Table 4. Quasi-Liquid Liabilities and Growth in Francophone and Anglophone SSA Countries
(Dependent Variable: Growth)
|
| Methodology |
|
Random
Effects
| Arellano – Bond |
|
One Step | Two Step |
|
Log Real GDP Per Capita
|
-.028
(.000)
|
-.138
(.001)
|
-.236
(.000)
|
|
Log Quasi-liquid
Liabilities (QLLY)
|
-.001
(.925)
|
.005
(.701)
|
.007
(.133)
|
|
Log Investment as Share of GDP
|
.007
(.170)
|
.017
(.028)
|
.017
(.000)
|
|
Log Primary School Enrollment
|
.020
(.021)
|
.044
(.012)
|
.055
(.000)
|
|
Log Government Share of GDP
|
-.016
(.013)
|
-.030
(.005)
|
-.044
(.000)
|
|
Terms of Trade Shock as % of GDP
|
.025
(.114)
|
.039
(.086)
|
.021
(.322)
|
|
Interaction: Franco-phone x log (QLLY)
|
-.002
(.594)
|
-.008
(.504)
|
-.013
(.070)
|
|
Constant
|
.103
(.126)
|
-.004
(.069)
|
-.005
(.018)
|
|
Observations |
140
|
116
|
116
|
|
R – Square
|
.150
|
|
|
|
Prob (Chi–Sq)
|
.0001
|
|
|
|
SarganTest(P value)1
|
|
.513
|
.864
|
|
Serial Correlation
Test (P value)2, 3
|
|
.002
.964
|
.010
.525
|
|
P – values of estimated regression coefficients in parentheses.
1 Test of the null hypothesis that the instruments are not correlated with the residuals.
2Test of the null hypothesis that the errors exhibit no first-order serial correlation.
3Test of the null hypothesis that the errors exhibit no second-order serial correlation.
|
|
Coefficient Estimates in Francophone and Anglophone SSA Countries1
(Dependent Variable: Growth) |
|
| | |
