California Agricultural Experiment Station Giannini Foundation for Agricultural Economics 1MELE: Maximum Entropy Leuven Estimators

Multicollinearity hampers empirical econometrics. The remedies proposed to date suffer from pitfalls of their own. The ridge estimator is not generally accepted as a vital alterna-tive to the ordinary least-squares (OLS) estimator because it depends upon unknown pa-rameters. The generalized maximum entropy (GME) estimator of Golan, Judge and Miller depends upon subjective exogenous information that affects the estimated pa-rameters in an unpredictable way. This paper presents novel maximum entropy estima-tors inspired by the theory of light that do not depend upon any additional information. Monte Carlo experiments show that they are not affected by any level of multicollinearity and dominate OLS uniformly. The Leuven estimators are consistent and asymptotically normal. Key words: multicollinearity, mean squared error, ordinary least squares, generalized maximum entropy JEL classification: C2, econometric methods: single equation models Quirino Paris is a professor of agricultural economics at the University of California, Davis and a member of the Giannini Foundation. This paper was written in honor and loving memory of my wife, Carlene Paris, who died of leiomyosarcoma---a very rare cancer---on May 5, 2001. I am indebted to Michael R. Caputo and Art Havenner for stimulating discussions and incisive comments that have contributed substantial im-provements of an earlier version. All errors are mine