While on the staff of the President's Council of Economic Advisors (1965-1966), Professor MacAvoy observed that both government and industry needed a broader economic evaluation of the breeder reactor program than that provided by the Atomic Energy Commission. Most salient, he found, was the necessity to consider national economic benefits from the breeder reactor as well as the promised technical improvements in uranium usage that apply to the reactor industry alone. As an economist delving into a field primarily reserved for the engineer, the author is able to introduce just such an economic evaluation of reactor development projects. Professor MacAvoy attended courses in nuclear engineering and consulted engineers throughout the country for this study – a cross-fertilization that has resulted in a provocative benefit-cost analysis of both engineering efficiency gains and consumers' surplus.
Research in advanced nuclear reactors requires most of the public expenditures on R&D in new energy systems. The questions of research strategy asked in relation to these large expenditures are how many different reactor systems should be developed? How many companies should construct separate demonstration reactors? When should the development be complete? To provide initial answers, the author completes forecasts of the costs of various fast breeder R&D programs and constructs a model of the energy sector of the American economy to evaluate benefits. Then, he focuses upon the first indications of the research project mix that will offer the largest present value of forecast economic benefits over and above research costs. The program that best meets this standard stresses duplicative research in separate companies at greater expense than that projected by the current Atomic Energy Commission Program. To pursue such a course would mean to increase the number of breeder reactor types in order to meet demands for varied sizes and performance characteristics in reactors, and in order to increase competition among reactor manufacturers.
Professor MacAvoy's method of evaluating net present benefits stresses the uncertainty of results from research programs. But equal stress on economy-wide benefits results in finding that a clear first round choice can be made among reactor development projects despite great uncertainty. The calculations of frequency distributions of net present value here can serve as examples of the method for presenting economic research critical for making choices between costly and complex scientific programs.