Data’s Role in Measuring Program Outcomes and Preventing Fraud, Waste and Abuse
istockphoto.com/designer491
Connecting state and local government leaders
COMMENTARY | The American Rescue Plan has shown a spotlight on government use of funds. Here are strategies state and local governments can use to ensure their programs generate positive community results.
On March 11, President Joe Biden signed a coronavirus relief package, called the American Rescue Plan Act, which includes $350 billion in aid to state and local governments. Two months later, the U.S. Treasury Department released guidelines on how that aid can be used. The guidelines are intentionally broad. As the Brookings Institute put it, they “retain ample room for creativity and the ability to address a variety of specific local needs.”
As state and local governments make decisions about how to spend their aid, they will be under scrutiny from constituents. New Orleans, for one, is using sentiment analysis to better understand how residents want the federal funding to be spent. But preventing fraud, waste and abuse is important too.
Going forward, state and local governments must be transparent and communicate with constituents about how relief funds are being spent and why. This communication is about more than sharing dollar figures; it requires showing how the money is driving positive community change. That isn’t possible without a strong foundation in data and analytics.
Laying the Foundation
When programs are inefficient and don’t drive outcomes, there is usually fraud, waste or abuse at the core. While fraud and abuse are malicious, waste can be unintentional, arising when
agencies have repetitive goals or programs are poorly defined. Investing in data integrity and sharing can help agencies combat wasteful spending. Without interagency data sharing, it’s almost impossible to pinpoint redundancies or programs that aren’t performing. Performance must be measured in outcomes; if a program does what it was designed to do but does not positively impact the community long term, that’s wasteful as well. From addiction treatment to education, it’s important to know which programs are truly effective.
Unfortunately, many programs are evaluated based only on their data, making it difficult to detect waste across programs. By centralizing the data and integrating it, agencies can get a holistic view of spending and outcomes. Agencies often cite privacy concerns as a key hurdle to data sharing, but those can be alleviated through the creation of data governance frameworks—an agreed-upon set of rules that govern data management, including cataloging, storage, security and access. With this legally binding framework in place, organizations can trust their data will be shared and managed appropriately.
Forming Data to Outcomes
After investing in data integrity and sharing, it’s time to layer in analytics. Once again, agencies need data and analytics to catch inaccuracies and pinpoint fraud. But those same datasets can also be used to demonstrate effective budgeting and use of funds. Creating a data strategy can help ensure that measurement and analysis is outcomes oriented. Consider unemployment claims.
After weeding out fraudulent activity, an important short-term measurement might be the number of checks mailed and claims responded to. But that must be complemented with a long-term measurement system to track how those successes create lasting impact. A data strategy will help agencies determine what those key metrics are and devise measurement dashboards and data analytics to capture them.
Sometimes, government agencies struggle with analytics because they aren’t used to handling large quantities of integrated data. Machine learning can help. Machine learning uses algorithms and models to analyze data, identify patterns, and predict outcomes—an indispensable tool for going from data integrity to insights. With machine learning, agencies can use huge amounts of data, as long as it’s of the right quality, to identify areas of fraud and duplicative programs. Meanwhile, reinforcement machine learning is particularly useful for assessing programs. As the system learns, it will eventually be able to predict which actions will lead to the desired outcomes.
The Bottom Line
Governments are under increased pressure to justify expenditures and to use taxpayer dollars more efficiently. The American Rescue Act is a huge opportunity for state and local governments—and not just because there are billions of dollars at stake. With this relief package, governments have the chance to use data to both drive investments and prove outcomes, both of which are crucial to building community trust. With a strong foundation in data and analytics, agencies can track and measure how their funds are being used, target programs with proven outcomes, and better serve the public.
Clarke Allen is senior director of Business Development and Strategic Partnerships at Qlarion, a GCOM company, specializing in data and analytics to drive government innovation.
NEXT STORY: Sharpening the focus for smartphone-based evidence