基于埃塔平方法（η2）的中国小麦生产驱动因素贡献份额研究

USING ETA SQUARE METHOD (η2) TO ESTIMATE THE CONTRIBUTION SHARE OF DRIVING FACTORS ON WHEAT PRODUCTION IN CHINA

关键词:  埃塔平方法（η2） 粮食安全 小麦 贡献份额 中国

Key words:  eta square  food security  wheat production  contribution share  China

摘要:

［目的］定量分析中国小麦生产驱动因素贡献份额，综合评价各因素对小麦生产的影响效应。［方法］采用埃塔平方法（η2）来确定气候与气象、科技与生产投入和社会经济3组因素对小麦生产影响的贡献份额。选取1978年以来小麦生产集中度波动上升的河南、河北、山东、江苏、安徽和新疆（春小麦）6省区作为研究案例。分别建立冬小麦生产和小麦生产两个模型群。每个模型群包含1个全因素模型和6个局部因素模型。［结果］两个模型群的R2×100的值分别为63.70和62.74，显示模型整体具有较强的解释能力，气候与气象、科技与生产投入和社会经济这3组因素中的X变量分别在两个模型群中解释了小麦生产变化的63.70%和62.74%。η2×100是相应因素或因素组合的解释力，由此得到的贡献份额显示了各因素对于小麦生产的独立和交互贡献份额。其中，冬小麦模型群的运算结果显示，科技与生产投入和社会经济因素的独立贡献份额分别为5.83%和4.30%，而交互贡献份额则高达40.57%； 气候与气象因素的独立贡献份额为2.43%，与科技和生产投入因素的交互贡献份额为0.14%，与社会经济因素交互贡献份额为-1.15%，说明气象气候与社会经济因素中共享了一部分信息； 整个模型所有因素交互贡献份额为11.58%。小麦模型群的运算结果支持冬小麦模型群，在3组因素中，科技与生产投入和社会经济因素的独立贡献及其交互贡献份额分别为5.05%、3.22%和44.86%； 气候与气象因素的独立贡献份额为2.48%，与科技与生产投入和社会经济因素的交互贡献份额均为负，分别是-0.41%和-1.24%； 3组因素交互贡献份额为8.78%。［结论］两个模型群的运算结果共同显示科技与生产投入和社会经济因素是影响中国小麦生产的主要驱动因素，其中科技与生产投入对于小麦生产的作用更为突出； 这两组因素的交互作用对于小麦生产有控制性影响，并在小麦生产模型中更加明显； 相比之下，气候与气象因素贡献份额相对较小，且在和科技与生产投入、社会经济因素的交互作用过程中，产生了一定的削弱作用，使得气候与气象因素的作用更加不明显。

Abstract:

The objective of this study was to estimate the contribution share of driving factors on wheat production in China and evaluate the integrative effects of the factors. An improved approach, eta square method, was used in the study to identify the contribution shares of three group factors that were climate and meteorological factors, technology and production input, social economic factors by which wheat production was affected. According to the concentration index of wheat production in China since 1978, wheat productions of the six main wheat producers in China were Henan province, Hebei province, Shandong province, Jiangsu province, Anhui province and Xinjiang Uygur Autonomous Region. Two groups of ordinary least square (OLS) models that included five winter wheat producers′ models and six winter-spring wheat producers′ models were constructed for further analysis. Each model group consisted of seven models, one full model and six partial models. The results of the two model groups showed that the values of adjusted R square multiply by 100 were 63.70 and 62.74,respectively, which indicated that the models had strong ability of explanation, and wheat production could be explained 63.70% and 62.74% respectively by these X variables such as climate and meteorological factors, Technology and production input, social economic factors. The winter wheat models showed that technology and production input, social economic factors individually and interactively contributed 5.83%, 4.30% 40.57% explanatory power to the variation of wheat production in five studied provinces. Climate and meteorological factors only contributed 2.43% individually and another 0.14% interactively together with technology and production input factors. When interacting together with social economic factors, the contribution share was -1.15% as they shared some same information. The three group factors interactively contributed the remaining 11.58% explanatory power. Both the winter and spring wheat models support the results of winter models. Among the three groups of factor, technology and production input, social economic factors individually and interactively contributed 5.05%, 3.22% 44.86% explanatory power. Climate and meteorological factors only contributed 2.48%, and interactively contributed -0.41% and -1.24% as sharing same information. The three group factors interactively contributed the remaining 8.78% explanatory power. The results of two groups of model showed that technology and production input, social economic factors were major factors affecting China′s wheat production, in which the role of technology and production input was more outstanding; In contrast, the contribution share of climate and meteorological factors was relatively small.