The hidden key to stopping public sector fraud

Digital identity verification tools should make it easier—not harder—to access benefits and interact with government. 

Done right, digital identity verification can ensure that money flows to the families and people who need assistance and away from bad actors. But done poorly, it can create friction and burdens for the very people that government services are designed to reach. 

One example: recent research from Georgetown University's Beeck Center for Social Impact and Innovation found that nine states require individuals to prove their identities online before even applying for government benefits. Yet, far too many people today aren’t able to pass legacy identity checks, whether due to a lack of credit history or some other reason. These states are making it harder, if not impossible, for such individuals to receive the benefits they are due. 

The stark reality is that, over the past three years, public benefit programs have been ripped off. More than $190 billion in improper payments, much of it fraud, was found in the pandemic unemployment insurance program.  To address this, some see only two paths forward – build in identity verification with high levels of friction, or don’t, and accept a certain amount of fraud as the price of doing business.

But this is a false choice. There are existing approaches that don’t accept fraud or friction as the answer. The truth is leaders in government, in close collaboration with private sector partners, need to know which approaches to digital identity work and which don’t. 

The key is to require transparent and measurable analytical performance of the digital identity platforms that the government chooses. This is something that the financial services industry is already doing with great success. They understand precisely where their friction is, they know the performance of their identity system, and they are constantly looking to improve them. As management consultant and author Peter Drucker famously stated, “If you can’t measure it, you can’t improve it.”

So what should be done? 

The federal government should leverage digital identity providers that offer performance-based measurements and regularly publish their auto-approval rates, with a focus on coverage, accuracy, precision, and equity.  These are the essential metrics that indicate what percentage of people can be identified with a high degree of confidence that they are who they say they are.  Conversely, the federal government should publish performance metrics for government-operated digital identity programs. 

Yet, this type of public reporting is too important to leave to the goodwill of companies, even those that are well-meaning. Today, assessing the performance of identity management tools is a matter of trust. Whether it’s the government or private sector providers, a performance report is not required. And the data that’s shared isn’t validated. Validated performance reporting should be required across the board to give federal and state agencies alike the ability to make meaningful apples-to-apples comparisons of their identity management solutions.

There are multiple levers that government can pull to create a transparent reporting environment for the digital identity ecosystem. Government agencies can insert this requirement in contracts with digital identity providers. Or the Biden Administration could include it in the anticipated executive order on digital identity. NIST could develop standards for performance reporting and the government can leverage third party assessment organizations to validate performance. OMB could update Circular A-11 and require performance data on IT investment be included in the budget request process. 

Performance dashboards drive transparency and accountability. The simple act of reporting will open the door to better outcomes, almost immediately, because leaders will see that data-science driven approaches are far more effective than those rooted in the past.

It will help eliminate fraud and speed the delivery of benefits to those who rely on them.  Despite what some people say, these two goals are not at odds with each other;  they go hand in hand. When government uses the latest in AI/ML and predictive analytics, it increases the ability to distinguish between good consumers and bad actors.  

It will lead to more equitable outcomes.  Credit invisible groups continue to be negatively impacted by traditional consumer identification programs’ primary reliance on credit header data. More modern approaches have been shown to be fairer and lead to better outcomes across the board.

And it will lead to better cooperation at every level of government.  The bad guys know that our government agencies are siloed and they can take advantage of one state after another.  A more modern approach will allow us to track bad actors across state lines and stop them from using the same false or synthetic identities to defraud the government over and over and over again.

Public sector fraud is a plague that will not be defeated overnight. But with transparent measurements and analytics, our nation has the tools to effectively tackle the problem.