University of Virginia-Led Pandemic Relief Efforts Help Prove Need For Stricter Applications

When it’s necessary for the government to come to the rescue during a pandemic or following a natural disaster, stricter application requirements can reduce potential fraud, according to research from the University of Virginia’s Karsh Institute of Democracy.

Karsh’s Corruption Lab on Ethics, Accountability, and the Rule of Law, known as the CLEAR Lab, researched the federal Paycheck Protection Program, the U.S. government’s unprecedented COVID-19 pandemic rescue plan.

The lab’s working paper was published June 19 by the National Bureau of Economic Research.

A centerpiece of pandemic relief efforts for the U.S. economy, the program offered forgivable federal loans to small firms to support payrolls and avoid mass layoffs during the pandemic. The urgency of the crisis caused the program’s rollout to be rushed with weaker documentation requirements for the applications.

In the program’s second phase, however, officials implemented stricter documentation requirements for applicants.

“Strategic evasion was systemic” in the first wave, said Daniel Gingerich, a professor in the UVA Department of Politics who co-directs the CLEAR Lab. “But we found that a little bit of screening can go a long way in reducing fraud.”

Karsh researchers used the new rules in the program’s second phase to compare applications in both phases. The same firms were applying for federal relief in both phases so the researchers had the opportunity to analyze the impact of stronger application requirements on the same set of borrowers.

CLEAR Lab researchers concluded that the stricter requirement reduced the total amount of potential fraudulent claims by at least $744 million. Savings, according to the team, would have been almost twice as much, at $1.5 billion, had the requirement been in place from the start.

At the start of the pandemic, the Paycheck Protection Program was set up with minimal application requirements to avoid placing a burden on small businesses in legitimate need of timely help. CLEAR Lab research showed that easily accessible documentation, such as tax documents and transcripts of employee payrolls, are highly effective at deterring fraud without creating an onerous burden for applicants.

“The fraud deterrent created by the requirement was large relative to its administrative burden,” the researchers wrote.

They found distinctive changes in behavior among applicants who applied in phase one and were subject to the screening requirement in phase two. Some of those applicants engaged in “bunching,” research showed.

“Borrowers strategically set their loan requests just below the threshold to avoid having to submit the documentation,” the researchers wrote, noting that the behavior is also found in tax reporting, money laundering compliance and other threshold-based programs.

“The bunched groups are a good place to start looking for fraud,” Gingerich said. “It’s not a smoking gun for any particular firm. But at the aggregate level, these groups are absolutely a smoking gun for evasion.”

The lessons learned from this research are valuable on a global scale going forward, as countries around the world have ever-expanding emergency relief portfolios due to floods, fires, droughts and other disasters. When governments need to issue large-scale funding in short order, finding the right balance of accessibility and fraud-reduction can save billions of dollars.

The authors note that tighter screening, such as requiring applicants to submit bank statements or documents that they would readily have on hand, would not be a burden. They noted that the original documentation requirements for the first phase of the Paycheck Protection Program were less strict than requirements for some social assistance programs.

“It’s painfully ironic that screenings for food assistance programs are more demanding than for businesses asking for $2 million loans,” Gingerich said. “Why is it that when businesses ask for something, we take them at their word?”

This research, and its quick turnaround for results, is part of the Karsh Institute’s agile model for pursuing timely research. The group of researchers were already Lab colleagues working on corruption issues in India (Sandip Sukhtankar, Department of Economics), Pakistan (Shan Aman-Rana, Department of Economics), and Brazil (Gingerich). When COVID-19 made their field work impossible, the lab’s remote capabilities allowed the research team to quickly pivot in order to jointly analyze the federal loan program.

“It was really helpful to work with the Karsh Institute team,” Gingerich said. “We were able to move funding quickly from other work, hiring a full-time data journalist and another data analytics expert. We’re all scholars in corruption, and we felt it was completely within our wheelhouse – and frankly, our responsibility – to look at a program that might have historic levels of fraud. This is exactly what we should be working on.”