Empirical Estimation of the Production Frontier

Estimating production frontiers is essential for performance benchmarking and productivity analysis. Current approaches include Data Envelopment Analysis (DEA), Stochastic Nonparametric Envelopment of Data (StoNED), and Stochastic Frontier Analysis (SFA). This study briefly reviews these existing approaches and proposes two distinct nonparametric methods for estimating the production frontier, based on data-fitting techniques. The first method offers an asymptotically consistent estimator for a piece-wise linear frontier, assuming a random inefficiency term. The second method combines a modified ordinary least squares approach with a resampling technique to account for the impact of random noise and enhance estimation precision. To illustrate the benefits of these proposed methods, simulated and empirical examples are included.