Research On The Pattern And Determinants Of Change In China's Labor Income Share

One of the main research topics in Economics is the “Income Distribution” or “Inequality”,and research on “Inequality” has been exploring the “Factor Income Distribution”,in other words,how national income is allocated between capital and labor.Labor income share,which is measured as the ratio of labor compensation out of national income,is the key indicator of factor income distribution.Classical researches exploiting both time series and cross-sectional historical data find that labor income share shows “constancy” property,which implies that labor income share has no observable trend.As one of the well-established “Kaldor's Stylized Facts”,this property is widely regarded as the cornerstone of modern Macroeconomics.However,recent researches show that there exists a global decline in labor income share since the early 1980 s,and China's labor income share has experienced a much more substantial decline than other countries since the mid-1990 s.Downward trend of labor income share not only contradicts the conventional “constancy” property,but also poses a great threat to sustained economic growth through enlarging income inequality or triggering other economic issues.This thesis tries to explore whether China's labor income share evolves in some specific pattern or not,and understand the underlying determinants of change in labor income share.Generally speaking,change in labor income share is determined by either its underlying pattern or economic environment change,and verify whether labor income share evolves in specific pattern or not will establish the benchmark for further discussions.Solid empirical analyses need reliable data,however,statistical adjustment is not universal across provinces,and the definition of labor compensation in China's System of National Account(SNA)does not reflect the real labor income.In this thesis,we construct China's provincial labor income statistics by applying coherent rules.The trend and scale of revised labor income share is quite different from the raw data.Besides,the direction of changes in provincial labor incomes diverges substantially even though these provinces share the similar economic development level.Further empirical analyses indicate that economic development has no significant impact on labor income share,which is consistent with classical researches.This finding implies that labor income share is trendless.However,when we use the raw data,we find that labor income share evolves in “U-shape” pattern alongside economic development,which is consistent with recent works.Compared with classical researches,recent works take the structural transformation into consideration,and China's economy is indeed in the process of transformation,hence,we develop a general equilibrium model to characterize this important feature.We find that the impact of economic development on labor share depends on two parameters,the relative labor intensity between agricultural sector and non-agricultural sector as well as the consumption elasticity of substitution between agricultural and non-agricultural goods.Labor share is either constant or changing monotonously with economic growth,and the changing direction is governed by these two parameters.Therefore,our empirical finding is a special case of our theoretical model's predictions.After confirming that labor income share has no specific pattern,this thesis then explores which factor cuts labor income share.We observe that “privatization” starts in the early 1980 s around the world,and China's government also initiates the state-owned enterprises reform by using the “Retain the Large,Release the Small” strategy in 1996.These policies definitely reduce the share of state-owned economy(SOE).Decline in the share of SOE as well as observed decline in labor income coincide with each other,which poses a natural question worth exploration: what is the relationship between SOE and labor income share? By exploiting the 2004 Economic Census raw data,we construct “Cell” as unit of analysis,and these cells can capture regional features and industrial features simultaneously.Empirical results indicate that the share of SOE has a substantial positive impact on labor income share.We also examine the impact of the share of SOE on labor income share by aggregating data within regions or within industries,and we find that the impact of the share of SOE is not robust,which is consistent with existing researches.One possible explanation for this discrepancy is that there exists significant difference in the structure of industries among regions,and aggregation among either regions or industries loses too much information.We then explore how SOE affects labor income share.By definition,labor income share is the ratio of labor compensation per capita and labor productivity,so we can examine the impact of the share of SOE on labor compensation per capita and labor productivity separately.However,the employees of state-owned enterprise are more educated than that of non-state-owned enterprise,and the average working hours of workers in state-owned enterprise is lower than that of non-state-owned enterprise,in other words,labor inputs among firms present great heterogeneity.We link the 2004 Economic Census data and the 2005 1% Population Sample Survey of China data,and make adjustment of labor input to education and working hours.Empirical results illustrate that,once we adjust working hours,the share of SOE has a significant positive impact on labor compensation per capita,and has no significant impact on labor productivity,which implies that SOE can improve factor income distribution only through the “income effect”.However,if we follow existing literatures and measure labor inputs by using the numbers of workers,we find that the share of SOE can significantly increase labor compensation per capita,and will significantly decrease labor productivity,which implies that SOE can improve factor income distribution through both the “income effect” and the “productivity effect”.State-owned economy offers higher labor compensation may be the result of higher worker's bargaining power in state-owned enterprise.In advanced capitalist countries,collective bargaining is conducted by union,and union definitely defends worker's interest.Union is also widespread in China,and implementation of China's labor protection policies highly depends on union.Therefore,we then explore how union affects China's factor income distribution.We derive the analytical relationship between union and labor income share by using McDonald and Solow(1981)'s wage bargaining model,and we find that the more bargaining power union poses to workers,the greater labor income share in equilibrium.We then conduct empirical analyses at both firm level and cell level by using the 2004 Economic Census data and the 2005 1% Population Sample Survey of China data.Empirical results indicate that union has a significant positive impact on labor income share,since union will increase worker's labor compensation mainly though increasing their welfare incomes,while decrease labor productivity mainly though decreasing working hours.This thesis firstly shows that labor income share has no specific pattern,which is consistent with classical researches,therefore,we cannot see factor income distribution improves automatically as economic growth.We find that decline in the share of state-owned economy cuts labor income share substantially,and SOE does not necessarily exhibit lower labor productivity,which helps us evaluate the value of state-owned economy in efficiency and equality more comprehensively.Finally,we demonstrate that China's union can improve factor income distribution,this finding not only helps us learn the function of China's union,but also sheds light upon that the empowerment of union may be an effective way to improve China's deteriorating factor income distribution.