Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
JAMA. Author manuscript; available in PMC 2013 February 1.
Published in final edited form as:
PMCID: PMC3544169

Potential Consequences of Reforming Medicare into a Competitive Bidding System

Zirui Song, Ph.D.,1,2,3,* David M. Cutler, Ph.D.,2,4 and Michael E. Chernew, Ph.D.1,2


The idea of a premium support (or voucher) system for Medicare has generated substantial debate. Under premium support, Medicare beneficiaries choose among health plans that compete in a market-based bidding system. In some models, the Traditional Medicare (TM) program is abandoned entirely in favor of private health plans. In other models, such as the Ryan-Wyden plan, TM becomes one option among many.

Proponents of premium support cite two potential strengths. First, competition may lower health care spending. Second, by pegging the Medicare contribution to one of the lower-cost plans, and limiting the increase in the government’s contribution over time, public spending on Medicare will slow.1 Critics maintain that bidding is largely a way to shift Medicare costs to beneficiaries by increasing their required premium payments.2

Competitive bidding is not new, even to Medicare. The Medicare Advantage (MA) program has used bidding to determine plan payments since 2006. In MA, plans submit a price (bid) they are willing to accept to insure a beneficiary. Payment is determined by comparing the bid to a “benchmark” payment rate set by Medicare based on the counties the plan serves. If the bid exceeds the benchmark, Medicare pays the plan the benchmark and the plan must collect the difference by charging a premium to enrollees. If the bid undercuts the benchmark, Medicare pays the plan its bid plus 75% of the difference—a “rebate”—which the plan must return to beneficiaries through lower premiums or additional benefits. Bidding low allows plans to attract beneficiaries through higher rebates. Currently, over 90% of MA plans offer some kind of rebate.

Based on the Ryan-Wyden plan, the bidding system proposed in the recent House Republican budget replaces the administratively-set benchmark with a market-determined benchmark.3 In every county, either the plan with the second-lowest bid or TM (whichever is lower) becomes the benchmark. Thus, every beneficiary would have at most one lower-cost option. Any beneficiary choosing a plan (including TM) that bids above the benchmark must pay a premium out-of-pocket, equal to the difference between that plan’s bid and the benchmark.

An estimate of what such a bidding system may mean for Medicare beneficiaries, using 2006–2009 data on MA plan bids and TM costs, is shown in the Table.

Table 1
Table Comparison of the 5 lowest plan bids to traditional Medicare costs, 2006–2009*

Nationally, in 2009, the benchmark plan under Ryan-Wyden bid an average of 9% below TM costs (TM was equivalent to about the tenth-lowest bid). Since TM is simply another plan option under Ryan-Wyden, a beneficiary in 2009 would have paid an average of $64/month (9% of $717) in additional premiums to stay in TM. Across the U.S., 68% of TM beneficiaries in 2009—roughly 24 million beneficiaries—lived in counties where TM spending was greater than the second-least expensive plan and would have paid more to keep their choice of coverage. (This share would have been 81% in 2008, 75% in 2007, and 67% in 2006.) Furthermore, over 90% of MA beneficiaries (roughly 6.6 million seniors, excluding those dually-eligible or in employer plans) would have also paid more for the plan they chose.

Private plans can cost less than TM for several reasons. Private plans may use medical resources more efficiently; they may enroll healthier patients relative to the risk-adjusted payment; or their negotiated prices may not fully reflect the costs of indirect medical education or payments for disadvantaged hospitals, which TM explicitly pays. The magnitudes of efficiency, selection, and avoided add-on payments are unclear; debate over whether add-on payments should be included in the TM amount for bidding purposes is ongoing. To the extent that the 9% cost advantage reflects efficiency, it suggests there are better ways to provide the TM benefit. Indeed, if plans are bidding above their cost of insuring beneficiaries, the 9% gap may underestimate the full efficiency gain.

The reforms to TM enacted in the Affordable Care Act (ACA) may change these estimates. Specifically, the ACA started TM on a path of improved incentives for cost and quality through Accountable Care Organizations, bundled payments, and strengthening primary care.4 The ACA also aims to slow the growth of TM costs by reducing fee increases for hospitals. If TM costs slow but do not close the 9% gap entirely, as currently projected, millions of beneficiaries will still have to pay more—although less than $64/month—to maintain their choice of coverage, assuming the benchmark stays the same. However, if the ACA reduces TM costs enough so that TM becomes the benchmark, beneficiaries would no longer pay more to keep TM; instead, the MA plans would be costlier than TM and require a premium.

These estimates may have potential implications for policymakers. Specifically, if competition or the ACA does not lower the benchmark, a bidding system like that under Ryan-Wyden faces the prospect of millions of Medicare beneficiaries being asked to pay more than they currently pay for the coverage they are currently enrolled in. Our estimates are broadly consistent with a prior analysis,5 although about 2.5 million more beneficiaries in our estimate would have paid more to stay in TM compared to the prior report, which used aggregated estimates of the 25th percentile of bids as the benchmark rather than actual plan bids. Our estimate of the $64 in additional premiums that TM beneficiaries would have to pay may also be higher than in the prior report, although a direct comparison is not possible. For high-income seniors, paying more may not be problematic. For low- and moderate-income seniors, however, $64/month could be very significant. Additional premium support for low-income seniors would help, but likely would not make up the difference.

Moreover, incentivizing beneficiaries to join private Medicare plans—even if those plans are less expensive—may have undesirable effects. In particular, the reliance on beneficiary shopping to discipline the market has been problematic. Beneficiaries are often slow to switch plans due to cognitive impairment, choice overload, consumer inertia, or other influences. For example, Medicare Part D plans, which also operate in a bidding system, have found it profitable to price low initially, attract many enrollees, then increase prices over time.6 Moreover, beneficiaries do not enroll in Part D plans that offer them the best coverage for their premiums and medical conditions.7 These market failures would likely be even greater in an entirely market-based Medicare system, where choosing plans would likely be even more difficult than in Part D. Additionally, the market requires beneficiaries to trade restrictions on care or limited physician and hospital networks for premium contributions, which is counter to how many beneficiaries view the Medicare program today.

Premium support based on competitive bidding offers a potential fiscal solution if ACA reforms fail, but it does so at the cost of making Medicare beneficiaries responsible for solving Medicare’s fiscal crisis. Success of the ACA can make premium support less risky by lowering TM costs and helping to monitor and improve quality in private plans. Without the ACA improvements, beneficiaries must either pay more for TM or join a private plan. Given the current fiscal pressures, this may be acceptable, but it is a major shift from TM that may have deleterious consequences.


The authors are grateful for funding from the National Institute on Aging (P01 AG032952). Dr. Song acknowledges support from a National Institute on Aging Predoctoral M.D./Ph.D. National Research Service Award (F30 AG039175) and a Pre-doctoral Fellowship in Aging and Health Economics from the National Bureau of Economic Research (T32 AG000186).



Zirui Song has no potential conflicts of interest to disclose.

David Cutler served on the Council of Economic Advisers and the National Economic Council during the Clinton Administration and was senior health care advisor to Barack Obama’s Presidential campaign.

Michael Chernew is a commissioner on the Medicare Payment Advisory Commission and a health advisor to the Congressional Budget Office.


1. Antos JR. The Wyden-Ryan proposal--a foundation for realistic Medicare reform. N Engl J Med. 2012 Mar 8;366(10):879–81. [PubMed]
2. Aaron HJ, Frakt AB. Why now is not the time for premium support. N Engl J Med. 2012 Mar 8;366(10):877–9. [PubMed]
3. House Budget Committee. Fiscal Year 2013 Budget Resolution. (
4. Department of Health and Human Services and Centers for Medicare and Medicaid Services. Final rule. U.S. Government Printing Office; 2011. Medicare Program; Medicare Shared Savings Program: Accountable Care Organizations. (
5. Feldman R, Coulam R, Dowd B. Competitive bidding can help solve Medicare’s fiscal crisis. Washington, DC: American Enterprise Institute; 2012. Feb, (
6. Ericson KMM. Market design when firms interact with inertial consumers: evidence from Medicare Part D. 2011 Dec 11; (
7. Abaluck J, Gruber J. Choice inconsistencies among the elderly: evidence from plan choice in the Medicare Part D program. Am Econ Rev. 2011;101:1180–1210. [PMC free article] [PubMed]
8. Centers for Medicare and Medicaid Services. Plan Payment Data. ( [PubMed]