Last week, a group of eight health policy wonks presented a plan for health care reform at the American Enterprise Institute. Writing at The Incidental Economist, Century Foundation fellow Harold Pollack objected to a central tenet of the AEI plan, namely the introduction of “individualized premiums” Pollack argues that the Affordable Care Act ended the practice of “explict or implicit discrimination against the sick, injured, chronically disabled, or others with high expected costs.” Undoing that change is “a gigantic blunder.” Boston University health economist Austin Frakt then took issue with Pollack’s post. Pollack and Frakt exchanged four posts in all.

We’ve reprinted Pollack’s initial post below. Interested readers can view the entire exchange at The Incidental Economist.

Eight distinguished figures in health policy presented an ambitious health plan at the American Enterprise Institute this week. I suspect that you’ll be hearing more about this at TIE. I want to make one quick comment.

Bullet-point slides present the central pillars of their plan. My friend Hank Aaron was brutal in his role as AEI discussant:

I find it exceedingly difficult to discuss this proposal for three reasons. First, we are in the midst of a great national effort to implement systemic health reform. That is the business of the day. This proposal seems irrelevant to that debate as it has little in common with any proposal now under discussion by either the political left or the political right. Second, it has design features that I find unappealing but, more importantly, that I am sure would have little appeal to the general public. Third, it leaves unspecified so many essential program elements, that I find it hard to know just what we are talking about.

And that was just the first paragraph! I share Hank’s basic perspective, though I’d be gentler. Political relevance is not the only lens through which such proposals should be viewed. It’s useful to see an imaginative alternative presented, from such accomplished figures. Even the flaws and the details left unspecified are helpful. If nothing else, the gaps give us greater sympathy for congressional staffers tasked to write the hundreds of pages of junk DNA in health reform bills.

It’s also useful to be reminded of the need to curb regressive tax expenditures for employer-provided coverage. The key obstacles here are interest-group politics and knotty transition problems, not lack of analytic clarity. ACA’s Cadillac tax was, in its way, a brave effort to bring greater efficiency to public policy.

Table 3 of their paper presents the deductible structure of a basic, income-related high-deductible plan as they conceive it. I find it clarifying. I suspect many TIE readers who admire the incentive-structure of high-deductible policies would hesitate once they viewed the actual details of at least the most basic of such policies.

Among families with no unusual health burdens, the AEI plan provides pretty full insurance for four-person families with incomes below $33,000. Things get daunting when incomes get much higher. A healthy family with an income of $67,050 would face a deductible exceeding $20,000. At the $89,000 income level, deductibles exceed $35,000, and things escalate after that. Presumably a supplemental market would layer something on top of this. Even so…

The authors are particularly exorcised about preventing “the forced transfer of resources from healthy poor consumers to the sick rich.”

To advance this principle, their basic high-deductible insurance plan provides protections on a sliding scale for “extremely burdened” families predicted to experience high medical costs. The plan would impose no deductible and low copayments up to 300% of the poverty line ($67,000), and then rapidly raising the deductible when incomes reach six-figure levels. An “extremely burdened” family at 700% of the poverty line (an income of $156,450) would face a deductible of about $55,000 with 20% copayments. An “extremely burdened” family with an income of $223,500 would face a deductible of $145,000.

I’m not sure how these extreme expected medical burdens would be calculated, how intermediate burdens would be addressed. Many details remain to be specified. Yet the underlying principle is clarifying: Low-income sick people would receive large financial subsidies. These subsidies would be dramatically reduced as one moved up the economic scale.

Eliminating community rating

I believe their plan’s most troubling feature is noted in one bullet point, which would overturn a central pillar of the Affordable Care Act.

Eliminate community-rating and allow individualized premiums to eliminate adverse selection and end the forced transfer of resources from healthy poor consumers to the sick rich.

Despite the ostensibly progressive frame, this strikes me as a gigantic blunder .

One of ACA’s signal accomplishments was to provide an emphatic statement that prevailing practices within the individual and small-group insurance market must change. It is no longer acceptable for firms to follow a business model which relies on individual-level risk-selection through explicit or implicit discrimination against the sick, injured, chronically disabled, or others with high expected costs. Such practices expose people to unacceptable classification risk. The underwriting process also exposes people to unwarranted indignities and implicit deterrents to seeking care.

ACA sent a clear signal to insurers—and to regulators, too—that the legal framework has changed. It will now enforce different and better social values. ACA does not impose complete community rating. Premiums are allowed to vary by age and smoking status. Beyond that, though, insurers are not to go. ACA is designed to dismantle the basic structure of medical underwriting. It rejects the idea that people should be charged higher premiums based on higher expected medical costs.

Whether health reform can fully deliver on this promise is another matter. Community rating requires an individual mandate and perhaps other mechanisms to ensure an adequate risk pool. Depending upon the details, young and healthy people face some incentives to “go bare” rather than to pay community-rated premiums. Ironically, community rating brings its own self-sabotage. Uninformed healthy consumers might believe—incorrectly, in the case of ACA—that they go uncovered, simply waiting to “sign up” in the event of serious injury or illness.

Community rating can fail in the absence of such effective mechanisms. New York and other states that implemented strong community rating in the absence of individual mandates have seen serious problems. ACA’s CLASS Act sadly imploded in the absence of such mandates, too, leaving a gaping hole in the areas of disability and long-term care.

ACA’s most immediate challenge in this area concerns allowable age-gradients in premiums. ACA already includes substantial subsidies for young and healthy people who have modest incomes. Some of these subsidies improve the risk-pool within the new exchanges. Others do not. Age-related options to purchase catastrophic coverage brings people in. Allowing young adults to remain on their parents’ insurance plans is sound policy and politically wise. Yet as Adrianna has noted, this does not help health insurance exchanges with their risk-pool difficulties.

It might be wise to allow greater age-gradients in premiums, with implicit or explicit general revenue subsidies into the new exchanges to make these arrangements work for older people. Allowing the system to bend in these ways is quite different from allowing insurers to individualize premiums outside the generic characteristic of age, or to legitimate a medical underwriting process that most Americans rightly consider cruel and unfair. This strikes me as a natural compromise between the ACA and what these authors are trying to accomplish. This may be the best remedy should too few young people respond to the mandate.

Individualized premiums reduce some incentives for cream-skimming, but create or perpetuate others. As with the ACA, any real-world insurance system modeled on the AEI plan would require (as-yet hypothesized) risk-adjustment systems, essential benefit regulations, and daunting fine-print. There is no simple and elegant solution to these inherent complexities.

References to “the wealthy sick” occur conspicuously often in the accompanying paper. I wonder how many such people there really are, and how many dollars we’re really talking about. Particularly if one considers things in a lifecycle perspective, there are plenty of other ways for the rich to help the poor, and plenty of ways to enact more progressive fiscal policies such as raising estate taxes on the more-than-sick rich to tightening the home mortgage deduction.

To my knowledge, no wealthy democracy implements universal coverage through this sort of individualized premium. One might ask why. Most of our peer democracies provide universal coverage at reasonable cost without undue administrative complexity within community-rated systems.

Maybe the “healthy poor” in France, Britain, and Canada unfairly subsidize the “sick rich” in the same pathological way these authors lament. Maybe some health policy experts visited the American Enterprise Institute lamenting the regressivity of these social-democratic arrangements. If so, I missed it.