中文摘要：

基尼系数测算本身并不复杂，然而由于方法对于数据要求高，传统测算方法无法在实证中应用．而现有文献采用的间接估计方法会导致结果与真实情况存在很大的估计偏差。本文的贡献在于通过中国健康与营养调查（CHNS）的个人实际总收入数据进行研究，弥补传统基尼系数测算方法的实证空白。同时，对基尼系数估计方法进行适当改进，首次在估计系数的同时可以得出估计标准误和置信区间，将这种方法应用于我国城乡基尼系数计算，并提出相关建议。

英文摘要：

Gini coefficient is straight forward itself but hard to use for rare data requirements. Few empirical literatures can be applied in traditional calculation methods, and the existing ones using indirect methods may lead a big deviation between their results and real situation＇s. This paper complements the existing literature of traditional method with the actual total personal income data from the China Heahh and Nutrition Survey（CHNS）. Meanwhile, we improve the Gini coefficient estimation method. For the first time, we can draw the estimated coefficient, the estimate standard errors and confidence intervals at the same time. This method is appliable to the urban- rural Gini coefficients, and feasible suggestions are put forward.