Evolution Of Income Inequality In China

This thesis exploits micro data to examine the evolution of income inequality among Chinese workers, as well as its causes and impact fators. We focus on three main aspects:(1) the rising return to education, due to technical change, and its impact on income inquality; (2) the role of college enrollment expansion policy, combining with labor market discrimination and skill-biased technical change, in changing the college non-college wage gap; (3) the impact of worker's unobserved latent ability and stochastic income shock on the evolution of residual income inequality.The thesis consists of seven chapters, which are organized as follows:The first chapter describes the micro data in details. It characterizes the unique features of the evolution of income inequality among Chinese workers, from which the main focuses of this thesis come.The second chapter reviews the broad literature on income inequality. It starts from studies of foreign researchers that cover a wide range of topics related to income inequality, including theoretical models describing the mechanism behind the rising inequality as well as statistical methods that decomposing inequality into contributions of observed labor characteristics. And then, we move to studies of domestic researchers which focus on China-specific factors that affect Chinese workers' income inequality.In third chapter, we establishes a general equilibrium model based on the framework of Skill-Biased Technical Change to illustrate that change in return to education is an important source of rising income inequality in China since early 1990s. The model shows that the total effect of education on income inequality can be decomposed into factor price effect and factor composition effect. While the former widens the magnitude of income inequality, the latter tends to narrow it. However, the total effect is still the rising of inequality. Using newly developed RIF regression techniques, our empirical results indicate that the model's predictions are highly consistent with the data.The fourth chapter introduces a model of signaling game. This model is used to explore the contribution of human capital accumulation effect to college wage premium, as well as the effect of university enrollment expansion to college-noncollege wage differential. Results from structural estimation indicate that the model performs very well in fitting wage distribution. Counter-factual simulation suggests that human capital accumulation effect account for nearly 72% of college wage premium in China. And, under very weak model condition, the model also predicts that university enrollment expansion can enlarge college-noncollege wage differential through statistical discrimination.In chapter five, we extend the model presented in chapter four by including Skill-Biased Technical Change. The new two-period model can also account for contributions of worker's individual characteristics on their wage. As a result, the new model can be used to predict the college non-college wage gap after college enrollment expansion more precisely.Chapter six studies the residual income inequality, which is a measure of within group inequality among workers with the same characteristics. We separates the impact of unobserved heterogeneity, stochastic income shock and their corresponding price effects on worker's residual income inequality in a unified framework. The results suggest that transitory income shock is the main source of residual income inequality, which accounts for more than 60% of its variance. However, the price effect related to worker's unobserved heterogeneity is responsible for the successive rising of residual income inequality in recent years.In chapter seven, we summarize the main finding of this thesis and discuss the policy implications of our study. Possible extensions of this study are also discussed.The main conclusions of this study can be summarized as follows:1. Skill-Biased Technical Change (SBTC) exists in China. It enhances the return to education and leads to increase of income inequality among Chinese workers. Moreover, SBTC enlarges income inequality mainly by increasing the dispersion of upper tail of the income distribution.2. Owning to the existence of Skill-Biased Technical Change, the college enrollment expansion policy may not necessarily narrow down the college non-college wage differential. However, the college enrollment expansion policy discourages worker's investment in human capital. As a result, the heterogeneity between college and non-college workers decreases.3. Transitory income shock is the main source of residual income inequality, which accounts for more than 60% of its variance. However, the price effect related to worker's unobserved heterogeneity is responsible for the successive rising of residual income inequality in recent years.This thesis improves the current study about Chinese income inequality in three ways.Firstly, although it is considered as the most import policy reform in higher education, the labor market consequence of college enrollment expansion policy has not been systematically studied by Chinese researchers. The same argument applies to the role of residual income inequality in China. Moreover, although existing literature has noted the increasing importance of return to education in wage determination, its impact on income inequality is still unclear. Thus, this study offers the first answers to all the three questions listed above.Secondly, our study has solid theoretical foundation. A general equilibrium model based on Skill-Biased Technical Change is developed to illustrate the impact of rising return to education on income inequality. Furthermore, we build up a signaling game based model to separate the impacts of human capital investment, innate ability and labor market signaling on college wage premium. The extended form of the model, which also accounts for SBTC, can be used to predict college non-college wage gap after the college enrollment expansion reform.Thirdly, our study exploits the cutting-edge econometric methods. Specifically, the newly developed Recentered Influence Function (RIF) Regression is adopted to decompose the price and structure effect of education on income inequality. Structural estimation techniques are used to estimate the model parameters of our signaling game, so that counter-factural simulation can be performed. Minimum Distance Estimator (MDE) is utilized to estimate the variance components of unobserved ability and random shock in the study of residual income inequality. All these techniques are seldom used by domestic researchers.