Ximing Wu, Texas A&M University
Semiparametric Estimations under Shape Constraints with Applications to Production Functions
Date and Location
Monday, January 14, 2013, 10:30 AM - 12:00 PM
3001
Plant and Environmental Sciences
Abstract
Economic theories often provide guidance on econometric modeling. We propose semiparametric estimators consistent with monotonicity and concavity constraints using the method of penalized splines. The shape constraints are maintained via integral transformations of spline basis expansions. We present an effective algorithm, the large sample properties, methods of inferences and of smoothing parameter selection of our estimators. We also consider multiple regressions under the framework of additive models. We conduct Monte Carlo simulations to illustrate the finite sample performance and usefulness of the proposed method. We applied the proposed methods to estimate: (a) the relationship between individuals' degree of optimism and risk tolerance; (b) a production function with multiple inputs.