Judgement and expert knowledge is widely thought to be important in the successful practice of forecasting. For policymakers, forecasts also embody judgements about the state of the economy, the transmission of shocks and optimal policy responses which necessitate the use of structural models. DSGE models estimated by Bayesian methods have become popular, partly because the models have become popular, partly because the models have well defined structural interpretations and because the Bayesian framework allows judgements about model properties to be embodied in priors. There is, however, uncertainty about both model parameters and the correct model.
This conference aimed to bring together policymakers and academics interested in these issues.