A0368
Title: Estimation and inference for extreme continuous treatment effects
Authors: Liuhua Peng - The University of Melbourne (Australia) [presenting]
Wei Huang - University of Melbourne (Austria)
Shuo Li - Tianjin University of Finance and Economics (China)
Abstract: Estimation and inference for the treatment effect are studied on deep tails of the potential outcome distributions corresponding to a continuously valued treatment, namely the extreme continuous treatment effect. Two measures are considered for the tail characteristics: the quantile function and the tail mean function, which is defined as the conditional mean beyond a quantile level. Then, for a quantile level close to 1, we define the extreme quantile treatment effect (EQTE) and extreme average treatment effect (EATE), which are, respectively, the differences of the quantile and tail mean at different treatment statuses. Estimators are proposed for the EQTE and EATE based on tail approximations from the extreme value theory. The limiting theory is for the EQTE and EATE processes indexed by a set of quantile levels and hence facilitates uniform inference for the EQTE and EATE over multiple tail levels. Simulations suggest that the method works well in finite samples, and an empirical study of its practical merits is illustrated.
