• Priceless: Chapter 15

    This post does double duty. It’s a TIE-U post associated with Karoline Mortensen’s Introduction to Health Systems (UMD’s  HLSA 601, Fall 2012). For other posts in this series, see the course intro. It’s also about Chapter 15 of John Goodman’s book Priceless.

    In Chapter 15, John covers Medicaid, as does Mortensen in her class this week. John’s treatment provides abundant grist for the adversarial* mill. Before I get to it, I want to make something perfectly clear. One can defend Medicaid as better than no insurance without implying that Medicaid is a great program needing no reform. That’s my position. By the same token, one can advocate changes to Medicaid without having to resort to preposterous claims that it is worse than no insurance at all. That’s how I choose to do it. But it is not the approach taken by some of the scholars and commentators that Goodman cites.

    Here are a few quotes from the chapter:

    Another study found that even the uninsured have an easier time making doctors’ appointments than Medicaid enrollees.4 [Reference 4 is to an ungated JAMA study by Brent Asplin and colleagues.] […]

    The academic evidence suggests that there is a severe quality problem in Medicaid. Here are some studies identified by American Enterprise Institute scholar Scott Gottlieb.13 [Reference 13 is to a Wall Street Journal column.][…]

    Here are some additional studies identified by Forbes health blogger [now Romney adviser] Avik Roy.18 [Reference 18 is to a post at National Review Online.] […]

    Frakt points to some studies finding that Medicaid makes a positive difference over being uninsured.27 But the results would probably have been just as good or better if we spent the money giving free care to vulnerable populations. Moreover, even with their Medicaid cards, enrollees turn to emergency rooms for their care twice as often as the privately insured and the uninsured. [Reference 27 is to this post by me.]

    John’s dismissal just above of my post (his ref 27) is, in the first sentence, speculation (hence his “probably”) and, in the second sentence, not germane. (It is germane to something else.) But, I do want to express kudos to John for raising both sides in the Medicaid outcomes debate, though he gives far more space to the Gottlieb-Roy view — that Medicaid is worse for health than no insurance at all — than to mine. The avalanche of studies provided by them, which John lists, sure gives the impression the body of evidence is on their side. It isn’t. Take a look at this post that goes through Gottlieb’s favorite studies. None of them provide a causal estimate of the effect of Medicaid on health, and they say that explicitly. They suffer selection bias. Same goes for Roy’s cited studies. As for the Asplin study (John’s reference 4 in the quote above), it doesn’t say what John wants you to think it does.

    As some readers may know, this battle over interpreting Medicaid outcomes studies motivated a thorough literature review by yours truly of studies that use methods capable of yielding causal estimates. I followed that with a (peer reviewed) paper in NEJM co-authored by Aaron, Harold, and Uwe Reinhardt. You can also consult the Oregon Health Study, which provides the least ambiguous illustration of Medicaid’s positive effects on health. Though it is true, as John points out, as do the study’s investigators, that much (but not all) of the effect documented in that study occurred after Medicaid enrollment but before much care was delivered, it’s still an effect of (randomized!) Medicaid selection. Winning the Medicaid lottery can reduce stress and, thereby, be good for health. The study investigators offer other explanations for why immediate effects aren’t comparable to later ones, but getting into them is beyond the scope of this post.

    With all the resources linked to above, you have all you need to inform yourself of the truth about Medicaid and outcomes. If you still want to believe John et al. that Medicaid is worse for peoples’ health than nothing, I can’t stop you. I am, however, still waiting for someone to articulate the mechanism. After all, nobody is obligated to enroll in Medicaid or to show their Medicaid card. One can always pay out of pocket to non-Medicaid providers.

    Pertaining to Medicaid block grants, John tells the story (or a story) of Rhode Island. Aaron tells a different one. About privatizing Medicaid saving money, the study that shows otherwise by Mark Duggan and Tamara Hayford is worth knowing about. About the size of refundable tax credits that John wants to offer consumers as part of a Medicaid reform, he writes,

    Some may worry that the $8,000 refundable tax credit [per year for a family of four] would not be sufficient to purchase adequate health insurance, and the private package could be worse than Medicaid itself. This is not something I worry about. Surely private entrepreneurs can produce a better health plan for $8,000 than what Medicaid offers. But just in case the critics are right, let’s keep traditional Medicaid around as a stopgap. Let people choose between Medicaid and private insurance, and let the government’s $8,000 check go to the plan they choose.

    Let’s get one thing straight, $2,000 per individual per year for health care is a legitimate concern. John notes in the chapter that in 2007, per person Medicaid spending was $5,163. So, John’s plan is to severely underfund Medicaid. You can’t say, “But this is for a catastrophic policy, which is cheaper,” because Medicaid beneficiaries can’t afford deductibles on those policies either. You can’t say, “But this is just for kids and adults, not the elderly,” because (a) John didn’t say that (or did I miss it?) and (b) even if he did, the fact is that adults cost more than $2,000 per year, and especially so if they are blind or disabled or are BCCA (Breast and Cervical Cancer Act) eligible women. Heck, foster kids cost more than that too. If critics are right, John’s stopgap is to let traditional Medicaid buy coverage for that same (too) low price. How is that a reasonable stopgap? It isn’t. If critics are right then $2,000 per individual on average is not enough and a stopgap is to spend more, not the same amount.

    In any case, why not let a competitive market involving public and private plans establish the right price? Maybe it is $2,000. Maybe more. Maybe less. Competitive bidding anyone? I’m surprised this idea never appeared in John’s book. Why not?

    I will post on Chapters 16 and 17 on Thursday. All posts in this series found at this link.

    * My ambition in this series on John’s book is not to be adversarial. I hope that’s clear. But I’m human, and sometimes I get annoyed when the same errors are made repeatedly by people who have the ability to know better.


    • I believe we established earlier that Medicaid spends $3,000 per adult and $2,000 for a child. The refundable tax credit I proposed is about $2,500 per adult and $1,500 per child. So if non-poor people want to buy into Medicaid, they would have to add $500 to their tax credit. If people on Medicaid want to leave the system and buy into a private plan, they would probably have to add additional money to the “Medicaid voucher” they would be entitled to.

      As for Medicaid’s quality, no one believes that it is as good as private insurance. I have been surprised at the people who not only defend two-tier medical care but actually advocate its expansion. I guess on this point I have more in common with British and Canadian socialists than I have with American liberals. Let’s put everybody in the same health care system.

      • Something isn’t right. At $2k per child and $3k per adult, how is Medicaid averaging $5,163, as you wrote in your book (which is consistent with what I’ve seen elsewhere)? One answer is that those are not all of the categories of people Medicaid covers. See Aaron’s post again: http://theincidentaleconomist.com/wordpress/medicaid-you-cant-save-much-money-by-dropping-work-eligible-adults/

        Are you leaving all those other folks out of Medicaid? Did you write that in your book (which is what I’m reviewing, not other blog comments)? All I recall seeing is $2k per person.

        Let’s see if we can get this basic math problem right. That would be a significant, collaborative achievement!

      • ” Let’s put everybody in the same health care system.”

        I like this idea, though I suspect we would disagree on the nature of that system. I think this strongly enough that I would happily take a system about which I have serious doubt, but puts us all in the same health care system, over our fragmented system. We would then have a better chance of reforming a uniform system if it did not work out.


    • I am using Aaron’s numbers, which you linked to. I wasn’t counting the disabled or the elderly or the dual eligibles. Although i would like to see more private sector competition for those groups as well.

    • Not in the book. I am using Aaron’s numbers at the link you have provided.

    • Yikes, there is a lot here. One of the reasons it is hard to even discuss Medicaid is because it is really three programs under one administration — low income families, the disabled, and LTC for the elderly. I have long thought the program needs to be divided into its three components. Other than the state/federal match, these three have nothing in common.

      I don’t agree with those who say Medicaid makes no difference. It must, but the money spent may not be warranted by the difference in outcomes. I would offer a different view. Health outcomes are largely a function of the age, education, and income of the population. These factors are (imo) more important than the insurance people may or may not have. That is why this quest for universal health insurance is (again, imo) misguided.

      Many many people simply can’t cope with insurance of any kind.That is why I admire John’s basic idea of tax credits (vouchers) that may be used to buy insurance, but otherwise go to safety-net providers for the direct provision of services.

      If we really care about the disadvantaged we would not try to force them into a financing system that is built around the sensibilities of the upper middle class. Clark Havighurst has done some interesting writing about this.

    • “Many many people simply can’t cope with insurance of any kind.That is why I admire John’s basic idea of tax credits (vouchers) that may be used to buy insurance, but otherwise go to safety-net providers for the direct provision of services.”

      I am post call again, so maybe I am fuzzy yet, but it seems you are saying they cannot handle insurance, so give them money to buy insurance. ?? If they cannot handle insurance, let them spend the money directly with some subset (?) of providers. Which providers? How much do they pay them? Who decides? If one does not have the skills to negotiate insurance, why would one necessarily have the skills to negotiate directly for services? Why would providers be willing to accept less (I think that is what you imply with the idea of safety net providers)?

      I think you had it close here.

      “Health outcomes are largely a function of the age, education, and income of the population. These factors are (imo) more important than the insurance people may or may not have. ”

      These factors really are important. Because of these, outcomes will likely always be worse for this group. But, they will be even worse if they have no access to health care, as you imply in the sentence preceding the quote. I can see making a case for a better ROI. If increasing expenditures stops delivering improved outcomes at some point due to the the social issues like the ones you suggest, and others, then we could look at capping those expenditures. Maybe we should look for another method besides insurance. Perhaps free public clinics is the way to go. Maybe something else. However, I dont see how we can just assume that giving them less money than Medicaid spends will result in better care.


      • Steve,

        No, the voucher would go towards buying insurance — for those who wish to buy it. For those who don’t, the money would go into paying for safety net providers (likely public hospitals and clinics) who would provide services directly. John’s usual example is the Parkland Hospital in Dallas (which otherwise provides free care to the indigent), but it would be more systematic than what happens today. It is worth looking at Goodman’s description of how it would work, though I don’t think he has gotten very detailed on exactly how the funds might be allocated and for what services.

        On your second paragraph, I am suggesting nothing of the sort — not less care, not no care. Actually better care and more care because we bypass the complicated insurance aspect and provide the care directly.

    • Arguing that Medicaid provides worse health outcomes than no insurance, then suggesting that we provide less money to those who qualify for Medicaid is simply ludicrous.

      If no insurance was truly “better” than Medicaid, then why provide people with Medicaid at all? But perhaps John is right, Medicaid, as a monopsonistic purchaser of health care has lower quality. Why might that be?

      “Monopsony power, however, does not necessarily trigger this negative welfare
      effect. If its exercise were confined strictly to capturing economic rents that
      would otherwise be earned by providers, then economic theory would not predict
      an inevitable reduction in the quantity or quality of health care. The effect might
      be merely to redistribute income from the providers of health care to the rest of society.
      Even then, however, it is possible that a monopsonistic payer might push
      this process too far and eventually trigger reductions in either the quantity or
      quality of health care, or both. Using monopsonistic payer systems in health care
      to procure just the mix of quantity and quality that is actually desired by the insured
      citizenry is a daunting task and not always achieved successfully in practice.”


      For sake of simplicity, lets assume that spending, as the above quote illustrates, is a major factor in determining quality with respect to insurance programs. Private insurance spends more money than Medicaid and receives higher quality outcomes as a result (at least as John argues). You don’t even have to assume causation here, correlation would be sufficient.

      That is to say if you spent more per person in Medicaid, quality would go up. You spend less, quality goes down.

      If however, the above assumption doesn’t hold, and there is no positive correlation, or at least not major positive correlation between spending and quality, then what is the deciding difference between private insurance and Medicaid that makes private insurance plans lead to better health outcomes?

      We know that Medicaid spends significantly less per person than private insurance, so right away people have to spend more for the same services in a private insurance scheme.

      Lets also not forget that those enrolled in Medicaid are also likely to be sicker than the average person, which means their premiums in a private insurance scheme are automatically higher.

      This means that anyone dumb enough to use their voucher for private insurance instead of Medicaid is paying significantly more for less.

      In order for John’s argument to work he would have to explain how private insurance provides better health outcomes controlling for not only demographics (people in Medicaid are by default more likely to be sicker) but also that amount of money spent by private insurers has no effect on outcomes at all.

      Otherwise, if you want higher quality health care for the people that can’t afford it, you have to spend more on Medicaid.

    • I have been in favor of more free clinics and more subsidized hospitals for years.

      There are a couple of challenges that must be faced in this regard.

      One is that the funding must be federal. States and cities have been running away from public hospitals for years. In the case of hospitals, they were sometimes seduced to sell off public facilities. This pleased local taxpayers and made some persons very rich.

      Federal funding will have to get past the anti-tax hawks. I suppose it can be presented as just another Medicaid initiative rather than a new tax.

      The second challenge is that America is a huge country. Not all the poor people live in big cities. A free clinic in rural Montana makes no sense. We need a way to pay local doctors and local small hospitals to give out free care. They are on the ground and ready to help.

      I guess this is a form of two tier medicine, where poor persons do not have their choice of doctor. Big deal, I thinik. Free care from an American clinic with no choice in the matter is better health care than 99.9 per cent of human beings throughout history have received.
      More to the point, I have been poor, and then I did not care about choice if the care I got was free and decent.

    • Bob Hertz, you wrote — “The second challenge is that America is a huge country. Not all the poor people live in big cities. A free clinic in rural Montana makes no sense. We need a way to pay local doctors and local small hospitals to give out free care. They are on the ground and ready to help.”

      I agree. This shouldn’t be too hard to arrange. For that matter, many places don’t have public facilities, and private ones could be paid to do the job. One thing Bush never got much credit for was a major increase in funding of local clinics.

      On federal funding, I believe John’s idea is to replace the current employer exclusion with this refundable tax credit. I suppose that could be matched with state money to top-off the low income.

    • Greg, I am glad to see we are in agreement on creating a safety net by paying local doctors and hospitals to provide free or next-to-free care.

      I am somewhat skeptical about the method that Dr Goodman has proposed. His plan would have the cash value of unused tax credits flow directly to safety net providers, who would use those funds to care for the uninsured.

      This seems unwieldy. First we have to wait a year and see how many credits are unused. Then we have to see how many credits are unused in Minnesota, and for how much.

      The providers need money on January 1st of each year. I would favor direct budget allocations on a per-population basis. Take the money right out of the Medicaid budget.

      Giving Minnesota 2% of the funds because Minnesota has 2% of the nation’s population is crude, I admit. But truthfully, I think that a lot of government funding for health care — and not just Medicaid — will have to come down to something like this, i.e. the block grant approach and then let local experts divide it up better.