He proposes,

20% Deductible/Out of Pocket Cap: The IRS snarfs 20% of your family economic income. 5% of it is an increase in taxes (but that replaces your and your employer’s current health insurance premiums). 15% of it goes straight into your Health Savings Account. That HSA is then used to pay all your family health bills. If your expenses in a year are less than what’s in your HSA, the balance is rolled into your IRA (or, if you prefer, returned to you with your tax refund check).

Single-Payer for the Rest: If your HSA is emptied and you still have more health bills that year, the federal government pays them. The main point, after all, is insurance: if you fall seriously sick, you want right then and there to be treated whether or not your wallet biopsy is positive.

Sin Taxes: on Tobacco, Gorgonzola, Three-Liter Bottles of Liquid High-Fructose Corn Syrup, Tanning Clinics (Melanoma), et cetera: Sin taxes (and, perhaps, someday general revenues) pay for an army of barefoot doctors and nurses and mobile treatment vans roaming the country, knocking on doors, and providing preventive and other long-run lifestyle services for free…

A Lot of Serious Research on Best Public-Health, Chronic-Disease, and Hospital Practices

The last point is something that I advocate in my book, so I can’t argue with it.

The idea of replacing all health insurance with a government-administered plan with income-based cost sharing is not so bad, either. It is reminiscent of Jason Furman’s approach. Except Jason has persuaded me that thinking of cost sharing solely in terms of a deductible is problematic–it creates perverse incentives in the vicinity of the deductible. Jason thinks of cost sharing primarily in terms of co-payment. (With a deductible, I pay for 100 percent of the cost of my health expenses up to, say, $10,000, and 0 percent thereafter. With a co-payment, I pay, say, 50 percent of the cost of my health expenses up to $15,000, 20 percent of my expenses between $15,000 and $25,000, 10 percent between $25,000 and $40,000, and so on.)

I don’t like the “sin taxes” and the army of doctors and nurses coming to examine lifestyles. The paternalistic aspect is troubling from a libertarian perspective, of course. But more importantly, my guess is that if you ran a pilot program with a control group getting no paternalistic interventions and a treatment group getting the paternalistic interventions Brad advocates, you would not see favorable benefit-cost results (maybe you would among the poorest 10 percent of the population). I would think you would want to determine cost-effectiveness before rolling out such a program on a national level, rather than the other way round.