摘要
首次将VAR模型应用到全国范围内的农业、林业、畜牧业和渔业为内生变量的农业动态结构系统,探讨我国农业结构的动态变化,Granger因果关系表明林业和畜牧业产值波动比较容易受到其他各业的影响,其他各业对农业产值的影响不显著,畜牧业与渔业存在双向的Granger因果关系;农业、林业、畜牧业和渔业分别对其他各业的新息冲击脉冲响应函数不同,随着时期数的增加对新息的冲击反应在逐渐减小,直至最后趋于原有的水平;畜牧业产值变化的预测方差主要由自身扰动引起,其次林业和农产产值扰动对其的影响各为约1/4,而渔业产值扰动对其的影响最小仅为12.34%。
A dynamic system VAR model was established including the variables of output value of farming, forestry, husbandry and fishery.The outcome of Granger causality test shows that output value forestry and husbandry are easily affected by other sectors of agriculture; output value of farming is hardly affected by other sectors; husbandry and fishery are mutual Granger reasons.Farming, forestry, husbandry and fishery have very different impulse responses when there are pluses from the four sectors of agriculture and the responses would get weaker as time went.Also conclusions were gained that output value of husbandry change had greater impact on itself and also output value of price of forestry, farming and fishery had a 25%, 25% and 12.34%impact on value of husbandry respectively through variance decomposition.
出处
《山西农业大学学报（社会科学版）》
2011年第11期1109-1115,共7页
Journal of Shanxi Agricultural University:Social Science Edition
关键词
VAR模型
农业结构
脉冲响应
方差分解
VAR Model
Agricultural structure
Impulse responses
Variance decomposition
