National Statistical

Domestic abuse can have devastating consequences yet often remains hidden. Our best measure for understanding the scale of this abuse is the Crime Survey for England and Wales (CSEW) with estimates used to inform policy decisions and ultimately improve outcomes and support for victims. The latest estimates show a large increase in the prevalence of domestic abuse. Meghan Elkin explains why it is too early to determine whether this represents a new, higher prevalence level and some caution around interpreting the size of the increase.

Questions about domestic abuse were included annually on the CSEW from 2004 and had remained largely unchanged, allowing for a long comparable time series. However, new laws on coercive and controlling behaviour, and the introduction of a definition of domestic abuse in the Domestic Abuse Act 2021, meant there was a need to improve the survey to provide a more comprehensive picture of abuse, as we explain here.

Following extensive research, we asked the new questions to half of the respondents to the survey from April 2023, alongside the previous questions as part of a pilot, before asking everybody from April 2025.  These new questions account for the sensitivities and challenges of gathering data on this topic. They better reflect the behaviours experienced by victims and provide policy makers with more information to tackle domestic abuse and put services in place to support victims.

Because we’re using new questions, we no longer have a long-term time series. All surveys have natural fluctuations because they’re based on a sample, but being a new time series makes it even more difficult to take a view on year-to-year change.

What the latest data shows

Today we’ve published domestic abuse estimates for the year ending (YE) March 2026. This is the first time estimates have been produced using the new questions for the full survey sample. We did not publish these in our annual release to allow for additional quality assurance and to better understand changes to the survey.

We’ve found around 9.6% of people aged 16 years and over experienced domestic abuse in the last year (around 4.7 million people), as estimated by the YE March 2026 survey. This represented a significant increase when comparing responses from equivalent questions from the YE March 2025 survey (when it was 7.8% or around 3.8 million people).

It was driven by a significant increase in partner abuse (7.9%) compared with the previous year (6.1%). However, there was no significant difference in experiences of family abuse (which excludes partners) in the YE March 2026 (2.8%) compared with the YE March 2025 (2.7%).

Some minor changes to the survey are likely to have influenced the size of the increase. With all survey estimates, there are various sources of uncertainty and with it being a new time series based on new questions, that is more difficult to interpret.  Our best assessment is that the increase we are seeing includes real change in the prevalence of domestic abuse, but it’s too early to interpret what this means for trends in domestic abuse.

What changes were made

Making improvements and changes to surveys over time is best practice to ensure the continued relevance of a survey. Numerous changes are made to the CSEW each survey year based on stakeholder needs and priorities, respondent feedback and ongoing research and testing. The changes can include both large-scale changes such as the addition or removal of questions, as well as smaller changes such as to question wording.

At the start of this survey year (April 2025), there were some changes made to the questionnaire itself and how the survey was implemented.

• The survey software system for delivering the crime survey questionnaire was changed and respondents are now doing the self-complete questions on a tablet rather than a laptop.
• Everyone completing the survey was asked the new domestic abuse questions for the first time (previously they were asked of a subsample while testing was conducted) along with everyone being asked the new sexual victimisation and stalking questions.
• Minor changes to some wording and presentation within the domestic abuse questions were made.
• An issue was identified with how questions about respondents’ relationship status were set up in the survey, which has now been corrected.

It is not possible to quantify the effect, if any, each of these changes has had on the overall estimates of domestic abuse. We will continue to capture people’s experiences of domestic abuse and will monitor and report on emerging trends over the coming years.

Next steps

In November we’ll produce our annual compendium of domestic abuse statistics, which will give more insight into the characteristics of victims and abusive behaviours experienced.

It will include data from the Crime Survey for England and Wales (CSEW), police recorded crime (PRC) and other organisations across the criminal justice system. The compendium will provide more detail on a greater number of abuse types as well as a measure of the scale of domestic abuse that victims experienced.

While challenging, producing domestic abuse statistics is vital for policy makers. Today’s figures directly feed into the combined measure of domestic abuse, sexual victimisation, and stalking which we’ve also published today.

This measure was produced as the main measure for monitoring the Government’s ambition to halve violence against women and girls (VAWG) in a decade, and refers to people who have experienced one or more of these crime types in the last year. We estimate that 10.9% of people aged 16 years and over (around 5.3 million people) experienced one of these crime types in the YE March 2026.

In December we will publish our evaluation on whether this estimate is comparable with the previous estimates we published. We will also provide information on the prevalence of experiencing more than one of the three crime types.

As ever, if you have any feedback on the statistics we’ve produced, please email crimestatistics@ons.gov.uk.