Asymmetric modeling of the fiscal policy–economic growth nexus in Somalia

Figures & data

Table 1. Description of variables and sources.

Variable	Code	Description	Source
Gross domestic product	GDP	USD (current prices)	OIC statistics
General government final consumption expenditure	GGFCE	USD (current prices)	OIC statistics
Domestic investment	DI	Gross fixed capital formation, constant 2015 prices (USD)	OIC statistics
Household consumption	HC	USD (current prices)	OIC statistics
Export	EX	USD (millions)	OIC statistics

Table 2. Descriptive statistics and correlation matrix.

	LRGDP	LGGFCE	LHC	LEX	LDI
Mean	20.7556	18.5007	20.4916	2.4293	19.3264
Median	20.8253	18.4984	20.5123	1.9400	19.2948
Maximum	21.6788	19.2374	21.3580	4.7962	19.8532
Minimum	19.6475	17.3672	19.4331	1.1522	18.9965
Std. Dev.	0.5370	0.4148	0.4895	1.1465	0.1976
Skewness	−0.2164	−0.6599	−0.2703	0.6175	0.4535
Kurtosis	2.1401	3.6257	2.4103	1.8962	2.6933
Jarque-Bera	1.9309	4.4451	1.3335	5.7151	1.9095
Probability	0.3808	0.1083	0.5134	0.0574	0.3849
Sum	1037.780	925.0333	1024.580	121.4662	966.3192
Sum Sq. Dev.	14.1300	8.43202	11.7394	64.4034	1.9125
Observations	50	50	50	50	50
LRGDP	1
LGGFCE	0.8243	1
LHC	0.9864	0.81605	1
LEX	−0.6607	−0.2364	−0.6101	1
LDI	0.1275	−0.0165	0.1543	−0.1313	1

Table 3. Results from unit root test.

level
	Augmented Dickey Fuller (ADF)		Phillips Perron (PP)
Variable	Constant Without Trend	Constant With Trend	Constant Without Trend	Constant With Trend
LRGDP	−2.1401	−3.6047**	−2.0466	−2.7297
LGGFCE	−3.563**	−3.7969**	−3.563**	−3.8616**
LHC	−2.2596	−3.814**	−2.1825	−2.6025
LDI	−2.1498	−2.1817	−2.0957	−2.137
LEX	−1.3395	−2.0312	−1.3395	−2.2229
First difference
LRGDP	−6.1337***	−5.721***	−7.2461***	−8.8169***
LGGFCE	−6.8209***	−6.8314***	−6.9787***	−6.9754***
LHC	−5.5784***	−5.574***	−5.9751***	−6.6687***
LDI	−8.3979***	−8.3108***	−8.39***	−8.3027***
LEX	−6.2556***	−6.2077***	−6.2339***	−6.1756***

Note: ***, **, and * indicate significance level of 1%, 5% and 10% respectively.

Table 4. Results from BDS test for nonlinearity.

Dimension	LRGDP	LGGFCE	LHC	LDI	LEX
2	0.1516***	0.1004***	0.1453***	0.1119***	0.1349***
3	0.2487***	0.1540***	0.2393***	0.1884***	0.2316***
4	0.3112***	0.1702***	0.2956***	0.2326***	0.2987***
5	0.3529***	0.1640***	0.3276***	0.2536***	0.3411***
6	0.3782***	0.1517***	0.3435***	0.2677***	0.3639***

Note: ***, **, and * indicates significance level of 1%, 5% and 10% respectively. In addition, all variables are based on natural logarithm.

Table 5. Results from Bound test for NARDL.

Model	F-statistic	Significance	Bound test critical value
			K = 5
			I(0)	I(1)
$\mathrm{\text{LGDP}}=\mathrm{f}(\mathrm{\text{LGGFCE}},\mathrm{\text{LHC}},\mathrm{\text{LDI}},\mathrm{\text{LFDI}},\mathrm{\text{LEX}})$	9.287387	1%	3.06	4.15
		5%	2.39	3.38
		10%	2.08	3.00

The critical values are based on Narayan (2004), while K presents the number of parameters in the model.

Table 6. Short-run and long-run results from NARDL.

Dependent variable is RGDP: NARDL (1,3,3,3,0,0) selected automatically using EViews
Coefficients for long run elasticity
Regressor	Coefficient	T-statistic
Constant	0.0785	1.8324*
LGGFCE_POS	0.5026	4.3147***
LGGFCE_NEG	0.5130	4.3951***
LHC	0.4370	3.0256***
LDI	−0.0836	−1.1700
LEX	0.0454	1.2868
Short term coefficients and error correction model
ΔLGGFCE_POS	0.2442	5.8116***
ΔLGGFCE_NEG	0.1920	5.6086***
ΔLHC	0.8202	21.7127***
ECT(-1)	−0.9999	−8.8579***

Note: ***, **, and * indicates significance level of 1%, 5% and 10% respectively.

Table 7. Results from various diagnostic checks.

Adjusted R-square	0.9879
LM Test	0.1438 (0.8667)
Reset Test	1.3868 (0.2489)
Normality Test	0.5850 (0.746398)
Heteroscedasticity Test	1.0887 (0.3027)
CUSUM Test	Stable
CUSUM of squares	Stable

Figure 1. CUSUM Test.

Figure 2. CUSUM of Squares.

Figure 3. Dynamic multiplier.

Narayan, P. (2004). Reformulating critical values for the bounds F-statistics approach to cointegration: an application to the tourism demand model for Fiji. Monash University.

 Google Scholar