Showing papers in "The Economic studies quarterly in 1973"

ON THE EXISTENCE AND THE APPROXIMATION FORMULAE FOR THE MOMENTS OF THE k-CLASS ESTIMATORS

Abstract: Quarterly or monthly data over ten to fifteen years have become available recently for a large number of economic and firm variables. This makes various dynamic econometric methods, such as distributed lag, auto-correlated errors, and optimal prediction, at least worth trying. At the present time, however, only the asymptotic theory is available for judging the performance of these statistical methods, whereas the length of data available to us is such that the asymptotic theory should neither be rejected on the basis of obvious irrelevance nor be accepted without a careful investigation. Professor Nagar in a series of papers developed the formulae for the bias to the order of T-1 and for the variance to the order of T-2, where T is the length of data, for a number of estimators that arise for various models. This kind of approach might be useful for examining the performance of dynamic econometric methods in our particular use. It is rather unfortunate therefore that the method that Nagar used has been subjected to a criticism by Srinivasan [22] denying its validity at least under situations as unrestricted as Nagar has.1)