Thursday, January 26, 2012

Tax policy colloquium on 1/24/12 - Amy Monahan paper on healthcare

This past Tuesday, Amy Monahan presented her paper, "Will Employers Undermine Healthcare Reform By Dumping Sick Employees?"

The basic argument is as follows. Under the recently enacted healthcare reform (HCR), employers have an incentive to seek the situation where employees who are likely to be high-cost health insurance customers opt NOT to be covered by the available employer-provided health insurance (EHI), but instead to opt out and seek coverage under the public insurance exchanges that HCR provided for. Employers can't, as a legal matter, directly tell their high-expected cost employees to go elsewhere, but perhaps they can push them in this direction indirectly (without also inducing low-cost or healthy employees to help out - after all, the preexisting tax subsidy for EHI makes this a good deal for employer and employee alike). An example might be not covering diabetes or AIDS at all, and being relatively stingy on hospitalization coverage, but being quite generous with regard to substituting gym memberships.

The paper uses the term "targeted dumping" to describe this strategy of getting the high-cost employees to opt out and use the public insurance exchanges instead. But despite the term's harsh tenor, this seemingly must be good for the employees who opt out, since they do it voluntarily. (Leaving aside scenarios in which the high-cost employees would have been better off still if the employer had designed its plan based on the fact that dumping was impermissible.)

So what could possibly be wrong with targeted dumping? Two groups in particular might be adversely affected: others on the public insurance exchanges if there is an influx of high-cost people whose risks are being pooled with those who are on the exchanges simply because they don't have EHI available, and taxpayers who must fund the subsidies for those exchanges.

My own view is somewhat as follows. Ex ante health insurance given your preexisting health status (high-cost or low-cost) is best accomplished by separating out the distinct expected-cost pools and not having cross-subsidization, in which the low-cost pay net transfers to the high cost. This would address adverse selection by the potential insured, as well as its flipside, favorable selection by the insurers.

As it happens, I also support transfers from low-cost to high-cost people, just as from high-ability to low-ability people through the fiscal system. Only, cross-subsidization is not, at least in theory, the best way to do this. Running it through employer plans is likely to have adverse efficiency consequences relative to financing it through general revenues. In addition, cross-subsidization may have odd and anomalous distributional effects on who ends up paying, relative to running it through general revenues. So targeted dumping may actually produce a better equilibrium - if one ignores the political economy problem of whether explicit taxpayer subsidies to the exchanges will be as close to the right level as the implicit and somewhat hidden ones that are delivered via blocking targeted dumping and thereby achieving greater rather than lesser cross-subsidization within EHI.

So I would offer at least one-and-a-half cheers for allowing targeted dumping, subject to the admitted political economy concern.

Further complications arise from the fact that HCR, for obvious political reasons, was constructed on top of the existing EHI system, with all of its horrendous defects. This makes everything more complicated. One of EHI's many defects is that, by carving out the covered employee group from everyone else, it greatly worsens adverse selection problems outside of its domain. So when you have these two somewhat segregated groups, the people in EHI and everyone else, shifting people around between them can worsen the problems in one sector or the other. Thus, for example, if targeted dumping shifts a significant high-cost population from the EHI to the non-EHI sector, then (with uniform premiums) cross-subsidization may actually become a bigger issue in the latter sector than it would otherwise have been.

No comments: