64, No. 6
It has been often argued that agriculture in Japan stagnated in the 1920s, and that hyperinflation caused by World War I aggravated shortages and led to the violent rice riots of 1918. The Japanese authorities are therefore said to have developed a policy of increasing rice production through designating Taiwan and Korea as rice-supply bases for the home country.
This paper argues that neither the available data nor theoretical considerations support the idea that Japanese agriculture stagnated during the 1920s and, therefore challenges the common interpretation of the relationship between Japanese rice policy and its colonial management. The empirical findings of this paper imply the following conclusions: (1) rice production in Japan did not stagnate during the 1920s, (2) farmers both in Japan and its colonies behaved rationally with respect to economic incentives, and (3) its was rice imports from the colonies which affected (in the sense of Granger causality) the level of rice production in Japan, rather than vice versa. That is, the behavior of farmers both in Japan and in its colonies can be explained in terms of economically rational motivation, and their attempts to take advantage of different sets of economic opportunities and constraints created by the division of rice production in the Japanese Empire.
The new version of the putting-out system gave putting-out masters greater flexibility for responding to fluctuations in demand. Putting-out masters were able to enjoy the flexibility in putting-out, while the chin'ori [agent-weavers] factories were assured both a regular supply of yarn and large-sized buyers of cloth. Mutual dependency fostered a feeling of trust in the interpersonal relationship between putting-out masters and chin'ori factories. This saved transaction costs, by reducing opportunistic behavior on the part of individual chin'ori factories under uncertain circumstances, and increased the efficiency of the overall economy of the producing area. In other words, •trustê built through the new version of the putting-out system acted as a kind of local public goods which benefited only local traders. Moreover, regulation by Dogyo Kumiai [local trade associations] after 1900 contributed to the maintenance of stable conditions for fair competition among the putting-out masters within a producing district.
Thus the new version of the putting-out system, regulated by Dogyo Kumiai, functioned as an institution which both allowed greater flexibility and channeled 'trust' which was able to act as a kind of local public goods.
As the relative estimation of work was one of the important features of this system, the efforts of young female workers were directed not toward absolute achievement, but toward improving their relative superiority in comparison to other workers, so that they became engaged in endless competition. In other words, the system functioned as a way of raising the incentive of workers without increasing the total sum of wages. By the end of the 1880s, the rank-order wage system was used in estimating labor productivity, as well as productivity in terms of raw material (cocoons), resulting in great increases in productivity in silk reeling. But after the mid-1890s, the American market began to demand higher quality, i.e. more even quality of filatures (machine-reeled silk), as a result of the increasing speed of power looms in the American textile industry. Rapid expansion of the scale of silk reeling factories was not sufficient to satisfy this demand. The silk reelers had to control and optimize the labor output of their female workers in terms of product quality as well as productivity. This was achieved by expanding the rank-order wage system to add the functions of monitoring and estimating product quality.
In this way, a very complex wage system had developed by the early 1900s, enabling the silk reelers to control and optimize labor output to a very high degree.