The federal government has always focused on Big Data and its usefulness in the business of governance, said a panel of officials in May – what has changed is the availability of computing power to delve into the data for insights.
For example, Joah Iannotta, the Government Accountability Office’s assistant director, Forensic Audits and Investigative Service, pointed out the watchdog agency has always used data to find evidence of problems.
“We use data to try to find red flags, indicators that could help other government agencies identify fraud, waste, and abuse,” Iannotta said. “GAO has special statutory authority to ask agencies to produce data … We can think about what datasets we can bring together that might offer fraud markers.”
One of the newest federal agencies, the Consumer Protection Financial Bureau, is really getting going just as Big Data tools are becoming commonplace.
“I spend a lot of my time working on data sharing with other agencies,” said Linda Powell, Chief Data Officer for CFPB. There are two large data efforts under way at the agency right now, she said. “We have a database of [consumer] financial complaints … We look at every single complaint we get, over 600,000 complaints [in] the past four years.”
The other big data push is in integrating disparate datasets that can help have a big impact on consumers’ lives, she said. For instance, medical debt is unlike any other type of debt, because it can be both unexpected and financially catastrophic. “Incurring medical debt didn’t actually have any correlation to paying other bills, [but] medical debt has as much weight as your mortgage or other debts” on consumers’ credit ratings, Powell said.
“CFPB went to the scoring agencies and made sure they periodically try to clean up” medical debt reports, she said. “Even more fascinating, the credit rating agencies read CFPB’s report that said medical debt shouldn’t have as much impact, [so] they voluntarily changed their FICO algorithms and now medical debt doesn’t carry as much weight.”
Healthcare services is another area ripe for the use of Big Data and analytics, but the need for data governance is critical, said Niall Brennan, CDO and director, Office of Enterprise Data and Analytics, Centers for Medicare and Medicaid Services.
“I am a believer in data and analytics, and the power of [big data] to transform our healthcare system. However, I do it through a healthy sense of pragmatism and cynicism,” Brennan said. “There are too many buzzwords, white elephants [and] magical solutions … To stand up a data governance process that [generates outcomes] is a lot of work.”
Brennan pointed out that CMS has always generated a vast amount of data, and always analyzed it. What is changing – that is, expanding – is what insights the data is used to provide.
“In the past we were more focused on paying claims, not necessarily on analyzing the data to see how the healthcare system was performing,” he said. “So data analysis now is in ‘real time’ … We’re routinely tracking drug spending in real time, especially high-cost specialty drugs, tracking re-admissions in near real time. [W]e’re looking a lot across the continuum of care. And externally we’re making more and more data available so healthcare providers can perform their own analysis and adopt a more patient-centric approach.”
Some resistance to the use of data analytics arises in organizations because of fears that they will show programs falling short in their missions, Iannotta said.
“Everybody’s concerned about the … exposure that analytics can bring to a program [because] of what it could imply,” she said. “Another key issue, especially if you’re trying to marry federal data with private industry data,” is that private data could reveal corporate practices. “That’s another big barrier to data sharing,” she said.
Tony Summerlin, Special Advisor to the CIO at the Federal Communications Commission, said the privacy horse is already out of the barn.
“I can find out almost anything on anyone, anywhere, at any time,” Summerlin said, adding, “Government is inherently rather crappy at it. … I admire that they’re doing their best at keeping [personally identifiable information] out of the public realm, [but] one part of government offers no anonymity, ther other part totally protected … You can’t have it both ways, and we keep trying to have it both ways.”