A0725
Title: Spatial heterogeneity in the effects of covariates on the distribution of income using Bayesian nonparametric methods
Authors: Ziyou Wang - Kings College London (United Kingdom) [presenting]
Maria Kalli - Kings College London (United Kingdom)
Abstract: The purpose is to extend a prior study that developed normalized latent measure factor models. It models a collection of distributions using linear combinations of latent factor measures. These latent factor measures are modelled as infinite mixtures, the weights of which are interpretable as characteristic traits shared by different distributions. The prior study focuses on estimating the density of personal income for the public use microdata sample (PUMS) classifications of the state of California. The mentioned study is extended by looking at how personal income is distributed in the other US states and by considering multiple discrete value covariates, such as gender and race. This allows gaining valuable insights into how the covariates affect the income distribution and how these effects vary over space.
