In April 2014, the Centers for Medicare and Medicaid Services (CMS) released data for payments Medicare made to doctors and outpatient providers nationally in 2012. The release touched off a series of discussions that focused on outlier physicians, have led to investigations of care patterns and calls for caution in using such data, all prompting CMS officials to defend their release.


It is unclear what the impact of such a release will be, but much of the public doesn't understand that the unprecedented data release excludes the payment data for three in 10 Medicare beneficiaries — those enrolled in Medicare's private insurance option, Medicare Advantage (MA). Even more surprising, Medicare Advantage data are not readily available to CMS, or researchers, hindering the public's understanding of this program.

Medicare Advantage is the current incarnation of a private insurance option that has been available to beneficiaries in one form or another for over three decades. The premiums paid by Medicare to private plans on behalf of beneficiaries are determined administratively, based in part on historical spending patterns where the patient lives, and in part on patient characteristics. Private plan payment data for the care of beneficiaries has never been readily available to researchers, or even to CMS, unlike the provider claims data from traditional Medicare, which is readily available in many forms, including the recent public release. There are proposed federal rule changes that could increase MA data availability, and some private insurance companies who offer MA plans are discussing private data-sharing options.

Beginning in the 1980s, Medicare paid private insurance companies 95 percent of the average historical cost that Medicare incurred for care in a given county, adjusted for age and sex. The initial logic of MA was to save money for Medicare. If Medicare paid 95 percent of the average cost of care as a premium on behalf of a beneficiary, with the private plan then responsible for paying for their care, then Medicare would save 5 percent if beneficiaries choosing MA were similar to those in traditional Medicare. However, persons choosing MA have been healthier and therefore less expensive than those remaining in traditional Medicare across the many reincarnations of the program, meaning the program has always overpaid for care.

The Medicare Modernization Act of 2003 greatly increased the premium payments to private plans for beneficiaries who choose to enroll in MA — in some cases up to 118 percent of the average cost of patients in traditional Medicare. The goal of the change was to increase the proportion of beneficiaries in private Medicare options, and it worked. In 1990, only one in four rural beneficiaries had a choice of MA plans, while around 85 percent of urban dwellers did. Currently, almost all beneficiaries have at least one MA option, and most have several.

Around 30 percent of Medicare beneficiaries are now covered by MA plans, and the proportion choosing such plans has been increasing the past few years as the baby boomers have begun moving into Medicare. It is reasonable to expect the Affordable Care Act will continue this trend toward MA, as people gain experience with exchange based health insurance.

Detractors of MA note that it has always overpaid for care, even more so the past decade, and they support planned cuts to MA plans imposed by the Affordable Care Act that have been delayed. Proponents note that the benefit packages of MA plans are more generous than that of traditional Medicare, and oppose these cuts. Research shows that between 14 and 20 cents of each dollar of overpayment to MA plans results in more expansive benefits for beneficiaries, with the rest mostly flowing to insurance companies.

Are overpayments of these magnitudes to MA plans worth it?

There are normative aspects to this question — a preference for private versus governmental insurance, for example. No evidence is likely to change anyone's view on that aspect of the question. However, there are many important questions about MA as compared to traditional Medicare for which the answer is unclear, and vitally important if we are to improve policy. What we need to answer these questions are the data.

Taylor is associate professor of public policy at Duke University. He blogs at