Increasing the transparency of macroeconometric forecasts: A report from the trenches

The acceptance of macroeconometric model results suffers from the procedures of model builders/users that are often opaque. This paper describes in detail the production of a macroeconomic forecast with a macroeconometric model for the German economy. The model used is the RWI-business cycle model, a medium sized macroeconometric model (41 stochastic equations, 86 identities). It has been used since the late 1970s in the RWI and in the biannual ‘Joint Diagnosis’ of German economic research institutes. The paper starts with a presentation of the analytical foundations of econometric model forecasting in general and of various ways and forms of incorporating outside-model information in particular. The presentation of the main features of the model is followed by the detailed description of the various stages of forecast production. The paper ends with suggesting a more intensive use of the analytical possibilities of the models and their forecast.