Introduction

Crime statistics for England and Wales

Police recorded crime statistics for England and Wales are published quarterly by the Office for National Statistics (ONS) and are a measure of the number of crimes that are recorded by the police in England and Wales. The Home Office collates recorded crime data from the 43 territorial police forces in England and Wales and the British Transport Police and supplies these data to ONS.

In England and Wales, police recorded crime statistics only cover notifiable offences. These are offences that could possibly be tried by jury, including violence against the person offences, sexual offences, robbery, theft, criminal damage and arson, as well as other crimes against society and the summary-only offence of common assault.

Police recorded crime statistics are one of two key sets of statistics on crime in England and Wales. The other set of statistics is the Crime Survey for England and Wales (CSEW) statistics. The CSEW is a statistical household survey that asks individuals about their experiences of crime; their attitudes towards different crime-related issues, such as the police and the criminal justice system; and their perceptions of crime and anti-social behaviour. ONS manages and owns the data from the CSEW.

Understanding and interpreting crime statistics for England and Wales is complex. Both sets of crime statistics have their individual strengths and limitations. When used together, they provide a more complete picture of crime.

The CSEW is the best source for understanding long-term trends in crime covered by the survey in England and Wales. The methods have changed little since the survey was first conducted in 1981, and it is not affected by changes to police crime recording practices or in the number of individuals reporting crimes to the police.

The police recorded crime statistics are a better indicator of police activity than trends in crime. Many crimes are not reported to the police, and the extent of under-reporting varies by crime type. For example, many victims of fraud do not report the incident to the police. In contrast, burglary offences are mostly well reported, as a police crime reference number is typically needed for home insurance claims. Because the CSEW includes crimes that respondents did not report to the police but did report in the survey, it provides a better picture of the extent of crime covered by the survey. However, the police recorded crime statistics provide insight on some higher-harm but less-common crimes, such as homicide or knife crime, which the CSEW does not cover or does not capture well.

The police recorded crime statistics cover a broader range of crimes than the CSEW statistics. For instance, the CSEW only captures crimes against individuals resident in households, whereas the police also record crimes against businesses and organisations such as shoplifting, and crimes against society and the state such as drug offences and public order offences. In addition, the police recorded crime statistics are more granular than the CSEW statistics – the number of offences recorded by the police is broken down by police force area.

ONS publishes information on which sets of statistics are thought to provide the most reliable measure of crime against individuals and households for the main crime types in England and Wales (Table 1). The police recorded crime statistics are the preferred data source for homicide, crimes involving a knife or sharp instrument (knife-enabled crime) and robbery.

Table 1. Overview of ONS’s preferred data source for the main crime types

Police recorded crime statistics and CSEW statistics show different long-term trends in crime. The CSEW statistics show a long-term fall in the extent of crime estimated by the survey since the mid-1990s. The latest estimates, for the year ending December 2023, show that the extent of crime estimated by the survey, excluding fraud and computer misuse, has decreased by 17% compared with the pre-coronavirus pandemic year ending March 2020, from 5.6 million offences to 4.3 million offences. If fraud and computer misuse are included, crime estimated by the survey decreased by 18% over the same period, from 10.2 million offences to 8.4 million offences.

In contrast, the total number of crimes recorded by the police, excluding fraud and computer misuse, has increased gradually since 2015, largely due to improvements to police recording processes and practices that we discuss in this report. For the year ending December 2023, the total number of crimes recorded by the police, excluding fraud and computer misuse, was 5.3 million offences, compared with 5.5 million offences in the pre-coronavirus pandemic year ending March 2020, a 4% increase. If fraud and computer misuse are included, the number of crimes recorded by the police increased by 10% over the same period, from 6.1 million offences to 6.7 million offences.

The Crime in England and Wales statistical bulletin is not the only crime statistics output for England and Wales. Subsets of the police recorded crime statistics for England and Wales are also published. ONS publishes statistical bulletins about specific crime types such as homicide, sexual offences and domestic abuse, which include police recorded crime figures. The Home Office publishes a statistical bulletin on the number of hate crimes recorded by the police.

The Home Office is responsible for publishing crime outcomes statistics for England and Wales. These statistics report on the actions police forces have taken in response to recorded crimes in England and Wales. They are related to the police recorded crime statistics and undergo a similar quality assurance process.

The Home Office also publishes the police recorded crime and crime outcomes open data tables. These provide more-detailed breakdowns of the police recorded crime and crime outcomes statistical series.

Why we did this review

Accreditation demonstrates that the public can have confidence in the quality of official statistics. The police recorded crime statistics for England and Wales are published as official statistics, not accredited official statistics. This review is the first step towards the statistics being considered for reaccreditation.

Accredited official statistics are a subset of official statistics that have been reviewed by the Office for Statistics Regulation as complying with the standards of trustworthiness, quality and value in the Code of Practice for Statistics. They are called National Statistics in the Statistics and Registration Service Act 2007.

We removed the National Statistics accreditation for the police recorded crime statistics for England and Wales in 2014 following an assessment which found that the quality and consistency of the underlying data may not be reliable. The evidence that supported this conclusion included:

• Inspections of crime data integrity by His Majesty’s Inspectorate of Constabulary and Fire & Rescue Service (HMICFRS, formerly HMIC), which showed that police forces in England and Wales were under-recording crime (nationally, 19% of all crimes reported to the police were found not to be recorded as crimes).
• Growing disparity between police recorded crime statistics and the Crime Survey for England and Wales statistics, which suggested that crime recording standards were falling.
• Concerns published by the then-Public Administration Select Committee (PASC) and the Home Affairs Select Committee about the reliability of the underlying police data. The 2014 PASC Inquiry into police recorded crime statistics criticised the UK Statistics Authority’s regulatory scrutiny of the statistics. PASC concluded that there had a been a “long-standing failure… to address the thoroughness of assessment of Police Recorded Crime”.

In our 2014 assessment, we advised that accreditation would only be reinstated when ONS, working with the Home Office or other bodies, was able to demonstrate that the quality of the underlying data and the robustness of the ongoing audit and quality assurance procedures were sufficient to produce statistics at a level of quality that meets users’ needs. Specifically, we required ONS to:

• Publish further information on data processing and quality assurance.
• Improve information about all aspects of the quality of crime statistics and the impact on their use.
• Respond appropriately to regular, systematic audits, particularly HMIC’s 2014 inspection.

ONS and the Home Office started implementing these requirements. However, due to the scale of crime under-recording being reported by HMIC, ONS requested that any reaccreditation assessment be postponed until the required improvements had been implemented. Thus, we placed on hold our regulatory work for the reaccreditation of the police recorded crime series.

We did restore the accreditation of the homicide statistics based on the Homicide Index in 2016. The Homicide Index is a separate, more-detailed, record-level database that the Home Office continually updates with revised information from the police and the courts. Statistics from the Homicide Index are published separately in ONS’s Homicide in England and Wales statistical bulletin. The homicide figures published in the quarterly Crime in England and Wales bulletin are from the main police recorded crime series and are therefore not accredited official statistics.

In the last few years, we have reviewed specific elements of police recorded crime statistics in England and Wales, including ONS’s knife-enabled crime statistics and the Home Office’s hate crime statistics, but we have not reviewed the full police recorded crime series since 2015.

This review revisits our earlier regulatory work on police recorded crime statistics for England and Wales but goes much further. Through extensive engagement and desk research it looks in depth at the quality of the underlying data used to produce the police recorded crime statistics for England and Wales. We cover the data process from beginning to end – from how police forces record crime, to how data quality is managed and assured by all those involved in their collection and processing, to the production of the final statistics.

Our recommendations for improving the quality of police recorded crime data are what we deem as critical to address before we undertake a reassessment of compliance with the Code of Practice for Statistics.

What do we mean by statistical quality?

A range of complementary tools exist for understanding and assessing the quality of administrative data used to produce official statistics.

The Code of Practice for Statistics (the Code) sets the standards for producers of official statistics. The Code states that quality means that statistics are fit for their intended uses, are based on appropriate data and methods, and are not materially misleading.

The supporting Regulatory Standard for the quality assurance of administrative data is the Quality Assurance of Administrative Data (QAAD) framework. This provides a toolkit for statistics producers in making judgements about the continued suitability of administrative data for producing statistics. The framework sets out four practice areas associated with data quality:

• Operational context and administrative data collection.
• Communication with data supply partners.
• Quality assurance principles, standards and checks by data suppliers.
• Producers’ quality assurance investigations and documentation.

The Administrative Data Quality Framework (ADQF) was developed by the Analysis Function to be consistent with the Code and the supporting QAAD standard. It provides a framework for statistics producers to assess the quality of administrative data. It describes:

• Quality dimensions for assessing input quality, or quality dimensions for data.
• Output quality, or quality dimensions for statistics.

The framework sets out six input quality dimensions: completeness, uniqueness, timeliness, validity, accuracy and consistency. We found the distinction between input and output quality and the quality dimensions to be particularly helpful for thinking about the quality of police recorded crime statistics.

Wider regulatory work on crime statistics for England and Wales

Our work on police recorded crime statistics is part of a wider programme of work on crime statistics for England and Wales.

At present, no crime statistics for England and Wales are published as accredited official statistics. In July 2022, at ONS’s request, we temporarily suspended the National Statistics accreditation of the Crime Survey for England and Wales (CSEW) statistics, due to concerns about the impact of a shorter data collection period and the lower response rates for face-to-face interviews (compared with the period before the pandemic) on the quality of the estimates. Therefore, users of crime statistics in England and Wales currently do not have a measure of the extent of crime of a suitable quality.

ONS is working hard to improve the response rates and intends to put forward the CSEW statistics for reaccreditation in 2024. Our accreditation review is a top priority; we deem it critical to ensure that users can have confidence in the CSEW statistics.

How police in England and Wales record crime

Police recorded crime data are administrative data that are collected as part of police operations. For a crime to be recorded, the incident must be brought to the attention of the police, and the police must decide to record the incident as a crime.

Though the data are not collected for the sole purpose of producing statistics, police forces are required to share data with the Home Office. The Home Secretary uses its powers in the Police Act 1996 to require chief constables to provide regular data on the number of crimes recorded by their police force. Police recorded crime data are one of many data types that form part of the Home Office’s Annual Data Requirement (a list of all requests for data made under the statutory powers).

The Home Office collates and quality assures the recorded crime data received from the 43 territorial police forces in England and Wales and the British Transport Police. It then shares these data with ONS, which carries out further quality assurance and publishes them quarterly as the police recorded crime official statistics. Figure 1 sets out the process for producing police recorded crime statistics in England and Wales.

Figure 1. An overview of the process for producing police recorded crime statistics for England and Wales

In England and Wales, crimes must be recorded in accordance with the Home Office Counting Rules. These rules set out whether an incident should be recorded as a crime, when a crime should be recorded and how many crimes should be recorded for any single incident (as a single incident may involve multiple crimes). The Counting Rules provide a framework for interpreting and classifying crime, and they standardise how crimes are recorded across police forces in England and Wales.

The purpose of the National Crime Recording Standard is to ensure greater consistency in recording crime and a more victim-focused approach to crime recording. The standard sets out requirements for recording crimes; for example, police forces must record a crime within 24 hours of it being reported, and any allegation of crime must be recorded unless there is credible evidence to the contrary.

Each police force has a College of Policing-accredited crime registrar, who makes the final decision on whether a crime should be recorded, how it should be recorded and if it should be ‘cancelled’ (removed from the police force’s crime recording IT system). The force crime registrar is responsible for ensuring that their force’s crime recording is compliant with the Home Office Counting Rules and National Crime Recording Standard.

HMICFRS has wide-ranging statutory powers that allow it to monitor the performance of all police forces in England and Wales. HMICFRS independently inspects and reports on the effectiveness, efficiency and legitimacy of police forces. This includes how well forces record crime.

Our approach

Our review set out to:

• Identify where quality improvements have been made to police recorded crime statistics and what key factors are important for driving further improvements.
• Identify the nature and extent of quality issues in the underlying police recorded crime data.
• Understand where responsibility for data and statistical quality lies.
• Develop recommendations to support quality improvements.

We focused on three key areas of statistical quality:

1. Crime recording processes and practices – how police forces record crimes, including their interpretation and application of the Home Office Counting Rules.
2. Crime recording IT systems – the systems police forces use to record crimes, and the tools they use to manage data quality.
3. The end-to-end quality assurance process – how police forces, the Home Office and ONS quality assure police recorded crime data.

We gathered evidence across four key stakeholder groups:

1. A sample of police forces in England and Wales

We had detailed meetings with nine police forces about their crime recording processes and practices (about one-fifth of all police forces in England and Wales). We were impressed by the professionalism and dedication of the police staff that we spoke to, and we appreciated the openness with which police forces discussed their recording practices and the challenges of recording crime accurately and consistently.

We selected forces based on a mix of characteristics, which included their geographic location, urban or rural mix and their HMICFRS inspection gradings of crime data integrity. Our aim was to have a reasonably representative group of forces with variations in approaches to crime recording.

We have chosen not to identify individual forces in this report unless they have agreed to be named or the information is already in the public domain. We did not have access raw police recorded crime data for this review, considering only published data and statistics.

2. His Majesty’s Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS)

Our review is not an audit of the crime data integrity of police forces in England and Wales. This activity falls within HMICFRS’s remit. We engaged with HMICFRS to understand its process for auditing crime data integrity and the improvements that police forces have made to crime recording since 2014. We thank HMICFRS for its openness about its processes.

3. Home Office and ONS analytical teams

We spoke to several teams in the Home Office to understand different aspects of the quality assurance and quality improvement of police recorded crime data. We spoke to:

• the Home Office Data Hub team, which is responsible for collating and quality assuring recorded crime data from police forces.
• the National Data Quality Improvement Service (NDQIS) team, which manages the development and roll-out of tools for improving the data quality of certain crime types.
• the National Crime Registrar.
• the Police National Database (PND) team, which produces data quality dashboards for police forces for a range of datasets that are shared with the Home Office, including police recorded crime.

We also spoke to the ONS crime statistics team to understand how it quality assures the police recorded crime data and the process for producing the final statistics.

4. Wider stakeholders – including national policing coordination bodies and devolved policing bodies

To understand the role that national policing bodies and groups play in data quality improvements, we spoke to the National Police Data and Analytics Board, the Centre for Data and Analytics in Policing and the Police Digital Service.

To understand how crime recording processes and practices and data quality differ across the UK, we spoke to Police Scotland, the Scottish Government and HM Inspectorate of Constabulary in Scotland; and the Police Service of Northern Ireland.

We also carried out extensive desk research to support the findings from our engagement. We reviewed:

• HMICFRS’s inspection reports and annual reports.
• Operation Soteria Bluestone Year One report.
• ONS’s and the Home Office’s statistical bulletins.
• ONS’s user guide to crime statistics and the Crime in England and Wales Quality and Methodology Information report.
• ONS and Home Office documentation on internal quality assurance.
• Academic papers.

What this review excludes

ONS also publishes statistics on the number of fraud and computer misuse crimes recorded by police. Our review did not examine the quality of police recorded fraud and computer misuse data as the process for recording these crimes is different from other crime types.

All police forces in England and Wales direct victims of fraud and computer misuse to Action Fraud, the national reporting centre for fraud and cybercrime, run by the City of London Police as the national lead force for fraud. The National Fraud Intelligence Bureau, also overseen by the City of London Police, processes these data and shares them with the Home Office. The police recorded fraud statistics published by ONS also include data from two industry sources: Cifas, the UK-wide fraud and financial crime prevention service, and UK Finance, the trade association representing UK finance and banking industry. Data on offences recorded by Cifas and Finance UK are sent directly to the Home Office by these bodies. The Home Office then collates fraud data from all three sources and sends the data to ONS.

We will review this subset of police recorded crime statistics separately as part of a wider review of the quality and value of fraud and computer misuse statistics for England and Wales later in 2024.