Economic Quarterly

We evaluate the Federal Reserve Bank of Richmond (FRBR) manufacturing survey and assess its contribution to explaining national and regional economic conditions. Specifically, we examine the predictive accuracy of a variety of static and dynamic models. The models include the composite diffusion index reported by the FRBR and other information readily available from the FRBR surveys but not currently employed in the calculation of the composite index. The paper concludes, first, that the diffusion indices currently reported perform reasonably well at explaining both the national and the regional economy. Second, it is possible to improve the predictive power of the indices by considering models that account for a richer dynamic structure given the high persistence of the series under study. Third, the predictive accuracy of the current FRBR composite index can be improved further by adjusting the weights used in its calculation and by including other diffusion indices. Also, the composite indices that track the national and regional economy would not necessarily be the same, and the paper provides a few insights on what those diffusion indices would look like.