Estimation in a disequilibrium model and the value of information

Abstract Suits' model of the watermelon market is reformulated as a disequilibrium model and the likelihood function is derived under various assumptions concerning the amount of available information. Such models are characterized by a min condition as in y t = min( D t , S t ) for an ordinary demand-supply model, where the observed quantity y t is the smaller demand and supply. Varying amounts of information may be available depending on which variables are observed and on whether prior information is available for separating the sample into subsets for which D t S t and conversely. The likelihood function corresponding to the least amount of prior information is shown to be unbounded without a certain variance ratio restriction. Computations are successfully carried out both with the original model and Monte Carlo experiments and the effect of different amounts of information on MSE's is analyzed.