What we found – Value

3.1 To understand the value of both the police recorded fraud and computer misuse statistics and the Crime Survey for England and Wales (CSEW) fraud and computer misuse estimates, we spoke to a small number of users of the statistics. We focused on their uses of the statistics; how the statistics are understood and interpreted; the insight of the statistics; and known data gaps.

Reliability of data sources

3.2 All users that we spoke to said they regularly use the CSEW fraud and computer misuse estimates. They emphasised the reliability of the CSEW estimates for understanding the scale and nature of fraud and computer misuse experienced by individuals. The CSEW is seen as the core source for monitoring changes in the level of fraud and computer misuse over time and is therefore central to the development and monitoring of government fraud policy. Other uses of the CSEW estimates include research into the experiences of fraud victims.

3.3 In contrast, users told us that the police recorded fraud and computer statistics are not a reliable indicator of trends in fraud and computer misuse because of the level of under-reporting to Action Fraud. In the year ending September 2024, ONS estimated that just one in seven fraud offences and one in fifteen computer misuse offences were reported to Action Fraud or the police.

3.4 Some users raised concerns about some aspects of the City of London Police’s fraud and computer misuse crime recording processes, which we described earlier. As explained in the Introduction, the Action Fraud statistics do not provide a complete picture of all fraud and computer misuse crimes reported to the police and other bodies such as the National Cyber Security Centre and the Serious Fraud Office; they should also be reported to Action Fraud but are not always reported.

3.5 Despite the level of under-reporting, we heard that the Action Fraud data provide a helpful barometer of the fraud and computer misuse crimes that victims do report to Action Fraud. One of the main strengths of Action Fraud (and industry bodies) data is that they capture fraud and computer misuse crimes against businesses. These victims are not covered by the CSEW as it is a household survey. Action Fraud data also give an indication of the demand on the City of London Police.

3.6 ONS should better articulate the value of police recorded fraud and computer misuse statistics by explaining what insight they add on the scale and nature of these crime types.

ONS’s statistical bulletins

3.7 ONS’s Crime in England and Wales statistical bulletin summarises short-term trends in the police recorded fraud and computer misuse figures and the CSEW estimates. Until recently, ONS also published a Nature of fraud and computer misuse bulletin with more in-depth commentary on trends and data sources. ONS developed this topic-specific bulletin, which brings the various data sources together and presents simple messages on the trends in these crime types, in response to the recommendation in our accreditation letter for CSEW fraud and computer misuse estimates.

3.8 Most users told us that the bulletins give a clear and helpful high-level view of the scale of fraud and computer misuse, but they also said that ONS could add more insight by providing more-detailed commentary. The Nature of fraud and computer misuse data tables contain a wealth of information about these crimes and their impact on victims, highlighting the richness of the CSEW data, but the commentary in the bulletin is typically restricted to trends in the number of offences.

3.9 In 2023, ONS temporarily paused the nature of fraud and computer misuse bulletin due to resource constraints. It still updates the fraud and computer misuse data tables annually. We recognise the significant resource pressures that ONS is currently facing, but we question the decision to pause the nature of fraud and computer misuse bulletin. Fraud is the most common crime type experienced by CSEW respondents, and there is substantial user interest in the topic.

Development of the CSEW

3.10 Since the fraud and computer misuse module was added to the CSEW in 2017, ONS has added new questions to better capture the nature of fraud and computer misuse and refined existing questions to make them simpler to answer for respondents.

3.11 For instance, in 2023, ONS introduced questions on phishing, a form of social engineering where criminals use scam emails, text messages or phone calls to trick their victims, which has become one of the main methods used to commit fraud. ONS has also added more specific questions, for example, on whether the fraudster acquired the respondent’s personal details by accessing their social media, computer or emails.

3.12 ONS told us that it can be challenging to simplify the questionnaire as fraud and computer misuse is constantly evolving. One way in which ONS has simplified the questionnaire is to remove questions that double-checked certain aspects of the crime with the respondent; these were originally included to check data quality. ONS said that changes to questions undergo cognitive and user testing to ensure that they can be easily understood by respondents and provide reliable data.

3.13 ONS has collected data on fraud and computer misuse victimisation on a consistent basis since the module was introduced. This allows the headline estimates of the number of fraud and computer misuse crimes to be compared over time. However, occasionally, changes to the questionnaire create a discontinuity in the lower-level estimates. For example, the proportion of fraud incidents that were flagged as cyber offences for the year ending March 2023 and onwards cannot be compared to earlier years due to changes to the questionnaire. Although ONS caveats the figures in the data tables, this change means that it has not been possible to see changes in cyber-enabled fraud since 2017.

Interpreting the statistics

Police recorded fraud and computer misuse statistics

3.14 As with other police recorded crime statistics, trends in police recorded fraud and computer misuse are affected by changes in the level of reporting by victims and changes in police crime recording practices, which can make them difficult to interpret. For example, the number of UK Finance intelligence reports increased substantially between 2022 and 2024 because of a drive by UK Finance to improve the volume of intelligence that is shared with the City of London Police. It is unlikely that this increase represents a genuine increase in the scale of fraud offences reported to UK Finance.

3.15 In addition, the Cifas and UK Finance figures skew patterns of police recorded fraud. All fraud incidents recorded by UK Finance and most fraud incidents recorded by Cifas belong to one fraud type – cheque, plastic card and online bank account fraud. This fraud type makes up a relatively small proportion of all fraud offences reported to Action Fraud (around eight per cent as of March 2025). Given that the Cifas and UK Finance intelligence reports comprise around three quarters of all police recorded fraud each year, this fraud type is overrepresented in the statistics and may not be representative of the nature of fraud experienced across all individuals and businesses in England and Wales. As we recommended earlier, we want ONS to clarify the definition of ‘police recorded fraud’. This should help users interpret trends in the fraud statistics.

3.16 The City of London Police’s recording practices for fraud and computer misuse crimes have remained broadly similar since Action Fraud was established. However, the new fraud and computer misuse reporting and recording system that is being rolled out, together with the review of the Counting Rules for Fraud, will significantly impact how these crimes are recorded. The force told us that it expects these changes to increase the number of fraud and computer misuse crimes reported to Action Fraud. As a result, it will be more difficult to determine whether a change in the number of crimes is genuine. The CSEW estimates are not affected by such changes, which partly explains why they are a more reliable measure of trends in fraud and computer misuse.

Crime Survey for England and Wales statistics

3.17 A common misunderstanding about the CSEW fraud and computer misuse statistics, and the CSEW statistics in general, relates to the coverage of the survey. We often see statements like ‘fraud accounts for 40% of all crime’ to describe the scale of fraud in England and Wales, including in the UK Government’s Fraud Strategy and Home Office press releases. However, the CSEW does not measure all crimes. The survey only covers crimes against individuals and households; crimes against businesses and organisations, and crimes against society and the state, are not covered. The scale of fraud in these other populations is not as well understood, and therefore the proportion of all crimes that fraud makes up could be lower or higher than 40%.

3.18 It is important that fraud and computer misuse statistics are used accurately. The misunderstanding about the scale of fraud is partly due to ONS’s communication of the CSEW statistics. Until recently, ONS used the term ‘total crime’ to describe the CSEW headline crime estimates, which we consider to be misleading because it implies that the survey measures all crimes. We therefore welcome ONS’s recent improvements to the communication of the CSEW statistics. The statistical bulletin now contains clearer information about the limitations of the statistics, in particular, which crime types are and aren’t included in the headline estimate, and ONS has replaced ‘total crime’ with ‘headline crime’.

3.19 When using CSEW headline crime estimates, it is good practice to state explicitly whether the figure includes or excludes fraud and computer misuse offences. This is particularly relevant for longer-term comparisons, as the headline estimate including fraud and computer misuse only goes back to 2017. Over the last few years, we have commented publicly on this aspect of the statistics on several occasions. CSEW headline crime estimates excluding fraud and computer misuse can be used to compare the level of crime before and after 2016, but it is helpful to say that the figure excludes fraud and computer misuse offences.

Data gaps

3.20 We found that fraud against businesses is the key data gap in ONS’s fraud and computer misuse statistics. As explained earlier, the CSEW cannot measure fraud against businesses. Action Fraud data contain a mixture of fraud against individuals and businesses, but the majority of offences are reported by individual victims (the NFIB Fraud and Cyber Crime Dashboard provides a breakdown of reports by the type of victim). Cifas and UK Finance data fill a gap, but these industry bodies only capture a subset of fraud crimes experienced by businesses.

3.21 The Home Office is currently undertaking the Economic Crime Survey 2024 to add further insight on fraud against businesses. The survey asks a sample of UK businesses about their experiences of economic crime and the approaches they are taking to tackle it. It follows a similar survey conducted in 2020, which found that one in five businesses had been a victim of a known fraud in the last three years.

3.22 Several users told us they welcome the latest survey as it will continue to provide information on the scale and nature of fraud against businesses, in the same way that the CSEW does for fraud against individuals. Results from the Economic Crime Survey 2024 will be published as a research and analysis output rather than official statistics. We encourage ONS to include the findings from the survey in its statistical bulletins to tell a more complete story of fraud.