Welfare effects of supplementary insurance: a comment.

This issue of the Journal features two papers, Community Rating of Health Insurance and Different Benefit Packages by Matthias Kifman and Alternative Health Insurance Schemes: A Welfare Comparison by Hansen and Keiding, that examine the welfare implications of permitting voluntary supplementary insurance in a regime with a community-rated, compulsory basic insurance package. This is a critical question in any country that either has adopted or is considering some mix of private and compulsory/public insurance, including Germany, the Netherlands and the US. The conclusions of these papers appear to be radically different. Hansen and Keiding (HK) conclude that a compulsory scheme with voluntary supplementation is likely to be welfare superior to the pure compulsory scheme, but that the compulsory scheme alone may also be welfare-dominated by a pure voluntary market insurance regime, using either a Hicksian or an average utility criterion of welfare improvement. By contrast, Matthias Kifman (MK) concludes that if the insurers that offer the compulsory basic benefit are also permitted to offer supplementary coverage, this can only benefit low risks at the expense of high risks. Hence, if one purpose of the compulsory insurance is to enforce a cross-subsidy from low to high risks, permitting supplementation is likely to be counterproductive. However, MK also concludes that high risks may be better off if the insurers that offer the supplementary benefit are subsidized while those that offer only the basic benefit are taxed, which seems counterintuitive, given the initial finding, that permitting supplementation without a subsidy benefits low risks at the expense of high risks. These apparently contradictory findings are possible because the two papers differ in their basic assumptions. Within their respective assumptions each paper makes a significant contribution. HK provide a welfare comparison between several alternative health insurance regimes: (1) an unregulated, voluntary market regime; (2) a compulsory and uniform universal regime with community rating; (3) a compulsory, universal, community-rated regime that permits private supplementary insurance. Their analysis offers two significant advances over previous papers on this topic. First, in order to make the welfare comparison between the community-rated compulsory insurance regime and the voluntary private insurance regime, HK derive the level of the compulsory insurance as the equilibrium choice in a median voter model of political choice. Modeling the compulsory coverage as an endogenous outcome of rational choice within the political process is an important step towards realism. More often, papers on this topic simply assume that the level of compulsory coverage is somehow optimally chosen by a benevolent, omniscient regulator (for example, Dahlby, 1981) or the political determinants of this choice are ignored (for example, Kifman