Identification of Causal Effects with a Bunching Design

Carolina Caetano, Gregorio Caetano, Leonard Goff +1

We show that causal effects can be identified when there is bunching in the distribution of a continuous treatment variable, without imposing any parametric assumptions. This yield…