Peter Schmidt Michigan State University SBR-9514750 The proposed research will cover three topics in econometric theory. The first part will deal with instances in which the effects of economic shocks die off completely, but very slowly. Research will be undertaken to determine which shocks are best characterized by this phenomena and how to estimate variables related to these shocks. The second part will develop estimation methods for panel data. Panel data involve repeated observations over time at the individual or firm level; they allow statistical control for certain types of unobservables and the identification of dynamic relationships. The third part will continue work on the estimation of production frontiers and the efficiency of economic units. The research will develop measures of the statistical precision of efficiency estimates and will extend knowledge of complications arising in efficiency measurement, The research undertaken will be useful to other economists who use econometric models in their research.
