November 20, 2018

Ed Dolan and Chris Pope Debate the Future of Health Care Reform



Exit polls from the recent midterm election suggest over 40 percent of voters consider health care as the top issue facing the country. Yet after a unified Republican government tried and failed to repeal and replace the Affordable Care Act, the conservative vision for health care reform remains something of an open question.

Here at the Niskanen Center, senior fellow Ed Dolan has been beating the drum for the reform path known as Universal Catastrophic Coverage (UCC) for more than two years. But not everyone agrees. That’s why we invited Chris Pope, a senior fellow of the Manhattan Institute and formidable conservative health care policy thinker, to join Ed Dolan in a debate.

Be it resolved:

Universal catastrophic coverage is a reasonable path to universal and affordable health care.

PRO — Ed Dolan

People have many basic needs, including food, shelter, education, and access to health care, which, if unsatisfied, can make our treasured rights to life, liberty, and pursuit of happiness meaningless. Of these, health care poses unique problems for public policy.

To understand why, imagine a society in which no one’s resources fall below the poverty level. Regardless of whether those resources come from a state-sponsored universal basic income or from some combination of work, family support, and private charity, I argue that such a society could adequately meet needs for food, shelter, and so on but not health care.

There are two reasons for that. One is the highly skewed distribution of the demand for health care services. In the United States, the healthiest half of the population account for just 3 percent of services consumed, while the sickest 5 percent account for 50 percent. For comparison, if the need for food were distributed similarly, half the population would need just 120 calories in their daily diet (one small potato), while the hungriest 5 percent would starve on anything less than 20,000 calories (35 Big Macs). For the top 5 percent, health care consumption exceeds the national median income.

The second reason health care is unique is its uninsurability. To be commercially insurable, a risk must be unpredictable and an actuarially fair premium must be affordable to the insured. Because of pre-existing conditions and advances in genetic testing, health risks are highly predictable. As a result, for many risk-prone individuals, actuarially fair premiums can exceed income.

The combination of skewed need for services and uninsurability means there is no simple market mechanism that would give everyone access to health care. That leaves us either with a nation of medical haves and have-nots, or a major role for government.

CON — Chris Pope

I agree that healthcare needs cannot be met in the same way as those for food and shelter – but for a different reason. The basic need for food and shelter is straightforwardly satiable, whereas that for healthcare services no longer is. That is a consequence of medical progress and fundamentally a good thing.

While it might have been possible a century ago to provide pretty much all that medicine could do through a universal publicly-financed benefit, in a world of heart transplants, immunotherapies, and knee replacements this is no longer the case.

Private financing provides the best reward for innovation and efficiency, and should be employed to the greatest extent possible. Most of healthcare is insurable, and privately-financed health insurance does, in fact, cover most Americans and bear the majority of healthcare costs. Although healthcare spending in any particular year is of course concentrated in the minority of people who get very sick, this is not the same people every year. Most of us will be in that group at some stage of our lives.

The vast majority of people are therefore able to afford to purchase actuarially-priced health insurance before they get sick. The minority that is not are best dealt with discretely, through the provision of targeted benefits as we currently do through Medicare, Medicaid, and the exchange (insofar as it is becoming a well-focused high-risk pool).

There is only so much money you can raise from taxpayers to fund healthcare (public assistance ranges from 6.4 percent of GDP in Spain to 9.0 percent in France; USA spends 8.3 percent). Yet, because America targets its public funds on those who are unable to provide for themselves, it serves to supplement rather than supplant private healthcare spending, which is worth a further 8.9 percent of U.S. GDP, compared to 2.6 percent in Spain and 2.5 percent in France.

Non-universality of benefits is therefore essential to expanding access to better quality care.

ED: First, let me say briefly, I am not as comfortable as you with phrases like “most is insurable” and “vast majority.” Kaiser Family Foundation estimates that 27 percent of nonelderly have a declinable pre-existing condition. NIHCM data (cited earlier) show substantial persistence of high spending by several measures, e.g., only a quarter of those in top half of spending migrate to bottom half each year. Not to mention the fear of developing a chronic condition, job lock for those in employer plans, and so on. In short, insurability is not a small problem; it is THE problem.

But I would rather focus where we agree, namely, that public health care spending should be targeted on those who need it, and should supplement rather than supplant private spending. My preferred way to accomplish that is through universal catastrophic coverage, or UCC (brief outline here, details here).

UCC would pay for coverage in full for the poor, and would impose income-based deductibles on the rest. For example, a family with income of (say) $25,000 or less would have no deductible in its UCC; one with income of $75,000 would have a deductible of $5,000 (a little less than under an ACA silver plan), and one with income of $400,000 would have a hefty deductible of $37,500. High-income families might consider supplemental insurance to help with deductible costs, for which premiums would be modest since they would be covered by UCC for catastrophic events.

I also agree that there is only so much you should try to squeeze out of taxpayers. A RAND study calculates that the cost of a well-designed UCC plan would not be greater than what the government now spends on Medicaid, Medicare, ACA, VA, and deductibility of employer sponsored insurance. No new taxes!

CP: I think we agree that a two-tier system is inevitable and desirable, but disagree on how the tiers should be designed to interact.

Currently, public and private revenues each fund about half of healthcare spending. I would prefer to better focus public spending on the uninsurable, to reduce crowd-out of private spending. The UCC proposal would push in the opposite direction – greatly increasing the crowd-out of private spending by public funds.

The UCC would yield a financing system similar to France. The basic tier would be similar to the French “CMU complémentaire”, and supplemental insurance would be similar to the “mutuelles,” which fill in cost-sharing and allow individuals to receive treatment from hospitals and physicians that do not accept public fees as full payment.

It would therefore be fair to imagine that this reform would yield a similar degree of crowd-out of private funds that is seen in France: even if public spending were increased from 8.3 to 9.0 percent of GDP, private spending would likely fall from 8.9 percent towards the 2.5 percent in France. As a result, the healthcare sector could contract by 5.7 percent of GDP – a cut of over 30 percent or around $1 trillion!

Would that mostly serve to cut costs or to reduce access to care? Healthcare is a labor-intensive industry. The average primary care physician in France earned $96,000 in 2011, relative to $186,000 in the United States. A move towards collective bargaining usually tends to push labor costs up, not down. It’s possible that, by suppressing private demand for healthcare services, shifting towards the French system could reduce the demand for medical labor and therefore costs – but any such cost reduction would only happen as a result of greatly reduced care delivery.

ED: Yes, UCC, like the systems of France, the Netherlands, and other OECD countries, would offer full first-dollar coverage only to people with low incomes. There, people with middle- and higher incomes face significant personal contributions through deductibles and copays, or through premiums for optional supplemental insurance to cover what the basic tier does not. I see that as a feature of those systems, and of UCC, not a bug.

It is true that by ensuring that non-poor consumers have “skin in the game,” UCC deductibles might make them more careful health care shoppers. That would be all to the good, as far as I am concerned. However, empirical studies of the effects of high-deductible insurance show an impact on demand orders of magnitude smaller than 5.7 percent of GDP, and only a small impact on quality of care, via neglect of needed services due to cost.

To the extent that high-income consumers decided to manage their deductibles with supplemental insurance, the impact on total private spending would be further muted. However, I would point out that might not be everyone’s first choice. Other middle-and high-income consumers would probably handle their UCC deductibles through health savings accounts or simply by paying cash, as many people do for home and auto repairs.

Finally, I would add that any health care reform should include a full range of cost-saving measures, such as increased price transparency, reduction of administrative fragmentation, moves toward bundled payments, reduction of barriers to competition, and malpractice reform. However, those measures would impact both the public and private components of health care spending without having much effect on the balance between the two.

Finally, let me add that the public/private balance under UCC could be fine-tuned by tweaking parameters like the low-income cutoff and the deductible percentage (see here).

CP: The crowd-out of private spending would likely be enormous. According to the MEPS survey, 43% of healthcare spending is on people who use more than $35,000, an additional 11% is by “middle income” individuals spending $5,000-$35,000, and another 18% is by poor or low income individuals spending under $35,000. So around 72% of existing healthcare costs would be assumed by the government under the proposed arrangement – with additional covered expenditures likely being induced for individuals whose premiums and cost-sharing are eliminated.

Deductibles are a bad cost-sharing design, which tends to yield skimping on care rather than better shopping. As most expenditures are beyond catastrophic thresholds (single hospital admissions often cost over $50,000), deductibles would do little to reduce costs and waste. Deductibles are upside-down versions of effective reference-pricing approaches, which pay first-dollar for cheapest providers, but not costlier alternatives. Having the government pay all costs beyond the catastrophic threshold would also eliminate the benefits of managed care, which has generated 25% savings through Medicare Advantage. The chronically ill, who hit their deductible year-after-year, would face an enormous bill – and likely be excluded from the purchase of supplemental insurance due to pre-existing conditions!

Although Medicare covers 76% of its beneficiaries’ healthcare costs, according to the MCBS survey, 91% of its beneficiaries with incomes over $50,000 in 2011 had some form of supplemental insurance. Supplemental coverage would also inflate costs for taxpayers – potentially by a further 27% as we’ve seen with Medigap. The UCC’s proposed catastrophic benefit design would also have a perverse incentive that Part D’s drug benefit has experienced: supplemental insurers would have an incentive to arrange “rebate” deals with providers to inflate costs and get to the catastrophic threshold artificially early.

There would also likely be a substantial cost from the adverse labor market impact from the incentives to work less to qualify for better subsidies and to avoid the higher taxes.

ED: Your calculations suggest that my illustrative UCC formula ($25,000 low-income threshold and a 10 percent deductible rate) would put 75 percent of all health care costs in the catastrophic category, to be paid by government. I am willing to accept those numbers for the sake of discussion, and to agree that a UCC plan should aim for roughly the 50/50 public/private division of health care spending we have now. I see three ways to do that.

  1. Tweak the UCC formula. As I explain in my long piece on affordability, the 25/10 formula is just a simplified illustration. Instead, say, put the low-income threshold at 135 percent of FPL, add some copays to increase maximum out-of-pocket costs to 15 percent of eligible income, and if needed, add a modest income-based premium. Based on the RAND study, that would get us to 50/50 and would still meet my criteria of universality and affordability.
  2. Control costs. Deductibles should not be the main tool, just one among many. As I said in my last segment, use a full range of cost control measures. By all means include reference pricing and measures to counter gaming of the system by providers through rebates or other gimmicks to push spending up into the catastrophic range. Throw in all your favorites. Cost control is hard, but we can learn from other countries that manage somehow to do a better job than we do. If we can cut a couple of points off total expenditures, then we could ease up on some of the UCC parameters like the low-income threshold and still hold the government share to 50/50.
  3. Give up on the goals of universality and affordability. But I don’t want to do that and I don’t think we have to.

CP: It might be helpful to evaluate the merit of the UCC by comparing it piecemeal with the status quo.

Firstly, it would see a big increase in taxes and out-of-pocket costs for higher-income individuals currently enrolled in Medicare and Employer-Sponsored Insurance. There might be a case for that, though I doubt it would be worth the political fight.

The lower tier would be pretty similar to the status quo on Medicaid, though would redress the inequality in resources between states. That would be a good idea, but would be politically difficult, as last year’s Graham-Cassidy effort found.

The middle tier seems similar in its structure of subsidies to that currently on the exchange, but would essentially overturn recent deregulatory reforms by the Trump administration. Although it is good to have a safety-net of subsidized coverage available to low-income individuals and those with pre-existing conditions, people should be allowed to purchase cheaper unsubsidized coverage if they prefer it.

Why not let people enjoy the savings from purchasing cheaper unsubsidized plans? The UCC implicitly mandates the purchase of very comprehensive insurance – especially for those with even modest incomes. A comprehensive benefit package, without preferred networks or gatekeeping of costly specialty providers, but with a high deductible, seems like the worst of both worlds from the point of view of efficiency. This is likely to cause the cost of the medical delivery system to increase, as Medicare has for the past 50 years – insulating the healthcare industry from most incentives to reduce costs.

In short, I believe the hope for improving the value of American healthcare likes with empowering managed care, not by increasing individuals’ exposure to out-of-pocket costs.

ED: Time to sum this up. Your comparison of UCC to what we have now is helpful. Yes, we have similar bits and pieces in place within our current kludge of a system, but it is rife with donut-holes, elevated administrative costs, and other problems born of fragmentation. One of the things I mean by universality is a simpler plan with fewer moving parts.

Gateways, preferred networks, and managed care? I don’t see why not, as negotiable parts of a cost control package. A “very comprehensive” set of benefits? Don’t get UCC mixed up with Sanders. We don’t need coverage for designer eyeglasses.

Should we let people opt out of UCC? That would be automatic. Unless there is a mandatory premium, everyone would be nominally covered, but people could just choose not to submit claims. The only caveat is that a UCC plan with a high mandatory premium (not the version I prefer) would potentially raise an issue of adverse selection. (Too complex to discuss here, but I touch on it briefly elsewhere.)

Allow providers to opt-out? Sure. Even Sanders allows that (see Sec. 303 of his bill). As for “cheaper, unsubsidized plans,” it is hard to see how there could be either supply or demand for anything other than supplemental policies that took advantage of the reinsurance implicit in UCC.

The bottom line: Yes, reforming U.S. health care is both technically and politically difficult. What we have now is such a mess that reformers need to juggle many problems at once. Still, I am not ready to give up on the goals of universality and affordability. Personally, I see UCC, with strong cost controls included, as a good way to get there. But, if you or anyone else has a different way, I am ready to listen.

CP: Healthcare policy is necessarily a kludge with many moving parts. The financing of over $3 trillion per year in medical services inevitably involves a vast array of political, labor market, geographic, and social considerations, which must be weighed against each other. Countries differ in the extent to which these messy trade-offs are determined by consumers, or made by government officials and hidden from public view.

Things like administrative costs, donut holes, and the fragmentation of care delivery are symptoms of trade-offs, rather than inherent evils. They are the consequence of attempting to give patients unconstrained access to providers, under a system of third-party payment, while attempting to control costs for taxpayers or those purchasing insurance. These trade-offs are made best if insurers are responsible for the full spectrum of individuals’ medical costs, and held accountable by consumers in competitive markets.

The most pervasive problem, as I see it, is that when government intervention displaces market forces, the logic of collective action in lobbying often leads it to favor the interests of providers over patients. This is why we have so many forms of hospital protectionism, such as network regulations that cripple the negotiating power of insurers, which generate ever-higher costs. If the UCC covers all expenditures above deductibles for major surgical procedures, it is hard to see how more cost-effective competitors could ever get a foothold in the marketplace.

The government can play a positive role by filling in gaps in funding, but this is best done through by risk-adjusted vouchers (as with Medicare Advantage), so that trade-offs are made for the benefit of beneficiaries rather than the profit of well-organized medical providers. To the extent that public funding serves to displace privately-funded healthcare, an expanded government role should clearly be seen as harmful.


With that, an extremely informative and good spirited debate comes to a close. To join the discussion, be sure to follow Chris and Ed on Twitter where the policy debate never ends.