Abstract: Although China has made great progress in urbanization，there remain a range of problems and challenges. One is the income disparity between migrant workers and urban locals. This paper utilizes the RUMiC baseline data to empirically analyze the income disparity problem using unconditional quintile and RIF regression decomposition method. The data show that the income gap increases from low to high quintile. The proportion of income disparity explained by structural effects also increases with the growth of income. However，there exists favorable income determination mechanism for migrant workers in the low income quintile. The main factors that contribute to the composition effects are human capital and occupational /industrial differences. Implications for policies in reducing income disparities are discussed.

Keywords: Urbanization, Income Disparity, Unconditional Quintile Regression, RIF Regression Decomposition

摘要： 我国的城镇化取得了巨大的成就，但在城市内部依然存在诸多需要解决的问题，其中之一是外来农村居民和本地城市居民的收入差距问题。文章利用2008 年中国城乡劳动力流动( RUMiC) 数据，基于无条件分位数回归以及RIF回归分解的方法，考察造成外来农村居民和本地城市居民收入差距的因素。数据显示两者的实际收入差距从收入分布的低端到高端逐渐扩大。随着收入的增加，由结构效应所解释的收入差异部分逐渐增大，但在收入分布的底端，却存在有利于外来居民的收入决定机制。人力资本和职业/行业的差异是解释由要素分布不同所引起的收入差距的主要原因。以人为本的城镇化应妥善处理好农民工和城市居民的收入差距问题。

关键词: 城镇化, 收入差距, 无条件分位数回归, RIF 回归分解