Year: 2012
Type: report
Abstract: Estimation of causal eects in regression discontinuity designs relies on a local\nWald estimator whose components are estimated via local linear regressions centred\nat an specic point in the range of a treatment assignment variable. The asymptotic\ndistribution of the estimator depends on the specic choice of kernel used in\nthese nonparametric regressions, with some popular kernels causing a notable loss\nof effciency. This article presents the asymptotic distribution of the local Wald estimator\nwhen a gamma kernel is used in each local linear regression. The resulting\nstatistics is easy to implement, consistent at the usual nonparametric rate, maintains\nits asymptotic normal distribution, but its bias and variance do not depend on\nkernel-related constants and, as a result, is becomes a more effcient method. The\neffciency gains are measured via a limited Monte Carlo experiment, and the new\nmethod is used in a substantive application.
Author: Eduardo Fé
Institution: University of Manchester
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1690023883