Andrew Atkeson - Consultant
Revised September 30, 2011
We present a tractable model of innovating firms and the aggregate economy that we use to assess the link between the responses of firms to changes in innovation policy and the impact of those policy changes on aggregate output and welfare. We argue that the key theoretical determinant of the relative long-run aggregate impact of alternative policies is their impact on the expected profitability of entering firms. We show that, to a first-order approximation, a wide range of policy changes have a long-run aggregate impact in direct proportion to the fiscal expenditures on those policies, and that to evaluate the aggregate impact of such policy changes, there is no need to calculate changes in firms' decisions in response to these policy changes.
We use these results to compare the relative magnitudes of the impact on aggregates in the long run of three innovation policies in the United States: the Research and Experimentation Tax Credit, federal expenditure on R&D, and the corporate profits tax. We argue that the corporate profits tax is a relatively important policy through its negative effects on innovation and physical capital accumulation that may well undo the benefits of federal support for R&D. We also use a calibrated version of our model to examine the absolute magnitude of the impact of these policies on aggregates. We show that, depending on the magnitude of spillovers, it is possible for changes in innovation policies to have a very large impact on aggregates in the long run. However, over a 15-year horizon, the impact of changes in innovation policies on aggregate output is not very sensitive to the magnitude of spillovers.
On the basis of these results we conclude that, while it is possible to make comparisons about the relative importance of different policies and sharp predictions about their aggregate impact in the medium term, it is very difficult to shed much light on the implications of innovation policies for long-run aggregate outcomes and welfare without accurate estimates as to the magnitude of innovation spillovers.
Download Paper (PDF)