Wednesday, October 3, 2012

Notes on a Healthcare Talk

Yesterday at the Ross School of Business, there was a panel discussion on health care in the United States. Jonathan Gruber and  Katherine Baicker, two well known health economists, in addition to David Leonhardt, a New York Times reporter who has reported extensively the health care debate, gave their perspectives on health care in the United States.

What I found was quite interesting was how much of a consensus there was among health care economists. Yes, the health care market fails for the 46 million uninsured Americans. No, providing more health care for people doesn't reduce spending on health care. But yes, there are major tax and market reforms that the government can do to increase the value of health care delivered in the United States.

Gruber was the first speaker, and I found his defense of the ACA quite witty. He characterized health care reform as a three legged stool that depends on three parts -- coverage expansion, individual mandate, and subsidies -- to fully function. Because pre-existing conditions can cause people to be kicked off the rolls when they need it most, it's important to reform coverage laws to allow more people to buy insurance. But then if there's no individual mandate, the market is subject to adverse selection. It is, as Gruber joked, "like forcing sports bookies to take bets at half-time". And to make sure that people can afford to purchase insurance, the government has to provide subsidies to increase affordability.

Something else that Gruber emphasized during the talk was that the health care debate is highly nuanced, and that talk of reform shouldn't be on partisan lines. The Affordable Care Act is not a government takeover of health care -- it merely creates a new market in which private insurers can innovate and thrive. Nor is it a federal takeover of health care delivery -- states are allowed large levels of discretion in how they implement their health care exchanges. Something that both Gruber and Baicker seemed excited about was the possible policy innovations the ACA may stimulate. The ACA, while it is certain in the coverage expansion, is much less certain on how to control costs. It is hoped that the variety of reforms -- capping the employer tax deduction, pilot programs for alternative health care delivery, and comparative effectiveness research -- can eventually discover something that the market will find valuable and later deploy.

We already see some of this innovation in the reform of fee-for-service payments. For two years, Blue Cross and Blue Shield of Massachusetts has already been trying out an alternative payment structure that focuses on patient wellness rather than the number of office visits. And then in the past week, Medicare took its first step away from fee-for-service by penalizing readmission and rewarding hospitals that deliver better care with higher payments. Later during the Q and A session, Baicker mentioned other possibilities, such as integrated health care units or even simple emails from your doctor, in order to improve health care effectiveness.

Gruber and Baicker were both optimistic that these kinds of reforms would be critical in bringing costs down. Baicker often mentioned that increasing coverage is easy -- the question is how to control costs. Yet, for both Gruber and Baicker, this ignorance about cost control was just a better reason to expand coverage first. We know how to expand coverage, we don't know how to expand costs. So, as Leonhardt noted, let's expand coverage to save lives now, and use what we learn from the new practices in the health care market to try to reduce costs on a national level.

I find this approach eminently sensible. As mentioned above, reform is critical to creating a larger market for health care, and it seems to be a natural result that the resulting market will have a greater variety of business models and payment structure innovations. Given that ideas can quickly spread, the efficiency of the system will ultimately be determined by the efficiency of the best, not the average. As such, increasing the number of health care consumers seems to be the first step towards discovering a solution to rising health care costs. Gruber also argued that this would help solve the primary care doctor shortage by changing the incentives for medical students. By changing the way insurers reimburse for services, this would reduce the demand for specialty medicine, and thus push more doctors out of specialty disciplines and into primary care.

Baicker made a statement that I feel is very important in framing discussions in health economics: "Health ≠ Health Care ≠ Health Insurance". From all the above analysis, we know this to be true. Just because something is a health care service, such as proton beam therapy, doesn't mean it improves health. And just because something improves health, such as phone calls from a physician, doesn't mean we think of it as "health care". And to properly fund all of this, we need a robust system of insurance that requires incentives and cost structures that are not directly related to a person's biological condition, but nonetheless change the way health care is delivered. The fact that this is true means that when we move forward with the health care system, we need to consider each of those three related, but distinct, parts.


  1. Hi Yichuan! Just read this post for the first time - interested to hear you comment on broad consensus, as a freshman I trust you to portray something like that without bias.

    I'm less pleased with your apparent endorsement of the notion that expanding coverage, especially under a scheme that reinforces the disconnect between provision and marginal cost to the consumer, is a smart and obvious first step to take before settling down to tackle cost-control. At any rate that needs some thought and consideration before you just slip it in like that. As Arnold Kling says, our system provides not so much insurance but insulation. (I recommend you take a look at all his writings on the subject, at EconLog and elsewhere - he did an EconTalk as well). Similarly Gruber's description of his preferred world sounds more like in-kind redistribution (at best) than actual insurance per se. David Henderson repeatedly emphasises that, contra Arrow (see also Avik Roy) the market has done a sufficient job of overcoming all the market-failures proper when allowed to try its own antibodies, such as price discrimination. Also I would have liked to see a shout out to Robin Hanson on the notion that most healthcare doesn't actually make us healthier. Not to mention some serious questioning (regardless of what you conclude) of the notion that healthcare should generally be bought as insurance; see John Cochrane's blog and Milton Friedman's essay "How to Fix Healthcare" for Hoover. I would love to see a reasoned response to these perspectives juxtaposed with the conventional wisdom of the profession as you have access to it at your university.

    1. Also, was that just a NYT-side consensus or a broad cross-partisan one, do you think?

    2. According to Gruber, it seemed to be a consensus across health care economists. Could you help me find some of those specific EconLog links? I'm browsing through a few of them:

      And I find them to be filled with an incredible amount of non-sequiturs. This might be an interesting post to write about in the future though.

    3. Hmmm, I agree that "Market Failure or Market Success" is a non-sequitur; just because the market correctly fails to undertake the prohibitive cost of ameliorating its own failure, doesn't mean it didn't fail in the first place to perform the task of efficiently matching buyers and sellers. But that isn't the post I was takling about; Rosen and Gayer are pointing out that markets can't ameliorate their failure to provide generalized poverty-insurance, not that they aren't quite successful in providing health-insurance when free to.

      The Bryan Caplan posts, however, still seem excellent to me (as usual). In "A Closer Look at Adverse Selection and Mandatory Insurance", he highlights the fundamental point that gets lost in the public discussions: the supposed market failure of adverse selection in health insurance is not, contrary to what most people think, that sick people don't get insured. The market failure is that healthy people won't get insured! Mandating insurance fixes the problem, not by taking money from certainly healthy people to "insure" certainly sick people, but rather to make sure that healthy people can "sell" their coverage to insurers along with the "lemons", so that all possible consumer surplus is realized. And in order to do so, the policy must simultaneously transfer welfare from the healthy to the sick, with a net loss to the former and net gain to the latter. The transfer itself is considered welfare-neutral, so the net effect on society is positive. (Cue deep discussion on philosophy of welfare economics.) Taking the transfer as a given, both groups now realize surpluses from buying insurance for themselves.

      But if, as in reality, markets can correct the information asymmetry on their own and sell separately to each group, then the economic argument for mandates is that they are a cheaper way of repairing the market failure. That's the real debate, where I am open to persuasion. Otherwise mandates are simply redistribution under the guise of market-repair, and in an honest world would be presented (and judged) as such.

      And I agree with everything he says in the other post. So if there were any logical non-sequiturs there that I missed, then I must be really stupid, or worse, really biased. The only one I can think of is that Matt Bandyck accuses markets of failing sick poor people, and then just sick people, and Bryan only answers the former. I disagree that markets "underserve" sick people in general (see above). For more on the guaranteed renewables idea he brings up at the end of the post, see John Cochrane's pieces, available from his blog.

      I'll try and find the relevant Arnold Kling links for you when I have time. And once again I recommed checking out Milton Friedman's article "How to Fix Healthcare", available here:

      I just sent you a couple of Twitter replies, by the way.

    4. Sorry, here's the Friedman link:

      And here's a Cochrane article:

      I'd love to see good counterarguments to these.

  2. I just fail to see why every sane economist doesn't regard Singapore's healthcare system (plus however much more redistribution you want, via the savings accounts) as a vast improvement on everyone else's. And if they do - then the rest is politics.

  3. And I wonder what you think of this: