ESRI Discussion Paper Series No.378 Measuring Income Inequality in Japan Using Accurate Sampling Weights

Masahiro Hori
Economic and Social Research Institute, Cabinet Office
Graduate School of Economics, Hitotsubashi University
Saeko Maeda
Economic and Social Research Institute, Cabinet Office
Fumihiko Suga
Economic and Social Research Institute, Cabinet Office
Department of Economics, Kyushu University


This study attempts to provide an overview of income inequality in Japan for the 1990s and the 2000s, using the Comprehensive Survey of Living Conditions (CSLC) data. We calculate four income inequality measures based on eight income definitions. To measure the income inequality with a precision greater than that of the previous studies, we create the sampling weights using micro-data from the Population Census. We find that income inequality measures calculated using Population-Census weights are higher than those without weighting adjustment and those adjusted with provided weights. Although the levels of inequality measures are higher, weighting adjustments do not seem to have a significant impact on the trend of inequality measures. We also find that the level and the upward trend are less pronounced if imputed rent is considered. Moreover, we attempt to find the cause of the rise in inequality. We find that, on an equivalized disposable income basis, 31.6 to 57.4% of the rise in income inequality can be explained by the changes in the demographic structure and the composition of households for the 1990s. Among others, the aging of the population, changes in household composition, and a decrease in the number of workers had a large impact on the rise in income inequality.

Structure of the whole text (PDF-Format 1 File)

  1. Full Textopen pdf in new window(PDF-Format 4.5MB)
  2. Part 1/5open pdf in new window(PDF-Format 777KB)
  3. Part 2/5open pdf in new window(PDF-Format 1.1MB)
  4. Part 3/5open pdf in new window(PDF-Format 1.1MB)
  5. Part 4/5open pdf in new window(PDF-Format 1.1MB)
  6. Part 5/5open pdf in new window(PDF-Format 1.3MB)
    • 1 Introduction
    • 2 Related Literature
    • 3 Data Sources
    • 4 Empirical Methodology
    • 5 Results
    • 6 Conclusion