中国工业企业全要素生产率的稳健估计

Robust Estimation of Industrial Enterprises' Total Factor Productivity in China

对企业全要素生产率的稳健估计,由于生产函数模型、样本范围与数据处理各不相同,估计结果差异较大。为此,本文研究了生产函数模型、样本范围和价格因子三种因素对不同估计方法的影响,并在统一的处理方式下对比了不同估计方法的估计结果。研究表明:(1)生产函数模型设定对估计结果影响最大,样本范围次之,价格因子影响相对较小。(2)不同的估计方法估计结果之间存在较大差异,意味着对企业全要素生产率的估计仍然存在大量干扰因素有待消除。(3)除非数据存在严重的测量误差,否则指数法是全要素生产率增长的一致精确估计。本文的价值在于清楚厘定了估计全要素生产率过程中各种要素的影响性质与影响大小,确定了不同估计方法的应用条件与适用范围。

The estimation of enterprises total factor productivity is important and fundamental. Due to different production function model, sample range and empirical strategies, estimation results from different researchers are quite distinct. Therefore, we study the impact of production function model, price factor and sample range on the estimation methods, and then compare the results of different estimation methods in a unified approach. We find that ： （ 1 ） Production function model has the most impact on estimation, followed by sample range, while the effect of price factor is relatively small ; （2） The huge variation between estima- tion results means that there are still a lot of interference factors to be eliminated ; （3） Unless there exists serious measurement error, index method is consistent with an accurate estimate of productivity growth. The contribution of our paper is to figure out the impact of different factors in the estimation of TFP, and deter- mine the scope of application of different estimation methods.