Dear Jon

Statistics on Police powers and procedures: Stop and search and arrests in England and Wales

We have reviewed your stop and search statistics, which are published in the Police powers and procedures: Stop and search and arrests, England and Wales statistical bulletin. 

There is substantial public interest in stop and search, particularly in the disparity in search rates of ethnic minorities. The Commission on Race and Ethnic Disparities (CRED) report highlighted a need for this disparity to be addressed and for more-granular data around ethnic disparities to be made publicly available. The Inclusive Britain Report sets out the UK Government’s response to the recommendations of the CRED report. One outcome is that the government’s Equality Hub has produced standards for ethnicity data describing best practice when collecting, analysing and reporting ethnicity data. 

Our review focused on the actions your team has taken to improve the way stop and search data are reported and to enable more-accurate comparisons to be made between different police force areas, to begin to address the recommendations of the CRED report. These actions are summarised in the Inclusive Britain update report, published in May 2023. 

The statistical bulletin, associated data tables and dashboards now include more detail on ethnic disparities in stop and search rates, allowing for more in-depth analysis and improving understanding of the scale and nature of ethnic disparities in stop and search rates. This information also enhances transparency about how stop and search is used by the police and which communities are most affected.  

We found a range of other positive features that demonstrate the quality and value of the statistics: 

• The bulletin identifies that self-defined ethnicity is missing in 20 per cent of stop searches, excluding vehicle-only searches. This is because self-defined ethnicity is provided by the searched person, who is not legally obligated to give it. To fill this data gap, your team has used officer-observed ethnicity when self-defined ethnicity is unavailable. Good-quality ethnicity data are important to understand the ethnic disparities in stop and search, so we welcome the work your team has done to overcome this limitation. However, we share the concern raised with us by a user about the quality of the officer-observed data, that there is potential under-reporting of those of mixed ethnicity in officer-observed ethnicity data. Your team should investigate the potential under-reporting in the data and use its findings to better communicate any limitations to users, and to further improve the statistics where possible.  

• The bulletin contains breakdowns by the sex, age and ethnicity of persons searched, and examines how the demographic profile of people stopped and searched compares with the general population of England and Wales. The analysis compares stop and search rates of non-white ethnic minorities with those of white people. These findings add insight on how sex, ethnicity, and age intersect and influence stop and search rates, by showing which age groupings, sex and ethnicities are most frequently searched.  

• We commend your team’s positive engagement with the Race Disparity Unit (RDU) to improve the statistics. Working with the RDU, your team has added more-detailed geographical information into the statistical bulletin. For example, the bulletin now includes analysis of stop and searches at the Lower Super Output Area (LSOA) for six police forces covering large urban areas and rural areas. This detailed analysis presents an in-depth and nuanced picture of how stop and searches are distributed throughout police force areas and how rural forces compare to urban forces. As highlighted in the CRED report, overall relative disparity between ethnic groups can be misleading if geographical differences are not considered.
   

These statistics were first designated as National Statistics in 2016. National Statistics are now known as accredited official statistics. These statistics can continue to be published as accredited official statistics on the proviso that you address the following points by December 2024: 

• As highlighted in our review of the quality of police recorded crime statistics for England and Wales, it is important for the Home Office to understand how police forces quality-assure their own data, so that it has sufficient oversight of data quality. This also applies to the stop and search data that the Home Office receives from police forces. To assure yourself and users of the quality of the underlying data, your team should improve its understanding of how police forces quality assure their stop and search data. Doing so will ensure that your team fully understands the limitations of the data, which can then be communicated to users.  

• In general, the published limitations of stop and search data are clear and helpful. However, this information about limitations is located after the analysis of the stop and search data. To Increase awareness of quality issues and aid interpretation of the statistics, caveats and information around the limitations should be made more prominent.  

• The intersectional analysis section of the bulletin contains analysis where people from ethnic minorities (not including white minorities) are grouped into one category. The bulletin explains that this grouping is due to the small sample sizes when the data sets are split by ethnicity. This is problematic as it can mask differences and disparities between ethnic groups. Your team should consider publishing intersectional breakdowns by ethnicity where possible, as this can provide added insights for the different ethnicities. This would support the CRED report recommendation to communicate racial issues more-effectively and to avoid grouping different ethnic minority groups together.
   

Thank you to your team for its positive engagement during this review. We plan to check in again with your team on progress in October 2024. Please get in touch if you would like to discuss this letter or if we can offer further assistance as these statistics continue to develop. 

I am copying this letter to Paul Trenell, Head of Police and Fire Analysis at the Home Office, and Rosanna Currenti, Head of Policing Statistics at the Home Office. 

Yours sincerely, 

 
Mark Pont
Assessment Programme Lead