Contents:

1. Summary

2. Background

3. HMRC’s statutory ability to publish this information

4. Comparison with the United States

5. Recommendations

6. Acknowledgements

7. Latest news

This briefing explains why HMRC can and should publish data on the employers that have received more than £41 billion of furlough support under the Coronavirus Job Retention Scheme.

HMRC has told Parliament it can’t share employers’ names for legal reasons. However, since this is the only realistic large-scale option for recovering up to £4 billion lost to fraud, we think HMRC’s governing legislation permits it to disclose the data.

You can read this briefing as a PDF.

To discuss any of the issues here, please contact contact@centreforpublicdata.org.

• Since April, more than 1.2 million UK employers have received grants of £41.4 billion to furlough employees​. HMRC estimates that up to 10% of grants may have been subject to fraud or error, meaning that up to £4.1 billion may have been defrauded​.

• MPs have asked HMRC to publish the names of employers receiving furlough, to allow public scrutiny of these employers. However, HMRC has told Parliament that it is not legally permitted to disclose information about individual taxpayers.

• However, our research suggests that HMRC can legally disclose such data when it is “reasonably necessary for HMRC to fulfil its primary function” of revenue collection and management.

• In this situation, publishing employers’ names to allow employees to identify and report fraudulent use of the scheme appears to be the only realistic option to recover very large losses to the taxpayer.

• HMRC therefore can and should publish the names of employers who have received furlough support since April.

• Similar data has already been published in the United States. As in the US, employers that have repaid their grants could be excluded from publication, to allow the option of privacy. Payments below a ​de minimis​ could also be excluded.

• It is unfortunate that this data was not published from the start of the scheme, when it could have prevented fraud. HMRC should update its processes to include appropriate consideration of data reporting when new schemes are implemented.

The Coronavirus Job Retention Scheme (CJRS) was launched on 20 April 2020, offering grants to UK employers to furlough employees unable to work because of Covid-19 restrictions. As of 18 October 2020, 1.2 million employers had received grants of £41.4 billion via the scheme​.

Concerns have been raised about significant fraud in the scheme, primarily via employers claiming grants but not passing them to employees. HMRC told the Public Accounts Committee that up to 10% of grants may have been affected by fraud and error​. The National Audit Office (NAO) found that 9% of surveyed employees reported working illegally while furloughed at the request of their employer.

Clearly, employees cannot report employer fraud if they are unaware that their employer has made a furlough claim. HMRC could have enabled employee whistleblowing, and thus reduced this type of fraud, by publishing the names of employers using the scheme, but it chose not to do so when the scheme was launched, fearing this could deter legitimate claims. (However, HMRC will publish the names of employers who claim under the CJRS from December 2020 onwards​.)

Requests for HMRC to publish the names of employers supported up until November, to allow public scrutiny and enable whistleblowing, have been denied. HMRC told journalists that it was not legally allowed to publish this data: ​“[our] duty of confidentiality under section 18(1) of the CRCA removes any possibility of disclosure on a discretionary basis”​. Parliamentary requests for the information have been denied on the same grounds.

It is unclear whether HMRC is correct to state that there is no legal possibility of disclosure. HMRC has cited Section 18(1) of the Commissioners for Revenue and Customs Act 2005 (CRCA), which​ s​ays ​“Revenue and Customs officials may not disclose information which is held by the Revenue and Customs in connection with a function of the Revenue and Customs”​. However, Section 18(2)(a) states: ​“Subsection (1) does not apply to a disclosure which is made for the purposes of a function of the Revenue and Customs”.

In other words, disclosure ‘for the purposes of a function’ of HMRC is permitted. HMRC’s legal functions include the ‘management’ of CJRS payments​. In 2016, the Supreme Court ruled that the law permitted HMRC to disclose information where “reasonably necessary for HMRC to fulfil its primary function” of revenue collection and management.

In the case of the CJRS, HMRC now plans to recover fraudulent grants via voluntary disclosure and spot checks, and estimates that this will recover around £275 million, albeit at an opportunity cost of £200 million. However, this is only a fraction of the potential losses, which HMRC is still estimating but believes may be up to £4 billion​.

We suggest that publication of CJRS recipients is now the only way for HMRC to recover significant sums of taxpayers’ money, as this is the only way that employees can identify and report fraudulent use of the scheme to HMRC. This method would also be far lower-cost than any other method of recovering fraudulent payments.

Therefore, this publication falls within the scope of information that can legally be disclosed, as it is necessary for HMRC to meet its function of managing revenue. Indeed it is the only way for it to achieve this function, insofar as this relates to recovering fraud losses.

We note that the United States Government has published data on all employers supported via the equivalent US support scheme, the Paycheck Protection Programme, since July 2020​. Published US data includes the name of the employer, the amount granted, their zip code and business type, plus self-reported data on the number of jobs supported.

Although the US makes this data public, it excludes grants that have been cancelled (i.e. repaid by the employer). We suggest that UK employers could be offered a similar amnesty before publication of the full dataset. If the privacy of sole traders is a concern, then such employers or payments below a ​de minimis​ could also be excluded.

We recommend HMRC should publish the names of employers that were supported under the furlough scheme between April and November 2020, to enable whistleblowing, reduce enforcement costs, and reclaim money for the taxpayer.

1. As in the US, employers could be offered the option to repay grants and be excluded from published data. Payments to micro-employers or below a ​de minimis could also be excluded.

2. When future support schemes are designed, HMRC should more carefully weigh up the benefits of data publication. The NAO has recommended that HMRC ​“should do more while employment support schemes are running to protect employees and counter acts of fraud”:​ we suggest that its processes should be updated to consider formally the publication of data, especially when this is the most straightforward method to prevent fraud. This would also reduce delays and administrative burdens from upfront fraud checks for legitimate businesses.

We are grateful to Francis Davey and the Good Law Project for advice. Any errors are ours.