A0690
Title: Bayesian dynamic modeling of Gini coefficient from grouped data
Authors: Kazuhiko Kakamu - Nagoya City University (Japan) [presenting]
Abstract: A dynamic model is proposed for income distribution, which enables us to examine the Gini coefficient directly. In the analysis of income distribution, the choice of the hypothetical distribution is crucial and reflects on the results. It means that the dynamics of the Gini coefficients may be different depending on the choice of the hypothetical income distribution. Our approach also enables us to examine several kinds of hypothetical income distributions. This model is applied to several simulated and real datasets. The results of a real dataset, which comes from a Family Income and Expenditure Survey, show the importance of the choice of the hypothetical income distribution.
