The premise for establishment of appropriate grain production policies is to analyze the contributions of different influence factors to the variation of grain production. However, the existing studies focus on the overall comparison of influence factors affecting the grain production in China, and thus cannot reflect the annual differences or changes. Therefore, in this study, we built a complete (zero residual) decomposition model based on logarithmic mean Divisia index (LMDI) method for investigation into grain production in China. With this model, the changes of grain production were decomposed into the contributions from four factors, including yield per cultivated area, planting structure, multi-cropping index, and changes of cultivated land area. Based on the data from 1996 to 2012, we decomposed and analyzed the factors affecting the changes of grain production in China during this period. The results showed that the total grain production in China generally increased from 1996 to 2012, but it also fluctuated at a severe rate inter-annually. From the perspective of increments, the effect size of yield per cultivated area was 85.9225 million t, and yield per cultivated area was a major influence factor promoting the increments of grain production in China. The effect size of planting structure was -39.0911 million t, and planting structure was a major influence factor inhibiting the increments of grain production in China. However, the effect of planting structure on grain production became strongly positive from 2004 to 2006, and planting structure was a major influence factor promoting the increments of grain production during this period. From the perspective of fluctuation, the fluctuation of grain production in China from 1996 to 2003 was mainly attributed to the effect of yield per cultivated area and the effect of planting structure. From 2004 to 2007, the effects of yield per cultivated area, planting structure and multi-cropping index on grain production were contributive alternatively, but no effect was absolutely dominant. Since 2008, the effect of yield per cultivated area and the total effect changed in very similar ways. The results above show that the effect of yield per cultivated area is mainly responsible for the changes of grain production in China at the current stage, while the effect of planting structure can be enhanced greatly in the future. Moreover, the effect of multi-cropping index will not bring about great changes, but its sudden reduction may cause fluctuation. In comparison, the effect of cultivated land area can be ignored. Thus, regarding how to enhance China's grain production ability in the future, the main developing trends are to stably improve the yield per cultivated area, to raise the enthusiasm of peasants for grain production, to modify the planting structure, to prevent natural disasters, and to avoid the abrupt changes of multi-cropping index. These findings provide some valuble information for relevant governmental departments to establish grain production development plans and to formulate related industrial policies.%为了明确不同因素对粮食产量变化影响的大小，分析粮食产量波动变化的机理，该文采用对数均值迪氏指数法建立了粮食产量无残差因素分解模型，将中国粮食产量的波动分解为播面单产、种植结构、复种指数和耕地面积变化的贡献。基于1996－2012年数据，对该期间影响中国粮食产量波动的因素进行了分解，结果表明：播面单产和复种指数的变化表现为增长效应，种植结构和耕地面积的变化表现为减量效应，其中播面单产变化促进粮食产量增长8592.25万t，对粮食产量增长的贡献率为101.03%；复种指数变化促进粮食产量增长6926.97万t，对粮食产量增长的贡献为率81.45%；种植结构变化导致粮食产量减少3909.11万t，对粮食产量增长的贡献率为−45.97%；耕地面积变化导致粮食产量减少3105.61万t，对粮食产量增长的贡献率为−36.52%。各影响因子的作用具有阶段性，1996－2003年种植结构效应对粮食产量波动的影响较大；2004－2007年，播面单产、种植结构、复种指数的贡献相互交替，2004年播面单产的贡献最大，2005年、2007年复种指数的贡献最大，2006年则是种植结构的贡献最大；2008年开始播面单产效应起着主导作用。该研究可为政府有关部门粮食生产发展规划和相关产业政策的制定提供数据参考和理论依据。
展开▼
