This paper considers the question of adopting "dynamic" revenue estimation methods, which would incorporate the predicted macroeconomic effects of policies. It describes the current revenue estimation process and how these estimates are used in the formation of policy. It then evaluates this process as an exercise in forecasting, considering not only its methodology but also its performance. This evaluation illustrates some of the difficulties that would accompany attempts to use dynamic methods. Finally, the paper relates the choice among different revenue estimation procedures to the political process and the use to which the revenue estimates are put.

Alan J. Auerbach, University of California, Berkeley

[Home]