基于空间Panel data分位数回归的粮食产量分析  

Analysis of Grain Yield Based on Spatial Panel Data and Quantile Regression Model

作　　者：赵佩佩[1] 袁永生[1] 吴楠楠[1] 

机构地区：[1]河海大学理学院,江苏南京211100

出　　处：《江西农业学报》2016年第8期115-120,共6页Acta Agriculturae Jiangxi

基　　金：江苏省水利科技创新基金项目(2011059);河海大学自然科学基金资助项目(2009426311)

摘　　要：在对空间面板数据和分位数回归基本原理进行全面分析说明的基础上,选用全国31个省市2000~2012年的面板数据,对其进行了平稳性检验,考察了粮食生产的空间相关性,利用分位数回归方法对影响我国粮食产量的各影响因素进行了实证分析,根据估计结果定量分析了我国粮食产量的主要影响因素及其影响程度。研究结果表明：各省粮食生产存在空间相关性,而且粮食播种面积、农用化肥使用量和农业劳动力对粮食产量有重要影响,农用机械动力和受灾面积对粮食产量影响在不同分位点处表现不一,受灾面积则是粮食产量的抑制因素。Based on the spatial panel data and the fundamental principles of quantile regression analysis,the panel data were collected from 31 provinces（ cities） in China during 2000 ～ 2012,and the spatial correlation of grain crop production was examined by stationarity test. The influencing factors of China’s grain yield were analyzed by quantile regression,and the main influencing factors and their impacts on China’s grain yield were quantitatively analyzed according to the estimated results. The results of the experiments showed that the spatial correlation of grain crop production existed in different provinces,and the sown area of grain crops,chemical fertilizer application rate,and agricultural labor force had important influences on grain yield. However,the effects of agricultural machinery power and damage area on grain yield were various in different quantiles. Comprehensively,the damage area was considered to be the inhibiting factor of grain yield in China.

关 键 词：面板数据 空间相关性 分位数回归 粮食产量 

分 类 号：F326[经济管理—产业经济] O212[理学—概率论与数理统计]