The Individual Mandate is Probably More Efficient than Insurance Subsidies

As Daniel pointed out in his last post, there are several ways to overcome the serious moral hazard issues that arise when insurers cannot discriminate against pre-existing conditions (“must issue” rules) and must charge most customers the same price (“community rating” rules). Congress chose the individual mandate. Alternately, again as per Daniel, Congress could simply have bribed low-risk individuals to participate, or paid insurers the extra costs of covering high-risk individuals.

While it’s true these are potentially viable alternatives, my claim (together with my two co-authors, Jake Brooks & Brendan Maher) is that they are less efficient than the existing system. In addition, although I won’t focus on this point, they are likely impossible to pay for under existing budget rules, especially if Congress first begins by repealing the NITT, which is the extra 3.8% tax on investment earnings. So any claim (not made by Daniel, but potentially by others) that the “repeal and replace” could be politically viable using one of these strategies is probably snake oil.

Ok, to my (even wonkier) point. Most public finance economists assume that the mandate is *less* efficient than Daniel’s alternatives, but that it was chosen for, to paraphrase Jon Gruber, political convenience. To make a very longs story a little less long, the mandate is said to be inefficient because it’s a differentiated consumption tax, and those are supposedly never as efficient as the income tax.

Let’s put this a slightly different way. It costs money to care for the sickest or most expensive members of an insurance pool. How should you raise that money? In Daniel’s proposals, the money comes from general revenues, so mostly income and payroll taxes. Under the ACA, the money instead comes from extra premiums on low-cost insureds: in essence, a sales tax on health insurance.

If you read this blog, you likely know that economists prefer the income tax to a sales tax because a sales tax carries all the distortions of an income, plus it changes our preferences between taxed and untaxed items. This is famously known in the legal literature as the “double distortion” argument.

Problem is, there are now a set of well-known exceptions to the double-distortion argument, and critics of the mandate haven’t sat down to work through whether these exceptions apply. I (and Brooks & Maher) have. And we think they do.

Just to take one example, consider the possibility that most of the distortion of the income tax is in tax avoidance, rather than in real changes in labor supply. If so, the mandate may be much less distortive than the alternative income-tax hike it would take to replicate the mandate’s effect. Why? Because efforts to avoid the mandate would largely not reduce income taxes, and vice-versa. There is, in other words, no double distortion. Of course, this depends on some key, and currently unknown, factors. If it turns out that both income taxes and the mandate strongly encourage people to work “off the books,” then there are overlapping distortions, and this income no longer holds as strongly.

We expect to have a more fully elaborated version of our argument available for the public in a month or two. Watch this space!