第七章练习题及参考解答
表中给出了1981-2015年中国城镇居民人均年消费支出(PCE)和城镇居民人均可支配收入(PDI)数据。
表 1981-2015年中国城镇居民消费支出(PCE)和可支配收入(PDI)数据 (单位:元)
年度 | 城镇居民人均消费支出PCE | 城镇居民人均可支配收入PDI | 年度 | 城镇居民人均消费支出PCE | 城镇居民人均可支配收入PDI |
1981 | 1999 | ||||
1982 | 2000 | ||||
1983 | 2001 | ||||
1984 | 2002 | ||||
1985 | 2003 | ||||
1986 | 2004 | ||||
1987 | 2005 | ||||
1988 | 2006 | ||||
1989 | 2007 | ||||
1990 | 2008 | ||||
1991 | 2009 | ||||
1992 | 2010 | ||||
1993 | 2011 | ||||
1994 | 2012 | ||||
1995 | 2013 | ||||
1996 | 2014 | ||||
1997 | 2015 | ||||
1998 | |||||
估计下列模型:
(1) 解释这两个回归模型的结果。
(2) 短期和长期边际消费倾向(MPC)是多少分析该地区消费同收入的关系。
(3) 建立适当的分布滞后模型,用库伊克变换转换为库伊克模型后进行估计,并对估计结果进行分析判断。
【练习题参考解答】
(1) 解释这两个回归模型的结果。
Dependent Variable: PCE | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 09:12 | ||||
Sample: 1981 2005 | ||||
Included observations: 25 | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
PDI | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | Schwarz criterion | |||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
收入跟消费间有显著关系。收入每增加1元,消费增加元。
Dependent Variable: PCE | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 09:13 | ||||
Sample(adjusted): 1982 2005 | ||||
Included observations: 24 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
PDI | ||||
PCE(-1) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | Schwarz criterion | |||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
(2) 短期和长期边际消费倾向(MPC)是多少分析该地区消费同收入的关系。
短期MPC=,长期MPC==
(3) 建立适当的分布滞后模型,用库伊克变换转换为库伊克模型后进行估计,并对估计结果进行分析判断。
在滞后1-5期内,根据AIC最小,选择滞后5期,其回归结果如下:
Dependent Variable: PCE | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 09:25 | ||||
Sample(adjusted): 1986 2005 | ||||
Included observations: 20 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
PDI | ||||
PDI(-1) | ||||
PDI(-2) | ||||
PDI(-3) | ||||
PDI(-4) | ||||
PDI(-5) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | Schwarz criterion | |||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
当期收入对消费有显著影响,但各滞后期影响并不显著。不显著可能是分布滞后模型直接估计时共线性造成的,也可能是真没显著影响。库伊克模型估计结果见上表,PCE(-1)部分回归结果t检验不显著。
表中给出了中国1980-2016年固定资产投资Y与社会消费品零售总额X的资料。取阿尔蒙多项式的次数m=2,运用阿尔蒙多项式变换法估计以下分布滞后模型:
表中国1980-2016年固定资产投资Y与社会零售总额X数据 (单位:亿元)
年份 | 固定资产投资 Y | 社会消费品零售总额X | 年份 | 固定资产投资 Y | 社会消费品零售总额X |
1980 | 1999 |
| |||
1981 | 2000 | ||||
1982 | 2001 | ||||
1983 | 2002 | ||||
1984 | 2003 | ||||
1985 | 2004 | ||||
1986 | 2005 | ||||
1987 | 2006 | ||||
1988 | 2007 | ||||
1989 | 2008 | ||||
1990 | 2009 | ||||
1991 | 2010 | ||||
1992 | 2011 | ||||
1993 | 2012 | ||||
1994 | 2013 | ||||
1995 | 2014 | ||||
1996 | 2015 | ||||
1997 | 2016 | ||||
1998 | |||||
【练习题参考解答】
直接估计结果如下:
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 09:32 | ||||
Sample(adjusted): 1984 2016 | ||||
Included observations: 33 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
X | ||||
X(-1) | ||||
X(-2) | ||||
X(-3) | ||||
X(-4) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | +09 | Schwarz criterion | ||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
使用阿尔蒙变换估计结果如下: Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 09:37 | ||||
Sample(adjusted): 1984 2016 | ||||
Included observations: 33 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
Z0 | ||||
Z1 | ||||
Z2 | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | +09 | Schwarz criterion | ||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
根据
直接使用软件结果:
Dependent Variable: Y | |||||
Method: Least Squares | |||||
Date: 03/10/18 Time: 09:39 | |||||
Sample(adjusted): 1984 2016 | |||||
Included observations: 33 after adjusting endpoints | |||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. | |
C | |||||
PDL01 | |||||
PDL02 | |||||
PDL03 | |||||
R-squared | Mean dependent var | ||||
Adjusted R-squared | . dependent var | ||||
. of regression | Akaike info criterion | ||||
Sum squared resid | +09 | Schwarz criterion | |||
Log likelihood | F-statistic | ||||
Durbin-Watson stat | Prob(F-statistic) | ||||
Lag Distribution of X | i | Coefficient | Std. Error | T-Statistic | |
. * | | 0 |
|
|
| |
. *| | 1 |
|
|
| |
. * | | 2 |
|
|
| |
.* | | 3 |
|
|
| |
* . | | 4 |
| |||
Sum of Lags |
|
|
| ||
利用表的数据,运用局部调整假定或自适应预期假定估计以下模型参数,并解释模型的经济意义,探测模型扰动项的一阶自相关性:
1)设定模型
其中
2)设定模型
其中
3)设定模型
其中
【练习题参考解答】
1)设定模型
其中
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 10:09 | ||||
Sample(adjusted): 1981 2016 | ||||
Included observations: 36 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
X | ||||
Y(-1) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | +09 | Schwarz criterion | ||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
根据回归结果,可算出h统计量为,明显大于2,表明5%显著水平下存在相关性。根据回归数据,可算出调整系数为
2)设定模型
其中
假设调整方程为:
Dependent Variable: LOG(Y) | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 10:11 | ||||
Sample(adjusted): 1981 2016 | ||||
Included observations: 36 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
LOG(X) | ||||
LOG(Y(-1)) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | Schwarz criterion | |||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
根据回归结果,计算h统计量时开方部分为负,没法计算。故没法根据h统计量判断相关性。根据回归数据,可算出调整系数为
3)设定模型
其中
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 10:09 | ||||
Sample(adjusted): 1981 2016 | ||||
Included observations: 36 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
X | ||||
Y(-1) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | +09 | Schwarz criterion | ||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
可算出调节系数为
表给出中国各年末货币流通量Y,社会商品零售额X1、城乡居民储蓄余额X 2的数据。
表中国年末货币流通量、社会商品零售额、城乡居民储蓄余额数据 (单位:亿元)
年份 | 年末货币流通量Y | 社会消费品零售总额X1 | 城乡居民储蓄年底余额X2 |
1989 | |||
1990 | |||
1991 | |||
1992 | |||
1993 | |||
1994 | |||
1995 | |||
1996 | |||
1997 | |||
1998 | |||
1999 | |||
2000 | |||
2001 | |||
2002 | |||
2003 | |||
2004 | |||
2005 | |||
2006 | |||
2007 | |||
2008 | |||
2009 | |||
2010 | |||
2011 | |||
2012 | |||
2013 | |||
2014 | |||
利用表中数据设定模型:
其中,
【练习题参考解答】
利用表中数据设定模型:
其中,
假设局部调整方程为:
Dependent Variable: Y | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 10:03 | ||||
Sample(adjusted): 1990 2014 | ||||
Included observations: 25 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
Y(-1) | ||||
X1 | ||||
X2 | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | Schwarz criterion | |||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
各回归系数在5%显著水平下均显著。可算出调整系数为
假设局部调整方程为:
Dependent Variable: LOG(Y) | ||||
Method: Least Squares | ||||
Date: 03/10/18 Time: 10:04 | ||||
Sample(adjusted): 1990 2014 | ||||
Included observations: 25 after adjusting endpoints | ||||
Variable | Coefficient | Std. Error | t-Statistic | Prob. |
C | ||||
LOG(Y(-1)) | ||||
LOG(X1) | ||||
LOG(X2) | ||||
R-squared | Mean dependent var | |||
Adjusted R-squared | . dependent var | |||
. of regression | Akaike info criterion | |||
Sum squared resid | Schwarz criterion | |||
Log likelihood | F-statistic | |||
Durbin-Watson stat | Prob(F-statistic) | |||
根据四川省1978—2014年的消费总额Y(亿元)和收入总额X(亿元)的年度资料,估计出库伊克模型如下:
试回答下列问题:
1)分布滞后系数的衰减率是多少
2)模型中是否存在多重共线性问题请说明判断的理由。
3)收入对消费的即期和长期影响乘数是多少
4)某同学查表发现,在显著性水平
【练习题参考解答】
1)分布滞后系数的衰减率为
2)模型中各斜率系数均显著,没有明显的多重共线性问题。
3)收入对消费的即期和长期影响乘数分别是:
即期乘数为; 长期乘数为=
4)该同学试图检验是否存在自相关性问题,但是此模型为自回归模型,模型中有滞后被解释变量
式中:d=;n=37;
h=,小于
利用某地区1980—2014年固定资产投资(Y)与地区生产总值GDP(X)的数据资料(单位:亿元),使用OLS法估计出如下模型:
(1)上述模型是否存在自相关性问题
(2)如果将上述模型看成是局部调整模型的估计结果,试计算调节系数
【练习题参考解答】
(1) 式中:d=;n=35;
h=,小于
(2) 如果将模型看成是局部调整模型的估计结果,
联系自己所学的专业选择一个实际问题,设定一个分布滞后模型或自回归模型,并自己去收集样本数据,用本章的方法估计和检验这个模型,你如何评价自己所做的这项研究
【练习题参考解答】
本题无参考解答
¥29.8
¥9.9
¥59.8