What do UK police and CPS say about religion/ethnicity data collection for sexual offences?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
The Crown Prosecution Service (CPS) publishes operational data that includes ethnicity of suspects/defendants but states that its ethnicity collection followed the Criminal Justice System’s 16+1 self‑defined ethnicity categories only up to January 2021 and that its datasets have structural limitations and coverage gaps [1] [2] [3]. By contrast there is no established, routine mechanism across police, CPS and central government for collecting perpetrators’ religion, and officials have warned that any religion data would largely rely on self‑declaration and therefore be weak as a basis for robust analysis [4] [5].
1. What the CPS formally says about ethnicity data
The CPS’s regular data summaries make clear that ethnicity of suspects and defendants has been recorded in CPS operational management information according to the agreed Criminal Justice System 16+1 self‑defined ethnicity categories, but the publication language repeatedly notes that this approach applied “up to the end of January 2021,” and that the CPS does not publish official statistics — its outputs are operational indicators with known limitations [1] [2] [3] [6]. The CPS also highlights technical constraints in its datasets: age is recorded in bands representing current age rather than age at offence, many records are allocated to a ‘not provided’ category when key fields are missing, and some specific breakdowns (for example, whether an offence involved gang rape) are not recorded centrally [1] [7].
2. What police and other official bodies say about religion data
Central government responses captured in public petitions and reviews have explicitly stated there is currently no mechanism to collect data on the religion practised by perpetrators beyond any voluntary declaration by the individual, and that relying on such declarations would “weaken both the accuracy and comprehensiveness” of any analysis [4] [5]. That conclusion was reflected in materials around the 2025 review commissioned from Baroness Casey, which reviewed ethnicity data on group‑based child sexual exploitation but did not mandate religion as a collected field [4].
3. Practical limits police and independent analysts identify for ethnicity recording
Researchers and inspectors have documented numerous operational reasons why ethnicity data for sexual offences can be incomplete: discomfort in asking victims or suspects, lack of clarity about whose responsibility it is to record the field, and police IT/reporting systems that make recording difficult, all of which reduce reliability of ethnicity statistics in this offence area [8]. HM Inspectorate of Constabulary and Fire & Rescue Services inspections have similarly flagged recording gaps and inconsistent practice across forces when examining race‑disparity issues in criminal justice decision‑making [8].
4. How other official statistical bodies frame the gap
The Office for National Statistics emphasises that its crime publications reflect victims’ experiences or police‑recorded crime rather than convictions datasets by ethnicity, and that it does not hold the convictions‑by‑ethnicity data some researchers seek — pointing users toward the Ministry of Justice for conviction-level enquiries [9]. Independent sector reports (e.g., the CSA Centre) continue to publish analyses that include defendant characteristics such as ethnicity, but these draw on a patchwork of sources and are transparent about methodological caveats [10].
5. Implications, debates and competing agendas
Advocates for mandatory collection argue that nationality, ethnicity, immigration status and religion should be statutorily recorded to inform safeguarding and policy; opponents and officials counter that unreliable religion data and patchy ethnicity recording risk misleading analysis and could stigmatise communities if presented without caveats [5] [4]. The CPS emphasizes prosecutorial fairness and data limitations rather than policy prescriptions, while policing inspections and independent researchers point to fixable administrative barriers if more consistent data collection is pursued [11] [8].
6. Bottom line
Officially, ethnicity has been collected in CPS management data historically under the 16+1 SDE framework but with acknowledged temporal and technical limits (notably “up to the end of January 2021” caveats and ‘not provided’ records), and police recording practices remain inconsistent; by contrast, there is no routine, reliable mechanism for recording perpetrator religion and government documents warn any such data would be largely self‑declared and therefore weak for robust policymaking [1] [2] [3] [4] [5] [8] [9]. Where sources are silent about specific proposals or recent changes, reporting cannot assert their existence.