Wallace and Hussain (1969) considered the use of an error components regression model in the analysis of time series of cross-sections and developed an estimator of the coefficient vector based on an estimated variance-covariance matrix of error terms. In this paper, we have shown that under the set of assumptions adopted by Wallace and Hussain there are an infinite number of estimators which have the same asymptotic variancecovariance matrix as the Wallace-Hussain estimator and also that it is not possible to choose an estimator on the basis of asymptotic efficiency. We have developed an alternative estimator of the variance-covariance matrix of error terms and have used this estimator in developing a feasible Aitken type estimator for the coefficient vector. We have derived some small sample properties of this estimator and have compared them with those of other estimators of the coefficient vector.

1. TO BE CLEAR from the beginning, this note does not provide practical aids to identification; the information is very unlikely to be available. Rather, this paper serves an expository purpose, that of seeking to clarify two ideas: first, multiple-identification, by presenting and analyzing an example within a linear system, and second, Working's and Fisher's ideas on relative variances as identification aids, by showing the consequences of knowing that one variance is a specific large multiple of another. Along the way we provide a correction to the econometrics literature.

[This paper deals with two single-equation estimators in a set of simultaneous linear stochastic equations--namely, ordinary least squares (OLS) and two-stage least squares (2SLS). Under the assumption that all predetermined variables in the model are exogenous, necessary and sufficient conditions are obtained for the existence of even moments of the above estimators. It is shown that for the general case with an arbitrary number of included endogenous variables, even moments of the 2SLS estimator are finite if and only if the order is less than K2 - G1 + 1. Furthermore, even moments of the OLS estimator exist if and only if the order is less than N - K1 - G1 + 1, where N is the sample size, G1 + 1 is the number of included endogenous variables, K1 and K2 respectively are the number of included and excluded exogenous variables in the equation to be estimated.]

[This survey of the use of structural equation models and methods by social scientists emphasizes the treatment of unobservable variables and attempts to redress economists' neglect of the work of Sewall Wright.]