搜索
中国猪肉需求计量经济模型分析08国际经济与贸易投资一班小组名单：200851102005罗元榜200851102040李永钊200851102006卓丹瑶摘要：文章运用了时间序列分析方法对中国猪肉需求情况进行分析，主要研究了人均猪肉消费量的影响因素，分析了相关因素，建立了人均猪肉消费量线性回归方程，并对所建立方程进行了平稳性检验、多重线性相关检验、自相关检验等，修复了回归方程。关键词：人均猪肉消费量计量经济学模型一、引言猪肉是我国重要的畜产品之一，也是我国城乡居民动物性蛋白的主要来源之一。自1985年以来，猪肉市场从计划流通体制向国家宏观调控下的自由流通体制过渡。始于2006年6月的这一轮价格的大幅波动，由于其幅度和速度都是前所未见的，更引起了社会各界的高度关注。而在市场经济条件下，猪肉价格有各自的供需均衡决定，本文目的在于研究猪肉需求的影响因素。二、计量模型1、模型的数学形式和变量的确定需求函数是以商品的需求量作为被解释变量，用影响需求量的因素，如收入、价格等作为解释变量的计量经济学模型。中国猪肉需求函数即选择收入和价格作为解释变量，同时考虑到，猪肉需求主要包括国内需求和国外需求，影响猪肉需求的因素主要是可替代品的产量。因此，笔者将上述对猪肉影响因素作为解释变量。收入选择的是城镇居民家庭每年人均可支配收入。价格选择的是猪肉生产价格指数即猪肉收购价格指数。模型中的被解释变量为国内人均猪肉消费量（Y）。根据其影响因素的大小和资料的可用性以及查阅的相关文献，本文选择以下指标作为模型的解释变量：城镇居民家庭人均可支配收入指数（x1）、猪肉收购价格指数（x2）、猪肉替代品牛羊肉人均产量（x3）、生猪出口量（x4）。参照单方程线性需求的表达式，国内猪肉需求函数模型的形式确定为：Y=β0+β1X1+β2X2+β3X3+β4X4+μ其中：Y代表国内人均猪肉消费量（千克）；X1代表城镇居民家庭人均可支配收入指数（1978=100）；2X2代表猪肉收购价格指数（1978=100）；X3代表猪肉替代品牛羊肉人均产量（千克）；X4代表生猪出口量（万头）；μ为随机误差项，描述变量外的因素对模型的干扰；β0为常数虚拟变量，包含政策等难以量化的因素的影响。2、样本数据收集整理通过中国国家统计局查询1996年~2009年统计年鉴得到了相关数据：年份猪肉消费量（万吨）牛肉（万吨）羊肉（万吨）人口（万人）出口（万头）城镇居民家庭人均可支配收入指数猪肉生产价格指数19851628.446.759.3105851.0315.0739.1121.119861776.758.962.2107507.0310.0899.6104.419871814.979.271.9109300.0302.01002.2118.619882000.695.880.2111026.0302.71181.4150.619892102.5107.296.2112704.0298.51375.7110.519902257.3125.6106.8114333.0299.91510.292.919912425.5153.5118.0115823.0285.01700.696.619922623.6180.3125.0117171.0291.32026.6106.319932839.4233.6137.3118517.0274.42577.4114.519943186.7327.0160.9119850.0270.43496.2154.619953638.2415.4201.5121121.0240.04283.0116.019963144.7355.7181.0122389.0228.14838.9102.219973577.6440.9212.8123626.0228.15160.3110.119983869.4479.9234.6124761.0220.45425.182.919993997.0505.4251.3125786.0196.15854.085.220003958.1513.1264.1126743.0203.96280.0100.220014037.0508.6271.8127627.0197.36859.6106.520024101.5521.9283.5128453.0188.97702.898.020034211.3542.5308.7129227.0188.78472.2102.920044301.0560.4332.9129988.0197.39421.6112.820054513.9568.1350.1130756.0176.910493.097.620064605.1576.7363.8131448.0172.311759.590.620074272.6613.4382.6132129.0160.913785.8145.920084635.7613.2380.3132802.0164.515780.8130.8整理得到所需数据：年份人均猪肉消费量（千克）城镇居民家庭人均可支配收入指数猪肉生产价格指数（收购价格指数）猪肉替代品人均产量出口（万头）198515.4739.1121.11.0315.0198616.5899.6104.41.1310.0198716.61002.2118.61.4302.0198818.01181.4150.61.6302.73198918.71375.7110.51.8298.5199019.71510.292.92.0299.9199120.91700.696.62.3285.0199222.42026.6106.32.6291.3199324.02577.4114.53.1274.4199426.63496.2154.64.1270.4199530.04283.0116.05.1240.0199625.74838.9102.24.4228.1199728.95160.3110.15.3228.1199831.05425.182.95.7220.4199931.85854.085.26.0196.1200031.26280.0100.26.1203.9200131.66859.6106.56.1197.3200231.97702.898.06.3188.9200332.68472.2102.96.6188.7200433.19421.6112.86.9197.3200534.510493.097.67.0176.9200635.011759.590.67.2172.3200732.313785.8145.97.5160.9200834.915780.8130.87.5164.53、模型的建立利用Eviews软件建立回归方程，结果如下：DependentVariable:YMethod:LeastSquaresDate:07/02/10Time:13:45Sample:19852008Includedobservations:24VariableCoefficientStd.Errort-StatisticProb.X1-9.72E-050.000117-0.8340070.4146X2-0.0221360.009965-2.2214430.0387X34.3848220.41907910.463010.0000X40.0597930.0216702.7592140.0125C-4.3046566.837869-0.6295320.5365R-squared0.989764Meandependentvar26.80417AdjustedR-squared0.987609S.D.dependentvar6.653634S.E.ofregression0.740642Akaikeinfocriterion2.420454Sumsquaredresid10.42247Schwarzcriterion2.665882Loglikelihood-24.04545Hannan-Quinncriter.2.485566F-statistic459.3042Durbin-Watsonstat2.137881Prob(F-statistic)0.0000004模型的参数估计对于理论模型运用OLS进行参数估计，再用Eviews软件进行运算得到的结果:SubstitutedCoefficients:=========================Y=-9.71811839598e-05*X1-0.0221364206268*X2+4.38482166958*X3+0.0597926879061*X4-4.30465630693（-0.834007）（-2.221443）（10.46301）（2.759214）（-0.629532）R2=0.989764DW=2.137881F=459.30424、统计检验异方差检验：用EViews软件进行White检验结果为：HeteroskedasticityTest:WhiteF-statistic2.022628Prob.F(14,9)0.1450Obs*R-squared18.21172Prob.Chi-Square(14)0.1973ScaledexplainedSS6.221619Prob.Chi-Square(14)0.9606TestEquation:DependentVariable:RESID^2Method:LeastSquaresDate:07/02/10Time:17:00Sample:19852008Includedobservations:24VariableCoefficientStd.Errort-StatisticProb.C-35.04612229.3940-0.1527770.8819X1-0.0035790.005181-0.6908760.5071X1^2-8.71E-085.51E-08-1.5805240.1484X1*X21.14E-059.44E-061.2103540.2570X1*X30.0005160.0003551.4560500.1794X1*X43.06E-061.62E-050.1883710.8548X2-0.0845860.412295-0.2051600.8420X2^2-0.0002890.000351-0.8242320.4311X2*X3-0.0067260.019157-0.3510790.7336X2*X40.0005320.0013550.3928600.7036X37.61879028.971930.2629710.7985X3^2-0.5492230.958339-0.5730990.5806X3*X4-0.0159940.090892-0.1759680.8642X40.2331851.4603370.1596790.8767X4^2-0.0004490.002328-0.1929710.8513R-squared0.758822Meandependentvar0.434270AdjustedR-squared0.383655S.D.dependentvar0.4631805S.E.ofregression0.363632Akaikeinfocriterion1.083820Sumsquaredresid1.190052Schwarzcriterion1.820104Loglikelihood1.994156Hannan-Quinncriter.1.279157F-statistic2.022628Durbin-Watsonstat2.161194Prob(F-statistic)0.144991查卡方分布表得χ2（23）=35.2TR=18.2135,2.所以该模型不存在异方差。自相关检验和序列相关检验由回归方程知：DW=2.137881查表得du=1.664-du=2.34dudw4-du所以该模型不存在自相