中国省域林业要素产出弹性空间异质性研究——基于地理加权回归模型的测度

Study on the Spatial Heterogeneity of the Output Elasticity in China's Provincial Forestry Elements: Based on the Evaluation of Geographical Weighted Regression Model

基于2015年中国31个省域林业投入产出数据,利用空间计量经济模型分析各省域之间的空间聚集效益,进一步建立地理加权回归模型对各个省域的要素产出弹性进行测算并比较。研究结果表明:各省域资本要素弹性系数估计结果介于0.378~1.647之间,劳动力要素弹性系数估计结果介于0.188~0.783之间,林业产出存在较为显著的空间相关性,要素产出弹性存在空间异质性,中国林业发展总体上处于规模报酬递增阶段。基于研究结论与各省域林业发展实际状况,提出加强不同地区之间的资源流动、因地制宜地制定林业生产发展规划和战略等政策建议。

(1)Background——Through the application of the relevant theories and methods of spatial econometrics,the paper establishes geographical weighted regression model. The model can help to the empirical study on the spatial heterogeneity of the output elasticity of China's provincial forestry production elements,and make clear the influence of the provincial spatial correlation on the output elasticity so as to make up for the research blank of the theoretical research in this field.(2)Methods——Based on the forestry input-output data in China's 31 provinces in 2015,Cobb-Douglas production function is applied to analyze the influence of forestry investment and forestry labor input on forestry output. It first calculates the global Moran's I index and the local Moran's I index of the whole China's provincial forestry output to examine the spatial correlation of China's forestry output,based on which the geographical weighted regression model is used to calculate the output elasticity of each province and then to compare them.(3) Results——The forestry output of China's 31 provinces has significant spatial correlation,and shows the positive spatial agglomeration effect which is much obvious in China's eastern coastal region. By using geographical weighted regression model,the variable parameter provincial forestry production function is found that local estimation results show that there are some differences in the elasticity of forest output in different provinces.The local R^2 of the provincial forestry production function model is between 0. 582 and 0. 926,and the average local R^2 is 0. 710,which shows that the input of two elements of capital and labor in the variable parameter GWR model has better explanatory ability to the variable forest output. The elastic coefficient of capital factors in31 provinces in China was estimated to be between 0. 378 and 1. 647 in 2015,and the elastic coefficient of the average capital factor was 0. 775. The elasticity coefficient of labor factor is between 0. 188 and 0. 783,and the elasticity coefficient of the average labor factor is 0. 363. They all show that the contribution of forestry investment to forestry output is greater than the contribution of labor factor to forestry output. In 2015,the total output elasticity of 23 provinces was more than 1,which indicated that the forestry production in most provinces was in the increasing stage of scale compensation.(4) Conclusions and Discussions——There is a significant spatial correlation of forestry output in China.In terms of the elasticity of factor output,the common characteristic of four regions is that the average labor output elasticity is less than the elasticity of capital output. With the advancement of industrialization and urbanization,the 3 elements of the productive forces in agriculture and forestry are constantly flowing to other industries,which makes China's forestry generally in a relatively low capital-intensive state. Therefore,Chinese forestry also presents the characteristic that labor share is below capital share in the production process. From the local view,both the average labor output elasticity and the capital output elasticity of the western region are significantly higher than other regions,because the western region is sparsely populated,which belongs to the typical area of capital investment,so the forestry production has not yet reached the level of the lowest economic scale; in contrast,the eastern region is more advanced in production technology,and the output increased by further capital investment will be limited in a certain degree so that the output elasticity of capital is relatively smaller than other areas. From the point of view of the elasticity of factor output,the forestry production process in most provinces is in the stage of increasing the returns of scale,which shows that these provinces have relatively low factor input in the process of forestry production,and can enlarge the investment scale to achieve the increase of scale returns.