Real-time reporting on bank bailout funds and stimulus spending needed

The federal government can save billions of dollars in fraudulent and wasteful payments by changing the way it collects and stores data submitted by companies receiving stimulus and bailout funds, a panel of technology specialists said on Tuesday.

As much as 2 percent of the money distributed through the Treasury Department's Troubled Asset Relief Program to shore up failing banks could be lost to waste, fraud and abuse, said Stephen Horne, vice president of master data management and integration services at Dow Jones.

Horne, who appeared on a panel hosted by the technology firm Teradata Corp. at the National Press Club on the role of technology in government, said the amount of fraud is relatively low given the program's size -- $700 billion. Still, taxpayers are losing $14 billion, some of which could be prevented by examining company filings in real-time.

He criticized the current audit-based approach to oversight that TARP and the Recovery Act board use, a process Horne compared to an autopsy. Under current practices, auditors find abuses well after they have been committed and funds have been spent. If bailout money and stimulus dollars are tracked in real-time, Horne says abuses could be caught when they occur, making it easier to recover the money or to stop payment.

Horne said the original Economic Stabilization Act, which created the TARP program, was written so quickly that lawmakers didn't have time to include the relevant transparency and accountability measures. As a result, banks receiving stimulus funds have filed thousands of paper documents with the Securities and Exchange Commission and other regulatory agencies, but it's all but impossible to track what has happened to the bailout funds.

Rep. Carolyn Maloney, D-N.Y., introduced a H.R. 1242, bill in March that would require banks and other bailout-fund recipients to report transactions to a government database as close to a real-time basis as possible. The bill has been referred to the House Financial Services Committee, where Horne said Dow Jones and other financial data companies have been asked to provide technical input and guidance.

Horne has been pressing for real-time reporting from TARP recipients since the program was introduced last fall, including writing letters to former Treasury Secretary Henry Paulson, urging the department to adopt real-time reporting.

"We could have solved the problem and tracked these funds," Horne said, adding that Maloney's bill would attempt to return some of the control of the funds from Treasury to Congress and establish a structured format through which companies can submit their data. If done appropriately, he said officials will understand how best to structure a similar program in the future.

Another technology that holds promise for saving the government billions of dollars is electronic health records, according to author Stephen Baker. He predicted that if the adoption rate of electronic health records increases, electronic records would lower the cost of health care by avoiding medical errors and reducing unnecessary procedures.

But Stephen Brobst, chief technology officer for Teradata, argued that the technology would not immediately produce the benefits that have been touted. "Electronic health records are not a panacea," he said. "It's about capturing the data in a meaningful way. If you look at them, electronic records are still fairly opaque. We need a lot more work to make use of the unstructured data."

Brobst said the benefit of e-health records is found in the vast amount of data that insurers could collect on patient behavior and outcomes to track anomalies and predict suspicious claims. Real-time data analysis would make it easier to detect fraudulent providers because these clinicians tend to file numerous bogus claims that a computer could detect.

Panel members agreed that privacy presents some obstacles to wider use of technology, especially because electronic health records and even the Smart Grid for electricity, could yield data on personal behaviors that would be useful, though potentially invasive.