Causal Dynamics Between Tourism Receipts, Economic Growth, and Real Exchange Rate in Algeria’s Economy Between 1990 and 2019

Variable	Characteristic	Definition
LNTOUR	The logarithm of international tourism receipts	International tourism receipts (currency US$)
LNGDPG	The logarithm of gross domestic product (GDP) growth	GDP growth (annual %)
LNREER	The logarithm of the real effective exchange rate	Real effective exchange rate index (2010 = 100)

1 Source: The variables are related to Algeria, and all data are from the World Development Indicators’ Data Bank by the World Bank https://databank.worldbank.org/reports.aspx?source=world-development-indicators, except for LNTOUR (1990–1994, 2018, 2019) and LNGDPG (2019), which we obtained from the Algerian National Bureau of Statistics (www.ons.dz).

Table 2:

Descriptive statistics of the variables in the study.

	LNGDPG	LNTOUR	LNREER
M	4.631941	4.731771	18.64382
Median	4.63574	4.637684	18.98015
Maximum	4.674714	5.396235	19.98303
Minimum	4.583947	4.535622	16.86003
SD	0.020151	0.188400	1.056439
Probability	0.693189	0.000001	0.225635
Sum	138.9582	141.9531	559.3146
Observations	30	30	30

1 Source: Output of EViews 10 software.

Table 3:

Correlation of the variables in the study.

	LNGDPG	LNREER	LNTOUR
LNGDPG	1
LNREER	−0.40829842	1
LNTOUR	0.43076231	−0.80311554	1

1 Source: Output of EViews 10 software.

Table 4:

Unit root test results.

		LNGDPG	LNREER	LNTOUR
At level	ADF test	–	–3.16**	–
At level	PP test	–	–4.49***	–
At first difference	ADF test	–8.63***	–	–5.73***
At first difference	PP test	–9.34***	–	–5.73***
Order of integration		I(1)	I(0)	I(1)

1 Note: ADF = augmented Dickey–Fuller test; PP = Phillips–Perron test. Source: Output of EViews 10 software.

2 **Rejection of the null hypothesis at 5%.

3 ***Rejection of the null hypothesis at 10%.

Table 5:

Computed F statistic for cointegration tests—ARDL bounds tests.

			] I(0)–I(1)]
Lower–upper bound (10%)		2.63–3.35	Cointegration
Lower–upper bound (5%)		3.10–3.87	Cointegration
Lower–upper bound (1%)		4.13–5.00	Cointegration
DV: LNTOUR	F statistics	4.44*
DV: LNGDPG	F statistics	2.25
DV: LNREER	F statistics	4.63
K		2

1 Note: DV = dependent variable; K = the number of regressors included in the models.

2 *Statistically significant at 1%.

Table 6:

Long-term coefficients estimation with ARDL bounds test model.

Model	Variable		Coefficient	p-value
ARDL	Dependent variable	LNTOUR
	Independent variable	LNREER	−3.42	0.1545
		LNGDPG	37.53	0.1193
		C	138.67	0.2423
ARDL	Dependent variable	LNGDPG
	Independent variable	LNREER	0.085	0.5482
		LNTOUR	0.0169	0.3703
		C	3.9	0.0008
ARDL	Dependent variable	LNREER
	Independent variable	LNTOUR	−0.05	0.4
		LNGDPG	−0.52	0.82
		C	8.03	0.45

1 Note: ARDL = autoregressive distributed lag.

Table 7:

Error correction representation of ARDL bounds test model.

Model	Variable		Coefficient	p-value
ARDL	Dependent variable	LNTOUR
		ΔLNREER	0.056	0.94
	Independent variable	ΔLNREER_t-1	1.07**	0.0305
		ECT_t-1	−0.263***	0.002
ARDL	Dependent variable	LNGDPG
		ΔLNTOUR	0.024***	0.0032
		ΔLNTOUR_t-1	−0.017**	0.0386
	Independent variable	ΔLNREER	−0.12***	0.0044
		ΔLNREER_t-1	−0.062***	0.0043
		ECT_t-1	−0.41***	0.0043
ARDL	Dependent variable	LNREER
		ΔLNREER_t-1	0.053	0.63
	Independent variable	ΔLNREER_t-2	−0.41***	0.000
		ECT_t-1	−0.25***	0.0002

1 Note: ARDL = autoregressive distributed lag.

2 **Statistically significant at 5%.

3 ***Statistically significant at 1%.

Table 8:

VECM Granger causality analysis.

IV/DV	Long-term			Short-term
IV/DV	ΔLGDPG	LNREER	ΔLNTOUR	ECT_t-1
ΔLGDPG	–	−0.25 (0.66)	5.9* (0.07)	−1.68*** (0.00)
LNREER	−0.12*** (00092)	–	0.35 (0.43)	−0.17*** (0.0002)
ΔLNTOUR	0.01 (0.14)	0.01 (0.71)	–	−0.98*** (0.00)
C	−0.166 (0.12)	0.79 (0.02)	−1.58 (0.45)	–
Observations	27	27	27	–
R²	0.71	0.92	0.16	–
DW	1.78	1.83	1.96	–
Fisher	6.81*** (00003)	48.70*** (0.00)	1.57(0.22)	–

1 Note: IV = independent variable; DV = dependent variable.

2 **Statistically significant at 5%.

3 ***Statistically significant at 1%.

Table 9:

Diagnostic tests.

Diagnostic test	LM test	ARCH test	RESET test	Normality test
F_TOUR(REER, GDPG)	0.9	0.25	0.67	0.057
F_GDPG(REER, TOUR)	0.75	0.066	0.79	0.45
F_REER(TOUR, GDPG)	0.38	0.55	0.9	0.44

1 Note: LM test = Breusch–Godfrey Lagrange multiplier test for residual serial correlation; ARCH test = autoregressive conditional heteroscedasticity test; RESET test = Ramsey’s test for functional misspecification; normality test = Jarque–Bera normality test.