Estimating the Distribution of Net Worth and Disposable Income Using Johnson’s S_U Distribution

While there has been increasing interest in the distribution of net worth and disposable income, these variables include negative values, making it impossible to use the distribution functions previously employed in estimating gross income distributions, as these distribution functions cannot accommodate negative values. In this study, we propose the Johnson’s S_U distribution to estimate the distribution of net worth and disposable income. The S_U distribution function is defined over the entire real number space and is one of the most flexible parametric distribution functions for capturing a wide range of skewness and kurtosis. Therefore, it is highly suitable for estimating the various shapes of distributions for net worth and disposable income, which can include negative values. We derive the Lorenz curve for the S_U distribution in a closed-form expression. As an illustrative example, we apply both the univariate and bivariate S_U distributions to estimate the distributions and Lorenz curves of net worth and disposable income using the survey data of Korea.