Partial identification in nonseparable binary response models with endogenous regressors

This paper considers (partial) identification of a variety counterfactual parameters in binary response models with possibly endogenous regressors. Our framework allows for nonseparable index functions multi-dimensional latent variables, and does not require parametric distributional assumptions. We leverage results on hyperplane arrangements cell enumeration from the literature computational geometry order to provide tractable means computing identified set. demonstrate how various functional form, independence, monotonicity assumptions can be imposed as constraints our optimization procedure tighten Finally, we apply method study effects health insurance decision seek medical treatment.