders can and do engage in a wide range of crimes often explained theoretically by the inter-relationships between several risk factors. For example, Wilson, Stover and Berkowitz’s (2009) meta-analysis of several studies found a relationship between exposure to violence and future antisocial behaviour.^{[footnote 57]} Far from distinct behaviours, offending is actually a complex arrangement of behaviours that cannot be understood through single-factor explanations.^{[footnote 58]}

What can be observed from these studies is a pattern that highlights how a series of interrelated factors appear to be able to predict broad patterns of offending to a reasonable level. The research identifies a series of individual and family level factors but exposes how these are interlinked with factors linked to economic deprivation and the community, social and individual harms that flow from that (for example, neighbourhood instability, job status, levels of education). For example, a lack of self-control, experience of victimisation, frequency of truanting are factors associated with adverse childhood experience (including abuse, neglect, parental criminality, substance abuse, being taken into care), poor educational attainment and school exclusion. In turn, these factors are all far more likely among communities in areas of socio-economic deprivation relative to areas of wealth.^{[footnote 59]}

Gaps in existing research and evidence to address known data collection and quality issues

As we argue above, a big problem with existing studies and reports is that they are essentially all based on the same datasets. While illustrative, these reports and studies have important and fundamental limitations. We summarise these below.

The methodology and data employed

Governmental and other administrative reports tend to be based on the same data. In turn, they rely on decontextualised statistics based on fragmented data taken from multiple agencies and organisations. While they can demonstrate broad patterns of disparities in CJS outcomes in relation to ethnicity, such aggregation cannot meaningfully be used to explore why these patterns exist. Indeed, the personal histories and perspectives of those who are drawn into the CJS are conspicuous by their absence. Such data tells us very little about the actual underlying levels of crime, given that the majority of offending goes unreported. Another issue relating to methodology is the fact that most of the research is correlational, so causal relationships cannot be deduced with certainty. There is often somewhat of a tautological relationship between ‘risk factors’ and associated behaviours. For example, ‘gangs’ are often identified as a risk factor’ for serious violence, yet serious violence offences are often seen as a predictor of ‘gang membership’.

The lack of detailed specificity in the existing datasets

Given limitations in the underlying data set, the majority of studies and reports that focus on ethnicity and crime use broad ethnic categorisations and do not tend to include a fine-grained analysis according to geographical location. They almost invariantly dedicate their analysis to patterns in England and Wales, and therefore preclude comparisons with Scotland and Northern Ireland. Those that do compare regions tend to do so by comparing London to the rest of England or the UK. It is difficult to ascertain patterns of disparity in relation to age since the reports tend to present data merely on those above and below 18 years of age. This lack of capacity to undertake fine-grained analysis is a major problem that cannot be easily overcome. For example, it is difficult to understand the localised drivers of crime as these relate to differing patterns of ethnicity both across and within the urban centres of the UK. The dominance of data from London often has the capacity to skew the national picture.

The victim and offender relationship

Governmental reports and other research which uses CJS data tends to focus either on the offenders or victims of crimes. This precludes an in-depth exploration of the complex overlap of and inter-relationships between these categories, in that offenders are also often victims of crime and vice versa. It also inhibits an analysis of how the relationship between victims and offenders may differ according to ethnicity, crime type and context.

A notable exception to this is the MoJ’s Statistics on Race and the Criminal Justice System report in 2018 which analysed homicides in England and Wales. The MoJ concluded that the association between homicide victim and suspect did vary according to ethnicity. For Asian and Other (including Chinese) victims, the principal suspect was more likely to be a family member (18% and 16% respectively) relative to White or Black victims (8% and 7% respectively). Asian victims had a higher proportion of cases where the principal suspect was a partner or ex-partner (19%) relative to Other (including Chinese), White and Black victims (14%, 14% and 6% respectively). Black victims had the highest percentage of homicides where the principal suspect is a stranger (35%) relative to 21% of White victims, and 26% of Asian and Other (including Chinese) victims. Understanding why such patterns exist is important but impossible if the focus of analysis is on victims or offenders as different populations.

BAME involvement in crime

The reliance on summary statistics, such as arrest figures, can present a misleading picture. For example, Home Office data in 2018 showed that in England and Wales only 8.2% of crimes recorded by the police resulted in a suspect being charged or court summoned. In 45.7% of offences, no suspect was identified at all.^{[footnote 60]} Measures such as arrest rates, as well as those prosecuted and convicted, can only give a limited and very partial picture of the overall patterns of crime and how these relate to ethnicity. The decontextualised figures supplied in many of the government-mandated annual or biannual statistical bulletins perhaps tell us more about disproportionate police practices (for example, use of stop and search) and potential disparities in the criminal justice system than they can ever reveal about genuine underlying variations in involvement in actual crime.

Disparities are intersectional and not restricted merely to ethnicity

Government reports tend to be limited in their exploration of the intersectionality of factors that combine to produce the patterns of disparity relating to CJS outcomes.

For example, the latest bulletin warns that: “It is important to note that for the majority of the report, no controls have been applied for other characteristics of ethnic groups (such as average income, geography, offence mix or offender history), so it is not possible to determine what proportion of differences identified in this report are directly attributable to ethnicity. It is not possible to make any causal links between ethnicity and CJS outcomes” (MoJ, 2019,^{[footnote 61]} page 6). Consequently, without simultaneously taking into account a wider range of factors, any analysis of how ethnicity relates to differential involvement of crime will be at best incomplete, and at worst dangerously misleading.

The academic literature reviewed in Section 2 has highlighted associations or ‘risk factors’ in relation to the likelihood of a person committing specific types of crime. However, like the governmental reports, academic studies also have important limitations, including:

Cross-cultural relevance

In comparison with other jurisdictions (such as the US), few studies specifically examine violence, gangs, drugs, property crime and antisocial behaviour in the UK context. The extent to which these findings can be applied to guide UK policies and practices is often uncertain.

Correlations and not the causes of crime

The current evidence base indicates the important ‘risk factors’ associated with committing specific crimes. Though as argued above, this does not mean that these factors are causative. For instance, a person may possess all the ‘risk factors’ identified for violent crime (for example, childhood abuse and neglect) and never commit a violent offence.^{[footnote 62]}

Offending over the life-course

In comparison with other jurisdictions such as the US, there are few UK-based studies that examine offending over the life-course of an individual. The most influential longitudinal study in the UK is Farrington’s Cambridge Study on Delinquent Development. This study, however, consists exclusively of men, and most of the men (87%) are white British.^{[footnote 63]} It is not possible to link its findings to other ethnic groups and to other genders.

The methodology employed

As we suggested in relation to the governmental reports above, there are relatively few academic studies in the UK that use a methodological approach that seeks to forefront the experiences and context of those who go through the CJS. Consequently, they are forced to focus on general patterns, and trends are often unable to shed light on exactly how or why ethnicity feeds into disparities in policing and criminal justice responses. In order to understand the drivers of crime, criminological research should seek to move beyond a risk-factor based approach that produces descriptive lists of the typical characteristics or circumstances of people who commit certain crimes. Instead, new research should be developed that would allow for the underlying drivers of crime and disparity within CJS in the UK.

What could the government and police force areas do to help us better analyse and understand the patterns and drivers of crime among different ethnic groups?

As we point out above, an issue is the reliance on data at the point where the criminal justice pathway begins, from point of contact with the police onwards. An alternative is to also study victim survey data, but taken in isolation this has powerful limitations, particularly when it comes to understanding ethnic disparities. Accordingly, an approach based on a range of methodologies allows for data triangulation where the weaknesses inherent in some datasets are offset by the strengths of others.

For example, CJS data can be used to assess the broad overall patterns of criminality and how this relates to ethnicity, but this data does not tell us why these patterns arise. Observations of police-public interactions and in-depth interviews can be used in conjunction with the statistics to help explain why the broad patterns exist.

In addition, the interrelated problems identified in the previous section revolved around:

• the need for systematic and standardised data capture by police forces and other stakeholders as this relates to crime and levels of offending
• an overreliance on summary CJS statistics
• a lack of capacity for fine-grained analysis (for example, patterns of offences by geographical area
• the need for UK-based studies

All these limitations point to the utility of a relatively large-scale, UK-wide, mixed-method study designed to gather both primary (new) and secondary (existing) data. Such a study might take around 3 years and begin by using quantitative data to identify a range of geographical ‘hotspots’ pertaining to the crimes of interest across a sample of several towns and cities in the UK. Having identified these research relevant geographical locations, there would need to be agreements reached between the research team and the relevant local stakeholders (for example, data sharing agreements with and between the local police force, relevant local authorities, and NHS) in order to allow the different stakeholders and the research team to systematically gather primary quantitative and qualitative data in a consistent and comparable way.

Data sources might include, but would not be limited to:

• police contact and use of force data, including logs from call handling centres and geographical deployment of officers and their activity (including stop and search data)
• localised socio-economic, health, and crime data
• hospital admissions and school exclusion data
• footage recorded by CCTV or police body-worn cameras
• direct observations of police-public interactions (for example, the use of stop and search powers)

In addition, it would be important to gain an understanding of both general experiences and details of a range of specific offences. This data could be obtained through the development of public surveys, where the data is appropriate to the localities under study, including local public perception surveys focused on specific offence types. It could also involve in-depth interviews with victims and perpetrators of crime, as well as police officers and other relevant agencies and stakeholders.

In order to explore the relationship and relative importance of the factors identified in the previous section, we recommend:

• conducting more randomised control trials and experiments in the UK context, as these research methods are capable of manipulating variables and help to attribute cause and effect (although this would be a longer-term goal)
• incorporating more ethnically-diverse samples when using quantitative methods
• conducting other major longitudinal studies of offending development in the UK with more ethnically and gender-diverse samples

Apart from utilising more quantitative research methods to examine drivers of crime, it is crucial to supplement these with qualitative methods. This is for 2 main reasons.

First, quantitative methods tend to give an incomplete picture of the drivers of crime. Second, minority ethnic groups and other marginalised groups may not be willing or able to engage with quantitative research methods (for example, they may not have access to a computer, or may not trust the authorities). We therefore suggest conducting in the UK context more, for example, ethnographic research in the style of Anderson’s (1999)^{[footnote 64]} Code of the Streets and Goffman’s (2014)^{[footnote 65]} On the Run, and using micro historical case studies as conducted by Ball et al (2019).^{[footnote 66]}

We suggest conducting more research involving victims of crime, not only because victims tend to be sidelined in the criminal justice process but also because offenders and victims tend to share similar profiles. By understanding why victims and offenders share similar profiles it is possible to gain a better understanding of the causes of crime.

Considering patterns of migration and settlement, as well as the demographic and socio-economic profiles of ethnic groups in England and Wales, is also important when conducting future analysis of official data.^{[footnote 67]} This would help to contextualise patterns of crime among different ethnic groups.

Appendix 1: Trust and its impact on crime

It is important to note that while we did not identify trust as a risk (or protective) factor for the crimes of interest, it is clear that a lack of trust is pervasive in the UK’s criminal justice system. As noted in the Lammy Review: “[t]his lack of trust starts with policing, but has ripple effects throughout the system, from plea decisions to behaviour in prisons.”

Trust is a social glue and lubricant which makes cooperation between individuals easier.^{[footnote 68]} A lack of trust can have a threshold effect in that too much distrust can result in mutual suspicion and hostility.^{[footnote 69]} We will reference a blend of UK, US, and European-based studies to examine the association between trust and offending in 2 areas of the criminal justice system: police and prisons.

Beginning with policing, Harcourt’s 2006 study in the US found that many interviewees carry weapons because they have limited confidence in the police to protect them from violence.^{[footnote 70]}

In the UK, however, mixed support for this explanation has been found.^{[footnote 71]} Linked to trust in the ability of police to protect individuals from violence is trust in the ability of police to performing their functions, and 2 UK studies are highly relevant. First, Jackson, et al (2012) tested a revised version of Tyler’s procedural justice model among a sample of 937 adults in England and Wales in the policing context.^{[footnote 72]} They found that trust in procedural fairness did not predict obligation to obey the police but predicted moral alignment. While moral alignment reduced offending behaviour, obligation to obey did not predict offending behaviour. In their model, perceived risk of sanction did not reduce offending behaviour. Second, Hough et al (2013) also tested a revised version of Tyler’s process-based model among a sample of 52,041 interviewees from the European Social Survey.^{[footnote 73]} Trust in procedural fairness, effectiveness, and in distributive fairness were all significant predictors of obligation to obey, moral alignment and legality, albeit with varying levels of significance.

They found that legality, deterrence, and moral alignment demonstrated significant and negative effects on offending behaviour, with obligation to obey showing no significant effect. These studies demonstrate that trust can impact on offending through the mediating variables of legality and moral alignment.

Other factors (apart from trust) are important in explaining crime. Legitimacy is one such factor.^{[footnote 74]} Tankebe tested a revised multidimensional model of Tylerian legitimacy among a sample of 5,120 London residents in the policing context.^{[footnote 75]}

He found that:

• legitimacy (as an aggregated scale) was a significant predictor of cooperation with the police
• procedural justice and distributive justice were significant predictors of cooperation with the police
• lawfulness was an important predictor of cooperation with the police
• perceived police effectiveness reduced cooperation with the police
• obligation to obey mediated the relationship between the aggregated legitimacy scale and the individual components of legitimacy

While this study did not focus on offending behaviour, it is reasonable to hypothesise that a lack of legitimacy and result in uncooperative behaviour. This is of utmost importance as police depend on the public’s cooperation to detect and solve crimes.

Turning to the prison context, trust has been identified as one of the aspects of prison life that matters most to prisoners.^{[footnote 76]} Prisons are already low-trust environments but trust in prison officers by prisoners, and trust in prisoners by prison officers can result in an orderly prison environment.^{[footnote 77]} While too little trust can negatively impact order in prisons, too much trust can also have a similar effect. When relationships between prisoners and prison officers are too close, too informal and lacking boundaries, it can lead to prison officers engaging in acts of corruption.^{[footnote 78]}

Appendix 2: Desistance

It is understood that reoffending is a major problem, and this is reflected in governmental statistics. For example, the MoJ’s most recent ‘Proven reoffending statistics’ bulletin covering adult and juvenile offenders showed that of those who were either released from custody, received a non-custodial conviction at court, or received a caution between October and December 2018, over a quarter (28.1%) went on to reoffend. Indeed, of adults released from custodial sentences of less than 12 months, nearly two-thirds (61%) had a proven reoffending rate. As shown in table X, it is important to recognise that in absolute terms by far the largest number of reoffenders are White. White reoffenders also consistently had the highest average number of reoffences. However, in relative terms the data shows that reoffending rates remained consistent across all ethnic groups between 2006 to 2007 and 2016 to 2017. Black offenders had the highest proportionate rates of reoffending for this period, offenders from the Other ethnic groups had the lowest rates.

Despite the fact that reoffending is a major problem, it is widely understood in academic literature that even the most persistent and prolific offenders can and generally do eventually desist from crime.^{[footnote 79]} It is relevant to focus on some important academic studies in this field because of what they further expose in terms of the situational drivers of crime.

We can reference 2 major and widely-cited academic studies on desistance, along with a report by HM Inspectorate of Prison (2016), and data from GOV.UK. Sampson and Laub (2017) analysed data from the USA gathered during a 3-wave longitudinal study of 1,000 ‘delinquents and non-delinquents’ matched on age, ethnicity, IQ, and low-income in Boston.

They analysed data at 3 points, when the individuals were 14, 25 and 32 years old. Their analysis was used to support the argument that desistance was not merely due to ageing and maturing character but related to 4 “turning points” that helped previous offenders desist. These were military service, marriage, employment and neighbourhood change. They argue these turning points helped offenders desist from crime because they changed the surrounding context for the individual by removing proximate opportunities for crime, created new social bonds, enabled new non-criminal activities, and provided a basis for identity transformation.

The second major study on desistance is a UK-based analysis known as the Sheffield Pathways out of Crime Study (SPOOCS). The SPOOCS is a longitudinal analysis of more than 100 persistent young adult offenders that was conducted between 2006 and 2007.

This study combined quantitative and qualitative methods to obtain an understanding of the processes of desistance among a sample of people who had begun offending in early adulthood.^{[footnote 80]}

The SPOOCS was distinctive in that it explored the early stages of desistance in a sample of mostly persistent offenders, and highlighted both the precariousness and the sense of struggle involved.1 This study showed that reoffending among this sample was high. For example, during the 3 follow-up years, 80% of the sample reoffended, and in the self-report section several individuals who were not convicted reported actually reoffending. One of the strongest predictors of reduction in offending was the perceived number of obstacles to desistance. These included continued drug use and lack of employment, combined with the opportunity’s crime afforded to make easy money and gain excitement.^{[footnote 81]} As with Sampson and Laub (2017) they also found that desistance was enabled through largely situational changes obtained through gainful employment, along with the absence of otherwise criminal peers.^{[footnote 82]}

An evaluation of the effectiveness of Youth Offender Teams identified factors that helped in the process of desistance, as well as factors that acted as barriers to desistance. The most important factors that helped young offenders to desist were:

• a balanced, trusting and consistent working relationship with at least one worker
• meaningful personal relationships and sense of belonging to family
• emotional support, practical help and where the worker clearly believed that the young offenders had the capacity to desist from offending
• changing peer and friendship groups
• problem solving interventions
• restorative justice interventions which are well planned^{[footnote 83]}

In contrast, the factors which acted as barriers to the process of desistance were:

• formal offending behavioural programmes not meeting individual needs
• poor relationships with, and frequent changes of, case managers
• identified needs not being addressed
• a lack of genuine involvement with their case manager in planning for work to reduce reoffending
• adopting a ‘one size fits all’ approach^{[footnote 84]}

This report complements and adds to the findings derived from the Sampson and Laub study, as well as the SPOOCS study.

Appendix 3: Relative rate index for BAME men relative to White men for drug offences in 2014

Ethnic category	Arrests rate	Proceeded Magistrate	Convicted Magistrate	Custodial Magistrate	Crown Court trial	Remand by Crown Court	Not guilty plea Crown	Convicted Crown	Custodial Crown
Black	5.4*	1.2*	0.9	0.7*	1.2*	1.8*	2.2*	1.0*	1.4*
Asian	1.4*	1.1*	1.1	1.1	1.4*	1.6*	1.7*	1.0*	1.4*
Mixed	3.2*	1.1*	1.0	0.7	1.1*	1.4*	1.6*	1.0	1.1*
Other	1.7*	1.0*	1.1	1.4	1.4*	2.2*	1.8*	1.0*	1.6*

Note: * indicates a statistically significant difference

Appendix 4: Different types of offenders

The academic literature of risk factors refers to 3 broad types of offenders:

Adolescent Limited (AL) Offenders: These are individuals who engage in minor offending or anti-social behaviour into their 20s.^{[footnote 85]} The main causes for AL offenders are thought to be delinquent peers and a disjunction between maturations and responsibilities. These offenders tend to naturally grow out of criminality after realising that more serious crimes can impede future job opportunities.

Life-Course Persistent (LCP) Offenders: In contrast to AL offenders, LCP offenders start offending in early in life and do not desist throughout their life-course, and often engage in violent behaviour.^{[footnote 86]} The main causes for LCP offenders are thought to be poor attention and hyperactivity (in early childhood), as well as family and societal disadvantages.^{[footnote 87]}

Late-Onset (LO) Offenders: LO offenders in contrast to LCP offenders seem to begin offending later on life, usually from the age of 21 onwards.^{[footnote 88]} There appear to be 2 groups of LO offenders.^{[footnote 89]} First, those who are LO because their levels of self-reported criminality extended over a long period of time and then increasing their level of offending in adulthood and who were then convicted.^{[footnote 90]} Second, those who self reported high levels of criminality in their youth but were either lucky or skilled enough to avoid a conviction until adulthood.^{[footnote 91]}

Footnotes

1. Bowling, B. and Phillips, C., 2007. Disproportionate and discriminatory: Reviewing the evidence on police stop and search. The Modern Law Review, 70(6), pp.936-961. ↩

2. Trust and desistance issues are dealt with in detail in Appendix 1 and 2. ↩

3. https://crimesciencejournal.biomedcentral.com/articles/10.1186/s40163-020-00132-7 ↩

4. See Appendix 3 ↩

5. MOJ (2015): Associations between ethnic background and being sentenced to prison in the Crown Court in England and Wales in 2015. ↩

6. MOJ (2016) ‘Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales’ ↩

7. Hopkins, K. (2015). Associations between police-recorded ethnic background and being sentenced to prison in England and Wales. Ministry of Justice, available online; Hopkins, K., Uhrig, N., & Colahan, M. (2016). Associations between ethnic background and being sentenced to prison in the Crown Court in England and Wales in 2015. Ministry of Justice, available online. ↩

8. See Appendix 3. ↩

9. It is important that the NCA (2017) report does not provide a definition of a ‘nominal’. It is therefore unclear as to whether this term refers to those suspected and/or convicted of county lines offences, which contributes to the ambiguity of the findings reported. ↩

10. MOJ (2015): Associations between ethnic background and being sentenced to prison in the Crown Court in England and Wales. ↩

11. Mayor of London Office of Policing and Crime (2018). Justice Matters: Disproportionality. ↩

12. Ministry of Justice (2016). Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales, table A2.1 in Appendix 2. ↩

13. Ministry of Justice (2016). Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales, table A2.3 in Appendix 2. ↩

14. Ministry of Justice (2016). Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales, table A2.9 in Appendix 2. ↩

15. Ministry of Justice (2016). Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales, table A2.11 in Appendix 2. ↩

16. Ministry of Justice (2016). Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales, table A2.5 in Appendix 2. ↩

17. s2(1)(a), Anti-social Behaviour, Crime and Policing Act 2014. ↩

18. Brown, J. and Sturge, G. (2020). Tackling Anti-Social Behaviour. House of Commons Library. ↩

19. Smith (2004). Explaining ethnic variations in crime and antisocial behavior in the United Kingdom. Ethnicity and Causal Mechanisms. Cambridge University Press. ↩

20. Mills & Ford (2018). Anti-social behaviour powers and young adults. Centre for Crime and Justice Studies. ↩

21. We provide a more detailed analysis of drivers of crime and issues of trust in Appendix 1. The data and analysis relating to desistance from crime is limited, particularly with regard to government and public sector reports, and tells us little if anything about ethnic disparities. Consequently, we provide analysis of this issue in Appendix 2. ↩

22. Ibid.; HM Government (2018). A whole system multi agency approach to serious violence prevention: A resource for local system leaders in England. ↩

23. Ibid.; National Crime Agency (2017). County lines Violence, Exploitation & Drug Supply; Bartol, C. R. and Bartol, A. M. (2011). Criminal Behaviour: A Psychological Approach. Upper Saddle River, NJ: Pearson Education, Inc. ↩

24. Brown, J. and Sturge, G. (2020). Tackling Anti-Social Behaviour. ↩

25. https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/focusonpropertycrime/yearendingmarch2016 ↩

26. Home Office and Early Intervention Foundation (2015). Preventing gang and youth violence: a review of the risk and protective factors. ↩

27. Ibid. ↩

28. Ibid. ↩

29. HM Government (2018). Serious Violence Strategy. ↩

30. Raby, C., & Jones, F. (2016). Identifying risks for male street gang affiliation: a systematic review and narrative synthesis. Journal of Forensic Psychiatry & Psychology, 27(5), 601-644; O’Brien, K., Daffern, M., Chu, C. M., & Thomas, S. D. (2013). Youth gang affiliation, violence, and criminal activities: A review of motivational, risk, and protective factors. Aggression and Violent Behaviour, 18, 417-425; Brennan, I. R., & Moore, S. C. (2009). Weapons and violence: A review of theory and research. Aggression and Violent Behavior, 14, 215-225; Gerard, J. F., Jacson, V., Chou, S., Whitfield, K. C., & Browne, K. D. (2014). An exploration of the current knowledge on young people who kill: A systematic review. Aggression and Violent Behavior, 19, 559-571; Farrington, D. P., Loeber, R., & Berg, M. T. (2012). Young Men Who Kill: A Prospective Longitudinal Examination from Childhood. Homicide Studies, 16 (2), 99-128; McVie, S. (2010). Gang Membership and Knife Carrying: Findings from the Edinburgh Study of Youth Transitions and Crime. Edinburgh: Scottish Government Social Research; Early Intervention Foundation & Cordris Bright Consulting (2015). Preventing Gang and Youth Violence. ↩

31. Home Office Report (2019). An analysis of indicators of serious violence: Findings from the Millennium Cohort Study and the Environmental Risk (E-Risk) Longitudinal Twin Study 2019. ↩

32. College of Policing Report (2019) - Knife crime evidence briefing 2019. ↩

33. Haylock, S., Boshari, T., Alexander, E. C., Kumar, A., Manikam, L., & Pinder, R. (2020). Risk factors associated with knife crime in United Kingdom among young people aged 10-24 years: A systematic review. ↩

34. Sutherland, A., Brunton-Smith, I., Hutt, O., and Bradford, B. (2020). Violent crime in London: trends, trajectories and neighbourhoods. ↩

35. For example, the meta-analysis by Pyrooz et al. (2016) of 179 empirical studies and 107 independent data explored the relationship between gang membership and offending and found that there is a fairly strong relationship between gang membership and offending. Pyrooz, D. C., Turanovic, J. J., Decker, S. H., and Wu, J. (2016). Taking stock of the relationship between gang membership and offending: A meta-analysis. Criminal Justice and Behaviour Vol. 43(3): 365-397. ↩

36. https://www.theguardian.com/uk-news/2018/dec/21/metropolitan-police-gangs-matrix-review-london-mayor-discriminatory ↩

37. Home Office and Early Intervention Foundation (2015). Preventing gang and youth violence: a review of the risk and protective factors. ↩

38. Ibid. ↩

39. Ibid. ↩

40. See Bjerregaard, B. (2010). Gang membership and drug involvement: Untangling the complex relationship. Crime & Delinquency, 56(1), 3-34. ↩

41. Pyrooz, David C., Jillian J. Turanovic, Scott H. Decker, and Jun Wu. “Taking stock of the relationship between gang membership and offending: A meta-analysis.” Criminal Justice and Behavior 43, no. 3 (2016): 365-397. ↩

42. Stone, A. L., Becker, L. G., Huber, A. M., & Catalano, R. F. (2012). Review of risk and protective factors of substance use and problem use in emerging adulthood. Addictive behaviors, 37(7), 747-775. ↩

43. Goldstein, P. J. (1985). The drugs/violence nexus: A tripartite conceptual framework. Journal of drug issues, 15(4), 493-506. ↩

44. Since robbery is an offence which involves theft with violence or threat of violence, its risk factors have been included in the section on violent crime. ↩

45. Goldsmid, S., & Willis, M. (2016). Methamphetamine use and acquisitive crime: Evidence of a relationship. Trends and Issues in Crime and Criminal Justice, (516), 1. Pierce, M., Hayhurst, K., Bird, S. M., Hickman, M., Seddon, T., Dunn, G., & Millar, T. (2017). Insights into the link between drug use and criminality: Lifetime offending of criminally active opiate users. Drug and alcohol dependence, 179, 309-316. Stewart, D., Gossop, M., Marsden, J., & Rolfe, A. (2000). Drug misuse and acquisitive crime among clients recruited to the National Treatment Outcome Research Study (NTORS). Criminal behaviour and mental health, 10(1), 10-20. Parker, H., & Newcombe, R. (1987). Heroin use and acquisitive crime in an English community. British Journal of Sociology, 331-350. ↩

46. Cromwell, P. F., Olson, J.F. and Avary, D. W. (1991). Breaking and entering: an ethnographic analysis of burglary. Newbury Park, CA: Sage; Santa Clara Criminal Justice Pilot Project (1972). Burglary in San Jose. Springfield, VA: U.S. Department of Commerce. ↩

47. Bennett, T., and Wright, R. (1984). Burglars on Burglary: Prevention and the offender. Brookfield, VT: Gower. ↩

48. Ibid; Nee, C., and Taylor, M. (1988). Residential burglary in the Republic of Ireland: A situational perspective. Howard Journal, 27: 105-116. ↩

49. Welsh, B. C., & Farrington, D. P. (2009). Public area CCTV and crime prevention: an updated systematic review and meta analysis. ↩

50. Welsh, B. C., & Farrington, D. P. (2004). Evidence-based crime prevention: The effectiveness of CCTV. Crime Prevention and Community Safety, 6(2), 21-33. ↩

51. It should be emphasised that CCTV while reducing crime in one area could increase crime in another due to displacement effects. Waples, S., Gill, M., & Fisher, P. (2009). Does CCTV displace crime? Criminology & Criminal Justice, 9(2), 207-224; Cerezo, A. (2013). CCTV and crime displacement: A quasi-experimental evaluation. European Journal of Criminology, 10(2), 222-236. ↩

52. Farrell, G., Tseloni, A. and Tilley, N. (2011) ‘The effectiveness of vehicle security devices and their role in the crime drop.’ Criminology and Criminal Justice 11, no. 1 pp. 21-35; Farrell, G., Tilley N. and Tseloni, A. (2014) ‘Why the crime drop?’, in M. Tonry (ed.) ‘Why Crime Rates Fall and Why They Don’t, volume 43 of Crime and Justice: A Review of Research Chicago: University of Chicago Press pp.421- 490; Morgan, N., Shaw, O., Feist, A., and Byron, C. (2016). Reducing criminal opportunity: vehicle security and vehicle crime. Home Office, London; Tilley, Nick, Graham Farrell, and Ronald V. Clarke. (2015) ‘Target suitability and the crime drop.’ In ‘The Criminal Act’, pp. 59-76. Palgrave Macmillan UK. ↩

53. Conduct disorder is a mental disorder which presents behaviours similar to anti-social behaviour. ↩

54. Farrington, D. P. (2005). Childhood origins of antisocial behavior. Clinical Psychology & Psychotherapy: An International Journal of Theory & Practice, 12(3), 177-190. ↩

55. Fitch, K. (2009). Teenagers at risk: The safeguarding needs of young people in gangs and violent peer groups. https://www.nspcc.org.uk/globalassets/documents/research-reports/teenagers-at-risk-report.pdf ↩

56. Murray, J., Farrington, D. P., & Sekol, I. (2012). Children’s antisocial behavior, mental health, drug use, and educational performance after parental incarceration: a systematic review and meta-analysis. Psychological bulletin, 138(2), 175. ↩

57. Wilson, H. W., Stover, C. S., & Berkowitz, S. J. (2009). Research Review: The relationship between childhood violence exposure and juvenile antisocial behavior: a meta-analytic review. Journal of Child Psychology and Psychiatry, 50(7), 769-779. ↩

58. Here it is important to note that the academic literature generally refer to three different types of offenders. These are set out in Appendix 4. ↩

59. Wikström, P. O. H., & Treiber, K. (2016). Social disadvantage and crime: A criminological puzzle. American Behavioral Scientist, 60(10), 1232-1259. ↩

60. https://www.bbc.co.uk/news/uk-48780585 ↩

61. MoJ (2019). Statistics on Race and the Criminal Justice System 2018 ↩

62. Wikström, P. O. H., & Treiber, K. (2016). Social disadvantage and crime: A criminological puzzle. American Behavioral Scientist, 60(10), 1232-1259. ↩

63. Farrington, D. P., Piquero, A. R., & Jennings, W. G. (2013). Offending from childhood to late middle age: Recent results from the Cambridge study in delinquent development. Springer Science & Business Media. ↩

64. Anderson, E. (1999). Code of the street: Decency, violence, and the moral life of the inner city. London: Norton. ↩

65. Goffman, A. (2014). On the run: Fugitive life in an American city. London: University of Chicago Press. ↩

66. Ball, R., Stott, C., Drury, J., Neville, F., Reicher, S. & Choudhury, S. (2019) Who controls the city? A micro-historical case study of the spread of rioting across North London in August 2011. City. https://doi.org/10.1080/13604813.2019.1685283. ↩

67. Phillips, C. and Bowling, B. (2017). ‘Ethnicities, racism, and criminal justice‘ in Liebling, A., Maruna, S. and McAra, L. (eds.) The Oxford Handbook of Criminology. Oxford: Oxford University Press. ↩

68. Liebling, A. with Arnold, H. (2004). Prisons and their Moral Performance: A study of values, qualities and prison life. Oxford: Oxford University Press. ↩

69. Sztompka, P. (1999). Trust: A sociological theory. Cambridge University Press. ↩

70. Harcourt, B. E. (2006). Language of the Gun: A Semiotic for Law & Social Science. ↩

71. Brennan, I. R. (2019). Weapon-carrying and the reduction of violent harm. The British Journal of Criminology, 59(3), 571-593. ↩

72. Jackson, J., Bradford, B., Hough, M., Myhill, A., Quinton, P., & Tyler, T. R. (2012). Why do people comply with the law? Legitimacy and the influence of legal institutions. British journal of criminology, 52(6), 1051-1071. ↩

73. Hough, M. Jackson, J., & Bradford, B. (2013). “Legitimacy, trust and Compliance: An Empirical Test of Procedural Justice Theory Using the European Social Survey” in Tankebe, J. and Liebling, A. (eds.) Legitimacy and Criminal Justice: An International Exploration. New York, NY: Oxford University Press (pp. 326-352). ↩

74. Legitimacy and trust are empirically similar yet conceptually distinct. While legitimacy focuses on the present (what is righful here and now), trust is more focused on the future. Bottoms, A., & Tankebe, J. (2012). Beyond procedural justice: A dialogic approach to legitimacy in criminal justice. The journal of criminal law and criminology, 119-170. ↩

75. Tankebe, J. (2013). Viewing things differently: The dimensions of public perceptions of police legitimacy. Criminology, 51(1), 103-135. ↩

76. Liebling, A. with Arnold, H. (2004). Prisons and their Moral Performance: A study of values, qualities and prison life. Oxford: Oxford University Press. ↩

77. Ibid; Liebling, A., Arnold, H. and Straub, C. (2011). An Exploration of Staff-Prisoner Relationships in HMP Whitemoor: 12 years on. ↩

78. Liebling, A., Price, D., & Shefer, G. (2011). The prison officer. Cullompton: Willan. ↩

79. Bottoms, A., & Shapland, J. (2014). Can persistent offenders acquire virtue?. Studies in Christian Ethics, 27(3), 318-333. ↩

80. Ibid. ↩

81. Ibid. ↩

82. Ibid. ↩

83. HM Inspectorate of Prisons (2016). Desistance and young people. ↩

84. Ibid. ↩

85. Bartol, C. R., & Bartol, A. M. (2011). Criminal behavior: A psychological approach. Upper Saddle River, NJ: Pearson Education Limited. ↩

86. Jolliffe, D., Farrington, D. P., Piquero, A. R., MacLeod, J. F., & Van de Weijer, S. (2017). Prevalence of life-course-persistent, adolescence-limited, and late-onset offenders: A systematic review of prospective longitudinal studies. Aggression and violent behavior, 33, 4-14. ↩

87. Ibid. ↩

88. Ibid. ↩

89. McGee, T. R., & Farrington, D. P. (2010). Are there any true adult-onset offenders?. The British journal of criminology, 50(3), 530-549. ↩

90. Ibid. ↩

91. Ibid. ↩