Improvement On The Measure Of China’s Urban-Rural Income Gap And Its Decomposition

The big income gap between urban and rural will hinder the integration of them. However, there are two kinds of problems about the measurement of China’s Urban-Rural Income Gap:First, the belongings of migrants’income. The reform of Hukou System can’t catch up with the flow of rural labor force. Therefore, the migrants can’t receive the same payments and public services as the urban residents. Theoretically, there are different opinions about the belongings of migrants’income. Meanwhile, migrants aren’t covered by the Household Survey. These all cause trouble when we measure Urban-Rural Income Gap. Thus, this paper compares the urban-rural income gap under different statistical belongings of migrants’income, and their implications.Second, the computation of the overall Gini coefficient. This is a special problem under the current statistical methods and the availability of the data, the major aspect is how to precisely calculate the overall Gini coefficient when there are only separated group data of urban and rural. Subgroup decomposition approach can’tlve the problem caused by the overlap between urban and rural data, functional aggregation approach solves this problem by aggregating the income distribution or density function of urban and rural. However, this approach relies on the maximum income of the country and each group, In this study, we solve these drawbacks by aggregating the Indirect Lorenz Curve of urban and rural. We use a new index to measure the change of the Urban-Rural Income Gap under the framework of aggregating Lorenz Curve.When solving these two problems, we make two improvements of the measurement of China’s Urban-Rural Income Gap. We found the Urban-Rural Income Gap in China are still very high. Under this circumstance, mean is not a good represent of the whole income distribution any more. Therefore, we use a distributional decomposition method based on Unconditional Quantile Regression (FFL Decomposition) to decompose the income gap between urban and rural workers, to see the contribution of endowments and return in different quantiles.To solve the questions mentioned above, we use subgroup decomposition, Gini coefficient’s calculation by aggregating the Indirect Lorenz Curve, distributional decomposition method, we found that:First, both mathematical analysis and statistical test shows, if we neglect the income of migrants, then the urban-rural income gap will be overestimate. The urban-rural income gap under different belongings of migrants’income, depends on the relative difference of migrants and urban residents’income, and that of migrants and rural residents’income. If the relative difference of migrants and urban residents’income is smaller, then urban-rural income gap based on migrants’income belongs to rural is smaller.Second, the new approach for calculating overall Gini Coefficient based on the aggregation of the Indirect Lorenz Curve, can make up the underestimate by using grouped data and error caused by the overlap between urban and rural. What’s more this approach don’t rely on the maximum income. When we use the new index to measure the urban-rural income gap, we found it is a little bit different from other indices. The new index shows it was decline in2002, while other indices show it was increasing. The reason is that our new index not only considered the mean difference of urban and rural income, but also the difference of their distributions.Third, FFL decomposition shows the proportion of unexplained factors in urban-rural income gap is about20-50percent, while it is50percent in75quantile, and20percent in90quantile. Thus Hukou discrimination is still very severe. Moreover, education plays a significant role in the urban-rural income gap. The return of education on urban residents are3times more than that of migrants.