Compilation and Analysis of Econometric Data for the Household Behavior in Japan

这项研究的动机是在纯经济/计量经济学理论的公理系统与用于理论验证的经验/统计数据之间经常观察到的深层次差距。该项目从寻找填补这一空白的方法开始，将我们的注意力集中在日本劳动力市场的家庭行为上，最终目标是在头脑中构建一个真正的经验理论。我们的两个阶段的项目的第一阶段主要包括初步研讨会：用现有数据进行文献回顾，并集思广益，寻找新的想法。第二阶段是对家庭行为的横断面数据进行汇编和分析，重点是家庭主妇的行为，以形成经验性解释的理论以及理论检验。我们从东京和大阪周围的两个特大城市地区的家庭中随机抽取了1200个样本。问卷被留在这些样本家庭中，让他们在一周后填写和收集。这种方法在一段时间后重复。两次试验的有效收益率都达到了约70%。对结果进行了编辑、分析，并以个人数据为基础存储在磁带中。这些数据包括每月收入、消费和个人每周用于工作、休闲和通勤的平均小时数等。磁带将根据要求提供。我们还开发了一个用于非线性估计的计算机程序，作为分析获得的数据的工具。这将使我们能够估计家庭生产函数或其双成本函数。

This study has been motivated by the oft-observed deep gap between the axiomatic system of pure economic/econometric theory and the empirical/statistical data for theory veryfication. The project started with a search for a device to fill this gap by focusing our attention upon the household behavior in the labor market in Japan with the final goal of constructing a genuinely empirical theory in mind.The first stage of our two-stage project consisted mainly of preliminary workshops: literature review for checking with available data and brainstorming for new ideas. The second stage featured compilation and analysis of the cross-sectional data on the household behavior with an emphasis on the housewives' behavior needed for the formation of an empirically interpreted theory as well as theory testing Our 1,200 random samples are chosen from the households in two megalopoleis areas around Tokyo and Osaka. Questionaires were left with these sample households for them to fill out and collected a week later. This method was repeated after an interval. The rate of effective return on both trials amounted to 70% approximately. The results have been edited, analyzed, and stored in the magnetic tape on an individual data basis. The data include the monthly income, consumption and the individual's weekly averages of hours spent on work, leisure, commuting, etc. The magnetic tape will be available upon request.We have also developed a computer program for nonlinear estimation as a tool to analyze the obtained data. This will enable us to estimate the household production function or its dual --- cost function.