Policy paper

Tackling error and fraud in the Covid-19 support schemes

Updated 13 October 2022

The COVID-19 support schemes helped millions of people and businesses through the pandemic.

The Self-Employed Income Support Scheme and Coronavirus Job Retention Scheme supported 14.6 million jobs and self-employed individuals. Over £98 billion was distributed in grants across both schemes.

The Eat out to Help Out Scheme was the first of its kind and was a key part of the government’s Plan for Jobs and was designed to support around 130,000 businesses and to help protect approximately 1.8 million employees.

Our priority was to get vital money to people quickly, to support people, businesses and the economy. We launched the schemes at unprecedented pace: just 4 weeks for the Coronavirus Job Retention Scheme and 7 weeks for Self-Employed Income Support Scheme.

We adapted our approach as the health and economic context changed to ensure the schemes continued to provide the support needed to help the economy recover as restrictions eased. We introduced a higher level of employer contributions in the Coronavirus Job Retention Scheme and put in place specific impact tests for the third and fifth grants for the Self-Employment Income Support Scheme. This allowed us to better target the grants at those most in need as the pandemic progressed, and then as restrictions eased.

Compliance activity

From the beginning we were clear that the schemes would be targets for fraud and that customers operating at pace and under pressure would make mistakes: that’s why we designed the schemes in such a way as to minimise fraud and error while not unnecessarily delaying payments.

We risk-assessed every single claim before they were paid: these pre-payment checks included matching lists of known suspect organisations and devices against claims. We looked at a variety of factors, including comparing the claims to historic data (for example, pre-pandemic payroll data), third-party information and other intelligence (like calls to our Fraud Hotline).

The speed of the government’s response and how the eligibility criteria were set helped ensure that organised criminals had very limited opportunity to exploit the schemes. As a result, the assessment of losses from organised crime are low.

Post payment compliance activity began in July 2020 once HMRC had been granted specific powers. To the end of March 2022, we have recovered more than £762 million lost to error and fraud in the COVID-19 financial support schemes through compliance activity.

Our compliance approach means that the total amount of money we have either blocked from being paid out (pre-payment) or recovered through our compliance work (post-payment) by the end of financial year 2021 to 2022 is more than £1.2 billion, with compliance activity still ongoing. We currently have 29 criminal investigations underway covering suspected fraudulent claims paid out for Coronavirus Job Retention Scheme, Self-Employed Income Support Scheme and Eat out to Help out worth around £15 million (although claims blocked or recovered on claims showing indications of criminal activity total over £150m).

HMRC’s compliance strategy is focussed on those who deliberately abused the schemes while providing a clear disclosure process for those who had overclaimed by mistake. Our compliance activity also aims to ensure that support reached those who needed it most; for example, we target cases where there are indications that employees have not been paid by their employer, even where the amounts are relatively small.

Taxpayer Protection Taskforce

At Budget 2021, the government announced a two-year investment of over £100 million in the Taxpayer Protection Taskforce to combat fraud in the COVID-19 financial support schemes administered by HMRC.

The Taxpayer Protection Taskforce is made up of tax specialists who have experience in responding to error and fraud. Tackling fraud in COVID-19 support schemes is a priority, which is why we moved more than 1,200 experienced staff from other tax compliance roles within HMRC, onto the taskforce. The investment has allowed us to replace the vacancies left by those who went to work on the taskforce.

HMRC ensured that experienced compliance officers were deployed to the taskforce, enabling us to effectively tackle fraud from the outset, while supporting customers where genuine mistakes have been made. We have opened over 40,000 compliance cases and contacted over 132,000 customers.

As more of these cases have been worked HMRC’s understanding of levels of error and fraud within the schemes has deepened. All available information points to this being at the lower end of our original estimates.

As a consequence, the amount the Taxpayer Protection Taskforce is expected to recover has been revised from between £800 million and £1 billion to between £525 million and £625 million across its lifecycle.

We found that the majority of large businesses got their claims right and that we were largely successful in stopping organised criminals gaining access to grants.

As planned, we will begin to wind our activity down from March 2023, with COVID scheme compliance activity moving to business-as-usual tax compliance work by September 2023. This will allow us to deal with all aspects of a customer’s potential non-compliance issues, related to the scheme and more widely. Moving the work of the Taxpayer Protection Taskforce into business-as-usual compliance activity is the most efficient way to ensure we protect and recover taxpayers’ money.

We are dedicated to tackling error and fraud in the COVID-19 financial support schemes. Our role is to make it easy to get tax right and hard to bend or break the rules. That is achieved by taking action against those who have deliberately sought to abuse the COVID-19 financial support schemes, while recognising there will be people who have made honest errors.

Throughout our compliance response, we’ve taken a reasonable and supportive approach to those who have made genuine mistakes – and have given these customers the opportunity to correct errors without fear of sanctions.

Error and fraud estimates

Across the full two-year lifecycle of the 3 schemes (covering 2020 to 2021 and 2021 to 2022), the total value of error and fraud is now estimated to be between £3.2 billion and £6.4 billion, with a most likely estimate of £4.5 billion (and error and fraud rate of between 3.3% and 6.5% with a most likely estimate of 4.6%), which is less than the previously published range for 2020 to 2021 alone (at a most likely figure of £5.8 billion).

To the end of March 2022, HMRC had recovered over £762 million (around £226 million in 2021 to 2022, and £536 million in 2020 to 2021, including activity prior to the establishment of the Taxpayer Protection Taskforce). The total amount of money we have either blocked from being paid out, or recovered through our compliance work, is therefore around £1.2 billion with compliance work on-going.

No amount of error and fraud is acceptable, but it is not possible to recover everything.

Estimates methodology

The error and fraud estimates were put together by experienced statisticians, with input and peer review from the UK’s Office for National Statistics and international colleagues. In the interest of transparency, we published detailed methodological notes for our estimates in July 2022, as we did in 2021 when we published provisional figures.

HMRC has worked hard to refine our error and fraud estimates as we gather more information, to ensure we are always using the best available evidence. New data from operational compliance activities focused on the Coronavirus Job Retention Scheme and Self-Employed Income Support Scheme error and fraud risks, a Random Enquiry Programme and updated Self-Assessment returns have provided an improved evidence base, allowing us to revise our estimates, which were published in the Annual Report and Accounts 2021-22. All estimates, by their very nature, come with a degree of uncertainty. We have recognised this uncertainty by using recognised statistical techniques to determine ranges for our estimates, rather than using single figures.

The available evidence supports our estimates, which were put together using best practice analytical methods. We are confident they are as accurate as they can possibly be given the data available, and we stand by them.