DSpace at VNU: Research note: Empirical assessment of the tourism-led growth hypothesis - the case of Vietnam

Keywords: TLGH; economic growth; cointegration; ECM; Granger causality; Vietnam In recent years, tourism business development in Vietnam has attracted research interest.. We empirically

Research note: Empirical assessment of the

tourism-led growth hypothesis – the case of

Vietnam

N GUYEN H O M INH T RANG

College of Economics, Hue University, 100 Phung Hung street, Hue City, Vietnam, and University of Economics and Law, Vietnam National University, Ho Chi Minh City.

E-mail: nguyenhominhtrang@gmail.com (Corresponding author.)

N GUYEN H UU C HAU D UC

Hue University of Medicine and Pharmacy, Hue City, Vietnam, and Department of International Health, Tokyo Medical and Dental University, Tokyo, Japan E-mail:

duc.ith@tmd.ac.jp.

N GUYEN T IEN D UNG

University of Economics and Law, Vietnam National University, Ho Chi Minh City,

Vietnam E-mail: ntdung@uel.edu.vn.

This study examines the tourism-led growth hypothesis (TLGH) in

Vietnam during the period 1992–2011 The authors use two-step

procedures to test the hypothesis They first apply cointegration and

the Granger causality test to identify the relationships between

tourism earning and gross domestic product (GDP) Second, they use

growth decomposition methodology to measure the contribution of

tourism to economic growth The results indicate that it is

worth-while for the government to implement economic policies to

stimu-late economic growth through the tourism sector in Vietnam.

Keywords: TLGH; economic growth; cointegration; ECM; Granger

causality; Vietnam

In recent years, tourism business development in Vietnam has attracted research interest Economists have emphasized the importance of tourism to the economy, with rapid tourism growth bringing about an increase of household incomes and government revenues through multiplier effects and improvements in the balance of payments As such, the development of tourism has usually been considered as a positive contribution to economic growth (Dritsakis 2004; Brida

et al, 2008; Cortes-Jimenez and Pulma, 2010; Cortes-Jimenez et al, 2011; Xie

et al, 2011; Ivanov and Webster, 2013) According to estimates of the World

T OURISM E CONOMICS

886

Tourism Organization (2009) the scale of the world tourism industry will reach roughly 11% of the world’s GDP in 2014

Vietnam, in South East Asia, has emerged as a tourist destination for backpackers, culture and nature lovers, sand and sun tourists, and attracts long-stay touring by veterans of the Vietnam War In the period 1992–2011, there was a breakthrough improvement in the number of visitors and tourism de-velopment Vietnam welcomed just over 250,000 foreign tourist arrivals in

1992, compared to 6,014,000 in 2011 The real tourism earnings increased from US$420 million to US$2,203 million

Although the tourism industry has grown significantly in Vietnam, tourism researchers have not paid much attention to the empirical assessment of the role

of the tourism sector to Vietnam’s economy This study therefore aims to further this important area of inquiry by answering the following questions Is there tourism-led growth in Vietnam? How much does tourism contribute to eco-nomic growth in Vietnam? We empirically investigate the relationships be-tween tourism, economic growth and the real exchange rate; the hypotheses are tested empirically by using cointegration and the Granger causality test Finally,

we quantify the contribution of the tourism industry to economic growth through the growth decomposition methodology

Literature review

Similar empirical analysis to this hypothesis has been conducted in different countries in employing different methods For instance, using Spanish data, Balaguer and Cantavell-Jorda (2002) discovered a stable long-run relationship between tourism and economic growth In Turkey, Gunduz and Hatemi (2005) also found empirical support for the tourism-led growth hypothesis (TLGH)

Akinboade and Braimoh (2010) and Brida et al (2010) confirmed empirical

support for the TLGH in South Africa and Uruguay However, most of these studies were based on a model that included three variables – real GDP, tourism earnings and the real exchange rate – and used the Johansen’s cointegration, the error correction model (ECM) and Granger causality test to examine the TLGH In addition, the Granger causality results suggested a bidirectional causality between tourism and economic growth (Seetanah, 2011), unidirec-tional causality with either the TLGH (Katircioglu, 2010, for Singapore;

Cortes-Jimenez et al, 2011 for Tunisia) or economic-driven tourism growth

hypothesis (Oh, 2005; Katircioglu, 2009) However, their common disadvan-tage is that they did not state how much of the economic growth was, in practice, attributable to tourism

In contrast, Ivanov and Webster (2007) used the growth decomposition methodology to measure the contribution of tourism to economic growth They also used the growth of real GDP per capital as a measure of economic growth and disaggregated it into economic growth generated by tourism and other industries The methodology is exemplified with an analysis of the contribution

of specific industries to economic growth (Ivanov and Webster, 2010) Brida

et al (2010) used decomposition to measure the magnitude of the contribution

of tourism to the economic growth for Costa Rica In the case of Colombia,

Brida et al (2008) used the Johansen’s cointegration, ECM and the Granger

Table 1 Tourism statistics of Vietnam, 1992–2011 (US$ million).

Year Real GDP Real GDP hotel and Real tourism earnings a

restaurant sector

Note: a Real tourism earnings of accommodation establishments and travel agencies.

Source: Authors’ data collection from the General Statistical Office of Vietnam and the Vietnam

National Administration for Tourism.

causality test and the growth decomposition methodology to examine the TLGH Considering the impact of tourism led-growth on the world’s economy, research in this field is of continuing importance

Data

Data used in the analysis were gathered from public sources The real GDP annual time series and the real population were collected from the General Statistical Office of Vietnam for the period from 1992 to 2011 Data on the contribution to GDP of the hotel and restaurant sector and tourism earnings were derived from the Vietnam National Administration of Tourism (2011) Finally, a time series of the real effective exchange rate between the Vietnam dong (VND) and other countries was obtained from the International Monetary Fund Table 1 shows the tourism statistics of Vietnam for 1992–2011

Methodology

Cointergration and Granger causality test

Following previous research and the literature (Brida and Risso, 2009; Akinboade

and Braimoh, 2010; Brida et al, 2010; Cortes-Jimenez and Pulina, 2010; Cortes-Jimenez et al, 2011; Seetanah, 2011) we assume that the tourism-growth

model in Vietnam takes the following form:

lnGDPt = β 0 +β 1 lnTOURt + β 2 lnERt + ut.

All the variables are expressed in natural logarithms so that elasticity can also

product; TOUR is tourism earnings (accommodation establishments and travel agencies); ER is the real effective exchange rate; u is the error term with the

conventional statistical properties

We first checked the stationarity of the data by applying unit root tests on the basis of the Augmented Dickey–Fuller (ADF) test (Dickey and Fuller,

T OURISM E CONOMICS

888

1981) If the series were found to be non-stationary, we tested for stationarity through the first or second differences Next, the cointegration test was used

to examine the long-run relationship between tourism and economic growth based on the work of Johansen and Juselius (1990) They propose two test statistics for testing the number of cointergrating vectors: the trace and maxi-mum eigenvalue statistics

The third step was used the ECM technique to find the error correction term The ECM equation for the model in this study is as follows:

ΔlnGDPt = β 0 + β 1 ΔlnTOURt + β 2 ΔlnERt + β 3ECt–1 + ε t,

lnTOURt–1 – β 2 lnERt–1 – β 3ut)

Finally, the Granger causality test exhibits the pairwise causal relationship between the variables under consideration It may be unilateral or bilateral So, this study also used the test to find the causality between GDP and tourism separately by simply running the following two regression models:

ΔlnGDPt =λ 0+Σ m

i=1 λ 1i ΔlnGDPt–i+Σ n

i=1 λ 2i ΔlnTOURt–i+Σ p

i=1 φ 1i ΔlnTOURt–i+ Σ n

i=1 φ 2i ΔlnGDPt–i +Σ q

where µt and ε t are white noise error processes; m, n, p and q denote the number

of lagged variables

Growth decomposition methodology

In order to support the TLGH in Vietnam, we used growth decomposition methodology to assess tourism’s contribution The proportion of GDP produced

by tourism is computed as in Ivanov and Webster (2007), Ivanov and Webster (2010) and Ivanov and Webster (2013) They used the growth of real GDP per

capita g r is:

Y q1(p0)

——

N1

Y q0(p0)

——

N0

where Y q1(p0) is the GDP in constant prices; Y q0(p0) is the GDP in the base year

and N is the average size of the population, and index 1 denote current period,

for which index 0 is the base period They disaggregate the nominator of Equation (1) to separate the tourism GDP in constant prices from the GDP

in constant prices of other industries (ΣY i

q1(p0)), and tourism GDP in base period

q0(p0)):

⎛

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎝

Y t

q1(p0) Σ i ≠ 1Y t

q1(p0) Y i

q0(p0) Σ i ≠ 1Y i

q0(p0)

——– +———— – ——– – ————

Y q0(p0)

——

N 0

They regroup the expressions in the nominator and come to:

Y t

q1(p0) Y t

q0(p0) Σ i ≠ 1Y i

q1(p0) Σ i ≠ 1Y i

q0(p0)

Y q0(p0) Y q0(p0)

And the first component in this expression:

Y t

q1(p0) Y t

q0(p0)

—— – ——

g t

Y q0(p0)

——

N 0

represents the direct contribution of the tourism sector to economic growth in

the period r.

Results and discussion

The results in Table 2 show that all variables were not stationary in their levels and first differences This is evident by comparing the observed values of the ADF test statistics with the critical values of the test statistics at the 5% significance level However, the second differences of all three variables are stationary under the ADF test Hence it is concluded that these variables are integrated of order I(2)

Given that integration of the three series is of the same order, we continued

to test whether the three series are cointegrated over the sample period Table 3 shows the existence of one or two cointegrating vectors at the 5% significance

Table 2 Unit root estimation (ADF test).

Variables Levels First differences Second differences

Statistic 95% CV Statistic 95%C.V Statistic 95%C.V

lnGDP –3.294231 –3.733 –1.999316 –3.691 – 4.990758 – 3.733

lnTOUR –0.944522 –3.676 –2.958250 –3.691 –5.057098 – 3.733

lnER –0.033436 –3.030 –2.908066 –3.040 –3.832942 – 3.067

Source: Authors’ calculations based on Eviews 6.0.

⎛

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎝

⎛

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎜

⎝

T OURISM E CONOMICS

890

Table 3 Johansen cointegration tests results.

Data trend None None Linear Linear Quadratic Test type No Intercept Intercept Intercept Intercept Intercept

No Trend No Trend No Trend Trend Trend

Source: Authors’ calculations based on Eviews 6.0.

Table 4 Pairwise Granger causality tests.

lnTOUR does not 1 24.5225 0.0001 * lnTOUR => Granger cause lnGDP 2 5.13897 0.0227 ** lnGDP lnGDP does not 1 1.01170 0.3295 lnGDP =>

Granger cause 2 1.05950 0.3747 lnTOUR lnTOUR

lnER does not 1 0.41050 0.5308 lnER => Granger cause lnGDP 2 1.17883 0.4363 lnGDP lnGDP does not 1 4.37913 0.05308 lnGDP => Granger cause lnER 2 6.08172 0.0137 ** lnER lnER does not 1 1.23188 0.2834 lnER =>

Granger cause 2 3.08596 0.0800 lnTOUR lnTOUR 3 10.4820 0.0020 *

lnTOUR does not 1 1.85776 0.1918 lnTOUR => Granger cause lnER 2 1.09500 0.3635 lnER

Note: * Significant at the 1% level; ** significant at the 5% level.

Source: Authors’ calculations based on Eviews 6.0.

level On the basis of these results, we can interpret that a unique cointegrating relationship emerges for GDP, tourism and the exchange rate

The dynamics of the cointegration technique were used to explore the long-run equilibrium among the variables Thus, the error term in our model can

be used as the equilibrium error Owing to two variables, GDP and tourism integrated I(2), and the cointegrating relationship, the ECM is:

The relationship between tourism and economic growth shows that tourism

is an exogenous variable in the coefficient of the error correction equation (0.4105) In particular, GDP is adjusted for long-run balance of tourism at a rate of about 41.05% This confirms that tourism is as a factor of long-run economic growth in Vietnam

Table 4 reports the statistical analysis of the causal relationships between

Table 5 Contribution of hotel and restaurant sector to Vietnamese economic growth in the period 1992–2011.

Year Real rate of Real rate of growth of GDP The contribution of

growth of GDP of hotels and restaurants hotels and restaurants to per capita (%) per capita (%) economic growth (%)

Source: Authors’ calculations based on data of the General Statistical Office of Vietnam.

lnGDP, lnTOUR and lnER for Vietnam All hypotheses were tested by a standard F-test The results also suggest that causality runs from tourism

earnings to GDP, but causality does not run from GDP to tourism earnings Table 5 shows the real variation of per capita GDP in the tourism sector and the contribution of tourism (hotel and restaurant sector) to the variation of total GDP This result confirms once again tourism-led growth in Vietnam

Conclusion and policy implications

The empirical results indicate that a long-run relationship exists between tourism and economic growth in Vietnam The causality experiment points out that there is tourism-led growth in Vietnam However, it is clear that the contribution of the hotel and restaurant sector is relatively low compared to its potential Therefore, in order to stimulate economic growth through tourism

in Vietnam, a detailed development strategy is required to give direction to the tourism sector This is beyond the scope of this research, but at the macro level the following measures are suggested in order to stimulate economic growth through tourism:

(ii) improvements to the means of transportation and communication to benefit both domestic and international travellers;

T OURISM E CONOMICS

892

(iii) government support for the developing tourism infrastructure systems and facilities to attract private investors;

(iv) enhancing tourism connections to other sectors, regions and localities to strengthen supply chains and boost ‘spillover’ effects

References

Akinboade, O.A., and Braimoh, L.A (2010), ‘International tourism and economic development in

South Africa: a Granger causality test’, International Journal of Tourism Research, Vol 12, pp 149–

163.

Balaguer, J., and Cantavell-Jorda, M (2002), ‘Tourism as a long – run growth factor: the Spanish

case’, Applied Economics, Vol 34, pp 877–884.

Brida, J.G., and Risso, W.A (2009), ‘Tourism as a factor of long-run economic growth: an empirical

analysis for Chile’, European Journal of Tourism Research, Vol 2, No 2 pp 178–185.

Brida J.G., Pereyra, J.S., Risso, W.A., Devesa, W.J.S., and Aguirre, S.Z (2008), ‘The tourism-led

growth hypothesis: empirical evidence from Colombia’, Tourismos, Vol 4, No 2, pp 13–27.

Brida, J.G., Lanzilotta, B., Lionetti, S., and Risso, W.A (2010), ‘The tourism-led growth hypothesis

for Uruguay’, Tourism Economics, Vol 16, No 3, pp 765–771.

Cortes-Jimenez, I., and Pulina, M (2010), ‘Inbound tourism and long-run economic growth’, Current

Issues in Tourism, Vol 13, No 1, pp 61–74.

Cortes-Jimenez, I., Nowak, J.-J., and Sahli, M (2011), ‘Mass beach tourism and economic growth:

lessons from Tunisia’, Tourism Economics, Vol 17, No 3, pp 531–547.

Dickey, D.A., and Fuller, W.A (1981), ‘Likelihood ratio statistics for auto-regressive time series with

a unit root’, Econometrica, Vol 49, pp 1057–1072.

Dritsakis, N (2004), ‘Tourism as a long-run economic growth factor: an empirical investigation for

Greece’, Tourism Economics, Vol 10, pp 305–316.

Gunduz, L., and Hatemi, A (2005), ‘Is the tourism-Led growth hypothesis valid for Turkey?’,

Applied Economics Letters, Vol 12, pp 499–504.

Ivanov, S., and Webster, C (2007),’Measuring the impact of tourism on economic growth’, Tourism

Economics, Vol 13, No 3, pp 379–388.

Ivanov, S., and Webster, C (2010), ‘Decomposition of economic growth in Bulgaria by industry’,

Journal of Economic Studies, Vol 37, No 2, pp 219–227.

Ivanov, S., and Webster, C (2013), ‘Tourism’s impact on growth: the role of globalization’, Annals

of Tourism Research, Vol 41, 231–236.

Johansen, S., and Juselius, K (1990), ‘Maximum likelihood estimation and inference on

cointegration-with application to the demand for money’, Oxford Bulletin of Economics and Statistics, Vol 52, pp

169–210.

Katircioglu, S.T (2009), ‘Revisiting the tourism-led-growth hypothesis for Turkey using the bounds

test and Johansen approach for cointegration’, Tourism Management, Vol 30, pp 17–20.

Katircioglu, S.T (2010), ‘Testing the tourism-led growth hypothesis for Singapore – an empirical

investigation from bounds test to cointegration and Granger causality tests’, Tourism Economics,

Vol 16, No 4, pp 1095–1101.

Oh, C.O (2005), ‘The contribution of tourism development to economic growth in the Korean

economy’, Tourism Management, Vol 26, pp 39–44.

Seetanah, B (2011), ‘Assessing the dynamic economic impact of tourism for island economies’,

Annals of Tourism Research, Vol 38, No 1, pp 291–308.

Vietnam National Administration of Tourism (2011) ‘International visitors to Vietnam in December and 12 months of 2011’ (http://www.vietnamtourism.gov.vn/english/index.php?cat=012040&itemid

=5143, accessed 26 January 2012).

World Tourism Organization (2009), ‘UNWTO world tourism barometer’ (http://unwto.org/facts/ eng/barometer.htm, accessed 1 April 2009).

Xie, F., Lacher, R.G., and Nepal, S.K (2011), ‘Economic impacts of domestic tourism in the rural

developing world: a case study of Zhangjiaje city, China’, Tourism Review International, Vol 14,

pp 29–42.