中文摘要：

研究目标：考察经典方法估计全要素生产率时是否因加成定价或规模报酬效应而出现估计偏差。研究方法：基于放松完全竞争和规模报酬不变假设的增长核算框架估计全要素生产率，并在分别估计加成定价率和规模报酬率基础上对古典索洛余值进行分解。研究发现：加成定价效应是造成古典索洛余值与全要素生产率偏差的重要原因；富有弹性的劳动力市场可以降低全要素生产率受到经济冲击的影响。研究创新：基于扩展后的增长核算框架估计了中国、韩国、日本和美国的全要素生产率，分析古典索洛余值是否存在估计偏差以及偏差的主要来源。研究价值：分离新古典索洛余值中的加成定价和规模效应，可以有效降低经验研究中全要素生产率的估计偏差。

英文摘要：

Research Objectives： To study whether the total factor productivity estimated by the classical method is biased due to the mark-up and the degree of return-to-scale. Research Methods： The total factor productivity is estimated based on the growth accounting frame- work, which is based on the assumption of relaxed competition and invariant scale, and the classical Solow residual is decomposed on the basis of the mark-up rate and the degree of re- turn-to-scale respectively. Research Findings： The mark-up effect is the main reason for the deviation of the classical Solo residual and the total factor productivity. The resilient labor market can reduce the impact of TFP on economic shocks. Research Innovations： Based on the expanded growth accounting framework, the total factor productivity of China, Korea, Japan and the United States is estimated, and we analyze whether the classical Solow residu- als exist the estimated deviations and the main sources of bias. Research Value： The separa- tion of the mark-up and the degree of return-to-scale effect from new classical Solo residual can effectively reduce the estimated deviation of total factor productivity.