Help is needed to learn how much people spend on drugs.

Photographer: Phil Walter/Getty Images

Where the Government Can Get Better Data

Peter R. Orszag is a Bloomberg View columnist. He is a vice chairman of investment banking at Lazard. He was President Barack Obama’s director of the Office of Management and Budget from 2009 to 2010 and the director of the Congressional Budget Office from 2007 to 2008.
Read More.
a | A

As the world grows ever better at gathering and making use of data, the question arises: Might the measurements and statistics that private businesses accumulate be used to help government statistical agencies that are under increasing budget pressure? Two recent studies suggest they can.

One of these, from Express Scripts, an online pharmacy, examined patients with extremely high prescription drug costs. Surprisingly high, really; I'd been aware that drug spending is concentrated, but until I read this study I hadn't realized quite how much. In 2014, the top 5 percent of patients ranked by drug costs accounted for more than 60 percent of total drug spending, the Express Scripts analysis found. Spending on the bottom 80 percent of people was less than 10 percent of the total.

This concentration is even more skewed than total health care spending, and it reflects very high drug costs at the top. Almost 140,000 Americans had drug spending of at least $100,000 on prescription drugs in 2014. One-third of them were being treated for 10 or more conditions, and almost two-thirds were taking 10 or more medications.

The other study, from the health analytics and technology company Inovalon, examined how quality ratings within Medicare Advantage (the private insurance component of Medicare) differed between those beneficiaries who are eligible for both Medicare and Medicaid and those eligible only for Medicare. Plans serving the dual eligibles rank lower in quality than other plans in the program do. Using a database of more than 2 million Medicare Advantage members, Inovalon examined why.

The differences turned out to mainly reflect special risk factors that dual eligibles experience -- including living in neighborhoods with higher poverty rates. The Inovalon researchers suggest that Medicare may want to adjust its quality ratings to account for these factors, to come up with a better apples-to-apples comparison of underlying plan quality.

Such  improved comparisons would have benefits for Medicare and for participants. That said, there is some legitimate concern about government agencies using private-sector sources of data rather than official surveys and administrative data: The private data might be in some ways biased toward the companies' interests or be unrepresentative in other ways. (For the sake of disclosure, both Express Scripts and Inovalon are clients of Citigroup, my employer.) Yet both these studies used large samples that do not appear to be skewed, and they provide new information useful to policy analysts.

The Robert Wood Johnson Foundation (on whose board I serve) is similarly exploring ways of using private data. It has teamed up with athenahealth, a provider of electronic health records and other services, to track the effects of health-care reform. A report from March, for instance, found no evidence that the expansion of health insurance under Obamacare has yet overwhelmed doctors with patient overload, as some had warned.

More than five years ago, Alan Krueger, who was then at the Treasury Department, argued that we should “leverage private sources of economic data to improve our statistical infrastructure.” By “making existing privately collected data more useful for policy makers and policy discussions," Krueger argued, “we could likely enjoy large improvements in our knowledge of the economy at relatively little cost.” It was a good idea then and, as the data pour in from the private sector, it keeps looking smarter.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Peter R. Orszag at

To contact the editor on this story:
Mary Duenwald at