15 December 2008

ryan health comments 2

Great comments.

The positions you represent, which you probably know, are what is generally called the Austrian school of economic philosophy; with strong skepticism of any government involvement, highest regard for the individual and their ability to optimize without much market correcting mechanisms. This school of thought is emphasized at Auburn, and follows the writings of Ludwig von Mises and Hayek.

  1. Yeah, I agree with the questioned need for government intervention. It has to do with efficiency of redistribution in general, an open question (see comment 5).

The discussion on moral hazard here is, if the government is already intervened in markets, and wants to minimize its expenditures, it should focus on reducing moral hazard. And, to do so, it should implement Medicaid and Medicare structures with higher co-pays and out-of-pocket payments, like the HMO approach. This is a major hope for any government program that wants to adapt its payee structure to reduce costs.

2. In any academic model I would write, it would be fatal to assume that individuals are not optimizing. That is a slippery slope which could explain anything. Away from academia and more realistically, I do believe the less-financially well-off forego buying adequate health insurance. My personal belief is that they do so because the effects aren't as tangible as physical commodities, but when they develop what appears to be a costly condition, they seek insurance; classic adverse selection. It's important because, doctors and hospitals and service missions help them, and ultimately, we all foot some of that bill, and that's my realistic general concern. My personal assessment of reality might be wrong. This is another good question.

3. Academically however, for empirical analysis of the degree of adverse selection and moral hazard in the US health insurance market, I like the model by Chiappori and Salanie, discussed in my re-production of that paper at http://mastroresearch.googlepages.com/ "Private Info for Health Care". It's a good complement to the Rand health experiment.

4. The big picture, like you mention, and I talked about in the last post, is the need to eliminate the information asymmetry, private information. That would indeed fix all the market problems. Unfortunately, health is more complicated than automobiles. Nonetheless, there are some institutional changes we could adopt to lessen the asymmetry. Having a more tiered approach; where patients first see a General practitioner before being allowed to seek a more expensive specialist, would greatly reduce costs by lessening the information asymmetry without government intervention. If the individual is not granted the General Practitioner's consent, the specialist visit either should not be covered, or less coverage than if they did receive consent. You could allow for payment recoupment if the individual did see the specialist and was vindicated in their complaints. The whole problem here is liability. Given the complexity of the human body, the malpractice coverage for General Practitioners would be expensive, and therefore they would pass those expenses to the patient. But, as time goes on, better records are kept, machines improve, and we learn more, those costs would drop. Plus, the Practitioners would have incentive to improve their service, reduce need for malpractice coverage, and get more patients. Reducing information asymmetry also eliminates adverse selection. These first tier practitioners would be hired by health insurers to provide screenings for pre-existing conditions.

5. Altogether, these are ways that current Govt Health Insurance programs could cut their costs. Private health insurance programs have already begun to adopt these measures. The need for existence of government health insurance programs is a different story, having to do with the general trend in our country to choose redistribution. Some argue that everyone is made better off by not having extremely sick cohabitating with extremely wealthy. Redistribution is costly, and if you think the costs outweigh the benefits, let your representatives know.

6. Your comment 3 is quite true and almost impossible to rebuke with a structural model; but it's not going to happen in the near term, and we are going to have to continue to pay the bills for government health insurance for at least a little while. Being pragmatic, a short term goal of economic policy analysis "how to achieve same coverage for less." The bigger picture is political and has to do with desired redistribution (comment 5).

7. Bad government policies do not necessarily imply that voters are not self-interested. That would assume that our political structure is perfectly representative, which we have reason to believe it is not. Given the improbability of actually being the decisive median voter, it is often justified as rational for a voter to be politically ignorant. When the costs of government ambivalence finally catch up with our country (not too far future in my perspective), we will see greater citizen involvement in political process. Another reason self-interested voter's policies might not get enacted are other imperfections in our political process, such as translucency in the quality of that individuals' representation. What exactly is the context of the legislation and all its amendments? What did the rep vote? What did the rest of your district want? Did the rep make a deal on that legislation that will actually help on more important policy tomorrow? Altogether, these two situations justify economist's position of why bad government policies get enacted while voters may still be self-interested. For a structural model of how special interest money can influence the political process from standard majority preferences, see my first paper FAMVM, again at http://mastroresearch.googlepages.com/


Ryan said...

I'm a bit confused that you equate my arguments with Austrian ones, or Austrian arguments with a belief in perfect individual optimization. I rather viewed everything I've said as straight-up neoclassicism -- just plain Mas-Colell, Whinston & Green. Sure, I don't believe in government intervention in a particular direction without a good reason; do many reasonable people think differently? Though I'm not terribly sure what economic philosophy labels have to do with whether my argument is right.

I'm not sure of your argument in re rational ignorance. First of all, it's not clear why higher costs from democracy imply people will be more informed -- there is zero probability of being decisive either way. Second, it's not clear why you think quality of policy has a connection to whether voters are self-interested. I mean, so what? There's no equivalent of the first fundamental theorem of welfare economics for government policy. Quite the opposite: voting for what would benefit others isn't rational, and good policy is a public good (and so we expect people, including voters, to under-provide it). Perfectly transparent government and perfectly self-interested voters would in no way equal good policy. This is sort of what I meant about the nirvana fallacy: if you don't have a model (formal or otherwise) of how the government works, it's not really clear why a good neoclassical economist should assert that government intervention WILL (as opposed to CAN) correct market failure. Don't let the fact that quite a number of neoclassicists forget this confuse you. That's just sloppy economics on their part, isomorphic to a proof that says "5 is not the biggest number in the world, ergo, x > 5" (that is, "there is an allocation that is better than the current equilibrium, ergo, the intervention we get must be better.")

Regarding the rest of your argument, I'm afraid I'm missing how it's responsive. Moral hazard argues against insurance and againstgovernment provision and/or subsidizing health care/insurance. Adverse selection also isn't an argument for gov't provision of general health care; if I grant you this point (and that it swamps positive selection), I'm only granting an argument for gov't providing catastrophic insurance for the poor-- i.e., this point gives no support whatsoever for the vast, vast majority of all gov't health care or for any modification being proposed. And the rest of your discussion seems to be how we could improve the interventions we have, as opposed to explaining why we shouldn't simply abolish them. You seem to be implicitly and sometimes explicitly answering my question "what's the economic rationale for all this gov't intervention" by saying "there isn't one."

I would finally like to make a point about redistribution. Make whatever assertions you want about the demand for redistribution -- they can't possibly explain why you'd want gov't provision of health care. To put this in Micro 1 terms, the only difference between giving people $100 of money and $100 of health expenditures is that the latter imposes an extra constraint in the lagrangian, and adding constraints always makes people weakly worse off. So arguing for redistributional goals works against health care: we could make people just as well off for cheaper by giving them money and getting rid of the general health care subsidy.

Ryan said...

One other comment regarding rational voting. The idea of rational ignorance implies that errors are random. Think about every student you've had in an intro section. Did they make random errors? Or were they systematically wrong? Now remember that many voters haven't gone to college at all and the vast majority haven't taken any economics classes -- that is, a bad intro student knows more than the average voter.

So in addition to the problem that rational and informed voters make socially suboptimal decisions, and the problem that rational voters aren't actually informed, we also have to remember that uninformed voters don't actually have rational beliefs. (Example: think about protectionism. Every such act is bad for the vast majority of voters. Rational, informed voters would vote against corn subsidies. Rational but ignorant voters would also vote against corn subsidies if errors were random. Yet we still have corn subsidies. Ergo ...)