Corruption, insécurité transfrontalière et dynamique du commerce intra-Cemac.

Annexes 2 : stationnarité des variables d'études

1-Exportations (Comij )

Panel unit root test: Summary

Series

: LO G ( S E R0 1

)

Date: 11 /24/ 1 5 Time: 05 : 2 5

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit ro ot (assumes common unit root proces s )
Levin, Lin & Chu t^* - 3. 1 7883 0. 00 07	15	195
Null: Unit root (assumes individual unit root process)
Im , Pesaran and Shin W-stat 1. 02 26 3 0. 84 68	15	195
AD F - Fisher Chi-square 15 . 56 1 3 0. 9862	15	195
PP - Fisher Chi-square 27 . 3776 0. 60 34	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.

2-logarithme du Produit Intérieur Brut du pays réel i (logPibreel i)

Panel unit root test: Summary

Series

: LO G ( S E R02

)

Date: 11 /24/ 1 5 Time: 05 : 2 8

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit root (assumes common unit root process)
Levin, Lin & Chu t^* - 0. 7 1 06 0 0. 2387	15	195
Null: Unit ro ot (assumes individual unit root proces s )
Im , Pesaran and Shin W-stat 2. 9582 9 0. 9985	15	195
AD F - Fisher Chi-square 14 . 744 1 0. 99 1 1	15	195
PP - Fisher Chi-square 16 . 856 1 0. 9743	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.

Page 94

Corruption, insécurité transfrontalière et dynamique du commerce intra- CEMAC

3-le logarithme du Produit Intérieur Brut réel du pays j (log pibreel j)

Panel unit root test: Summary

Series

: LO G ( S E R03

)

Date: 11 /24/ 1 5 Time: 05 : 3 1

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit ro ot (assumes common unit root proces s )
Levin, Lin & Chu t^* - 3. 30 692 0. 00 05	15	195
Null: Unit root (assumes individual unit root process)
Im , Pesaran and Shin W-stat 0. 57888 0. 7 1 87	15	195
AD F - Fisher Chi-square 22 . 0475 0. 8523	15	195
PP - Fisher Chi-square 30 . 0 56 0 0. 46 28	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.

4-le logarithme de la distance entre le pays i et j(logdist _ij )

Panel unit root test: Summary

Series

: LO G ( S E R04

)

Date: 11 /24/ 1 5 Time: 17 : 0 3

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit root (assumes common unit root process)
Levin, Lin & Chu t^* - 3. 30 692 0. 00 05	15	195
Null: Unit ro ot (assumes individual unit root proces s )
Im , Pesaran and Shin W-stat 0. 57888 0. 7 1 87	15	195
AD F - Fisher Chi-square 22 . 0475 0. 8523	15	195
PP - Fisher Chi-square 30 . 0 56 0 0. 46 28	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.

Page 95

Corruption, insécurité transfrontalière et dynamique du commerce intra- CEMAC

5-le logarithme de la corruption dans le pays i(logcorrup i)

Panel unit root test: Summary

Series

: LO G ( S E R05 )

Date: 11 /24/ 1 5 Time: 05 : 34

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit ro ot (assumes common unit root proces s )
Levin, Lin & Chu t^* -4. 00 830 0. 00 00	15	195
Null: Unit root (assumes individual unit root process)
Im , Pesaran and Shin W-stat -4. 20 858 0. 00 00	15	195
AD F - Fisher Chi-square 68 . 4 04 2 0. 00 0 1	15	195
PP - Fisher Chi-square 55 . 4498 0. 00 32	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.

6-le logarithme de la corruption du pays j (logcorrupj)

Panel unit root test: Summary

Series

: LO G ( S E R06

)

Date: 11 /24/ 1 5 Time: 05 : 36

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit root (assumes common unit root process)
Levin, Lin & Chu t^* - 5. 1 3494 0. 00 00	15	195
Null: Unit ro ot (assumes individual unit root proces s )
Im , Pesaran and Shin W-stat - 5. 1 4 075 0. 00 00	15	195
AD F - Fisher Chi-square 79 . 97 1 9 0. 00 00	15	195
PP - Fisher Chi-square 54 . 374 1 0. 0042	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.

Page 96

Corruption, insécurité transfrontalière et dynamique du commerce intra- CEMAC

7-l'insécurité transfrontalière

Panel unit root test : Summary Series: SE R0 7

Date: 11 /2 5/ 1 5 Time: 05 : 34

Sa	m pl e : 1998 20 1 2

Exogenous variables: Ind ivi d u al effects User-specifiedlags: 1

Newey-West automatic bandwidth selection and Bartlett kernel Balanced observations for each test

Method Statistic Prob . **	Cross- sections	Obs
Null: Unit root (assumes common unit root process)
Levin, Lin & Chu t^* -4. 08 1 36 0. 00 00	15	195
Null: Unit root (assumes individual unit root process)
Im , Pesaran and Shin W-stat - 2. 3585 1 0. 00 92	15	195
AD F - Fisher Chi-square 46 . 2 1 59 0. 02 96	15	195
PP - Fisher Chi-square 34 . 9679 0. 24 38	15	21 0

** Probabilities for Fisher tests are computed using an asymptotic Chi -square distribution. All other tests assume asymptotic normality.