What FOI or academic studies have tried to estimate crimes by immigration status in specific UK police forces since 2000?
Executive summary
A modest body of academic work since 2000 has attempted to estimate offending by immigration status at the level of police force areas, and Freedom of Information (FOI) exchanges with statutory bodies have clarified what national data do — and do not — exist; however, national administrative systems and MoJ practices mean there is no routinely published, reliable national dataset of convictions or prison populations by immigration status, forcing researchers to rely on police force and bespoke datasets [1] [2] [3] [4] [5].
1. Academic use of police-force and Metropolitan Police data — targeted empirical studies
A series of academic studies have exploited local and police-force‑level data to ask whether inflows of specific migrant groups changed recorded crime or arrest patterns: LSE and collaborators modelled the impact of A8 (2004 EU accession) migrants on crime using panel data across Police Force Areas, explicitly using force-level arrest shares and local authority crime counts for 2002–2012 to identify relationships [2] [3], and an IZA Journal paper used spatial panel models and unique Metropolitan Police arrest data to compare arrest rates of natives and migrants for the 2000s [1]. These studies therefore do not rely on a national register of offenders by immigration status but instead reconstruct patterns from police force records and census-derived denominators [1] [2].
2. FOI routes, ONS replies and the limits of national statistics
Researchers and journalists seeking raw counts have frequently used FOI channels; the Office for National Statistics records FOI inquiries and publishes material on crimes by race and immigration status and related FOI responses, while signalling that its regular crime outputs focus on victims and police‑recorded crime rather than offender immigration status [6] [7] [8] [9]. The practical implication communicated repeatedly by official sources is that routine national series do not capture immigration status for offenders in the way needed to answer “how many crimes were committed by asylum seekers or recent migrants” without bespoke linkage or local police extracts [7] [8].
3. Government publications, foreign-national offender statistics and data quality caveats
The government publishes statistics on foreign national offenders and on immigration-system interactions, but the official 2025 release warns the system that aggregates details for deportation/return purposes contains quality issues which affect the reliability of variables such as convicted offence type, sentence length and nationality — a problem that undermines simple national tabulations by immigration status [10]. Parallel commentary from the Migration Observatory stresses that while prison snapshots and foreign‑national counts exist, there are no reliable national data linking convictions/prison population to immigration status and duration of residence — a gap that forces academics to rely on local arrest records, force-level arrests data, or bespoke linkages [11] [5].
4. What these methodological choices mean for findings and public debate
Because the Ministry of Justice does not routinely record immigration status for all offenders in a way that produces national statistics, much published research focuses on natural experiments (large migrant waves), spatial panels, and police-force or Metropolitan Police arrest records to infer patterns; these approaches produce informative estimates for specific times/places but cannot deliver a single, definitive national rate of offending by immigration status [4] [1] [2]. The political context matters: commentators and politicians sometimes frame isolated incidents as representative, while researchers caution that data limitations and different denominators (resident population vs. arrested population) complicate inference — an implicit agenda in some public narratives is to press for policy change on immigration without acknowledgment of the statistical shortcomings [4] [11].
5. Practical takeaways for researchers and FOI requesters
Anyone seeking police-force‑level estimates since 2000 will find the most productive paths are (a) academic papers that used PFA or Met Police arrest records (examples above) and (b) FOI approaches to local forces or to ONS for archived force-level data, but should proceed with caution about comparability, missingness and the government’s own admission of data quality problems in administrative systems that record nationality or “foreign national offender” status [1] [2] [10] [6].