Revised: March 24, 1999 <p> The original draft of this paper was presented to the American Economic Association Meetings session "In Honor of Edwin Mansfield," held in New York, N.Y., on January 5, 1999. <p> This paper is a first step toward closing the analytical gap in the extensive literature on the results of interactions between public and private R&D expenditures, and their joint effects on the economy. A survey focusing on econometric studies in this area reveals a plethora of sometimes confusing and frequently contradictory estimates of the response of company financed R&D to changes in the level and nature of this category of public expenditures. Yet, a theoretical framework seldom is provided within which the empirical results are to be interpreted. Some such structure is necessary, in view of the multiple channels through which public research can affect private R&D performance, especially as not all the effects flow in the same direction. A major cause of "inconsistencies" in the empirical literature is the failure to recognize key differences among the various policy "experiments" being considered depending upon the economy in which they are embedded, and the type of public sector R&D spending that is contemplated. Using a simple, stylized structural model, we identify the main channels of impact of public R&D. We thus can characterize the various effects, distinguishing between short-run and long-run impacts that would show up in simple regression analyses of nominal public and private R&D expenditure variables. Within the context of our simple model it is possible to offer interpretations that shed light on recent cross-section and panel data findings at both high (i.e. national) and low (specific technology area) levels of aggregation. <p>JEL Classification: H41, H42, O31, O38 <p> Keywords: R&D, public goods, crowding out, spillovers, supply of scientists and engineers.
