Some recent developments in Bayesian statistics and econometrics

After a brief review of the past and current status of the Bayesian approach to econometrics and statistics, a review of some recent results in Bayesian forecasting and their relation to time series model-building is presented. Then some remarks on the growing importance of information theory in Bayesian analysis are provided and illustrated with selected examples. It is pointed out that various information theoretic criterion functionals are being utilized to produce models for observations, prior densities, measures of the information provided by experiments, information processing rules, including Bayes' Theorem. Information processing when the form of the likelihood function is unknown via the new Bayesian method of moments and generalized maxent procedures is discussed. Last, the future of Bayesian analysis is discussed.