中文摘要：

利用1990-2012年的统计数据,测算了中国内地各省区的物质资本存量和异质人力资本存量。基于＂索洛余值法＂,构建了能源和异质人力资本也作为投入变量的面板回归模型,验证了能源消费和异质人力资本因素对经济增长率单门限效应的存在。若欲提高全要素生产率（TFP）的测量精度,应该考虑门限效应的影响,能源消费的门限值是0.1508,异质人力资本的门限值是0.0110。计算了各省区1990-2012年的TFP值,并进行了TFP收敛性和门限效应分析,除西部地区外,国家整体、东部及中部都有明显的σ收敛效应。全国及三大地区均存在条件β收敛效应,只有中部地区具有绝对β收敛效应,即只有中部地区具有俱乐部收敛现象。在TFP收敛中,不存在异质人力资本的门限效应,但能源消费的门限效应是存在的。

英文摘要：

This paper estimates the stock of the physical capital and heterogeneity of human capital in China＇s mainland provinces using statistical data during 1990-2012. Based on the ＂Solow residual method＂, the paper chooses energy and heterogeneous human capital as an input variable and builds a panel regression model. If you want to improve the measurement accuracy in total factor productivity（TFP）,it should consider the impact of threshold effects. Then calculate the value of TFP of provinces in 1990-2012, and conduct the TFP convergence and threshold effects analysis. Except the western region, the country as a whole, as well as the eastern and central part of China all have significant σ convergence effects. The whole nation and China＇s three major regions present conditions β convergence effect,and only the central region has an absolute β convergence effect. In convergence TFP, there is no threshold effect of heterogeneous human capital, but the energy consumption threshold effect exists.