The National Audit Office (NAO) has slammed the lack of take-up of data sharing provisions in the Digital Economy Act to help prevent fraud.
Its Tackling fraud and corruption against government report says government is only at an early stage of using intelligence, data matching and data analytics to prevent fraud and corruption – a situation that has helped lead to the taxpayer losing as much as £21 billion to fraud since the start of the pandemic. The figure is four times as much as the prior two years.
But most of government still does not know how much fraud and corruption it faces, and still cannot show that it is tackling it effectively.
That’s particularly poor, says the public spending watchdog, as it is now almost six years since the Digital Economy Act of 2017 was passed into law.
It specifically allows sharing of information between public organisations through Information Sharing Agreements under its anti-fraud provision.
“There have been 20 agreements under the Digital Economy Act 2017 for the purposes of preventing or disrupting fraud,” says the report,
“However, officials working in counter-fraud told us that the processes in place for sharing data, both between and within departmental groups, are often slow and burdensome, often resulting in incomplete or time lagged data being shared.”
Focus and prioritisation
Last year, NAO acknowledged that the benefits of better data are not simple to achieve in practice, and will require focused effort, funding and prioritisation to achieve – plus that legacy systems and ageing data “present a barrier” to effective use in government, as they create complex practical issues for data sharing and the challenges involved in using these data is not well understood.
It has welcomed the setting up of the Public Sector Fraud Authority (PSFA), which intends to improve the use of intelligence across government and provides an opportunity for government to improve “both its understanding of fraud” and “the action it takes to prevent it”.
Nonetheless, that team has been in operation for less than a year and will need to (among other actions) require a new cadre of counter-fraud and corruption professionals capable of working with other professions to assess risks and develop digital and operational means to tackle fraud.
A key early win here could be looking at using technology to perform real time risk assessment of transactions – with the report noting that DWP uses HMRC real time payroll data to verify benefit payments.
However, it says, most public bodies using data approaches to detect fraud use time lagged data, which has to be manually formatted.
To use real time data, systems need to be integrated, but, states the report: “Department counter-fraud staff told us that the processes in place for sharing data, both between and within departmental groups, are often slow and burdensome, often resulting in incomplete or time lagged data being shared.”
There is also variability in the quality and format of data as there are no standards across government, it warns.
Crime and error
That £21 billion figure, by the way, is short of tax lost to evasion and criminal attacks and excludes loss from error.
All in all, the PSFA’s estimate of the extent of fraud and error across all of government in 2020-21 was actually £33.2 billon to £58.8 billion out of £1,106.1 billion of expenditure and £608.8 billion of tax income.
The report also points out that no less than two thirds of central government expenditure is not subject to any direct fraud and error measurement.
