Race and policing: An inspection of race disparity in police criminal justice decision-making

Foreword

Some decisions that the police make daily about criminal justice matters can have life‑changing consequences for the people concerned.

For example, the power to arrest and deprive someone of their freedom is one of the strongest powers available to police officers. Subsequent decisions on whether to charge someone with an offence, or deal with a case on a more informal basis, often also fall to the police.

Police decisions also profoundly affect other parts of the criminal justice system. We know, for example, that children from ethnic minority backgrounds are disproportionately represented in youth custody. And this disparity has been getting larger over the past decade.

So it is of the utmost importance that the police and relevant Government departments gather information about whether these police decisions are being made fairly and consistently. And they should use that information to help them understand whether the decisions may be influenced by the characteristics of the people involved. To gain the trust and confidence of all communities, this information should be made easily available, and the public encouraged to help scrutinise it.

The police have done much work in recent years to identify race disparity in stop and search, and some use-of-force powers. This is important. But we wanted to know whether there had been an equal focus on other important interactions between the police and the public.

In 2017, the Lammy review kickstarted wider work to understand race disparity in the criminal justice system. The review highlighted how “Black, Asian and Minority Ethnic” people are disproportionately treated by the criminal justice system. But it didn’t consider in detail how police decisions may contribute to this.

We want our report to build on what the Lammy review has achieved. Our inspection examined whether there is race disparity in police criminal justice decisions. This is the first time that race disparity has been considered for some of the decisions.

In many cases, we found clear differences between the experiences of people from different ethnic backgrounds. For example, people from ethnic minority backgrounds were less likely than White people to receive a caution in 16 out of the 17 forces that we had data for.

A lack of consistency among forces in the recording, collection and analysis of relevant data makes it difficult for us to draw definitive conclusions, and for forces to understand and either explain or address any potential disproportionality. But we have identified opportunities for the police, and organisations they work with, to better understand and, where required, act to address race disparity in the police elements of the criminal justice system.

It is an important principle that police officers can exercise discretion when making criminal justice decisions. But we found the accompanying requirements – good supervision and effective organisational checks and balances – were often absent. These are essential so that forces can reassure themselves and the public that decisions are being made legitimately.

We wish to make it clear that evidence of disparity isn’t in itself evidence of discrimination. But it should warrant further investigation.

Police forces and Government have a legal duty to have due regard to how different people will be affected by their activities. This is set out in the Equality Act 2010 and is called the public sector equality duty. The duty applies equally to protected characteristics. This inspection focused only on how this duty was being applied to people from ethnic minority backgrounds.

We found that in England there was a need for a greater central government focus on considering disparity more closely in some of the criminal justice decisions we inspected. This would give much needed impetus to policing and other agencies.

The statutory requirement to gather and publish ethnicity information for some aspects of the operation of the criminal justice system dates back to 1991. But, despite progress in some areas such as stop and search, we found no evidence that the list of police criminal justice decisions that data needs to be published for had been updated since 2002. Given the many calls over the years for increased transparency, and the developing nature of many areas of policing, this should be addressed without further delay.

Set against this background, and the range of recommendations previously made in various statutory and non-statutory reviews and inspections related to race and policing specifically, we conclude there is also much greater scope for the National Police Chiefs’ Council (NPCC) and the College of Policing to consider potential race disparity in police criminal justice decisions. The NPCC’s diversity, equality and inclusion strategy doesn’t specifically address the need for this focus, and the accompanying toolkits give insufficient guidance to forces. In addition, neither the strategy nor the toolkits have been incorporated well enough into police work.

Police and crime commissioners (PCCs) have an important role in governance and oversight, and, in particular, the scrutiny of disparity in forces. PCCs can also bring together other agencies via Local Criminal Justice Boards (LCJBs) in a whole-system approach to understanding the problem. In 2006, the then Government rejected a recommendation by an independent review for LCJBs to play a larger role. This decision preceded the introduction of PCCs and, given their pivotal role, we would propose that this decision is revisited.

In our inspection, we did find some green shoots of encouraging activity. Some forces have introduced custody scrutiny panels as one way to gather, analyse and scrutinise information about some important criminal justice decisions. Similarly, some PCCs have introduced their own independent scrutiny panels to complement the work of forces. This is very positive, and we would like to see all forces adopt this approach.

The first iteration of the Police Race Action Plan (PRAP), published for consultation in 2022 by the NPCC and the College of Policing, states:

“Our intention is to remove bias, discrimination and racism from the use of police powers. Where disparities remain, we will explain why and set out what we believe can be done to address them.”

The police will be held to account for this commitment, but we are concerned that the production of the PRAP has been blighted by delays. We will work with the NPCC and other interested parties to consider what more we can do to help policing achieve the plan’s ambitions.

The Home Office has also introduced a requirement for police forces to provide it with information in the same format – where they have it – about protected characteristics in the data they regularly request from police forces.

It has been more than five years since the Lammy review concluded that “bringing decision-making out into the open and exposing it to scrutiny is the best way of delivering fair treatment”. But progress in achieving better scrutiny has been far too slow in response. And the police have failed to be proactive in trying to understand how their actions may directly contribute to criminal justice race disparity.

We have made 13 recommendations in this report. We are confident that with the right commitment by the police, working with Government and other organisations, significant progress can be made, and quickly, in this important area.

Terminology used in this report

Section 9 of the Equality Act 2010 states that race includes colour, nationality and ethnic or national origins. In our report, we also use the word ‘ethnicity’ in a broader sense to cover all aspects of race.

Our reports contain references to, among other things, ‘national’ definitions, priorities, policies, systems, responsibilities and processes. In some instances, ‘national’ means applying to England and Wales. In others, it means applying to England and Wales and Scotland, or the whole of the United Kingdom.

For the purposes of this report, we don’t consider it necessary to provide further differentiation between national, regional and local matters.

In March 2021, the Commission on Race and Ethnic Disparities recommended that the Government stop using the term ‘Black, Asian and minority ethnic (BAME)’. It recommended this because the terms ‘BAME’ and ‘Black and minority ethnic’ (BME) emphasise certain ethnic minority groups (Asian and Black) and diminish others (Mixed, Other and White ethnic minority groups). The terms can also mask disparities between different ethnic groups and create misleading interpretations of data. Any instances of ‘BME’ or ‘BAME’ are used only in the context of direct quotes in this report.

Summary

Leadership and governance

Over many years, there has been insufficient leadership from central government departments or senior police officers on race disparity in the police’s criminal justice decision-making. There are no relevant specific Government strategies, and the first published version of the Police Race Action Plan didn’t mention criminal justice disparity.

The statutory requirement to publish information on ethnicity for some aspects of the criminal justice system dates back to 1991 and section 95 of the Criminal Justice Act 1991. The list of police criminal justice decisions that should be published hasn’t been updated since. This is despite numerous reviews and inquiries calling for more transparency in how people from certain groups are treated by the criminal justice system.

In comparison, in June 2022 the Welsh Government published the Anti-racist Wales Action Plan. The plan included a crime and justice section, with a commitment to decide by June 2023 what relevant data relating to disparity should be gathered and published. This was followed by a Criminal Justice Anti-Racism Action Plan for Wales.

The lack of published data is unhelpful

There was a lack of published data on disparity, both at a force level and throughout England and Wales, for most of the police criminal justice decisions we looked at. This is surprising and worrying, as most of these decisions have serious consequences for the people and communities concerned.

In 2005, the then Government published the independent Root and Branch Review of Race and the CJS Statistics. The report made 32 recommendations. One of these was that Local Criminal Justice Boards should collect, collate and analyse criminal justice data relating to “Black and Minority Ethnic” people in their areas. The Government rejected this recommendation.

Police powers aren’t consistently scrutinised

The police and relevant Government departments have done effective work to better understand disparity in stop and search. But there hasn’t been comparable scrutiny of other important police powers considered in this report.

The Home Office has led on the need to understand the extent and causes of disparity in stop and search, and the use of force. But there has been much less work to explore the part that police decisions play in the race disparity that exists at each stage in the criminal justice process.

For example, we found that all forces published information about stop and search on their websites. But none of the forces we inspected published any information about disparity in police criminal justice decision-making.

There is clear evidence of race disparity in police criminal justice decision-making

We found that when police forces had analysed police criminal justice data, there was evidence of race disparity. But in most cases forces hadn’t carried out work to try to understand and explain the reasons for this.

Government data shows Black people are three times as likely to be arrested as White people. And our analysis of cases examined in our fieldwork also found clear differences between the experiences of people subject to police criminal justice decisions, depending on their ethnicity. For example, after arrest, Asian people were substantially less likely to receive a caution than White people.

The Commission on Race and Ethnic Disparities made it clear in its 2021 report that “differences – or ‘disparities’ – are not always sinister and do not always arise from discrimination”. But forces should do more to analyse and explain why these differences exist.

The police need to improve the collection of ethnicity data relating to decision-making

At the time of our inspection, police forces weren’t recording ethnicity in a consistent way. This is needed for the effective and efficient gathering of data, which is an important first step towards managing and improving performance. Progress has been too slow on creating a single recording standard, and there hasn’t been enough strategic oversight of this important area.

Limitations of police IT systems have contributed to the failure to gather enough data, and there is a need for greater impetus to make things better. In 2005, the Government’s Root and Branch Review of Race and the CJS Statistics recommended that changes should be made to the Police National Computer so that ethnicity is better recorded. These are finally intended to be introduced in 2023.

Equally, police officers aren’t given adequate guidance and training about why it is important to gather accurate information on the ethnicity of the people they come into contact with.

There are some green shoots of activity

The Association of Police and Crime Commissioners (APCC) published race action plans for 2020/21 and 2021/22 that include specific actions relating to criminal justice matters. These can help police and crime commissioners (PCCs) make sure police forces contribute to our understanding of whether race disparity exists, and in what forms. The APCC also publishes a Race Disparity Toolkit twice a year. It helps police to promote equality and diversity principles locally and to fulfil their duties under the Equality Act 2010. And since December 2021, the APCC has annually published an Equality Framework, to provide guidance and help its members self-assess how they are meeting their equality duties locally. The framework is endorsed by the Equality and Human Rights Commission.

This range of resources for PCCs contrasts with a lack of information and guidance aimed directly at police forces themselves. Despite this, some forces have started to think about how they can develop processes for gathering relevant information on disparity, and what they should do with the results.

In April 2023, the Home Office introduced a requirement that data it requests from forces should contain information about protected characteristics where this is held. This won’t compel forces to make changes to their systems to collect data that they currently don’t gather, but it is a positive first step.

There is little external community scrutiny of police decision-making

There is little community involvement in the scrutiny of race disparity for most criminal justice decisions. Where there is scrutiny, it usually only involves out-of-court disposals. Even then, there is no consistency in the scrutiny processes used by forces.

None of the forces we inspected published any information about the scrutiny of disparity in criminal justice decision-making. There was some limited information published on PCCs’ websites about out-of-court disposal scrutiny.

Recommendations

Introduction

Fair decision-making, and working and interacting positively with the public, can improve perceptions of the police and increase trust in them. (This is often referred to as procedural justice.) This, in turn, helps efforts to reduce crime by encouraging greater respect for the law. It makes people less likely to break the law and fosters a social responsibility that leads to more co-operation with the police. Research has shown that procedural justice is likely to increase public trust in the police – more so than neighbourhood policing or low crime rates.

When the police use their powers disproportionately – in different proportions for different ethnic groups – it causes suspicion among some communities that they are being unfairly targeted. This can undermine confidence in, and the legitimacy of, the police, which are fundamental aspects of the British model of policing by consent.

Adequate explanations for disproportionate use of criminal justice decisions are likely to help reassure communities – including those that have low levels of trust and confidence in the police. Without those explanations, the public may conclude that discrimination may play a part in the way the police make these decisions.

Disproportionate criminal justice decision-making leads to more people from ethnic minority backgrounds being drawn into the criminal justice system. The effects of this can include disrupting education, reducing work opportunities, and breaking down families and communities. It can contribute to perceptions among the public and police officers about people from ethnic minority backgrounds and crime. It may also influence how the police allocate resources. This in turn can increase imbalances in the criminal justice system, and perceptions of a link between ethnicity and criminality.

Some police powers, such as depriving a person of their liberty by arresting them, may have a considerable effect on the person concerned. So it is particularly important that the Government, the police and other interested parties understand how such powers are being used.

In this report, we examine what published demographic data is available for a range of police criminal justice decisions. We then detail the policy context that exists for these, before setting out the findings from the fieldwork that we carried out.

Background

The negative effect of public perceptions of disproportionate decision-making, and the resulting poor relations between the police and certain communities, shouldn’t be underestimated. The damage can be far-reaching and long-lasting.

In 2021, the chair of the National Police Chiefs’ Council (NPCC) acknowledged that it was an “operational necessity to boost racial justice” because the issue was affecting the trust and confidence of Black communities in policing.

Disparity doesn’t necessarily mean there has been discrimination or misuse of powers. But forces should be able to explain disparities and show the public evidence that their use of the powers is fair, lawful and appropriate. At the introduction of the race disparity audit in October 2017, the then Prime Minister Theresa May said: “If the disparities can’t be explained, they must be changed.”

And there are other important reasons why the police should gather, analyse and act upon information about the powers they use. Police forces are specified public bodies under the Equality Act 2010 and have a duty to comply with the public sector equality duty. The duty is to have ‘due regard’ to the need to achieve these three goals:

• eliminating unlawful discrimination, harassment and victimisation, and other conduct prohibited by the Act;
• advancing equality of opportunity between people who have a protected characteristic and those who don’t; and
• fostering good relations between people who have a protected characteristic and those who don’t.

Section 96 of the Police Act 1996 and section 14 of the Police Reform and Social Responsibility Act (PRSRA) 2011 require that arrangements for obtaining the views of the community should be made by the local policing body, in consultation with the chief officer.

There are further requirements under sections 34 and 36 of the PRSRA for chief officers to fulfil their statutory duties to engage with local people, make arrangements for obtaining their views on policing, crime and disorder, and provide information to them and to the elected local policing body or police and crime commissioner.

In 2016, the then Prime Minister David Cameron asked David Lammy MP to investigate evidence of possible bias against Black defendants and other ethnic minorities. In 2017, David Lammy published The Lammy review: An independent review into the treatment of, and outcomes for, Black, Asian and Minority Ethnic individuals in the Criminal Justice System. The review concluded:

“Those who are charged, tried and punished are still disproportionately likely to come from minority communities.”

The review contained 35 recommendations. None of these were specifically for the police. But there were several areas where the police’s involvement was needed to support an appropriate response.

Recommendation 4 of the review stated:

“The government should introduce a new principle of ‘explain or reform’ for racial disparities across the CJS. If governments cannot provide an evidence-based explanation for apparent disparities, then reforms should be introduced to address them.”

The Police Race Action Plan (PRAP) published in 2022 by the NPCC and the College of Policing commits to adopting this ‘explain or reform’ approach.

We discuss some other relevant recommendations and the responses to them later in this report.

In 2020, the Ministry of Justice published its Tackling Racial Disparity in the Criminal Justice System: 2020 Update report. The report also included an update about progress on responding to the Lammy review.

The then Lord Chancellor and Secretary of State for Justice the Right Honourable Robert Buckland MP said in the foreword:

“The current over-representation of people from racial and ethnic minorities in the Criminal Justice System (CJS) is a serious concern and something I have seen first-hand over a 25-year career working in the system. Making strides to address this is a focus for me personally as the Lord Chancellor and Secretary of State, and for the Ministry of Justice as a whole, and the many agencies and partners with whom we work.”

Also in 2020, the House of Commons Justice Committee published the Children and Young People in Custody (Part 1): Entry into the youth justice system report, which concluded that:

“The youth justice population has changed considerably in the past 10 years, but children from BAME backgrounds continue to be disproportionately represented, with outcomes getting worse in some areas.”

In March 2021, the Commission on Race and Ethnic Disparities (CRED) published its report on race and ethnic disparities in the UK. The report made 24 recommendations towards 4 overarching aims:

• to build trust between different communities and the institutions that serve them;
• to promote greater fairness to improve opportunities and outcomes for individuals and communities;
• to create agency so individuals can take greater control of the decisions that impact their lives; and
• to achieve genuine inclusivity to ensure all groups feel a part of UK society.

There are three specific recommendations for policing which are relevant to this inspection:

• Recommendation 4: Bridge divides and create partnerships between the police and communities.
• Recommendation 5: Improve training to provide police officers with practical skills to interact with communities.
• Recommendation 12: Prevent harm, reduce crime and divert young people away from the criminal justice system.

The first two recommendations are mainly rooted in the commission’s findings on stop and search. But there is a link with this report’s findings in relation to the need for better scrutiny, accountability and training.

Recommendation 12 is directly relevant to this inspection, as we examined the use of out-of-court disposals – the main way the police divert people from the criminal justice system.

The commission proposed the following framework to distinguish between different forms of racial disparity and racism:

• Explained racial disparities: this term should be used when there are persistent ethnic differential outcomes that can demonstrably be shown to be as a result of other factors such as geography, class or sex.
• Unexplained racial disparities: persistent differential outcomes for ethnic groups with no conclusive evidence about the causes. This applies to situations where a disparate outcome is identified, but there is no evidence as to what is causing it.
• Institutional racism: applicable to an institution that is racist or discriminatory processes, policies, attitudes or behaviours in a single institution.
• Systemic racism: this applies to interconnected organisations, or wider society, which exhibit racist or discriminatory processes, policies, attitudes or behaviours.
• Structural racism: to describe a legacy of historic racist or discriminatory processes, policies, attitudes or behaviours that continue to shape organisations and societies today.

We discuss in the section Findings from our fieldwork why the results of our data analysis in this inspection presently amount to unexplained racial disparities.

Also in 2021, His Majesty’s Inspectorate of Probation carried out an inspection of the experiences of Black and mixed heritage boys in the youth justice system. The report contained several recommendations for the police relating to the need to publish data broken down by ethnicity (as well as gender and age). The report concluded that partners in the criminal justice system were:

“not collating data and using it effectively to analyse and address the barriers that contribute to the over-representation of black and mixed heritage boys in the criminal justice system”.

In March 2022, the Government published its response to the CRED report. The policy paper included responses to the CRED recommendations and actions to be taken. We discuss one of these actions at the Scrutiny section below.

In February 2023, the House of Commons Committee of Public Accounts published the Support for vulnerable adolescents report, which concluded that:

“The Ministry of Justice and Home Office seem to lack curiosity about the increase in the proportion of children from ethnic minority background in youth custody and appear to have no current plan to address the situation.”

The committee recommended that within six months, the Ministry of Justice and the Home Office should report back on their understanding and planned actions in relation to this issue. In May, the Government responded that it agreed with the committee’s recommendation.

So it is important for relevant organisations to work together to consider whether there is disparity at crucial police criminal justice decision points and, where it exists, its extent.

About us

His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS) independently assesses the effectiveness and efficiency of police forces and fire and rescue services, in the public interest. In preparing our reports, we ask the questions that the public would ask, and publish the answers in accessible form. We use our expertise to interpret the evidence and make recommendations for improvement.

Our commission

In July 2021, the Home Affairs Select Committee published a report summarising what progress had been made on the recommendations of the 1999 Macpherson report.

The report concluded that there had been some important improvements in some aspects of the police response. For example, in the policing of racist crimes, in the commitments made to promoting equality and diversity, and in good examples of local community policing.

The report recommended that HMICFRS carries out thematic reviews of race and policing to provide scrutiny of the police service response. Our programme of work on race and policing demonstrates our response to this recommendation.

This inspection is the second in our race and policing programme that we started in 2021. The first inspection, focusing on leadership and governance , is published alongside this report. We draw on some of the findings from that inspection in this report.

Our overarching question for the race and policing programme is:

How effectively are the police responding to race-based inequalities and disparities that negatively affect community confidence?

In this inspection, our aim was to explore race disparity in the police forces selected for fieldwork, and in particular how well they understand, scrutinise and communicate their relevant work.

Methodology

What we inspected

This inspection focused on decisions made during criminal justice processes that affect individuals who are the subject of criminal investigations. For the purposes of this inspection, ‘criminal justice processes’ mean those activities that take place after the police have decided that a criminal act may have taken place. The decisions we focused on included those that may take place after something is found in a stop and search encounter.

But we didn’t specifically consider disparity in stop and search, and use of force powers. This is because we include these subjects in our core policing inspections, which we call the PEEL assessment framework (PAF). In February 2021, we also published the separate Disproportionate use of police powers report on stop and search, and the use of force.

So the question of how police actions and decisions affect alleged victims wasn’t included. This is because we also already consider this matter in our PAF.

The police criminal justice decision points that this inspection covered are:

Arrest

Police officers mainly make arrests using powers specified in the Police and Criminal Evidence Act 1984. A lawful arrest requires two elements:

• a person’s involvement, suspected involvement or attempted involvement in the commission of a criminal offence; and
• reasonable grounds for believing that the person’s arrest is necessary. It can never be necessary to arrest a person unless there are reasonable grounds to suspect them of committing an offence.

Both elements must be satisfied.

Officers must always consider whether a person’s arrest for an offence is lawful and necessary. When officers think that an arrest isn’t necessary, they can decide to use voluntary attendance to seek a person’s explanation of events (see below). For example, this could be where an offence has taken place some time ago and there is no need to secure evidence immediately.

As arrest is a strong power, it is important that police forces reassure themselves and the community that it isn’t used differently for different ethnic groups.

Voluntary attendance

Police can use voluntary attendance (VA), also known as a voluntary interview, to interview a suspect who isn’t under arrest. VA can be used for both adults and young people, and be carried out at or away from police buildings. It is a formal interview to gather material about an allegation of crime and, as such, may have significant consequences for the suspect. A suspect in a VA interview has the same rights and entitlements as they would have in an interview carried out in police detention under arrest. The difference is that the suspect in a VA interview has the right to leave.

In general, and when used appropriately, VA can have less of an impact on a person than an arrest. For example, appointments for VA are made at times that are mutually convenient, and people aren’t placed in police cells. So police forces need to be sure that VA isn’t used less often for some ethnic groups than others when similar circumstances, such as offence types, are present.

Community resolution

Community resolution is an alternative way of dealing with less serious crimes that allows the police to use their professional judgement when dealing with an offence. It typically seeks to resolve an offence informally, promptly and in a way that brings closure for the victim where one has been identified. It should take account of the victim’s views and may involve the offender doing something that ‘puts right’ the offending to the victim’s satisfaction.

Community resolution is generally seen as an informal way of dealing with a crime. So if information suggested that one ethnic group was more likely than others to receive community resolutions for similar offences, this should warrant further attention.

Cautions

At the time of our inspection, there were two types of cautions – simple and conditional. Cautions can be given to anyone aged 10 or over, usually for minor crimes. The person has to admit committing an offence and agree to be cautioned.

A simple caution is a formal notice issued by a police officer that warns a person not to reoffend, and about what could happen if they do. It is used in cases where it isn’t in the public interest to prosecute the person who has offended. This usually relates to relatively low-level offending.

A conditional caution sets specific conditions that the person is required to keep to. If the person doesn’t comply with the conditions, then the police may choose to recommend prosecution. A conditional caution is likely to be used in cases where prosecution isn’t in the public interest, but the police consider that specific actions are required. This may be to reduce the likelihood of further offending, to protect a specific victim, or where previous out-of-court disposals haven’t been effective.

In 2023, forces will begin to move towards using two new types of cautions, called ‘community’ and ‘diversionary’ cautions.

Police officers should use a ‘gravity matrix’ framework to help them make decisions about whether an adult is suitable for a caution or a community resolution. There is a similar process for young people, but these decisions are mostly taken in partnership with local Youth Justice Services.

Cautions provide an opportunity for offences to be dealt with, but avoid a person attending court and being given a criminal record. So it is important that forces monitor whether decisions about cautions are being made fairly and consistently.

Charge

Charging is, in effect, sending someone to have the case against them heard at court. The police only have the power to charge someone with an offence in some cases.

The Director of Public Prosecutions issues guidance on which cases the police can make decisions on, and which are the responsibility of the Crown Prosecution Service. In general, the police have the power to charge for less serious offences.

In this inspection we only considered charging decisions made by the police. As these decisions can have profound consequences, any differences between ethnic groups should be identified and carefully considered.

Post-charge bail

Under the Bail Act 1976, when a person who is in police custody has been charged with an offence, the police must release the person on bail (with or without conditions) unless certain conditions apply. For example, the police may refuse to grant bail if the custody officer considers that the charged person may not appear at court.

Depriving someone of their freedom is a serious step. It is also more likely that a person will be remanded into custody by a court if that person has been refused bail by the police. So it is particularly important that for these decisions, the police understand whether any differences between ethnic groups exist.

We chose all the above decision points because they are likely to have a significant impact on the person concerned. More generally, there is also considerable public interest in whether the police are acting fairly, and whether the criminal justice system acts equally in the interests of all members of the community.

We didn’t set out to examine the particular circumstances of the people involved in the cases, or whether or not the police’s decisions were in our view correct. So we have generally not attempted to explain any apparent disparity. Our findings should only be taken as broad indications of where more work is needed by police forces and other organisations to gather and analyse information about race disparity in police criminal justice decisions.

What do we mean by disparity?

In this report, we use the words ‘disparity’ or ‘disproportionality’ to mean that a certain group of people is affected by police action in a way that is substantially different from people not in that group. The group that this report focuses on is people from ethnic minority backgrounds.

We have taken the same approach to considering disproportionality as in our other inspections. This uses information from the most recent census of England and Wales about the proportion of people from different ethnic backgrounds in each police force area.

In this report, we sometimes discuss the relative likelihood of a person from an ethnic minority group being subject to a police decision compared to a White person (the majority group). In these instances, we have followed Ministry of Justice guidance and applied tolerance zones to disparities so that we could be confident that these weren’t due to chance. See Annex A for a detailed explanation of this approach.

How do police forces record ethnicity?

Crime recording systems are used when a person has been accused of a crime and then subject to a criminal justice decision, such as a caution. Custody systems are used when a person has been arrested and is taken into police custody. In some forces, one IT system is used for both of these purposes. In others, there is a different IT system for each.

Ethnicity is recorded by the police in two ways:

• self-defined ethnicity (SDE); and
• officer-defined ethnicity.

SDE doesn’t relate to visible appearance but to a person’s self-image in relation to his or her own cultural origins. A police officer should record the SDE category chosen by the person encountered, even if the category chosen is clearly at odds with the officer’s visual assessment. Officers are required to ask the person concerned what they consider their SDE to be and record it in the relevant data field.

The NPCC policy for recording SDE dates back to 2002, when it was introduced as the 16+1 ethnic classification system.[1] There were 16 specific ethnic classifications, while the “+1” referred to the category “not stated”.

This brought the police into line with census data that started using the same ethnic classifications from 2001.

In 2012, in line with changes in the 2011 census, the 18+1 ethnic classification system was introduced. The additional two ethnic classifications were Arab, and Gypsy or Irish Traveller.

From April 2023, following consultation between the Home Office and the police, the 19+1 ethnic classification system will be introduced. The additional classification is Roma. This is again linked to changes in the census, in this case from 2021. The full list of categories is included at Annex B. Some forces have already made this change.

Despite this policy, police forces record SDE in different ways. The NPCC Protected Characteristics Working Group (which sits within the NPCC diversity, equality and inclusion portfolio) is working with partners and other interested parties to understand the situation and agree one set standard.

It is important that ethnicity data is correctly and consistently gathered and recorded. So we were surprised to find that the work needed to standardise this within policing wasn’t included in the NPCC’s diversity, equality and inclusion strategy or the associated toolkits. It was also not included in the main body of the PRAP, published for consultation in 2022. Appendix B of the PRAP does however detail the need to “work across policing to improve the consistency of capture, application, and use of data and information relating to race and inclusion”.

We will continue to monitor the work of the NPCC to make sure that it gives this important subject the appropriate priority.

Officer-defined ethnicity, also called visual assessment descriptive monitoring, requires officers to make their own judgment about the visible ethnic appearance of the person concerned. There are six categories plus an additional one of “not known”. The list of categories is included at Annex B.

In launching the new approach to recording SDE in 2002, the then Association of Chief Police Officers leads made a joint statement:

“Effective policing in the 21st century is dependent upon productive partnership between the service and all members of society. To achieve this, we must treat everybody proportionately, according to their needs and operate in a culture of transparency and openness.

It is important that we gather information to indicate fairness in operational policing and to highlight priority areas for closer scrutiny and decisive action.”

This statement is still true today.

How we inspected

We began our inspection by asking all forces to complete a survey. We discuss the findings from the survey below.

Reviewing the survey responses helped us to determine what we should look for in our fieldwork. We then developed the criteria and indicators for the inspection. These are stated at Annex C.

We also used the survey to help us decide which forces to visit.

We carried out our force fieldwork in two phases. Phase one consisted of interviews and focus groups in five forces. These were Hampshire Constabulary, Gloucestershire Constabulary, Gwent Police, Northamptonshire Police and Staffordshire Police.

In phase two we carried out a case file review in the same five forces as well as Merseyside Police. We asked these six forces to give us information about a range of offences that had been recorded between 1 April 2021 and 31 March 2022, and decisions made in relation to them. The offences and the police decision points we reviewed are shown in the table below, and described in more detail in Annex E.

Table 1: Decision points and offences that we reviewed

Decision point	Offences we reviewed these decisions for
Arrest	Possession of an offensive weapon, and criminal damage
Voluntary attendance	Possession of an offensive weapon, and criminal damage
Community resolution	Possession of cannabis, and shoplifting
Caution	Possession of cannabis, and shoplifting
Charge	Possession of cannabis, and shoplifting
Post-charge bail	Burglary, and actual bodily harm (non-domestic abuse related)

We chose to review the offences above because the police often make decisions on crime outcomes for these without reference to the Crown Prosecution Service. We also wanted to examine some offences, such as possession of cannabis, that are often relevant to stop and search encounters.

The fieldwork forces provided us with information about a total of 22,624 relevant crimes that had been recorded between 1 April 2021 and 31 March 2022. We conducted analysis of these records. Then we randomly selected 1,470 for our case file review. This was calculated to be a statistically significant sample, representative of the volumes of each decision point submitted by forces, and the ethnicities of the people dealt with by the police. We discuss the results of these reviews in the Findings from our fieldwork section below.

We also interviewed leaders of relevant organisations, and officials from the Home Office and the Ministry of Justice.

Our inspection was supported by an external reference group (ERG), which was set up to advise us on our wider race and policing programme. A list of members of the ERG is included at Annex D.

More details about the methodology for this inspection can be found at Annex E.

Published data

Background: Published data

It is very important that all forces gather and publish good quality data. This also helps Government and other organisations be aware of any potential for disparity and the potential impact on community confidence.

The public sector equality duty, which we discuss elsewhere in this report, applies to Government departments, police forces, and police and crime commissioners and their offices.

The first statutory requirement to gather and publish information on ethnicity in relation to certain aspects of the operation of the criminal justice system was introduced by section 95 of the Criminal Justice Act 1991. The purpose of the legislation was to gather and publish information on ethnicity to help Government and public sector organisations “avoid discriminating against any persons on the grounds of race, sex or any other improper ground”.

We weren’t able to establish the exact arrangements that existed in 1991. The earliest record of information required by the Home Office from the police that we were told about is from 2002, a ‘Guide to Self-Defined Ethnicity and Descriptive Monitoring’[2] produced by the Association of Chief Police Officers (ACPO) – the forerunner of the National Police Chiefs’ Council (NPCC). This lists information, for the Home Office to publish, on:

• stop and search;
• arrest;
• cautions, reprimands and final warnings;
• police complaints;
• police manpower (and personnel);
• deaths in police custody; and
• homicide.

To coincide with this ACPO guide, the Home Office issued revised guidance to ensure consistency between forces. The Home Office told us this guidance is no longer widely available as it has been archived.

The 2002 ACPO guide also describes how, in October 2000, ACPO voluntarily extended the ethnicity monitoring requirement on forces to include certain traffic‑related stops.

Although it hasn’t since been updated, the ACPO guide contains useful information for forces relating to the collection, analysis and sharing of information on ethnicity. The guide also has a chapter titled ‘The imperative for training’. We discuss what training is currently available in the Training and guidance chapter below.

In 2005, the Government published the Root and Branch Review of Race and the CJS Statistics: Final Report. The report emphasised that statistics on race and the criminal justice system are important because:

“They represent one of the main ways in which accusations of institutional discrimination by CJS agencies can be judged.”

The report made 32 recommendations to various agencies and organisations to bring about improvements in the availability and use of data on people from ethnic minority backgrounds. One of the recommendations related to the better use of Local Criminal Justice Boards. We discuss this further in the Local accountability section.

The report also made two recommendations that directly affected the police. We have included these below, together with the subsequent Government response:

“Recommendation 22: Quality thresholds for completion rates should be set for the Police, as the collecting agency:

• To encourage more complete recording of BME classifications
• To encourage the police to pass more complete data to other agencies.”

Government response:

“We will work to formalise our requirement for expected levels of data quality by issuing a minimum data quality and timeliness. This will be the subject of guidance.”

We haven’t been able to establish whether this guidance was ever produced.

“Recommendation 27: Notwithstanding the variation between police-defined and self-defined ethnicity, the Police National Computer (PNC) should change its BME classification system to a classification compatible with the 16+1 system, to improve the quality of BME data.”

Government response:

“The Association of Chief Police Officers who are the owners of the data held on the Police National Computer (PNC) have agreed that the 16+1 BME classification should be considered as part of the work to be incorporated in PNC modernisation.”

The PNC still uses the six officer-defined ethnicity categories. The replacement for the PNC is called the Law Enforcement Data Service (LEDS). We have been told by the NPCC lead for PNC and LEDS that the 16+1 ethnicity recording system will be introduced into LEDS in 2023. This is long overdue.

Emphasising the importance of publishing good quality data on ethnicity, the Lammy review referred to the Criminal Justice Act 1991 and contained the following recommendation:

“Recommendation 3: The default should be for the Ministry of Justice (MoJ) and CJS agencies to publish all datasets held on ethnicity, while protecting the privacy of individuals.”

The recommendation related to publishing existing data sets. But it didn’t ask the Ministry of Justice and other criminal justice service agencies to consider what other datasets could and should be published.

In 2020, the Ministry of Justice responded to the recommendations. The response included several examples of new ethnicity data that had been published since the recommendation was made in 2017. None of these related to police data. The Ministry of Justice also stated:

“We will work closely with the Cabinet Office Race Disparity Unit to identify other justice areas where further ethnicity data may be published on gov.uk.”

The Ministry of Justice told us that none of the other justice areas considered for further work had included police criminal justice decision points.

The Home Office leads on collecting data from police forces in England and Wales through its annual data requirement (ADR). Some of the decision points we considered in our inspection are covered by regular Home Office data collections. At the time of our inspection fieldwork the only requirement for all police forces to provide ethnicity data for the decisions we examined related to arrests and voluntary attendance.

In April 2023, the Home Office introduced a separate requirement. This requires forces to provide information about protected characteristics, where they have it in a consistent way, for all data returns. This requirement will initially be voluntary because forces may not currently collect this information.

Appendix B of the Police Race Action Plan details the plan’s data approach, including the ambition to “improve the consistency of capture, application, and use of data and information relating to race and inclusion”. While not specific to police criminal justice decisions, the plan supports:

“Recommendations for data capture gaps to be closed via Home Office requirements and the mechanisms available, ADR and the Crime Survey of England and Wales (CSEW).”

We strongly support the view that publishing data on race disparity allows scrutiny and is likely to lead to activity to provide explanations for any apparent disparity. But currently, it isn’t mandatory for forces to collect ethnicity data about most criminal justice decisions. So the data isn’t complete or robust enough to be published for all forces.

In our inspection we considered six important criminal justice decision points. Below, we outline what published data is available for each point. (See Annex A for an explanation of the tolerance zones we have used for the data in the charts below.)

Arrests

The Home Office publishes annual data on arrests as part of its Police powers and procedures statistical series. As stated above, the statutory requirement to do so dates back to the Criminal Justice Act 1991.

Figures published in March 2023 show that, for the year ending 31 March 2022, Black people were 2.4 times as likely to be arrested as White people. There were 21 arrests for every 1,000 Black people, and 9 arrests for every 1,000 White people (see figure 1 below). These figures were calculated using resident population data from the 2021 census; we also used 2021 census data for our fieldwork analysis.

Figure 1: Likelihood of people being arrested compared to White people by ethnicity, all forces in England and Wales, in the year ending 31 March 2022

Source: Home Office Police powers and procedures, year ending 31 March 2022

There are also very wide, and unexplained, variations between forces. For example, Dorset Constabulary had the highest rate of disparity, with Black people being 8.2 times more likely to be arrested than White people in the year ending 31 March 2022. Meanwhile, Northumbria had the lowest rate of disparity, with Black people 1.4 times more likely to be arrested than White people.

In contrast, Asian people are generally less likely to be arrested compared to White people, with less variation between forces.

Every two years the Ministry of Justice publishes information relating to arrests as part of its ‘Statistics on Ethnicity and the Criminal Justice System’ report. This reproduces information included in the Home Office Police powers and procedures statistical series discussed above.

The Youth Justice Board (YJB) for England and Wales publishes information relating to the arrests of children annually.

In January 2022 the chair of the YJB published a blog called ‘Annual statistics: a youth justice system failing Black children’. This highlighted statistics showing that Black children are overrepresented throughout the criminal justice system, including arrests. For example, Black children account for 4 percent of the 10–17-year-old population (Census data, 2011) but 15 percent of arrests.

There is a measurable disparity in the arrest rate for all Black people, and a clear disparity among children.

Voluntary attendance

In 2021, the Home Office started to require data on voluntary attendance (VA). This data had been collected on a voluntary basis since April 2020. Since 2021, it has formed part of the department’s ADR from forces, and some of the results are published as experimental statistics in the Police powers and procedures statistics.

But just over half of forces don’t give this data to the Home Office. We generally found racial disparity among the 21 forces that submitted data to the Home Office for the year ending 31 March 2022. For 18 of these forces, Black people were more likely to have been subject to VA compared to White people. People identifying as being from an Asian or other ethnic background were generally less likely than White people to be dealt with through VA.

The likelihood of VA being offered varies between forces and between ethnic backgrounds. Again, there is higher disproportionality, and unexplained variation between forces, for Black people compared to other ethnic groups.

Community resolution

There is no publication of race disparity information relating to these police decisions that covers all police forces.

Cautions

The Ministry of Justice publishes quarterly criminal justice system statistics. Information on out-of-court disposals (excluding community resolution) is included in this publication. As with data on arrests, the requirement to publish data on cautions dates to the Criminal Justice Act 1991.

The published caution statistics include demographic information, including data on ethnicity. This is in an inaccessible format (a spreadsheet pivot table) that many people won’t be familiar with. And there is no analysis or summary to help the reader understand this data.

Every two years, the Ministry of Justice also publishes statistics on ethnicity and the criminal justice system. But this doesn’t include information about cautions.

So demographic data is gathered in relation to cautions. But it is a missed opportunity that there is no analysis of this and that the data isn’t at least summarised in the Ministry of Justice’s biennial ethnicity publication.

We have carried out our own analysis of published data gathered by the Home Office about the outcomes of VA and arrests.

We found racial disparity in how cautions were used in 11 of the 17 police forces that submitted data on VA in the year ending 31 March 2021. At the time of our inspection, more recent similar data for the year ending 31 March 2022 wasn’t available.

In eight forces, Black people were less likely to receive a caution compared to White people. And in 11 forces, Asian people were less likely to receive a caution than White people. Overall, people from ethnic minority backgrounds were less likely than White people to receive a caution in 16 out of the 17 forces that had submitted data to the Home Office.

In 2023, a new two-tier system of out-of-court disposals is being introduced. Simple and conditional cautions will be replaced by community and diversionary cautions.

The Ministry of Justice told us that the accompanying statutory guidance would address the need for better gathering and analysis of demographic information. The NPCC lead for out-of-court disposals is working closely with the Ministry of Justice to put in place systems and processes for forces to make sure this activity happens.

There is more information available about cautions in relation to children. The YJB has a summary disproportionality tool, for which Youth Justice Services provide information. The tool gathers demographic information on young people that the services deal with, and the outcomes for them.

The YJB publishes data annually on the proportion of children receiving a caution or sentence by ethnicity. This shows that the proportion of children cautioned or sentenced who are Black has been increasing over the past 10 years – from 7 percent in the year ending March 2011 to 12 percent in the year ending 31 March 2021.

Police charging decisions

There is no publication of race disparity information relating to these police decisions that covers all police forces.

Decisions to bail or not after charge

There is no publication of race disparity information relating to these police decisions that covers all police forces.

Conclusion

For most of the important criminal justice decision points that we examined in this inspection, there is very limited published information about whether race disparity exists throughout England and Wales.

The Home Office publishes full data on disparity for arrests and limited data for VA.

The Ministry of Justice publishes data relating to cautions – although, as we discuss above, it isn’t presented with accompanying analysis. This approach makes the data difficult to interpret and needs to be clarified.

The Government has a website called Ethnicity facts and figures. The purpose of this is to gather in one place information about the experiences and outcomes of people from a variety of ethnic backgrounds.

At the time of our inspection, little information relating to policing was published on the website. It included information on arrests and youth cautions, but not on any of the other decision points considered in this report.

Requirements for data gathering and publication of police criminal justice decisions remains unchanged since 2002. Much has changed since then, including an increase in awareness of the effects of disparity on individuals and community confidence.

The policy context

Strategies, policy guidance and action plans

Central government

There are no Home Office or Ministry of Justice (MoJ) strategies or guidance that specifically include a requirement for forces to gather, analyse and publish information about disparity in criminal justice decision-making.

While the MoJ doesn’t have a specific strategy, there is a clear expectation on the police to improve the collection and publication of criminal justice data on protected characteristics.

In 2007, after changes within Government, the MoJ took over responsibility for publishing data on race disparity from the Home Office. The biennial Race and the criminal justice system report continues to fulfil this requirement. We discuss this further under Published data.

MoJ policy officials we spoke to told us that responsibility for collecting data from police forces for the criminal justice decisions we examined rested with the Home Office.

The Home Office makes it clear that it is important for the police to collect and publish information about race disparity. The Home Office leads on collecting data from police forces in England and Wales. We consider this further under Published data, above.

The Home Office then publishes some of this data. This mainly relates to the use of powers (such as stop and search) and the use of force (for instance, the use of tasers).

Policy officials in the Home Office told us that, of the decision points we examined, the department only had policy responsibility for arrests, voluntary attendance and bail after charge. The Home Office leads on the publication of information about arrests and voluntary attendance.

The Home Office also oversees the work that takes place in police custody suites, although not the decisions about criminal justice outcomes that take place there.

So in England, both departments are concerned with the need for police forces to gather information about race disparity in the criminal justice decision points we examined in this inspection.

The complex arrangements described above may partly explain why (as we discuss elsewhere) there have been no changes to what police criminal justice disparity information is published by Government since 2002.

In its report Ethnicity and the Criminal Justice System, 2020, the Government acknowledges there is race disparity in arrests. Likewise, it accepts that people from ethnic minority backgrounds make up a disproportionate number of people in the prison system. But there has been little work to explore the part that police decisions may play in any race disparity that exists at the intervening stages in the criminal justice process. For example, we don’t know whether there is a progressive increase in disparity after the arrest of people from ethnic minority backgrounds.

The Welsh Government

In September 2020 the Welsh Government agreed with criminal justice and other bodies to a “one public service response to advance race equality across Wales”. The resulting Anti-racist Wales Action Plan was published in June 2022.

One of the goals of the action plan is as follows:

“To respond to the clear evidence from stakeholders that strengthening the evidence base to identify racial disparity within the justice system is vital, we will collect and report robust, granular, timely and consistent criminal justice data on ethnicity and race, together with working closely with ethnic minority people to improve their knowledge/trust in how their data will be used to increase their confidence to share data with government and partner organisations.”

The action plan also contains a commitment to create a specific Criminal Justice Anti‑Racism Action Plan for Wales, and for the Criminal Justice Board for Wales to provide oversight of the introduction of the plan. There will also be an independent scrutiny panel, which will allow for more transparency and accountability for the organisations involved. The criminal justice action plan was published in December 2022.

We think that the co-ordination of activity between different organisations in Wales is a good model. The ambition in Wales is to have better and more accurate data that will be put to better use and support improvements in performance. This model is therefore well-placed to lead to improvements in the collection and publication of data when the action plan has been implemented.

National Police Chiefs’ Council

The National Police Chiefs’ Council (NPCC) is the co-ordinating body for police forces in the UK. It is organised into committees that each deal with a broad area of policing. Reporting to each committee are various portfolios that have responsibility for different themes. Each portfolio is led by a different senior officer.

In the same way that the criminal justice decision points in our inspection don’t fall neatly within the remit of one Government department, they also don’t correspond directly to one NPCC portfolio.

There is no one officer who holds portfolio responsibility for disparity in criminal justice decision-making. This is because this matter is relevant to both the Criminal Justice Committee and the Diversity, Equality and Inclusion (DEI) Committee.

The DEI Committee has responsibility for the NPCC’s diversity, equality and inclusion strategy, action plan and associated toolkits. We describe these further below.

Within the Criminal Justice Committee, there are four different portfolio leads with responsibility for the decision points we considered in this inspection. In addition, the NPCC custody portfolio lead would also retain an interest in some aspects of criminal justice activity that takes place within a custody environment. This is because most of the activity relating to the decision points takes place in police custody centres.

In our race and policing leadership and governance review we highlighted that the complicated NPCC structure is unhelpful when race overlaps many different portfolio areas.

We acknowledge that policing can be very complex, with many connections between different elements. And we didn’t expect to find one single NPCC lead for disparity in criminal justice decision-making. But while we found that NPCC portfolios had been recently working together more effectively (particularly within the NPCC DEI Protected Characteristics in Policing Systems Working Group) the complexity of the working arrangements could create barriers to progress.

We believe that relevant activity could be better co-ordinated within the new NPCC diversity, equality and inclusion strategy and its associated documents. The different elements of police interactions with people from ethnic minority backgrounds could be brought together more coherently. This would help the police better understand the use of all relevant powers and the links between them.

Police Diversity, Equality and Inclusion Strategy

The ‘NPCC’s diversity, equality and inclusion strategy’ was published in 2018. The strategy says that it provides “a framework to accelerate progress on diversity, equality and inclusion, the delivery of positive equality outcomes and to ensure legal compliance”.

It has three main elements: “our partners”, “our organisation”, and “our communities”. There are three associated toolkits. These toolkits suggest actions for forces to consider adopting. We consider how forces have responded to this strategy and the relevant suggested actions later in our report.

The two elements of the strategy most relevant to this inspection are “our communities” and “our partners”.

“Our communities” has two relevant commitments: “Understanding our communities” and “Engagement and good relations”. The commitments in this element of the strategy are to do with gathering and analysing information so forces can better understand how their actions affect the communities they serve. And then work with these communities in a transparent way.

“Our partners” has three pertinent commitments: “Understanding the partnership landscape”; “Tackling disparity”; and “Joint service delivery”. These commitments are to do with recognising the need for collaboration between the police and the organisations they work with. And the need for these partners to work together to identify and understand disparity, and reduce and/or eliminate it where possible.

All the above commitments are directly relevant to the police’s role in making criminal justice decisions. But, as we discuss below, progress on the collection, analysis and scrutiny of data on race disparity in police aspects of the criminal justice system has been very limited.

The three associated NPCC DEI toolkits contain suggested activities for forces.

In the “Our communities” toolkit, one workstream under the “Understanding our communities” commitment is:

“Equality data on policing performance should be analysed to identify and address any areas of disproportionately [sic] and negative impact on diverse communities”.

There are four associated actions for forces to consider. These relate to making sure that communities are involved in the scrutiny of data and that data is explained and published.

It isn’t explicitly stated what information forces should be gathering, analysing, scrutinising and publishing. But it is a reasonable expectation that information relating to ‘policing performance’ would include data on at least some of the criminal justice decisions included in this inspection.

Another workstream, this time under the “Increasing confidence” commitment, is for “forces to undertake meaningful involvement and consultation with their communities”. There is a suggested action to involve communities in the scrutiny of coercive powers such as stop and search, use of force and strip searches. But there is no mention of the scrutiny of criminal justice decisions.

In the “Our partners” toolkit, one workstream under the “Tackling disparity” commitment is:

“The Police is the entry into the criminal justice system, as such forces to undertake analysis to identify any disproportionately [sic] in people brought into custody or dealt with by out of court disposal.”

But the toolkit only contains information about collecting data on arrests and out-of-court disposals. So the toolkits aren’t comprehensive enough to help forces understand what other criminal justice data they should be collecting.

In summary, the strategy and toolkits convey a clear intent and expectation that forces should gather, analyse and scrutinise demographic information about some aspects of the criminal justice process. But the strategy and actions give insufficient detail, and aren’t specific enough, to achieve this.

As part of our inspection, we asked all forces in England and Wales to complete a survey on their approach to race disparity in criminal justice decision-making. The full results are detailed at Results from our survey of all forces.

One of the questions we asked forces was: Does the force have an action plan that reflects the NPCC 2018–25 diversity, equality and inclusion (DEI) toolkits? Of the 43 police forces in England and Wales, 38 said they did.

In our inspection fieldwork we looked more closely at the action plans of the five selected forces. Only two of the five forces had a DEI action plan that incorporated the criminal justice elements of the NPCC toolkits. One other force had a Disproportionality in Criminal Justice Improvement Strategy with associated actions.

So we aren’t convinced that, even the limited actions relating to criminal justice race disparity contained in the NPCC DEI, toolkits have been incorporated into all police force DEI action plans. We also found in our race and policing leadership and governance review that the ‘NPCC diversity, equality and inclusion strategy 2018–25’ and its associated toolkits didn’t receive widespread support from forces.

Police Race Action Plan

In May 2022, the NPCC published a new Police Race Action Plan (PRAP) for consultation. The stated aim of the plan is to:

“Give officers the tools they need to build trust and confidence with Black communities, so that they are better equipped to challenge racism and to identify and address any ingrained cultural biases that may be fuelling racial disparities across policing.”

The published plan contains four workstreams, with the aim of building an anti-racist police service.

Most relevant to this inspection is workstream two – “Not over-policed (use of powers)”. There are eight associated actions, but none relate to criminal justice decisions. The workstream mainly relates to stop and searches, and similar initial interactions with the public.

Appendix B of the PRAP includes details of work that should take place to improve aspects of data and performance. The NPCC told us that this included work to achieve consistency across the methodology applied to proportionality calculations.

We were also told that the PRAP programme team were seeking opportunities to include race and ethnicity information in a future iteration of the criminal justice system (CJS) scorecard to improve police transparency.

None of this work is specifically linked in the PRAP to the need to improve information about police criminal justice decision-making. So we are concerned that the current version of the PRAP presents a missed opportunity. We will work with the NPCC to refine the actions in the plan so decisions made about criminal justice outcomes are given equal importance to stop and search, and the use of force.

In addition, we know the NPCC intends to review and refresh the DEI strategy and its associated toolkits. We welcome this development as a means of supporting forces in gathering and analysing demographic information about some important interactions with the public.

We will continue to work with the NPCC DEI lead to make sure the findings from this inspection are considered in the new strategy and associated documents.

The College of Policing

The College of Policing (the College) is responsible for developing written guidance for the police on a range of topics, in the form of authorised professional practice (APP).

The College doesn’t give any specific guidance about recording ethnicity on police systems and why this is important. We cover the training provided by the College, and what we were told by officers about this, in the Training and guidance chapter.

In our race and policing leadership and governance inspection we said that the College should establish national standards for police diversity, equality and inclusion training. The College told us it has started work on this and is working closely with the NPCC. We will follow this work closely to make sure that it develops with the required urgency.

The Association of Police and Crime Commissioners

The Association of Police and Crime Commissioners (APCC) is the organisation that offers support and guidance to police and crime commissioners (PCCs). This includes facilitating governance structures.

In 2020 the APCC published a ‘Race Disparity Working Group Action Plan 2020/21’.

For the APCC criminal justice portfolio, the plan set out three priorities and five actions. Two of the desired outcomes were:

• tackling race disparity to become a priority for the criminal justice system (CJS) nationally; and
• more and better-quality data published locally and nationally regarding ethnicity in the CJS.

The APCC 2021/22 ‘Race Disparity Working Group Action Plan’ also contained relevant actions for PCCs to carry out to improve the understanding of race disparity in criminal justice decision-making.

In 2021, the APCC signed a data sharing agreement with the Youth Justice Board (YJB) to allow it to access information in the YJB’s summary disproportionality tool. We discuss this tool further, below. When launching the agreement, YJB chair Keith Fraser said:

“With access to the summary tool, PCCs will now be better able to interrogate the evidence and reveal whether children from ethnic minorities are over-represented in their area. And if the evidence points that way, then there is huge potential for PCCs to forge new alliances with local community-based groups and help to build trust, as they seek to explain ‘why it is happening’ and ‘what can be done about it’.”

The APCC publishes a race disparity toolkit twice a year. This collates publicly available data relating to race disparity and information on how to access other data sources, such as the YJB summary disproportionality tool. The toolkit also highlights examples of the ways that PCCs are tackling race disparity in their own areas.

It is right that PCCs carry out this activity because they have a public sector equality duty to do so. PCCs also have a role in holding chief constables to account for how they carry out policing activity.

There is an obvious contrast between the leadership and direction shown by the APCC in this regard, and that of the NPCC and College of Policing. At the time of our inspection, the APCC’s work in relation to criminal justice disparity was clearer, more specific and more up to date than that of the NPCC. But as we have pointed out in our race and policing leadership and governance review, PCCs have an additional opportunity to prioritise race and equality in their police and crime plans.

Findings from our fieldwork

Results from our survey of all forces

In February 2022, we asked all police forces in England and Wales to complete a short survey containing 13 questions. As we discussed above, we used the results from this to help us design our inspection and choose which forces to carry out fieldwork in. We include some references to the survey in other chapters where relevant.

We describe the main survey responses below, with some commentary.

We wanted to know whether forces had recognised their public sector equality duty and how this duty relates to police criminal justice decisions.

• 24 of 42 forces have strategies, policies and actions for criminal justice processes that include the need to take account of the public sector equality duty.
• 22 of 42 forces have a strategy to explain, where necessary reduce, and where possible eliminate, disparity within the criminal justice system.

As we have discussed above, there are no policies or guidance on the need to understand disparity in criminal justice matters. So it isn’t surprising that more than 40 percent of forces haven’t recognised its importance.

We consider that forces should routinely gather, analyse and scrutinise information relating to criminal justice disparity so that they understand whether any disparity exists. These processes should be linked to existing policies and guidance that detail what is done, by whom and when. We asked forces about their approach.

• 22 of 42 forces have policies and guidance that reflect the need to gather information about, scrutinise and understand disparity in criminal justice processes.
• 30 of 42 forces gather information and data about disparity in criminal justice decision-making.
• 28 of 42 forces have processes in place to monitor any disparity in decision‑making on a force level.

Most forces told us they gathered and monitored at least some data relating to disparity in criminal justice decisions. But, in some cases, it seems these processes aren’t linked to any force policies and guidance. Without this clear link, there could be a lack of clarity about what information is being gathered and why.

The survey was designed to be quick and easy for forces to complete. We didn’t intend the survey to be a comprehensive review, so we didn’t give forces the opportunity to give detailed responses. Because of this, the survey results should be treated with caution. It is likely that the results reflect an optimistic assessment of the situation, and that the true picture in forces is less reassuring than some of the survey responses suggest.

For example, we asked forces whether they gathered information and data about disparity in criminal justice decision-making. We weren’t prescriptive about what information this was, and we didn’t include for reference all the relevant decision points. So if the force gathered information about one of the decision points but not the five others, this may still have generated a positive response.

In our fieldwork, we found that most forces gathered some information about cautions. But we didn’t find any forces that gathered information about all the decision points included in our inspection. So it is possible to conclude that in many forces there is insufficient focus on, and activity relating to, race disparity in criminal justice decision‑making.

Analysis of data from forces

Following the survey, we asked the six fieldwork forces to give us information about a range of offences and decisions made in relation to them. From a total of 22,624 relevant records provided to us, we randomly selected 1,470 for review. This was calculated to be a statistically significant sample, representative of the volumes of each decision point submitted by forces, and the ethnicities of the people dealt with by the police.

Data on arrests and voluntary attendance

In our analysis of data provided by the fieldwork forces, we examined rates of voluntary attendance and arrest for certain offences. (See Annex A for an explanation of relative rate indices we used in our analysis.)

For the offences of possession of an offensive weapon and criminal damage, the arrest rate for Black people was 33 per 10,000 population. In contrast, the arrest rate for White people was 14 per 10,000 population. For Asian people, and people from Mixed and Other backgrounds, the arrest rates were 7, 21 and 10 people per 10,000 population, respectively.

For the same offences, the decision to use voluntary attendance as opposed to arrest also varied. The rate was:

• 5 per 10,000 population for Black people;
• 4 per 10,000 population for White people;
• 2 per 10,000 for Asian people;
• 5 per 10,000 for people from Mixed backgrounds; and
• 2 per 10,000 for people from Other backgrounds.

Data on outcomes following arrest or voluntary attendance

We also examined the likelihood of a person being assigned one of four outcomes after arrest or voluntary attendance. These outcomes were community resolution, caution, charge and no further action.

Our analysis suggested that, after arrest, Asian people were less likely to receive a caution than White people (with a relative rate (RR) of 0.7), but 1.3 times more likely to receive a community resolution. Black people were less likely to receive a community resolution than White people (an RR of 0.6). People from other ethnic groups were less likely than White people to be charged (an RR of 0.8), and 1.3 times more likely to receive no further action.

After voluntary attendance, compared to White people, Asian people were 2.1 times more likely to be cautioned and 1.3 times more likely to be charged. They were less likely to receive no further action (an RR of 0.8). Black people were 1.3 times more likely to be cautioned than White people, but less likely to be charged (an RR of 0.7) and less likely to receive no further action (an RR of 0.3).

These results should be viewed with caution. This is because we only examined a small number of cases for specific combinations of decision point and ethnicity. Nonetheless, they indicate that more work needs to be done to understand why such differences exist. The results of this work should be published.

Data on charging decisions

The information we asked for included data about cases where the police had made a decision to charge a suspect with an offence. The forces provided this, but most of them told us that this was difficult and time consuming to do.

When an investigation is concluded, forces assign a crime outcome type to the crime record. For charges, this is outcome type one. But this doesn’t differentiate between whether it was the police or the Crown Prosecution Service (CPS) that made the decision. This is unhelpful because it makes it more difficult for forces and the CPS to understand whether the police have complied with guidance published by the Director of Public Prosecutions. This guidance lists which offences forces can charge without referral to the CPS. For example, the police shouldn’t make charging decisions about cases involving domestic abuse.

Because the crime outcome type for charges doesn’t allow forces to understand what cases are being charged by which organisation, it is hard for forces to gather information about, and analyse, whether any disparity exists in these cases.

We have found in previous inspections that the police sometimes incorrectly charge offences without referral to the CPS. This can cause problems for the CPS as it may need to change the charge to a different offence if it considers other charges are more appropriate. Without good ways to identify the nature and scale of the problem, it is difficult for organisations to learn and make improvements.

The Home Office has told us that a review is taking place into the use of the crime outcome types. This review should consider the findings in this report.

Analysis of our case file review

How decisions were recorded

In the case file review, we examined whether, and how well, the police recorded the rationale for the criminal justice decisions. A rationale is the police’s record explaining why decisions are made and the other options considered. It is important that the rationales for criminal justice decisions are recorded so that:

• supervisors can check whether decisions made are fair, proportionate and legitimate;
• a force can more easily quality assure the decisions and the work of officers and their supervisors;
• forces can analyse and understand any trends that emerge and act on them quickly; and
• where scrutiny processes exist, partners can more easily understand the decisions and the reasons for them.

We were disappointed to find that a rationale for the criminal justice decisions we examined was recorded in only 75 percent of the 1,470 cases we reviewed.

The proportion of cases where there was no recorded rationale varied depending on the type of decision-making point. This is shown in the table below.

Table 2: Proportion of cases reviewed with no recorded rationale by decision-making point

Decision point	Percentage of cases with no recorded rationale
Arrest	2.2
Voluntary attendance	73.2
Community resolution	12.1
Caution	9.4
Charge	13.3
Post-charge bail	29.3

In other cases, the rationale was recorded. But, among these, we found that in just over half (55 percent), there was insufficient rationale recorded to understand why the decision was taken. (As we discussed in the Methodology section, we didn’t set out to examine whether or not the police’s decisions were in our view correct.)

Again, there were differences depending on the type of decision-making point.

Table 3: Proportion of cases where the recorded rationale didn’t justify the decision, by decision point

Decision point	Percentage of cases where recorded rationale didn’t justify the decision
Arrest	8.5
Voluntary attendance	63.8
Community resolution	85.2
Caution	87.3
Charge	76.8
Post-charge bail	28.4

Overall, we found no evidence that the decision had been supervised in 62.7 percent of cases we reviewed. This is a very disappointing level of supervision being applied to criminal justice decisions, and is in keeping with a general finding of low levels of supervision in relation to various areas of policing, as noted in HM Chief Inspector’s State of Policing annual report. This finding is worrying because supervision is the best way for forces to make sure that decisions are fair.

In previous inspections, we have similarly found problems with the recording and supervision of rationales for criminal justice decisions. In 2015 we carried out a joint inspection with HM Crown Prosecution Inspectorate of the provision of charging decisions. We found that the quality of some criminal justice decisions was inconsistent, and that the decisions were sometimes not properly supervised.

Inconsistencies in how decisions were recorded

We also examined whether recording practices varied depending on the ethnicity of the person subject to a decision.

Taking all the decisions together, there was little difference in whether a rationale had been recorded. But for arrests and voluntary attendance it was less likely that a rationale was recorded for the decision if the person was White (64.0 percent) than if they were from an ethnic minority background (73.5 percent).

We don’t know the reasons for this difference. But this is a good example of the need for forces to do more work to understand why there is such a variation, and to use the Lammy review’s principle of ‘explain or reform’ (see Background section) to reassure communities.

We considered the quality of the rationale recorded in two ways: whether the rationale justified the decision made, and whether it set out what other options were considered.

When we considered all the decisions together, we found that the quality of the recorded rationale was generally lower for White people than for those from an ethnic minority background. The recorded rationale was insufficient to understand why the decision was taken for White people in 54.3 percent of cases, compared to in 7.2 percent of cases for people from ethnic minority backgrounds.

There were also variations between decision points. For example, we found that for arrests and voluntary attendance the recorded rationale was sufficient to understand why the decision was taken in 47.1 percent of cases we reviewed for White people, compared to 65.7 percent for those from ethnic minority backgrounds. So in this respect it was more likely that decisions were better recorded for those from ethnic minority backgrounds.

When we looked at whether the recorded rationale for the decision considered all available other options, we found different results. When we considered all the decisions together there were no differences between people from different ethnic backgrounds. And when we looked specifically at arrests and voluntary attendance, it was much more likely that the recorded rationale considered all options available for White individuals (60.3 percent of cases), compared to those from an ethnic minority background (20.8 percent).

In our fieldwork, we also carried out interviews and focus groups with officers. All the officers we spoke to told us that all their decisions were based only on the facts of the case. Some decision-makers told us that sometimes they wouldn’t be aware of the ethnicity of the person they were making the decision about. For example, the decision to charge is sometimes taken by officers that have had no direct contact with the person concerned.

Conclusion

Our analysis of data from forces found that there were clear differences in the recording of the rationale for decisions between the experiences of those subject to police criminal justice decisions depending on their ethnic background.

Using the framework suggested by the Commission on Race and Ethnic Disparities, these would be classified as unexplained racial disparities.

Data collection and analysis carried out by fieldwork forces

Race disparity analysis carried out by fieldwork forces

We asked the five fieldwork forces to tell us which of these decision points they gathered ethnicity information about. We then asked whether this information was subject to any analysis to determine whether there was evidence of race disparity. We have shown the results below.

• Arrests – all forces gathered ethnicity information relating to arrests, and all carried out some form of analysis of this.
• Voluntary attendance – only one force gathered ethnicity information relating to voluntary attendance, and didn’t carry out analysis of this.
• Community resolution – only two forces gathered ethnicity information relating to community resolutions, and they didn’t carry out analysis of this.
• Cautions – four forces gathered ethnicity information relating to cautions, and all four carried out analysis of this.
• Police charging decisions – only two forces gathered ethnicity information relating to charging decisions. Both forces carried out analysis of this. But neither of these forces routinely split this data between decisions made by the police and the CPS. For the reason we explained earlier, most forces found it too time-consuming for them to do.
• Bail decisions after charge – only two forces gathered any ethnicity information relating to decisions to bail or not after charge, and of these only one carried out analysis. An additional force gathered and analysed data on all bail granted, but didn’t divide this into different types of bail.

We believe that the inconsistent and incomplete activity we found in the fieldwork forces is likely to be replicated in most – if not all – forces.

Examples of analysis by fieldwork forces

One force we inspected had, through an analysis of its data, determined that people from ethnic minority backgrounds were less likely to receive an out-of-court disposal (OoCD):

“Cautions remain lower for VEM (visibly ethnic minority) subjects both over the last year and for Q4 (quarter four).

Black subjects have seen the greatest increase in the use of OOCD over the last 12 months (was also higher in Q4). However, all ethnicities see lower use of OOCD when compared to the rate of arrest, when compared to White British subjects.

All other ethnicity groups see OOCD used around half to a third less often per arrest than White British subjects over the last 12 months.”

These findings are consistent with the picture across England and Wales we have described above at Published data.

In a different force, analysis produced for a force-wide governance board chaired by a senior officer concluded:

“Black & Mixed [heritage] detainees are more likely to be charged while in custody than White DPs [detained persons].

25.9% (70) of Black detainees were charged, and 20.5% (25) [of] Mixed [heritage detainees], compared to 18.1% (849) of White detainees. The percentage of Black detainees charged has increased since last quarter, and [the percentage of] White detainees has decreased.

Once charged, Black detainees are more likely to be remanded in custody [than White detainees].

48.6% (34) compared to 43.3% (368).”

We didn’t find any evidence to suggest that this force had taken any action because of this analysis to seek an explanation for the apparent disparities.

Another fieldwork force gave us a table containing some ethnicity data it had gathered about people detained in custody. This showed the proportions of people from different ethnic backgrounds who had been refused bail after charge.

The table showed that Black detainees were more likely than White detainees to be refused bail after charge. The same was true of detainees of Mixed heritage. But those of Asian origin were less likely than both White and Mixed heritage detainees to be refused bail after charge.

In the table, ethnicity was recorded as “not stated” for a substantial proportion of records (more than 20 percent). This is unhelpful. We discuss this further in The problem of missing ethnicity data section.

We didn’t find any evidence that the force concerned had carried out any further analysis of the apparent disparity in outcomes to find the reasons for it. Nor did it share the data with the community, such as through a forum that could scrutinise the force and hold it to account.

Specific force analyses about young people

Two of the five forces we inspected had carried out analyses of some aspects of criminal justice disparity for young people. In both instances, this was part of wider work with other organisations.

One analysis report concluded:

“The indicators would suggest [force name deleted] is disproportionate in relation to Ethnic Minority children in relation to custody and criminal justice.”

The report also stated that caution should be exercised when interpreting the conclusions, due to the small sample size.

Another force had carried out analysis of outcomes for children and young people who had been taken into police custody:

“Where a C/YP from an ethnically diverse community is detained they are more likely to receive a charge outcome than a White counterpart (20.2% vs 12%). Also less likely to receive an NFA [no further action] outcome (49.3% vs 54.2%). However, it should be noted that the data set is still in motion [development] and a sizeable number of those detained remain on bail or RUI [released under investigation] which could affect the final outcome figures.

Where a C/YP from an ethnically diverse community is detained they are less likely to receive an out of court disposal (OOCD) when compared to those of a White background (8.0% vs 12.4%).”

We couldn’t tell from these analyses whether outcomes for young people in these forces were more disproportionate than for adults. The analysis carried out by each force would have been enhanced by such comparative research. Similarities or differences could then have been explored further to help the force learn more about how it acted towards people from different ethnic minority backgrounds.

The Youth Justice Board publishes data about outcomes for young people. We discuss this further in the Published data section. Its analysis corroborates the findings from the two fieldwork forces. That is, that there is disparity in some criminal justice outcomes for young people from ethnic minority backgrounds.

Conclusion

For some of the decision points we examined, we found that some forces had carried out an analysis of disparity. It is encouraging that this activity is taking place. But, without any further work to try to explain the disparities and then take action to address the causes where possible, it has limited effect.

It is clear from our fieldwork that, where analysis in forces had taken place, in each of those forces there was evidence of disparity. But in most cases, there had been little, or no, work carried out to try and explain this. When evidence of disparity exists, but no further work is done to try and explain it, the police can miss opportunities to engage effectively with the public and understand the potential implications for local policing. This in turn can lead to assumptions that may damage public confidence.

As we discuss above, the Lammy review made a relevant recommendation:

“If CJS agencies cannot provide an evidence-based explanation for apparent disparities between ethnic groups then reforms should be introduced to address those disparities. This principle of ‘explain or reform’ should apply to every CJS institution.”

We would like to see all forces adopt the principle of ‘explain or reform’ to the criminal justice decisions they make.

We are also aware that in some forces that we didn’t inspect work has taken place (sometimes with criminal justice partner organisations) to try to understand different outcomes for different ethnic groups. This is also encouraging. But, as in our fieldwork forces, not all the decision points in our inspection have been included in this work to date, so it necessarily presents an incomplete picture.

We weren’t told there were insufficient resources available to carry out research into disparity. Indeed, in one force we inspected there was a dedicated analytical criminal justice disparity resource. It was also clear that all the forces we inspected recognised that race disparity was an important issue. So we have concluded that the lack of analysis of the reasons for race disparity is probably connected to the lack of both local and national impetus and direction, as well as competing demands.

While some forces have carried out some analysis of race disparity, there doesn’t seem to be any co-ordination of this work by the National Police Chiefs’ Council (NPCC). This could include setting standards about how best to gather and analyse information and sharing information about the results.

We have made a recommendation elsewhere in this report for the NPCC lead for criminal justice to work with the College of Policing on such co-ordination and standard-setting. Specifically, this would include establishing ways for forces to gather, analyse, scrutinise and publish information on criminal justice race disparity.

The Youth Justice Board summary disproportionality tool

The Youth Justice Board (YJB) produces a summary disproportionality tool. This is a way for Youth Justice Services (YJS) to collect demographic information about outcomes for young people in the criminal justice system.

The summary disproportionality tool is a Microsoft Excel spreadsheet that provides information about how members of all ethnic groups were represented in the youth justice system in England and Wales. It aims to support YJSs in addressing any issues of over-representation of different ethnic groups in their local youth justice systems.

The tool should be a valuable resource for the police. Forces should use it to analyse how their actions affect young people and help them understand whether some criminal justice decisions may be disproportionate.

So we were surprised to find that none of the forces we inspected used the tool when analysing data on criminal justice outcomes. Force leads we spoke to seemed unaware of the tool’s existence.

His Majesty’s Inspectorate of Probation (HMIP) is responsible for inspecting YJSs. Since 2018, HMIP has seen variable use of the tool within some YJSs’ overall approach to addressing disproportionality.

HMIP has found that some YJSs have used the tool to gather data and then develop plans and actions to address the identified areas of disproportionality. Others have used it to gather information, but then not translated this into a meaningful action plan. Several other YJSs have used other methods to address disproportionality, and some have done nothing to identify areas of disproportionality.

HMIP has concluded that the use of the disproportionality tool varies between YJSs. This variability extends to YJS management boards, some of which are more proactive than others in adopting a data-led approach.

In Wales, at the time of our inspection, the tool hadn’t been considered as part of the data-mapping process to understand and develop the Criminal Justice Anti-Racism Action Plan for Wales.

In contrast, as we outline above, in its 2020/21 action plan the APCC advocated that police and crime commissioners (PCCs) make use of the tool. The APCC also later signed a data-sharing agreement with the YJB and continues to suggest PCCs use the YJB tool to hold chief constables to account for any apparent disparity. But as it isn’t within our remit to inspect PCCs, we don’t know how many do this.

The problem of missing ethnicity data

In our inspection reports we have frequently stated that police forces should make improvements to the way they gather data on ethnicity.

For example, in our 2018 hate crime inspection we concluded that the police didn’t pay enough attention to gathering ethnicity information about perpetrators.

In our rolling programme of custody inspections we often find forces don’t have accurate information about detainees’ ethnicity. Custody officers don’t always ask detainees to self-define their ethnicity. This means the information gathered is incomplete and not accurate enough to assess whether treatment of detainees is fair.

In 2022 we published a response to a super-complaint made by the Tees Valley Inclusion Project. We concluded that police-recorded data was too patchy, poor‑quality and inconsistent for us to be able to compare the experiences of people of different ethnicities.

The above findings, and this report, specifically relate to the recording of ethnicity about people by the police because they had been arrested or suspected of committing offences. But in other reports we have made similar comments and recommendations about the recording of ethnicity in different aspects of policing. The findings in this report are part of a wider problem, which we cover in more detail in our race and policing leadership and governance inspection.

As we have said above, the police rely on the confidence and trust of the public to carry out their duties effectively. The police need to demonstrate that their use of powers and decision-making is fair. So gathering and analysing data about the ethnicity of people dealt with by the police is important. And this data should be as complete and accurate as possible.

As discussed above in the Methodology section, in the inspection fieldwork the 6 forces gave us information about 22,624 cases. We found that in 2,779 of those cases (12.3 percent), the force hadn’t recorded the ethnicity of the person involved.

In addition, the information available about the ethnicity of the person concerned varied between the types of case, as we have shown in the table below:

Table 4: Proportion of cases where self-defined ethnicity was unknown by decision point

Decision point	Percentage where self-defined ethnicity was unknown
Arrest	3.3
Voluntary attendance	13.9
Community resolution	35.1
Caution	8.5
Charge	12.1
Post-charge bail	5.8

We don’t know why there is such a substantial variation in the recording of ethnicity and such a large absence of recording the ethnicity of people receiving a community resolution. As a community resolution is a less formal outcome, it is possible that police officers consider it less important to gather accurate information on them.

In our fieldwork, we also looked in more detail at the information gathered by individual forces.

In one force, a report on the use of voluntary attendance found:

“Of the 128 PDE (person defined ethnicity) assessments, 32 were Blank and 30 were recorded as Unknown. This suggests that officers may not place much importance on getting this professional assessment correctly recorded.”

In another force, a report into criminal justice outcomes said that for 50 percent of cases the ethnicity wasn’t recorded or was recorded as “not stated”.

In a different force, the minutes of an out-of-court scrutiny panel meeting recorded:

“The panel reviewed the latest out of court disposal data. Concerns were raised in relation to the outcome 8 (community resolution) ethnicity data. Ethnicity was not recorded in 58 cases out of 186.”

When there are such large numbers of records where the ethnicity of the person concerned isn’t recorded, the worth of any subsequent analysis is questionable.

In our opinion there are several reasons why ethnicity may not be recorded at all, or recorded as “not stated”:

• Officers may be busy and could forget.
• Officers may not recognise the importance of recording ethnicity and see it as overly bureaucratic.
• Officers may not feel comfortable asking the person for their self-defined ethnicity or making their own judgment about the person’s ethnicity.
• In some situations, such as violent incidents, it may not be appropriate to ask a person for their self-defined ethnicity.
• Some people may not want to tell the police about their ethnicity.

Some of these reasons apply to both the recording of officer-defined and self-defined ethnicity data fields. Some reasons relate to self-defined ethnicity information, but can’t explain why officers don’t record their own assessment of the ethnicity of the person they are dealing with.

A contributory factor is that in the forces we inspected, the ethnicity data fields weren’t mandatory. If officers couldn’t move on from the record until both the self-defined ethnicity and officer-defined ethnicity fields had been completed, this would make a big difference to the number of crime records with absent ethnicity fields.

Of the forces we inspected, five used the Niche crime recording system, and one used a system bespoke to the force. This latter force plans to transfer to Niche in the future.

Of the 43 forces in England and Wales, 28 either currently use Niche or plan to do so in the future. The remaining forces use a variety of different systems, but the intention is that these will be brought together under one system called Connect. Each system has a business user group, where forces are represented and changes to systems are approved. So bringing consistency to changes to force record management systems should in future become easier.

In this inspection we found that despite previous recommendations we have made about the need to record accurate information, little has changed. In all the forces we inspected there were large gaps in the data, where the ethnicity of the person concerned was either “not stated” or not recorded. The situation is unsustainable, and more urgency and co-ordination are now required.

Making changes to police systems can be complicated, expensive and time‑consuming. It is also sometimes difficult for individual forces to make changes to IT systems, which will have implications for other forces that use the same system. For these reasons, some of the forces we inspected had created work-arounds to improve their recording of ethnicity information, which could themselves be costly and inefficient.

We consider part of the solution to the problem of missing ethnicity data is to make the self-defined ethnicity field on police record management systems mandatory. This would make it harder for officers to avoid recording this important information. It would also remove the need for forces to make changes to other systems to achieve the same aim.

Even if the self-defined ethnicity field became mandatory, supporting activity would also still need to take place. For example, there would need to be good supervision to make sure that there were good reasons if officers recorded that the person concerned didn’t state their ethnicity.

And supervision needs to be backed up by effective force quality assurance systems and processes. This could include dip-sampling to make sure that relevant procedures have been followed and that the decision has an adequate rationale recorded.

Police officers also need to be given the necessary training and guidance. In our Training and guidance chapter, below, we describe what guidance and training is given to police officers to help them understand the importance of recording ethnicity.

The NPCC diversity, equality and inclusion portfolio has formed a protected characteristics working group. One of the purposes of the group is to review how protected characteristics, including ethnicity, should be recorded. The working group includes members from various NPCC portfolios, the College of Policing and the Home Office. This is an encouraging step, and we consider that this group can certainly help to co-ordinate the changes required.

Demonstrating legitimacy

Research by the College of Policing has shown that for the police to be considered legitimate by the public, people need to believe that the police will:

• treat them with respect;
• make fair decisions and take the time to explain these decisions; and
• be friendly and approachable.

The research also indicates that when officers behave fairly, this can encourage greater respect for the law, foster social responsibility and lead to people co-operating with the police. This is often called procedural justice.

Police forces can demonstrate legitimacy in criminal justice decision-making by showing they are examining whether decisions have been appropriately made and haven’t been subject to any bias. When this is done by working with people who are independent of the police it is more likely to reassure the public, particularly people from ethnic minority backgrounds.

Publishing the results of any scrutiny of cases, or analysis of any data, means these are widely available. This transparent approach is more likely to give the public confidence in the legitimacy of the police. The reverse is also true. A lack of transparency is likely to lead to a lack of trust and confidence in the police.

In this inspection, we examined how well forces scrutinised criminal justice decision‑making. We also looked at whether forces published information about this scrutiny and/or wider information about disparity in their criminal justice decision‑making.

Scrutiny

In the context of this inspection, by scrutiny we mean the involvement of local communities and partner organisations in the examination and investigation of both individual cases and higher-level data on criminal justice decision-making.

There appears to be a consensus among Government, the police and local communities that scrutiny of police decisions and use of powers is a good thing and should be encouraged.

In 2002, with the introduction of the Association of Chief Police Officers policy of recording self-defined ethnicity, one of the stated aims of gathering information was:

“Sharing local data with local community representatives and groups to inspire confidence through transparency and openness. Discussions arising from sharing facts with community partners may help to dispel false perceptions and assist in identifying explanations for data and gain shared ownership of proposals for action.”

The College of Policing doesn’t give guidance on how scrutiny of criminal justice decision-making should take place. But it does give detailed guidance on the scrutiny of stop and search powers.

In the context of stop and search, the College states that independent scrutiny opens “practices to communities for close examination, with a view to them providing constructive oversight, dialogue and challenge”. The same should also be true for criminal justice decisions.

Later in this chapter, we discuss the Commission on Race and Ethnic Disparities’ March 2021 report and the Government response to this, Inclusive Britain, which was published in March 2022. Both publications detail the importance of scrutiny of the police’s use of powers.

In our inspection, we examined what scrutiny of criminal justice decisions took place in our five fieldwork forces.

Out-of-court disposals (cautions and community resolutions)

All five forces had scrutiny panels set up to examine out-of-court disposals (OoCDs), but only two of these considered community resolutions as well as cautions. One of the scrutiny panels only examined cautions where conditions had been attached. This meant that community resolutions and cautions without conditions were excluded.

So, while it is positive that all the fieldwork forces had scrutiny panels, some types of OoCDs weren’t included in the scrutiny process.

The National Police Chiefs’ Council (NPCC) and the Ministry of Justice published OoCD scrutiny panel guidance[3] in 2019. This sets out good practice for how panels should operate. The guidance isn’t mandatory, but it states principles so forces can show they are accountable for decisions, and that the decisions made are justifiable. We found that not all panel chairs were aware of this guidance.

The guidance focuses on the examination of individual cases and there is no mention of the scrutiny of higher-level data, particularly on race (or other) disparity. This is a missed opportunity to make best use of a forum that usually includes influential and representative community members.

Despite this, in one of the forces we inspected we found the scrutiny panel was discussing force-wide data. In another force, the panel was given high-level data but no discussion about it took place.

The guidance states that cases should be selected at random. But in our inspection, we found that in two forces the cases weren’t randomly chosen for the panel by the force. Instead, the force chose cases it thought were worth the panel considering. We didn’t find any evidence that this was done to avoid providing cases that didn’t show the force in a good light. But it could undermine the principle of the independence of the panel process. And it is unlikely to reassure the community that the scrutiny process is open and transparent.

The guidance also says:

“Panels should ensure that there is examination of disproportionality with respect of OOCDs issued to both BAME (Black, Asian, Minority & Ethnic) communities and females. This is to ensure greater transparency and trust that these are issued appropriately.”

The guidance suggests that these themed reviews take place at least annually. As the table below shows, this had only taken place in two of the five fieldwork forces.

In one force, the scrutiny panel chair told us the force frequently didn’t provide information about the ethnicity of the people in the cases chosen to be scrutinised. This was because this hadn’t been recorded by the officer that had dealt with the case.

We noted that in two of the five forces, the scrutiny panel was chaired by a member of the Office of the Police and Crime Commissioner (OPCC). The guidance says:

“To maximise transparency and public confidence, the panel should, wherever possible, elect an independent (i.e., non-PCC or police force) chair. Where this is not possible, the panel may be chaired by a representative of the police force. It would not, however, be appropriate for a PCC to chair a panel.”

We don’t agree with the guidance that a police chair is preferable to an OPCC chair. We think it is likely the public would consider there to be more transparency if a member of the OPCC chaired the panel rather than the police. We spoke to one police and crime commissioner (PCC) who told us they saw the scrutiny panel as an integral part of their role in holding the local chief constable to account. We make a recommendation below about how other changes to the guidance could be made.

The table below shows how scrutiny arrangements varied in the five fieldwork forces.

Table 5: Variations in out-of-court disposal scrutiny arrangements in fieldwork forces

Force	Chair	Adult and youth cases?	Frequency	Race-themed reviews?	Community resolutions included?	Number of cases per meeting
Force one	Independent (judiciary)	Yes, by rotation	Half-yearly	No	No	10
Force two	OPCC	Yes, at same meeting	Every two months	Yes	No	10
Force three	Independent (judiciary)	Yes, at separate meetings	Quarterly	Yes (but not within past 12 months)	Yes	10
Force four	OPCC	Yes, at same meeting	Quarterly	No	Yes	20
Force five	Independent (judiciary)	Yes, at same meeting	Quarterly	No	No	12

It is likely that similar variation exists between other police force areas.

We recognise that a one-size-fits-all approach to the scrutiny of OoCDs isn’t appropriate and that areas should be free to adapt the guidance to local circumstances. For example, this variation may include the number of cases considered and the types of crime these involve.

But some panels aren’t scrutinising community resolutions at all. As a way of dealing with cases quickly and informally, sometimes these may not receive the same supervision and quality assurance as more formal outcomes. So it is important that some form of scrutiny takes places to make sure these outcomes are being used fairly and appropriately.

Some panels don’t regularly consider whether OoCDs have been given appropriately to people from ethnic minority backgrounds. This does little to assure local communities that the police are working transparently and fairly.

In 2023, a new two-tier system of adult OoCDs is being introduced into law. We have been told by the NPCC lead for OoCDs that the current scrutiny guidance will therefore be updated. Some principles for the scrutiny process will be included in the statutory guidance that accompanies the legislation. We will continue to work with the NPCC and the Ministry of Justice to make sure the findings from this inspection are considered as part of these revisions.

Arrests, voluntary attendance, charges and bail decisions after charge

At the time of our inspection, two of the five forces had set up custody scrutiny panels and two had plans to do so. There were variations in all the established and proposed arrangements.

For example, not all the arrangements included examining high-level data to inform the panel’s work. One arrangement only included force representatives and didn’t involve independent community members. None of the existing or proposed panels included all our decision points.

In March 2022, the Chief Constables’ Council agreed to a proposal by the NPCC custody lead that all forces should support the implementation of independent custody scrutiny panels.

Regarding the panels’ objectives, the proposal stated:

“1. The aim of Independent Scrutiny Panels is to increase trust and confidence by:

Ensuring partnership scrutiny of police custody procedures, representative of key minority interests in the community, and supporting the work of independent agencies that operate within the custody space;

Using data trends to identify and understand any apparent disproportionality;

Considering wider issues where the panel feels it could add value, for example concerning the treatment of detainees, issues of safety and dignity;

Increasing public confidence in the Criminal Justice System (CJS) and building trust with ethnic minority communities through transparency, accountability, engagement, and understanding of the custody processes.

Reviewing and commenting on the application of police powers including strip search, use of force, and remands in custody, among other custodial areas where there is evidence of disproportionality.

2. The panel should use their independent perspective to scrutinise, guide and inform policing. Policing in turn should learn, develop, act on where appropriate and feedback to the panel having considered their comments, but also invite them to look at particular aspects of police custody that they feel would be beneficial.”

We support the implementation of independent custody scrutiny panels. We think that they have great potential to help forces gather, analyse and scrutinise data on potential race disparity in criminal justice decision-making.

In our rolling programme of custody inspections, we work closely with the NPCC, inspected forces and other interested parties. Through this inspection programme, we will work with these organisations to help forces understand the benefits of independent custody scrutiny panels.

We propose that custody scrutiny panels include as wide a range of relevant activity as possible, including the custody-related criminal justice decision points we examined in this inspection, to make the most of the forum. It is also important that there are strong links between the custody scrutiny panels and those that exist for out‑of-court disposals.

In 2022, the Government responded to the Commission on Race and Ethnic Disparities (CRED) report. The response included the following action:

“Action 10

The police need the powers to tackle crime – but there also needs to be effective local scrutiny of these powers in order to enhance trust and strengthen relations between police and communities. The Home Office, with policing partners including Police and Crime Commissioners, will develop by summer 2023 a new, national framework for how the use of police powers – including stop and search and use of force – are scrutinised at a local level. This framework will ensure that local scrutiny panels are independently-led, reflect the diversity of the areas they represent and give police officers the confidence to use their powers with the backing of local communities.”

The action to create a framework for scrutiny was included in the context of CRED recommendations about the use of stop and search, and use of force powers by the police. The Home Office has now accepted that this framework should also include other important police powers that are used in the criminal justice process. Our inspection has shown there is little scrutiny of some of these decisions. And where scrutiny does take place, there is too much variation between forces to promote consistent improvements.

Local publication of data

In the Published data section, above, we have explained what published information is available about race disparity in criminal justice decision-making throughout England and Wales. In our fieldwork we also examined what information is published by forces or their partner organisations.

It is important that forces publish local information relating to race disparity because:

• it helps reassure communities that forces are operating legitimately;
• it allows forces to be held to account;
• it shows forces are aware of the concerns of different communities and are responding to these; and
• it allows comparisons between forces.

The Lammy review called for an emphasis on open decision-making and the publication of data. It advocated “publishing data in much more detail, thereby enabling outsiders to identify and scrutinise disproportionate treatment”.

All the forces we inspected published information relating to disparity in stop and search in an easily located place on their websites. But none of the forces published any statistical information about race disparity in criminal justice decision-making.

This supports our view that much more emphasis has been placed on stop and search – and that criminal justice disparity, despite its importance, hasn’t been given the prominence envisaged by other reviews.

As we have said above, not all forces are able to publish data about race disparity in criminal justice decision-making. This is because they don’t currently gather the data, and/or can’t easily gather it.

We also found that none of the forces we inspected published on their websites any information about arrangements for the scrutiny of OoCDs.

Instead, there was some information on the websites of the relevant OPCCs, but this information varied considerably and in some cases was very limited. It didn’t include any high-level data or any information about race disparity.

While it is appropriate for OPCCs to publish information on the scrutiny of OoCDs, this is undermined by the fact that there were no links to the relevant OPCC web pages (or other information) from the force websites. We believe it is more likely that members of the public seeking information about how OoCD decisions are scrutinised will first look for this information on their local police force websites.

We were told by the NPCC that its performance management co-ordination committee had formed a data transparency publication group. One of the purposes of the group is to provide a consistent process and governance for data publications for all forces. We believe this group is best placed to co-ordinate the work of chief constables in relation to criminal justice disparity data, and help them to publish it in a consistent way.

As we have said above, we are working closely with the NPCC lead for OoCDs on new guidance for forces. We will propose that the publication by the police of scrutiny arrangements is considered as part of this work.

Local accountability

In 2005, the Government published the Root and Branch Review of Race and the CJS Statistics: Final Report. We discuss this further in the Published data section, above.

One of the recommendations in the report was that Local Criminal Justice Boards (LCJBs) should take responsibility for collecting, analysing and publishing data on ethnicity from criminal justice agencies, including police forces. Other recommendations concerned the need to adequately resource LCJBs for this purpose.

The same year, the Government responded to the recommendations. While acknowledging the need for better collection of criminal justice statistics, it rejected the suggestion that LCJBs should lead this locally. This was because it was felt that LCJBs at that time didn’t have the capacity.

The Government response to this recommendation concluded:

“LCJBs face an important and challenging task to deliver the CJS PSA [public service agreement] targets and making them responsible for the development of ethnicity statistics could divert their resources and focus. Even if additional resources were made available to LCJBs, the serious issues raised by the review about completeness and timeliness of the statistics will require significant effort and impetus from central departments to address.”

It wasn’t within the remit of this inspection to examine the effectiveness of alternative solutions the Government provided to make sure that the police (and other agencies) produce good data on race disparity in the criminal justice system. But there are now opportunities for information to be published locally (and in some cases for England and Wales) about this important subject.

Since 2005, changes have been made to the model of governance and accountability in police force areas. In 2019, the Government published an operating framework setting out how PCCs (and their mayoral equivalents) can play an active role in local criminal justice system (CJS) forums to improve co-ordination and achieve shared CJS outcomes.

Although this guidance emphasises that PCCs should use their position to provide leadership and transparency for the CJS at a local level, it doesn’t specifically mention disproportionality. Given the pivotal role PCCs play in local justice partnerships, we think they are in a good position to work with LCJBs to challenge the police (and other agencies) on any apparent disparity in criminal justice decision-making.

As we have described elsewhere in this report, LCJBs (or their regional equivalents) are already carrying out some activity to consider disproportionality in their areas. Formalising this approach so that all LCJBs take responsibility for criminal justice disparity would:

• use an established forum with the right mix and seniority of membership, including those from community groups;
• hold agencies to account, promote greater co-ordination and increase transparency;
• enable criminal justice agencies to share data with each other and explore links and trends; and
• provide good links to national organisations such as the National Criminal Justice Board and the Association of Police and Crime Commissioners. This would mean good practice and areas for improvement could be shared easily and widely.

PCCs produce police and crime plans that detail important local priorities for the police to focus on. Chief constables are held to account for putting this plan into practice and performance against it.

The APCC provides information on national policing policy issues and legislation, and consults with PCCs to develop relevant policy and influence change. Since December 2020, the APCC has published a race disparity toolkit approximately every six months. This aims to help support its members to tackle race disparity locally, in part through sharing good ways of working.

In this inspection, we examined the police and crime plans for the six forces we inspected. Of these, only two forces had diversity, equality and inclusion as an area of priority.

In our leadership and governance review, we noted that of the eight police and crime plans we reviewed, race was prioritised in just one. In the review, we noted that if race and policing isn’t a feature of PCCs’ plans, chief constables may be less inclined to give it the priority it requires.

Training and guidance

In this inspection we considered three aspects of the training and guidance provided to police officers and staff, and forces:

• What training and guidance is given to officers and staff to make sure that decisions are made fairly? This is often known as unconscious bias
• What training and guidance is given to officers and staff to make sure they understand the importance of collecting information on ethnicity? This typically includes how best to explain to people why this information is being sought, and how to make sure that the exchange is handled sensitively.
• What guidance is given to police forces about what information relating to race disparity should be gathered and why?

The College of Policing, working closely with the National Police Chiefs’ Council (NPCC), is responsible for setting standards for policing. This includes through authorised professional practice, evidence briefings and guides.

What training and guidance is provided to officers and staff to make sure that decisions are made fairly?

Unconscious bias training seeks to raise awareness of potential biases that someone may have but be unaware of. These biases may result in judgments being made and decisions being taken that prejudice people with certain characteristics.

In 2020 the Government Equalities Office commissioned the Behavioural Insights Team to provide a summary of the evidence on unconscious bias and diversity training. The report concluded that there was no evidence that the training changes behaviour in the long term and that there was emerging evidence of unintended negative consequences.

In view of this report, the Government concluded that unconscious bias training didn’t achieve its intended aims. It therefore phased out the training in the Civil Service and encouraged other public sector employers to do the same.

The College told us that it didn’t advocate standalone unconscious bias training, as it believed that awareness training on its own wouldn’t change behaviour. But it said it is important that police officers are aware that unconscious bias exists, and that the subject will continue to form part of the national curriculum.

Stop and search training for new student officers includes an element on unconscious bias in those specific circumstances. All five forces we inspected confirmed this. The stop and search training also includes making officers aware of the Equality Act 2010. It makes it clear that it is never appropriate to stop and search someone solely based on their protected characteristics.

A stand-alone distance-learning training package, available to all officers, about the Equality Act 2010, also includes a slide about unconscious bias. This course isn’t mandatory. We discuss this further on in this chapter.

Despite the position of the College, some of the fieldwork forces told us that they either had provided more general unconscious bias training to existing officers or were intending to do so.

One force told us it had secured funding from its police and crime commissioner to work with a company to carry out online unconscious bias scenario ‘tests’ on all officers. The force told us that it intended to present different scenarios to officers, and ask them to give their likely response.

The force hopes this will highlight areas of policing activity for it to focus on for training and/or awareness raising. Analysis of the results should inform an assessment of its training needs.

This is an interesting but untested initiative. We have given further details to the College so it can monitor this work and decide whether to explore it further.

Some forces provided training covering diversity, equality and inclusion to some of their non-student officers. None of this training specifically covered the need to make sure that decisions were fair, although this was clearly implied through the provision of the training itself.

The College has recently introduced an additional one-day custody training package. The training is designed to help custody officers and staff better understand vulnerability and make good decisions about detainees who may have specific needs. While the training isn’t specifically about race disparity, it more generally emphasises the importance of treating people fairly, making good decisions and recording rationales. As many of the criminal justice decisions we examined in this inspection are made in a custody environment, this training is a welcome addition.

What training and guidance is given to officers and staff to make sure they understand the importance of collecting information on ethnicity?

In the section The problem of missing ethnicity data, we discuss that training and guidance is needed to help officers and staff understand why it is important to effectively gather data on ethnicity.

It is important that officers understand why they need to collect information on ethnicity. Without this understanding, officers may fail to gather the information and/or fail to add it to crime recording systems. They may think the task is unimportant and just a bureaucratic exercise.

In 2002, the Association of Chief Police Officers (ACPO) produced a Guide to Self‑Defined Ethnicity and Descriptive Monitoring.

This guidance states:

“It is proposed that training specific to the completion of documentation and administrative processes relating to self-defined ethnicity and descriptive information monitoring is developed in-force as it will need to conform to local forms, systems, and protocols.”

We think it is likely that in most forces this training was never developed. If it was, it doesn’t appear to have persisted to this day. This is unsurprising, as the system for recording self-defined ethnicity is now well established and known by all officers. The only differences that have occurred are the moves from the 16+1 to the 18+1 system after the 2011 census; then the recent move from 18+1 to 19+1 after the 2021 census.

None of the forces we inspected provided any specific training or guidance on this subject.

We understand that, for some people, being asked about their ethnicity may be a sensitive topic. So it is important that police officers and staff are given the understanding and confidence to handle such situations.

In our fieldwork, some officers told us they were sometimes embarrassed to ask a person about their ethnicity. In our report into the super-complaint made by the Tees Valley Inclusion Project, we found something similar:

“The task of asking victims about their ethnicity appears to be a sensitive one for some officers and staff. Some may choose not to ask a victim their ethnicity because they believe it might offend them or make them think they will be treated differently.”

In these situations, it is possible that officers won’t ask the relevant questions. This could lead to more crime records where the self-defined ethnicity of the person concerned is shown as “not stated”. This may be one explanation for the large numbers of such cases that we found in our data gathering.

In our inspection fieldwork we found that most officers knew they had to collect ethnicity information. But we heard varying explanations from officers as to why they thought they had to. Knowledge about the public sector equality duty was very low indeed. This may be because specific guidance about the Equality Act 2010 is given during student officer training, and the officers we spoke to were more experienced. So the officers may have received this training some time ago, or in some cases not at all.

The training about the Equality Act 2010 is given to new student officers as a specific module within the Valuing Diversity and Inclusion module. This is part of the Ethics and Values programme of learning. The subject of gathering information about ethnicity and other protected characteristics, and why this is important, isn’t currently covered in this student officer training module. We think that this training module would be a good place to include this important subject.

The NPCC diversity, equality and inclusion (DEI) portfolio told us it recognised the importance of giving officers training and guidance to help them to overcome their lack of knowledge and confidence in this area. But we are concerned that even if changes are made to student officer training, other officers will still be ill-prepared for dealing with this important element of policing.

The College also provides a distance-learning training package on the Equality Act 2010. The College told us that the training was developed before 2010, when it was called Introduction to Diversity. As it had only received some minor updates since its introduction, the module was due for review.

The College told us that since this training package was introduced, it has been completed by just over 82,000 police officers and staff. At the time of our inspection there were around 220,000 officers and staff.

The course isn’t mandatory. And none of the stated learning outcomes for the training refer to the importance of gathering accurate information about protected characteristics, and why this is important. None of the officers we spoke to had completed this training.

What guidance is given to police forces about what information relating to race disparity should be gathered and why?

As we have discussed above, section 95 of the Criminal Justice Act 1991 provided the first statutory requirement to gather and publish information on ethnicity in relation to certain aspects of the operation of the criminal justice system.

The Home Office produced accompanying guidance to make sure there was consistency. This guidance was then updated in 2002. As we discuss above, this was to coincide with ACPO’s Guide to Self-Defined Ethnicity and Descriptive Monitoring.

The Home Office now issues guidance to police forces about what information should be collected using the annual data requirement (see Published data, above). This includes a commentary about why this information is important.

As we have discussed above, the NPCC’s diversity, equality and inclusion strategy and toolkits, and the Police Race Action Plan, all contain references to the importance of collecting information about people from ethnic minority backgrounds. But we have set out reasons for including greater detail about some of the important criminal justice decisions that the police make.

Working with other organisations

In our inspection we asked forces if they worked with criminal justice partner organisations on a local and/or regional basis. We wanted to understand whether forces had tried to develop a wider and more system-based approach to understanding race disparity in the criminal justice system.

Four of the five forces we inspected were actively involved in regional partnership work, including with other forces. This is positive because it helps forces share good ideas and compare different approaches.

The regional groups were at various stages of development. Some had only recently been established and at the time of the inspection didn’t have any agreed terms of reference or action plans.

We have discussed, above, the approach in Wales led by the Criminal Justice Board for Wales. This is a good example of how organisations can come together to collaborate and encourage each other to improve, and to try to develop a more holistic understanding of disparity in the criminal justice system.

We would encourage agencies to work together more to try and determine where disparity exists in the criminal justice system, and how prevalent it is. It is also important that there is national oversight and co-ordination of such projects because the findings are likely to have wide relevance beyond individual forces.

Annex A – About the data

Background

Data in this report is from a range of sources, including:

• the Home Office;
• the Ministry of Justice; and
• our inspection fieldwork.

When we collected data directly from police forces, we took reasonable steps to agree the design of the data collection with forces and with other interested parties such as the Home Office. We gave forces several opportunities to quality assure and validate the data they gave us, to make sure it was accurate. We shared the submitted data with forces, so they could review their own and other forces’ data. This allowed them to analyse where data was notably different from other forces or internally inconsistent.

We set out the sources of this report’s data below.

Data in the report

Arrests data

We took this data from the May 2022 and October 2022 releases of the Home Office ‘Police powers and procedures: Stop and search and arrests, England and Wales, year ending 31 March 2021 and 31 March 2022’. The Home Office may have updated these figures since we obtained them for this report.

Voluntary attendance data

We took this data from the November 2021 and 2022 releases of the Home Office ‘Police powers and procedures: Other PACE powers, England and Wales, year ending 31 March 2021 and 2022’. The Home Office calls these experimental statistics to indicate that they potentially have a wider degree of uncertainty. For the purposes of this report, the only relevant issue is that some forces didn’t provide data.

We indicate which forces submitted data in the report. The Home Office may have updated these figures since we obtained them for this report.

Cautions data

We took this data from the Ministry of Justice Criminal Justice System Statistics Publication: Out of Court Disposals 2011 to 2021: Pivot Table Analytical Tool for England and Wales Time Period: 12 months ending December 2011 to 12 months ending December 2021 (PDF document).

Population

For likelihood ratios for arrests, voluntary attendance and cautions, ONS census 2021 population estimates were used and police force area demographics were calculated by the Home Office.

Methodology

Relative likelihoods (relative rate indices)

We calculated relative rate indices (RRI) or relative likelihood to compare ethnic disparities.

To calculate a relative likelihood, we used the following formula:

Relative likelihood = percentage (or proportion) of one group experiencing an outcome, divided by percentage (or proportion) of the comparator group experiencing an outcome (typically the majority group – here the White ethnic group).

A relative likelihood greater than 1 suggests the outcome is more likely in ethnic group ‘X’ than in the comparator ethnic group. A relative likelihood less than 1 suggests the outcome is less likely in ethnic group ‘X’ than in the comparator group.

We applied the Ministry of Justice tolerance zone of 0.8 to 1.25 to the relative likelihoods, as set out in its Methodology review report: Statistical reporting from a diversity perspective (PDF document). The zone is marked on the charts with dashed lines.

This tolerance zone represents the industry standard rule, which has the benefit of flagging up whether an RRI value has practical as opposed to statistical significance. Only if an RRI value is outside the tolerance zone between 0.8 and 1.25 is there considered to be a disparity of outcomes from a practical perspective.

For more information on relative likelihoods, see this Government guide.

Annex B – Ethnicity codes to be used by the police from 2023

Self-defined ethnicity codes

Aligned to the categories used in the 2021 Census of England and Wales.

Asian or Asian British

• Indian
• Pakistani
• Bangladeshi
• Chinese
• Any other Asian background

Black, Black British, Caribbean, or African

• Caribbean
• African
• Any other Black, Black British, or Caribbean background

Mixed or multiple ethnic groups

• White and Black Caribbean
• White and Black African
• White and Asian
• Any other Mixed or multiple ethnic background

White

• English, Welsh, Scottish, Northern Irish or British
• Irish
• Gypsy or Irish Traveller
• Roma
• Any other White background

Other ethnic group

• Arab
• Any other ethnic group

In Wales, ‘Welsh’ is the first option in the White category.

Officer-defined ethnicity codes

1. white European

• dark European

2. Afro-Caribbean

• Asian

3. Oriental

• Arab

4. unknown

Annex C – Criteria and indicators

1. There are effective strategies in place and strong leadership is demonstrated at local and national level

1.1 There is effective strategic leadership at national level

1.1.1 There is a NPCC lead with responsibility for disparity in criminal justice processes.

1.1.2 There is an understanding at a national level of the nature and extent of disparity in decision-making in criminal justice processes.

1.1.3 National leads work together effectively to drive improvements in the way that information about disparity in criminal justice decision-making is gathered and understood.

1.1.4 There is effective data collection at a national level which is used to shape strategy.

1.1.5 NPCC leads work with a wide range of partners to better understand and scrutinise disparity on a national level.

1.1.6 There are clear mechanisms in place to monitor any disparity on a national level.

1.1.7 There is effective work on a national level to address any disparity in criminal justice processes.

1.1.8 There is an understanding at a national level of the links between disproportionality in stop and search and subsequent criminal justice decision-making.

1.2 Police policies and guidance are available and fit for purpose

1.2.1 There is effective policy and guidance for forces about what they should do to gather information about, scrutinise and understand disparity in criminal justice processes.

1.2.2 Local police policies and guidance reflect the need to gather information about, scrutinise and understand disparity in criminal justice processes.

1.3 There is effective force strategic leadership

1.3.1 There is a force lead with responsibility for disparity in criminal justice processes.

1.3.2 The force has strategies, policies, and actions for criminal justice processes to take account of the public sector equality duty.

1.3.3 The force has a strategy to explain, where necessary reduce, and where possible eliminate disparity within the criminal justice system.

1.3.4 There is an understanding at a force level of the nature and extent of disparity in decision-making in criminal justice processes, including by comparison with other forces.

1.3.5 Force leads work together effectively to drive improvements in the way that information about disparity in criminal justice decision-making is gathered and understood.

1.3.6 There is effective data collection at a force level which is used to shape strategy and reduce disparity.

1.3.7 There are clear mechanisms in place to monitor any disparity on a force level.

1.3.8 There is effective work on a force level to address any disparity in criminal justice processes.

1.3.9 There is an understanding at a force level of the links between disproportionality in stop and search and subsequent criminal justice decision-making.

1.3.10 There is a force action plan that reflects the NPCC 2018-25 EDI toolkit.

1.4 Police forces provide training to officers and staff in dealing with unconscious bias and ensure that all relevant staff have received this training

1.4.1 Appropriate training is available for all staff dealing with decisions in criminal justice processes.

1.4.2 The force ensures that training is mandated at appropriate levels for all staff.

2. There are effective partnership arrangements in place to scrutinise decision-making and improve community confidence

2.1 Partnership working

2.1.1 Forces work with a wide range of partners to better understand any disparity that exists.

2.1.2 Forces work jointly with partners to identify ways of reducing and addressing any disparity that exists.

2.2 Scrutiny

2.2.1 Forces work jointly with partners to scrutinise individual cases by way of a scrutiny panel or similar arrangement.

2.2.2 Forces work jointly with partners to scrutinise disparity on a force level, including by comparison with other forces.

2.3 Community confidence

2.3.1 Forces publish and explain the findings of any disparity to the public to increase confidence and legitimacy.

2.3.2 Force publications of disparity specify levels of community involvement and do so in a user-friendly way.

3. Police decisions are proportionate and legitimate

3.1 Police decisions are legitimate

3.1.1 Police decisions comply with national and force guidance.

3.1.2 Police decisions are well-evidenced with a clear recorded rationale.

3.1.3 Police decision are well supervised and there is evidence of challenge where appropriate.

Annex D – External Reference Group

• Home Office
• College of Policing
• Wales Criminal Justice Board
• National Police Chiefs’ Council (NPCC)
• Association of Police and Crime Commissioners (APCC)
• Race Action Plan Programme Board
• Black Police Association (BPA)
• Police Federation
• Independent Office for Police Conduct (IOPC)
• Criminal Justice Alliance
• Unjust
• Open Society
• Faith Matters
• Academia

Annex E – Methodology

Introduction

Our inspection assessed whether there was any race disparity in decisions made by police forces in England and Wales. The inspection focused on decisions made during criminal justice processes that affected individuals that were the subject of investigations, not the alleged victims.

For the purposes of this inspection, ‘criminal justice processes’ means those activities that take place after it has been decided by the police that a criminal act may have taken place.

The decision points that the inspection covered are:

• arrest;
• voluntary attendance (often used as an alternative to arrest);
• community resolution (Home Office crime outcome type 8);
• caution;
• charge (police only decision); and
• post-charge bail.

So this inspection didn’t look specifically at decisions made to stop and search individuals. But it may have included the actions of the police when stop and search encounters resulted in the need for the police to take further action. For example, when something had been found in a search.

The inspection comprised three phases:

• Phase one: All-force information collection and analysis.
• Phase two: Case file review and fieldwork in six forces.
• Phase three: Interviews with relevant national leads.

More detail on each of these phases is given further below.

The findings from this inspection will be added to findings on disparity from the other criminal justice inspectorates. This will result in a compendium report that will be published separately.

Other sources of information

The inspection drew on information that is available from other HMICFRS inspection activity. This included PEEL and other inspection reports, most pertinently those related to super-complaints and the HMICFRS race and policing programme.

Other open-source information was also used in the inspection process where it was relevant.

External Reference Group

HMICFRS has established an external reference group (ERG) to help inform and guide the inspection. The membership of this is listed at Annex D.

Methodology detail

Phase one

All forces in England and Wales were asked to provide information on the force and partnership response to disparity in criminal justice decision-making. It was intended that the request would be as succinct and easy to complete as possible, using Smart Survey.

The Technical Advisory Group was consulted in advance of the survey, to make sure that the information being requested should be available to forces, and/or to gauge the level of activity required by forces to source the information.

Analysis of the survey was used to select the six fieldwork forces and provide detail of the national position to inform national interviews and the overall report.

Phase two

This phase consisted of a review of case files and fieldwork in six selected forces. The inspection specifically sought areas of apparent positive practice.

We aimed to achieve a mix of force sizes, including urban and rural forces. We also included a Welsh force. We consulted widely about force selection and considered other recent inspection activity in any proposed forces, to help regulate inspection demand.

Pre-fieldwork document request

Forces were asked to provide documentary evidence of their approach to disparity in criminal justice decision-making, to include strategy and policy documents as well as minutes from recent relevant meetings.

Case file review

HMICFRS inspectors examined a statistically significant number of cases for each decision-making point in each of the six forces. The reviews considered a representative sample of cases based on the ethnicity of the person concerned.

Inspectors considered whether the decision was appropriate, adequately justified, subject to supervision, and whether the case was subsequently quality assured and/or scrutinised.

Fieldwork

Fieldwork included the following:

• interview with force disparity lead;
• focus group of response and neighbourhood officers, and investigators;
• focus group of supervisory officers from response and neighbourhood teams and investigation teams;
• focus group of custody officers and dedicated decision makers.

The inspected forces were provided with a verbal debrief (setting out emerging findings and themes from the fieldwork) at the conclusion of the fieldwork. This included the preliminary results from the case file review.

Phase three (national interviews)

These were carried out after the force fieldwork and comprised:

• interviews with national police leads;
• interview with College of Policing lead; and
• interviews with national stakeholders including policy officials at the Home Office and the Ministry of Justice, and representatives from groups advocating for greater scrutiny of disparity in the criminal justice system.

References

[1] ‘Guide to Self-Defined Ethnicity and Descriptive Monitoring’, Association of Chief Police Officers of England, Wales and Northern Ireland, 2002.

[2] ‘Guide to Self-Defined Ethnicity and Descriptive Monitoring’, Association of Chief Police Officers of England, Wales and Northern Ireland, 2002.

[3] Scrutiny panel guidance, National Police Chiefs’ Council and Ministry of Justice, July 2019.

Back to publication

Race and policing: An inspection of race disparity in police criminal justice decision-making