Improving data capture and analysis regarding stop and search

• Improve the force’s understanding and analysis of stop and search. This means analysis at a force, Community Safety Partnership (CSP), Lower Layer Super Output Area (LSOA) and individual level.
• Improve the force’s understanding of the causes of stop searches.
• Improve the recording and analysis of disproportionality in relation to stop and search, leading to the development of an action plan.
• Deliver increased transparency and accountability to the wider community by sharing the data produced through the community scrutiny panel. The data is shared with the community scrutiny panel via a newly formed, complementary data scrutiny panel. The outcomes are published publicly via a number of methods.

• Reduce rates of disproportionality relating to stop and search, or be able to provide the community with a full explanation for why disproportionality exists, linking this to operational context and seeking to reassure the community. Disproportionality includes race, gender and age.
• Explain to the community via the community scrutiny panel and the policing context the reasons for any disparity on a quarterly basis.  A series of regeneration roadshows are being held to discuss police legitimacy with wider communities, which encompasses stop and search. This has a particular focus on Black and Black heritage communities, in accordance with the Police Race Action Plan (PRAP).
• Increase the volume of positive outcomes relating to stop and searches in Bedfordshire leading to increased community confidence in the power.
• Reduce the volume of simple possession stop and searches, and disproportionality linked to simple possession searches.
• Increase community trust and confidence in relation to stop and search through increased transparency and improved scrutiny.
• Improve the quality of stop and searches completed across Bedfordshire.

There is no measure for community confidence, other than feedback from the community scrutiny panel. The panel’s feedback included them being content that stop and search in Bedfordshire is used fairly and effectively in the vast majority of cases. The force collates all feedback from dip sampling which provides an evidence base of effectiveness. A community confidence survey led by Office of the Police and Crime Commissioner (OPCC) is planned for 2024.

The motivation to investigate data quality and improve the collation and analysis of data, comes from recognition that this is an integral element of improving the effectiveness of stop and search by the force, while also increasing transparency and accountability with the Bedfordshire public. Prior to undertaking this programme of work, there was limited central collation of data, and the only public-facing data shared was via the police.uk website.

Delivery

The following steps were delivered to improve data capture and analysis regarding stop and search:

1. Reviewed the Tu-Serv recording platform and ensured that key data was mandated ensuring that officers captured this data. Tu-Serv is a digital recording platform used in Bedfordshire Police, which officers and staff can use via mobile devices. The platform enables officers to complete statements, record crimes, record use of force and stop and searches alongside other functionality. Tu-Serv then links with wider force systems, and data is collated via a digital warehouse and presented via the Power BI platform.
2. Amended the Tu-Serv recording platform to include the initial reason for each stop and search with categories defined as follows: intelligence led, linked to an incident, information from a third party or officer initiated.
3. Stop and search data was placed on the Microsoft Power BI platform, with a full review and enhancement to data presentation, enabling more meaningful analysis.
4. Qualitative review of supervisor reviews was undertaken to ensure that initial supervisor reviews were completed, providing suitable scrutiny and assessment and leading to improved standards.
5. Developed staff stop and search continuing professional development (CPD), based on evaluation of the data, and direct feedback from the community scrutiny panel.
6. Shared data relating to stop and search with the community scrutiny panel, redacting personal information. At present, the wider community scrutiny panel have used this data to target their dip sampling and scrutiny of stop and searches, as opposed to this being a randomised process. Launching in January 2024, the community panel will have a dedicated data sub-panel which will review data, setting direct questions for the force to answer. It is anticipated that this will be published along with responses to the wider community.
7. Established a community stop and search sub-panel to focus on data analysis, enabling the force to be held to account.
8. Improved the feedback mechanism to the community scrutiny panel in relation to stop and search, with dip samples being undertaken by the panel. This demonstrated what the force had done, and how learning was captured.

The evaluation looks at the impact of the intervention and how measures change over time. It is anticipated that evaluation of the effectiveness of the new method of data collation and analysis will be achieved in June 2024. This provides a six-month timeframe for the launch and embedding of the new data panel, alongside an anticipated academic study which is aimed to take place in 2024. This includes the:

• evaluation of data and trends relating to stop and search.  The overarching aim is to increase the effectiveness of stop and search through an increased level of positive outcomes, the reduction of disproportionality and increase in public confidence
• undertaking of a community survey to evaluate confidence in stop and search
• a survey with community scrutiny panel members to measure trust and confidence, to be undertaken once the data panel is launched in January 2024

The initial outcomes, as a direct comparison of 2022–2023 reporting period (1 April to 17 September) revealed:

• the total number of searches conducted increased from 1939 (2022) to 2243 (2023)
• the percentage of positive outcomes (including arrest, community resolution, summons, words of advice, caution) increased from 29% (2022) to 34% (2023)
• the percentage of simple possession searches decreased from 44% (2022) to 42% (2023)
• the percentage of disproportionality relating to the Black community in relation to simple possession searches decreased from 3.3% (2022) to 2.9% (2023)
• the percentage of disproportionality decreased from 3.6% Black ethnicity self-defined and 3.4% officer defined (2022), to 3.3% self-defined and 3.0% officer defined (2023). This is calculated based on the number of searches per 1000 of population, which is then divided by the census population figure for the ethnic group
• the percentage of disproportionality relating to Asian ethnicity decreased from 1.6% Asian ethnicity self-defined and 1.7% officer defined (2022), to 1.3% Asian ethnicity self-defined and 1.5% officer defined (2023)

Challenges 

• To deliver the digital change it took six months for the developer to update. 
• Ensuring that CPD products land with officers, carrying the right message and authenticity from guest speakers so that key messages are not lost.

Learning 

• Encouraging continuous engagement and communication to staff and providing reasons for the importance.  
• Maintaining engagement with the community, including them in change and including their views, as this will support building trust and confidence.  
• Engagement has to be widespread and include impacted communities. This was achieved in Bedfordshire through the community scrutiny panel and the independent advisory groups (IAG), work commissioned through the violence and exploitation reduction unit (VERU), including legitimacy surveys with young people in Luton, and regeneration roadshows across Bedfordshire seeking engagement with a variety of age ranges from Black and Black heritage communities. This was to gather their lived experiences relating to stop and search and seek their views on how this can be improved moving forwards.  

Barriers

• Initially, the ability for core data to be transferred to Power BI and presented. 
• Engaging with technical experts to ensure this is achievable.

HMICFRS identified practice