Medicare Advantage as a form of managed competition

Can health insurance competition work? Evidence from Medicare Advantage

By Vilsa Curto, Liran Einav, Jonathan Levin, Jay Bhattacharya
National Bureau of Economic Research, December 2014, NBER Working Paper 20818


We estimate the economic surplus created by Medicare Advantage under its reformed competitive bidding rules. We use data on the universe of Medicare beneficiaries, and develop a model of plan bidding that accounts for both market power and risk selection. We find that private plans have costs around 12% below fee-for-service costs, and generate around $50 dollars in surplus on average per enrollee-month, after accounting for the disutility due to enrollees having more limited choice of providers. Taxpayers provide a large additional subsidy, and insurers capture most of the private gains. We use the model to evaluate possible program changes.

From the Introduction

Introducing managed competition into healthcare has been an alluring idea to economists and policy-makers. Proponents argue that effectively designed market mechanisms can avoid the inefficiencies of an administrative price system.

The final part of the paper combines our estimates to calculate welfare effects from Medicare Advantage. A useful way to think about this is in terms of the total (dollar) surplus created by a private plan enrollment, and then its division among the plan, the enrollee and taxpayers. As mentioned, MA plans achieved cost savings of 77 dollars in providing the standard FFS insurance benefit. Taxpayers captured none of this. Instead, they provided an additional subsidy of around 94 dollars per month relative to government costs under fee-for-service Medicare. The beneficiaries of these 171 dollars (cost savings of 77 plus subsidy of 94) were insurers and consumers. We estimate that insurers captured 95 dollars per enrollee-month (not accounting for any fixed administrative costs). MA enrollees gained the rest, but also incurred a disutility of 27 dollars from having a limited network of providers, so their net surplus was lower, 50 dollars (per enrollee-month). In short, we estimate fairly substantial efficiency benefits, but an even larger net cost to taxpayers.


By Don McCanne, MD

This technical paper is not worth reading. It was apparently written with the intent of proving that private Medicare Advantage plans show that health insurance competition works. Numerous assumptions were made while largely ignoring the gaming of insurers through favorable risk selection and upcoding for risk adjustment. With all of the authors’ efforts to prove the efficiencies of managed competition through private Medicare Advantage plans, they still showed that this program resulted in “an even larger net cost to taxpayers.”

We need to lay managed competition to rest. Single payer, now.