Implications of changing domestic abuse measurement on the Crime Survey for England & Wales
The Office for National Statistics (ONS) is making a major decision this month on the future of Crime Survey for England and Wales (CSEW) Domestic abuse measurement and monitoring.
Last year, ONS ran an experiment where half of the CSEW sample got the domestic abuse module used since 2005, and the other half got a new module that is not comparable with the previous one. ONS intend to move over entirely to the new module in the next data collection (2025/26).
Loss of the existing module has major implications: it is world-leading, uses globally comparable items, and with trend data going back to 2005. Without consistently administered core items from that module, it will no longer be possible to:
- Produce long-term trends over time in domestic abuse for England and Wales.
- Group a decade of survey years together to have enough cases to robustly examine domestic abuse in particular regions, minoritised groups, and by other protected characteristics for many years. This is essential for understanding inequalities in violence and subsequent service contact, and whether these are changing.
The new module is problematic for many reasons:
- Is not a standardised measure, has undergone little validation or psychometric testing, and is not comparable with anything used previously or in any other country or study.
- It separates data collection between former and current partner based on relationship status at the time of the interview, not at the time of abuse. This distinction creates confusion for interpretation of analysis and may be misinterpreted. The distinction is also problematic for classification of casual and other relationship types.
- The overhaul of the module was intended to align measurement with the Domestic Abuse Act 2021 definition, but it appears that domestic abuse as recognised by that Act cannot be identified by this module.
We urgently recommend that before losing this world-leading time series and relying on an untested, not comparable, and flawed new approach to DA measurement in England and Wales, that ONS:
- Pause: continue the split-sample data collection for one more year.
- Test the new approach: fully compare data collected using the new and old modules data so the validity and utility of the new measures can be evaluated appropriately, and its impact on inequalities assessed.
- Publish these results publicly: and fully consult once stakeholders understand all the implications of having data collected in each way before the decision to roll out new data collection is finalised.
- With this information, then compare all options: such as maintaining some of the existing questions alongside adding new coercive control items. This straightforward approach would ensure the utility of the survey for national trends (in both England and Wales) and analysis of inequalities and minoritised groups, while also improving the measurement of coercive control.
We urge others who feel similarly to contact ONS at CrimeStatistics@ons.gov.uk or contact us at VISION_Management_Team@city.ac.uk if you would like to discuss.
Note that ONS is planning a raft of further changes with similar implications for trends and analysis of minoritised groups, including:
- Removal of the sexual victimisation module from next data collection (2025/26), with redevelopment at some future date.
- Removal and redevelopment of the nature of partner abuse questions, which cover DA survivors service use and police contact and are essential to understanding whether some groups are underserved by services.
These will further undermine continuity of data for trends and the ability to analyse minoritised groups or by protected characteristics.
For researchers interested in combining CSEW waves to enable robust analysis of inequalities by protected characteristics and for minoritised groups, VISION researcher Niels Blom has published syntax: https://vision.city.ac.uk/news/new-possibilities-created-by-crime-survey-wave-integration/.
Photo from Adobe Photo Stock subscription
