摘要：The model of Stiglitz and Weiss (American Economic Review, 1981, 71(3)) is the seminal analytical work on credit rationing. However, in a recent paper, Arnold and Riley (American Economic Review, 2009, 99(5)) claim that the distributional assumption on which that model’s main result depends cannot hold. This paper shows that Arnold and Riley’s result is an outcome of their implicit assumption of a one-period Bertrand game between banks. In more realistic modes of bank competition, in which banks have some degree of monopoly power, Stiglitz and Weiss’s result can hold.
