The National Audit Office has just published our overview of the impact of fraud and error on public funds for the new parliament.
It’s aimed primarily at members of parliament to brief them on the issue of fraud and error and how government can best tackle it. We have sent it to every MP and made it available for all to read on our website.
I would urge anybody working in this space to read it.
The level of fraud and error across government reached £55bn to £81bn in 2023-24.
This is our latest estimate using the Public Sector Fraud Authority’s methodology and the most recent figures from the 2023-24 accounts and spending statistics.
This includes:
- £39bn of fraud and error in taxation (part of the tax gap), of which HMRC expects to recover £9bn;
- £10bn of fraud and error in welfare, of which DWP has detected and can attempt to recover £3bn;
- £3bn of income and expenditure across the rest of government where there is some sort of fraud-and-error measurement, of which £1bn is detected so they can attempt recovery; and
- an assumption that there is between £3bn and £28bn (0.5% to 5.0%) across parts of government income and expenditure without proper fraud and error measurement.
The estimate has risen significantly since the government published its 2021-22 estimate, which stood at £39.8bn to £58.5bn. The level of government income and spend has increased significantly, welfare fraud has risen, and reporting has improved while fraud awareness has increased.
Our overview is based on our previously published work and information in the public domain. We decided to update the government’s estimate because it’s made up of various parts that have been updated since the estimate was last published – and we thought this was necessary to give MPs and the public a full picture of the scale of the problem.
In practice, we found that although we used PSFA’s methodology, updating the estimate was far from simple and required tricky adjustments and judgements about how to merge the parts together.
We had help from various experts across government, so I’d like to say a big thank you to them all. We’ve also seen saw a general improvement in reporting on fraud and error since the Covid-19 pandemic due to greater fraud awareness.
But there remain opportunities to refine the estimate – particularly to focus in on the risks in the £3bn to £28bn element of the estimate, where there’s no proper fraud-and-error measurement.
So why does this all matter?
The revised estimate brings to light the huge opportunity in tackling fraud and error. As we have long argued, tackling fraud and error isn’t about “compliance”, but about freeing up vast sums of money to deliver public services.
"As we have long argued, tackling fraud and error isn’t about 'compliance', but about freeing up vast sums of money to deliver public services"
The new government has set out its ambition to tackle both the tax gap and welfare fraud. It’s also announced the appointment of a new covid corruption commissioner to see if it can recover some of the £10.5bn of fraud and error lost in temporary Covid-19 schemes.
There’s also a significant annual problem across the rest of government – but it’s going to take rigorous risk assessment and fraud-and-error measurement processes to work out where and how best to tackle that.
What else does our overview draw out?
The overview sets out how government is structured when it comes to fraud and error and the setup of the PSFA. This includes details about its 2024-27 functional strategy for counter fraud. The strategy has built on our insights on the government’s strategic challenges to tackling fraud and error, including the need to develop the profession, harness the power of data sharing and analytics, and the need to build in fraud risk assessment and prevention into new programmes from the start.
We draw the parallel between the increased attacks against government and the prevalence of fraud in wider society. This has decreased since its peak during at the height of the Covid response (was it too many fraudsters stuck at home with nothing else to do?), but the Crime Survey for England and Wales still estimates there were 3.2 million frauds committed against adults and private households in the year to March 2024. The fraud-prevention service Cifas did a survey that found 12% of people in the UK admitted to committing fraud against a business or a public body in 2023, up from 8% from 2021.
We also set out some of the emerging fraud risks that we see. A lot of the frauds we see or have been reported to us through our whistleblowing line relate to simple failures of internal control – things that would be entirely preventable if public bodies improved their financial management. But we’re also seeing new threats emerge, whether it be the so-far theoretical threat of new technology (is it just that we can’t detect its use yet?), phoenixism (dissolving and resetting up companies to avoid paying obligations to government), and increased payroll fraud as more staff and contractors work from home.
Do take a look at the overview here and reach out on LinkedIn to tell us what you think.
Joshua Reddaway is director of fraud and propriety at the NAO. Find out more about the NAO’s work on fraud and error.