ESRI Discussion Paper Series No.358 In Search of Accurate Measures of Income Inequality across Japanese Households

Masahiro Hori
Economic and Social Research Institute, Cabinet Office, Japan;
School of International and Public Policy, Hitotsubashi University, Japan
Saeko Maeda
Economic and Social Research Institute, Cabinet Office, Japan
Fumihiko Suga
Economic and Social Research Institute, Cabinet Office, Japan;
Department of Economics, Kyushu University, Japan


Using microdata from the National Survey on Family Income and Expenditure (NSFIE), this study tries to provide new accurate measures of income inequality across households in Japan from 1990s to 2000s. In order to correct for potential biases in conventional measures, we use microdata from the Population Census to construct original sampling weights. We calculate multiple income inequality measures, such as the Gini coefficient, the relative poverty rate, the mean log deviation (MLD) of income, and the log variance (LV) of income, with our original sampling weights and find that actual income inequality across Japanese households likely is larger than suggested in earlier studies. Further, while our new estimates confirm the findings of previous studies that income inequality increased throughout the 1990s and 2000s, the rate of increase for disposable income was quite moderate due to the redistributive effects of the tax and social security system. We also find that approximately 40-50% of the increase in income inequality in the 1990s, and 30-40% of the increase in the 2000s, resulted from changes in household compositions, such as a decrease in the number of family members living together and increases in the shares of jobless households and dual-income households.

JEL Classification: D31, N35, C83
Keywords: income distribution, inequality, sampling weight adjustment

Structure of the whole text(PDF-Format 3 File)

  1. Introduction - Referencesopen pdf in new window(PDF-Format 404KB)
  2. 1. Introduction
  3. 2. Data Sources
  4. 3. Empirical Methodology
  5. 4. Results
  6. 5. Conclusion
  7. References
  8. Tables & Figuresopen pdf in new window(PDF-Format 2.8MB)
  9. Appendixopen pdf in new window(PDF-Format 3.6MB)