世界各国の穀物自給率と耕地賦存，所得，農業保護

自給率の基礎的規定要因と日本の位置付け

抄録

We investigated the relationship between the self-sufficiency ratio of cereals and its three determinants: land resource endowment, income level and support for agriculture. The relationship also provides an explanation of the trade pattern of cereals. Cross-country analysis was carried out for 157 countries in the period 1994-1998. We also used data from 27 countries in the period 1982-1987 for the analysis of % PSE.
Non-parametric regression and quadratic regression analysis revealed the non-linearity of the relationship among variables and an interaction of GDP (per capita) and arable land area (per capita). (1) When GDP increases, cross-country difference in the self-sufficiency ratio increases. The difference depends on arable land area. When GDP increases, the self-sufficiency ratio increases in land-abundant countries, while the ratio decreases in land-scarce countries. (2) In land-scarce countries, the slope of self-sufficiency ratio to GDP changes from down-ward to upward when GDP exceeds a certain level. The more scarce arable land, the higher the level of GDP where the slope of self-sufficiency ratio turns positive.
The pattern of (1) suggests that relatively abundant factor among labor, land and capital tends to be used more intensive for cereal production. When per capita GDP increases, the cross-country difference in yield decreases, while the cross-country difference in land-labor ratio remains. Both cross-country differences depend on arable land area per capita.
Probably government support for agriculture causes (2) . In fact % PSE is higher in the countries with more scarce arable land, lower self-sufficiency ratio, and higher GDP per capita. But more direct evidence was not found.
The Analysis of the ratio of harvested area of cereals to arable land also supports our interpretation of (1) and (2).