Summary: Attempts to develop measures of output which can be used consistently among federal research and development projects must try to deal with the following problems: the variety of reasons for which research and development expenditures are undertaken; their diverse results; their speculative nature; their poorly understood economic effects; and the fact that research and development is only one input into the complex process of discovering and implementing new ideas. Federal commercially directed research and development is one type of federally financed research and development in which progress in improving the allocation of resources is most likely to be made. These programs are justified on the grounds that results will ultimately have an economic value. GAO suggests a conceptual framework for appraising commercially directed programs based on two prinicples: (1) federal contributions to the innovation process need to be evaluated in the context of all other factors affecting commercial acceptance of innovation; and (2) estimates of value to consumers can be used to calculate the rate of return on resources invested in federal research and development. In estimating the value of products to consumers one of two approaches for measuring the output of the federal research and development expenditures could be used: comparing the measured output with the total program costs which consist of research and development costs and all other program input cost; or comparing the marginal output of research and development spending with research and development costs. The second of these approaches is the most desirable, because budget constraints indicate that research and development expenditures should be allocated to projects in which the value received per dollar invested is the highest. The conceptual framework was applied to Department of Energy's Solar Photovoltaics program to illustrate the importance of this framework in identifying and attempting to quantify the regulation, tax, and other factors that affect the value of federal research and development spending.
