On Saturday, the New York Times published an op-ed by Yuval and Ramesh that calls for reviving the 2008 McCain campaign’s health reform proposal.

Unlike wages, employer-provided health benefits are untaxed, giving employees an incentive to get the most expensive coverage available through their jobs. This drives up health care costs, but nobody in politics wants to fix the problem by simply taxing benefits.
In his 2008 presidential campaign, Senator John McCain proposed to untangle this knot through an insurance tax credit. Whether they get insurance from their employer or buy it themselves, workers would receive the same credit, creating an incentive to shop around for the cheapest coverage.
The proposal has a lot of virtues. Many without employer coverage would be able to get insurance on their own. With coverage less dependent on employment, people wouldn’t worry about losing coverage if they lost or changed their jobs. They wouldn’t have to worry, either, about their company cutting their benefits. And, as more Americans bought their own insurance, consumer pressure would bring down costs for everyone. Ultimately, companies could get out of the business of managing their employees’ health insurance altogether.

As the authors go on to observe, however, there is a danger that this tax credit would be disruptive, e.g., that it would encourage employers to drop coverage for their employees. Indeed, that’s part of the point of the proposal — to separate coverage from employment, thus encouraging transparency and cost consciousness while easing the problem of job-lock.

Yet the threat of disruption made the proposal a non-starter, and Yuval and Ramesh offer a modest revision of the proposal with that in mind:

So far, then, Republicans have proposed a big-bang insurance tax credit and no tax credit at all. The middle course is to make a gradual move toward an individual market. It would make the tax break a flat credit, as Mr. McCain proposed, to end the incentive to overspend. But it would let people use that credit to buy their own insurance only if they do not have access to a company plan. Those whose employers offer insurance coverage could use the credit only to offset the cost of that coverage.
This alternative sacrifices some advantages of the McCain plan. It would not do as much to shift control over insurance to workers. They would have to stay in their jobs to keep their existing plans. But it would cut costs and help people the tax code now pushes out of insurance markets. And it would do so, critically, without threatening the insurance arrangements of the satisfied majority. Over time, this reform could help the individual market grow and become more attractive to more Americans. Voters might then become receptive to relaxed restrictions on using the tax credit to exit the employer market.

What are the problems that might arise? Matt Yglesias suggests that adverse selection will prove a serious obstacle:

If you simply do what Ponnuru and Levin propose, every insurance company will be competing to make a product that’s attractive to young men with no chronic health problems and unappealing to everyone else. To turn this idea into an idea that actually works for people with medical needs you need to do three things. One, you need to prevent firms from turning customers away because of their health status or demographic characteristics. Two, you need some kind of regulatory definition of the minimum benefits that need to be offered in order to qualify as “health insurance” that’s eligible for the tax credit. And three, you need some kind of penalty for failing to enroll yourself in a plan to ensure the existence of a viable risk pool. What you need, in other words, is the Affordable Care Act and its regulate/subsidize/mandate tripod structure.

There are three separate issues here.

(1) Do we need to prevent insurers from turning customers away because of their health status or demographic characteristics? This isn’t necessarily true, particularly in a world in which the tax subsidy is offered on a more equitable basis, as Yuval and Ramesh propose. Consider the following from Mark Pauly:

The purpose of costly underwriting is to identify the high-risk applicants, but underwriting requires that every applicant be reviewed. It only makes sense to spend resources on a careful review if you think there is a large enough fraction of applicants you would want to decline. If subsidies enrich the pool of applicants with many more good risks, it becomes much less cost effective to screen, and so screening efforts will decline. It would also no longer pay as much for low risks to search aggressively for policies with premiums tied to their risk levels, if high risks are not segmented out.

The subsidies devised by Yuval and Ramesh can help shift insurance markets to a new, low-loading equilibrium. Well-designed high-risk pools, or perhaps a public reinsurance program, would go a long way towards solving what is left of the problem. James Capretta and Tom Miller have offered one possible approach.

(2) I agree that we need a definition of what constitutes insurance that is eligible for the tax credit.

(3) Do we need a mandate? A March GAO report offers a number of alternatives. I am comfortable with Paul Starr’s suggestion:

The law could give people a right to opt out of the mandate if they signed a form agreeing that they could not opt in for the following five years. In other words, instead of paying a fine, they would forgo a potential benefit. For five years they would become ineligible for federal subsidies for health insurance and, if they did buy coverage, no insurer would have to cover a pre-existing condition of theirs.

Under Levin-Ponnuru proposal, perhaps there could be a one-time enrollment period instead of Starr’s five years in the penalty box. Capretta and Miller suggested the following:

[W]hen these reforms are first implemented, there will need to be a one-time open-season enrollment period to allow people who have fallen through the cracks over the years to re-establish their rights by maintaining continuous coverage. Those who have forfeited their coverage would get just one chance to become insured under the new rules (though perhaps at higher rates than those who had not forfeited their rights); once the enrollment window closed, everyone would know that people who remain continuously insured are protected, and that those who choose not to become insured have taken a risk.

I’m not sure that this approach will work. Perhaps Starr’s penalty box is more realistic. But it is far from clear that a mandate is the only workable solution.

What we end up with is something that looks very different from PPACA, yet that doesn’t have one of its fatal flaws. As Daniel P. Kessler argued in yesterday’s Wall Street Journal, the subsidies on the exchanges create a serious problem:

Starting in 2014, subsidies will be available to families with incomes between 134% and 400% of the federal poverty line. (Families earning less than 134% of poverty are eligible for Medicaid.) For example, a family of four headed by a 55-year-old earning $31,389 in 2014 dollars (134% of the federal poverty line) in a high-cost area will get a subsidy of $22,740. This will cover 96% of an insurance policy that the Kaiser Family Foundation predicts will cost $23,700. A similar family earning $93,699 (400% of poverty) gets a subsidy of $14,799. But a family earning $1 more—$93,700—gets no subsidy.
Economists call large, discontinuous changes in program benefits like this “notches.” Although notches might be administratively convenient, they have terrible incentive effects.

These notches make climbing the income ladder a tricky proposition for middle-income households:

Fixing the notch is not so easy. To phase out the subsidy smoothly for families with incomes of 134% to 400% of poverty, the law would have to take away $22,700 in subsidies as a family’s income rose to $93,700 from $31,389. In other words, for every dollar earned in this income range, a family’s subsidy would have to decline by 36 cents. On top of 25% federal income taxes, 5% state income taxes, and 15% Social Security taxes, this implies a reward to work of less than 20 cents on the dollar—in economists’ language, an implicit marginal tax rate of over 80%. Although economists may differ on the effect of taxes on work effort, it is hard to fathom how anyone could argue that this will not reduce economic activity.

Yuval and Ramesh have offered a distinctive approach that should unite conservatives, and that should also appeal to moderates who understand that PPACA is unstable, expensive, and economically damaging.

One good outcome of the health reform debate could be that the sweeping nature of PPACA could defang criticisms of a McCain proposal once condemned as radical.