Page 287

vances in knowledge or the introduction of new products in measuring outcomes. Although inherently imperfect, the use of multiple proxy indicators in combination with peer review probably represents the most likely (and most reasonable) performance measurement approach for fundamental research under GPRA.

This discussion of performance measures raises a key issue: the rate of economic return on R&D investments. Specific data are not available for condensed-matter and materials physics, although powerful evidence abounds, such as the economic impact of the transistor, magnetic materials, fiber optics, and the solid-state laser. Numerous studies of economic return have been performed for the broader R&D enterprise. These studies, many of which are listed in Table 7.4, indicate average rates of return of 15 to 20 percent per year. This is an extraordinary indication of the value of research. Unfortunately, these data do not provide information on the impact of adjustments (either up or down) in the level of R&D investment and are therefore not useful in determining absolute funding levels.