About 90 percent of the data in the world today has been created in the past 2 years alone, according to IBM. Yes, we live in the era of Big Data.
Data is vital to our work as an OIG. We use data analytics – including data mining, risk assessment models, and predictive analytics – to help focus our audits and investigations on high-risk areas of the U.S. Postal Service that yield the largest financial impact and/or efficiency improvements. For our organization, data analytics is a game-changer. Using a single data interface, investigators no longer have to comb through different programs and network folders, saving considerable time. Our predictive model lets us identify cases involving a high probability of fraud, before beginning an investigation.
While the data game is rapidly evolving, federal laws governing data use have moved at a slower pace. The recently enacted DATA Act provides a powerful weapon in combatting fraud and waste in government by standardizing and opening up federal spending information for all to see. But agencies still face bottlenecks in uncovering fraud and abuse. Notably, the Computer Matching and Privacy Protection Act of 1988 – written before Big Data and intended as an extension of the Privacy Act – added procedural steps that agencies must follow when matching federal, state, and local electronic databases.
Say an agency wanted to check its payroll data against the Department of Labor’s (DOL) workers’ compensation records to determine if an individual is collecting both a paycheck and a workers’ compensation check. Under the 1988 law, the requesting agency would need to draft a formal matching agreement to be reviewed by the data integrity boards at both the requesting and responding agencies (in this example, DOL). The complicated process can take 6 months or more, during which time fraud can continue.
The Computer Matching Act was passed at a time when people were unfamiliar with computers and worried about their privacy. Privacy is still a major concern, but is privacy protection inadvertently skewed in favor of criminals? Data analytics allows investigators to root out fraud and abuse early and find those responsible before they can make a long-term habit of it. But the most effective uses of data analytics are often obstructed with administrative hurdles.
What is the right balance between protecting federal employees’ privacy and equipping agencies to quickly detect fraud and abuse? If you accept money from the government – such as a paycheck, disability check, grant award, or contractor payment – should you expect more scrutiny? Would you be willing to share your data to help combat fraud? Or is an overabundance of protection necessary in this age of Big Data?