51茶馆

Introduction

Finding and reducing the impact of fraud is an essential part of government efficiency. In order for public bodies to understand and tackle public sector fraud, they must be able to find it.

This report sets out levels of detected, prevented and recovered fraud and error, as reported by departments and public bodies to the Government Counter Fraud Function鈥檚 Centre of Expertise ^{[footnote 1]}. Data in the report corresponds to the 2021-22 financial year, and includes fraud and error relating to business-as-usual activity ^{[footnote 2]} and COVID-19 support schemes.

We know that the public sector is a target for those who commit fraud, and we cannot mitigate every risk. However, government is taking action to ensure that as much taxpayers鈥� money as possible goes to delivering vital public services, rather than to criminals.

The government has a long-standing commitment to find more fraud in the public sector. We recognise that fraud is inherently a hidden crime and to fight it you have to find it. As such, the government sees an increase in detected fraud and error as a sign of progression. Detecting fraud requires a proactive effort to look for that which has been hidden.

The continued increase in detected fraud and error is an indication that good work is being undertaken to understand and identify the problem. The better the public sector detects fraud and error, the better it is able to both find and recover losses, and build stronger controls to prevent such losses in the future.

The report also provides an update on the 鈥淪cale of the Challenge鈥�. This is the estimate for the overall level of fraud and error in central government ^{[footnote 3]} for 2021-22. This report also provides an update on the estimate of fraud and error loss in the COVID-19 support schemes as reported by government departments, which also refer to reporting on the 2020-21 financial year.

Since the start of the reporting period covered in this report, the government has invested over 拢1bn in taking action on fraud and error across government. Most of the investment has been into the Department for Work and Pensions (DWP) to increase action on the level of fraud and error in the benefits system.

The government also created the new Public Sector Fraud Authority (PSFA) to modernise the counter fraud activity in the public sector - taking into account the experiences of the pandemic.

The PSFA was launched in August 2022, and works to better understand and reduce the impact of fraud across the public sector. Its mandate, detailing what it will do and how it will work with its partners, was published in September 2022.

The PSFA, along with government departments, recognises that there is more to do to modernise the government鈥檚 approach to countering fraud. The PSFA is working with departments and public bodies to increase the ambition for tackling fraud, and to increase central government鈥檚 capability to take action. As a result of this work, 100% of the departments prioritised by the PSFA have now created counter fraud financial targets. We provide further detail on this in the final section of the report.

The PSFA was set a target to deliver 拢180m in audited benefits in 2022-2023. Working closely with departments and public bodies, the PSFA was able to exceed this target and delivered a total of 拢311m in audited counter fraud benefits.

Scope of the Data

The detected, prevented and recovered data in the report excludes fraud and error within welfare (including any impact of COVID-19 on fraud and error levels), tax and COVID-19 support schemes administered by HM Revenue & Customs (HMRC) and the Department for Work and Pensions (DWP). These figures are reported separately by HMRC and DWP ^{[footnote 4]}.

It should be noted that in respect of recoveries, there is not always a direct alignment between the detection of fraudulent activity and the recovery of money. Recovery takes time and may be reported across multiple years.

In addition to the reported data, we have provided an update on the estimate of fraud and error within COVID-19 support schemes. This is based on post-event assurance ^{[footnote 5]} activity undertaken by departments. In contrast to the data on detected, recovered and prevented fraud and error, this data is provided over the two financial years, 2020-21 and 2021-22.

As part of this work, we considered measurement, where available, of COVID-19 support schemes operated by HMRC and DWP.

Facts & Figures

• 52% increase in detected fraud and error in 2021-22, compared to 2020-21
• 拢823m in detected fraud and error in 2021-2022, up from 拢541m in 2020-2021
• 45% COVID-related of detected fraud, 45% related to COVID-19, and 55% related to departments鈥� business-as-usual work, increased 389%
• Detected fraud associated with COVID-19 increased 389% from 拢54m to 拢264m between 2020-21 and 2021-22
• 拢190m fraud and error recovered in 2021-2022
• down 9% from 拢209m in 2020-2021

NB All figures from prior to the establishment of the PSFA. All figures contain outliers - exceptional instances that we do not expect to occur again in the future.

Detected, prevented and recovered fraud

The data presented here sets out the detected, prevented and recovered fraud and error that departments have reported to the Government Counter Fraud Function鈥檚 Centre of Expertise.

HMRC and DWP separately publish data on fraud and error relating to the tax and welfare systems, for example tax evasion and benefit fraud. Therefore, this data on detected, prevented and recovered fraud and error only includes HMRC and DWP鈥檚 internal fraud and error - that is, where it has been carried out by a government employee or contractor. It also does not include data from HMRC鈥檚 COVID-19 support schemes.

Government departments and public bodies report on both fraud and error. In this context, the difference between fraud and error is the intent of the individual or organisation involved in the incorrect payment. This intent is established on the balance of probabilities, in line with the government鈥檚 definition for reporting fraud. As such, there are elements of both subjectivity and uncertainty in the reporting, as it will be dependent on the extent to which a public body has evidence of intention, and their assessment of this evidence.

In 2021-22, reported levels of detected fraud and error increased by 52% from 2020-21, rising from 拢541m to 拢823m. This follows a general upward trend since reporting began in 2014-2015 (see fig. 1).

In 2021-22, fraud made up 71% of the total (including outliers ^{[footnote 6]}), with 29% classified as error. This is in contrast with the previous financial year, where fraud made up 45% of the total detected figure.

COVID-19 Reporting

From the start of the COVID-19 pandemic in January 2020, the government ran initiatives to support individuals and businesses. It created 393 schemes, and as of July 2023, ^{[footnote 7]} 拢358bn of public funding had been provided ^{[footnote 8]}. During this period, the government accepted that departments and public bodies would be operating in a riskier context than in non-emergency times, and that the level of fraud and error would be higher.

Fraud and error reported against COVID-19 consists of fraud and error within specific schemes that were introduced to support the public and businesses through the pandemic. As noted above, this excludes HMRC-administered schemes, which are reported separately.

Detected fraud and error relating to COVID-19 made up 43% (拢356m) of the total reported in 2021-22 (拢823m). This is an increase from 2020-21, both in proportion and value, when fraud and error relating to COVID-19 made up 38% (拢205m) of the total reported level ^{[footnote 9]}.

When the reported data from 2021-22 is combined with that reported in 2020-21, ^{[footnote 10]} the total detected fraud and error relating to non-HMRC COVID-19 schemes across the two years is 拢561m.

Government departments continue to work to find and recover fraud and error in COVID-19 schemes. The PSFA will continue to update the data on detected fraud and error in COVID-19 spending in future Fraud Landscape Reports.

Looking only at reported detected fraud in COVID-19 schemes, there was a near four-fold (389%) increase between 2020-21 and 2021-22, from 拢54m to 拢264m.

拢235m (89%) of the detected fraud in COVID-19 schemes for 2021-22 was reported by the then Department for Business, Energy and Industrial Strategy ^{[footnote 11]} (BEIS). A further 拢28m (11%) was reported by the Department for Health and Social Care.

Fig. 1 Detected Fraud, Detected Error and Recoveries reported by government departments since 2014-15 ^{[footnote 12]}

Recovered Fraud and Error

Departments are required to report the amount of fraud and error they recover each year.

In 2021-22, government recovered 拢190m worth of fraud and error. Out of the 16 contributing departmental groups, the Department for Education (拢47m), Department for Transport (拢49m) and Department for Health and Social Care (拢41m) contributed significant proportions of the recovery figure.

2021-22 recovery figures were down 9% from 2020-21, when departments reported recovering 拢209m overall. While this is an increase on pre-COVID-19 recovery levels, as a proportion of detected fraud and error, the recovery rate dropped from 39% in 2020-21 to 23% in 2021-22. However, it should be recognised that there can be a lag in reporting recovery of detected fraud and error, due to the timescales involved in recovering debt.

COVID-19 Reporting

Recoveries relating to COVID-19, excluding HMRC-administered schemes, decreased from 拢88m in 2020-21 to 拢29m in 2021-22.

Of the 拢88m fraud and error recovered in 2020-21, 拢69m of this can be attributed to error within COVID-19 support schemes, recovered by the then Department for Business, Energy and Industrial Strategy.

It is a government priority to recover money which has been lost to fraud and error. The government is committed to recovering fraud and error losses from COVID-19 schemes where it can.

Prevented Fraud and Error

In 2021-22, government prevented 拢334m in fraud and error, including both business-as-usual and COVID-19 reporting. This is a 60% decrease on 拢827m in 2020-21. Business-as-usual prevented fraud and error fell 43% from 拢559m in 2020-21 to 拢319m in 2021-22.

When considering the previous year鈥檚 data (2020-21), 拢174m of the 拢827m was considered as arising from outlier instances. These are exceptional reporting occurrences that we do not expect to occur again in the future, and therefore partly explain the significant drop between years.

COVID-19 Reporting

Reported prevented fraud in 2021-22 relating to COVID-19 schemes dropped 94% from 拢268m to 拢15m. This is in line with reduced expenditure on COVID-19 schemes in 2021-22, meaning there were fewer opportunities to prevent fraud than in 2020-21.

In 2021-22, there was a 66% decrease in government spending on COVID-19 related support measures, ^{[footnote 13]} compared with 2020-21. The Office for Budget Responsibility estimates that the government spent 拢229bn on COVID-19 support measures in 2020-21, and 拢78bn in 2021-22.

Estimates of fraud and error in COVID-19 support schemes

In response to the COVID-19 pandemic, the government launched support schemes for individuals and businesses. It created 393 schemes and, as of July 2023, their spending totalled 拢358bn.

The Coronavirus Job Retention Scheme (CJRS) was set up to support employers in retaining their employees through the COVID-19 pandemic, and the Self-Employment Income Support Scheme (SEISS) supported self-employed individuals (including members of partnerships) whose businesses were adversely affected by COVID-19. These employment support schemes achieved their primary aim of protecting jobs and businesses, with unemployment peaking at just 5.2%.

The Bounce Back Loan Scheme was introduced in May 2020 and provided over 拢46bn to 1.5 million businesses to support them with the finance they urgently needed to get through the COVID-19 pandemic.

In the creation of these schemes, and in response to the pandemic, the government accepted that there would be an increased risk of fraud and error.

While accepting the increased risk, departments were expected to understand that risk and to take action to prevent loss where possible. Fraud measurement was undertaken in the areas that departments identified as the highest risk. Departments were also expected to find and recover detected fraud and error, where it showed value for money.

HM Treasury required departments to make an evidence-based estimate in their 2021-22 Annual Reports and Accounts of the level of fraud and error in the COVID-19 support schemes that they administered.

The estimates published in Annual Reports and Accounts ^{[footnote 14]}, and separate fraud and error publications ^{[footnote 15]}, show a total estimated level of fraud and error in COVID-19 schemes of 拢10.34bn. This data has already been published separately by departments, but this is the first time we have published an aggregate figure.

Published departmental fraud and error estimates cover 拢171bn of the estimated 拢358bn of COVID-19 spending.

In wider post-event assurance work on COVID-19 support schemes, where departments sought to measure fraud and error levels, they reported estimates totalling a further 拢144m to the PSFA ^{[footnote 16]}. This additional reporting related to 拢6bn of spending, with measurement therefore carried out across 拢177bn (49%) of the estimated 拢358bn of COVID-19 spending.

The schemes which did not participate in measurement are likely to be lower risk and therefore experience lower levels of fraud and error. This aligns with the Global Fraud Risk Assessment, which the Counter Fraud Centre of Expertise established to build a deeper understanding of the fraud landscape across government during the pandemic.

The PSFA estimates that the level of fraud and error in government expenditure and income where no measurement has been conducted is between 0.5% and 5% ^{[footnote 17]}. However, this relates to non-emergency spending. Schemes delivered during emergency situations - such as the COVID-19 pandemic - contain an 鈥渋nherently high risk of fraud鈥� ^{[footnote 18]}. As such, it follows that this range may be exceeded in COVID-19 support schemes.

The 拢10.48bn estimated level of fraud and error in COVID-19 schemes equates to 5.9% of the 拢177bn spend covered by measurement exercises (published and unpublished). However, this does not all equate to a loss for the taxpayer. Significant work continues to go into recovering payments that have been made to the incorrect recipients, either through fraud or error.

As of the end of September 2023, HMRC had recovered 拢1.2bn ^{[footnote 19]} in fraud and error on their COVID-19 schemes. It also blocked over 67,000 claims, worth a total of 拢430m, via pre-payment checks. HMRC remains committed to tackling error and fraud in the COVID-19 support schemes where this is the most cost-effective use of resources.

In its 2022-23 Annual Report and Accounts, the Department for Business, Energy and Industrial Strategy forecast that of the 8.9% fraud level (拢4.2bn) in the Bounce Back Loan Scheme, only 5.9% of loans with an estimated value of 拢1.7bn will go unrecovered. ^{[footnote 20]}

2021-2022 Scale of the challenge

Total estimated fraud and error (outside of COVID-19 schemes) is between 拢39.8bn and 拢58.5bn.

Fraud against public bodies takes money away from the vital public services that citizens rely on and can damage trust in government. The Public Sector Fraud Authority estimates that, in 2021-22, the level of fraud and error was between 拢39.8bn and 拢58.5bn ^{[footnote 21]}, outside of COVID-19 specific schemes. This calculation includes tax and welfare and is based on established and published estimates where fraud and error measurement is undertaken to a high standard, along with an estimated range of 0.5%-5% ^{[footnote 22]} on the areas where high-quality measurement work has not been carried out.

In 2020-21, the PSFA estimated that the level of fraud and error per year was between 拢33.2bn and 拢58.8bn. The 鈥榝loor鈥� of the estimate has risen slightly for 2021-22. This is due to higher fraud and error estimates in welfare and tax in 2021-22 compared to those in 2020-21. The 鈥榗eiling鈥� has remained broadly flat. The monetary value of fraud and error in the benefits system subsequently reduced in 2022-23.

Our fraud and error estimate relates to between 2.6% and 3.8% of overall government expenditure and income ^{[footnote 23]}.

Counter fraud colleagues across government play a key role in finding, preventing and recovering this fraud, continuously striving to improve the methods used to reduce the impact of fraud against the public sector.

Whilst 2021-22 showed that progress has been made across government in relation to detecting fraud, there is still work required to better protect public funds.

Public Sector Fraud Authority

In the Spring Statement 2022, the Chancellor announced the creation of a new authority to counter fraud against the public sector.

The Public Sector Fraud Authority (PSFA) was launched on 3 August 2022 as an integrated partnership between the Cabinet Office (CO) and HM Treasury (HMT). The PSFA engages with government departments and public bodies to understand and reduce the impact of fraud.

Since its launch, the PSFA has worked hard to modernise how the public sector counters fraud, based on experiences and lessons learned in both the UK and internationally. This will help the government to better understand, prevent and take action on fraud - reducing the cost of public services and protecting the public sector.

Below we summarise just some of the PSFA鈥檚 achievements since launch. In Section 7, we provide further detail on the PSFA鈥檚 work with departments to set ambitious counter fraud targets.

拢311m in audited counter fraud benefits delivered

The PSFA was set a target to deliver 拢180m in audited benefits in 2022-2023. Working closely with departments and public bodies, the PSFA was able to exceed this target and delivered a total of 拢311m in audited counter fraud benefits.

The 拢311m in audited counter fraud benefits represent prevented and recovered fraud and error as a result and is made up of a variety of data analytics projects, which include:

• the National Fraud Initiative, which is an exercise that works with more than 1,100 public bodies to compare sets of data against other records to identify discrepancies that may be evidence of fraud;
• the delivery of the COVID-19 Loan Schemes Fraud Analytics Programme, providing counter fraud flags to identify instances of potential fraud. These were then shared with accredited lenders to help them recover fraudulently obtained money.

Building a data platform with Quantexa

After a competitive tender, UK technology firm Quantexa was selected as the PSFA鈥檚 lead data analytics partner to deliver the Single Network Analytics Platform (SNAP).

This new platform uses artificial intelligence to connect millions of data points and provides a holistic view of risk involving UK-registered companies.

As part of this work, the PSFA has built several detection models to identify shell companies, suspect formations and sanctioned entities.

The SNAP makes it easier for fraud teams to look closely at UK-registered companies and to prevent public contracts, loans and grants from falling into criminal hands.

Government Counter Fraud Profession

The Government Counter Fraud Profession ^{[footnote 24]} (GCFP) was launched in October 2018 to develop a common structure for developing counter fraud capability across government

A key strategic aim for the Profession is to build deep counter fraud capability. As well as publishing standards and guidance, the Profession develops, delivers and signposts to, counter fraud training. The Profession aims to double the number of officers trained in Fraud Risk Assessment by March 2024 from 100, to 200.

In 2022-23, the Profession launched the bespoke Counter Fraud Leadership Programme, bringing together the first ever qualification for fraud leaders at Senior Civil Servant level. Having seen the first cohort of counter fraud leaders graduate, the Profession is already working to welcome the next cohort of senior leaders in January 2024, as well as year upon year.

The Centre of Learning for the Profession also recently collaborated with fraud subject matter experts and learning experts to reinvigorate the Fraud, Bribery & Corruption Awareness course, developed in partnership with Civil Service Learning. This all-staff course increases fraud awareness across the public sector and reminds participants that it is everyone鈥檚 responsibility to protect taxpayers鈥� money.

Engagement with international partners

The PSFA leads the International Public Sector Fraud Forum (IPSFF) of counter-fraud experts from the Five Eyes countries 鈥� America, Australia, Canada, New Zealand and the United Kingdom. The IPSFF has recently developed guidance on and .

In September 2023, the PSFA hosted 14 international delegates for a public sector fraud summit in London. The summit featured in-person strategic and technical discussions and workshops dedicated to improving our collective understanding of fraud and to share insights and best practice.

Target setting

As part of its work to modernise counter fraud activity, the government asked departments to set targets for their counter fraud work based on the financial impact they have. This forms part of the government鈥檚 wider drive to increase productivity and specific intent to increase its ambition in finding and dealing with fraud against the public sector.

Eighteen departments were prioritised into a tier-based system, based on their size and perceived level of risk, with a focus on ensuring departments in the top two tiers (15 departments - see fig. 2 below) set credible targets.

Tier 3 organisations (Cabinet Office, HMT, UK Export Finance) were seen as the smallest and had the lowest level of perceived risk of fraud and error. As such, these were not formally part of the target setting process. However, the Cabinet Office took the decision to still develop a target. The PSFA monitors these departments鈥� detected, prevented and recovered fraud and error through standard reporting.

A financial target for impact on fraud comprises prevented and recovered fraud and error. It was expected that targets would both show a positive return on investment and an uplift on previous performance. In cases where this was not possible, a clear rationale was expected.

In 2022-23, only two departments had set a financial impact target for their counter fraud work. For the 2023-24 financial year, 100% of Tier 1 and 2 departments created financial targets supported by the PSFA.

HMRC agrees its targets for activity on fraud and error in the tax system with HM Treasury Ministers.

DWP set a target to achieve savings in 2023-24 of at least 拢1.3bn from dedicated counter fraud and error resource. The target represents an increase on the savings achieved by these activities in 2022-23.

The other departments worked on targets directly with the PSFA. If met, the performance from the agreed targets would result in a 64% increase in counter fraud outcomes - a marked shift in ambition and committed outcomes across government.

The target setting process is developing. The government has clearly stated that its ambition is to continue to increase its impact from counter fraud activity.

Fig. 2 Departments broken down into tiers

Tier 1

• Department for Business and Trade
• Department for Education
• Department for Energy Security and Net Zero
• Department for Environment, Food and Rural Affairs
• Department for Health and Social Care
• Department for Transport
• Department for Work and Pensions*
• HM Revenue & Customs*
• Ministry of Defence
• Ministry of Justice

Tier 2

• Department for Culture, Media and Sport
• Department for Levelling Up, Housing and Communities
• Department for Science, Innovation and Technology
• Foreign, Commonwealth and Development Office
• Home Office

Tier 3

• Cabinet Office
• HM Treasury
• UK Export Finance

*HMRC and DWP are mature in their approach to counter fraud. HMRC agrees its targets for activity on fraud and error in the tax system with HM Treasury Ministers. DWP agreed its target with PSFA and HM Treasury.

Annex A: data tables

The tables below show the total amount of detected fraud and error reported by government departments and their arm鈥檚 length bodies in 2020-21 and 2021-22, both including and excluding COVID-19. This was reported to the Government Counter Fraud Function鈥檚 Centre of Expertise, prior to the establishment of the Public Sector Fraud Authority.

The reporting for DWP and HMRC is internal fraud and error only - that carried out by a government employee or contractor - and as such excludes welfare benefits, tax and HMRC-administered COVID-19 schemes.

Fig. 3 2021-22 Detected fraud and error by department, including outliers ^{[footnote 25]}

Department	Total Detected Fraud and Error 20-21	Total Detected Fraud and Error 21-22
Department for Business, Energy & Industrial Strategy	拢215.9尘	拢252.1尘
Ministry of Defence	拢148.9尘	拢225.2尘
Department for Health & Social Care	拢58.5尘	拢175.3尘
Department for Education	拢63.5尘	拢69.2尘
Department for Transport	拢12.3尘	拢53.7尘
Department for Digital, Culture, Media & Sport	拢0.9尘	拢12.4尘
Home Office	拢17.1尘	拢11.3尘
Department for Environment Food & Rural Affairs	拢16.3尘	拢11.2尘
Foreign, Commonwealth and Development Office	拢4.9尘	拢4.4尘
Department for Levelling Up, Housing and Communities	拢0.2尘	拢4.3尘
HM Revenue & Customs*	拢1.7尘	拢1.5尘
Ministry of Justice	拢0.5尘	拢0.6尘
HM Treasury	拢0.2尘	拢0.5尘
Department for Work & Pensions*	拢0.3尘	拢0.4尘
Cabinet Office	拢0.0尘	拢0.3尘
Department for International Trade	拢0.0尘	拢0.0尘
Total [footnote 26]	拢541.1尘	拢822.5尘

*Internal fraud only

Department	Total Detected Fraud and Error 20-21	Total Detected Fraud and Error 21-22
Ministry of Defence	拢148.9尘	拢225.2尘
Department for Health & Social Care	拢55.8尘	拢70.3尘
Department for Education	拢63.5尘	拢64.2尘
Department for Transport	拢12.3尘	拢53.0尘
Department for Business, Energy & Industrial Strategy	拢18.1尘	拢16.7尘
Home Office	拢17.1尘	拢11.3尘
Department for Environment Food & Rural Affairs	拢12.0尘	拢11.0尘
Foreign, Commonwealth and Development Office	拢4.9尘	拢4.4尘
Department for Levelling Up, Housing and Communities	拢0.2尘	拢4.3尘
Department for Digital, Culture, Media & Sport	拢0.9尘	拢3.3尘
HM Revenue & Customs*	拢1.7尘	拢1.5尘
Ministry of Justice	拢0.5尘	拢0.6尘
HM Treasury	拢0.2尘	拢0.5尘
Department for Work & Pensions*	拢0.3尘	拢0.4尘
Cabinet Office	拢0.0尘	拢0.3尘
Department for International Trade	拢0.0尘	拢0.0尘
Total [footnote 27]	拢336.2尘	拢467.0尘