摘要
不同于传统的成本函数实证研究,本文关注企业间异质的全要素生产率对成本函数估计的影响。为此,本文首次从CES生产函数推导出一个包含全要素生产率的完整的成本函数模型,并以我国大型重化工企业2003—2007年的样本进行回归,同时使用传统模型与SFA模型回归,以进行方法论的比较。结果表明,全要素生产率确实对成本函数估计产生了重要影响,改良前的传统模型可能出现"遗漏变量"问题,使似不相关回归产生内生性问题。三种方法的回归结果表明:非国有企业的规模经济效应是显著的;但国有企业的规模经济检验结果却不稳健,不同方法的测度结果不一致。因此,对规模经济的测度应该尽量考虑到企业异质性,并使用多种检验方法,以保证政策干预的有效性。
Different from the traditional empirical research of cost function, this paper is concerned with the effects of heterogeneous total factor productivity of enterprises on cost function estimation. Therefore, this paper derive a complete cost function model including the total factor productivity from the CES production function for the first time. And regression is done using traditional model and the SFA model based on sample from 2003 to 2007 of Chinese large chemical enterprises so as to achieve methodological comparative analysis. The results show that the total factor productivity has important influence on the cost function estimation, and the traditional model may have the problem of omitting variables and bring endogenous problems to seemingly unrelated regression. The regression results of three methods show that the scale economy effect of non-state enterprises is remarkable, but the test results on scale economy of state-owned enterprises are not robust the results of the different methods are not consistent. Therefore, before regulations and support and subsidy policies are made, heterogeneity of enterprises should be taken into account when measuring economies of scale and a variety of test methods should be used to ensure the effectiveness of policy interventions.
出处
《统计研究》
CSSCI
北大核心
2015年第11期26-35,共10页
Statistical Research
基金
国家自然科学基金青年项目“行政垄断产业的政府管制体系研究”(71203078)和重点项目“推动经济发达地区产业转型升级的机制与政策研究”(71333007)的阶段性成果
教育部哲学社会科学重大课题攻关项目“进一步扩大现代产业开放的模式和路径研究”(14JZD021)
广东产业发展与粤港澳台区域合作研究中心及暨南大学跨越计划的资助
关键词
超越对数成本函数模型
似不相关回归
全要素生产率
规模经济
Translog Cost Function
Seemingly Unrelated Regression
Total Factor Productivity
Economies of Scale