• Community rating has an economic rationale

    I had a bit of a Twitter flurry this morning, with some back-and-forth with Avik Roy. One of the things that came up in the exchange that I agreed with turns out to be incorrect. So, let me correct the record.

    In the tweet, Avik is talking about the age-based pricing regulations of the ACA that constrain exchange-based premiums for older consumers to be no more than three times those of younger consumers. This modified community rating scheme is one of the aspects of the law that he and Douglas Holtz-Eakin would like to see removed. They prefer unconstrained pricing. Well, if there’s no moral or economic reason for community rating, I agree it should go, though the replacement demands considerable income redistribution, which I find politically unlikely. Nevertheless, this is the barest of hints at an idea advocated by Mark Pauly long ago. (See, for example, his eBook.)

    Coincidentally, not long after the Twitter exchange I turned to a developing working paper by Keith Ericson and Amanda Starc. (It exists as an NBER working paper and a more recent, ungated PDF.) Turns out, they offer an economic reason for age-based pricing regulations. So, Avik and I were wrong.

    [C]onsumer welfare is higher with the [age-based] pricing regulation in place than it would be in the absence of regulation; the positive compensating variation for the older consumers is larger in magnitude that the negative compensating variation for younger consumers, who must be paid to be made whole. […]

    With the ACA’s maximum allowable price ratio and minimum loss ratios, consumer surplus is increased relative to unconstrained age pricing.

    The reasoning is pretty intuitive. In an imperfectly competitive health insurance market, insurers can exercise market power and charge markups over the competitive price. The degree of markup is related to consumers’ price sensitivity. In the insurance market, older consumers are less price sensitive than younger ones. That is, older consumers are willing to pay more for the same product. Consequently, in the absence of age-based pricing regulation, markups would be higher for older consumers relative to younger ones. (Note, this is true apart from health status differences. Older consumers have other reasons to be less price sensitive. They may be more risk averse, for example.)

    What age-based pricing regulations do is cap the markups that could be charged to older consumers by yoking prices to those faced by younger consumers. This decrease in price for older consumers increases their consumer surplus (the value they obtain from the product less the price they pay for it). Of course, younger consumers experience an increase in price, decreasing their consumer surplus.

    The crucial finding is that the decrease in consumer surplus of younger consumers is more than offset by the increase in surplus for older ones. So, though it is true that younger consumers are getting a bad deal (until they grow older), consumer welfare is increased overall. That’s not a bad policy outcome.

    One need not like this economic rationale. Perhaps one could cook up a model different from Erikson and Starc’s that produces a different result. But, one can no longer say that there is no economic rationale for age-based pricing regulations. Modified community rating is justifiable on those grounds. I’m not touching the moral issues.

    P.S.: I expect this and other issues to arise at the Health Insurance Exchange Conference at Penn’s Leonard Davis Institute, April 11-12. I’ll be there and will have more to say about it another day.

    @afrakt

    Share
    Comments closed
     
    • Isn’t age a preexisting condition? I mean, one day I woke up and I was 60 and had no choice in the matter. What both sides of this argument omit is that age-based premium discrimination was included in ACA because it lowered, and lowered substantially, the cost of universal coverage (i.e., the subsidies). Universal coverage was deemed a higher priority than community rating. Or put another way, a significant part of the cost of universal coverage was imposed on the nearly old. For those not in the age 60 to age 65 cohort, I can assure you that it is health insurance hell.

    • The answer to this problem is competition, not regulation.

    • One can look at consumer surplus with community rating.
      One can also consider insurer surplus.
      With the 3 to 1 community rating of the ACA, insurers will garner more surplus from the younger workers to offset the deficits of the older workers.
      The ACA does allow for lower community rating, even pure community rating of 1 to 1 – everyone is charged the same rates.
      In our plan design, we have pure community rating.
      One may think that is unfair for the young and healthy, and they would be right.
      The good news is that for $300 a month, the young and healthy can buy paid-up coverage every month, thus reducing the amount of pure community-rate coverage, by the balance of the increasing paid-up coverage.
      In 36 months, paid-up coverage can builed to 25,000, cutting premiums 60%.
      In 50 months, paid-up coverage can build to $50,000, cutting premiums 80%.
      We worked with Milliman, an actuarial firm o n this concept, and we have their actuarials as support for the plan.
      As far as the older and healthier, they can a significant price break, and they deserve it! Dealing with health issues is enough of a problem twhich is exacerbated by experience-rated premiums.
      Don Levit

    • Dr. Goodman, you know health care superbly but you have never been a health insurance actuary.

      No insurance company competes for unhealthy people, most of whom are over age 55, unless they are subsidized heavily by government.

      The correct actuarial premium for an unhealthy 60 year old might be $15,000 a year, even with a high deductible. The only persons willing to pay that much are almost certain to file a claim.

      The vast majority of Americans who have secure health insurance have no age rating at all. The federal government, Medicare, and large corporate plans have no age rating.

      In a voluntary market, insurers will compete for young healthy persons and charge them low preimiums.

      In a voluntary market, many persons over age 55 will drop their insurance right before they need it most.

      This is not a peculiarity of health insurance. It happens in voluntary LTC insurance, voluntary disability insurance, and in term life insurance.

      If it were not for large group insurance and government coverage, about 80% of hospital patients would show up uninsured, and hosptials would go broke.

    • I don’t think it’s the same product.

      An insurance product for a young person and an insurance product for an older person, assuming median levels of health for both, represent different things, even if the facial terms of the policies are identical. Some of the differences work in favor of the older person, on average; the older person might reasonably expect to have a minimum of $10,000 in semi-routine health care costs over the course of a year, for example, and with that expectation the $12,000 annual premium seems pretty reasonable. The younger person might expect $500 in minimum costs, for a couple of checkups and maybe a walk-in to a clinic for a mashed finger or toe; to her, the annual premium is an incredibly expensive hedge against some very unlikely catastrophic scenario.

      Some of the variance will work in the opposite direction, of course. All that mandatory reproductive health coverage will be an amusing jest to the 65-year old woman, but might be a huge boon to the 24-year old getting ready to start (or strongly wanting to delay) a family.

      It’s not the facial terms of the policy that make it an identical product, it’s the utility derived from the product by each individual. It is precisely because those utility functions vary so widely that the appropriate locus for price is the individual market, or the voluntarily collective market, where providers can bid for customers. If some insurers are using “market power” to overcharge for their products, the answer is not to attempt to hide real prices and real utility behind some obscuring involuntary communitization of the product, it’s to remove barriers to entry and regulatory obstacles that are preventing other entrants from reducing those unearned rents down to zero, where they belong.

      There may well be lots of areas of price-setting where theoretically a busybody government could increase utility by obscuring the information conveyed by prices – “you young folks will be happy that we rob the young people in thirty years, when you’re old and the beneficiary of the robbery” is not a convincing argument to those who become aware of the disutility being inflicted on them, and find that injustice itself to be a source of negative utility. I gain no satisfaction from knowing that in a few years, the collectivists will be attempting to rob my children’s generation on my behalf, and I was just as graveled as a youngster to see my paychecks being slurped up so that retirees making more than I was could get another bump.

      Let markets work and let prices be what things cost. In the long run it is the only way to preserve the moral viability of an economic system. Every twiddle, every tweak, every veil drawn and every price obscured represents a destruction of information and a reduction in the amount of knowledge people have about the world. I see no payoff in that strategic approach to reality.
      \

      • I love corn flakes so much, I’d by $20 a box. I just get so much value from them! My friend, on the other hand, doesn’t care for them, but is willing to buy them at $3 per box. Is it the same product or not?

        • Not in the minds of the individuals concerned and perhaps differ in use as well.

          Al

        • Is the box of corn flakes the same product?

          Well, does it cost $100 to provide you your box, which you value at $20, but cost $2 to provide me with my box, which I value at $5?

          Insurance is not a fungible good. Every single aspect of the product differs from one consumer to the next – the use they will get out of it, the return on the investment they will see from it, the cost of providing it to them, the psychic benefit they derive from the ‘security’ it delivers, etc.

          Bob Hertz, if only a few people can actually afford to pay what health care would cost, then collectivizing the problem does not solve it; it simply adds a layer of abstraction and obfuscation.

          But it is not true that nobody other than the rich could afford premiums. It may well be true that relatively few people can afford premiums for insurance policies that cover all the health care that the person would ideally like to receive. But that is the case with just about every area of economic life; I’d like three different Bugatti roadsters, one for weekends, one for commuting, and one specially for date night. The fact that society cannot afford to provide that for everyone, or even for many people, does not mean that we can’t afford cars; it means that demand for a good is effectively infinite, which is what prices are good for. They prompt us to allocate our resources to the things we prioritize, and the prices tell us how other people prioritize their own resources, so that we can self-sort into arrangements that maximize the benefit to everyone.

          Does that suck for poor people? Yes, it very often does. The solution, however, is to identify the areas where our sympathy for the poor people reaches the level of desiring to assist them materially by providing a subsidy. Health care certainly seems like one such area; I have no objection to health programs to ensure care is given to the very downtrodden.

          But we cannot all be downtrodden; if we must subsidize everyone, then all we are really doing is saying that we want more than we can afford, and so we demand that other people give us free money. The supply of free money is very, very limited.

          • Just because one’s interaction with a product differs from that of another doesn’t mean the product itself is different. Otherwise, we can say the same thing of everything, including commodities.

            • “Just because one’s interaction with a product differs from that of another doesn’t mean the product itself is different. Otherwise, we can say the same thing of everything, including commodities.”

              But if in combination with that, the cost of producing and delivering the product varies, not trivially because of something like geography, but fundamentally and by several orders of magnitude because of huge differences in the consumers’ characteristics, then the product IS different.

              If healthcare-for-Austin costs $1,000,000 per year to provide, and healthcare-for-Robert costs $10,000 per year to provide, then a certificate promising to provide healthcare for one generic person for one year does not have a predictable value and cannot have a predictable cost. Efforts to give it one, invariably and inevitably, mean obscuring information about actual costs of provision so that some sucker will take the certificate for you.

              When has obscuring information ever led to greater economic efficiency or correct decisionmaking concerning the allocation of finite resources?

            • The predictability of the value of an insurance product is a population-level concept.

    • Robert, in the free market health insurance that you describe, the only older persons who could afford coverage would be those who had large incomes and/or large savings accounts.

      What would happen to the millions who could not afford the actuarially correct premiums?

      In other areas of American life, we let social insurance fill in the gap to some extent. Very few Americans buy disability insurance, and it is not even offered to those with health problems like bad backs So Social Security has a disability benefit. Most Americans cannot afford LTC insurance, so Medicaid has a skimpy and troubled benefit.

      But leave aside the humanitarian issue. If private insurance were all we had, the hospitals would go broke. Their best customers would all be uninsured.

      Incidentally, I do agree that charging young people more for health insurance may not be the best way to correct this problem. A tax on rich young people and rich old people is a better solution.

      • Unfortunately, some people who want a free market health insurance situation, think the way to solve this is to deny people care. Simple as that. Turn them away from the hospitals.

        Of course that method of containing costs would come with other loss of freedom far more egregious than insurance regulation, and various security expenses raising the costs of health care.

        I think young people will be a lot happier with a system where they don’t diligently keep insurance for 20+ years, and then lose their good paying job at age 48, need an appendectomy at age 50, and wind up with $26,000 worth of debt at age 52 working a low wage job… with all those years of paying insurance amounting to NOTHING.
        Because that’s what’s happening now.

        • Maybe a big mistake was made when healthcare, because of the tax benefit, was tightly linked to one’s job.

          • Elhauge argued cogently for the de-linking of health coverage and employment nearly 20 years ago. See “Allocating Health Care Morally” 1994.

            We remain stuck in a Prisoner’s Dilemma. Some people are perfectly fine with that, though.

            • I agree, the whole employer provided health care insurance thing seems like it was doomed from the start… or does it seem that way only in 20/20 hindsight & it’s all unintended consequences?

              Of course it seems perfectly fine for those who remain continuously employed with good benefits throughout their lives, with no major illnesses, and no major extended job loss.

              But I think COBRA was the biggest indicator of a flawed system. IE: The need for COBRA should’ve been a big red flag, no?

            • Of course healthcare should be de-linked from employment. The day that happens we will see what is happening with HSA’s today, high deductibles, more careful utilization of resources and falling costs.

            • W:”Of course it seems perfectly fine for those who remain continuously employed with good benefits throughout their lives”

              Would it seem so fine if one recognized that some of the bells and whistles that were not needed were coming out of salary?

              Do you think food stamps is the biggest indicator of a flawed system in the delivery of food?

    • There are indeed both fundamental moral and economic reasons for re-drawing the envelope of social-health insurance. In any given “open enrollment” period (typically annual), ~5% of individuals account for half of health care expenditures, ~50% spend nil, and the rest of us are somewhere in between, sliding inexorably toward that high-util 5%. The core coverage frame should minimally be, say, 50-60 years, from the onset of adulthood/wage earner forward, with rational incentives for the private purchase of ancillary coverages — would could in fact remain private market actuarial/risk based. (See, e.g., Elhauge, 1994 “Allocating Health Care Morally.” None of this is exactly news.)

      The fact that Avik chose to douse his assertion with inflammatory FoxNews “generation warfare” allusions is telling. And unhelpful.

      • The immorality stems from the fact that no one sees or knows the prices, until after they’ve agreed to the services.

        Normally, that consideration is what keeps people from buying things they can’t afford, and, more importantly, spurs businesses to a) higher efficiencies and b) limits what they can charge.

        Since no one’s spending their own money, no one knows or cares what medical services cost. Hospitals charge accordingly.

        Free markets don’t work (aren’t free) without free flow of information, especially price information. No one can make a wise decision without knowing cost and benefits.

        If milk were sold the same way with government support, it’d cost $97.33 a gallon. (Or $0.50 a gallon, but be completely unobtainable.)

        Here’s a positive, real-life example of the benefits of competition:
        http://www.ocpathink.org/videos/264

        • “Free markets don’t work (aren’t free) without free flow of information, especially price information.”
          __

          Yeah. But, Opacity = Margin. “Efficient Markets Hypothesis” and all that.

        • Well, that and it doesn’t matter if someone is presented with a price quote for $26,000 before the surgeon does the appendectomy. It’s not like you’re in a position to say “Oh we can’t afford that.”
          :o(

    • We could solve the hardest cases of medical debt with one simple reform:

      Make Medicare the payor of last resort for hospital charges.

      If someone did get a hospital bill and had no insurance or a very limited policy, here is what would happen:

      – the patient would have to make some payment based on their income.
      If their income was $30,000 , the payment might be $100 a month.

      – the remainder of the bill would be sent to Medicare. The hospital would get paid immediately, although at Medicare rates and not the grotesque charges they now invent for the uninsured.

      Hospitals would wind up collecting more money than they do now, and theno vicious collection agencies would be involed.

      The cost to Medicare would be about $25 billion a year. This is less than one half of one percent of payroll.

      Bob Hertz, The Health Care Crusade

      note to Watermelonpunch……..you are correct that a form of
      Social Darwininsm lurks in the Republican ideology. Their unspoken assumption is that those who fail to get a good job with health insurance and to save money in an HSA are not very valuable citizens anyways, who cares if they die sooner?

      Actually they don’t die sooner at all. They live on in nursing homes at taxpayer expense anyways. Thus the Darwinist pressure to cut Medicaid also

      • And you could solve the problem in the first place by providing medicare for everyone, with the appropriate taxes and premiums.
        Everybody could then also figure out how much their care was costing (in theory at least).

        But this is not realistic.

        Under your plan, I would be highly tempted to drop my insurance plan. Or at least reduce it drastically. Thus, I would increase the chance of needing your plan. Because the only way personal plans really make economic sense is if I expect a hospital event (even to those with chronic illnesses). And even then they tend to be “catastrophic” to finances. In other words, the price tag on your plan might be a bit low.

    • In the fine points of my plan, Medicare for anyone would have a large income-based deductilble.

      Therefore, you might still keep a mini-med health policy for the $5000 or $10000 of exposure you still had.

      But you would not need a policy with a $1 million maximum benefiit. That is what Americans cannot afford.

    • “The predictability of the value of an insurance product is a population-level concept.”

      I am not entirely sure what you mean by this.

      The cost of providing insurance is predictable at the individual level, by reference to population-level actuarial data. IE, people who are male, nonsmokers, X years of age, with such-and-such medical history, incur expected median treatment costs of $Y, ergo the expected cost to provide the coverage this insurance policy promises, for this individual, is $Z. Charge the guy $Z plus fifteen percent, you’re in business.

      If you mean the cost of providing insurance can be averaged across a population of size N, by taking the costs for everybody and dividing it by N, well, sure. Except then you’re charging the young and healthy way more than their insurance is worth to them, and charging the old and sick way less, and that creates automatic resentment and automatic attempts to evade the market. It’s very difficult to socialize things that not everyone wants socialized; the people whose chumphood is relied on to make the system work tend to dechumpify pretty darn fast.

      • ‘It’s very difficult to socialize things that not everyone wants socialized”
        __

        Which, nonetheless, does not invalidate the principle.

        Fan the flames of resentment, you get what you get.

    • Austin,

      This assumes that the ACA mandate is strong enough. Interesting note in the paper:

      “Finally, modified community rating can lead to severe welfare losses in the absences of an effective mandate. Even if risk-adjustment were perfect, so that insurer costs did not differ
      by enrollee age, differences in preferences alone can lead to this market to partially unravel. If consumers are allowed to opt out of coverage, modifed community rating restrictions will lead the most price-sensitive consumers to opt out.”

      A weak mandate can negate the surplus gain (and there’s reason to think that the ACA’s mandate, at least for the first two years may be too weak). Also unclear how generalizable this is to the national level.

    • Key Insights Thus Far:

      1) “Except then you’re charging the young and healthy way more than their insurance is worth to them, and charging the old and sick way less, and that creates automatic resentment and automatic attempts to evade the market.”

      2) “Finally, modified community rating can lead to severe welfare losses in the absences of an effective mandate. Even if risk-adjustment were perfect, so that insurer costs did not differ
      by enrollee age, differences in preferences alone can lead to this market to partially unravel. If consumers are allowed to opt out of coverage, modifed community rating restrictions will lead the most price-sensitive consumers to opt out”

      It’s far from clear that the aggregate-welfare-surplus calculations derived from the model are generalizable beyond the boundaries of the model, much less in all hypothetical age-based transfer-schemes, still less in under the actual incentives and constraints that will shape that insurer, employer, and consumer behavior under the ACA.”

      Given 1) and 2) – have you really concluded that the statement below holds for systems beyond the boundaries defined by the paper – such as real consumers acting under the incentives and constraints under the ACA as it will actually be implemented?

      “The crucial finding is that the decrease in consumer surplus of younger consumers is more than offset by the increase in surplus for older ones.”