The Financial Cost Of Fraud 2021

2. Overview2. Overview2.12.22.312Our research has now reviewed807 loss measurement exercisesundertaken from 1997 to 2020.The exercises took place across40 different types of expenditurein 49 organisations from 10countries considering losses inexpenditure with a total valueof 25.9 trillion. The value of theexpenditure examined has notbeen uprated to 2020 values.The losses referred to are apercentage loss of expenditure.This report is based on extensiveglobal research, building onpreviously established directknowledge, to collate informationabout relevant exercises.The data was then analysedelectronically. Exercises werecollated from Europe, NorthAmerica, Australasia and Africa.None were found in Asia.The report has excludedguesstimates, figures derived fromdetected fraud losses, and figuresresulting from surveys of opinion.It has also excluded some lossmeasurement exercises whereit is clear that they have not metthe standards described below.2.4It has included exercises which: have considered astatistically valid sample ofincome or expenditure have sought and examinedinformation indicating thepresence of fraud, erroror correctness in eachcase within that sample have been completedand reported have been externally validated have a measurable levelof statistical confidence have a measurable levelof accuracy.2.5 There are a number of caveats.2.6Some of the exercises haveresulted in estimates ofthe fraud frequency rate,some of the percentage ofexpenditure lost to fraud andsome have measured both.2.7It is also the case that someexercises have separatelyidentified and measured fraudand error, and some have not.

2.82.9Sometimes, once such exerciseshave been completed, theorganisations concerned havemistaken the view of the authorsof this report and decided not topublish their results. Transparencyabout the scale of the problemis a key factor in its solution,because attention can be focusedand a proportionate investmentcan be made to address the issue.In some cases, those directlyinvolved in countering fraud havedecided, confidentially, to provideinformation about unpublishedexercises for wider consideration.In those cases, while the overallfigures have been included in thefindings of this report, no specificreference has been made tothe organisations concerned.The financial cost of fraud 20212.10 We are also aware of a verysmall number of other exerciseswhich have been completed,but which have not beenpublished and where nothingis known of the findings.2.11 Finally, it is important toemphasise that this research willnever be complete. More evidencebecomes available each year.However, the preponderance ofthe evidence does point clearly inone direction, as is explained later.2.12 While it is necessary to makethese caveats clear, theimportance of the evidencecollated in this report shouldnot be underestimated. Itshows that losses to fraud anderror represent a significant,damaging and cruciallyunnecessary business cost.13

3. Data from around the world3. Data from around the world3.1 The 10 countries in which the authors are awarethat fraud loss analysis exercises have taken place are:UnitedKingdomCanadaThe NetherlandsIrelandBelgiumUnited Statesof AmericaFranceZambiaAustraliaNew Zealand3.214By value of income or expendituremeasured, the United States hasundertaken the greatest amountof work in this area. This is adirect reflection of the ImproperPayments Information Act of 2002(IPIA) which requires designatedmajor US public authorities toestimate the annual amount ofpayments made where fraud anderror are present, and to reportthe estimates to the President andCongress with a progress reporton actions to reduce them. TheImproper Payments Eliminationand Recovery Act of 2010 furtherstrengthened this requirement.

3.3The guidance relating to theoriginal IPIA stated ‘The estimatesshall be based on the equivalentof a statistical random sample witha precision requiring a sample ofsufficient size to yield an estimatewith a 90% confidence intervalof plus or minus 2.5%2 . Thisremains the case although manyUS agencies undertake work tothe higher standard often foundin the UK and Europe – 95%statistical confidence and /- 1%.3.5In the UK, the government is onrecord as requiring this work tobe undertaken. Indeed in late2014, the government’s FraudError and Debt Taskforce, withthe agreement of Ministers, askedall government departments toundertake ‘random sampling’loss measurement exercises.This work has proceeded rapidlysince then, resulting in a majorstep forward to countering fraudin UK central government.3.4In other countries, while therehas not hitherto been any legalrequirement, there is a growingunderstanding that the keyto successful loss reductionis to understand the natureand scale of the problem.For example, in Europe, theEuropean Healthcare Fraud andCorruption Declaration, agreed byorganisations from 28 countriescalled for ‘the development of aEuropean common standard ofrisk measurement, with annualstatistically valid follow upexercises to measure progressin reducing losses to fraud andcorruption throughout the EU.’33.6These developments are partof a consistent trend. In each ofthe five year periods between1997 and 2016 – the growth inthe number of loss measurementexercises was marked, with atenfold increase in prevalence,as shown in the graph below.3.7In the four years since 2017there have been 249 exercisescompleted and this is projectedto reach 311 by the endof the five year period. 2 Appendix C to Office of Management and BudgetCircular A-123. 3 European Healthcare Fraud and CorruptionDeclaration 2004.Number of Fraud Loss Measurement (FLM) exercises against timeNumber of FLM e financial cost of fraud 20212002–20062007–20112012–201615

4. Types of income and expenditure and the nature of the figures16

4. Types of income and expenditureand the nature of the figures4.1 The types of income and expenditure wherelosses have been measured include:PayrollHealthcareSocial SecurityMining4.2ProcurementThe financial cost of fraud 2021InsuranceAgricultureThe key figures which havebeen produced concern thepercentage loss rate (PLR –i.e. the proportion of expenditurelost to fraud and error).Housing4.3EducationTax CreditsConstructionPensionsCompensationThere is more research stillto be done and it is intendedthat this report will beupdated on a regular basis.17

5. Fraud and error losses5. Fraud and error losses.1%38The range of percentage lossesacross all the exercises reviewedbetween 1997 and 2020 werefound to be between 0.02%and 63.96%4, with averagelosses of 6.42% (68.25% ofthe exercises showed lossfigures of more than 3%).31.7%5.130.2%Percentagelost 3%Since the start of the globalrecession in 2008, there hasbeen an increase in averagelosses from 4.57% to 8.58%for the period of 2019 to2020 – an increase of 88%.And remember, this is beforethe advent of COVID-19.We describe the impact ofthis later in the report. 4 An area of expenditure in the US Departmentof Veterans Affairs.Percentagelost 8%9.00%8.00%8.58%7.00%6.00%Average % lost5.2Percentagelost 1997–2007182019–2020

5.3The reasons for theseincreases, whether over thelast two years or over thelonger period since 2007, seemto go beyond the economiccycle. Previous research hassuggested some evidencethat certain frauds increaseduring recessions and plateauor decrease slightly duringperiods of economic growth.55.4This does not explain why thecost of fraud has continued toincrease since economies havereturned to growth. Furtherresearch will be needed but itmay be that longer term socialand technological factors are anunderlying cause of the growthof fraud, in addition to theeffect of the economic cycle.5.5Such factors might include: greater individualisation(less belief that we shouldall be bound by commonmoral and ethical ‘norms’) greater complexity ofprocesses and systems(it’s becoming easier todisguise fraud amidst thiscomplexity, which itself isharder to understand) most transactions arenow being undertakenby computer and thereare fewer face to facetransactions (fraudstersfeeling more distant from thevictims of their dishonestyand thus less concernedabout any response)The financial cost of fraud 2021 a perception that manysocietal ‘role models’ (bethey Members of Parliament,senior executives orother public figures) aredishonest, weakeningthe anti-fraud culture the increasing pace of changein business (with controlsstruggling to keep up).5.6The evidence demonstratesthat organisations which haveundertaken repeated exercisesto measure losses in thesame areas of expenditure,have reduced the lossesover time. This suggests thatorganisations that know theextent of their fraud losses arebetter at reducing the losses.5.7The global average loss rate for the entire period of theresearch (6.42%), when taken as a proportion of theglobal Gross Domestic Product (GDP) for 2020 ( 83.85trillion or 68.04 trillion6), equates to 4.37 trillion ( 5.38trillion). This sum is over twice the UK’s entire GDP.Even reducing such losses by 40%, which individualorganisations have achieved, would free up more than 1.55 trillion – a sum larger than the GDP of all but theseven largest global economies. 5 Gill, M. (2011) Fraud and Recessions: Views fromFraudsters and Fraud Managers. InternationalJournal of Law, Crime and Justice, 39, 204-214. 6 International Monetary Fund figures.19

5. Fraud and error losses5.8In the UK, applying that globalaverage loss rate to GDP7would imply total losses of 137 billion each year (althoughmore detailed UK-focused,sector by sector researchmakes this total even larger atnearer 190 billion8). Reducingsuch losses by 40% wouldfree up almost 55 billioneach year. This sum is greaterthan what the UK governmenthave spent on defence ( 50.6billion9) in 2019 to 2020.5.9On the basis of the evidence,it is clear that fraud and errorlosses in any organisation shouldcurrently be expected to be atleast 3%, probably almost 6.5%and possibly more than 10%. Itwould be wrong to go too muchfurther in terms of predictingwhere in this range lossesfor an individual organisationwill be. This cannot be donewithout some organisationspecific information aboutthe strength of arrangementsto protect it against fraud(its ‘fraud resilience’).5.10 Crowe UK and the CCFS,in parallel research, havedeveloped Europe’s mostcomprehensive database offraud resilience information,with data recorded concerningmore than 1300 organisationsfrom almost every economicsector. By combining the datawhich underpins this report andorganisation-specific informationabout fraud resilience, Crowe UKand CCFS are able to predict:the likely scaleof lossesthe key improvementswhich reduce themthe related costof making thoseimprovements.5.11 Crowe UK and CCFS can alsoaccurately measure losses ortrain client organisations to doso. The practical experience ofCrowe UK specialists, combinedwith the academic rigour ofCCFS researchers, provides anunparalleled expert resource.7 International Monetary Fund figures estimate UKGDP for 2020 to be 2.64 trillion or 2.14 trillion. 8 Annual Fraud Indicator 2017. 9 UK Defence Spending - Analysis, Charts Tables(ukpublicspending.co.uk).20

The financial cost of fraud 202121

6. The position since COVID-196. The position since COVID-196.1Since the advent of COVID-19 we have seen a surge in fraud. Thishas included both COVID-19 specific fraud and general economiccrisis-driven fraud.6.2 Examples of COVID-19 specific fraud Fraudsters have pretendedto be undertaking researchfor the World HealthOrganisation (WHO). Theyclaim to provide the victimwith a list of active infectionsin their area but to accessthis information the victimneeds to either click on a linkwhich redirects them to acredential-stealing web page. Emails asking for donationsto buy ‘medical preparationsand supplies’ for the NHSto cope with coronavirus. Other scams purporting tobe official messages fromthe government includetexts telling people theyhave been fined 250 forleaving their home more thanonce during lockdown.22 Fraudsters have beenproviding articles about thevirus outbreak with a link to afake company website wherevictims are encouraged toclick to subscribe to a dailynewsletter for further updates. Fraudsters have been sendinginvestment scheme andtrading advice encouragingpeople to take advantage ofthe coronavirus downturn. Fraudsters have purportedto be from HMRC offeringa tax refund and directingvictims to a fake website toharvest their personal andfinancial details. The emailsoften display the HMRC logomaking it look reasonablygenuine and convincing.

6.4The risks to organisations are bothinternal and external. Possibleeconomic pressures may lead peopleto radically re-evaluate loyalties andto rationalise behaviour which, innormal times, they would not considerappropriate. Previous recessionshave seen a new generation offraudsters created where the ‘normal’dishonest minority is enlarged by thosedetermined to protect their incomeand assets no matter what. This isparticularly the case because of someover-extending with a post-Brexit surgeof expenditure, house buying andinvestment, all of which came to anabrupt halt with the advent of COVID-19.6.5The impact of this has been seen inthe latest Office for National Statisticscrime statistics10 which have showna 19.8% increase in the incidenceof fraud in England and Wales.6.6It is likely that this will impact on thecost of fraud and error, and we willsee what this is in the next FinancialCost of Fraud report in 2022.6.3 Examples of economiccrisis driven fraud Employees who have eitherbeen made redundantalready, or think that theywill be, stealing client dataand intellectual property toingratiate themselves withprospective future employers. Fraudulent invoices submittedby under pressure suppliersfor goods and serviceswhich either haven’t everbeen ordered or havenever been delivered. Senior managers, unbeknownto their employers underfinancial pressure,using their authority tomanipulate accountsand misdirect money. Mandate fraud, where bankaccount details are changedto misdirect money.10 ONS CSEW statistics for the 12 months up to theend of March 2020 versus the same statistics forthe 12 months up to the end of September 2020.The financial cost of fraud 202123

7. Conclusion and recommendations7. Conclusion and recommendations7.17.224This is the eighth reportsince 2009 in an area where,for too long, the accuratemeasurement of losses wasconsidered either impossible ortoo difficult. It no longer is. Inmany areas, loss measurementhas become routine. Lossesto fraud and error can now betreated as a business cost likeany other – to be measured,managed and minimised.It is also the case that workto measure losses is highlycost-effective. Efforts toreduce losses are helped bygreater knowledge about thescale of the problem. Thedata shows that organisationswhich re-measure the samearea of expenditure haveconsistently lower loss rates.7.3Where losses have been measured, andthe organisations concerned have accurateinformation about their nature and extent, thereare examples, especially in the UK and US,where losses have been substantially reduced.

7.4Three things are clear.37.5Losses to fraud and error can bemeasured – and cost effectively.On the basis of the evidence it is likely thatlosses in any organisation and any area ofexpenditure will be at least 3%, probablynear to 6.5% and possibly more than 10%.Losses can be significantly reducedwhen accurate information about theirnature and extent is available.In any economic climate, not to consider the financial benefits of makingrelatively painless reductions in losses to fraud and error is foolhardy.Doing so can mean more money for better public services, more profitablecompanies and charities better able to fulfil their charitable purposes.The financial cost of fraud 202125

8. About the authors8. About the authorsJim Gee,Partner and National Head ofForensic Services, Crowe UKVisiting Professor and Chair of the Centre forCounter Fraud Studies, University of PortsmouthJim is a Partner and National Head of Forensic Services at Crowe UK.He is also Visiting Professor at the University of Portsmouth and Chairof the Centre for Counter Fraud Studies (Europe’s leading centre forresearch into fraud and related issues) and Chair of the UK Fraud CostsMeasurement Committee (a cross-sector body) which, each year,develops and publishes the UK Annual Fraud Indicator.During more than 25 years as a forensic specialist, Jim has advised Ministers,Parliamentary Select Committees and the Attorney-General, as well as nationaland multi‑national companies, major public sector organisations and some of themost prominent charities. To date he has worked with clients from 41 countries.Jim specialises in helping organisations to reduce the cost and incidence of fraudthrough strengthening the resilience to fraud of relevant processes and systems.26

Professor Mark Button,Partner Director of the Centre for CounterFraud Studies at the Institute of CriminalJustice Studies, University of PortsmouthMark is Director of the Centre for Counter Fraud Studies at theInstitute of Criminal Justice Studies, University of Portsmouth. Markhas written extensively on counter fraud and private policing issues,publishing many articles, chapters and completing eight books withone forthcoming.Some of Mark’s most significant research projects include leading the researchon behalf of the National Fraud Authority and ACPO on fraud victims; the NuffieldFoundation on alternatives to criminal prosecution, the Department for InternationalDevelopment on fraud measurement, Acromas (AA and Saga) on ‘Cash-for-Crashfraudsters’, the Midlands Fraud Forum and Eversheds on ‘Sanctioning Fraudsters’.Mark has also acted as a consultant for the United Nations Office on Drugs andCrime and on Civilian Private Security Services. Mark also holds the positionof Head of Secretariat of the Counter Fraud Professional Accreditation Board.He is a former director of the Security Institute. Before joining the Universityof Portsmouth he was a Research Assistant to the Rt. Hon. Bruce George MPspecialising in policing, security and home affairs issues. Mark completedhis undergraduate studies at the University of Exeter, his Masters at theUniversity of Warwick and his Doctorate at the London School of Economics.The financial cost of fraud 202127

9. Crowe UK’s Forensic Services and the Centre for Counter Fraud Studies9. Crowe UK’s Forensic Services andthe Centre for Counter Fraud StudiesCrowe’s Forensic Services aredesigned to help clients whatever theproblem, wherever the place. We helpclients to react to an adverse event orto better protect themselves againstsuch events in the future. We havedelivered such services across mostcontinents, and in some of the mostdifficult countries in which to operate.O

the GDP of all but the seven largest global economies. 6 What is the current financial cost of fraud? Global stats Reducing losses by 40% would free up more than Global losses of fraud equate to This equates to 6.05% (2019) 3.89 trillion (2019) 1.55 trillion 4.37 trillion (USD 5.38 trillion) 6.4% of