Lawmakers mull a fraud-focused scorecard for oversight

Lawmakers on the House Oversight Committee are constructing a new “scorecard” to track fraud and improper payments in government. 

Reps. Pete Sessions, R-Texas, and Kweisi Mfume, D-Md. — the top Republican and Democrat, respectively, on the committee’s government operations subcommittee — said during a Tuesday hearing that they’ve been working for about a year on the effort. 

That hearing was supposed to serve as the launch for the assessment, but “it’s become apparent that we simply lack the data to determine whether programs and agencies are getting better,” said Sessions.

Definitions can also get confusing. While lawmakers discussed both improper payments as well as fraud — which can, at times, overlap — the two categories are distinct. 

Improper payments are any payments that shouldn’t have been made or were made in the wrong amount, meaning that the fault may lie with the government, not the recipient of a government benefit. 

The Social Security Administration, for example, has faced public scrutiny for its policies — which it has been changing — of clawing back payments sent to beneficiaries by mistake, some due to the fault of the agency and more due to beneficiaries not reporting information.

Fraud involves wilful misrepresentation. Some government programs, like unemployment insurance, saw upticks in the amount of fraudulent program applications from bad actors exploiting identity theft vectors during the pandemic. 

The oversight committee majority also released a staff report on unemployment insurance fraud during the pandemic on Tuesday.

In April, the Government Accountability Office estimated that governmentwide fraud accounted for $233 billion to $531 billion in annual losses, although the Office of Management and Budget has called the number “unrealistic” and argued that the methodology is flawed. 

“Things have to change,” said Sessions, who promised to continue the focus on the issue in the next Congress. “That includes the way the oversight community — and that is this subcommittee — will look at and track progress not only on federal agencies, but also at the state level, to make our anti-fraud measures amount to more than just a screen door on a submarine, meaning it's just open for fraud.”

Mfume said that the pair would be focusing their oversight scorecard on programs with the highest fraud risk, pointing to the Earned Income Task Credit and Supplemental Security Income as likely examples. 

It’s important to track improper payments and fraud in the programs, he said, “because they've been so successful in keeping our country afloat, particularly during the worst of the COVID-19 pandemic.”

Michael Horowitz — chair of the Pandemic Response Accountability Committee and Inspector General for the Department of Justice — said that the scorecard is an “important step forward.”

He and other witnesses also implored lawmakers to ensure that the data analytics center set up for pandemic oversight in 2021 is extended past its sunset date, scheduled for the end of September next year. 

“Unless Congress takes action, one of the most significant tools to improve program integrity and prevent fraud will be lost,” said Horowitz, touting the work of the Pandemic Analytics Center of Excellence in fraud prevention and recovery.

“It’s already resulted in recoveries that far exceed its operating costs, and it's assisting over 40 law enforcement agencies as they pursue hundreds of fraud cases involving over $2 billion,” he said. 

A similar analytics center was set up to oversee 2009 stimulus spending, but it expired in 2015. 

There is legislation in progress to establish a successor analytics center to replicate the work of the PACE. A Senate bill passed committee and is on the calendar for a vote of the full chamber. The House bill is still in the committee stage. 

The center differs from other government anti-fraud resources, like the Treasury Department’s Do Not Pay service, because of its cross-agency purview and use of machine learning, said Horowitz, who also emphasized the fact that the PRAC has access to law enforcement data because of its authority.

Better data within the agencies that deliver programs to prevent fraud would also be helpful, several witnesses said. 

Agencies could use data to cross check claims and prevent problems before they happen, but sharing data across agencies can be an arduous process and is bound by existing statutory limits on data sharing in government. 

“Agencies can’t or don’t try to access the data that they need to find fraud,” said Linda Miller, co-founder and CEO of the new, fraud-focused Program Integrity Alliance.

It took the PRAC eight months to get a data-sharing agreement with the Social Security Administration to check applicant information coming into the Small Business Administration for programs like the pandemic-era Paycheck Protection Program, said Horowitz, who also cited the cost of getting that information.