Distribution of academic research funds: a case of Japanese national research grant
- A database of competitive research funds in the Japanese academia is constructed.
- The distribution of research grants at the university and individual levels is examined.
- Results indicate high inequality at the university level and slightly lower inequality at the individual level.
- Over the last three decades, the total grant budget has greatly increased and more researchers have received any grant.
- Simultaneously, large-size grants have become more common and multiple awarding has become more frequent.
- The extent of inequality largely differs between scientific fields,cohorts, and career stages.
- The funding distribution is more unequal than the distribution of publications as an output indicator.
Drawing on a database of the competitive research funds in the Japanese academia, this study examines the distribution of research grants at the university and individual levels. The data indicates high inequality at the university level and slightly lower inequality at the individual level. Over the last three decades, the total grant budget has greatly increased and an increasing number of researchers have received the funds. Simultaneously, large-size grants have become more common and multiple awarding (i.e.,one researcher receives more than one grant simultaneously) has become more frequent. These changes taken together, the level of inequality has not been changed substantially. The extent of inequality largely differs between scientific fields; especially high in basic natural sciences and relatively low in social sciences. A close examination of inequality over researchers’ career indicates different patterns of transition between fields and cohorts. Finally, both at the university and individual levels, the funding distribution is found more unequal than the distribution of publications as an output indicator.
Growth of National Funds
In the Japanese academia, Grants-in-Aid for Scientific Research (GIA,
hereafter) have been playing the most fundamental role. GIA is the largest funding system, covering all scientific fields from social sciences to natural sciences, and is broadly awarded to university researchers from Ph.D. students to full professors. The system of GIA dates back to the 1950s. Since then, the budget size of GIA has been consistently increased in an attempt to improve the capability of university research, and it amounted to 200 billion JPY (2.2 billion USD) in 2010.
Funding Inequality at University Level
This graph shows two Lorenz curves of the university-level funding distribution in 2005; one for all universities and the other for only national universities. It clearly shows high inequality with a Gini coefficient of 0.919 for all universities. I also examined the transition of the Gini coefficients and found that the extent of inequality has not noticeably changed over 30 years (0.900 in 1995, 0.899 in 1985, and 0.914 in 1975).
Transition of Grant Size
I examine the transition of grant size since the 1970s. This graph distinguishes three size groups and shows their proportions: large (greater than 100 million JPY), middle (10-100 million JPY), and small (less than 10 million JPY). The graph indicates the tendency to enrich larger grants. In 1976-1980, small-size grants account for 66% and middle-size grants account for 34%, whereas small size is only 32% but middle size is 46% and large size is 21% in 2001-2005. oticeably, since the 1990s,large-size grants have constituted about 20% of the total budget. However, the recent two periods show a slight decrease in the proportion of large grants (from 23% in 1996-2000 to 21% in 2001-2005), while the budget amount for large-size grants have increased by 35%.
I counted the number of grants that each researcher received as a PI
simultaneously in a single year. This graph shows that most grantees obtained only one grant and 10-20% received two or more grants simultaneously. The ratio of multiple awarding increased from 9.5% in 1975 to 22% in 1995. The maximum number of multiple awarding to a single PI was seven and that to a single PI or member was 27 in 2005.
Funding Inequality at Individual Level
At the individual level (full and associate professors), funding
inequality is analyzed. The Gini coefficients are 0.685 (all universities) and 0.746 (top seven universities) in 2001-2005. The inequality has increased in the 1970s and the 1980s and then decreased in the 1990s and the 2000s.
Inequality of Funding vs. Inequality of Publication
I compare the individual-level distributions of funding and publications. The Gini coefficient for the publications of full and associate professors in all universities in 2001-2005 is 0.592, which is smaller than that for funding, or 0.685. The figure shows that publication inequality is smaller than funding inequality in all fields except for social sciences. The difference is especially noteworthy in mathematics and physics (0.523 vs. 0.726) and basic biology (0.570 vs.0.728).