中国制造业劳动生产率增长来自全要素生产率变动还是要素积累效应——基于状态空间随机前沿面板模型的计量分析

Factor Accumulation Effect of Total Factor Productivity Change,Which Determine Labor Productivity Growth of China's Manufacturing Industry:Empirical Test Based on State-space Stochastic Frontier Panel Data Model

基于超越对数生产函数构建具备时变截距项和时变斜率参数的随机趋势模型,采用卡尔曼滤波和卡尔曼平滑算法估计模型参数,并运用制造业劳动生产率增长分解方法分析制造业劳动生产率增长来源,识别全要要生产率变动、资本和中间投入要素积累效应对制造业劳动生产率增长的贡献。结果表明:资本投入和中间投入的一阶斜率参数具有显著的时变特征,且中间投入的一阶斜率参数远高于资本投入,各省中间投入要素积累效应对制造业劳动生产率增长的贡献远高于资本投入要素积累效应和全要素生产率变动,全要素生产率变动对制造业劳动生产率增长的相对重要性强于资本要素积累效应;由于资本和中间投入一阶系数显著时变上升,资本和中间投入的技术进步偏离成为全要素生产率变动的重要来源,且对制造业劳动生产率增长具有显著的正向影响。

In the paper, we construct a state-space stochastic frontier production function with translog praduetion function. Definitely the stochastic trend model which has time varying intercepts and slopes are estimated with Kalman filter and Kalman smoothing algorithms, and the growth account of labor productivity in manufacturing industry is taken to analyze the source of labor productivity growth： The relative contribution of TFP, labor factor accumulation, and capital factor accumulation are separated. The empirical results show that, the first order slope parameters of labor and capital are basically time variant and parameters of labor are greater than that of capital the relative contribution of intermediate inputs factor accumulation effect is greater than that of capital factor accumulation effect as well as total factor productivity change, total productivity change playing a more important role in labor productivity growth rather than capital factor accumulation effect; since the first order slope parameters of capital and intermediate inputs have a significant Time-varying rising trend, the bias of labor and intermediate inputs in technical change appear as a important contributor to total factor productivity change, and each of them has a significant positive contribution to the manufacturing labor productivity growth.