Investment in R&D, Costs of Adjustment and Expectations

This paper proposes a framework which integrates convex costs of adjustment and expectations formation in the determination of investment decisions in R&D at the firm level. The model is based on cost minimization subject to the firm's expectations of the stream of output and the price of R&D, and results in equations for actual and multiple-span planned investment in R&D and for the realization error as functions of these expectations. The model accommodates alternative mechanisms of expectations formation and provides a methodology for testing these hypotheses empirically. We derive estimable equations and testable parameter restrictions for the rational, adaptive and static expectations hypotheses. The empirical results using pooled firm data strongly reject the rational and static expectations hypotheses and generally support adaptive expectations.