• Erosion of confidence in the confident market solution

    In a prior post I described James Capretta’s plan that could replace Obamacare’s exchanges with lighter-touch regulations. I also wrote that his plan seemed to be identical to that offered by Yuval Levin and Ramesh Ponnuru. An essential, common feature of the approach offered by these gentlemen is that they do not include competitive bidding. That is, the premium subsidy for consumers in the individual and small-group markets would not be a function of market premiums; it’d be a flat tax credit.

    To these authors, I presume this is a feature. By stepping away from the competitive bidding design of the Affordable Care Act’s (ACA’s) exchanges — in which subsidies are tied to to the premium of the second cheapest silver rated plan — their proposal does not require an administrator to take and manage bids from qualified plans. With this obviation, many regulations fall away. The market is unleashed! (To be fair, a few, other regulations would or could be put in their place.)

    And yet, I think a flat tax credit is far from the most efficient design. One should not abandon competitive bidding and exchanges eagerly. The fundamental reason is that we do not know the right level of tax credit. It cannot be determined administratively. Capretta suggests starting levels “of about $5,000 [for families], and for individuals, about $2,500,” and then indexing them “to grow with some measure of inflation.” Is this exactly how much support should be offered individuals and families for purchase of health insurance? Why should it be identical across markets, which exhibit considerable variation in health care costs and premiums? To what measure of inflation should it be indexed?

    Similar puzzles have arisen with respect to Medicare Advantage plans, for which subsidies are set administratively. We know what happened. They shot up well above what plans reported was necessary to provide the Medicare benefit. The ACA was supposed to change that, and to a large extent it did. But the administration also layered on a quality bonus program — the value of which has been questioned by MedPAC and the GAO — that helped keep payments above cost, providing yet another example of the difficulty of administratively setting and maintaining payments at an efficient level. Capretta’s argument against exchanges, which necessarily drove him away from competitive bidding, is that they invite price controls. The lesson from Medicare Advantage is that administratively set subsidies invite bureaucratic meddling too, which is no less damaging to the efficiency of markets.

    Competitive bidding, which I’ve covered extensively, is an alternative to guessing the right subsidy level, as it ties that level to the actual cost of plans (or a plan) in a market. Moreover, if plan bidding is conducted on a market-by-market basis, it permits subsidies to vary with the geographic cost of care. This provides protections for consumers, as it keeps subsidies in line with their actual premium costs; and it provides protections for taxpayers, as it prevents subsidies from running amok, as they have for Medicare Advantage plans.

    One might be concerned that competitive bidding implies a defined benefit, as opposed to a defined contribution (e.g., in the form of a pre-determined tax credit), but it need not. One can retain competitive bidding in a defined contribution regime, though possibly at the expense of the consumer protection I just described. My point is that if one’s interest in a pre-set tax credit, such as that Capretta proposed, is because it’s a defined contribution, one need not abandon competitive bidding to achieve that.

    To be sure competitive bidding requires some additional regulations. Plans must bid their costs for some set of benefits. Thus, one needs to specify a minimum benefit standard and establish standardized plans for the purposes of bidding. But this also aids competition. Choice among many different options that vary in many different dimensions is difficult for consumers. Such an environment weakens competition, as consumers are forced to rely more heavily on inaccurate heuristics. This is the basic premise of a managed competition design, of which competitive bidding a one variant. Exchanges so designed are good for competition and consumers. One can always argue what the minimum benefit and standardized designs should be, of course.

    Perhaps one might be reasonably pessimistic about exchanges, given the poor roll-out of healthcare.gov and some of those in specific states. There’s no denying this has been a disaster. And yet, we also have examples of well-functioning exchanges. Kentucky’s works well, for example. Enrollment in the California exchange has been brisk. Medicare has a perfectly fine, exchange-like environment on medicare.gov for Medicare Advantage and Part D drug plan selection. There are also private exchanges that serve businesses and individual markets. I’m no fan of a big bureaucracy screwing up a large IT project such as healthcare.gov, but I don’t think that condemns the exchange concept; it condemns the administration overseeing that bureaucracy.

    My final concern about abandoning exchanges that harness a competitive bidding design is that I do not find doing so to be a conservative notion at all. For one, we know competitive bidding works. It’s a sound theory that has been tested. Capretta, Ponnuru, and Levin know this. Medicare’s drug benefit program — Part D — has a competitive bidding design that has performed well, and Capretta himself has praised it. The Federal Employees Health Benefits Program also has competitive bidding design elements and has been cited by conservatives, Capretta included, as a model.

    Competitive bidding is also central to any Medicare premium support proposal worth considering. In this context, Levin once called it “the confident market solution.” I agree! With so much evidence of its soundness, and having expressed confidence in it, I’m surprised to see thought leaders for right-of-center policy abandon it. Have they lost the courage of their convictions? If there is a clear line of conservative logic from the confident market solution to an administratively set subsidy, I am not aware of it.


    Comments closed
    • Having a flat tax credit will probably cause insurers to offer plans exactly at that price point. The idea is probably to have a flat rate for the poor and working class, then make insurance companies to trim benefits to get to that level.

      The approach in the ACA is to have the government set the minimum standards on benefits while the market determines the prices. The “conservative” idea is to make the government set the pricing standards and the market set the level of benefits.

      My theory is that the the flat tax credit approach is appealing to conservatives because they care more about government spending than they do about providing comprehensive coverage to those who don’t earn enough. They set a ceiling on government spending even if it creates a less efficient market. They would argue that if someone doesn’t make enough money for a comprehensive insurance plan, then it is their fault and not the government’s responsibility. They should be lucky to be receiving any subsidy at all.

      • Not to mention that whenever they actually flesh out their plans with numbers, the flat tax credit is ludicrously low for someone who actually wants to buy insurance with it.$2500? That’s nice. And with another $3500, why, I could actually buy the cheapest Bronze plan offered in my region. $5000 for my whole family? Great, now all I have to do is come up with the other $9000 and I’m golden.

    • A lot depends on who the tax credit is for – universal, only those who don’t have employer-provided care, or only the poor (however defined). Different target groups will generate different tax credit levels, I’d assume.

      My own 2 cents is that we’d be better off with a blend of tax credits for those that can’t afford insurance on their own (including converting Medicaid to tax credits) and something the Bush administration proposed, giving a ‘standard deduction’ for the purchase of health insurance. As to what qualifies as health insurance, I’d propose that whatever any given state chooses to recognize as insurance qualifies for residents of that state, with sufficient flexibility to include health sharing ministries, fixed-benefit policies, and other arrangements.

    • Austin
      One advantage of competitive bidding, as you note, encompasses the incorporation of local factors such as geography and how such factors impact paid price.

      Have any studies been done prospectively looking at implementation of a bidding system and the ultimate premium paid vs, a predicted premium using as an index x local adjuster?

      If Capretta’s plan pays $2500 and adjusts, would the simplicity both ameliorate complexity of administration, albeit not perfectly, but get closer to your ideal?

    • Great column! You’ve hit most of the key issues. I just want to add that there is a middle ground on the issued of standardized vs free-form benefits. In the FEHBP plans bid (all bids accepted) against a benchmark taken from the all-plan average of bids. Plans of course want to hit the “sweet spot” but they don’t know it with precision. The key point is that this is a target that moves over time and that is driven by plan benefit design and consumer purchasing decisions. The consumer pays the full marginal cost of a plan with rich benefits. Depending on how the credit is designed, savings will be split between plans, consumers, and the government (in the FEHBP 75% of savings go to the government, in Medicare Advantage and Part D 75% go to consumers, a better ratio in my view, and refundable credits should be considered). In this model, benefits need not be standardized and widely different types of plans may compete (e.g., HMOs vs high deductible). One could impose a minimum actuarial value standard without specifying benefits at all (my own preference would be this plus a tough standard for catastrophic protection, loophole free, and not much else). There can also be risk adjustment mechanisms that adjust the government share based on the enrollees in each plan. So there really is high flexibility in system design and level of regulation to be imposed. Since there are ways to simplify consumer information even in a complex system of choices, that need not be a problem either.

    • As Cardinal Fang points out, the greatest inequality in American health insurance is between those who have employer paid coverage and those who do not.

      Assume that two workers each make $60,000.

      One is in a strong union or generous govt agency or big company, and gets $15,000 of family health insurance for free.

      The other does NOT get a raise to $75,000 to help him buy health insurance. He has to find a much cheaper policy or gets no insurance at all.

      The Capretta/McCain flat credit would give the second worker $5000, and he would still have to find a cheap policy but he is a little better off.

      The ACA is more complex.

      If the second worker is below 400% of poverty, then he pays about $6000 for health insurance and the government pays the other $9000.

      If the worker makes over 400% of poverty, he is just as screwed as he is today.

      No one has a perfect formula!

    • My complaint about competitive bidding is different (but then, I’m sane, unlike Capretta/McCain).

      My complaint is collusion, monopolism, cartels. There are precisely three companies offering health insurance on the individual market in my county. This is likely to drop to two, or even one. When it drops to one, that one can *set* the price of the “second lowest cost silver plan” and write its own paycheck. Already, it only requires two phone calls for the three companies to collude to raise prices (and therefore raise subsidies).

      Health insurance is quickly turning into a monopoly industry. Health care has already been taken over by large doctor/hospital conglomerates, and there are now few enough that cartel-like behavior is quite possible.

      If the alternative is monopolies and cartels, then I prefer outright government-run central planning.