John F. Geweke University of Minnesota - Twin Cities SBR-9514865 This research will be in the area of Bayesian econometrics, which uses priors to improve model results. The research has three parts. The first part will attempt to make Bayesian methods computationally less difficult. The second part will investigate the possibility that Bayesian results can be made publically available and reused by other researchers through reweighting. The third part will explore ways of relaxing distributional assumptions. Bayesian estimation is sufficiently difficult such that most economists currently do not use it. The main thrust of this research is to overcome the problems so that more economists use Bayesian results and methods.
