Dear Jen
Accreditation of the estimates from the Crime Survey for England and Wales
We have now completed our review of the estimates from the Crime Survey for England and Wales (CSEW). You asked us to consider reinstating the accreditation of the CSEW estimates following the temporary suspension of the accreditation in July 2022 due to quality concerns.
We are pleased to confirm the reaccreditation of the CSEW estimates, starting from the next release of the statistics on 24 October 2024. This is to allow time for you to make the improvements to the presentation of the statistics detailed below. You have committed to implementing these in the next release.
In June 2024, you published a quality review, which explored the impact of sample size and response rate on CSEW data quality. We are satisfied that you demonstrated that the quality of the current estimates is broadly similar to the pre-pandemic estimates and that you provided sufficient assurance to users about the limited impact of the lower response rate. Your team should continue to monitor the quality of the estimates, including the impact of the measures you have put in place to increase response rate.
In July 2024, we advised you that user concerns about the presentation and communication of CSEW statistics meant that we needed to carry out further work to ensure their value. To understand how you addressed these concerns, we reviewed the three most recent Crime in England and Wales statistical bulletins (released in January 2024, April 2024 and July 2024).
We found that you made substantial improvements to the clarity and insight of the CSEW statistics. Notably, since April 2024 the term ‘total crime’ has been replaced with ‘headline crime’ to reduce the likelihood of misinterpretation that the survey captures all crimes. In later releases the focus has shifted towards trends in individual crime types rather than aggregating them, providing more meaningful insights.
The bulletins now more-clearly explain both short- and long-term trends using written commentary and visualisations. Furthermore, later releases place greater prominence on data about domestic abuse, sexual assault, stalking, and harassment, in line with identified user need. Importantly, in making these changes you have actively engaged with users to address their concerns and improve the presentation of the statistics.
However, our review also identified several areas for improvement to further enhance the value of the CSEW statistics:
- The bulletin should be more explicit about which crime types are included and excluded in the ‘headline crime’ estimates. In particular, it should explain why sexual assault is excluded yet domestic violence is included, despite the interviewer-administered parts of the survey underestimating both crime types.
- To help users understand the change from ‘total crime’ to ‘headline crime’, a disclaimer should be added to previous bulletins.
- To aid user interpretation of the statistics, the bulletin should clarify the difference between prevalence and incidence measures along with associated limitations of using each approach.
- To improve clarity, the summary of short- and long-term changes in crime in the ‘main points’ section of the July 2024 bulletin (and future bulletins) should highlight the source of the statistics.
In addition, we reviewed a preliminary draft of the findings and recommendations from your internal review of the presentation and communication of the Office for National Statistics’ crime statistics for England and Wales. We understand that you plan to publish this review in the next few months. The review involved extensive user and stakeholder engagement to identify what you are doing well and how to make the statistics easier to understand and interpret. We fully support the review’s recommendations, which cover the accessibility of the statistics as well as clarity and insight.
You told us that you are working closely with Home Office statisticians and policy teams on how best to meet the emerging needs of government ministers, in particular the metrics required to measure the government’s pledge to halve violence against women and girls in the next decade. It is good that you are developing a strategy that brings together these emerging needs with the internal review’s findings. As part of this work, we encourage you to explore whether a crime severity index (a metric that weights offences by their seriousness) may be appropriate for measuring changes in violence over time.
Since the Covid-19 pandemic, you have made major changes to the design of the CSEW. The survey is now a longitudinal, multimodal survey, with some respondents taking part in multiple waves. As the current CSEW estimates are based on wave 1 data only, the accreditation applies to wave 1 data only.
We understand that you are investigating the quality and comparability of wave 1 and wave 2+ data and are exploring the feasibility of using data from different waves to produce crime estimates. We will continue to engage with your team as it develops these new statistics.
Yours sincerely
Ed Humpherson
Director General for Regulation
Related Links:
Ed Humpherson to Jen Woolford: Quality of the estimates from the Crime Survey for England and Wales