Response to Jason Furman

We are pleased that Jason Furman responded to our proposal by recommending that the book (on which we base the proposal) should be “required reading by specialists and non-specialists alike” and noting that he “would be perfectly happy if [our] proposal were adopted.” Both comments are extremely gratifying to receive from a skilled and insightful economist, and particularly from someone who was involved—at the highest levels of the Obama Administration—in crafting national health care policy. We're tempted to stop our response here.

If only Furman had, too.

But in the remainder of his essay, Furman critiques our proposal… and also complains that it can never get enacted. This pairing reminds us of the old joke about people critiquing the culinary options at a resort: the food is terrible… and such small portions!

We offered a two-part proposal for U.S. health insurance policy: (i) universal, automatic, basic coverage that is free for the patient; and (ii) the option to buy supplemental coverage in a well-designed market. Furman appears to have only one substantive critique with this proposal, which is the lack of cost sharing in the basic plan. He asks why everything in the basic plan should be covered for free, given the substantial body of evidence that cost-sharing is a “proven tool for reducing costs without worsening outcomes.”

We have no disagreement with Furman's description of the evidence on the impacts of cost-sharing. But cost-sharing cannot serve its cost-reducing function if most people and/or most expenses end up exempted from it. And, as we describe in our original Point piece, this is what has happened in countries around the world that have tried to introduce cost-sharing in their basic plan.

Furman takes note of our answer but argues that “their discussions of the experience of other countries feels like adding a political constraint to their optimization exercise, something they eschew in developing their overall approach.” We certainly agree that it would be inconsistent—and unpalatable—to pick and choose only some political constraints to respect. But our argument is not that we shouldn't have cost-sharing in the basic plan because it's politically unsustainable. Rather, our argument is that cost-sharing in the universal basic plan is substantively in tension with the very purpose of the existence of universal basic coverage: to provide access to essential health care, regardless of resources. There are always going to be people who cannot—or may not—be able to afford even small copays. This is why—in order to accomplish the purpose of the universal basic coverage—countries have found themselves compelled to issue an enormous set of exemptions. As we say in our original piece, cost-sharing in the basic plan is on a collision course with itself.

Furman further counters that a better approach would be income-related cost-sharing as originally proposed by Feldstein (1971) and expanded upon in Furman (2007). This is a well-taken comment that we have received from a number of policymakers whom we have talked with about our proposal. In principle, we have no objection: if we could find an efficient way to identify everyone who is constrained by cost-sharing for basic coverage, and exempt only them, the tension is resolved. In practice, however, we worry that the “administrative obstacles and complications to implementing this in practice” (that Furman recognizes) may get us back in the same position we are currently in: trying to administer income-related health insurance coverage through Medicaid, and getting people confused about what they are eligible for, what they have to do to document that eligibility, and what they have to do to periodically recertify and maintain eligibility. For this reason, we opted for the administratively simpler version. But this is not the hill we want to die on: if Furman and others like him are enthusiastic about trying to build momentum for a proposal for universal, automatic, basic coverage with income-related cost-sharing and the option to supplement it, we would be happy to pull up stakes and declare victory.

We suspect, however, that Furman is not yet ready to lead this battle, because of what he sees as the political infeasibility of our plan.

As two academic economists who have never worked in government, to argue with Furman about what is and isn't politically feasible feels a bit like Don Quixote lecturing Sancho Panza on what constitutes chivalry. Yet, like Don Quixote, we cannot resist.

Our biggest objection is to Furman's critique that while it may be defensible for us to ignore political constraints, it is not defensible for us to ignore “social constraints”—by which he means that if basic were to remain actually basic it would be “subject to serious backlash” of the type that HMOs received a few decades ago. Furman suggests that a “basic” plan might look something like the current Medicaid plan, and that most people wouldn't be satisfied with that. We agree with him on both these points.

What he has missed, unfortunately, is that most people would in fact have much more than the basic plan. As we note in our original piece, we suspect that about two thirds of Americans—those who are covered by Medicare or by private health insurance through an employer—would want to supplement beyond the basic coverage. The basic coverage is a floor, not a ceiling. Those who want to supplement can do so through a “top up” system. In this system, the funding the government would have used to pay for basic coverage can go to a private insurer who provides the basic benefits but can also provide more—including shorter wait times, greater choice of doctor, less “gatekeeping,” and coverage of additional services not provided through the basic package. This is the way that supplemental coverage operates in a number of countries with universal basic coverage, including Singapore, Israel, Germany, Switzerland, and the Netherlands. And it's the approach the U.S. has taken in allowing people to opt out of the public Medicare program for the elderly and the disabled (and about half of Medicare enrollees currently do so). Thus, we think Furman's concerns about popular backlash are misplaced, and reflect a misunderstanding of our proposal that we have now hopefully clarified.

Much of the rest of Furman's counterpoint is devoted to the political infeasibility of radical reform and his own preference for incremental health reform, both because it is more likely to be feasible and because of the dangers of unforeseen disruptions with more radical reform. Here's where we stop playing Don Quixote and defer to Furman's expertise.

We simply note that even if our proposal is not currently politically feasible, there is still value to stepping back and trying to articulate—and hopefully creating broad-based consensus around—the ideal design. Who knows when the Overton window may change, and the first step in shifting it is to build support for policies outside of it. Moreover, even for those pessimistic about the possibility of ever achieving radical change—or committed like Furman to incremental reform for substantive reasons—it seems important to try to agree on the North Star towards which we are trying to gradually move.

In that respect, we end this exchange on a note of optimism. There seems to be broad-based consensus between Furman and us on what that North Star should look like. We trust he'll continue to try to help steer us there.