5. Results of estimates
By using the cross-section method on 96 countries6 and
by making the average for each variable over the period 1996-2003, we obtain
the results presented in the tables below.
6 The choice of the 96 countries was made according to
the availability of the data.
For the first table, there is a regression in function with
the variables of bases only (Model 1 :M1). For the remainder of the
regressions, we integrate each time one of the six indicators of governance
from the data base of KAUFMANN7.
Table 1. Dependent variable. In y
(Governance Indicators. KA UFMANN)
|
M1
|
M2
|
M3
|
M4
|
M5
|
M6
|
M7
|
|
Cste
|
-4.280
(-3.400)
|
2.790
(2.023)
|
1.648
(1.237)
|
3.473
(2.788)
|
2.617
(2.253)
|
3.498
(2.902)
|
2.905
(2.491)
|
|
Ln(n+g+ä)
|
-2.047
(-4.088)
|
0.065
(0.134)
|
-0.142
(-0.289)
|
-0.400
(-0.986)
|
-0.373
(-0.93 1)
|
-0.305
(-0.76)
|
-0.340
(-0.855)
|
|
Lnk
|
0.324
(8.165)
|
0.260
(7.961)
|
0.291
(8.912)
|
0.171
(5.124)
|
0.208
(6.717)
|
0.184
(5.793)
|
0.205
(6.659)
|
|
KAUFMANN
|
|
VA
|
|
0.689
(7.444)
|
|
|
|
|
|
|
PS
|
|
|
0.633
(6.957)
|
|
|
|
|
|
GE
|
|
|
|
0.711
(9.171)
|
|
|
|
|
RQ
|
|
|
|
|
0.893
(9.429)
|
|
|
|
RL
|
|
|
|
|
|
0.725
(9.617)
|
|
|
CC
|
|
|
|
|
|
|
0.648
(9.603)
|
|
N. Obs
|
96
|
96
|
96
|
96
|
96
|
96
|
96
|
|
R2
|
0.597
|
0.748
|
0.736
|
0.789
|
0.795
|
0.799
|
0.799
|
7 For each indicator of gouvernance one makes the
average of the data available: 1/5 * (Value 96+ Value 98+ Value 2000+ Value
2002+ Value 2003).
For the second table and for each regression, we integrate each
time one of the five institutional variables8 selected by the
ICRG9.
Table 2. Dependent variable . In y
(Institutional Quality. ICR G)
|
M8
|
M9
|
M 10
|
M 11
|
M 12
|
|
Cste
Ln(n+g+ä)
Lnk
GS COR LO BQ DA
|
-5.338
(-3.557)
|
-1.521
(-1.334)
|
-2.110
(-1.660)
|
1.726
(1.476)
|
-1.873
(-1.42)
|
|
-2.085
(-4.172)
|
-0.850
(-1.850)
|
-1.285
(-2.591)
|
-0.844
(-2.116)
|
-1.052
(-2.00)
|
|
0.313
(7.743)
|
0.287
(8.492)
|
0.269
(6.943)
|
0.135
(3.674)
|
0.292
(7.761)
|
|
ICRG
|
|
|
|
|
0.130
(1.285)
|
|
|
|
|
|
-0.42 1
(-6.366)
|
|
|
|
|
|
0.276
(4.183)
|
|
|
|
|
|
0.656
(8.641)
|
|
|
|
|
|
0.243
(4.031)
|
|
N. Obs
R2
|
96
|
96
|
96
|
96
|
96
|
|
0.604
|
0.720
|
0.661
|
0.777
|
0.658
|
The results above are significant since they come to confirm
the existence of a close connection between the quality of governance and
economic growth. It is noted that when one adds the governance variables of
Kaufmann, the R2 passes from 0,597 to 0,748 % with the introduction
of the Voice and Accountability indicator (VA), to 0,736% with the indicator of
Political Stability (PS), to 0,789% with the Government Effectiveness indicator
(GE), to 0,795% with the Regulatory Quality indicator (RQ), to 0,799% with the
Rule of Law indicator (RL) and finally to 0,799 % with the Corruption Control
indicator (CC). The corresponding estimated coefficients are positive and
statistically significant. The most significant are respectively the Regulatory
Quality (RQ), the Rule of Law (RL), the Government Effectiveness (GE), the
Voice and Accountability (VA), the Control of Corruption (CC) and finally
Political Stability (PS).
8 The choice of these variables has been made to
remain in the same context that the variables of KAUFMANN.
9 For each institutional variable, one makes the
average over the period 1996-2003.
These results come to corroborate those obtained by Kaufmann
and al. (1999) and Knack and Keefer (1995) who show that the indicators of
governance constitute relevant elements of the long run economic growth.
For the ICRG indicators, we also note that when the
institutional variables are added, the R2 passes from 0,597 to
0,604% with the introduction of the government stability indicator (GS), to
0,720 % with the corruption indicator (COR), to 0,66 1% with the laws and
orders indicator (LO), to 0,777% with the bureaucratic quality indicator (BQ)
and finally to 0,65 8 with the democratic accountability indicator (DA). All
these indicators present expected significant signs except the indicator of
government stability which is not significant.
One can note that these empirical results for all the 96
countries come to confirm those obtained by several authors like Barro (1991)
and Londregan and Poole (1992) for political stability, Mauro (1995) for
corruption and Kaufmann, Kraay and Mastruzzi (2004) for the indicator of rules
and laws.
That makes it possible to conclude that the determinants of
good governance, the capacity of the State to manage the resources effectively
and to formulate and implement policies and quality regulations, all of them
explain for the most part the long run economic performances of nations.
| 
