As is generally known, since Z. Griliches did a pioneering study of how far the profitability of R&D investment is, many an intensive investigation has been put into practice on this topic. Granted the significance of profitability analysis to measure an efficiency of resource allocation, however, it cannot provide a standard for passing fair judgment on the income distribution problem. Accordingly, this paper, taking account of R&D tax-bearings explicitly, sheds some light on the distribution of surpluses yield by technological progress among economic entities.

　It is easily comprehended that some parts of surpluses will revert to consumers through a decrease in product prices. And on the other hand, other surpluses which are due to producers will depend, to a large extent, on the degree of cost-reduction and price decrease. That is, there exists a trade-off between consumers and producers through the medium of technological progress, price decline, and tax burden. And as a result of it, government is confronted with three political options in determining the level of R&D investment : maximization of consumers' utility, producers' profit, and social welfare.

　Some concluding remarks can be summarized as follows.

　1) Theoretical considerations reveal that the criterion for maximizing the social welfare is the marginal value product of R&D knowledge stock, which provides the criterion for inferring whether resources are allocated effectively or not. Judging from the fact that the rate of return of R&D investment has been drawing near to an interest rate, we can conclude that resource allocation has been effectively achieved. However, value judgment inheres in this conclusion in itself.

　2) One important conclusion that I draw from the Pareto's criterion is as follows : Until the beginning of the 1970s, an additional investment in R&D improved the welfare of both consumers and producers, and thus constituted a Pareto improvement. However, afterward, an additional investment brought about an increase in consumer's welfare and, to the contrary, a decrease in producer's. Therefore, the allocation of R&D investment triggers the conflict of interest.

　3) It is a decline in the marginal product of R&D stock which gave rise to a decrease in the producers' profit owing to an additive R&D investment.

　The objective of this paper is to analyze the evolution of rice variety improvement and its contribution to the reduction of the rice production cost in California State. The "Rice Research Program" was formulated in 1969 with the principal objective to expand the size of rice research and strengthen the competitive power of the Californian rice industry. As a source of research funds for the implementation of this research program, a mandatory assessment on the rice growers was introduced based on the Marketing Order. The Rice Research Board carefully screened the funding requests from research agencies, and distributed research funds to each of the research projects in each calendar year.

　Since the introduction of the Rice Research Program, a total of 21 varieties have been developed. The development of the semidwarf varieties accelerated the increase of rice yields in California during the 1980s to the extent that California assumed the number one position in terms of land productivity in the world. The production cost per cwt in California is the lowest in the United States.

　It can be concluded that rapid technological innovation through rice research and extension in California contributed significantly in transforming a less-developed rice-growing state in the late 1960s to the most efficient U. S. rice-growing state in the 1980s.