統計と日本経済

Property of Japanese Family Income and Expenditure Survey (JFIES) has been examined. JFIES would not be a good business cycle indicator due to sampling error, while it is sufficiently accurate for academic analysis. JFIES suffers sample selection biases, with which dual-income and lower-income households are under-represented. Also, there are substantial under-reporting for expenditures on expensive goods and services and receipts of asset income. Those biases results in upward bias in household saving rate. In general, however, the quality of JFIES is high and comparable to the household surveys in other developed countries such as US and UK.

This paper analyzes aggregate flows in and out of unemployment as well as individual level job search behavior from 2002 to 2007 when non-standard employment markedly increased in the Japanese labor market. The paper is among the first to match the individual level data of Labor Force Survey to make one year panel data and supplement the data with the individual level information available from 2002. Our analysis shows that males with high school education involuntarily lost job during the recession, and males who are just out of college as well as other male job seekers seem to have sought for better jobs resulting in markedly lower outflow from the unemployment pool compared with females, resulting in a gender gap in the unemployment rate.

The population data of women by parity (i.e. number of children ever born) are important information for projecting the fertility rate. They are normally obtained from current population statistics such as the Population Census. In Japan, however, such data have not been collected in the Population Census since 1970 in response to the privacy concerns of the people. This paper presents the results of an estimation of the number of women by parity based on the own-children table of the Population Census, including an evaluation of the accuracy of the estimation, and comments on the trend of cohort fertility in recent Japan. The results of the estimation of the data by parity are found to be consistent with another estimation based on the vital statistics. It is concluded that the estimation is useful for the fertility analysis in Japan.

The principle of “evidence based policy” has been emphasized, but is too often ignored in practice. In the past few decades, “neo-classical” economists and “Keynesians” have fiercely disputed over necessity and efficacy of fiscal policy, monetary policy, deregulation and marketization etc, but both sides based their arguments mainly on pure logic of their theories, but not on deep analysis of statistical data. But any policy, whether intervenient or regulatory, has its own effects on economy depending on circumstances, and they can be predicted or analyzed only through empirical observations. It is necessary to implement the principle of EBP in the discussion of any short, medium and long range economic policies.

Consider this a report "from the trenches" of a government committee responsible for economic statistics. For the first time in 60 years the Japanese Diet revised completely the Statistics Act. It created a "Statistics Commission," and the Commission formed plan for improving official statistics. It faced a formidable challenge indeed. Observers had identified a wide variety of ways in which it could -- and should -- improve the data it collected and disseminated. This article is an on-the-ground report from the official working group charged with reforming the collection and dissemination of those economic statistics. Unfortunately, the process is not going well. Given the government control of the commission process, it is not likely to go well any time soon. Only through a continuous process of candid, open, and free discussion among the scholars and observers interested in the information -- scholars and observers such as the readers of this e-journal -- will the statistics available improve. On this front, Japan is substantially behind the world standard.