您好,欢迎访问三七文档
当前位置:首页 > 商业/管理/HR > 企业财务 > 斯托克计量经济学第十章第十一章实证练习stata
E10.1(1)(2)(3)(4)(5)lnviolnviolnviolnviolnvioshall-0.443***-0.368***-0.0461*-0.288***-0.0280(0.0475)(0.0348)(0.0189)(0.0337)(0.0278)incarc_rate0.00161***-0.00007100.00193***0.0000760(0.000181)(0.0000936)(0.000114)(0.0000720)density0.0267-0.172*-0.00887-0.0916(0.0143)(0.0850)(0.0139)(0.0485)avginc0.00121-0.009200.01290.000959(0.00728)(0.00591)(0.00796)(0.00729)pop0.0427***0.01150.0408***-0.00475(0.00315)(0.00872)(0.00252)(0.00781)pb10640.0809***0.104***0.1000***0.0292(0.0200)(0.0178)(0.0182)(0.0183)pw10640.0312**0.0409***0.0401***0.00925(0.00973)(0.00507)(0.00912)(0.00538)pm10290.00887-0.0503***-0.0444*0.0733***(0.0121)(0.00640)(0.0175)(0.0129)_cons6.135***2.982***3.866***2.948***4.348***(0.0193)(0.609)(0.385)(0.569)(0.435)N11731173117311731173R20.0870.5640.2180.5800.955adj.R2StateEffectsTimeEffects0.0859NoNo0.5613NoNo0.1771YesNo0.5690NoYes0.9525YesYesStandarderrorsinparentheses*p0.10,**p0.05,***p0.01(1)①回归(2)中shall的系数是-0.368,这意味着隐蔽武器法律,也即“准予”携带法律,约使暴力犯罪减少36.8%。从“现实意义”上讲,这个估计值很大。②回归(1)中shall的系数是-0.443,回归(2)中shall的系数是-0.368,加入(2)中的控制变量后“准予”携带法律的效应略微减小。系数估计显著性都很高,两者均在1%水平下显著。③不同州的人们对待枪支和暴力犯罪的态度。(2)加入州固定效应的回归结果如上表第(3)列所示。回归(3)中shall的系数是-0.0461,和回归(2)相比,“准予”携带法律的效应减小许多,显然说明回归(2)中存在遗漏变量偏差。个体固定效应的差异来源于随个体变化但随时间不变的遗漏变量,回归(3)中,变量shall,density,pb1064,pw1064,pm1029至少在5%水平下显著,可信度较高,加入州固定效应得到的回归结果总体上较回归(2)可信。(3)加入时间固定效应的回归结果如上表第(5)列所示。回归(5)中shall的系数是-0.028,和回归(2)(3)相比,“准予”携带法律的效应减小许多。回归(5)同时包含个体和时间固定效应以控制州间不同但时间上相同的变量和随时间变化但州间相同的变量,adjustedR2的值为0.9525,比前面的回归更接近1,因此回归(5)的结果更加可靠。(4)(1)(2)(3)(4)(5)lnroblnroblnroblnroblnrobshall-0.773***-0.529***-0.00782-0.341***0.0268(0.0693)(0.0510)(0.0253)(0.0457)(0.0243)incarc_rate0.00101***-0.00007630.00177***0.0000314(0.000187)(0.000125)(0.000155)(0.000112)density0.0905***-0.186-0.00918-0.0447(0.0154)(0.114)(0.0189)(0.0737)avginc0.0407***-0.0175*0.0643***0.0144(0.00927)(0.00791)(0.0108)(0.0101)pop0.0778***0.01630.0720***0.0000164(0.00549)(0.0117)(0.00343)(0.0118)pb10640.102***0.112***0.167***0.0141(0.0266)(0.0238)(0.0248)(0.0270)pw10640.0275*0.0272***0.0557***-0.0128(0.0135)(0.00679)(0.0124)(0.00720)pm10290.02730.0112-0.189***0.105***(0.0150)(0.00857)(0.0238)(0.0222)_cons4.873***0.9042.446***1.792*3.587***(0.0279)(0.889)(0.515)(0.772)(0.645)N11731173117311731173R20.1210.5960.0370.6530.961adj.R20.12010.5934-0.01350.64340.9593Standarderrorsinparentheses*p0.05,**p0.01,***p0.001用lnrob代替lnvio后所得回归如上表所示,分析如下:①回归(1)中shall的系数是-0.773,回归(2)中shall的系数是-0.529,这意味着隐蔽武器法律,也即“准予”携带法律,约使暴力犯罪分别减少77.3%和52.9%。从“现实意义”上讲,这个估计值很大。系数估计显著性都很高,两者均在1%水平下显著。②加入州固定效应的回归结果如上表第(3)列所示。回归(3)中shall的系数是-0.00782,和回归(1)(2)相比,“准予”携带法律的效应减小许多,显然说明回归(1)(2)中存在遗漏变量偏差。加入时间固定效应的回归结果如上表第(5)列所示。③回归(5)中shall的系数是0.0268。和回归(1)(2)相比,明显回归(3)(5)“准予”携带法律的效应减小许多。回归(5)adjustedR2的值为0.9593,比前面的回归更接近1,因此回归(5)的结果更加可靠。(1)(2)(3)(4)(5)lnmurlnmurlnmurlnmurlnmurshall-0.473***-0.313***-0.0608*-0.198***-0.0150(0.0485)(0.0357)(0.0258)(0.0340)(0.0297)incarc_rate0.00210***-0.000360**0.00260***-0.000116(0.000154)(0.000128)(0.000115)(0.000148)density0.0397***-0.671***-0.0134-0.544***(0.0118)(0.116)(0.0141)(0.117)avginc-0.0773***0.0243**-0.0698***0.0566***(0.00875)(0.00807)(0.00803)(0.0136)pop0.0416***-0.0257*0.0393***-0.0321**(0.00351)(0.0119)(0.00255)(0.00902)pb10640.131***0.03070.188***0.0220(0.0188)(0.0242)(0.0184)(0.0434)pw10640.0471***0.01030.0739***-0.000489(0.00909)(0.00693)(0.00921)(0.0119)pm10290.0655***0.0392***-0.0502**0.0692*(0.0137)(0.00874)(0.0177)(0.0289)_cons1.898***-2.486***0.460-2.831***0.657(0.0220)(0.615)(0.525)(0.574)(0.726)N11731173117311731173R20.0830.6060.1530.6420.921adj.R20.08260.60320.10870.63280.9167Standarderrorsinparentheses*p0.05,**p0.01,***p0.001用lnmur代替lnvio后所得回归如上表所示,分析如下:①回归(1)中shall的系数是-0.473,回归(2)中shall的系数是-0.313,这意味着隐蔽武器法律,也即“准予”携带法律,约使暴力犯罪分别减少47.3%和31.3%。从“现实意义”上讲,这个估计值很大。系数估计显著性都很高,两者均在1%水平下显著。②加入州固定效应的回归结果如上表第(3)列所示。回归(3)中shall的系数是-0.0608,和回归(1)(2)相比,“准予”携带法律的效应减小许多,显然说明回归(1)(2)中存在遗漏变量偏差。加入时间固定效应的回归结果如上表第(5)列所示。③回归(5)中shall的系数是-0.015。和回归(1)(2)相比,明显回归(3)(5)“准予”携带法律的效应减小许多。回归(5)adjustedR2的值为0.9167,比前面的回归更接近1,因此回归(5)的结果更加可靠。(5)“准予”携带法律会对犯罪率产生影响,同时,一个州可能因为犯罪情况而决定是否颁布“准予”携带法律,这就产生双向因果关系偏差,威胁内部有效性。(6)综上可知,上述三个表格的回归(5)结果最为可信。用lnrob,lnmur代替lnvio所得回归与lnvio所得的回归结果相类似,以lnvio为例进行分析。Shall效应的95%置信区间为-8.6%至3.0%,其中包括shall的系数取0,根据显著性分析,没有证据能表明“准予”携带武器法律对犯罪率有明显影响。E10.2(1)(2)(3)(4)(5)fatalityratefatalityratefatalityratesb_useagesb_useagesb_useage0.00407**-0.00577***-0.00372*(0.00123)(0.00116)(0.00145)speed650.000148-0.000425-0.0007830.0751***0.0228(0.000408)(0.000334)(0.000580)(0.0115)(0.0205)speed700.00240***0.00123***0.0008040.009710.0120(0.000472)(0.000329)(0.000457)(0.0133)(0.0206)ba08-0.00192***-0.00138***-0.0008220.0463***0.00376(0.000361)(0.000373)(0.000443)(0.0104)(0.0176)drinkage210.00007990.000745-0.00113-0.003780.0107(0.000987)(0.000507)(0.000622)(0.0183)(0.0272)lnincome-0.0181***-0.0135***0.006260.337***0.0583(0.00109)(0.00142)(0.00670)(0.0218)(0.256)age-0.000007220.000979*0.00132-0.00497*0.0138(0.000164)(0.000382)(0.000694)(0.00242)(0.0231)primary0.300***0.206***(0.0125)(0.0232)secondary0.148***0.109***(0.0106)
本文标题:斯托克计量经济学第十章第十一章实证练习stata
链接地址:https://www.777doc.com/doc-6297342 .html