What data sources report suspects' nationality for sexual offences in the UK and how reliable are they?
Executive summary
Official UK statistics that touch on suspects’ nationality in sexual offences come from police forces’ FOI disclosures, Home Office and Ministry of Justice (MoJ) offender datasets, and crime prevalence work by the Office for National Statistics (ONS) using the Crime Survey for England and Wales (CSEW); each source has important caveats about coverage and data quality (for example, police FOIs note nationality fields can be inaccurate and MoJ/official publications warn of limitations) [1] [2] [3]. Available reporting shows the Home Office/ONS police-recorded crime and CSEW provide victim and offence context, while MoJ and some police FOI releases provide offender nationality or “foreign national offender” breakdowns — but those datasets are partial, sometimes experimental, and have known reliability problems [4] [2] [3].
1. What official sources publish nationality or “foreign national” breakdowns?
The main, publicly available streams are: police force Freedom of Information (FOI) releases which sometimes publish counts by nationality for arrests/charges (examples include Metropolitan Police and Devon & Cornwall FOIs) [1] [5]; Ministry of Justice quarterly offender statistics and its “Outcomes by offence” tools which include defendant/conviction tables by characteristics and ethnicity (the MoJ is the primary source for offender data) [2] [6]; and the Home Office collections and government publications on Foreign National Offenders which report FNO numbers and known data issues [3] [7]. The ONS compiles and publishes sexual offences prevalence and victim-characteristic datasets drawing on the CSEW and police-recorded crime but explicitly says it does not hold full offender nationality files and points researchers to MoJ/Home Office for offender nationality data [8] [4].
2. What the CSEW and ONS do — and what they don’t — about nationality
The Crime Survey for England and Wales (CSEW) — presented by ONS publications — is the best source for victim prevalence and some perpetrator characteristics reported by victims (relationship, age, sex), but it does not routinely provide robust offender nationality breakdowns and ONS directs users to MoJ for offender nationality/conviction stats [9] [10]. The ONS has recently redeveloped sexual victimisation questions (October 2025) and flags reduced sample-size cautions for some recent estimates; it stresses police-recorded crime and the CSEW capture different phenomena and that police counts reflect recorded demand rather than true prevalence [11] [12] [13].
3. Police FOIs: more granular but inconsistent and prone to missingness
Individual police forces have responded to FOI requests by publishing nationality-specific charge/chargeable-arrest spreadsheets (for example the Met’s release on people charged by nationality for 2022–23), but these releases commonly carry force-level caveats: nationality is self‑reported or administrative, ethnicity/nationality fields are sometimes inaccurate, and many records have “unknown” nationality — all of which undermine simple rate comparisons [1]. Investigative or advocacy groups then aggregate force FOIs to estimate shares (e.g., claims that a large share of London charges were foreign nationals), but those aggregated figures often exclude unknowns or make assumptions about them, amplifying uncertainty [1] [5] [14].
4. MoJ and Home Office FNO statistics: formal but labelled experimental and limited
The MoJ produces offender and defendant tables (including ethnicity) and the Home Office publishes statistics on foreign national offenders; however, official notes warn that systems used to compile nationality and deportation-linked data “face a number of issues” affecting quality and completeness, and some nationality breakdowns are being released as experimental statistics [2] [3]. The Migration Observatory also highlights that there are no reliable data on convictions or prison population by immigration status or length of residence — so nationality counts do not tell you whether a person is recently arrived, naturalised, or a long‑term resident [15] [16].
5. What reliability problems should users keep front of mind?
Available official sources repeatedly flag limitations: nationality is often self-reported or inconsistently recorded; many cases have “unknown” nationality; police-recorded crime trends are strongly affected by recording practices and changes to law (e.g., Online Safety Act offences); and population denominators used to compute rates (by nationality) can be outdated, biasing per-capita comparisons — for example critiques of high rate claims for small national groups note use of 2021 population snapshots against 2021–23 offence data can hugely overstate risk [13] [17] [3].
6. Practical advice for reporters, researchers and policymakers
Use multiple sources: combine MoJ offender/conviction tables, Home Office FNO statistics, police FOIs (with force notes), and ONS/CSEW prevalence data — and always report the denominator and how “nationality” was defined (country of birth, passport, self-identified nationality, or “foreign national offender”) [2] [1] [3]. Explicitly state data gaps: where sources do not record immigration status, length of residence, or have large “unknown” fields, say so [16] [3]. When seeing high per-capita ratios for small national groups, check whether population counts and offence years line up — critiques in national coverage flagged this mismatch as a major driver of exaggerated multipliers [17].
Limitations and missing items: available sources do not provide a single, fully reliable national dataset that cleanly links nationality, immigration status and sexual‑offence offending across all stages of the justice process — researchers must therefore triangulate across ONS, MoJ, Home Office and force-level FOIs and be transparent about uncertainty [8] [3].