Disequilibrium econometrics in dynamic models

Abstract In previous articles on disequilibrium econometrics, the need as well as the difficulty of dealing with dynamic models has been stressed. In this paper, we show how models with autocorrelated disturbances or unobservable lagged endogenous variables can be estimated with maximum likelihood techniques. We also propose simple algorithms to solve the maximization problem and we provide limited-information estimators which can be used as startingpoints for the maximization procedure.