The Zilcha criteria for dynamic inefficiency reconsidered (Q1028550)

The goal of the paper is to correct, clarify and complete the inefficiency conditions in competitive stochastic general equilibrium models (stationary and nonstationary case) given by \textit{I. Zilcha} [J. Econ. Theory 52, No. 2, 364--379 (1990; Zbl 0731.90020); J. Econ. Theory 55, No. 1, 1--16 (1991; Zbl 0743.90034)] and then discussed and criticized in \textit{P. Rangazas} and \textit{S. Russell} [Econ. Theory 26, No. 3, 701--716 (2005; Zbl 1127.91045)]. After stating the model and basic definition of inefficiency the authors prove that necessary and sufficient conditions for inefficiency given by Rangazas and Russell are in fact equivalent to Zilcha's conditions in [loc. cit. (1990)] . Moreover a corrected proof of necessity and sufficiency is given since the original proof of Zilcha is erroneous. Next, a counterexample to Zilcha's conditions for non-efficiency in stationary context is presented. At the end of the paper some remarks on applicability of inefficiency/efficiency conditions for constructing statistical test are given.