Developing a combined measure of domestic abuse, sexual assault and stalking, England and Wales: July 2025

This article outlines our method for producing a new combined prevalence measure of domestic abuse, sexual assault and stalking from the Crime Survey for England and Wales (CSEW).

In this article, the term "combined measure of domestic abuse, sexual assault and stalking" refers to people who have experienced one or more of these crime types in the last year. We refer to domestic abuse, sexual assault and stalking as "crime types", but recognise that, in some cases, a criminal offence may not have occurred.

The combined measure has been produced as the main measure for monitoring the Government's ambition to halve violence against women and girls (VAWG) in a decade. Home Office will provide more detail on the use of this combined measure and other metrics that will be used to monitor progress against this ambition, in a new cross-government VAWG strategy that will be published later this year.

While testing a new set of domestic abuse questions on the CSEW, not all respondents were asked the full set of questions on domestic abuse, sexual assault and stalking. This means we have developed an innovative approach to produce the combined measure, as described in Section 4: Method used to produce the combined measure. Particular caution should be taken when interpreting estimates from the measure, and we are unable to calculate whether differences observed are statistically significant. All estimates presented in this article are official statistics in development.

• For the survey year ending March 2024, we estimate that 11.3% of people aged 16 years and over were victims of at least one of the crime types of domestic abuse, sexual assault or stalking in the last year (14.0% of women and 8.6% of men).

• For the survey year ending March 2025, we estimate that 10.6% of people aged 16 years and over were victims of at least one of the crime types of domestic abuse, sexual assault or stalking in the last year (12.8% of women and 8.4% of men).

In the future, we will be able to produce the measure directly by following standard survey processes. This will allow us to conduct statistical testing.

We have been working closely with Home Office on how crime statistics can be best used to support the Safer Streets mission. This article focuses on how the Crime Survey for England and Wales (CSEW) can be used to provide insight on the Government's ambition to halve violence against women and girls (VAWG) in a decade.

As the main measure for monitoring this ambition, Home Office have asked that we combine the prevalence measures of domestic abuse, sexual assault and stalking from the CSEW. These are considered "some of the biggest crimes that most heavily affect women and girls" (see Question 52, Home Affairs Committee, Oral evidence 17 December 2024).

Home Office will provide more detail on the use of this combined measure and other metrics that will be used to monitor progress against this ambition, in a new cross-government VAWG strategy that will be published later this year.

Crime Survey for England and Wales

The CSEW is a face-to-face victimisation survey asked to people aged 16 years and over resident in households in England and Wales. The main aim of the CSEW is to provide robust trends for the crime types and population it covers.

Self-completion modules are used in the CSEW to collect information on topic areas (such as domestic abuse, sexual assault and stalking) that respondents may feel uncomfortable talking about to an interviewer. Respondents answering questions directly on a tablet allows them to feel more at ease, because of increased confidence in the privacy and confidentiality of the survey. Previous analysis has shown that rates for domestic abuse derived from the self-completion module are substantially higher than those obtained from the face-to-face interviews. See Annex 1: Measuring domestic abuse in our Domestic abuse prevalence and trends: year ending March 2019 article for more information.

One of the main strengths of the CSEW is that it covers many crimes that are often not reported to the police. This is particularly relevant for VAWG-associated crime types, like domestic abuse, which have low reporting rates. For more information on the CSEW, see Section 2: CSEW of our User guide to crime statistics for England and Wales.

As crime types evolve over time, we carry out continuous improvement work to the CSEW to ensure the statistics remain relevant and accurate. Following research to develop a new set of domestic abuse survey questions, and a split-sample trial between April 2023 and March 2025, we concluded the new questions provided a better measure of domestic abuse. As a result, we made the decision to fully replace the existing questions with the new questions from April 2025. Further information is available in our Evaluating a new measure of domestic abuse article.

The new questions provide a better measure, as they include more abuse types and cover a wider range of abusive behaviours. For example, the new questions include behaviours concerning marital status-related abuse and health abuse, which were not previously covered. Other abuse types were partly reflected in the previous questions (for example, economic abuse), but the new questions include a wider range of abusive behaviours that respondents may identify with. For more information on the new questions and abuse types, see Section 11: Concepts and definitions of our Evaluating a new measure of domestic abuse article.

As the new questions provide a measure that more accurately captures victims' lived experiences, it is important that the combined measure includes the estimates from these new questions. Using the previous domestic abuse questions in the combined measure would be misleading because the time series going forwards would be inconsistent.

Through conversations with Home Office, we agreed that the combined measure should be produced as far back as possible, and that the new domestic abuse questions should be used. As a result, the measure has been produced back to year ending March 2024, when the new questions were first included on the survey.

Our aim is to produce an estimate of the proportion of the population aged 16 years and over in England and Wales who have experienced domestic abuse, sexual assault or stalking in the last year.

It is possible that some victims will have experienced a crime type multiple times or experienced more than one crime type.

Such a statistic is known as a prevalence measure, as it identifies the proportion of victims without considering the number of times people were victimised.

We already publish separate prevalence measures for domestic abuse, sexual assault and stalking in Tables A9 to A11 of our Crime in England and Wales: Annual supplementary tables, year ending March 2025. As some people may have experienced more than one crime type, adding the measures together would result in estimating too many victims. This is because a victim of more than one crime type would be counted more than once.

If people completing the Crime Survey for England and Wales (CSEW) are asked about their experiences of all crime types, we can directly produce a combined measure. However, while testing the new, longer set of domestic abuse questions, limited space available on the CSEW meant that we could not ask respondents about non-domestic sexual assault and non-domestic stalking as well. Therefore, we could not use the direct approach.

Instead, to create a combined prevalence measure of domestic abuse, sexual assault and stalking for year ending (YE) March 2024 and YE March 2025, we have developed an alternative method to make best use of the information available from the survey. This method applies a different and innovative approach, compared with the direct method, to estimate the proportion of victims in multiple categories.

When we produce the combined measure for YE March 2026 next year, changes to the survey will allow us to use the direct method instead. We will then review whether it is possible to use this data to improve the estimates produced for YE March 2024 and YE March 2025.

Following research to redevelop our survey questions on domestic abuse, new questions were added to the CSEW in April 2023, as part of a split-sample trial that ran until March 2025.

From April 2023 (for the survey years ending March 2024 and March 2025), half of survey respondents (split-sample 1) were asked the existing set of domestic abuse questions (DA), alongside questions on sexual assault (SA) and stalking (ST). Questions on sexual assault and stalking captured information on both domestic crimes (perpetrated by a partner, ex-partner or family member) and non-domestic crimes (perpetrated by someone other than a partner, ex-partner or family member).

The other half of survey respondents (split-sample 2) were asked the new set of questions on domestic abuse (DA2). Because of limited survey space, respondents who were asked the new set of questions on domestic abuse were not also asked questions on non-domestic sexual assault and non-domestic stalking. Questions on domestic sexual assault and domestic stalking were asked as part of the new domestic abuse questions.

As the two halves of the split-sample did not overlap, it was not possible to know whether someone who was identified as a victim of domestic abuse through answering the new domestic abuse questions, was also a victim of non-domestic sexual assault, or non-domestic stalking, or both.

As the sexual assault and stalking questions were asked of the same people, we know how many people were victims of both sexual assault and stalking (SA and ST). However, we do not know the true overlap between the new domestic abuse questions and the sexual assault and stalking questions (DA2 with SA and ST).

Through consultation with topic experts and key internal and external stakeholders, we explored and assessed numerous options. Our chosen method makes best use of the available data, prioritising using the data collected from the new domestic abuse survey questions. It brings together data on domestic abuse from the new domestic abuse questions (including domestic sexual assault and domestic stalking), with data on non-domestic sexual assault and non-domestic stalking from the sexual assault and stalking questions (see Image 1).

Image 1: Method for creating a combined measure of domestic abuse, sexual assault and stalking, using the new domestic abuse questions

Source: Office for National Statistics

Notes:

“Other types of domestic abuse” include abuse types such as economic abuse and physical abuse. See Section 11: Concepts and definitions of our Evaluating a new measure of domestic abuse article for more information on the new questions and abuse types.

Download this image Image 1: Method for creating a combined measure of domestic abuse, sexual assault and stalking, using the new domestic abuse questions

.png (47.5 kB)

Step 1: Include respondents from split-sample 2 who experienced domestic abuse

By initially using data from the new domestic abuse questions (DA2), anyone who had experienced any type of domestic abuse (including domestic sexual assault and domestic stalking) was counted in the combined measure. These people may have also experienced non-domestic sexual assault or non-domestic stalking. However, as respondents to split-sample 2 were not asked questions about their experiences of these crime types, it is not possible to know.

Step 2: Add respondents from split-sample 1 who experienced non-domestic sexual assault or non-domestic stalking (but not domestic abuse)

To complete the combined measure, it was necessary to add respondents from split-sample 1 who had experienced non-domestic sexual assault, or non-domestic stalking, or both, to the DA2 data. Respondents from split-sample 1 who had experienced non-domestic sexual assault or non-domestic stalking were only included if they had not also experienced any type of domestic abuse (including domestic sexual assault and domestic stalking).

Step 3: Ensure respondents are only counted once

It was necessary to ensure that anyone who had experienced domestic sexual assault or domestic stalking was removed from the non-domestic sexual assault and non-domestic stalking estimates in split-sample 1. This lowered the risk of overcounting respondents, as those who had experienced any form of domestic abuse (including domestic sexual assault and domestic stalking) were accounted for in step 1.

Although this method provides the best estimate for the combined measure, it does have limitations. The new domestic abuse questions capture a wider range of abusive behaviours. Therefore, we expect that some respondents in split-sample 1 who had experienced only non-domestic sexual assault or non-domestic stalking, may have also experienced domestic abuse, had they been asked the new domestic abuse questions. While this would lead to a risk of double-counting respondents across the two separate samples, we expect this overcount to be low and the impact on the combined measure to be minimal.

The data presented in this section are official statistics in development and are subject to change as we evaluate future data and finalise methods. In addition, the method used to produce these estimates adds an additional level of uncertainty and prevents us from using standard statistical testing to assess differences. Caution should be taken when interpreting these estimates, particularly when making comparisons between years.

The Crime Survey for England and Wales (CSEW) year ending (YE) March 2024 estimated that 11.3% of people aged 16 years and over were victims of at least one of the crime types of domestic abuse, sexual assault or stalking in the last year (14.0% of women and 8.6% of men). This equates to 5.4 million people aged 16 years and over (3.4 million women and 2.0 million men).

Data for YE March 2025 showed that an estimated 10.6% of people aged 16 years and over were victims of at least one of the crime types of domestic abuse, sexual assault or stalking in the last year (12.8% of women and 8.4% of men). This equates to 5.1 million people aged 16 years and over (3.2 million women and 2.0 million men).

There was a 0.7 percentage point difference in prevalence between the latest data and YE March 2024. However, limitations with the survey and the resulting method used to create the estimates prevent us from using standard statistical tests. It is not possible to assess whether the difference is statistically significant. For more information, see Section 4: Method used to produce the combined measure.

From next year (YE March 2026), changes to the survey mean we can use a direct method. This means we will be able to conduct significance testing between years. In addition, we aim to publish the combined measure broken down by a range of characteristics.

CSEW data on the combined prevalence measure for domestic abuse, sexual assault and stalking can be found in Table A8 of our Crime in England and Wales: Annual supplementary tables, year ending March 2025.

From April 2025 onwards, all Crime Survey for England and Wales respondents are asked all questions on domestic abuse, sexual assault and stalking at the same time. This means we will be able to produce a direct combined measure of domestic abuse, sexual assault and stalking in the future.

Alongside our Crime in England and Wales annual bulletin in July 2026, we will publish a combined measure of domestic abuse, sexual assault and stalking for year ending (YE) March 2026. At this point, we will consider whether it is possible to revise data for YE March 2024 and YE March 2025, based on the known overlaps in the questions for YE March 2026.

If you have any questions about our combined measure, please contact us at crimestatistics@ons.gov.uk.