HMRC needs to ramp up checks on furlough fraud

The Treasury Committee has conducted a hearing with the National Audit Office about the reasons for fraud related to bounceback loans and the furlough scheme

Mel Stride, chair of the Treasury Committee, said: ‘The pandemic has created an extraordinary environment, what is your view as to where fraud has been most noticeable and where this has happened?’

One of the major areas of concern raised by MPs was the level of fraud related to the furlough scheme.

Gareth Davies, comptroller and auditor general, National Auditor Office (NAO), said: ‘One obvious area is employers overstating the furlough scheme – the scheme is vulnerable to employers inflating number of staff, number of claims; the other thing is claiming the money and requiring your staff to work; and third is organised crime, ether coercing employers to make claims or through agents, false agent arrangements with tax agents.

‘For the self employed scheme the controls are a lot tighter and so the opportunities for fraud are much more limited. As the schemes have been extended there is more opportunity for fraudsters to adapt their methods. HMRC has responded with greater awareness of how people are pushing the boundaries of the schemes, and has introduced tighter definitions of how businesses and individuals are ‘adversely affected by the pandemic’.

‘There have been improvements in control – so for example, since January HMRC has been publishing lists of employers claiming furlough. That transparency was a deterrent for future fraud. That is an area where control has improved, but increasing complexity certainly makes error more possible and there is more opportunity for fraud.’

HMRC works on the assumption of 5-10% fraud on furlough scheme, 1-2% for the self employed scheme. Davies said that part of the HMRC compliance process has seen the specialist furlough taskforce at HMRC contacting around 10,000 employers to question them on their use of furlough schemes.

Davies added: ‘Estimates need to be constantly refined. We have been in dialogue with HMRC on reviewing these estimates. They need to be updated more regularly. Some of the anti fraud measures used on grants are probably more readily used on these schemes.

Andy Morrison, director, HMRC, at NAO said: ‘Although their [HMRC] approach is designed at getting a robust estimate by the end of 2021, we would like to see updated estimates while the schemes are operating. Almost 10% of people were working in the first phase of furlough while employers had claimed furlough and some people were working part time.

‘Like all fraud, whistleblowing is one of best ways to identify fraud. There’s been a big investment in extra compliance staff at HMRC and that has taken some time to bring people in.

‘ It is just worth pointing out that whistleblowing had to be done through an online form until August. HMRC has always made clear to employers that they need to keep records and will be carrying out compliance checks. Most recent figures showed that HMRC were going to check 10,000 companies and more funding was announced in the Budget but we haven’t seen the results of that yet.’

Davies added: ‘HMRC has a good track record on investment in compliance staff, for every pound spent on tax recovery nine in £10 is recovered. But it is unlikely to be so high for CJRS.’

While furlough fraud is one issue affecting the pandemic support package, there are also concerns about abuse of bounceback loans with estimates that between 35% and 60% of bounceback loans may not be repaid.

Davies said: ‘The fraud risk has been elevated by the pandemic and it has affected bounceback loans. This is the area we are most concerned about – there is a credit risk where people cannot pay back loans and in the fraud control area.

‘The big challenge is the way that those schemes were accelerated – this meant that the process was accelerated removing credit checks – this meant that lenders did not go into detail about the provenance of the claims, and also there was a high risk of identity theft for fraudulent claims. These are some of the reasons why bounceback loans were vulnerable.’

As the loans were underwritten by the government, rather than the lenders, qualifying criteria were not as robust as usual. The British Business Bank has been carrying out random sampling work to identify risk levels and this will produce two data streams – those paying back loans and those not, which will give a clearer picture of how to identify risk.