Government Insurance Advantage?

Comments on Tuesday’s post itemizing medical market failures suggest that many think private insurance fails via excess administrative costs, and via excess costs of enforcing contact details.  Private insurers, many think, charge too much and try too hard to renege on their promises.  Let me explain why this sounds odd to an economist.

In a typical industry, producers use tech to convert inputs into outputs, and price those outputs to induce consumers to buy them.  Markets typically fail by inducing consumers to buy the wrong product mix, and inducing producers to make the wrong product mix.  In such cases we consider regulating prices and/or product mixes to mitigate this failure.  In principle, a market could also fail if private producers only had inferior tech available to them, while the government had superior tech only usable via direct government production.  In such an apparently odd hypothetical case one might prefer direct government production.

Strangely enough, this possibility seems to be the reason many want government provided medical insurance.  Let me elaborate.  There are usually many trade-offs to make between costs and various product features and methods of production.

In medical insurance, the cost to estimate customer risk is reduced if all customers must buy your product.  That and other administrative costs are also reduced when the insurance is renegotiated less often, and when one insures a large group at once instead of one at a time.  The cost of enforcing contract details, such as what treatments are covered in what situations, is higher when the contract is complex, excluding many particular things while including many others.  Contract enforcement is also harder when the agents who are supposed to comply with the contract have strong contrary financial incentives.

Many think that government provided insurance would be cheaper and more dependable.  But the question is: would this be because the government would choose a different product mix, using the same insurance-making technology available to the private sector, or is this because the government is assumed to have some superior tech for making insurance?

If government insurance is better because it is infrequently renewed, is required to all or offered to large groups, has simple coverage definitions, or is implemented by agents with weak financial incentives, then we must wonder: why couldn’t private insurers make those same production choices?  If they could have done so, but didn’t expect consumers to want such a product, then there would have to be some market failure that induces a bad product mix.   If they could not have done so, then we have to postulate some extra insurance technology only available to government insurers.  But what could that extra tech be?

GD Star Rating
Tagged as: ,
Trackback URL:
  • Ben

    “If government insurance is better because it is infrequently renewed, is required to all or offered to large groups, has simple coverage definitions, or is implemented by agents with weak financial incentives, then we must wonder: why couldn’t private insurers make those same production choices?”

    It seems to me that “required to all” is an option only available to government (or of course a government-backed monopoly).

    • Remember I discussed the possibility of regulating private insurance. Regulation could require everyone to buy private insurance, just as it could require folks to get public insurance. Also, consider that we can get most of the gain from this, reduced effort to estimate risk, by going to longer term contracts and larger groups. If the public isn’t now much interested in such products, we need a market failure explanation for that disinterest before we conclude they would be better off being forced to get every last bit of such gains possible.

      • we can get most of the gain from this, reduced effort to estimate risk, by going to longer term contracts and larger groups. If the public isn’t now much interested in such products

        I don’t see how you can determine this lack of interest. It’s not obvious whether the lack of long term contracts is due to lack of interest or obstacles to supply. I see a preference for getting insurance via large groups where I see it being offered (employers). It seems quite likely that obstacles to supplying other types of group insurance are sufficient to explain their absence. How would insurance covering a new type of group coexist with employer-based group insurance?

        Also, the historical reluctance to adopt new types of insurance (especially life insurance) suggests that if there is a lack of demand it may be due to irrational attitudes toward risk.

        I agree that economists advocating government insurance ought to address these issues. Maybe that would help reveal regulatory obstacles to longer term contracts.

      • Benquo

        A requirement to insure everybody could be gamed by parties who want to buy (or sell) less insurance than the required minimum, unless it’s very carefully written and enforced. If one insurer figures out how to offer less coverage within the limits of the law, they’ll suck up most of the low-risk market (at least until their competitors catch up in the race to the bottow). My guess is this problem would be less relevant in a single-seller market.

        BTW I think this is a very weak advantage — not because the problem isn’t real, but because the creation of a single-seller market is a poor solution (as gov’t is no more likely to be populated w/ angels etc.)

  • Psychohistorian

    This doesn’t seem like a hard question.

    The market can always profitably discriminate in favor of low-risk people. A non-discriminating company can have it profits taken by a discriminating company that offers low-risk members insurance at a much lower price.

    The outcome of a highly competitive market is that insurers will offer people insurance as at close to the actual cost of their insurance as they can, because this cost (unlike, say, fire insurance) is quite predictable, varies tremendously between individuals, and in some cases has no real upper bound. This pretty much defeats the purpose of insurance-as-spreading-out-risk. Even if people prefer a monolithic insurer that insures everyone, the market just can’t make that happen on its own.

    • Psychohistorian

      I’d add that this doesn’t necessarily mean the government will be able to hold up in a truly competitive market, only that a specific (and arguably preferable) equilibrium can be brought about by single-payer that cannot be brought about by competition. There is some serious risk of companies fishing out all the low-cost buyers after the government enters the market, but the amount of research and advertising necessary to do this, combined with public distrust of insurance companies, may prevent them from being effective.

    • Your “insurance as spreading out risk” isn’t really insurance, it is redistribution. Yes of course the government can redistribute, but there is not obvious reason to mix that activity with providing insurance.

      • Randall Randall

        But “health insurance” isn’t really insurance; it’s just spreading the cost of healthcare across more people. If it were really insuring against something, that something wouldn’t be an expected and regular event, like yearly checkups and monthly prescriptions for allergy drugs. People call it “health insurance” in an attempt to obscure the redistributionist nature of the industry — an attempt that’s apparently succeeding. 🙂

      • komponisto

        What Randall said. Since medical care is the rule rather than the exception, the appropriate model is that of a service to be provided, rather than a risk to insure against. So we should either have folks pay directly for medical services, or just have them provided by the government. Paying a third party to pay the provider seems like a pathological inefficiency.

      • Granite26

        Wait… isn’t that what insurance is doing? I’ve got a 10% chance of a $10,000 cost, so I pay $1,000 to cover it in case it happens.

        There’s only a difference between redistribution and insurance if people aren’t paying proportionate to their expected costs/risk (which may be your point, I don’ t know)

      • Psychohistorian

        If your idea of efficient insurance is that everyone pays roughly the actuarial cost of their insurance plus a small premium for management etc., then your system is entirely nonredistributive and the government can’t really have an advantage. Plenty of people will probably choose to go uninsured in this equilibrium, particularly given upper bounds and bankruptcy incentives, though this depends on the legal responsibilities of health care providers.

        If you believe that people who have higher actuarial risk should have their costs defrayed by those fortunate enough to have lower actuarial risk, the market will not achieve an efficient outcome with out at least moderate intervention. This system could redistribute based on income, but it need not necessarily. I’d think that most people believe those who are unfortunate (especially for exogenous reasons) should get some kind of subsidy. Hence all the complaints about rejection or discrimination due to pre-existing conditions.

        Admittedly, people want others to get this subsidy, but they are not so keen on paying for it.

      • anon

        “Plenty of people will probably choose to go uninsured in this equilibrium, particularly given upper bounds and bankruptcy incentives”

        First of all, bankpuptcy is itself a form of redistributist social insurance. Secondly, if we hold redistribution as a normative goal, it’s not clear why the government couldn’t specialize in insuring the highest-risk portion of the population (so as to set a standard of care) and let everyone else be served by private insurance.

      • Granite26

        Thanks, Psychohistorian… I think I understand what the people who disagree with the current system are thinking a little better now. Don’t agree, but I don’t think they’re crazy.

  • Mario

    I don’t know that I’d call it “tech,” though that’s an interesting way to look at it. The government would have a lot of advantages in the insurance business, but mostly because it dictates laws and regulations to insurance providers that it will not have to abide by. For instance, a government plan would probably not be regulated on the state level, nor would it be taxed. The current health care plan increases these regulations on legitimate insurance providers, so the advantage will grow. The plan also wouldn’t need a large stockpile of reserves on hand in case of a sudden surge in claims, it would not be at risk for default, and it wouldn’t need its own HR division. There are quite a few more, none of them being inherent advantages in the insurance industry, just the typical ways a government could muscle its way into a business.

    The point, I suppose, is that the government could easily become the major or sole provider of health insurance in a very short time, but we shouldn’t confuse that with the idea that it is actually providing a better product.

    • Kenny Evitt

      This is the first thing that came to my mind – the federal government doesn’t have to abide by all the regulations that affect private insurers, nor are they (quite as) limited by the other myriad constraints inherent in private organizations (whether for-profit or not-for-profit).

  • Unnamed

    One thing that the government has is a respected brand name, which may allow it to attract a relatively large and not especially high-risk share of the market. Other companies devote a lot of effort to competing with each other to avoid getting stuck with the high-risk customers, and the government could skip out of that competition and focus on other things without getting stuck with the riskiest customers. It may not work – the government could get stuck with the risky folks – but the government’s chances of avoiding that fate are much better than those of any other company that tried the same approach.

    • I don’t see what the brand name as to do with the relative fraction of high vs low risk customers who are attracted to it. Are you suggesting only low risk people like the government brand name?

    • Kenny Evitt

      I can’t disagree more. Where’s the evidence that the government indeed has a respected brand name? Did you have a specific field or industry in which the government is so well respected?

      • Granite26

        No strings charity

  • Mike

    It’s a good point above that govt might not be able to hold up in a competitive market. This situation might be seen as similar to that with public schools. If public schools had to raise funds by charging tuition, they would be at a significant disadvantage, because they agree to take all students, including special needs students who demand relatively large resources, while private schools can be selective.

    Plus, private insurance has the benefit of being able to externalize costs. That is, it increases efficiency within the private insurance company if it spends resources to deny “legitimate” medical claims. But if the medical procedure is going to be purchased anyway, and it’s a type we want people to have, then this policy introduces inefficiency into the system as a whole.

    To me there is something akin to environmental regulation in this. It is terribly inefficient for a company to dump waste on pristine land, if eventually people will want to use that land. But it’s a very good deal for the company, if it won’t be held responsible, and therefore it does so. It would seemingly not help to create a govt option, since presumably the govt option would seek efficiency for the system as a whole, would not so pollute, and thus would have to charge more. We could nationalize the whole industry, but we choose instead to create regulation to change business incentives.

    It seems an interesting debate whether we should (1) simply devise new and better regulation for health insurance companies, (2) create a public option, or (3) nationalize the entire market. Maybe this is precisely the debate Robin is trying to have. An interesting aspect of the debate is that perhaps we are not debating realistic options. That is, Robin is arguing against (3), which I guess is the option most liberals prefer, yet it is not on the table, from what I understand. My understanding is the ideas floating around Congress involve (1) and (2). For reasons I said above, I suspect (2) is not going to work well unless there is subsidy. I suspect many people like this option anyway, because they see it as a first step toward (3).

    • Government insurance can also externalize cost by denying treatments. That is the whole point of my discussion of contract enforcement. Also, public schools could charge tuition.

      • Mike

        Govt insurance *could* externalize cost by denying treatment, but I presume it does so for different reasons. A private company does this to increase the efficiency of the the company, even if it is not efficient for the economy as a whole. But govt can in principle act with regard to efficiency for the economy as a whole, and presumably this would be the reason to “externalize” costs. Yet I would agree that, realistically, there are a number of less desirably reasons why our govt would deny treatment.

        I don’t understand the reference to public schools charging tuition. Maybe my own reference wasn’t clear. I meant only to suggest that they are at a disadvantage insofar as they do so, because their public mission incurs costs that private schools can avoid. Something similar might inflict a public insurance option, relative private insurers.

      • Chris

        I already pay tuition. Its called paying taxes…

      • lxm

        Government insurance should externalize cost by denying treatment. However, it should be done
        . You should know up front what treatments the government will pay for and what they won’t. In today’s health care world, you do not know what insurance will cover or deny until you ask. Let the private insurance industry fill in the gaps.

        If you’re 85 years old with an incurable cancer the government should not pay for a hip replacement. That’s just stupid. Government should pay for palliative care, but that’s all.

        There’s got to be limits. One of the many problems with the current system is that the limits are not explicit. I wish there was more discussion of limits to health care in the current discussions, but apparently those conversations are off limits for political reasons.

    • Kenny Evitt

      Why would you presume that governments would deny treatments to increase the efficiency of the healthcare industry? They deny treatments for exactly the same reason as private companies – lower costs. One of the main selling points of a “public option” is that it will cost less than existing private insurance.
      A (truly) free market in healthcare would allow insurers to compete on treatments covered, etc. and information about the denial of treatments would be available for a secondary market of reviews, guides, competitor advertising, etc. It’s bad enough that education is so heavily political; do we really need political conflict over treatments covered by the federal government’s public option?

      • Mike

        I guess the question is, why does the govt want to lower costs? I presume this incentive comes from voters, who do not want to pay more for health care than they have to. But, establishing a policy where the govt denies to pay for treatment that the public thinks people should have, this does not decrease the cost of health care. It merely shifts it from govt to individual. So, presumably people get together and decide how the burden of paying for health care should be distributed, and establish policy to reflect that. Efficiency might be gained because the govt doesn’t spend money fighting paying for treatment that people intend to receive regardless.

        This is very different than with the private insurer. The private insurer is insensitive to the total cost of health care — they pay only what they pay, which makes it profitable for them to spend money fighting claims.

  • Kenny Evitt

    Replies can only be nested to two levels?

    • Jess Riedel

      Yes. It’s annoying. I assume it is to keep the page from getting too cluttered or the conversation too off-topic. The standard operating procedure seems to be replying to the same level once you run out of room.

  • Thanks, Robin. This post helps confirm something I’ve been considering for a while, namely that if government is to take a stronger role in health care it must do so in a way that goes outside of what the market could do. In particular, whenever I hear proposals for a “public option” health insurance plan, this just seems dumb to me because if the government is going to provide health insurance that is better than what’s currently on the market, why hasn’t the market already provided it. I suspect that it wouldn’t be profitable, and any public option is going to require government subsidies to remain viable.

    That said, there seem two interesting ways we might go to fix the problem. There are others, but these are the ones that seem best to me.

    One is to outlaw health insurance. Without anyone having the option of it, only a very small minority of people will be able to pay the current prices for health care. To avoid losing their livelihoods medical professionals, their support staffs, the educational institutions that train them, the pharmaceutical industry, and others associated with providing health care or services to health care providers will be forced to lower prices. In the short term this is a rather nasty option because it destroys a huge amount of expected earnings and would thus damage the entire economy, but on the other it restores the health care market in a way that would likely be sustainable at an affordable rate.

    The other is to eliminate any use for health insurance by nationalizing health care. By paying for health care mostly with taxes (you must have some usage charges, though, to prevent the increased costs that are associated with totally free care that doesn’t result in improved outcomes) we ensure that as a society we all have access to care, thus improving public and thus individual health, and reduce costs through a combination of strong government negotiation, widely distributed risk, and time to gradually bring health care into a proper valuation.

    I’m not a health economist, though, so I could always be missing some key details in my analysis, but it does seem to me we have a case of “the emperor’s new clothes” when it comes to proposals to reform health care.

  • Granite26

    Could it be that part of the problem with sic ‘Health Care’ (really health insurance, since we have great health care) is that people don’t understand what they’re paying for?

    I had a long debate with an otherwise intelligent person complaining that his insurance didn’t cover his wife’ pregnancy. I had to explain to him that pregnancy is a 99.9% consumer controllable cost, and if it was included in the insurance package, you could expect your rates to go up by 10k (his cost) a year because no one that wasn’t planning on getting pregnant would pay the extra money.

    People think ‘I have insurance’ but not ‘I have insurance that covers X or Y’, except maybe a few common items. They think that when they purchase healthcare, all costs are covered (or should be).

    Could be just me, though

    On the other hand, can you really expect people to make rational decisions about what health services they need or are effective when it’s their life on the line? People aren’t wired to say ‘It’s a 1 in a million shot, so it’s not worth taking, I’ll just enjoy the time I’ve got left and not spend that money that isn’t mine to begin with’.

  • It is best to think of private insurance as piggybacking on government insurance. Government is the largest player in the market and dictates coverage and costs through medicare. Private insurers largely just tack on overhead charges and use these to limit coverage and select healthier than average clients, even if they have to do it after the fact as some recently did in California. They really don’t have an advantage, only a disadvantage due to their smaller size, one that they can only partially overcome by being direct products of their affiliated producers with no real interest in reducing costs but only in passing them along.

  • Jim Babcock

    Private insurance is more expensive because it creates an adversarial relationship between the insurance companies, which want to pay for as little as possible, and patients and hospitals, who want to have stuff paid for. Insurance companies employ large bureaucracies for the sole purpose of finding excuses not to pay, and hospitals employ their own bureaucracies for the sole purpose of fighting with the insurance companies. Both groups are unnecessary, and would vanish under a single-payer system.

    I estimate that about one-fourth of the current cost of health care is solely due to this unnecessary bureaucracy, based on the medical loss ratios I found googling, plus an assumption that hospital costs for dealing with them is smaller but on the same order.

    • Constant

      Explanations of why single-payer systems are better than private systems need to explain why this does not extend to everything. For example, by the same reasoning, single-payer food supply should be superior to private food supply. Granted, there aren’t insurance companies involved, but there are relationships whose logic is just as adversarial, since the adversarial logic of the relationship derives from the adversarial logic of buyer/seller relationships, which logic is present in any buyer/seller relationship.

      Arguments that medicine should be socialized need to show what is special about medicine, how it is fundamentally different from food, clothing, computers, auto repair, and so on. The adversarial logic of the buyer/seller relationship is something that medicine has in common with every other market and is thus not a way in which it is fundamentally different.

      • Doug S.

        Arguments that education should be socialized need to show what is special about education, how it is fundamentally different from food, clothing, computers, auto repair, and so on. The adversarial logic of the buyer/seller relationship is something that education has in common with every other market and is thus not a way in which it is fundamentally different.

      • Constant

        Doug – yes, that’s correct. And here’s what I think people think is special about education: it’s so important that children get it, and that it have a certain content, that the choices of the parents must be overriden, and if the parents cannot afford it (or choose not to pay for it) then it must nevertheless be paid for (meaning the taxpayer pays).

        That is nothing about the adversarial relationship. Similarly, I expect that the argument about the adversarial relationship in medicine is a rationalization, and is one of many arguments that people come up with after they have already decided, for other reasons, to advocate single-payer.

      • Jeff

        I’m not sure that single-payer is clearly better, but I’ll support it being fundamentally different than food.

        In the market for food, I shop for the food, I choose the food, I pay for the food, and I eat and enjoy or not the food. In the market for healthcare, Food is a 2-way interaction.

        I purchase insurance from competing options whose differences I can’t really tell — that’s if they let me. I ask my doctor — except the Insurance company tells me which doctor to ask — who then tells me what services I need. The doctor gives me the services, and then asks the insurance company to pay for the services.

        This is also fundamentally different from other insurance in the inclusion of the doctor, and the fact that I’m not insuring against risk or loss — I’m not buying a certain amount of money in response to certain actions that happen to me — I’m insuring against the doctor determining that I need a certain service. Since the insurance company determines whether a service is paid for and the doctor determines whether a service is needed, and when, I cannot properly evaluate whether the insurance is adequate to my needs. Health insurance is a 3-way interaction.

        Now, if you have another 3-way interaction to compare it to, this might get interesting.

  • “some extra insurance technology only available to government insurers. But what could that extra tech be?”

    Huh??? Taxes and machine guns, obviously.

  • undergoingbias

    “why couldn’t private insurers make those same production choices? If they could have done so, but didn’t expect consumers to want such a product, then there would have to be some market failure that induces a bad product mix.”

    Exhibit A: market failure that induces a bad product mix. combine the investor community pressure with the decreased competition (as seen by concentrated regional market shares).

    WENDELL POTTER: Well, there’s a measure of profitability that investors look to, and it’s called a medical loss ratio. And it’s unique to the health insurance industry. And by medical loss ratio, I mean that it’s a measure that tells investors or anyone else how much of a premium dollar is used by the insurance company to actually pay medical claims. And that has been shrinking, over the years, since the industry’s been dominated by, or become dominated by for-profit insurance companies. Back in the early ’90s, or back during the time that the Clinton plan was being debated, 95 cents out of every dollar was sent, you know, on average was used by the insurance companies to pay claims. Last year, it was down to just slightly above 80 percent.

    So, investors want that to keep shrinking. And if they see that an insurance company has not done what they think meets their expectations with the medical loss ratio, they’ll punish them. Investors will start leaving in droves.

    I’ve seen a company stock price fall 20 percent in a single day, when it did not meet Wall Street’s expectations with this medical loss ratio.

    For example, if one company’s medical loss ratio was 77.9 percent, for example, in one quarter, and the next quarter, it was 78.2 percent. It seems like a small movement. But investors will think that’s ridiculous. And it’s horrible.

  • Good explanation for the failure of failure of private insurance.
    This must be realized by all private insurance company which introducing their products in network marketing companies for mass moving of their products, actually this will give all credits and publicity to the networkmarketing company. But if insurance company realize this all these benifits are for them only.

  • Doug S.

    I’ll let Paul Krugman answer this one.

    Why Markets Can’t Cure Healthcare

    • Jeff

      That’s a terrible op-ed. He argues that health insurance isn’t like food, it’s like insurance, and then fails to discuss how it’s fundamentally different from other insurance markets, which work successfully. It’s true that you can’t predict whether your house will burn down, but you can and do buy insurance against it and that market works fine.

      His point about trust is an interesting one — we don’t trust HMOs to make healthcare decisions for us, and points toward one of the problems (doctors), but doesn’t really explore it.

  • lxm

    Here’s more about malfunctioning health insurance markets at baseline scenario:

    Your answer please.

  • Jeff


    I’m really interested in your using this lens to evaluate Medicare. Medicare is a large government-run health “insurance” system that has incredibly happy users. In what ways does it fail to live up to your ideals?

  • David

    All of the projections I’ve seen for Medicare are dismal- the costs are projected to go up so much, that we would have to either double taxes or eliminate all other spending to fund it. So obviously, Medicare will have to be slashed. My question is, what would a single payer system do that Medicare doesn’t, to avoid the same fate?