Skip to content
Risk Fraud & Compliance

Forum: A better way to fight fraud with data analytics

Jon Coss  Vice President / Risk, Fraud & Compliance / Thomson Reuters

· 6 minute read

Jon Coss  Vice President / Risk, Fraud & Compliance / Thomson Reuters

· 6 minute read

For criminals who specialize in defrauding the government, 2020 was a dream come true

In order to keep the US economy afloat during the COVID-19 pandemic, more than $4 trillion was pumped into the economy through various government programs. The money had to be distributed as quickly as possible, so in just a matter of weeks, pandemic stimulus checks were sent, new emergency loan programs such as the Paycheck Protection Program (PPP) and others started processing loans by the thousands and applications for unemployment compensation skyrocketed.


Check out all the articles in the latest issue of Forum magazine.


Fraudsters know a good opportunity when they see one. Almost instantaneously, fictitious businesses began applying for PPP loans; bogus websites started selling fake or nonexistent Personal Protective Equipment; and thousands upon thousands of people tried to collect unemployment checks in any way they could – through synthetic identities, stolen Social Security numbers, shell company shenanigans and more.

In 2020, the US Department of Justice reclaimed $360 million in fraud related to the Coronavirus Aid, Relief, and Economic Security (CARES) Act, but hundreds of additional cases are currently under investigation, and the Small Business Administration’s Inspector General (SBA IG) has reported receiving tens of thousands of fraud tips. By comparison, the total number of fraud tips the SBA IG received in 2019 was 800.

Considering that the federal government is preparing to inject another $1.9 trillion into the US economy in 2021, it is perhaps unsurprising that Thomson Reuters “2021 Government Fraud, Waste, and Abuse Study” found that 93% of frontline government administrators at the state and local level expect fraud in 2021 to be the same or worse as in 2020.

“The same” in this case may not sound all that bad, but it is really quite alarming. Indeed, it would mean that government employees will continue to be overwhelmed and have insufficient resources and inadequate technology to do their jobs. Further, billions of taxpayer dollars will land in the hands of criminals worldwide, and thousands of American citizens in need of assistance will not receive it.

A “worse” scenario is hard to imagine, but more than half of government employees surveyed in the report said that’s precisely what they expect in 2021.

Attack of the cyber-gangs

Government administrators often get a bad rap, but to be fair, they faced an almost impossible task last year. When the shutdowns began in March 2020, more than 32.5 million people filed for unemployment within the first two months of the pandemic, and administrators handling those claims were processing an average of 10 to 15 times their normal workload. In order to distribute unemployment funds as quickly as possible, controls had to be relaxed, which – not surprisingly – created a golden opportunity for fraudsters.

Even in the best of times, however, most government agencies are using outdated technology to fight a class of criminals that is getting smarter and more technologically sophisticated by the day. Almost two-thirds of government employees said their most pressing issues were budget constraints and lack of resources, the report noted. Only half of those respondents said they felt confident they had adequate resources to fight fraud, waste and abuse, and 45% said their primary tool for screening new vendors and contractors was a simple Google® search.


A 2018 ACFE report estimated that risk-assessment programs using proactive data analytics had reduced fraud losses by more than 50%.


Clearly, government administrators do the best they can with the resources they have – but most of the technology available to them is incapable of detecting or preventing fraud before a great deal of money already has been stolen. Indeed, most fraud-prevention efforts are focused on investigations after a suspected fraud scheme has been uncovered and the stolen funds are long gone.

The power of data analytics

Most government agencies rely on anonymous tips, whistleblowers and alert administrators to identify suspicious patterns of behavior that may indicate fraud. The problem with that approach is that program administrators don’t necessarily know what to look for, don’t have enough time to devote to fraud prevention and aren’t technologically equipped to identify anything but the most common and egregious indications of fraud.

Behavioral analytics, a discipline that can be applied very effectively to identify patterns of fraudulent behavior that might otherwise go undetected, is a growing specialty among fraud investigators. For example, when sophisticated fraudsters target unemployment benefits, they don’t create one or two bogus applications – they flood the system with hundreds or thousands of fake applications in the hope that someone will slip up and send them money. But if, say, 6,000 applications use the same address or P.O. box (which is not uncommon), a system equipped with the proper analytical tools can easily flag the anomaly and send an alert.

Take another example: Suppose it takes an average of eight minutes to fill out an unemployment application online, and a flood of applicants suddenly start filling out the form in 30 seconds. Again, behavioral analytics can identify and alert authorities to such a deviation from the norm well before any human observer might catch on.

In actual practice, properly programmed data-mining software can identify all sorts of suspicious or anomalous patterns that an overworked, under-equipped administrator might never detect. Further, the software can run 24/7 in the background or be integrated with a case-management system to identify trends and patterns affecting the program, providing an extra layer of security. Unfortunately, almost two-thirds of state and local government employees don’t use any kind of matter-management software or analytics to help them prevent fraud, according to the report.

The question is: Why?

A more cost-effective option: preventive analytics

It’s no secret that preventive fraud measures are more cost-effective than after-the-fact investigation. For years, the US Government Accountability Office as well as industry groups like the Association of Certified Fraud Examiners (ACFE) have been issuing reports on best practices for fraud risk management, where the benefits of using proactive analytics as part of a dedicated fraud-prevention program are almost always highlighted. In fact, a 2018 ACFE report estimated that risk-assessment programs using proactive data analytics had reduced fraud losses by more than 50%.

Given that the US Treasury may end up pumping more than $2 trillion into the economy in 2021, and fraudsters already have had a year to test vulnerabilities in current government systems, it’s a fair bet that billions more will be lost in the pandemic shuffle.

A modest investment in data analytics could save some of this loss. It could also give government administrators on the front lines of the pandemic crisis a fighting chance against criminal opportunists who view government benefit programs as their own personal ATM.


You can download the Thomson Reuters 2021 Government Fraud, Waste, and Abuse Study here.

More insights