The Estimation Of Elasticity Of Factor Substitution:Theoretical Review And Empirical Analysis

As an important parameter in economics research, the elasticity of substitution (a) which reflects the impact of the factor relative price change to the factor allocation plays a crucial role in economic growth and income distribution. Based on careful combing the theoretical and empirical research of the elasticity of substitution and Chinese provincial panel data 1978-2011, we used the CES production function with general factor-augmenting technical progress to estimate σ series by variable parameter panel data model. Through a series of robust tests, and compared with the relevant research, it showed that the results of this paper were very reliable. We found that:The elasticity of substitution (a) in China was obviously less than 1 (0.23-0.55), and during the reform period it showed a rising trend, so the C-D production function was not the reasonable approximation of macro-economy in China. Then, with Chinese provincial panel data, we gave the empirical test about the "Solow conjectures" for a, we found that a has obvious time variation trend, so the "Solow conjecture 1" was true in China. But the "Solow conjecture 2" that total factor elasticity of substitution is greater than industry elasticity of substitution was rejected. And, at the same level of industry, the elasticity of substitution also showed significant differences, the first industry had low elasticity (0.289-0.499), the second and the third industries had high elasticity (0.357-0.864; 0.394-0.713).Finally, taking the employment problem as an example, we studied the application value of the improvement factor substitution elasticity estimation for other fields of economic problems. Basing on the existing researches, we relaxed the assumption that the elasticity of substitution was 1, and we used the CES-CES economy growth model to study the structure changes of the employment, which was decomposed into the intensity effect, the output effect, the productivity effect, and the labor-augmenting technical progress effect. This paper showed that:whether it was on the "agriculture and non-agriculture" level or on the "manufacture-service" level, the productivity effect and the labor-augmenting technical progress effect were very significant, but the intensity effect was not significant; at the same time, there were significant regional differences between various effects.