Nonparametric Estimation of Oligopsony Power in First-Price Auction

This paper proposes a nonparametric procedure to estimate market power for first-price auction data and applies the procedure to cattle procurement markets. Most previous studies have used parametric methods, which require specific functional forms for retail demand, input supply and processors' cost equations. However, researchers often find that market power estimates from parametric methods are sensitive to the choice of functional forms and specifications. An application to data from cattle procurement experiments shows that our nonparametric approach greatly outperforms commonly used parametric methods in estimating the degree of market power. While parametric estimates are sensitive to functional form specification and are at least 90% smaller than 'true' market power indices, estimates from our nonparametric procedure deviate from the actual value by no more than 25%. Copyright (c) 2009 The Authors. Journal compilation (c) 2009 The Agricultural Economics Society.