difficult issue, since standard discounting procedures are likely to distort the value of programs implemented in early years relative to that of the more recent programs. Therefore, estimates of the time path of benefits and costs are desirable (all expressed in 1999 dollars) whenever possible. Several different approaches to summing benefits and costs across programs and across time periods are possible, including the approach that reports undiscounted aggregates of future and past benefits and costs.
Options benefits are the positive and negative consequences of technologies for which R&D has been completed, that are ready to be commercialized but are constrained by current economics or other circumstances, and that could be adopted under some plausible future economic, regulatory, and tax conditions. The types of technologies classified as yielding options benefits include the following:
Deployable technologies not likely to be commercialized under the most likely economic, policy, and tax condition but likely to be commercialized under some set of reasonably plausible conditions.
Technologies for which the technological challenge has been met but for which the normal costs of improving a technology in the course of its commercialization have not yet been expended and for which the commercialization process can be expected under some set of reasonably plausible conditions.
An options benefit is closely associated with a technology that is “on the shelf” and is not commercially viable under current economic conditions. Thus, for example, indirect coal liquefaction may have significant options value because it has been developed and may become commercially viable if oil prices reach and remain well above $30 or $40 per barrel. On the other hand, magnetohydrodynamics may not have options value because the R&D program was terminated before the technology was fully developed.
Knowledge benefits are defined as scientific knowledge arising from a technology for which R&D has not been completed but that holds promise for future application, perhaps in completely unforeseen ways. These benefits are qualitative descriptions of advances in knowledge based on research over and above the research that developed specific technologies. The advances could lead to other technologies, but at this time those technologies have not been developed. Knowledge benefits include unanticipated and not-closely-related technological spin-offs that are made possible by research programs. For example, the Office of Fossil Energy’s coal R&D programs have had many significant technological spin-offs. These spin-offs represent knowledge benefits.
The category “knowledge benefits” probably has by far the greatest diversity of economic, environmental, and security benefits and is, accordingly, probably the hardest to evaluate with any confidence. For some classes of knowledge benefits, it will be impossible to quantify in any manner that would allow an objective overall assessment of importance. For example, improvements in our knowledge of basic physical processes would fall into this category. However, other knowledge benefits do allow some quantification. This is particularly true for some well-defined technology development programs currently under way. The Partnership for a New Generation of Vehicles may fall into this category.
The matrix approach developed here is useful for placing the benefits and costs of energy R&D programs in a consistent and comprehensive framework that will permit objective comparison across programs and technologies. However, several caveats are in order with respect to the use of this approach.
First, there may be a tendency to concentrate on the information contained in the northwest cell of the matrix—realized economic benefits and costs—because it is often the simplest to identify and quantify. Nevertheless, when evaluating federal R&D, it would be shortsighted to concentrate excessively on the data in this cell of the matrix. The other criteria developed here are also meaningful and important in assessing the costs and benefits of the DOE R&D programs and must be objectively valued in context of the national interest. In addition, technology developments promising to provide short-run economic benefits are more likely than technology development providing only environmental or security benefits to be pursued by private sector corporations. Therefore, it is programs that promise those environmental or security benefits that are most likely to need government support.
There is another problem with concentrating on realized economic net benefits: doing so tends to favor R&D programs that were successfully completed many years ago and had time to produce substantial realized economic benefits, at the expense of more recent or current programs. Thus, the energy efficiency R&D program to develop electronic ballasts for fluorescent lights, conducted in the late 1970s and the early 1980s, had produced substantial realized net economic benefits by 2000, whereas the PNGV program, which began in the late 1990s, is not expected to begin generating economic benefits until after 2005. Focusing on realized economic benefits alone would inappropriately bias the assessment in favor of R&D on electronic ballasts and against R&D through the PNGV program. In other words, estimates orga-