Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Dependent Variable: Y
Method: Least Squares
Sample: 1947Q1 1972Q4
Included observations: 104
Variable
Coefficient
C
70.67592
X
0.860012
R-squared
0.998711
Adjusted R-squared
0.998698
S.E. of regression
18.04716
Sum squared resid
33221.38
Log likelihood
-447.4306
F-statistic
79005.80
Prob(F-statistic)
0.000000
i.
Std. Error
t-Statistic
6.019729
11.74071
0.003060
281.0797
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
Prob.
0.0000
0.0000
1687.933
500.1494
8.642897
8.693750
8.663499
0.867650
47.06273
33.05710
Prob. F(1,101)
Prob. Chi-Square(1)
0.0000
0.0000
Since prob(Obs*R2) = 0.0000 < = 0.05, Reject H0, there is evidence of first-order
autocorrelation.
ii. Obtain the second round estimate of
Round 1
Dependent Variable: UHAT
Method: Least Squares
Sample (adjusted): 1947Q2 1972Q4
Included observations: 103 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
UHAT(-1)
0.566704
0.082246
6.890374
R-squared
0.317619 Mean dependent var
Adjusted R-squared
0.317619 S.D. dependent var
S.E. of regression
14.90456 Akaike info criterion
Sum squared resid
22658.89 Schwarz criterion
Log likelihood
-423.9200 Hannan-Quinn criter.
Durbin-Watson stat
1.943467
Prob.
0.0000
-0.038358
18.04287
8.250873
8.276453
8.261234
Dependent Variable: Y1
Method: Least Squares
Sample (adjusted): 1947Q2 1972Q4
Included observations: 103 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
C
32.43388
5.016498
6.465442
X1
0.857654
0.005780
148.3811
R-squared
0.995434 Mean dependent var
Adjusted R-squared
0.995388 S.D. dependent var
S.E. of regression
14.96512 Akaike info criterion
Sum squared resid
22619.44 Schwarz criterion
Log likelihood
-423.8302 Hannan-Quinn criter.
F-statistic
22016.95 Durbin-Watson stat
Prob(F-statistic)
0.000000
Prob.
0.0000
0.0000
743.9043
220.3700
8.268548
8.319708
8.289269
1.941469
Prob.
0.0000
0.235964
18.09842
8.249108
8.274688
8.259468
= 0.570629
(b)
Use Engle-Granger test for cointegration to determine whether the Y and X series are
cointegrated.
t-Statistic
0.547536
-4.049586
-3.454032
-3.152652
t-Statistic
0.547536
Prob.*
0.9993
Prob.
0.5852
C
@TREND(1947Q1)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
-2.760483
0.100935
0.187213
0.170957
15.28975
23377.63
-425.5282
11.51670
0.000032
13.72443
-0.201136
0.261847
0.385476
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
0.8410
0.7007
17.13107
16.79236
8.320936
8.397675
8.352018
1.989219
The computed ADF test-statistic (0.5475) is greater than the critical "tau", thus we do not reject the Ho.
That means the "Y" is non-stationary at level.
Null Hypothesis: D(Y) has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 0 (Fixed)
t-Statistic
Augmented Dickey-Fuller test statistic
-9.833661
Test critical values:
1% level
-4.050509
5% level
-3.454471
10% level
-3.152909
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(Y,2)
Method: Least Squares
Sample (adjusted): 1947Q3 1972Q4
Included observations: 102 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
D(Y(-1))
-1.011126
0.102823
-9.833661
C
4.139859
3.147858
1.315135
@TREND(1947Q1)
0.251048
0.056406
4.450762
R-squared
0.494817 Mean dependent var
Adjusted R-squared
0.484612 S.D. dependent var
S.E. of regression
15.33975 Akaike info criterion
Sum squared resid
23295.48 Schwarz criterion
Log likelihood
-421.7149 Hannan-Quinn criter.
F-statistic
48.48435 Durbin-Watson stat
Prob(F-statistic)
0.000000
Prob.*
0.0000
Prob.
0.0000
0.1915
0.0000
0.455882
21.36736
8.327742
8.404947
8.359005
1.943602
The computed ADF test-statistic (-9.83366) is less than the critical "tau", thus we reject the Ho. That
means the "Y" becomes stationary at first-differenced.
t-Statistic
0.749614
-4.051450
-3.454919
-3.153171
Prob.*
0.9997
Prob.
0.4553
0.1645
0.4169
0.6530
0.9608
20.51683
20.65390
8.764719
8.894180
8.817128
1.761294
The computed ADF test-statistic (0.7496) is greater than the critical "tau", thus we do not reject the Ho.
That means the "X" is non-stationary at level.
Null Hypothesis: D(X) has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 1 (Fixed)
t-Statistic
Augmented Dickey-Fuller test statistic
-5.859654
Test critical values:
1% level
-4.051450
5% level
-3.454919
10% level
-3.153171
*MacKinnon (1996) one-sided p-values.
Augmented Dickey-Fuller Test Equation
Dependent Variable: D(X,2)
Method: Least Squares
Included observations: 101 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
D(X(-1))
-0.910160
0.155327
-5.859654
D(X(-1),2)
0.083873
0.112453
0.745846
C
4.459000
4.059313
1.098462
@TREND(1947Q1)
0.268397
0.073085
3.672403
R-squared
0.367076 Mean dependent var
Adjusted R-squared
0.347501 S.D. dependent var
S.E. of regression
18.85992 Akaike info criterion
Sum squared resid
34502.55 Schwarz criterion
F-statistic
18.75229 Durbin-Watson stat
Prob(F-statistic)
0.000000
Prob.*
0.0000
Prob.
0.0000
0.4576
0.2747
0.0004
0.904950
23.34800
8.750753
8.854322
1.751929
The computed ADF test-statistic (-5.85965) is less than the critical "tau", thus we reject the Ho. That
means the "X" becomes stationary at first-differenced.
Prob.*
0.0000
Prob.
0.0000
-0.227816
16.80896
8.250873
8.276453
8.261234
The computed ADF test-statistic (-5.268299) is less than the critical "tau", thus we reject the Ho. That
means the "UHAT" is stationary at level.
If both sets of data are I(1) (non-stationary), then if the regression produces an I(0) error
term, the equation is said to be cointegrated.
QUESTION 2
The following table contains an annual time series data on the demand for cigarettes in Turkey
and its determinants for the years 1946 - 1994. The definitions of the variables are as follows;
Q
Y
P
There has been considerable debate about the impact that government restrictions have had on
the demand for tobacco. In Australia all tobacco advertising in the electronic media was banned
after 1976. Further bans on advertising in the print media and on billboard advertising were
initiated in 1987.
(a) Based on all the given variables and information regarding tobacco advertising bans,
estimate a model that will enable you to test the effectiveness of the tobacco advertising
bans in the electronic media after 1976, and in the print media and on billboard in 1987.
Dependent Variable: Q
Method: Least Squares
Sample: 1946 1994
Included observations: 49
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
Y
P
D1
D2
D1Y
D2Y
D1P
D2P
-0.182390
-1.932014
2.385953
12.16804
-22.25672
-0.783607
1.151386
-1.959297
3.813171
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.693288
0.631946
0.505873
10.23631
-31.16395
11.30195
0.000000
3.086080
0.872523
0.438545
7.555939
8.281885
1.534964
1.645440
1.091236
1.359541
-0.059101
-2.214283
5.440609
1.610394
-2.687397
-0.510505
0.699744
-1.795485
2.804748
0.9532
0.0326
0.0000
0.1152
0.0104
0.6125
0.4881
0.0801
0.0077
7.598282
0.833846
1.639345
1.986822
1.771177
1.722676
(b) Test the null hypothesis that there is no difference between marginal effect of Q with
respect to Y and marginal effect of Q with respect to P.
Q
1 5 D 1 6 D2
Y
Q
2 7 D 1 8 D2
P
For the period 1946 1976; D1 = D2 = 0; we need to test H0: 1 = 2
Wald Test:
Equation: Untitled
Test Statistic
t-statistic
F-statistic
Chi-square
Value
df
Probability
-4.739684
22.46460
22.46460
40
(1, 40)
1
0.0000
0.0000
0.0000
Value
Std. Err.
-4.317966
0.911024
Wald Test:
Equation: Untitled
Test Statistic
Value
df
t-statistic
-2.907776
40
F-statistic
8.455163
(1, 40)
Chi-square
8.455163
1
Null Hypothesis: C(2)+C(6)=C(3)+C(8)
Null Hypothesis Summary:
Normalized Restriction (= 0)
C(2) - C(3) + C(6) - C(8)
Probability
0.0059
0.0059
0.0036
Value
Std. Err.
-3.142276
1.080646
Value
df
t-statistic
-3.700887
40
F-statistic
13.69657
(1, 40)
Chi-square
13.69657
1
Null Hypothesis: C(2)+C(6)+C(7)=C(3)+C(8)+C(9)
Null Hypothesis Summary:
Normalized Restriction (= 0)
C(2) - C(3) + C(6) + C(7) - C(8) C(9)
Probability
0.0006
0.0006
0.0002
Value
Std. Err.
-5.804061
1.568289
(c) Test the impact of these government restrictions i.e. whether they are successful or
otherwise in reducing cigarette consumption?
We need to test H0: 3 = 4 = 5 = 6 = 7 = 8 = 0
H1: at least one of the 0
Wald Test:
Equation: Untitled
Test Statistic
Value
df
F-statistic
4.254457
(6, 40)
Chi-square
25.52674
6
Null Hypothesis: C(4)=C(5)=C(6)=C(7)=C(8)=C(9)=0
Null Hypothesis Summary:
Normalized Restriction (= 0)
C(4)
C(5)
C(6)
C(7)
C(8)
C(9)
Probability
0.0021
0.0003
Value
Std. Err.
12.16804
-22.25672
-0.783607
1.151386
-1.959297
3.813171
7.555939
8.281885
1.534964
1.645440
1.091236
1.359541
END