（1.南开大学经济学院；2.南开大学经济与社会发展研究院）

    【摘要】本文将反映行业间生产率联动的购买距离矩阵和销售距离矩阵引入空间自回归模型，研究了1998年、2003年和2008年三个时期行业间生产率联动对我国工业生产率增长的影响。为了克服引入社会经济距离矩阵带来的异方差和矩阵的行标准化问题，本文采用空间GMM法进行模型的估计。结果表明，行业生产率联动对我国工业生产率增长具有显著的正影响，并且在资源密集型、劳动密集型和资本密集型工业行业中，行业间生产率联动对工业生产率增长的影响相对于其他因素的影响更为稳健。此外，由销售距离矩阵所体现的联动作用效果整体上大于购买距离矩阵体现的相关效果。

关键词  生产率联动  经济距离  空间GMM估计

中图分类号  F404.3             文献标识码  A

The Role of Productivity Comovement Across Industries in China’s Industrial Productivity Growth

Abstract: This paper uses the spatial autoregressive model with buy distance matrix and sell distance matrix, analysis the role of productivity comovement across Industries in China’s industrial productivity growth in 1998, 2003 and 2008. To avoid the heteroskedasticity and matrix’s row-normalized problem caused by social economic matrix, this paper uses spatial GMM method for model estimation. The empirical result shows that inter-industrial productivity comovement has a positive and significant effect on the industrial productivity growth. The effects of resource-intensive industries’, labor-intensive industries’ and capital-intensive industries’ productivity comovement on industries’ productivity growth is more robust than other factor’ s effect. The comovement effect by sell distance matrix totally greater than the comovement effect by buy distance matrix.

Key words: Productivity Comovement；Economic Distance；Spatial GMM Estimation