Osaka Workshop on Economics of Institutions and Organization (#133)

This study aims to measure income disparity in ancient Japan before the onset of modern economic growth. Specifically, the analysis relies on estimating the “institutional” income of peasants and government officials under the Ritsuryō regime in Nara-period Japan from historical documents such as ancient laws and regulations. The results indicate that 30-40% of peasants’ income was collected as land tax (so), poll taxes (yō and chō) and compulsory public loans (suiko), while they earned income from the cultivated land allocated by the Ritsuryō government. In particular, the tax burden ratio of poll taxes and compulsory public loans was high. The class government officials who did not have any tax burden can be divided into upper nobles, middle and lower nobles, and general officials by their income level. The income level of the upper nobles was remarkably high, and the results indicate income disparity among government officials. The incomes of general officials were higher in the local branch offices than in the central government. A comparison between social classes does not confirm a large difference between peasants and the general officials, but does show an extremely high income disparity between peasants and upper nobles. Though the result is based not on actual income, but on institutional income, the ancient Ritsuryō regime was established based on a system that produces a very high income disparity.Abstract: Masanori Takashima

In this paper we investigate relational contracts when players are allowed to save/borrow so that their wealth change over time. In contrast to the existing studies about relational contracts which have assumed no savings/borrowings, we show that efficiency of relational contracts is drastically improved by saving/borrowing opportunities; even in a finite horizon environment the lack of verifiable information causes no efficiency losses during certain periods in the beginning when agents make hidden savings which are not observed to principals. We also show that the efficiency loss vanishes as the time horizon is sufficiently long.Abstract: Shingo Ishiguro