Abstract
The level of the government funding of university research is a fundamental and important indicator of a government's attitude toward basic research. However, a definition of government funding of university research and its estimation method are not fixed in Japan. There are several important statistic resources concerning the government funding of university research: (i) the survey of Research and Development (Statistics Bureau, Management and Coordination Agency), (ii) OECD's science and technological indicators, (iii) the budget for universities of the Ministry of Education, Science and Culture (Monbusho), (iv) the budget for science and technology of Monbusho, (v) the accounting data of universities, (vi) the estimation pursued by Irvine and Martin. This paper describes data handling and estimation manners in these statistic resources to clarify mutual relationships among them. Especially, methodological errors of the estimation (ii), and the estimation method (iv) are introduced for the first time in this paper. This paper points out following questions concerning estimation methods: (i) technical problems in FTE (Full-time Equivalent) adjustment such as double adjustment, contents of wages and salaries, and a coefficient of FTE, (ii) institutional problems, (iii) perception gaps between actual R&D activities and statistical data. The estimation of the government funding of university research is very confusing. Every statistical resource has defects. To correct the defects, it is necessary to make clear the definitions of the Survey of R&D and to reexamine the manner of data aggregation and classification.
