ESTIMATING COST-FUNCTIONS IN REGULATED INDUSTRIES CHARACTERIZED BY ASYMMETRIC INFORMATION

In this paper a structural model is developed to estimate a cost funct ion that includes both an adverse selection and a moral hazard variabl e. It is shown how traditional cost function estimates may produced bi ased results if moral hazard is present. This gives an alternative rea son to Feinstein and Wolak [The econometric implications of incentive compatible regulation, In: G.F. Rhodes, ed., Advances in econometrics, Vol. 9, 159-204 (JAI Press, Greenwich, CT, 1991)] of why scale econom ies may be overestimated in regulated industries. Parameter estimates from our model can be used to design an optimal contract or any other regulatory mechanism. We illustrate our approach using data from the N orwegian bus transport industry. (C) 1997 Elsevier Science B.V.