[The paper discusses the problem of pooling cross-section and time-series data in the framework of likelihood functions and the associated posterior distributions and shows how this analysis can be useful in deciding whether or not to pool.]
[This paper is an attempt at overcoming various logical inconsistencies which have been found in the Leontief dynamic model. It is based on an interpretation of the Leontief model in the framework of a general model of capital accumulation of the Walras-Hicks type. A particular case of such a model, corresponding to the hypothesis of static expectations, turns out to be the competitive capital theory underlying the Leontief model. This theory, capable of defining a meaningful equilibrium even when the Leontief equilibrium based onfull employment of all stocks is not possible, allows us to overcome the difficulty in retaining the economic interpretation of solutions. In the terminology of Hicks (Capital and Growth), this theory solves the problem of the traverse between two full employment equilibria. It will be evident, contrary to current opinion, that output levels and prices are indissolubly connected also inside the Leontief model. Finally, Solow's prices, which give rise to the striking "dual stability paradox," will be discussed. The ultimate reason for such a paradox will be identified as an intrinsic inconsistency in the assumption of correct expectations on which these prices are based.]
The paper argues that variance components models are very useful in pooling cross section and time series data because they enable us to extract some information about the regression parameters from the between group and between time-period variation-a source that is often completely eliminated in the commonly used dummy variable techniques. The paper studies the applicability and usefulness of the maximum likelihood method and analysis of covariance techniques in the analysis of this type of model, particularly when one of the covariates used is a lagged dependent variable.