An Empirical Analysis of the Effects of Population Growth on Economic Growth in Ethiopia Using an Autoregressive Distributive Lag (ARDL) Model Approach

Table 3

ARDL model selection criteria.

Dependent variable: LNRGDP
Method: ARDL
Date: 06/12/20 time: 02:20
Sample (adjusted): 1983 and 2018
Included observations: 36 after adjustments
Maximum dependent lags: 3 (automatic selection)
Model selection method: Akaike information criterion (AIC)
Dynamic regressors (3 lags, automatic): LNPOP, FDI, PREM, LNPOPGR, RI, LNGCF
Fixed regressors: C
Number of models evaluated: 12288
Selected model: ARDL (2, 0, 1, 1, 0, 3, 3)
White–Hinkley (HC1) heteroscedasticity consistent standard errors and covariance
Variable	Coefficient	Std. error	t-Statistic	Prob.
LNRGDP (−1)	0.281058	0.131681	2.134392	0.0461
LNRGDP (−2)	−0.437488	0.167134	−2.617585	0.0169
LNPOP	0.451623	0.097259	4.643495	0.0002
FDI	−2.62E − 11	3.21E − 11	−0.817119	0.4240
FDI (−1)	1.77E − 10	4.58E − 11	3.864522	0.0010
PREM	−0.000226	6.21E − 05	−3.648837	0.0017
PREM (−1)	0.000419	4.26E − 05	9.844719	0.0000
LNPOPGR	−0.158448	0.201627	−0.785846	0.4416
RI	0.002391	0.000805	2.970558	0.0079
RI (−1)	0.002891	0.001129	2.560639	0.0191
RI (−2)	0.004362	0.001468	2.971370	0.0078
RI (−3)	0.003631	0.001127	3.221178	0.0045
LNGCF	0.159154	0.041861	3.801961	0.0012
LNGCF (−1)	0.033373	0.056122	0.594656	0.5591
LNGCF (−2)	0.166142	0.073842	2.249969	0.0365
LNGCF (−3)	0.214824	0.050341	4.267364	0.0004
C	−0.115940	1.072006	−0.108152	0.9150

Source: authors’ own computation. Note. The selected model is ARDL (2, 0, 1, 1, 0, 3, 3).