Copyrights to papers in the Duke Economics Working Paper Archive remain with the authors or their assignees. Archive users may download papers and produce them for their own personal use, but downloading of papers for any other activity, including reposting to other electronic bulletin boards or archives, may not be done without the written consent of the authors.

Duke Economics Working Paper #00-03

Effective Nonparametric Estimation in the Case of Severely Discretized Data


Mark Coppejans

Abstract

Almost all economic data sets are discretized or rounded to some extent. This paper proposes a regression and a density estimator that work especially well when the data is very discrete. The estimators are a weighted average of the data, and the weights are composed of cubic B-splines. Unlike most nonparametric settings, where it is assumed that the observed data comes from a continuum of possibilities, we base our work on the assumption that the discreteness becomes finer as the sample size increases. Rates of convergence and asymptotic distributional results are derived under this condition.

Key Words: Discretization, Nonparametric Density Estimators, Nonparametric Regression Estimators, B-splines, Kernels

JEL:

Forthcoming in Journal of Econometrics.

Retrieve document:

47 pages

  Help on downloading documents