Malcolm Sparrow, a professor of public management at Harvard’s Kennedy School and the author, most recently, of License to Steal: How Fraud Bleeds America’s Health Care System, has long argued that government agencies underestimate the scale of a fraud epidemic that he believes costs taxpayers hundreds of billions of dollars annually. His work suggests that better fraud control systems can help significantly trim government budgets without impacting services. Such ideas are of peculiar importance in this era of public-sector austerity.
I recently spoke with Sparrow by phone about prospects for improving audits, the value of transparency in stimulus spending and what investigators can learn from chess strategy. Below is an edited transcript of our conversation. —Sasha Abramsky
Do you have a sense of the exact scale of Medicare fraud and, more generally, healthcare fraud today?
No, we don’t. It’s not because we don’t know how—we absolutely do. Basically, the audits they’re using on a random sample are nothing like fraud audits. The difference between a fraud audit and a medical review audit—a medical review audit, you’re taking all the information as if it’s true and testing whether the medical judgment seems appropriate. You can use these techniques to see where judgments are unorthodox or payment rules have not been followed, but almost nothing in these methods tests whether the information you have is true.
These fraud perpetrators know the value of the sweet spot; they aim their claims dead center, right in the middle of medical orthodoxy. It looks kosher every step of the way—except it’s not true. It’s an inflated procedure, or one that wasn’t done, or an entire medical episode is being created and a completely fake diagnosis has been put on a patient’s record. The stories are legion of people getting a Medicare explanation of benefits statement saying “We’ve paid for this operation in Colorado.” And those people have never been in Colorado. And when you complain about it, nobody seems to care. There’s not enough capacity in the system to investigate.
You’ve said that we might have to jack up the investigative resources within Medicare to 2 percent of the total to protect the integrity of the other 98 percent.
The average spending on medical integrity across the system is somewhere between one- and two-tenths of 1 percent of what the program costs. Meanwhile, [there] could be an 8 to 10 percent loss rate, and if you did the measurements it’d probably be much worse than that. We’ve seen some segments go completely rotten—95 percent. What often happens is that everybody knows a segment has gone bad; it might be DME [durable medical equipment] in Florida, or home healthcare in Texas, or community mental health clinics in Florida. For a good long time it was HIV infusion therapy. Then the authorities wake up and make one criminal case against one person, at which point you can sometimes see the whole billing for this procedure drop off by more than 90 percent. This tells you something about the level to which such practices had become widespread before law enforcement paid attention.
Presumably many of the billing scams apply equally to private healthcare systems?
That’s absolutely right.
What sort of crime do you find most interesting?
The kind that are most interesting intellectually are the deliberate and well-orchestrated fraud schemes. They definitely involve conscious opposition; it’s a substitution for my chess-playing life. You should assume the opponents are playing the counterintelligence game and studying carefully what you do—where the normal values of openness, transparency and predictability are, in fact, losing strategies, and where artfulness and creativity count best. Whoever wins this game is the one who knows best about what the opposition is up to and adapts to it. It makes it a dynamic and intelligent game. It’s fascinating.
Regarding your work on the Recovery Independent Advisory Panel, which is monitoring whether federal stimulus spending is being diverted by fraud, waste and abuse: obviously, there are terrific sums of money flowing in all kinds of directions. How do you track down where it has gone wrong?
The stimulus funds are divided into three major categories [tax benefits; entitlements; and contracts, grants and loans]. Contracts, grants and loans is $275 billion. That’s the piece everybody is aware of, because you see the signs up on the highway or the airport. The big deal in terms of government transparency is mostly oriented around that third [piece], because the bulk of the reporting obligations involve private-sector contractors and states and other recipients reporting on a regular basis and the data being available in an easily searchable form on the website. You can look up your ZIP code, and if you don’t see work that is supposed to be being done or it doesn’t look genuine or it’s a useless project, there are reporting mechanisms built into the site. The contractors have to present all this new data, and they can get excluded if they’re persistently noncompliant. It’s a remarkable depth of accountability.
The news to date is fairly positive, and the level of fraud connected with stimulus spending seems fairly small. Are you confident that is, in fact, the case?
I’m not completely confident of that, because I know the nature of the fraud control challenge—and I know that fraud, if perpetrated well, remains invisible.
It seems that under-resourced agencies become magnets for fraud, and because everyone knows they’re magnets for fraud public confidence continues to deteriorate. It’s the opposite of a virtuous circle. How do you break that?
By establishing how much is being lost, which would be a political firestorm but would move the discussion along—to what should we do, what should we tolerate in terms of additional checks and controls, in order to eliminate this cancer.