How do reported sexual assault rates vary between white, Black, Asian, and mixed-ethnicity groups in the UK?
This fact-check may be outdated. Consider refreshing it to get the most current information.
Executive summary
The Crime Survey for England and Wales (CSEW), published and summarised by the Office for National Statistics (ONS), shows that reported sexual assault victimisation varies across ethnic groups but the publicly described pattern depends heavily on which dataset, age range and time-period are used and on significant data limitations; the ONS provides ethnicity breakdowns in specific tables (for example Table 5 for year ending March 2020 and Table 4 for year ending March 2022) rather than a single clear, up-to-the-minute ranking of rates by White, Black, Asian and Mixed groups [1] [2] [3]. Detailed per‑group numeric comparisons therefore require consulting those ONS tables or the CSEW annual supplementary tables (Table A11c) because headline prose in media and advocacy reporting often omits the sampling caveats the ONS itself highlights [4] [1].
1. What the official surveys actually provide
The ONS uses the CSEW to estimate the proportion of people aged typically 16–59 (and sometimes 16–74 in combined publications) who were victims of sexual assault in the previous 12 months and publishes breakdowns by ethnic group in specific tables—examples cited by the ONS include Table 5 (combined CSEW for year ending March 2018 to March 2020) and Table 4 in the year ending March 2022 release—so comparisons by ethnicity are available but sit inside those formal tabulations rather than in a single summary narrative [1] [2] [3].
2. Variation exists but the direction depends on metric and period
ONS material makes clear that the prevalence of sexual assault differs between ethnic groups in the CSEW outputs, but the precise ordering (which group has higher or lower reported rates) is driven by which years and age-ranges are combined, and whether the measure is “any sexual assault” or a sub-type; the ONS therefore cautions readers about reduced sample sizes and lower response rates in more recent releases (notably the six‑month‑based year ending March 2022 tables), which can affect apparent differences between White, Black, Asian and Mixed groups [1] [2].
3. Important data limitations and missing pieces
Crucial caveats: the CSEW captures both reported and unreported experiences but is a household survey with sampling limitations; ethnicity breakdowns are for England and Wales only (Scotland and Northern Ireland are separate) and the ONS flags that some tables are underpowered for fine-grained breakdowns or by type of assault, while reliable defendant/offender ethnicity data are patchier and not held centrally in the form some FOI requesters seek [3] [5] [6]. The Ministry of Justice and Crown Prosecution Service hold complementary but differently structured records, and the CPS says it does not centrally record some offence attributes such as gang involvement in the way FOI requesters sometimes assume [7] [6].
4. How media and political debates complicate interpretation
High‑profile claims linking migrant status or particular ethnic groups to higher sexual offence rates have circulated in recent years, but outlets and fact‑checks note imperfections in nationality and ethnicity recording (self‑reported nationality, differences between recorded suspects and convicted defendants) and stress that police or conviction figures are not equivalent to survey prevalence rates—data caveats that the ONS and independent fact‑checks emphasise [8] [4]. This means political narratives (for example about “foreigners” or specific community patterns) can outpace what the ONS datasets actually quantify, and selection of particular datasets or time windows can produce very different impressions [8] [9].
5. Bottom line for comparative claims and how to get precise numbers
The authoritative public sources note that sexual assault prevalence does vary by ethnic group in England and Wales, but the exact comparative rates for White, Black, Asian and Mixed groups require consulting the ONS/CSEW tables (notably Table 5 for combined 2018–2020, Table 4 for the 2022 release, and the A11c supplementary series for 2005–2025) because headline summaries omit the fine print and recent tables may have limited sample sizes; offender ethnicity and conviction data are available in other MoJ/CPS datasets but are incomplete and recorded differently from victimisation survey data [1] [2] [4] [6]. Where the sources are silent or limited—such as up‑to‑date national comparisons that fully adjust for age, location or reporting differences—this reporting does not assert a definitive numeric ranking and instead points readers to those specific ONS tables for exact rates [1] [4] [3].