Linde Institute/SISL Seminar: Itai Ashlagi, Stanford University
Abstract: Kidney exchange facilitates over 800 transplants per year in the United States. We show that, despite this success, kidney exchange markets suffer from market failures that cause the loss of hundreds of transplants per year, but could be addressed with simple policies. Our argument has three parts. First, we document that the market is highly fragmented, with 65% of transactions happening in small platforms, often within hospitals, as opposed to in large, national platforms. Moreover, there is smoking-gun evidence these small platforms often perform inefficient matches. Second, we propose a simple model, showing that inefficiency arises for two reasons: hospitals do not fully internalize their patients' benefits from participation, and current mechanisms do not give hospitals adequate incentives. Third, we estimate the transplant production function, the key primitive of the model, to quantify the inefficiency and design practical mechanisms. Our estimates show that hospitals' production scale is too small to match patients efficiently. Consequently, market fragmentation leads to a deadweight loss of at least 200 transplants per year. Simple optimal mechanisms and policies that encourage participation are likely to have large positive effects, but eliminating this inefficiency requires a combined approach.