Error Correction Mechanisms

The interface between economic theory and applied econometrics is often one of uneasy compromise, with the pragmatic justification for many accepted procedures resting on a theoretical base. This paper examines the surprisingly strong arguments that exist in terms of economic theory, for the use of error correction mechanisms in the specification of short run dynamic adjustment. A common heresy exists that while economic theory provides a detailed analysis of comparative static equilibria it can offer no guidance as to the appropriate specification of dynamic adjustment towards an equilibrium. perhaps in consequence it is not uncommon to find examples where the necessary dynamic specification is achieved by "tacking" onto an existing equilibrium specification some relatively ad hoc short run adjustment scheme. The intercession of stochastic arguments in this process is confused and critical implications are frequently ignored in practice, but perhaps more importantly there will typically be no guarantee that the dynamic specification is consistent with the prescribed equilibrium. Consistency in this sense requires that the short run dynamic adjustment be directed by the perceived disequilibrium and that eventual convergence to the equilibrium position be ensured. That two separate theoretical arguments, co-exist within the final specification is the root cause of many difficulties both theoretical and empirical.