Manajemen | Fakultas Ekonomi Universitas Maritim Raja Ali Haji 073500105000000045
Journal of Business & Economic Statistics
ISSN: 0735-0015 (Print) 1537-2707 (Online) Journal homepage: http://www.tandfonline.com/loi/ubes20
Correction
To cite this article: (2005) Correction, Journal of Business & Economic Statistics, 23:3, 363-363,
DOI: 10.1198/073500105000000045
To link to this article: http://dx.doi.org/10.1198/073500105000000045
Published online: 01 Jan 2012.
Submit your article to this journal
Article views: 28
View related articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ubes20
Download by: [Universitas Maritim Raja Ali Haji]
Date: 12 January 2016, At: 23:53
Correction
Bun, M. J. G., and Carree, M. A. (2005), “Bias-Corrected
Estimation in Dynamic Panel Data Models,” Journal of Business & Economic Statistics, 23, 200–210.
what was reported earlier. In general, bias is neglible or small,
whereas only for γ and T small is there a moderate size distortion. Based on an MSE criterion, this estimator can compete with bc and gmm estimators. Regarding the third design
(Table 4), especially for γ , bias in gmm carries over to bias
in ac, demonstrating the dependence of additive bias correction on preliminary consistent estimators. Previously this remark was erroneously made in the discussion of Table 2. Also,
bias in estimating γ increases with increasing γ , contrary to
what was reported earlier. Finally, the correct empirical results
of the ac estimator for the unemployment growth model (Table 5) are γ̂ac = .624 and β̂ac = −.056 (with standard errors
.038 and .012), which are close to the bc and gmm estimates.
Downloaded by [Universitas Maritim Raja Ali Haji] at 23:53 12 January 2016
Due to misinterpretation the simulation and empirical results
for the additive bias corrected estimator (labeled ac) do not do
full justice to this method. In this erratum we correct these results. In the original study we used three simulation designs,
of which the simulation results have been reported in Tables 2,
3, and 4 respectively. Here we repeat the corrected tables (for
the ac estimator only).
The corrected results show that the ac estimator performs
well in the first two designs (Tables 2 and 3), contrary to
Corrected Results, ac Estimator for Table 2
(N, T)
(300, 2)
(200, 3)
(150, 4)
(100, 6)
(60, 10)
(40, 15)
bias γ
RMSE γ
bias β
RMSE β
% actual size γ (nominal is 5%)
% actual size β (nominal is 5%)
.005
.068
.002
.081
14.9
5.9
−.000
.043
.002
.060
12.5
5.4
−.000
.033
.001
.052
11.3
5.2
−.001
.023
−.001
.044
8.6
5.2
−.002
.016
.001
.038
6.8
5.5
−.002
.013
.001
.034
6.8
5.2
Corrected Results, ac Estimator for Table 3
(N, T)
(300, 2)
(200, 3)
(150, 4)
(100, 6)
(60, 10)
(40, 15)
bias γ
RMSE γ
bias β
RMSE β
% actual size γ (nominal is 5%)
% actual size β (nominal is 5%)
.024
.073
.007
.082
16.1
5.8
.014
.046
.004
.061
13.2
5.4
.011
.035
.001
.052
12.4
5.2
.007
.024
−.002
.044
9.5
5.1
.003
.016
−.002
.038
7.1
5.3
.002
.013
−.002
.034
6.5
5.1
Corrected Results, ac Estimator for Table 4
Design nr.
bias γ
RMSE γ
bias β
RMSE β
% actual size γ (nominal is 5%)
% actual size β (nominal is 5%)
I
II
III
IV
V
VI
VII
VIII
IX
−.003
.037
.000
.053
6.7
5.2
−.011
.046
.002
.050
10.6
5.3
−.123
.136
.005
.108
76.0
6.2
−.009
.049
.000
.211
8.1
5.1
−.030
.063
.007
.222
16.9
5.5
−.135
.148
.013
.622
81.5
6.5
−.017
.099
.001
.095
17.9
5.4
−.002
.063
.000
.046
10.6
5.3
−.025
.105
.001
.095
21.4
5.4
X
−.002
.064
.000
.046
10.8
5.3
© 2005 American Statistical Association
Journal of Business & Economic Statistics
July 2005, Vol. 23, No. 3, Correction
DOI 10.1198/073500105000000045
363
ISSN: 0735-0015 (Print) 1537-2707 (Online) Journal homepage: http://www.tandfonline.com/loi/ubes20
Correction
To cite this article: (2005) Correction, Journal of Business & Economic Statistics, 23:3, 363-363,
DOI: 10.1198/073500105000000045
To link to this article: http://dx.doi.org/10.1198/073500105000000045
Published online: 01 Jan 2012.
Submit your article to this journal
Article views: 28
View related articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ubes20
Download by: [Universitas Maritim Raja Ali Haji]
Date: 12 January 2016, At: 23:53
Correction
Bun, M. J. G., and Carree, M. A. (2005), “Bias-Corrected
Estimation in Dynamic Panel Data Models,” Journal of Business & Economic Statistics, 23, 200–210.
what was reported earlier. In general, bias is neglible or small,
whereas only for γ and T small is there a moderate size distortion. Based on an MSE criterion, this estimator can compete with bc and gmm estimators. Regarding the third design
(Table 4), especially for γ , bias in gmm carries over to bias
in ac, demonstrating the dependence of additive bias correction on preliminary consistent estimators. Previously this remark was erroneously made in the discussion of Table 2. Also,
bias in estimating γ increases with increasing γ , contrary to
what was reported earlier. Finally, the correct empirical results
of the ac estimator for the unemployment growth model (Table 5) are γ̂ac = .624 and β̂ac = −.056 (with standard errors
.038 and .012), which are close to the bc and gmm estimates.
Downloaded by [Universitas Maritim Raja Ali Haji] at 23:53 12 January 2016
Due to misinterpretation the simulation and empirical results
for the additive bias corrected estimator (labeled ac) do not do
full justice to this method. In this erratum we correct these results. In the original study we used three simulation designs,
of which the simulation results have been reported in Tables 2,
3, and 4 respectively. Here we repeat the corrected tables (for
the ac estimator only).
The corrected results show that the ac estimator performs
well in the first two designs (Tables 2 and 3), contrary to
Corrected Results, ac Estimator for Table 2
(N, T)
(300, 2)
(200, 3)
(150, 4)
(100, 6)
(60, 10)
(40, 15)
bias γ
RMSE γ
bias β
RMSE β
% actual size γ (nominal is 5%)
% actual size β (nominal is 5%)
.005
.068
.002
.081
14.9
5.9
−.000
.043
.002
.060
12.5
5.4
−.000
.033
.001
.052
11.3
5.2
−.001
.023
−.001
.044
8.6
5.2
−.002
.016
.001
.038
6.8
5.5
−.002
.013
.001
.034
6.8
5.2
Corrected Results, ac Estimator for Table 3
(N, T)
(300, 2)
(200, 3)
(150, 4)
(100, 6)
(60, 10)
(40, 15)
bias γ
RMSE γ
bias β
RMSE β
% actual size γ (nominal is 5%)
% actual size β (nominal is 5%)
.024
.073
.007
.082
16.1
5.8
.014
.046
.004
.061
13.2
5.4
.011
.035
.001
.052
12.4
5.2
.007
.024
−.002
.044
9.5
5.1
.003
.016
−.002
.038
7.1
5.3
.002
.013
−.002
.034
6.5
5.1
Corrected Results, ac Estimator for Table 4
Design nr.
bias γ
RMSE γ
bias β
RMSE β
% actual size γ (nominal is 5%)
% actual size β (nominal is 5%)
I
II
III
IV
V
VI
VII
VIII
IX
−.003
.037
.000
.053
6.7
5.2
−.011
.046
.002
.050
10.6
5.3
−.123
.136
.005
.108
76.0
6.2
−.009
.049
.000
.211
8.1
5.1
−.030
.063
.007
.222
16.9
5.5
−.135
.148
.013
.622
81.5
6.5
−.017
.099
.001
.095
17.9
5.4
−.002
.063
.000
.046
10.6
5.3
−.025
.105
.001
.095
21.4
5.4
X
−.002
.064
.000
.046
10.8
5.3
© 2005 American Statistical Association
Journal of Business & Economic Statistics
July 2005, Vol. 23, No. 3, Correction
DOI 10.1198/073500105000000045
363