关键词: 全要素生产率, ACF方法, 时变产出弹性

Abstract: Solow Model, Stochastic Frontier Production Function Model and Data Envelopment Analysis Model are the three types of classical models in measuring Total Factor Productivity (TFP). All of these models ignore the endogenous problem of production function and the timevarying output elasticity of the model parameters, but ACF model improves and provides a more reliable way to estimate TFP. We deduct the endogenous problem and timevarying output elasticity of the parameters, and propose a new method to estimate timevarying output elasticity based on the ACF model. Using ACF model with our new method we reestimate the TFP in China with the data of 28 provinces of China from 1990 to 2017. The results show that the ACF model is more accurate in measuring TFP. From a national perspective, capital investment growth has the largest contribution to economic growth, and the contribution of TFP growth to China’s economic growth is decreasing. The contribution of labor input to economic growth is relatively weak and showing great volatility. From a regional perspective, the TFP level of all regions shows a downward trend. The growth rate of TFP in the northeast region of China was the lowest in recent years, and the TFP growth rate from 2012 to 2017 was negative. Meanwhile, the loss of labor in Northeast China is very serious.

Key words: Total Factor Productivity, ACF Method, Time-varying Output Elasticity