Japan's Inequality in Income Distribution

橘木(1998)の日本の所得格差についての見解に対して，多数の研究者が統計の利用の仕方に問題があり適当でないと指摘した．しかしながら，不平等度の比較に際して焦点となった，利用する統計データ毎になぜ計測結果が乖離するかの問いに対して充分な解明はなされていない．また，近年の所得の不平等化がいかなる要因によるかについて納得しうる結論は得られていない．本稿では所得分布の研究に利用されてきた統計調査のミクロデータにもとづく結果を使用して，計測結果の乖離が統計調査の精度上の問題に起因するのか，それとも調査方法等の相違によって説明しうるのかを分析した．乖離は対象世帯，定義，捕捉率，推計方法の相違を主たる原因としており，これらを調整すれば計測結果は接近することが明らかになった．さらに世帯所得の不平等化に親と子の同居率の低下が大きく影響していることを指摘し，世帯の変容と不平等の関係ならびに不平等の捉え方を検討している．Tachibanakis (1998) discussed Japan's inequality in income distribution, but several scholars pointed out that his handling of statistics was questionable and rather inadequate. Nevertheless, from my point of view, the question of why measured outcomes estrange from each statistical data, which was the focal point in comparing the degree of inequality, has not yet been sufficiently investigated. Further, no conclusive or satisfactory basis for the recent increase in income inequality has been presented. This paper uses results from income distribution research based on survey micro-data to analyze whether the discrepancies in measurement outcomes are caused by inaccuracy of the statistical surveys or by differing survey methodologies. As a result, the discrepancy seems to be mainly caused by households surveyed, definitions, the sample rate and inference methods. It is clarified that if these are rectified, measurements are more accurately approximated. Furthermore, this paper also emphasizes that the increasing inequality of household income is influenced by the decreasing rate of the less shared living, and examines how we should understand the inequality and the interaction between changing composition of households and inequality in income distribution