• Office Hours: Competitive Bidding

    Setting: my office last week, talking with an undergrad in my Intro to the U.S. Health System course.

    Student: Have I got this straight? Everyone dies. Age is the strongest predictor of death and Medicare covers around 8 in 10 persons who die in a given year. Health care costs rise near death since people are sick. And within Medicare, a small proportion of beneficiaries account for a large fraction of total spending. Correct?

    Me: Yes. But there is no guarantee a low user will remain low the next year. And since everyone dies, most eventually shift into the high use camp.

    Student: You said that recent work that you had read has made you be more open-minded about competitive bidding in Medicare. This would mean private insurance companies covering some of the elderly, correct?

    Me: Yes.

    Student: We don’t make anyone choose private Medicare plans, right?

    Me: Correct, it is their choice. And private plans are only available in areas that insurance companies desire to offer such coverage.

    Student: It seems to me that the only way private insurance companies would be interested in Medicare business is if they can somehow only sign up younger and healthier beneficiaries and hope they go back to regular Medicare before they get very sick and die. If we have private options, it seems like the patient and the company should have to agree to stick together until the end.


    I stuttered and stammered a bit on this one, in part because the insight of the student surprised me, but also because I have been trying to convince myself that competitive bidding can work. Two years ago, I really thought the Medicare Advantage program should be abolished even though I didn’t say it clearly here, because I thought it was nothing but cherry picking/selection. As I have said before, Austin has done lots to convince me it is possible for competitive bidding to work with private insurance alongside a government Medicare option. I still believe that it is possible in theory.

    However, in some basic “gut” way it just doesn’t make any sense that a private insurance company would be interested in seeking great market share among the elderly if there was a truly competitive situation brought about. My recent writing about the Federal Flood Insurance has helped to crystallize these thoughts. Sometimes there may just be risks that only government can undertake, and the only question is whether the benefits of the government doing so outweigh the costs. My head believes competitive bidding can work in theory, but my heart/gut is not so sure in practice.

    Update: nice post from Austin this am; I would add that my heart/gut reticence is based on ability of private to reduce costs. Maybe wrong metric. Want to see the below study….

    What I would like to see: a long term case control study (10 years!) starting with a representative sample at age 65 that tracks a cohort of people moving in and out of Medicare Advantage and FFS Medicare (and death) as well as comparing patients by overall spending (say by quintiles). I don’t think such a study has been possible because Medicare claims don’t exist (publicly?) for those in Advantage plans. We need such a study.

    update: from Austin via email ” The point is not that competitive bidding certainly beats a public plan (though it may). The point is that it is a way to economize within the current political equilibrium of a public-private hybrid Medicare. It’s better than that status quo and has elements that should appeal to the left and right, as I’ve explained in posts.” Let me add that I personally believe that a political deal on some way forward on health reform is what our country most needs to have any hope of developing a sustainable health care system and federal budget; I suspect this could be part of such a deal and noted that in my book.

    • Hi Don-
      Yes, an interesting study idea. Although CMS would have to conduct it, as you are correct – the Medicare Advantage claims are not publically available in their entirety at this time. Your student is sharp and on the right track! Hope all is well-

      Cordt T. Kassner, PhD
      Principal, Hospice Analytics

      • Hey Cordt. Do you know if private plans give claims to CMS? I know they are not public, but once CMS sends premiums, is that the end of the sharing or does CMS have the claims to actually do this study retrospectively?

        • Hi Don-
          I checked with ResDAC about this, and it appears CMS has Medicare Advantage claims similar to the LDS and RIF claim file structure – so CMS staff or contractors would have the ability to conduct this study. Addtional information about this can be found at . Hope all is well-

          • @Cordt Kassner
            interesting. Will nose around and talk to a few people. Would like to see an intake cohort of age eligible Medicare folks move in and out of Med Advantage for a longish follow up period

    • Why doesn’t the risk adjustment for Medicare Advantage plans address this concern?

      • @SL
        not sure which problem you mean. Risk adjustment of MA plans can account for differences across patients, but the basic problem that competitive bidding would fix is how the basic premium is set: currently it is administratively set, whereas in comp bidding it would be done based on bids from plans to deliver a set of benefits. Each plan would then only get this amount of premium and if they charged more they would have to find consumers will to pay more. This is the simplest explanation I have seen from Austin’s excellent FAQ on the topic http://theincidentaleconomist.com/wordpress/medicare-advantage-payment-vignettes/ Here is link to full FAQ http://theincidentaleconomist.com/wordpress/competitive-bidding-faq/

        • I was (ineptly) trying to suggest that the student’s question might not be a huge concern if the administratively set risk adjustment is not very inaccurate.What I am asking then is how imperfect is that risk adjustment? As a theoretical matter, it should be worse than the adjustment arising from a comp bid, but how much worse? Is there any empirical evidence to suggest that we see any adverse selection in MA? Is the answer just that would just need the study you would like to see done?

          • @SL
            two issues. First, risk adjustment (moving premium up or down due to patient characteristics is getting better and could continue to do so). Austin or Aaaron linked to a doc on this in the last day or so in one of these comment strings. Second, competititve bidding is about how the “base” premium for a given area is set. Historically has been administrative pricing. The discussion has been about moving it toward competitive bidding, with base set via competition among insurance providers to deliver a defined package of benefits in a given area. I would like to see a long term follow up of movement in and out of MedAdv and FFS Medicare, looking at costs and quality to fully see what impacts competitive bidding could have.