Relief losses in 24 countries ran from 0.1% found to 21.5% estimated, counted four ways

United States

Paycheck Protection Program (PPP)

Lenders made the loans and the government stood behind all of them. SBA launched the program in April 2020 under the CARES Act: "fully guaranteed loans that are forgivable under certain circumstances" for small businesses, individuals and nonprofits, to cover payroll and benefits, rent, utilities and mortgage interest. Borrowers obtained nearly $800 billion through those "third-party lending partners". To move the money fast, the rules let SBA "reduce or eliminate key internal controls and rely on self-certification of eligibility."

In June 2023 SBA's Office of Inspector General put "potential fraud" in the program at $64 billion, 8 percent of the money disbursed, built from fraud indicators in the loan data such as hold codes, shared IP addresses and bank accounts. SBA disputed the method. Its own reviews found about $36 billion of loans "likely to be fraudulent" across PPP and the disaster-loan program together. SBA said 86% of that likely fraud originated in the programs' first nine months, before controls it added from January 2021: automated screening of PPP applications and checks against Treasury's Do Not Pay system. The Inspector General's figure counts loans that trip an indicator; SBA's counts loans its reviewers judged likely fraud.

Across PPP and the disaster loans, nearly $30 billion had been seized or returned to SBA by June 2023, most of it disaster-loan money.

Sources: COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape, Report 23-09 (SBA Office of Inspector General, 27 June 2023); Report 23-09 summary page (SBA Office of Inspector General); SBA, Response to COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape (U.S. Small Business Administration, 21 June 2023).

COVID-19 Economic Injury Disaster Loans (EIDL)

Here SBA lent directly. The agency was given authority to make low-interest, "fixed-rate, long-term" loans for working capital to meet "ordinary and necessary operating expenses", and paid emergency grants it called advances, then targeted and supplemental targeted advances. SBA disbursed over $400 billion through the program.

The Inspector General's June 2023 estimate of "potential fraud" in the program was more than $136 billion, 33 percent of the money disbursed: about four times the PPP's 8 percent, on loans SBA made itself with no lender in between. Its combined figure for both programs was over $200 billion, "at least 17 percent" of about $1.2 trillion. SBA's answer, about $36 billion of likely fraud across both programs, applies here too.

By June 2023 financial institutions had returned over $8 billion of EIDL money to SBA and borrowers another $20 billion.

Sources: COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape, Report 23-09 (SBA Office of Inspector General, 27 June 2023); Fraud Landscape one-pager (SBA Office of Inspector General, June 2023); SBA, Response to COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape (U.S. Small Business Administration, 21 June 2023).

Pandemic unemployment insurance (regular UI, PUA, FPUC and PEUC)

States paid the benefits; the pandemic laws widened who could claim. Pandemic Unemployment Assistance "extended unemployment benefits to self-employed, independent contractors, those with limited work history, and other individuals not traditionally eligible". More than $888 billion in federal and state benefits was paid for weeks in the pandemic period.

In February 2023 the Labor Department's Inspector General applied the department's latest improper payment rate, 21.52 percent, to that $888 billion. The result was "at least $191 billion" in payments that "could have been improper payments, with a significant portion attributable to fraud." The department applied that rate to two of the pandemic programs, PEUC and FPUC, but it "does not include the PUA program". The Inspector General applied it to PUA too, because PUA "had control weaknesses that may have facilitated comparable or greater improper payments." Improper payments count errors as well as fraud. The regular program's estimated rate had been above 10 percent for 15 of the previous 19 years.

By January 2023 the Inspector General's investigations had charged over 1,200 people with crimes related to UI fraud, producing more than 500 convictions and $905 million in investigative monetary results.

Sources: Testimony of Larry D. Turner, DOL Inspector General, House Committee on Ways and Means (U.S. Department of Labor Office of Inspector General, 8 February 2023).

Economic Impact Payments (stimulus checks)

Three laws, three rounds of checks. The CARES Act, the Consolidated Appropriations Act, 2021 and the American Rescue Plan Act of 2021 authorized Economic Impact Payments that went to around 165 million Americans. From April 2020 to December 2021 the government's direct payments to individuals totaled $931 billion, a figure that also includes the advance Child Tax Credit payments of July to December 2021. The checks were refundable tax credits with no earned-income requirement, so people with little or no income qualified.

The errors ran both ways. In September 2020 GAO put the improper payment rate for the checks at around 2%, and the Treasury Inspector General for Tax Administration later listed further steps IRS should take to recover them. At the other end, "some eligible Americans never received payments", especially nonfilers, people not required to file tax returns.

Sources: Stimulus Checks: Direct Payments to Individuals during the COVID-19 Pandemic (GAO-22-106044) (U.S. Government Accountability Office, June 2022).

Employee Retention Credit (ERC)

A credit against employment taxes, paid out as refunds, for employers who kept paying staff. Running it through employment taxes made it available to employers without tax liability. By June 2025 it had provided about $283 billion, and IRS had processed nearly 5 million claims, most filed on paper amended returns, under "complex and retroactive eligibility criteria".

The money arrived late. About 83 percent of refunds, about $235 billion, were issued from 2022 through June 2025, "well after unemployment had returned to its pre-pandemic level." Claims surged; GAO found that "some marketing companies convinced confused employers to file claims in order to receive a portion of the money," and IRS imposed a processing moratorium in September 2023. IRS never completed the improper payment estimate the law required; Treasury said it would not do so for short-term pandemic programs. A July 2025 law disallowed certain unpaid claims made after 31 January 2024, and IRS officials said most claims were closed by 31 December 2025. The deadline for assessing tax on certain improper paid credits has passed; fraud cases can still be pursued indefinitely.

Sources: COVID-19 Relief: IRS Can Use Lessons Learned to Address and Prevent Improper Payments in Future Tax Programs (GAO-26-107456) (U.S. Government Accountability Office, 10 February 2026).

Restaurant Revitalization Fund (RRF)

Demand ran at two and a half times the money. Beginning in May 2021 SBA received 278,300 applications asking for $72.2 billion, approved about 101,000 and disbursed all $28.6 billion the program was authorized to pay. SBA wrote a risk framework for the awards, but in March 2024 its Inspector General found that the plan "did not include all program requirements" and that SBA "did not always adhere to the plan".

The Inspector General questioned awards that missed eligibility checks and, counting each award once, found SBA had disbursed "nearly $6.7 billion to applicants without sufficiently verifying they were eligible", about 23% of the fund. Questioned costs are awards paid without the required verification, not a fraud count. SBA agreed or partially agreed with all five recommendations, which include reviewing and recovering awards that went to ineligible applicants or exceeded statutory limits. In 2023 SBA had estimated the combined fraud rate of this program and the venue grants "near half of one percent."

Sources: Report 24-09: SBA's Restaurant Revitalization Fund Program Award Practices (SBA Office of Inspector General, 26 March 2024); SBA, Response to COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape (U.S. Small Business Administration, 21 June 2023).

Shuttered Venue Operators Grant (SVOG)

Grants for businesses that "engage in venue operations". Congress created the program on 27 December 2020 in the Economic Aid Act and amended it in the American Rescue Plan Act on 11 March 2021, which added money and cut grants for some recipients who had also taken PPP loans. Together the laws gave it $16.25 billion, paid out in 13,011 awards.

SBA began post-award reviews in July 2022 and assigned 30 staff to review 2,162 awards; by 5 February 2024 it had completed 165. At that pace, the Inspector General estimated in July 2024, the remaining reviews would take 19 years, until February 2043. It also found SBA had picked 155 awards for eligibility reviews while another 1,818 high-risk awards, totaling $1.6 billion, warranted one. SBA partially agreed with two of the recommendations and disagreed with the third.

Sources: Report 24-21: Improvements Needed in SBA's Shuttered Venue Operators Grant Post-Award Review Process (SBA Office of Inspector General, 10 July 2024).

United Kingdom

Coronavirus Job Retention Scheme (furlough)

Furlough paid the wages of staff employers kept on the books. Employers could put workers on temporary leave, and the government paid "cash grants of up to 80% of employees' salaries, up to £2,500 a month", with the share later cut to 70% and 60% as employers were asked to contribute. CJRS furlough payments came to £68.9 billion, paid to 1.3 million employers for 11.7 million jobs: £5,887 per furloughed job. The National Audit Office found the schemes met their main aim of "preventing millions of job losses" and were "introduced at commendable speed".

In July 2023 HMRC put error and fraud in the scheme at £3,498 million, 5.1% of its £69,224 million net cost, within a range of £2,045 million to £5,418 million. The figure counts mistakes and fraud together. Across its three schemes HMRC had collected £1,056 million of what it calls COVID yield by then, and the Counter-Fraud Commissioner's December 2025 report counted over £1 billion in unprompted disclosures and voluntary repayments for CJRS alone. Details of 195 people who deliberately overclaimed CJRS or Eat Out to Help Out have been published.

Sources: Error and fraud in the COVID-19 schemes: methodology and approach (an update for 2023) (HM Revenue and Customs, updated 17 July 2023); Delivery of employment support schemes in response to the COVID-19 pandemic (National Audit Office, HC 656, 13 October 2022); Pursuing Recoveries and Preventing Reoccurrence: Final Report (Covid Counter-Fraud Commissioner, CP 1462, 9 December 2025).

Self-Employment Income Support Scheme (SEISS)

The self-employed got grants sized to their past profits. Grants paid up to 80% of average monthly trading profits for a three-month period, up to £7,500 a grant in the most generous rounds, worked out from tax returns. SEISS paid £28.1 billion to 2.9 million people, £9,700 each, against £5,887 per furloughed job; one reason, the NAO noted, is that "SEISS claimants could receive the grant while continuing working, while CJRS only covered hours not worked". By the fifth grant 1.7 million self-employed people were not eligible, including those with profits over £50,000, a group whose profits generally rose in the pandemic.

HMRC's July 2023 estimate of error and fraud in SEISS was £1,472 million, 5.2% of its £28,066 million net cost (range £1,189 million to £1,776 million). Unprompted disclosures and voluntary repayments for SEISS came to £51 million.

Sources: Delivery of employment support schemes in response to the COVID-19 pandemic (National Audit Office, HC 656, 13 October 2022); Error and fraud in the COVID-19 schemes: methodology and approach (an update for 2023) (HM Revenue and Customs, updated 17 July 2023); Pursuing Recoveries and Preventing Reoccurrence: Final Report (Covid Counter-Fraud Commissioner, CP 1462, 9 December 2025).

Bounce Back Loan Scheme (BBLS)

Banks lent the money and the state guaranteed all of it. The Chancellor announced the scheme on 27 April 2020 and the business department launched it on 4 May: loans of up to £50,000, capped at 25% of a firm's annual turnover, at a fixed 2.5% for up to ten years, with the government paying the first year's interest and guaranteeing 100% of capital and interest to the lender. To lend fast, the scheme dropped credit and affordability checks and let businesses self-certify their applications.

In December 2021 the National Audit Office set the department's "most likely" estimate of fraudulent loans, £4.9 billion, against the £32 million added for counter-fraud work and a target to recover at least £6 million from organized crime, and called both "inadequate". Its summary of the design: the government "prioritised payment speed over almost all other aspects of value for money."

By 31 December 2025 businesses had drawn £46.47 billion. Lenders had flagged £1.88 billion of it as suspected fraud, a count of cases reported rather than an estimate, which is why it sits below the £4.9 billion. The government had paid out its guarantee on 27.54% of the loans by number, 66.88% were repaid or on schedule, and the guarantee had been withdrawn from 14,107 loans worth £499.29 million.

Sources: The Bounce Back Loan Scheme: an update (National Audit Office, HC 861, 3 December 2021); COVID-19 loan guarantee schemes repayment data: December 2025 (Department for Business and Trade).

Coronavirus Business Interruption Loan Scheme (CBILS)

Unlike Bounce Back loans, these left lenders carrying a fifth of the risk. Over 100 lenders made CBILS loans, and the government guaranteed 80% of each; on a claim, "the 20% of the outstanding balance that was not guaranteed" was not paid by government. Businesses drew a total of £25.83 billion.

By 31 December 2025, 87.64% of CBILS facilities by volume were fully repaid (60.40%) or on schedule (27.23%). The Counter-Fraud Commissioner put 4% of CBILS lending by value in default, in arrears or claimed on the guarantee as of September 2025, against around 27% for Bounce Back loans. The guarantee had been removed from 622 CBILS loans worth £155.7 million. Lenders had flagged £0.06 billion of CBILS drawn value as suspected fraud, against £1.88 billion for Bounce Back loans.

Sources: COVID-19 loan guarantee schemes repayment data: December 2025 (Department for Business and Trade); Pursuing Recoveries and Preventing Reoccurrence: Final Report (Covid Counter-Fraud Commissioner, CP 1462, 9 December 2025).

Eat Out to Help Out (EOHO)

A discount "to encourage people to return to eating out in August 2020", paid through participating venues and run by HMRC. The Counter-Fraud Commissioner puts it at £0.8 billion, under 1% of the £98.1 billion the three schemes cost.

It had the highest error-and-fraud rate of the three. HMRC estimated £71 million, 8.5% of the scheme's £840 million net cost, in a range of 5.1% to 11.8%. Unprompted disclosures and repayments came to £2 million, and the NAO counted £18 million of compliance yield from the scheme. HMRC also used EOHO claims to test furlough claims: "a claim through EOHO by an employer could signpost an excessive claim for CJRS."

Sources: Error and fraud in the COVID-19 schemes: methodology and approach (an update for 2023) (HM Revenue and Customs, updated 17 July 2023); Delivery of employment support schemes in response to the COVID-19 pandemic (National Audit Office, HC 656, 13 October 2022); Pursuing Recoveries and Preventing Reoccurrence: Final Report (Covid Counter-Fraud Commissioner, CP 1462, 9 December 2025).

Brazil

Auxílio Emergencial (emergency aid), 2020

Cash for people with no formal income, paid on their own word. Eligibility turned on household income of up to half the minimum wage per person, a limit on members per family and an extra share for mothers heading a family, and applicants declared their own income and household. Automated cross-checks rejected about 57 million applications; about 68 million people received at least one payment, and R$330 billion was paid out. The Federal Court of Accounts (TCU) found the aid timely and said it reached workers without formal income.

It also found the self-declared limits were close to uncheckable. The TCU compared the eligible population implied by the national household survey with the people actually paid. It estimated that 7.4 million people outside the legal requirements received at least one payment and 6.4 million mothers received an extra share they were not due. Together that came to about R$45.7 billion, 14% of the R$330 billion, after payments that controls had already stopped. That is a survey-based estimate of payments to the wrong people, not a count of fraud cases. Requirements were checked only when the aid was granted, so people who later found jobs, drew another benefit or died stayed on the rolls. Later cross-checks against tax, defense-ministry, state and other databases found more improper payments; the TCU named the missing integrated data system as the main cause of those data gaps. Audit bodies had identified 4.9 million improper benefits, since cancelled, by July 2021, more than 7% of recipients, which the TCU said prevented R$12.3 billion in further payments.

Sources: Balanço da Fiscalização do Auxílio Emergencial (Tribunal de Contas da União, updated September 2022); TCU atualiza balanço da fiscalização do Auxílio Emergencial (Tribunal de Contas da União, 14 September 2022).

Benefício Emergencial (BEm), job-retention payments, 2020

Brazil's furlough paid the workers directly. Under the Emergency Employment and Income Maintenance Program, employees who signed agreements to suspend their contracts or cut hours and pay received the Benefício Emergencial from the federal government. From May to July 2020 it reached 12.4 million workers and cost R$18.2 billion, 39% of the authorized budget. Of the agreements processed, 54.4% suspended the contract and 44.4% cut hours and pay; the rest covered workers on intermittent contracts. The average monthly payment was R$863.

The TCU judged the program well managed but found problems with the publication of data and delays in building the appeals process. Cross-checking databases, it identified about 90,000 possible cases of improper receipt, worth more than R$150 million, under 1% of the R$18.2 billion paid in those three months. It told the managers to act on the payments and ordered the government to publish the list of beneficiaries by state and municipality (Acórdão 2.025/2020-TCU-Plenário).

Sources: Acompanhamento do Programa Emergencial para Manutenção do Emprego e da Renda (Tribunal de Contas da União, Acórdão 2.025/2020).

Canada

Canada Emergency Wage Subsidy (CEWS)

Ottawa paid employers up to 75% of qualifying wages, to a cap of C$847 per employee per week, to keep staff on the payroll. The Auditor General's 2021 audit counted 1,741,919 applications and quoted a Finance Canada estimate of about C$97.6 billion to the end of 2021–22; the Canada Revenue Agency later put the total paid at C$100 billion. The audit found that the agency "delivered the subsidy payments rapidly" but "chose not to establish tighter controls", and lacked the current earnings and tax data it needed to check applications efficiently.

The checking came after payment. By 31 March 2023 the agency had finished audits of C$5.53 billion of claims, chosen because they looked higher-risk, and reduced or denied C$325 million: 5.9% of the money audited and 0.325% of the C$100 billion paid. A further C$10.35 billion was under audit. The agency put most adjustments down to "unintentional errors and lack of supporting documentation" or to employers who did not reply, and said over 86% of penalties fell on files linked to certain third-party preparers it suspected of "willful and/or aggressive non-compliance". Large businesses had 2% of audited amounts reduced or denied; businesses with fewer than 25 employees, 35%, mostly as adjustments to the amount rather than findings of ineligibility.

The 5.9% comes from files picked as higher-risk, so it is not a program-wide rate; the 0.325% counts only audits finished by March 2023, so it is a floor.

Sources: Report 7 - Canada Emergency Wage Subsidy (Office of the Auditor General of Canada, 2021); Canada Emergency Wage Subsidy (CEWS) – Post-payment audits – Status Update – March 2023 (Canada Revenue Agency).

Canada Emergency Response Benefit (CERB)

A flat C$500 a week for workers who had lost income to the pandemic, first for 16 weeks and, after two extensions, for up to 28. Applicants attested that they had earned at least C$5,000 in the previous 12 months or in 2019, had not quit voluntarily and had stopped working because of COVID-19; they sent no documents, and the revenue agency paid in an average of 3 days. The Canada Revenue Agency and Employment and Social Development Canada (ESDC) both took applications. By 31 March 2022 the benefit, counted with its Employment Insurance twin, had paid C$74.8 billion to 8.5 million people.

The Auditor General's 2021 audit found that both offices "understood the risks of emphasizing quick delivery instead of their usual approaches to validating eligibility". Before a control stopped it, some people drew the benefit from both; the two put that at about C$500 million, less than 1% of payments. The 2022 audit tested recipients against the rules with data that arrived later. By its count, 562,540 people paid C$5.2 billion had earned less than C$5,000; 1,414,835 paid C$3.8 billion had earned more than C$1,000 in a four-week benefit period; 655,066 paid C$1.4 billion had received more than one benefit for the same period; and 1,845,073 had received a C$3.1 billion advance lump sum that counted as an overpayment. Those are payments that were ineligible or that the auditors said should be investigated, not proven fraud. The agency had flagged over 2.1 million recipients who might have earned more than C$1,000 in a benefit period; its April 2022 plan picked 150,000 of them, 7%, for verification. Across all the individual benefits the two offices reported recovering about C$2.3 billion.

Sources: Canada Emergency Response Benefit (Office of the Auditor General of Canada, Report 6, 2021); Specific COVID-19 Benefits (Office of the Auditor General of Canada, Report 10, December 2022).

Canada Emergency Business Account (CEBA)

Small businesses got interest-free loans of C$40,000 to C$60,000 through their banks, with C$10,000 to C$20,000 forgiven if the rest was repaid on time. Export Development Canada (EDC), a Crown corporation, ran the program. A business qualified with a 2019 payroll of more than C$20,000 and less than C$1.5 million or, from June 2020, with non-deferrable expenses such as rent, insurance and utilities. About C$49.1 billion went to about 898,000 businesses.

The Auditor General's 2024 audit found that about 91% of loan recipients met the eligibility rules and estimated that C$3.5 billion went to ineligible ones. Running the program had cost C$853 million by 31 March 2024, including C$575 million to financial institutions and C$248 million to EDC; of C$230 million in vendor contracts, 91% went to Accenture without competition. By the same date C$28.1 billion had been repaid and C$12.4 billion forgiven on loans repaid on time; C$0.1 billion had been written off and C$8.5 billion was still outstanding.

Sources: Canada Emergency Business Account Report 8 (Office of the Auditor General of Canada, 2024).

Canada Emergency Rent Subsidy (CERS) and Lockdown Support

The rent subsidy ran in 14 four-week claim periods, from 27 September 2020 to 23 October 2021, with a top-up called Lockdown Support. The Canada Revenue Agency's claims page, with data to 3 September 2023, counts 2,134,290 applications received, 2,080,200 approved and 223,540 applicants with approved claims, 94,030 of whom also drew Lockdown Support. Approved subsidies came to C$7.66 billion: C$6.55 billion of rent subsidy and C$1.11 billion of Lockdown Support, about C$3,680 per approved application. Most approved applications were small: 1,190,300 of them, 57%, were for less than C$2,000, and 141,350 were for more than C$10,000. Approved claims fell from 170,400 in the fifth period to 97,590 in the last.

The agency's page gives no audit or recovery figures.

Sources: Claims to date - Canada Emergency Rent Subsidy (CERS) and Lockdown Support (Canada Revenue Agency, data to 3 September 2023).

France

Activité partielle (short-time work)

France kept workers on payrolls at reduced hours through a scheme it already had, activité partielle, opened wide by the emergency law of March 2020. The State's share cost €21.98 billion over 2020–2022. Unédic, the unemployment-insurance fund, contributed about €10.8 billion more; the Cour des comptes, France's audit court, put the total at about €32.8 billion. Of the aid, 57% was requested in the first wave, from March to September 2020.

Up-front checks were few and refusals rare, the price of speed in the Cour's account. It found that fraud risks had not been properly identified in advance and that control of the scheme improved significantly only in July 2021, when public bodies began cross-matching their data. Checks aimed at organized fraud, such as chains of companies with no activity or stolen identities of real firms, blocked or rejected claims of more than €329 million from 2020 to April 2023. By April 2023 the labor ministry and the payments agency (ASP) had run at least one check on 10.3% of the establishments that used the scheme, and €217 million had been recovered or was being recovered. The Cour also found aid paid to firms whose turnover rose sharply over the same months; a voluntary repayment procedure run with Unédic from February 2021 brought back €6 million.

Sources: Garantir l'efficacité des aides de l'État aux entreprises pour faire face aux crises (Cour des comptes, July 2023).

Fonds de solidarité (solidarity fund)

Grants for firms hit by the closures, claimed through the tax website individuals already used. From the second wave in autumn 2020 the rules multiplied: eligibility widened to more sectors and to larger firms, the criteria grew more complex and payments slowed. By 31 December 2022 more than two million firms had received €41.03 billion, the largest single item in France's €82.2 billion budgetary cost of business support from 2020 to 2022.

The tax administration (DGFiP) ran automatic checks from the start: each firm's SIREN number had to match a bank account it had on file, to stop identity theft, and duplicate payments to the same firm for the same month were blocked. From July 2020 it added after-the-fact checks, targeting large sums, fraud networks, repeat applicants and dormant companies revived during the crisis. In 2021 and 2022 those checks identified about 187,670 irregular claims and led to recovery orders of €472.6 million. By the end of 2022, €73.2 million had been recovered and €161.8 million handed back voluntarily.

Sources: Garantir l'efficacité des aides de l'État aux entreprises pour faire face aux crises (Cour des comptes, July 2023).

Prêt garanti par l'État (PGE, state-guaranteed loans)

The State guaranteed bank loans that were cheaper than other financing, and the first repayment fell due a year after disbursement. Firms could then repay in full, repay over one to five years, or defer another year and repay over one to four, within a six-year limit. Loans totaled €143 billion. The Cour des comptes counts them in France's €249.7 billion of business support from 2020 to 2022, 10.9% of 2020 GDP, which cost the budget €82.2 billion.

A Banque de France study of 579,778 borrowers holding €105 billion, 73% of the total, found that 10% repaid in full a year after the loan and 73% chose the longest term. Size mattered: 26% of intermediate-sized firms repaid at once, against 10% of micro and very small firms, and small firms that repaid at once had held more cash going into the crisis. By the end of June 2024, 62% of the amount lent had been repaid and €4.3 billion of guarantees had been called, 3% of the total; the Banque de France judged that repayment "raises no particular concerns". The Cour des comptes credited the banks, which carried part of the risk, with picking borrowers in a way that avoided mass fraud of the kind seen in the United Kingdom and the United States.

Sources: State-guaranteed loans: how do firms choose to repay? (Banque de France, Eco Notepad, October 2024); Garantir l'efficacité des aides de l'État aux entreprises pour faire face aux crises (Cour des comptes, July 2023).

Denmark

Lønkompensation (wage compensation)

Denmark paid part of the wages of employees whom firms would otherwise have laid off. A firm qualified if it faced giving notice to at least 30% of its staff or to more than 50 employees. The State then paid 75% of salaried employees' pay and 90% of hourly workers' and apprentices' pay, up to DKK 30,000 a month per full-time employee. In the first phase, March to August 2020, 34,042 firms filed 66,631 applications and DKK 12.7 billion was paid, an average of DKK 394,500 per approved firm. Across all four phases, to October 2021, the scheme paid DKK 19.4 billion.

Parliament set aside about DKK 102 billion for temporary business compensation, DKK 96 billion of it for this scheme and the two below; by October 2021 the three had paid about DKK 48 billion. Rigsrevisionen, the national audit office, examined their first phase in November 2021. Automatic checks let low-risk applications through and sent riskier ones to manual review. Of the applications paid out at once without manual review, Rigsrevisionen found a risk that about 16% would have been rejected had they been checked by hand. In a sample of 88 manually reviewed applications, the ministry had established for 86% whether the firm met the criteria; it made processing errors on 5% and only partly established eligibility on 8%. The main control was a check after payment, which the ministry had postponed by about 11 months, to as much as a year and a half after the money went out. Follow-up on tips of possible fraud from the Money Laundering Secretariat was deferred without clear reason, although 76% of the tips were considered relevant. Parliament's public accounts committee warned that some firms may have received compensation they were not entitled to and that could not be recovered.

Sources: Beretning om Erhvervsministeriets forvaltning af hjælpepakker under COVID-19 (Rigsrevisionen, report 6/2021, November 2021).

Kompensation til selvstændige (self-employed)

Self-employed people and freelancers whose revenue or fee income fell at least 30% against 2019 received 90% of the drop, up to DKK 23,000 a month; a business could have at most 25 full-time employees. In the first phase 73,753 businesses filed 114,870 applications and DKK 5.9 billion was paid, an average of DKK 82,200 per approved business. All four phases paid DKK 11.0 billion.

Rigsrevisionen's case review found 5 approved applications, 6% of those it examined, where the 30% fall in revenue had only partly been established; each was approved on a low-detail extract from the applicant's accounting system, such as one revenue total for the period, or on a spreadsheet the firm had prepared itself. Its findings on the postponed check after payment, under wage compensation above, cover this scheme too.

Sources: Beretning om Erhvervsministeriets forvaltning af hjælpepakker under COVID-19 (Rigsrevisionen, report 6/2021, November 2021).

Kompensation for faste omkostninger (fixed costs)

Firms whose revenue fell at least 35% against 2019, or that were barred from opening, were compensated for documented fixed costs such as rent, interest and contractual expenses they could not cancel. In the first phase 27,761 firms filed 32,826 applications and DKK 7.3 billion was paid, an average of DKK 270,000 per approved firm. All four phases paid DKK 17.4 billion.

Rigsrevisionen found processing times satisfactory overall except here: 7% of firms that applied for fixed-cost compensation waited more than 60 days for a decision.

Sources: Beretning om Erhvervsministeriets forvaltning af hjælpepakker under COVID-19 (Rigsrevisionen, report 6/2021, November 2021).

Australia

JobKeeper Payment

From 30 March 2020 the government paid employers A$1,500 a fortnight for each eligible employee, money the employer had to pass on. A business qualified if it had, or expected to have, a fall in GST turnover of 30% (50% for businesses with turnover of more than A$1 billion). In the first phase, to 27 September 2020, the expected fall was enough; the extension that followed tapered the payment and required an actual decline. The first six months were costed at A$130 billion, later revised to A$70 billion. By 15 August 2021 net payments came to A$88.82 billion, A$69.97 billion in the original period and A$18.85 billion in the extension, and 1,068,856 entities had had applications processed. The independent evaluation published by Treasury in 2023 cites estimates that the scheme preserved between roughly 300,000 and 800,000 jobs.

The Australian Taxation Office estimated a net payment gap, the difference between what it paid and what fully compliant payments would have been, of 2.4%: 1.5 million payments worth A$2.2 billion. Treasury's review of the first six months found A$27 billion paid to businesses whose turnover did not fall by at least 30% (or 50%) on a year earlier, and at least A$4.9 billion paid to growing or restructuring businesses. By 10 September 2021 the tax office had identified A$470 million of overpayments and waived A$180 million of it under a legislative discretion. The Auditor-General's 2022 audit found that the office "did not implement all key compliance and integrity measures as intended": of 22 it tested, 8 were implemented fully as intended, 7 largely and 2 partly, and the audit could not conclude on 5.

Sources: Independent Evaluation of the JobKeeper Payment: Final Report (Treasury, 2023); Administration of the JobKeeper Scheme (Australian National Audit Office, Auditor-General Report No. 22 2021–22).

Coronavirus Supplement

An extra A$550 a fortnight on top of JobSeeker and other selected income-support payments from 27 April 2020, cut to A$250 from 25 September and to A$150 from 1 January 2021, until 31 March 2021. Eligibility for social security was widened at the same time, and the weekly number of JobSeeker recipients rose from 807,291 on 13 March 2020 to 1,649,794 by 15 May. From March to July 2020 Services Australia paid A$17.5 billion to eligible support recipients, and it spent over A$345 million of its own funds in 2019–20 to meet the extra workload.

To meet demand the agency took on large numbers of relatively inexperienced claims processing staff; the audit records that the resulting risk to payment accuracy and quality was documented as "accepted" by a deputy chief executive. Mutual obligations were suspended, and on 3 April 2020 the government paused debt raising and recovery to October 2020; the agency stopped issuing debt notices and recovering debts but kept raising debts in its system. As it wound the response down it planned remediation for the debt pause and changes to identity verification, and began matching its data against the tax office's.

Sources: Services Australia COVID-19 Measures and Enterprise Risk Management (Australian National Audit Office, Auditor-General Report No. 23 2020–21).

Cash Flow Boost

The Australian Taxation Office (ATO) estimates that $943 million of Cash Flow Boost credits, 2.7% of the total, went to taxpayers that did not meet the eligibility criteria. The boost was introduced in the 2019–20 financial year for small and medium businesses and not-for-profit organizations, and it was paid as credits in the activity statement system. Because the program was temporary, the ATO made a single estimate.

The ATO's table starts from $35,400 million of credits distributed. Taking out $97 million recovered, $891 million of revised credits and the $943 million it counts as ineligible leaves $33,469 million of eligible credits. The overview on the same page says over 97% of credits, "totalling $34.5 billion", were correctly distributed. Subtracting only the ineligible credits from the $35,400 million gives $34,457 million, which rounds to that figure. The table's gross gap, which adds the recovered credits back, is $1,040 million, or 2.9%.

The estimate is statistical. The ATO split pay-as-you-go (PAYG) withholding lodgers into 16 groups by lodgment profile and built a series of 22 quarters for each, from December 2015 to December 2020. It then predicted how many would lodge in January to March 2020, to measure the change in behavior the boost announcement caused. Eleven groups departed significantly from the prediction. In one of them the model expected 36,657 lodgers and 85,634 lodged; the ATO turns that difference of 48,977 into a net gap of $802 million. Across all eleven groups the net gap came to $1,834 million, and subtracting the $891 million of revised credits gives the $943 million.

The ATO lists techniques the analysis leaves out: restructuring to gain access to the boost, inflating reported PAYG withholding amounts to maximize it, resurrecting dormant entities or illegal phoenix activity, fraudulent attempts to create an entitlement, and identity crime. It says external experts assessed the estimate as "reliable, defensible and credible."

Sources: Cash flow boost payment gap (Australian Taxation Office, last modified 31 October 2022; Internet Archive copy of 5 February 2023).

South Korea

Employment Retention Subsidy (고용유지지원금)

A business in a temporary slump could keep its staff on paid leave or furlough instead of cutting jobs. The employer paid a leave allowance of 70% of average wages and the Employment Insurance Fund repaid up to 90% of that cost for up to 180 days. In 2020 the subsidy reached 72,000 firms and 773,000 workers and paid ₩2.2779 trillion, which the Ministry of Employment and Labor put at 34 times the 2019 figure. The Kangwon Ilbo put the 2020–2022 total at about ₩4 trillion.

Ministry figures given to National Assembly member Lee Ju-hwan were reported by the Kangwon Ilbo in June 2023. Fraudulent claims detected went from ₩800 million in 2019 to ₩9.3 billion in 2020 and ₩22.9 billion in 2021, then fell to ₩13.1 billion in 2022, with ₩3.0 billion more in January–May 2023. The three pandemic years come to ₩45.3 billion, about 1.1% of the ₩4 trillion paid.

In a six-month investigation from May to October 2022, 210 employment-insurance investigators looked for organised fraud across the fund's programmes. They found 269 people who had taken ₩2.57 billion, among them 38 employers and 5 brokers, and the ministry ordered ₩6.01 billion repaid including additional charges. Twenty-five employers, in separate cases, had filed papers saying they had put staff on the leave set out in their advance plans when they had not, and took ₩1.047 billion in retention subsidy. The ministry lists 11 of those employers for criminal punishment; across all programmes it sent 177 people to prosecutors with a recommendation to indict. The release reports referrals, not convictions. From November 2022 the ministry began a special inspection of 10,739 workplaces suspected of employment-subsidy fraud, centred on the retention subsidy, which it says surged with COVID-19.

Sources: 코로나19 위기 속에서 고용보험은 어떤 역할을 했을까 (“What role did employment insurance play in the COVID-19 crisis?”, Ministry of Employment and Labor on korea.kr, 19 August 2021); [사설]줄줄 새는 ‘고용유지지원금’, 부정 수급 근절해야 (“Leaking employment retention subsidy: stamp out fraudulent claims”, Kangwon Ilbo editorial, 20 June 2023); 2022년 고용보험 부정수급 기획조사 결과 (“Results of the 2022 planned investigation into employment-insurance fraud”, Ministry of Employment and Labor, 14 December 2022).

Emergency Disaster Relief Fund (긴급재난지원금)

On 22 September 2020 the Ministry of the Interior and Safety announced that the government had finished paying its emergency disaster relief money: ₩14.2357 trillion to 22.16 million households across the country. That is an average of about ₩642,000 a household.

Sources: 정부 긴급재난지원금, 전국 2,216만 가구에 14조 2,357억원 지급완료 (“Government emergency disaster relief payments completed: ₩14.2357 trillion to 22.16 million households nationwide”, Ministry of the Interior and Safety on korea.kr, 22 September 2020).

Emergency Employment Stabilization Subsidy (긴급고용안정지원금)

The subsidy was set up in 2020 for workers in special forms of employment and freelancers who fell outside employment insurance and other parts of the safety net; the ministry's examples are jazz pianists, fitness trainers and freelance broadcast producers. By 27 May 2021 four rounds had paid ₩3.4 trillion to 1,792,000 people, and someone paid from the first round onward received up to ₩3 million. The fourth round reached 715,000 people: 672,000 earlier recipients paid in March and 43,000 new applicants paid up to ₩1 million each on 26 May. Minister An Kyung-duk said the ministry would analyse recipients' incomes and use the results to extend employment insurance to workers in special forms of employment.

Sources: 특고·프리랜서 71만 5000명에 4차 긴급고용안정지원금 지급 완료 (“Fourth emergency employment stabilization subsidy paid to 715,000 special-form workers and freelancers”, Ministry of Employment and Labor on korea.kr, 27 May 2021).

Small-business disaster funds (소상공인 재난지원금)

From September 2020 to May 2022 the government paid small-business owners ₩52.9 trillion in seven rounds of disaster support: the New Hope Fund (새희망자금, September 2020), the Buttress Fund (버팀목자금, January and March 2021), the Hope Recovery Fund (희망회복자금, August 2021), quarantine support payments (방역지원금, December 2021 and February 2022) and the Loss Compensation payment (손실보전금, May 2022). They were grants, not loans, so the money comes back only where a payment is found to be ineligible, and the five-year limitation on recovery began to run out in September 2025.

The Ministry of SMEs and Startups had picked 11,547 payments, worth ₩25.4 billion, for recovery, mostly for errors in confirming the fall in sales or for unmet conditions such as the closure date. A Board of Audit and Inspection audit in July 2024 added 24,269 cases worth ₩35.2 billion, bringing the total to 35,816 cases and ₩60.6 billion: three times the cases, about 2.4 times the money, and about 0.1% of the ₩52.9 trillion paid. The auditors found fraud (forged quarantine-compliance certificates, ghost companies, businesses in excluded industries) and double payments, where an owner had also received the Emergency Employment Stabilization Subsidy or a grant for reopening after closing a shop. Double payments were the largest group, 15,490 cases; excluded industries were the largest by value, ₩19.83 billion. Some recoveries have been challenged in administrative lawsuits. The figures come from the ministry and the Small Enterprise and Market Service, obtained by Democratic Party lawmaker Kwak Sang-eon and reported by the Hankook Ilbo on 3 October 2025.

Sources: “사장님 코로나19 지원금 돌려주셔야”...환수 대상, 감사 거치자 세 배로 늘었다 (“Recovery targets for small-business COVID-19 grants tripled after audit”, Hankook Ilbo, 3 October 2025).

Italy

Cassa integrazione guadagni (short-time work)

Italy's wage-supplement funds pay workers whose hours are suspended or cut, and the pandemic made them free to use: a firm that gave COVID-19 as the reason paid nothing for access. In 2020 INPS, the national social security institute, paid €18.7 billion to 6.7 million workers, 57% of those covered, an average of €2,787 each. In 2019 it had paid €1.4 billion to 613,000. Contribution income fell over the same year, to €4.4 billion from €4.8 billion, and INPS noted that benefits on that scale could not be balanced against ordinary income. In 2021 payments came to €10.0 billion for 2.97 million workers. The hours authorised in 2020 and 2021, 7.15 billion, were more than the 6.2 billion of the six crisis years from 2009 to 2014. These figures cover the schemes INPS runs; artisans' and agency workers' schemes are separate. The INPS annual report gives no figure for improper payments.

Sources: XXI Rapporto Annuale — Conoscere il Paese per costruire il futuro (INPS, July 2022).

Fondo di garanzia per le PMI (state loan guarantees)

The state guarantee fund for small and medium-sized firms was widened by the Liquidity decree of April 2020: simpler access, higher cover and more eligible borrowers, including self-employed people and professionals. The ceiling per firm rose from €2.5 million to €5 million, and guarantees were granted without the fund's usual economic and financial assessment of the firm. After the May 2021 amendments, loans of up to €30,000 carried 90% cover for up to 15 years with no credit assessment. Such a loan could not exceed 25% of revenue or twice the annual wage bill.

From 17 March 2020 to 30 June 2022 the fund accepted 2,719,377 guarantee applications: more than €198.6 billion of guarantees on about €250.9 billion of loans. Of those, 1,184,751, or 43.6%, were for loans of up to €30,000, and 694,901 backed loan moratoria under the Cura Italia decree. Over 2020–2022 the fund granted more than €215.6 billion of guarantees, 78.9% of all it granted from 2016 to 2022, on about €271.7 billion of loans.

The report counts guarantees called, without amounts: 51,284 from 2016 to 2022, rising from 7,386 in 2020 and 8,934 in 2021 to 13,534 in 2022, an increase of 51.4% in a year. It says the cost of a public guarantee is hard to estimate until a default occurs.

Sources: Relazione sugli interventi di sostegno alle attività economiche e produttive (Ministry of Enterprises and Made in Italy, September 2023).

Indennità 600/1000 euro (self-employed and seasonal workers)

The Cura Italia decree of March 2020 gave a one-off €600 for March to five groups: professionals and collaborators in INPS's separate scheme, the self-employed in its special schemes, seasonal tourism and spa workers, farm workers and entertainment workers. Later decrees added months, raised payments to €1,000 and added categories, from on-call workers to door-to-door sellers and fishermen. Applications opened on 1 April 2020. The first five categories drew 5.1 million, almost 60% of them in the first four days, and 3.5% of applicants applied under two or more categories although the payments were mutually exclusive.

In all INPS received 6.7 million applications, almost 1.5 million of them duplicates or redundant, from just under 5.2 million people. It paid 4,228,000 of them (82%) and rejected 956,000 (18%). Refusal rates varied by category: 7% of the self-employed, 50% of seasonal workers, 73% of fixed-term tourism and spa workers. It made 8.8 million payments worth almost €6 billion, more than 70% of it to the self-employed and seasonal workers. The institute's statisticians trace many refusals to the rules themselves. An active employment contract was grounds for refusal. In the paper's textbook case it shut out, twice, a shopkeeper who also held an idle on-call contract in the theatre, until a later decree created a category for on-call workers. A mismatch between the rules for the separate scheme and the self-employed schemes was fixed by re-examining the refused applications centrally.

Sources: INPS paper on the COVID-19 €600/€1,000 indemnities (INPS, Coordinamento Generale Statistico Attuariale, 2021).

Contributo a fondo perduto (non-repayable grants): checks and recovery

Italy's non-repayable grants to businesses were claimed by application to the tax agency, the Agenzia delle entrate, which checks the data declared under the rules for auditing tax returns. It also matches the applications against electronic invoices, electronic till receipts, periodic VAT settlements and VAT returns. Whatever the size of the grant, it runs further checks against attempts at criminal infiltration under a protocol signed with the interior and economy ministries. The agency passes the data in the applications for grants paid to the Guardia di Finanza, the financial police. The agency's page gives no count of grants paid, checked or recovered.

A grant found wholly or partly not due, including after the anti-mafia checks, is recovered with a penalty of 100% to 200%, and the reduced settlement is not available. Article 316-ter of the criminal code, on wrongful receipt of public payments, provides six months to three years in prison or, for a grant under €4,000, an administrative fine of €5,164 to €25,822, capped at three times the grant. Whoever signed the anti-mafia self-certification faces two to six years, and a grant already paid can be confiscated under article 322-ter. A recipient may repay of its own accord, with interest and reduced penalties. Payment goes only by the F24 form, with no offsetting, under tax codes set by resolution no. 37 of 26 June 2020.

Sources: I controlli e l’eventuale restituzione (Agenzia delle entrate).

Netherlands

NOW (wage-cost subsidy)

The NOW paid employers an advance based on their wage bill and their own estimate of lost turnover, so that they could keep paying salaries. What they were actually owed was settled later, on application, once real turnover and the wage bill were known; the advance was 80% of the calculated subsidy. In the first period about 139,500 employers received advances totalling about €7.9 billion. Of the roughly 136,000 still in the register when UWV, the agency that ran it, published final amounts, 125,000 (92%) had been settled. For 90,000 of them (71%) the settlement meant repaying €2.3 billion between them; almost 36,000 (29%) were due more than €455 million in further payments. For 26%, more than 32,000 employers, the final subsidy was zero, in most cases because turnover had fallen by less than 20%, the threshold for the scheme. UWV says the gap between advance and final amount has nothing to do with abuse: many employers estimated high losses early in the pandemic and ended up with smaller ones.

Across all rounds employers had received about €23.9 billion in wage-cost support by 30 June 2022, according to Statistics Netherlands (CBS). The Algemene Rekenkamer, the national audit office, concluded in 2023 that the minister had carried out the NOW and the Tozo support for the self-employed well, with UWV and the municipalities, and had to keep watching how both were wound up.

Sources: UWV publiceert register definitieve tegemoetkomingen eerste periode NOW (UWV); Overzicht gebruik coronasteunmaatregelen (CBS, Twee jaar coronasteun, 2022); Resultaten verantwoordingsonderzoek 2022 Ministerie van Sociale Zaken en Werkgelegenheid (Algemene Rekenkamer, May 2023).

Bijzonder uitstel van belastingbetaling (special tax deferral)

Throughout the support period, to the end of March 2022, any business could ask the tax office to defer almost any tax it owed. Interest on late payment was cut to 0.01%, no default penalties were imposed, and penalties already imposed were reduced where a deferral had been granted. Deferring income tax deferred health-insurance contributions, corporate tax, payroll tax and VAT with it, and the other way round. In 2020 between 127,000 and 140,000 businesses had a deferral open at any one time; on 30 June 2022 almost 205,000 did, about 11% of all businesses, two-thirds of them sole proprietorships. About €47 billion of tax was deferred over the whole period. By mid-2022 €20.8 billion was still outstanding, so more than half had been paid; repayment of the rest began on 1 October 2022 and is due within five to seven years.

Sources: Twee jaar coronasteun (CBS, 2022); Overzicht gebruik coronasteunmaatregelen (CBS, Twee jaar coronasteun, 2022).

Tozo (income support for the self-employed)

The Tijdelijke overbruggingsregeling zelfstandig ondernemers (Tozo), the temporary bridging scheme for the self-employed, started with retroactive effect from 1 March 2020 and was paid by the municipalities. It offered a benefit toward living costs and a loan for working capital. The benefit was paid to households, so a partner counted as a recipient too. In 2020 municipalities paid the living-costs benefit to 501,000 people, 3.5% of adults in the Netherlands, according to final figures from Statistics Netherlands (CBS). Of these, 401,000 lived in households of more than one person. The working-capital loan went to 42,000 self-employed people, 35,000 of whom also received the benefit.

The 501,000 recipients had about 330,000 businesses between them, against 1.6 million self-employed in the country in 2020. Other services, such as hairdressers, beauty salons and repairers, made up 5.3% of the self-employed but 13.8% of recipients' businesses; culture, sport and recreation 4.2% against 10.7%. Rental and other business services, the largest sector with almost 17% of the self-employed, accounted for 4.5% of recipients' businesses. Men were 53.0% of recipients and women 47.0%; among all self-employed the split was almost 51 to just over 49. Nearly half, 242,000, were aged between 45 and the state pension age of 66 years and 4 months; 35,000 were under 27.

Of the four largest cities, Amsterdam had the highest share of adults on the benefit, 7.8%, and half of the big-city recipients lived there. Three small municipalities had higher rates: Vlieland 8.7%, Terschelling 8.3% and Ameland 8.2%.

Sources: Half miljoen mensen kregen Tozo-uitkering in 2020 — 11-11-2021 (Dutch) (CBS, 11 November 2021).

TVL (fixed-costs grant)

Across all rounds of the Tegemoetkoming Vaste Lasten (TVL), the fixed-costs grant, €7.5 billion was awarded and €128 million of it, 1.7%, was paid out wrongly because of established misuse and improper use. The figures come from the government's progress report to the Tweede Kamer (document 2022D04872). By the time of the report €12.3 million of that had come back, 9.6%. The report expected no more than 25% to be recovered in the end, because misuse often ran through money mules and the money was often gone within a few hours.

The agency RVO paid an advance of 80% of the grant and settled the rest later. At settlement 82% of businesses received a further payment and 15% had to repay the advance in whole (10%) or in part (5%). For another 3% the amount owed was below €500, which RVO does not collect. The TVL produced about 590,000 unique decisions, which the report puts at about twice what RVO handles in a normal year across its roughly 600 schemes. Of the applications, 96% were decided well within the legal deadline, and 48% of businesses had their advance within one week.

Every application went through automated checks, and 28% were also checked by hand before an advance was paid. About 9,000 of the manual checks raised serious suspicion of deliberate misuse, such as forged evidence or identity fraud, and went to a team of specialist fraud investigators. In about 23% of those cases no deliberate misuse was found and the advance was paid. Where a business did not repay after a demand, RVO had court bailiffs seize bank balances and assets. By the time of the report that had happened in 2,217 cases, covering €121 million of granted subsidy.

Businesses filed 17,000 objections against the 590,000 decisions, and 4,800 were past the legal deadline for a response. For TVL Q4 2021, which opened on 20 December 2021, the turnover-loss threshold was lowered from 30% to 20% and the state-aid ceiling raised from €1.8 million to €2.3 million. Businesses that missed even the 20% threshold, the report says, experienced it as a great injustice.

Sources: Voortgangsrapportage van de Tegemoetkoming Vaste Lasten (TVL), Tweede Kamer document 2022D04872 (Tweede Kamer).

Spain

Prestación extraordinaria por cese de actividad (PECATA, self-employed)

From 14 March to 30 June 2020, the end of the month in which the state of alarm ended, Spain paid a benefit to self-employed workers and exempted them from social contributions. To qualify, their activity had to have been suspended, or their turnover in the previous month had to have fallen by at least 75% against their average for the previous half-year. 1,486,924 self-employed workers received it. Net spending came to at least €3.74 billion on the benefit, plus €1.50 billion of contributions waived; the average benefit per person over the whole period was about €2,513. Most awards were made provisionally by the mutual insurers (mutuas) that ran the scheme, which were to review every one once the state of alarm ended. On 30 June 2023, three years later, the review had not begun.

The Tribunal de Cuentas, Spain's audit court, found the rules had been amended seven times in three years. The ministry relaxed the requirement to keep meeting the conditions while being paid by unsigned letter and email. Some mutuas read a letter of 29 April 2020 as dropping it and others did not; an email of 31 July 2020, after the benefit had ended, dropped it for those who had qualified on turnover. Some recipients gave the benefit up while others in the same position kept it. Later decrees let a fall of 7.5% in the average daily number of workers in the claimant's sector count as proof of the turnover drop; that presumption covered at least 91.64% (643,563) of the beneficiaries who had qualified on turnover. The court counted at least 125,292 beneficiaries of the mutuas whose benefits were paid wrongly or miscalculated, a net €119.5 million.

The largest items were €87.0 million paid to 35,764 people who were behind with their contributions and €24.4 million overpaid to 51,317 people whose benefit had been calculated on the wrong base, against €12.6 million underpaid to 20,699. At least 156,066 more people paid on the ground that their activity had been suspended were registered in activities on neither list of suspended activities. They received €386.4 million, and the court could not check whether they had in fact stopped work. Claims for wrongly paid social-security benefits lapse after four years. In May 2024 the parliamentary committee for relations with the Tribunal de Cuentas adopted the report and urged the ministry's social-security directorate to start the review.

Sources: Informe de fiscalización sobre la gestión y control de la prestación extraordinaria por cese de actividad de los trabajadores autónomos (Tribunal de Cuentas, published in the BOE, 12 August 2024); Resolución de 21 de mayo de 2024 de la Comisión Mixta para las Relaciones con el Tribunal de Cuentas (BOE-A-2024-16741).

ERTE (temporary lay-offs: unemployment benefit and contribution exemptions)

Spain kept workers attached to their jobs by letting employers suspend them or cut their hours under a temporary employment regulation file, an ERTE. The state paid the suspended workers unemployment benefit through the public employment service (SEPE) and exempted their employers from social-security contributions. The measures ran from 18 March 2020 to 31 March 2022. At least 558,350 firms used them and at least 4,157,854 workers drew the benefit. SEPE paid 24,042,861 monthly benefits at a cost of €18.27 billion, and employers were excused €7.96 billion of contributions. In May 2020 the number of ERTE benefits paid equaled 22.07% of all employees affiliated to social security.

The Tribunal de Cuentas approved its audit on 28 July 2025. The labor directorate (DGT) that handled national ERTE files had no computer system for them until November 2023. It could not say how many files it had processed or how it had decided them, and it issued contradictory and duplicate decisions. It did not know how many force-majeure files had been approved by default, because it could not decide them within the five-day deadline set by Royal Decree-law 8/2020.

SEPE granted benefits on the employer's word, sometimes before it had the labor authority's confirmation of the force majeure the employer claimed; the court says it put workers' social protection ahead of checking first that the legal requirements were met. Its later checks were not planned and its corrections were made by hand, so their results could not be documented or counted. The court found at least 10,896 recipients whose inactivity their employer had never reported to the social-security treasury (TGSS) and who were not on SEPE's list of debtors for undue benefits; they were paid €24,656,594. The DGT says it emailed its decisions to SEPE every day. A cyber-security incident at the labor ministry from 9 June to 9 July 2021 lost those email archives, and the DGT says they could not be recovered.

The treasury granted the contribution exemptions on employers' sworn statements and checked later. Its checks found signs of non-compliance that could affect 278,757 firms and 1,769,253 workers, which it passed to the labor inspectorate (ITSS). By the end of the audit the inspectorate had inspected 11,688 firms and regularized €75,949,433 of exemptions covering 55,558 workers. SEPE paid for most of the extra spending out of its own reserves, €13.53 billion of its unallocated cash surplus. The €21.32 billion Spain received from the EU's SURE instrument for unemployment risks had not been transferred to SEPE; the Treasury replied that its cash is fungible.

Of the 558,350 firms that used an ERTE, 154,833 (27.73%) had no workers registered with social security by May 2024. Firms that used one remained active at a rate of 72.27%, against 67.47% for the 788,489 firms that did not.

Sources: N.º 1.633 — Informe de Fiscalización Operativa Relativa a las Prestaciones por Desempleo (COVID-19) (Tribunal de Cuentas, approved 28 July 2025).

Líneas de Avales ICO COVID-19 (state loan guarantees)

Under the COVID-19 guarantee lines of the Instituto de Crédito Oficial (ICO), banks lent €140,736.7 million in 1,192,484 loans to 674,922 self-employed people and companies, and the State guaranteed €107,186.9 million of it. Those are the figures in ICO's monitoring report with data to 30 September 2023. By then ICO had paid banks €701 million under the guarantees, on 55,439 loans with unpaid instalments belonging to 38,781 borrowers. Recoveries of €32.7 million brought the net figure to €668.3 million. Over the same period the State had collected €1,925.79 million in guarantee fees.

The banks made the loans under their own credit policies. The guarantees belong to the State, through the economy ministry, and ICO manages them. Banks report unpaid instalments between 90 and 270 days old in quarterly windows, on the 1st to the 15th of January, April, July and October. ICO checks each loan against the eligibility conditions after the fact and pays within 60 days. It pays the guaranteed share of each unpaid instalment on the loan's original repayment schedule, not the whole guaranteed principal at once, even when a bank calls in the full loan. The bank must pursue the borrower for the whole debt and share recoveries with the State pari passu.

ICO's default ratio therefore counts the outstanding principal on every loan with an unpaid instalment, net of recoveries: €3,231.6 million, of which €2,540.3 million is the guaranteed part. Against total lending it gives 2.30%. By borrower, the ratio was 3.27% for the self-employed, 3.45% for micro-enterprises, 2.53% for SMEs and 0.80% for larger companies; by sector, it ranged from 1.15% in real estate services to 3.17% in energy. Tourism, leisure and culture, the sector with the most loans in arrears, was at 1.98% and construction at 3.11%.

Guarantees of €721.6 million on 1,451 loans were excluded or cancelled. In 88% of those loans nothing had been paid. On 178 of them, with €8.4 million guaranteed, ICO had already paid €1,923,777.67 of unpaid instalments; the report says that amount was refunded in full after the guarantees were cancelled.

Sources: Informe de seguimiento de importes abonados y recuperaciones ICO Avales Covid-19 (Instituto de Crédito Oficial, data to 30 September 2023).

Sweden

Omställningsstöd (fixed-cost support)

The Swedish Tax Agency (Skatteverket) paid omställningsstöd to help firms hit by the pandemic cover their fixed costs. Applications went through a new e-service that checked each one for completeness and against the firm's earlier tax returns; the agency says most errors were caught there and the money was never paid. By February 2021 it had paid SEK 5.8 billion to 27,000 firms and had stopped SEK 546 million before payment. Follow-up checks of 1,700 firms led to repayment claims of SEK 32 million, which the agency put at less than half a percent of the support paid. The commonest error was paying a dividend, which a firm claiming the support was not allowed to do; the next was overstating fixed costs or turnover for the same period a year earlier. Construction had the highest share reclaimed, 2.33% of the support it received. These are the agency's own figures, from a press release of 23 February 2021; the scheme was extended to further periods after that.

Sources: Återkrav på 32 miljoner av omställningsstödet (Skatteverket press release, 23 February 2021).

Företagsakuten (state loan guarantees)

In March 2020 the government told the Swedish National Debt Office (Riksgälden) to run a state guarantee program for new loans, mainly to small and medium-sized firms hit by the pandemic but otherwise viable. The state took most of the risk. Meant to run for three months, it was extended and closed to new guarantees on 30 September 2021. By then lenders had reported SEK 2.7 billion of loans, SEK 1.9 billion of it guaranteed, across 764 loans; Riksgälden's account gives no figure for losses or fraud. At the end of 2021 the government allowed loans already granted to be extended from three years to at most five.

Sources: Åtgärder under coronapandemin 2020 (Riksgälden).

Tillfälliga anstånd (tax deferral)

From 2020 firms could defer a year's employer contributions, tax withheld from employees' pay and VAT, and pay it later with interest and a deferral fee; small firms with low turnover could defer two years of VAT. No credit check like a bank's was required. Deferrals granted from 2020 to 2023 came to about SEK 111 billion, an estimate the Swedish National Audit Office (Riksrevisionen) grossed up from the SEK 105.7 billion granted to limited companies, and 63,472 taxpayers received one. The government put the net cost to public finances at SEK 2.6 billion, expecting interest and fees to roughly cover the state's borrowing cost and extra collection losses.

Riksrevisionen concluded in February 2025 that the net cost would very probably be much higher and that SEK 10 billion was a more reasonable estimate. The audit found that firms with both low profitability and low equity before the pandemic were clearly over-represented among those with deferrals. The audit found no support for the idea that deferrals had prevented bankruptcies up to September 2024, and signs that they had partly displaced bank lending. Skatteverket expects collection losses of SEK 20.2 billion (± 3.5 billion) among firms with deferrals over 2020 to 2028, SEK 8.6 billion of it already incurred by September 2024. The last installments fall due on 12 September 2027, or 17 January 2028 for deferrals based on annual VAT returns. The audit office called the measure relatively effective overall and recommended limiting future deferrals to periods when credit markets are not working normally.

Sources: Tillfälliga anstånd med inbetalning av skatt (Riksrevisionen, RiR 2025:1, February 2025).

Korttidsstöd (short-time work support)

By June 2021 Tillväxtverket, the agency that paid Sweden's short-time work support, had paid out more than SEK 35 billion. A page it published on 11 June 2021 says it had also stopped payments of more than SEK 1 billion, made about 8,000 recovery claims worth about SEK 1 billion and filed nearly 130 police reports. The police reports are the cases the agency treated as suspected crime.

The support was for companies hit by the pandemic, and a company had to show a negative break in the trend of its finances to qualify. Automated checks ran against data from Skatteverket, the tax agency, Bolagsverket, the companies registry, and other suppliers of company data: whether the company had paid employer contributions, had tax debts or a trading ban, or was insolvent. Applications that showed deviations went to manual handling, where staff contacted the company. A unit for in-depth control took cases with serious deviations or tips worth investigating, with help from about 50 Skatteverket employees, and staff on loan from the police helped prepare the police reports.

The agency's examples of grounds for a police report were sham furloughs, deliberately overstated wages, and missing agreements that the company had attested to in its application. Richard Källstrand, head of quality and control, said the agency has laws to follow and is paying out taxpayers' money, "så det måste bli rätt" (so it has to be right). It is also relatively common, the agency said, for companies to repay the support once it starts asking questions.

Sources: Tillväxtverkets arbete mot fel och bedrägerier (Tillväxtverket, 11 June 2021; Internet Archive copy of 7 December 2022).

New Zealand

COVID-19 Wage Subsidy Scheme

The Wage Subsidy paid employers up front so that they could keep paying staff through COVID-19 restrictions. It was a "high-trust" scheme: the Ministry of Social Development (MSD) approved applications on the employer's declaration that it was eligible, and the employer undertook to pass the money on in wages. Before paying, MSD checked each application against Inland Revenue's records, and tens of thousands were declined because the details did not match. It paid on average within three and a half days of an application, and on one day paid out nearly NZ$1.8 billion. About $18.8 billion went out in 2020 and 2021 for wages for more than 1.8 million jobs.

The Auditor-General reported in May 2021 that one condition, that an employer had "taken active steps to mitigate the impact of COVID-19", was open to interpretation and hard for MSD to verify. MSD's post-payment reviews consisted mainly of verbal confirmation from employers; MSD called them audits, and the Auditor-General said they were not. By 5 March 2021, $703 million had been paid back voluntarily and $23 million recovered compulsorily, and 1,017 cases had been referred for investigation. MSD reports that as of 30 June 2026 there had been 25,589 repayments totaling $837 million, 7,895 allegations of misuse had been resolved, 73 people had been sentenced and 36 were before the courts. MSD says many employers chose to repay subsidies they had been entitled to keep. It rejects the Integrity Institute's claim of $10 billion in overpayments as having no reasonable basis.

Sources: Wage Subsidy Integrity and Fraud Programme (Ministry of Social Development); Management of the Wage Subsidy Scheme (Office of the Auditor-General, B.29[21a], May 2021).

Small Business Cashflow (Loan) Scheme

From May 2020 Inland Revenue lent to small and medium businesses hit by the pandemic. More than 129,000 businesses were issued loans totaling NZ$2.4 billion; the average amount approved was $17,000. A high proportion of borrowers were sole traders, followed by businesses with 2 to 5 staff; most borrowed in May and June 2020 with five years to repay. By 31 December 2024, 51,021 borrowers had repaid in full and $953 million was still owed. About 10,000 loans, owing just over $161 million, were already in default. From June 2025 Inland Revenue was to put any loan not paid off at its cut-off into default and charge default interest: the use-of-money interest rate of 10.88% plus the standard rate of 3%.

Sources: Final repayment date nears for SBC loans (Inland Revenue, 17 March 2025).

Resurgence Support Payment and COVID-19 Support Payment

New Zealand paid businesses two revenue-drop grants besides the wage subsidy. The Resurgence Support Payment (RSP) went to viable and ongoing businesses or organizations that had lost 30% of their revenue over seven days because of higher COVID-19 alert levels. The Treasury says NZ$2.9 billion was paid to 231,000 eligible businesses as at 20 January 2022, or about NZ$12,600 per business on those figures. The COVID-19 Support Payment raised the bar to a drop of 40% or more. Three payments between February and April 2022 paid NZ$1.3 billion to 113,000 eligible businesses as at 30 June 2022. Divided the same way, that is about NZ$11,500 per business.

In the Treasury's table of spending from 2019/20 to 2021/22, the COVID-19 Resurgence Support Payment appropriation spent NZ$2,912 million and the COVID-19 Support Payment appropriation NZ$1,307 million. The Business Support Subsidy appropriation, which funded the wage subsidy, spent NZ$17,996 million. The Treasury page gives no figure for wrong or fraudulent payments under either scheme.

Sources: COVID-19 Response and Recovery - What has been achieved? (New Zealand Treasury, last updated 24 April 2025; Internet Archive copy of 5 February 2026).

South Africa

Social Relief of Distress grant (R350)

The R350 social relief grant was part of a R500 billion package for the health response and for social and economic relief, and the South African Social Security Agency (Sassa) paid it to people in distress. By 31 July 2020 Sassa had paid R19.6 billion in social grants; the Auditor-General's release does not split that between the R350 grant and the top-ups to existing grants. The first report of the Auditor-General's real-time audit of the package was released on 2 September 2020. It found that the application process did little to check whether applicants had other income, so that students older than 18, for example, could get the grant. Sassa checked applicants against some government databases. The auditors' own data analysis still flagged payments to over 30,000 beneficiaries for further investigation, among them people employed by the government or receiving other social grants, a government pension or UIF payments. The auditors also found Sassa's databases outdated, which could have led it to reject applicants who should have been paid. The release puts no amount on the flagged payments.

Sources: Auditor-General: Covid-19 Relief Package Landed in an Environment with Many Control Weaknesses (Auditor-General of South Africa, media release, 2 September 2020).

Temporary Employer/Employee Relief Scheme (TERS)

The Unemployment Insurance Fund (UIF) paid TERS benefits to help employers keep paying workers through lockdown; by 31 July 2020 it had paid just over R37 billion. The Auditor-General found that a new system had miscalculated benefits for the first lockdown period, 27 March to 30 April, by ignoring the actual period of inactivity and the part of the salary employers still paid, which caused significant overpayments. Checking payments against other government databases flagged a high number for investigation, including payments to people below working age, people who had died, government employees, grant recipients and state-funded students.

In February 2025, Independent Online reported, the Special Investigating Unit (SIU) told Parliament's standing committee on public accounts that UIF had paid approximately R57.4 billion to 240,000 entities. The SIU had investigated 5,763 matters involving civil servants and finalized 4,699, and R114 million had been paid irregularly to civil servants. It had completed 25 cases on payments to 52 dead people and 36 on payments to 53 prisoners. Across the investigation it had recovered R229.4 million and made 738 referrals to the National Prosecuting Authority and 174 for disciplinary proceedings against officials. It had begun investigating 1,100 more entities, with a potential recovery of R7.2 billion.

Sources: Auditor-General: Covid-19 Relief Package Landed in an Environment with Many Control Weaknesses (Auditor-General of South Africa, media release, 2 September 2020); SIU recovers R229 million in UIF Ters payment probe (Independent Online, February 2025).

COVID-19 Loan Guarantee Scheme

The loan guarantee scheme, a partnership between the National Treasury and the South African Reserve Bank, lent through the banks to small businesses for salaries and other operating and start-up costs. The banks' November update says the scheme "does not extend non-repayable grants to companies in financial difficulties." By 24 October 2020 banks had lent R16.71 billion under it, beside R33.61 billion of payment breaks on existing credit agreements, which had ended after six months.

Participating banks had received 44,912 applications by that date. Of these, 27% had been approved and taken up and 15% were still being assessed. Thirty-seven percent were rejected, either for failing the eligibility criteria set by the Treasury and the Reserve Bank or for failing the banks' own risk criteria. The main reasons given were a poor record of paying debt and loans too large for the business to repay. Banks had been given more flexibility to assess ability to repay from the beginning of August 2020, but the update says demand "remains below original expectations". On present trends the banks expected they would probably have lent R24.41 billion by January 2021, a shortfall they put down largely to weak business and economic conditions; business owners were reluctant to take on more debt.

Sources: November Update: COVID-19 Loan Guarantee (The Banking Association South Africa, 12 November 2020; figures to 24 October 2020).

Philippines

Small Business Wage Subsidy (SBWS)

From 16 April 2020 the Department of Finance ran the SBWS through the Social Security System (SSS) and the Bureau of Internal Revenue. It paid two months of wage subsidy to formal-sector employees of small businesses who could not work because quarantine had closed their workplaces. Employers applied online, and the names, tax identification numbers (TINs), bank accounts and mobile numbers they entered were treated as final; SSS management told the auditors no procedure had been set up to check them. The Commission on Audit (COA) reported in September 2023 that ₱45.1 billion of the ₱51 billion budget was paid between 29 April and 25 June 2020, to 3.1 million of the 3.4 million employees targeted. In non-essential businesses 717,442 targeted employees, 41%, were not served, while quasi-essential businesses exceeded their target by 412,077.

The auditors found ₱1.92 billion paid to 132,506 beneficiaries with duplicate or invalid TINs. Payouts of ₱6.7 billion failed or were rejected and were not reprocessed, and ₱5.7 billion of unused funds went back to the Bureau of the Treasury. Management relied on employers' undertakings and self-attestation to check eligibility, and the COA judged that recovering payments to ineligible beneficiaries would be difficult. The bank account for voluntary returns also took deposits from other programs, so returns to the SBWS could not be reliably counted.

Sources: Performance Audit Report PAO-2023-03: Small Business Wage Subsidy (SBWS) Program (Commission on Audit, Philippines, September 2023).

Social Amelioration Program: Emergency Subsidy Program (ESP)

The Bayanihan to Heal as One Act (Republic Act No. 11469) made the social welfare department, DSWD, responsible for an emergency subsidy to around 18 million families during the 2020 lockdowns. The subsidy ran from ₱5,000 to ₱8,000 a month for two months, set by region from the local minimum wage: ₱8,000 in the National Capital Region, where the daily minimum wage was ₱537, and ₱5,000 in the lowest-wage regions. Families already in the Pantawid Pamilyang Pilipino Program (4Ps) had their cash grants and rice subsidy, estimated at ₱2,150 a month, topped up to the regional amount.

DSWD's August 2020 summary gives the payouts. The first tranche paid ₱99,982,104,250 to 17,695,811 beneficiaries: ₱80.62 billion to 13,310,259 low-income families outside 4Ps, ₱18.57 billion to 4,287,420 4Ps households and ₱785 million to 98,132 ride-hailing and public utility vehicle drivers in Metro Manila, ₱8,000 each. The second tranche paid ₱46,449,503,600 to 7,333,891: 5,946,887 non-4Ps families and 1,387,004 4Ps households. During the first tranche 1,075,970 beneficiaries were paid twice; DSWD counted them as having received both tranches, from the first tranche's budget.

Families left off the first lists were paid later. By August, ₱21.52 billion had gone to 3,246,325 waitlisted low-income families, and 1,758,441 waitlisted families in areas under enhanced community quarantine had received a second tranche, ₱12.13 billion. A further 150,695 drivers nationwide received ₱1.06 billion, and 33,336 drivers in those areas ₱203 million. The payments listed add up to about ₱181.3 billion.

Sources: DSWD Social Amelioration Program/Emergency Subsidy Program (Department of Social Welfare and Development, as of August 2020).

Colombia

Ingreso Solidario (cash transfers)

Ingreso Solidario was the cash transfer through which Colombia temporarily widened its social protection system during the pandemic. The World Bank lent the government US$700 million to support its COVID-19 response. Its results brief, the lender's account of the operation it financed, gives no cost for Ingreso Solidario and no figure for payments to ineligible households. It reports that the number of vulnerable households benefiting from social programs through Ingreso Solidario increased by 3 million. Alongside it, the government made additional transfers in existing programs to 2.6 million poor households, 296,000 young people and 1.6 million poor elderly people. The brief says these and other measures are estimated to have kept approximately 632,000 people out of poverty, while at least 1.45 million people are estimated to have fallen into it during 2020.

Sources: Supporting Colombia’s COVID-19 Crisis Response (World Bank results brief).

Kenya

COVID-19 emergency cash transfer

Kenya kept paying its existing cash-transfer program, Inua Jamii, through the pandemic and added an emergency transfer for "newly" vulnerable households that were not Inua Jamii beneficiaries. The government appropriated Ksh 10 billion in the 2019/20 financial year, and 333,200 households in 47 counties received Ksh 1,000 a week by mobile money for four months. The Kenya Institute for Public Policy Research and Analysis (KIPPRA), which reports these figures, gives no total paid out under the emergency transfer and no figure for ineligible payments. It describes the new group as mainly urban households with younger, better-educated heads, unlike the households poverty-targeted programs already supported. A microsimulation it cites put the pandemic's addition at about 2 million newly poor Kenyans. An EU-funded consortium led by the Kenya Red Cross and Oxfam paid monthly top-ups of Ksh 5,668 to Inua Jamii households in informal settlements in Nairobi and Mombasa. KIPPRA reports that poor coordination between the national single registry and non-profit institutions led to duplication and double dipping.

Sources: Social Assistance a Lifeline for Households During COVID-19 Crisis — KIPPRA, July 20, 2021 (Blogs) (KIPPRA blog, 20 July 2021).

Tax relief (PAYE, VAT and turnover tax cuts)

When the pandemic reached Kenya in March 2020, the government cut Pay As You Earn and corporate tax rates from 30 to 25 percent, Value Added Tax from 16 to 14 percent and turnover tax from 3 to 1 percent. People earning below Ksh 24,000 a month got 100 percent tax relief. Ukur Yatani, Cabinet Secretary for the National Treasury and Planning, recounted the measures in his Budget Statement to the National Assembly on 10 June 2021.

The statement says the reliefs, introduced in April 2020, reduced revenue collection and hurt the implementation of the budget, and that in January 2021 the government reversed some of them, which improved collection. Containment measures returned in March 2021 with the third wave and lasted one month. The statement does not say which of the measures were reversed, and it gives no figure for the revenue the cuts cost.

Sources: Kenya National Treasury and Planning, "Budget Statement FY 2021/22" (National Treasury and Planning, 10 June 2021).

Nigeria

Targeted Credit Facility (TCF)

The Central Bank of Nigeria (CBN) introduced the Targeted Credit Facility in March 2020 to support households and micro, small and medium enterprises hit by COVID-19, lending through NIRSAL Microfinance Bank at an interest rate of 5%. The money came from the Micro Small and Medium Enterprise Development Fund. Households could borrow up to ₦3 million and small businesses up to ₦25 million. According to Nairametrics, ₦419.42 billion had been disbursed by 30 September 2023; principal repaid came to about ₦41.39 billion, roughly ₦378.03 billion was outstanding and ₦261.07 billion was overdue, which the site reported as over 62% of the amount disbursed. A document seen by Nairametrics counted 792,936 beneficiaries: 117,964 small businesses and 674,972 households. Among the problems the document lists are a widespread perception of the loans as grants, which left beneficiaries unprepared to repay, and inflation. The CBN and NIRSAL Microfinance Bank have sought to recover arrears by deducting from defaulters' bank accounts under the Global Standing Instruction mandate, and beneficiaries have reported sudden, unexpected deductions. NIRSAL Microfinance Bank says the loans were not grants.

Sources: Nigerians fail to repay over 62% of CBN and NIRSAL COVID-19 loans (Nairametrics, 28 March 2024); Targeted Credit Facility (TCF) (Central Bank of Nigeria).

CACOVID food palliatives

The Private Sector-led Coalition Against COVID-19 (CACOVID) bought food relief for almost 2 million of the most vulnerable families, over 10 million people, across Nigeria's 774 local government areas, and delivered it to the state governments to distribute. By late October 2020, amid the violence that came with the #EndSARS protests, warehouses holding pandemic palliatives had been looted in several states and the Federal Capital Territory.

The coalition's statement, issued through the Central Bank of Nigeria's acting director of corporate communications, Osita Nwanisobi, and reported by Nairametrics on 27 October 2020, called for calm. It blamed the delay on "the sheer scale of this nationwide food programme" and on the timing of orders and deliveries, which coincided with the lockdowns and forced it to roll out distribution "in a staggered manner". The large order and the production cycle had delayed delivery to the states and so, it said, the governors' own distribution. Some intended beneficiaries, it said, would now not receive the food. It added that not every warehouse of palliatives held CACOVID's donations, since the federal government, the states and other organizations had donated too. Nairametrics added that Nigerians angry that the food was still undistributed months after the programme was announced had accused politicians of keeping it for themselves, their cronies and possible handouts during election campaigns.

Sources: CACOVID reacts to looting of food items, explains why distribution of palliatives was delayed (Nairametrics, 27 October 2020).

COVID-19 Cash Transfer and the Rapid Response Register

On 19 January 2021 Vice President Yemi Osinbajo launched the Rapid Response Register (RRR), a list of the urban poor for Nigeria's COVID-19 cash transfer. The federal information ministry's release on the launch said: "Each beneficiary will receive five thousand naira for a period of six months." Neither the release nor Osinbajo's launch speech gives the number of people paid from the new register or its total cost.

Nigeria already had a National Social Register, built under the World Bank-supported National Social Safety Nets Project (NASSP). Osinbajo said it listed about 24.3 million poor and vulnerable individuals, about 5.7 million households, as of 31 December 2020, and that about ₦10 billion a month was going directly to about 2 million of those households. Because registration had started in the poorest areas, the people on it were largely rural. The lockdown hit urban informal workers and daily-wage earners who were not on it, and the RRR was designed for them. It targets urban poor wards selected with satellite remote sensing, machine learning algorithms and big data analysis; Osinbajo said it was the first such targeting strategy in Sub-Saharan Africa and that Nigeria would be the first country to use it. The Ministry of Humanitarian Affairs, Disaster Management and Social Development, under Minister Sadiya Umar Farouq, designed it with NASSP.

Osinbajo called the existing national transfer "about the largest evidence-based effort by any administration on poverty reduction." He set an aim of a social security program for at least twenty million Nigerians within two years and said the only constraint was funding, which the government had to look for.

Sources: Office of the Vice President of Nigeria, "Launch of the Rapid Response Register (RRR) for the Economic Sustainability Plan COVID-19 Cash Transfer" (Office of the Vice President, 19 January 2021); VP Osinbajo Lauds Humanitarian Minister For Covid-19 Rapid Response Register Cash Transfer Project (Federal Ministry of Information and National Orientation, 20 January 2021).

Germany

KfW coronavirus loans (Unternehmerkredit and Schnellkredit)

Germany's emergency credit ran through KfW, the state development bank, but a firm did not apply to KfW. It applied to its own bank, its Hausbank, which approved the loan through KfW. By 25 May 2020 KfW had 50,085 loan applications, and 47,743 had been approved through the banks with commitments of €25.6 billion. The product in most demand was the Unternehmerkredit for small and medium-sized firms: over 35,000 applications and €8.9 billion committed. The Schnellkredit, which the state secured 100%, had 8,559 approvals worth about €2.9 billion.

The figures come from the federal government's own page of 28 May 2020, whose opening line reads "Die Corona-Hilfe kommt an" (the coronavirus aid is getting through). Manufacturing, maintenance and repair, and hotels and restaurants applied most often; by state, North Rhine-Westphalia, Bavaria and Baden-Württemberg. Its graphic, with KfW data as of 20 May 2020, counts 48,120 loan applications, 45,910 approvals and 26 rejections.

Sources: Kredite der KfW: Schnelle Hilfe für die Wirtschaft — KfW Kredite gegen die Krise (Corona-Hilfe) (Bundesregierung, 28 May 2020).

Corona-Soforthilfe (emergency grants)

At the end of March 2020 the federal government launched Soforthilfen: one-off non-repayable grants for solo self-employed people, small firms, freelancers and farmers in economic difficulty. A business with up to five employees could receive €9,000 and one with up to ten €15,000, to cover three months from an application made in March, April or May 2020. Where that period overlapped with the later bridging aid, the grant was to be set off against it.

The first supplementary budget of 27 March 2020 set aside €50 billion for the grants; the second, of 14 July 2020, put the line at €18 billion. The federal government's page of 28 May 2020 says in its opening paragraph that twelve billion euros had been paid "an mehr als zwei Millionen Antragsteller", to more than two million applicants. Its question-and-answer section, with figures to 25 May 2020, gives more than 2 million applications nationwide, €12.5 billion paid out and 1.62 million approvals.

Sources: Kredite der KfW: Schnelle Hilfe für die Wirtschaft — KfW Kredite gegen die Krise (Corona-Hilfe) (Bundesregierung, 28 May 2020); Abschließende Mitteilung — Corona-Überbrückungshilfen: Ausgestaltung und Zugangsvoraussetzungen (Bundesrechnungshof, 25 May 2022).

Überbrückungshilfe I–IV (bridging aid)

Bridging aid took over from the emergency grants in June 2020 and paid a share of a firm's fixed costs, scaled to how far its turnover had fallen. Where turnover had fallen by more than 70%, round I (June to August 2020) paid 80% of eligible fixed costs and round II (September to December 2020) 90%. Rounds III (November 2020 to June 2021) and III Plus (July to December 2021) paid 100%. From round III larger firms could apply too. Rounds III and III Plus added an equity grant for firms whose turnover had fallen by at least 50% in at least three months. Applications went in through a tax adviser, auditor or lawyer, the "prüfende Dritte", on a federal online platform, and offices named by the Länder approved and paid.

The money was a Billigkeitsleistung under § 53 of the federal budget code: a payment nobody has a right to, made as a rule to offset hardship from an event the recipient could not foresee. By 26 April 2022, according to the government data platform the audit office cites, €69.5 billion of coronavirus business aid had been paid, €39.7 billion of it as bridging aid I to IV. Round I paid €1.4 billion on 123,340 applications, round II €2.8 billion on 203,987, round III €29.9 billion on 445,069, III Plus €5.0 billion on 194,868 and round IV €0.6 billion on 25,431.

The Bundesrechnungshof, the federal audit office, sent its findings to the economy ministry (BMWK) on 25 May 2022. The Bund paid the whole cost, although supporting business is normally the Länder's job; the ministry acknowledged it had not examined whether they should contribute. By the audit office's reckoning, with an equal split the Bund could have avoided spending up to €19.2 billion. The list of reimbursable fixed costs included investment in digitization and building work for hygiene plans. The ministry said the investments kept firms trading and so reduced the damage. The audit office held that a hardship payment makes good damage already done, and its word for using one to fund investment that prevents damage is "haushaltsrechtswidrig", unlawful under budget law.

When a firm closed or filed for insolvency, repayment claims could hardly be enforced. The ministry, citing low recovery rates in insolvency, thought it acceptable to waive them, and acknowledged that it could not rule out aid to firms that were not competitive regardless of the pandemic. It could not say whether, or by how much, firms had taken money they did not need. It had set no criteria for judging whether the program worked, and said that review could probably begin in the second half of 2022.

Sources: Abschließende Mitteilung — Corona-Überbrückungshilfen: Ausgestaltung und Zugangsvoraussetzungen (Bundesrechnungshof, 25 May 2022).

Japan

Koyō chōsei joseikin (Employment Adjustment Subsidy): fraud and recovery

The health and labor ministry's page on the Employment Adjustment Subsidy sets out what happens to an employer who claims it falsely. The employer must repay everything from the assessment period in which the fraud occurred onward, plus a 20% penalty and late interest at 3% a year. For five years it can receive no employment subsidy of any kind. The prefectural labor bureau publishes the employer's name. It may withhold it if the employer owned up before the bureau investigated, repaid in full within a month of the repayment order and the case was not especially serious. Below ¥1 million it withholds the name unless the case was especially serious or malicious. If a labor and social-security consultant or other agent took part, the name is published whatever the amount.

The page gives totals to the end of March 2026, with no start date, no breakdown by year and no split between the pandemic special measures and ordinary claims. By the end of March 2026, according to the ministry, 4,640 subsidy decisions had been canceled for fraud, totaling about ¥115.3 billion, and about ¥89.6 billion had been recovered.

Sources: 雇用調整助成金（不正受給関係） (Ministry of Health, Labour and Welfare).

Tokubetsu teigaku kyūfukin (Special Cash Payment)

Japan paid ¥100,000 to every person on the basic resident register on 27 April 2020. Municipalities ran the payment and the national government reimbursed all of their costs. The money was claimed by, and paid to, the head of each household. Claims went in by post or online with a My Number card, and payment was by transfer to a bank account in the claimant's name. The first supplementary budget for 2020 provided ¥12.88 trillion: ¥12.73 trillion for the payments and ¥145.9 billion for administration.

Payments reached ¥3.85 trillion by 5 June 2020 and ¥12.32 trillion by 31 July. By 31 March 2021 they stood at ¥12.67 trillion. On 23 April 2020, four days before the reference date, the internal affairs ministry asked four telecoms industry bodies to warn their users about frauds riding on the payment.

Sources: 特別定額給付金(新型コロナウイルス感染症緊急経済対策関連) (Ministry of Internal Affairs and Communications).

Jizokuka kyūfukin (Sustainability Benefit for businesses)

The Sustainability Benefit was a grant for firms and sole traders whose monthly business income had fallen by at least 50% on the same month a year earlier. It paid up to ¥2 million to a company and ¥1 million to an individual. The amount was the 2019 annual income minus twelve times the income of the month the applicant chose. Applications ran from 1 May 2020 to 15 February 2021. They came to 4.41 million, 1.43 million of them in May 2020 alone. By the end of March 2021 the Small and Medium Enterprise Agency had paid 4.23 million of them, ¥5.51 trillion in total. Over 60,000 were refused and over 100,000 withdrawn. Of those paid, 68.4% received the money within the two weeks the agency aimed for.

The agency contracted the paperwork out. In April 2020 it signed a contract worth up to ¥76.9 billion with an incorporated association, サービスデザイン推進協議会 (literally, the Service Design Promotion Council). At signing, 99.8% of the contract value was to be passed on to subcontractors and outside suppliers, and 97.4% to Dentsu. The final cost was ¥66.9 billion, of which 96.0% went to subcontractors and suppliers and 95.9% to Dentsu. Across 61 other projects of the ministry the average subcontracted share was 54.6%, according to the Board of Audit. By 1 September 2020, 92 participants across six tiers of subcontracting were reported as handling applicants' personal data. Not a few had started before the agency approved it, and the agency did not know until the paperwork arrived.

Anyone found to have claimed falsely must repay the whole grant, plus late interest at 3% a year, plus 20% of that total, and their names are published. By 31 August 2026 the agency had received 25,390 offers to repay voluntarily and completed 28,787 repayments, about ¥20.4 billion.

Sources: 持続化給付金事業の実施状況等について (Board of Audit of Japan, report on fiscal 2020); 不正受給及び自主返還について (Ministry of Economy, Trade and Industry).

Go To Travel

Go To Travel paid part of the cost of trips: a discount on the travel price and regional coupons to spend locally. It started on 22 July 2020, with Tokyo excluded until 30 September; coupons were usable from 1 October. From 15 October the tourism agency asked travel firms to fix the discount at 35%, and at the end of October it announced a cap of seven nights a trip. Nationwide suspension began on 28 December 2020 and from 8 March 2021 was extended with no end date. The tourism agency's preliminary count, published on 10 February 2021, was at least 87.81 million person-nights and ¥539.9 billion of support, both approximate: ¥408.2 billion in discounts and ¥131.7 billion in coupons.

The whole Go To campaign had a 2020 budget of ¥3.08 trillion; that year it paid ¥721.3 billion in benefits and ¥221.7 billion under its contracts, which for travel included ¥115.4 billion advanced for cancellation charges. For travel, the agency paid its secretariat ¥621.2 billion by the end of March 2021, and the secretariat paid ¥620.7 billion on to 34,672 registered travel businesses and 376,500 coupon shops. The difference, more than ¥400 million, came from duplicate claims the secretariat had made by mistake for the same trips. At the end of September 2021 it still held the money, which the Board of Audit found not appropriate. Coupon spending from October to December 2020 went mostly to souvenirs and other retail (27.4%), restaurants (17.4%) and convenience stores and supermarkets (12.3%).

In October 2020 the agency learned that people were booking trips, using the electronic coupons on the day and never arriving. It told the 315 travel firms selling such trips to check identities and take payment before departure, and added SMS verification in November. The agency says that once it confirmed the 315 firms' controls in December 2020, no more such no-shows occurred. No-show coupon use reported by the end of March 2021 came to ¥21,147,000. Where it could identify the booker the agency demanded repayment: ¥643,000 by the end of September 2021, of which ¥144,000 had been repaid.

Sources: サービス産業消費喚起事業（GoToキャンペーン事業）の実施状況等について (Board of Audit of Japan, report on fiscal 2020); Go To トラベル事業の利用実績等について (Japan Tourism Agency, 10 February 2021).

India

Pradhan Mantri Garib Kalyan Anna Yojana (free foodgrains): the Chhattisgarh audit

Under the Pradhan Mantri Garib Kalyan Anna Yojana (PMGKAY), the central government gave every beneficiary of the National Food Security Act five kilograms of free rice a month on top of the regular ration. Each family also got one kilogram of pulses. It ran from April to November 2020 and again from May 2021 to March 2022. Chhattisgarh had been giving its priority households extra rice at ₹1 a kilogram since July 2019. A card with three members, for example, drew 35 kg a month. In April 2020 the state government reset the totals so that, with the central rice included, the same card still drew 35 kg. The audit by the Comptroller and Auditor General of India found that 31.05 lakh beneficiaries (3.1 million) in households of one to three members received no more rice than before. Another 136.27 lakh (13.6 million) in larger households received three kilograms a person extra instead of five: 167.32 lakh people (16.7 million) in all. The state replied in May 2022 that it had distributed at least the combined regular and additional entitlement; the auditors did not accept the reply.

The free chana (chickpeas) fell short as well. Of 3.60 lakh quintals allotted for the seven months from April to November 2020 (June excepted), 3.42 lakh were distributed, leaving 17,803 quintals undistributed to 2.54 lakh households. In the areas where families already bought two kilograms a month at ₹5 under a state scheme, 25.34 lakh households received the same quantity as before the pandemic.

Under the Atma Nirbhar Bharat Scheme for migrants without ration cards, the central government allocated 20,076 tonnes of rice for May and June 2020; the state lifted 2,108 tonnes and distributed 1,964.41. Because IDs were generated more than once for the same person or family, 2,644 people and 1,641 families appeared on the list more than once and drew 283.01 quintals of rice and 35.98 quintals of chana beyond their entitlement. The audit says the possibility that this grain was diverted "cannot be denied". Of 222,605 registered migrants, 30,218 (14%) received no rice, and of 109,179 families, 20,395 (19%) received no chana; the state said they did not turn up at the shops. The auditors inspected 120 fair price shops with the department's staff. In 106 (88%) there were no separate PMGKAY stock and distribution registers, 96 (80%) had not been inspected monthly, and 117 (98%) had never had the social audit the state's food security act requires.

Sources: Chapter II Compliance Audit of General, Social and Economic Sectors (Report for the Year Ended 31 March 2021) (Comptroller and Auditor General of India, Chhattisgarh).

Emergency Credit Line Guarantee Scheme (ECLGS)

The Emergency Credit Line Guarantee Scheme guaranteed loans to businesses through the National Credit Guarantee Trustee Company (NCGTC). The NCGTC's state-by-state figures, given in an annexure to Lok Sabha unstarred question 1822, run to 31 January 2023. By then 11,353,254 loans to micro, small and medium enterprises had been guaranteed, for ₹239,147.53 crore (about ₹2.39 trillion): an average of about ₹2.1 lakh (₹210,000) a loan.

The averages differ widely by state. West Bengal had the most loans, 2,007,453, averaging about ₹68,000 each. Maharashtra had the most money, ₹33,439.94 crore on 971,628 loans, about ₹3.4 lakh each. Delhi's 93,672 loans averaged about ₹13.3 lakh.

Sources: ECLGS Support to MSMEs (Lok Sabha unstarred question 1822, annexure I; data NCGTC).

Indonesia

Bansos (social assistance) in Jabodetabek: the Juliari Batubara case

Indonesia's COVID-19 social assistance included staple-food packages, bansos sembako, bought for Greater Jakarta: Jakarta, Bogor, Depok, Tangerang and Bekasi, together Jabodetabek. The former social affairs minister, Juliari Peter Batubara, was convicted of corruption in the procurement of those packages by the anti-corruption court at the Central Jakarta District Court. On Monday 23 August 2021 a panel chaired by Judge M Damis sentenced him to 12 years in prison and a fine of Rp500 million. It also stripped him of the right to be elected to public office for four years after serving his sentence. Prosecutors from the Corruption Eradication Commission (KPK) had asked for 11 years.

The court convicted him under article 12(b) or article 11 of the anti-corruption law, each read with article 18. The judges said he had denied receiving bribes worth Rp32.482 billion in the procurement. One of them, Yusuf Pranowo, listed the denial among the aggravating factors and called it "lempar batu sembunyi tangan": throwing a stone and hiding the hand. The court also weighed against him that the offence was committed during a declared non-natural disaster emergency, the pandemic, and that corruption in the court's jurisdiction was rising in both number and seriousness.

Sources: Eks Mensos Juliari Divonis 12 Tahun Penjara dalam Perkara Bansos (hukumonline.com, 23 August 2021).

Bansos PKH, Sembako/BPNT and BST: the 2021 BPK audit

Indonesia's Supreme Audit Board (BPK) found that Rp6.93 trillion of social assistance paid under the Program Keluarga Harapan (PKH), Sembako/Bantuan Pangan Non Tunai (BPNT) and Bantuan Sosial Tunai (BST) was "terindikasi tidak tepat sasaran", indicated as mistargeted. The finding comes from BPK's compliance audit of the social protection program for fiscal 2021, summarized in its IHPS II 2021 and reported in the May 2022 issue of its magazine, Warta Pemeriksa.

BPK traced the problem to the recipient lists. Aid went to beneficiary families (KPM) who were not in the October 2020 integrated social welfare data (DTKS) or among the local governments' proposals entered through the SIKS-NG application. Families with problems in 2020 were still set as recipients in 2021, as were families already deactivated, families reported dead and duplicate recipients. BPK also found welfare cards (KKS) that had never been distributed and families that had not transacted, leaving Rp1.11 trillion of PKH and Sembako/BPNT balances that had not been paid into the state treasury. It recommended that the Ministry of Social Affairs (Kemensos) repair and validate its PKH recipient data and instruct the distributing banks to return the Rp1.11 trillion to the treasury and send BPK the proof.

Kemensos, through its director general for social protection and security, said recipients for the first and second stages were set from the 2020 DTKS. That list, it said, was not inclusive because it required 47 variables that were hard to meet; from the third stage the problem did not recur.

Sources: BPK, Warta Pemeriksa, Edisi 5 Vol. V (May 2022) (Badan Pemeriksa Keuangan, May 2022).

Kartu Prakerja: the 2021 BPK audit

BPK found that 165,544 Kartu Prakerja participants, with assistance worth Rp390.32 billion, were added to the blacklist of people barred from the program after they had been designated as participants, and called the payments waste (pemborosan). They came from batches 12 to 22, which enrolled 6,019,842 people, so 2.7% of those participants were on the list. BPK's further check of the database found that the participants had been designated before the blacklist data was obtained from the agencies that hold it. It recommended that the Coordinating Minister for Economic Affairs order the head of the program's management unit to account for the Rp390.32 billion in payments made after the blacklist data, or updates to it, had been received.

A management unit (MPPKP) under the Coordinating Ministry for Economic Affairs runs the program. Its rules exclude state officials, regional legislators, civil servants, soldiers, police officers, village heads and village officials, and the directors, commissioners and supervisory boards of state-owned and regional enterprises. Under an operating procedure in force from 19 February 2021, applicants' national ID numbers are screened automatically against blacklists supplied by ministries and agencies.

BPK also found that 119,494 participants, with assistance of Rp289.85 billion, were indicated as mistargeted because the aid went to workers earning more than Rp3.50 million a month, and that the identity cards of 42 participants were invalid. Another 68,733 participants, worth Rp181.6 billion, were also registered as recipients of the 2020 wage subsidy for teachers and education staff at the Ministry of Education and Culture.

MPPKP replied that the accuracy and timeliness of blacklist data is each ministry's responsibility and outside its control. It said it had offered to connect its system to theirs through an API, that no ministry had built one, and that the data still arrived by hand each period. All the overlapping recipients, it said, had been designated before their data entered the blacklist. MPPKP said it had frozen their Prakerja accounts.

Sources: BPK, Warta Pemeriksa, Edisi 5 Vol. V (May 2022) (Badan Pemeriksa Keuangan, May 2022).

Peru

Bonos (household cash grants): Yo me quedo en casa, Independiente, Rural and Familiar Universal

Peru paid its lockdown relief to households as four cash grants, each created by emergency decree and aimed together at more than 6.8 million households. The first, "Yo me quedo en casa" ("I'm staying home"), paid S/380 to 2.75 million poor and extremely poor households in the areas the health ministry ranked most vulnerable, and was doubled in April. The Bono Independiente (27 March) paid the same to about 780,000 households of self-employed workers classed as not poor, whose members earned under S/1,200 and were on no public or private payroll. The Bono Rural (19 April) paid S/760 to 830,000 rural households; the report notes it came 33 days after the national health emergency was declared. The Bono Familiar Universal (5 May) paid S/760 to poor households and to those in the Juntos, Pensión 65 and Contigo programs, where no member earned over S/3,000 a month and none had received another grant. Its first lists, approved on 14 May, covered about 600,000 urban and 250,000 rural households.

Every list was built from the household targeting system, Sisfoh, and its general household register, which the national ombudsman, the Defensoría del Pueblo, found out of date. Its 2020 report counts 1,132 complaints it handled on Yo me quedo en casa and the Bono Rural, and 168 on the online registration platform for households left off the lists. It learned of 10 cases in which employees of eight municipalities appeared to have received Yo me quedo en casa, a grant for the poor. Some were local officials, or headed or worked in the municipal units that classify households. The report questioned paying mainly through banks when, by a 2016 survey it cites, 41% of people had an account (24% in rural areas); by INEI figures, 6.7% of rural households had internet access, against 39.7% of urban ones. The report found that the grants' beneficiary lists had not been published.

Sources: Entrega de Bonos a Hogares en el Contexto de la Emergencia por la COVID-19: Dificultades y Recomendaciones (Defensoría del Pueblo, Serie Informes Especiales Nº 025-2020-DP).

Mexico

Crédito a la Palabra (Programa de Apoyo Financiero a Microempresas Familiares)

Mexico's Secretaría de Economía (SE) handed out 25,000 pesos each to 1,401,636 beneficiaries in 2020, 35,040.9 million pesos in all, on a promise to pay it back. By September 2021 they had repaid 2,083.9 million pesos, 5.9% of the total, according to the Auditoría Superior de la Federación (ASF), which audited the program for the 2020 public accounts. Only 442,110 beneficiaries, 31.5%, had made any repayment.

The SE announced the program, also called Crédito a la Palabra, on 24 April 2020 as a response to the COVID-19 emergency. It had two strands: Microempresa Familiar covered municipalities chosen by the SE, and Apoyo Solidario a la Palabra could be national, as the social security institute (IMSS) determined. The guidelines named formal and informal microbusinesses, the self-employed, firms registered with IMSS that had kept their average first-quarter 2020 workforce as of 15 April 2020, and domestic and independent workers with rights in force on 30 April 2020. The money was paid through financial institutions. The repayment terms were three months' grace and then 33 monthly repayments, which the ASF notes came with no penalty or sanction for paying late or not at all.

The ASF audited 35,492.5 million pesos, 99.9% of the 35,534.7 million pesos the SE spent on the program in 2020. It found that the SE had not determined the potential population of each selected municipality before operations began in April 2020. It paid 3.1 million pesos to 124 beneficiaries aged between 2 and 17, although the rules required recipients to be adults. Another 134 beneficiaries received the support twice, worth 3.35 million pesos, and for 550,000 pesos paid to 22 beneficiaries the SE could not show bank confirmation of delivery.

The SE had no controls to identify which beneficiary made each repayment paid directly to the federal treasury (TESOFE), the ASF found. Of the 2,083.9 million pesos, 1,701.0 million went into an SE bank account and 382.9 million directly to TESOFE. Repayments of 111.4 million pesos from 31,019 people could not be matched to anyone the program had paid. The ASF closed its work on 27 January 2022 with 12 results, 3 recommendations, 5 referrals for administrative sanctions and 3 pliegos de observaciones (formal findings of presumed damage to the federal treasury; together they come to 7.55 million pesos). At the time of the review, 563.7 million pesos of repayments were still in the SE's bank account and had not been returned to the treasury.

Sources: ASF, Auditoría de Cumplimiento 2020-0-10100-19-0379-2021 (379-DE), "Apoyo Financiero a Microempresas Familiares" (Auditoría Superior de la Federación, Cuenta Pública 2020; Internet Archive copy of 25 September 2023).