A1256
Title: Semiparametric efficient estimation of causal effects defined by statistical functionals
Authors: Kuan-Hsun Wu - Pennsylvania State University (United States) [presenting]
Abstract: The problem of estimating a general class of causal effect is explored. In particular, the causal effects can be represented as a Hadamard differentiable statistical functional of the underlying distribution. The semiparametric lower bounds are found and attained by a nonparametric estimation approach. Commonly used cases, including but not limited to average, quantile, and distributional causal effects, are shown to be special cases in the general class. Moreover, a new causal effect defined on expectiles is investigated, which is also fully covered by the theory.
