Twenty-five years ago macroeconomics seemed poised for a substantial advance in quantitative precision and usefulness. The statistical tools of simultaneous equations modeling had reached maturity and appeared ready to allow application of then accepted Keynesian theory with steadily increasing scale and precision. But these expectations of dramatic progress proved overoptimistic. Keynesian theory no longer represents a consensus; the elegant tools of simultaneous equations theory are less used than a few years ago. Progress in quantitative macroeconomic research is currently due to the research of several different groups with disparate aims and methods. These include work by commercial forecasting organizations on sustained, practically useful, large-scale contact with data; rational expectations models that provide realistically complex interpretative explanations of how observed patterns in the data could arise from individual and institutional behavior; and methods of probability modeling that provide an increasingly accurate picture of the uncertainty associated with model forecasts and policy projections. Communication among these groups is limited, and progress is slow for all of them because of the shrinkage in recent years of funding for sustained, scientifically oriented research in quantitative macroeconomics. This grant supports the creation of the Institute for Empirical Macroeconomics. The purpose of the Institute would be to encourage research in empirical macroeconomics and related fields that is large scale, sustained over time, and reproducible, with quantitative measures of uncertainty attached to the results. A renewed effort can be expected to make progress, given recent developments in macroeconomics and in computational capacity. The institute can play a unique and potentially important role in drawing together the different strands of empirical macroeconomic research and pushing out more rapidly the frontiers of computational research in macroeconomics.
