Date of Award
Doctor of Philosophy
This thesis is concerned with the single-equation errors-in-variables estimation of rational expectations models. More specifically, we propose some new single-equation errors-in-variables estimators for a certain class of models which occur frequently in the macroeconomics literature. The asymptotic properties of these new estimators are derived and a Monte Carlo study is used to investigate the small sample performance of these new estimators relative to that of the existing errors-in-variables estimators and the single-equation substitution method estimator.;This Monte Carlo study is particularly useful in providing, as a by-product, the first small sample comparison of certain estimators used widely in the literature. An important finding is that, to the extent to which we can generalise on the basis of our limited experiments, there are many instances in which statistical inference associated with all of these estimation techniques can be highly misleading. In addition, we illustrate the use of some new non-parametric density estimation techniques in the summary of the Monte Carlo results.;The thesis also provides a useful survey of the literature on the solution of rational expectations models together with a reasonably comprehensive survey of estimation methods.
Power, Simon A., "Single-equation Errors-in-variables Estimation Of Rational Expectations Models" (1986). Digitized Theses. 1483.