Annexes 3: résultats du test
d'hétéroscédasticité et sa correction
1-test
d'hétéroscédasticité
Source
|
SS
|
df
|
MS
|
Number of obs =
|
225
|
F( 6, 218) =
|
6882.61
|
|
|
|
|
Model
|
86250.008
|
6
|
14375.0013
|
Prob > F =
|
0.0000
|
Residual
|
455.314171
|
218
|
2.08859711
|
R-squared =
|
0.9947
|
Adj R-squared =
|
0.9946
|
|
|
|
|
Total
|
86705.3222
|
224
|
387.077331
|
Root MSE =
|
1.4452
|
residus2
|
Coef.
|
Std. Err.
|
t
|
P>|t|
|
[95% Conf. Interval]
|
logpibreeli
|
20.66821
|
.1258484
|
164.23
|
0.000
|
20.42018
|
20.91625
|
logpibreelj
|
-3.027739
|
.130537
|
-23.19
|
0.000
|
-3.285015
|
-2.770462
|
logdistij
|
-5.908226
|
.1551594
|
-38.08
|
0.000
|
-6.214031
|
-5.602422
|
logcorrupi
|
-1.146092
|
.7431219
|
-1.54
|
0.124
|
-2.610715
|
.3185314
|
logcorrupj
|
-9.602522
|
.4680009
|
-20.52
|
0.000
|
-10.52491
|
-8.680136
|
instraij
|
-9.916242
|
.2632378
|
-37.67
|
0.000
|
-10.43506
|
-9.397425
|
_cons
|
-284.0963
|
3.56105
|
-79.78
|
0.000
|
-291.1148
|
-277.0778
|
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity Ho:
Constant variance
Variables: fitted values of residus2
chi2(1) = 3.71
Prob > chi2 = 0.0540
Page 97
Corruption, insécurité transfrontalière et
dynamique du commerce intra- CEMAC
2-la correction de
l'hétéroscédasticité
Linear regression Number of obs =
|
225
|
F( 6,
|
218)
|
=
|
515.79
|
Prob > F
|
|
=
|
0.0000
|
R-squared
|
|
=
|
0.8638
|
Root MSE
|
|
=
|
.55807
|
logcomij
|
Coef.
|
Robust
Std. Err.
|
t
|
P>|t|
|
[95% Conf. Interval]
|
logpibreeli
|
1.473571
|
.0509322
|
28.93
|
0.000
|
1.373189
|
1.573954
|
logpibreelj
|
-.251845
|
.04233
|
-5.95
|
0.000
|
-.3352735
|
-.1684166
|
logdistij
|
-.385963
|
.0671101
|
-5.75
|
0.000
|
-.5182306
|
-.2536954
|
logcorrupi
|
-.242403
|
.3670455
|
-0.66
|
0.510
|
-.9658151
|
.4810091
|
logcorrupj
|
-.6793408
|
.1975841
|
-3.44
|
0.001
|
-1.06876
|
-.2899213
|
instraij
|
-.7701079
|
.1091191
|
-7.06
|
0.000
|
-.9851713
|
-.5550444
|
_cons
|
-16.1904
|
1.748728
|
-9.26
|
0.000
|
-19.63698
|
-12.74382
|
Annexe4 : d'autocorrélation
RE GLS regression with AR(1) disturbances Number of obs =
Group variable: code Number of groups =
|
225
15
|
R-sq: within = 0.6987
|
|
Obs per group: min =
|
15
|
between = 0.5257
|
|
avg =
|
15.0
|
overall = 0.5629
|
|
max =
|
15
|
|
|
Wald chi2(7) =
|
138.11
|
corr(u_i, Xb) = 0 (assumed)
|
Prob > chi2 =
|
0.0000
|
logcomij
|
Coef.
|
Std. Err.
|
z P>|z| [95% Conf.
|
Interval]
|
logpibreeli
|
.4053621
|
.0536522
|
7.56 0.000 .3002056
|
.5105185
|
logpibreelj
|
.2484885
|
.049396
|
5.03 0.000 .1516741
|
.3453029
|
logdistij
|
-.8204785
|
.3412215
|
-2.40 0.016 -1.48926
|
-.1516966
|
logcorrupi
|
.040483
|
.1621586
|
0.25 0.803 -.277342
|
.358308
|
logcorrupj
|
.2632605
|
.1688906
|
1.56 0.119 -.0677589
|
.5942799
|
instraij
|
-.8408219
|
.5387677
|
-1.56 0.119 -1.896787
|
.2151435
|
_cons
|
-1.808704
|
2.683344
|
-0.67 0.500 -7.067962
|
3.450554
|
rho_ar
|
.87858731
|
(estimated
|
autocorrelation coefficient)
|
|
sigma_u
|
.80674674
|
|
|
|
sigma_e
|
.2329313
|
|
|
|
rho_fov
|
.92305035
|
(fraction of variance due to u_i)
|
|
theta
|
.5987641
|
|
|
|
modified Bhargava et al. Durbin-Watson = .52784527 Baltagi-Wu LBI
= .9575195
Page 98
Corruption, insécurité transfrontalière et
dynamique du commerce intra- CEMAC
|