Growth of live-donor kidney exchange has stagnated although an increasing number of transplant centers are facilitating exchanges through large national exchanges, in isolation or in small consortiums with other centers.
We first discuss the problem of dynamic matching in an kidney exchange program. We show that matching greedily essentially does not harm the match but the lack of collaboration between kidney exchange programs does.
Next we discuss the kidney exchange market in the U.S., which enables about 800 transplants annually. Using novel administrative datasets, we document that the market still shows clear signs of inefficiency because of fragmentation due to hospital participation behavior. We analyze these facts using a model that views an exchange platform as a producer of outputs (transplants) using inputs (donors and recipients) supplied by participants (hospitals). It identifies two market failures that can cause inefficiency: sub-optimal rewards for supplying inputs and agency problems in supply decisions. We then estimate a production function to provide quantitative results in the kidney exchange market. Our results show that the market produces about 400 fewer transplants than feasible because individual hospitals that conduct the majority of transplants are at an inefficiently low scale. While an optimal mechanism increases the number of transplants, it is also necessary to solve agency problems in hospital decisions to eliminate this inefficiency.