Abstract
Many important decisions about the allocation of research and development resources are made at the project level, but previous investigators have had very limited access to project-level research and development data. This paper contains analyses of the duration and intensity of investment in individual projects reported in the Defense Department's “Independent Research and Development” Data Bank, which includes longitudinal data on thousands of projects conducted by contractors during (approximately) 1980–1985. These data enable us to achieve three objectives: (1) to determine the statistical properties (moments and quantiles) of the distributions of completed projects by two measures of duration (calendar years and man-years) and by a measure of intensity (employment of scientists and engineers); (2) to discriminate between alternative hypotheses concerning the nature of the “hazard function” of independent research and development investment; and (3) to determine the properties of project sponsors' estimates (forecasts) of the duration and intensity of investment.
Get full access to this article
View all access options for this article.