Sepulveda, Facundo2007-06-182011-01-052007-06-182011-01-052006-091 921262 07 91442-8636http://hdl.handle.net/1885/45256http://digitalcollections.anu.edu.au/handle/1885/45256When the government cannot commit to withdraw from providing charity health care, as is the case when it faces the Samaritan's Dilemma, a public health insurance scheme can be Pareto improving. However, the large heterogeneity in the design of such schemes observed around the world begs the question of what characterizes the optimal public health insurance plan. In this paper, we examine the distortions created by three plans, nested in terms of the constraints they place on the individual's decision problem. We find that linking public health insurance benefits to the use of a certain type of health care, such as treatment in public hospitals, creates incentives against the efficient use of higher quality health care. When such constraint is lifted, but the public insurance scheme still determines a minimum level of coverage for each illness, first best efficiency is achieved. It turns out that placing constraints in the form of minimum levels of coverage for each illness is necessary for e±ciency. Removing such constraint decreases the relative price of high quality care for a subset of illnesses, and leads to too much high quality care used in equilibrium. This analysis suggests that the widespread practice of determining illness by illness coverage in public health insurance systems has an efficiency rationale, despite the administrative and informational difficulties that it entails.enSamaritan's DilemmaHealth insurance2006-09