Research On Human Capital Heterogeneity And Threshold Effect

This thesis analysis the macroeconomic relationship between education of human capital and economic growth within the framework of endogenous growth theory using Chinese provincial panel data, also explore the chanel and the structural change of human capital impact on economic growth. We focus on several issues. Firstly, the impact on technological progress caused by the composition of human capital and the distance to the frontier of tecknology, both of them will determine the direction of human capital between the innovation and the imitation sectors, eventually determine how fast the technical progress will happen. Secondly, education of human capital threshold effect in economic growth, human capital threshold effect is not only reflected in the direct effect, more significantly reflected in its indirect effect. The comprehensive results of human capital threshold effect is to promote the expansion of the production possibility frontier. The last issue we are interested in is the classification of human capital threshold effects. Because the skilled and unskilled human capital plays different role in the technological progress, we need to distinguish their threshold effect. Besides the existence of a significant threshold effect suggest that there is a bias of the social return of human capital, and the classification of human capital threshold effect helps to identify the direction of the bias.The article contains seven chapters. By the perception of China’s rapid accumulation of human capital along with the potential of China economic growth is a gradual decline, and then by means of literature study and comparative analysis of the domestic and foreign as well as explore the mechanisms and channels of human capital’s impact on China economic growth by theoretical models and empirical analysis to get a rational understanding of the relationship between education and economic growth. The main contents of each chapter is organized as follows.The first chapter is about the basic statistical data summary. We focus on six aspects to describe the employees’ s characteristic such as educational attainment, human capital variance and education Gini coefficient, funding for education at all levels of school enrollment rates and graduation rates, family and social investment on education, as well as R&D personnel. The data summary helps us to form an initial understanding of China’s human capital development, and propose starting point for our reseach.The second chapter is a literature review. This chapter centre on several aspects of human capital and economic growth in the representation of documents to sort out, including the measurement of human capital, the impact of human capital on technological progress and economic growth and other issues. And then I summarize the motivation for this study by comparing the foreign theoretical and empirical literatures to the domestic ones.The third chapter is aim to estimate the essential data and do the growth accounting preliminarily. I discuss the measurement issues of Chinese provincial capital stock in detail and focus on the selection of depreciation rates s, and propose an approach based on statistical data to estimate the average depreciation rate. Using the estimated depreciation rate I re-estimate the Chinese provincial capital stock. In addition, this chapter also raised a human capital model with which efficiency adjusted. On the basis of the adjusted human capital, this paper comes out a conception of fundamental human capital. After the essential data has been estimated correspongdingly, each parameter setting will be calibrated in this chapter, such as the share of capital in income, the depreciation rate, and the parameter used to adjust efficiency. We need to know how the parameter will affect the explanating power of income differences among regions. The main purpose of this chapter is aim to combined the calibration procedure used in Caselli(2005) to calibrate the parameters and analyze whether the setting of these parameter have any impact on the subsequent empirical study, which will pave the way for robust test. After this work, we may identify the sensitive parameter, so the robustness test part of empirical research will be more targeted.The main study in the fourth chapter is about the impact of the compositon of human capital and the technological distance to the frontier on the technical progress. Using China provincial panel data(1996-2012) to verify the theory proposed by B&S(1994), we may find that the imitation effect of human capital is negative by the whole sample. Huang et al(2013) also suggest that there is a very big difference among the regional performance after zoning eastern, central and western, the imitation effect of human capital in eastern is also negative(but not significant), the empirical results shows a departure with theoretical expectations so that we must further explore the mechanisms of human capital affect economic growth. VAM(2006) studied the impact of human capital composition on the technological advance with theoretical and empirical model, effectively solved the B&S(1994) empirical study of sub-sample regression, and answer the K & L(2001) puzzle. This chapter using the framework of VAM(2006) to investigate how the structure of human capital affect China’s technological progress(total factor productivity growth). This study find that there is a significant impact on technological progress from the structure of human capital. Indeed, a higher total level of human capital does not necessarily accompanied by acceleration of technological progress, an increase of skilled labor is a source of technological progress.In the fifth chapter, we study China’s human capital threshold effects in detail with the structure of human capital is not considered. The study finds that the threshold effect of human capital in economic growth does exist and is significant. By the mesure of education attainment, the threshold of human capital per capita up to 9 years, which is higher than the current compulsory education level required. And the existence of human capital externalities threshold suggest that a linear model may make the social return of human capital biased. Because the threshold effect shows that human capital will attract more physical capital and FDI and other factors, the following functions in linear model in this regard is often overlooked. Human capital accumulation drives innovation and imitation from technology frontier, so it may improve the the rate of return of physical capital and FDI and other elements. A region with its the human capital exceed the threshold level will accompanied by the rate of return of physical capital and FDI increased significantly. However, the innovation effect of human capital is not consistent with expectation, after crossing the threshold of human capital, innovation effect seems weakening. On the contrary, the imitation effect is slightly strengthened. After all, the human capital threshold effect is mainly reflected in the indirect effects but not direct effects, the overall effect is to promote a substantial expansion of production possibility frontier outward.Chapter VI revisited the threshold effect with considering the composition of human capital by the decomposition of unskilled and skilled human capital. The result shows that unskilled human capital threshold effect is mainly reflected in the direct effect to improve its own rate of return, but the indirect effect of skilled human capital threshold effect is significant to improve the factor’s rate of return such as human capital, physical capital. Both types of human capital help to improve the rate of return on FDI, and promote attract FDI further. After decompositing human capital into unskilled and skilled types then re-investigate their threshold effect, we find that the social rate of return of unskilled human capital tend to be overvaluted due to the existence of physical capital "crowding out" effect. However the corresponding rate of skilled human capital is apt to be underestimated for further attracting physical capital and FDI to promote economic growth, and this effect is often ignored in linear model.The main conclusions are concluded in chapter VII, which also summarize some policy implication as well as the prospect to expand this research. Thesis main conclusions and policy implications can be summarized as follows.(1) When we use "the proportion of skilled human capital" to measure the sutructure of human capital, the technological distance to frontier shows negative and significant impact on TFP. That is, keeping the other conditions remain unchanged, the distance between the technology and the technological frontier region shrink would accompanied by the economy growth slowed down. The proportion of skilled human capital on growth effect is significantly positive. The interaction term shows positive but not significant effect on economic growth suggest that the propotion of skilled human capital is growth-enhancing. After decomposing the human capital into skilled and unskilled of human capital, the regression result shows that both skilled and unskilled human capital help to promote economic growth with the technological distance effect is still negative. And the negative marginal effect of the technological distance will be enhanced by unskilled, but weakened by skilled human capital which are consistent with VAM(2006).(2) The threshold effect of human capital in economic growth present significantly. There are double threshold levels of human capital for immediate effect. A region across the first threshold into a transition stage may appear physical capital, FDI substituted by moderate level of human capital then the output growth shows alternative slowdown. When the region’s human capital across the second later, the output elasticity of physical capital, FDI and other are rising sharply. Human capital externalities estimates of 0.25 or more, that is means after crossing the threshold of human capital, 1 percent increasement of physical capital, FDI will get more 0.25 percent output growth. The immediate effect of imitation was not significant, the raise of the level of human capital significantly promote technological innovation indeed outputs, but the marginal effect declines as a higher stage of development. The lag effect presents a single threshold, and the threshold externalities up to 0.24. It means that the elasticity of physical capital and FDI increased dramatically ater crossing the threshold. The lag effect of innovation and imitation of human capital is positive but insignificantly for innovation. Both immediate and lag effect, the direct effect of human capital are negative. A further decomposition helps to find that the innovation effect decline and imitation effect increases after crossing the human capital threshold. This phenomenon may result from the inefficiency for intersectoral allocation of human capital. China’s human capital may occur allocative distortions, lack of skilled human capital but heavily concentrated in imitation sector leads to a weakening innovation but imitation effect is enhanced.(3) Without considering the threshold externality of human capital, then the raise of human capital contribute little to economic growth. Nevertheless, when the threshold externality is considered then the accumulation of human capital plays an important role in economic growth. The law of compulsory education and expansion of university enrollment policy allows China surpass the threshold of human capital which threshold externality may promote long-term growth for regions. Taken in this sense, both policies indeed contributed to economic growth. However, a problem needs to be noticed is that the allocative distortions for Chinese human capital. How to encourage innovation through institutional mechanism to drive the skilled human capital transfer from imitation to innovation sector remains a major challenge for the sustainability of China’s growth. Furthermore, the estimation results show that the threshold of per capita education attainment is higher than 9 years, that is beyond the current age of compulsory education, and therefore extend the compulsory education from junior to senior hight school should be considered now.(4) There is a big difference between the threshold effect of unskilled and skilled human capital in economic growth. The decomposition of unskilled and skilled human capital to investigated their threshold effects helps to find that the threshold effect of unskilled human capital directly contribute to its own rate of return but crow-out the rate of return of physical capital. On the contrary, the threshold effect of skilled human capital may not directly contribute to accelerate the rate of return on human capital increase, but driven by innovation and technological progress makes the output elasticity of physical capital increased significantly.(5) Compared to the threshold model, the social yield of human capital is biased in linear model. Based on the changes for the yield of various elements before and after the human capital across the threshold level, it can be seen that human capital helps to attract physical capital and other factors thus promote economic growth. However, this kind of "external effects" in linear regression model may not identified, which makes the social return underestimated in linear model. The rate of return of unskilled tends to be overstated due to the existence of "crowding out" effect for physical capital.After all, the rate of return of unskilled exist upward bias yet downward bias for skilled.