The Demographic Transition Theory in Japanese Perspective

这个项目的目的是检验日本视角下的人口转变理论的有效性，重点是1920年至1960年期间。第二次世界大战后近半个世纪以来，人口变迁理论在社会科学的各个领域得到了广泛的接受，由于其对人口与发展的影响，该理论也在帮助发展中国家的国际政治中发挥了重要作用。然而，最近的研究，特别是普林斯顿大学的欧洲生育项目，对这一理论提出了严重的质疑。本项目的主要发现可概括如下：(1)从人口统计学角度来看，日本绝不是铁板一块，而是具有明显的区域格局。例如，日本东部的婚姻生育率很高，而日本西部的生育率很低。(2)生育率随时间的变化趋势也因地区而异。并不是所有的县在1920年以后生育率都有下降的趋势。欧洲和日本在人口行为上存在差异，有的呈上升趋势，有的没有。在欧洲人口结构转型期间(1870-1930)，婚姻生育率急剧下降，而结婚率变化很小。另一方面，日本的结婚率和婚姻生育率同时下降。(4)多元回归分析表明，日本生育率与农业变量和婴儿死亡率密切相关。1920-1950年间，日本的生育率普遍存在下降的结构。但这个结构在1960年完全消失了。与其使用婚姻生育率本身作为因变量，不如使用生育率下降的速度来得出更好的结果。

The aim of this project is to examine the validity of the demographic transition theory in Japanese perspective, focusing on the period from 1920 to 1960. The demographic transition theory has been widely accepted in various fields in social sciences for almost half a century after World War II. Because of its implications on population and development, the theory also played an important role in international politics aiding developing countries. However, recent studies, particularly Princeton's European fertility project, throw serious doubt on the theory. The main findings of this project can be summarized as follows :(1) Demographically speaking, Japan was by no means monolithic but had clear regional patterns. For example, marital fertility was high in eastern Japan but low in western Japan.(2) Fertility trends over time also varied regionally. Not all prefectures showed a declining trend in fertility after 1920. Some exhibited an increasing trend and others did not show any trend at all.(3) There existed a difference in demographic behavior between Europe and Japan. During the demographic transition in Europe(1870-1930), marital fertility declined precipitously whereas marital rates changed very little. In Japan, on the other hand, both marital rates and marital fertility declined simultaneously.(4)Multiple regression analyses show that fertility in Japan had a close relationship with agricultural variables and infant mortality. There existed a structure diterming fertility of Japanese which was common to the period from 1920 to 1950. But this structure completely disappeared in 1960. Instead of using a level of marital fertility itself as a dependent variable, using the speed of fertility decline can yield a better result.