TOURIST ARRIVALS AND ECONOMIC GROWTH IN SARAWAK

Transcription

1 MPRA Munich Personal RePEc Archive TOURIST ARRIVALS AND ECONOMIC GROWTH IN SARAWAK Evan Lau and Swee-Ling Oh and Sing-Sing Hu Universiti Malaysia Sarawak, Universiti Malaysia Sarawak, Universiti Malaysia Sarawak 6 August 2008 Online at MPRA Paper No. 9888, posted 9 August 2008 :07 UTC

2 TOURIST ARRIVALS AND ECONOMIC GROWTH IN SARAWAK Evan Lau ψ, Swee-Ling Oh and Sing-Sing Hu Departent of Econoics, Faculty of Econoics and Business, Universiti Malaysia Sarawak Kota Saarahan Sarawak, Malaysia Abstract This study epirically investigates the cooveents and the causality relationship between tourist arrivals and econoic growth in Sarawak during the period of 972 to The epirical evidence clearly shows that the long run causality running fro tourist arrivals to econoic growth in the estiation period. As one of the incoe generator for Sarawak, the findings are consistent with econoic theory and proffer iportant policy conclusions. Keywords: Tourist arrivals; Econoic growth; Sarawak; causality.. INTRODUCTION Touris is one of the fastest growing sectors in an econoy and the ajor source of incoe for ost of the developing countries. According to the World Touris Organization (2007) ( the tourists expenditure on goods and services represented about 8 percent of total world export receipts and 5 percent of the world GDP. Nuerous studies have been conducted dealing with the ipact of touris toward the econoic growth in developing countries. Many of these studies have focused on testing whether touris sector leads to iproveent in econoic growth perforance of a particular country (see for ψ Corresponding author: Tel: , Fax: , E-ail: lphevan@feb.unias.y WTO is the central body that plays a vital role in prooting the developent of responsible, sustainable and universal accessible touris and pays particular attention to the interest of developing countries. This organization encourages the ipleentation of Global Codes of Ethics for touris with the view that eber states axiize positive econoic, social and cultural affects of touris and fully reap its benefits and iniize the negative social and environental ipacts. As of 2007, WTO s ebership includes 50 countries, seven territories and 350 affiliate ebers, which represent the private sectors, educational institutions, touris associations and local touris authorities in the world.

3 exaple, Lea, 988; Akis, 998; Sinclair, 998; Lanza et al., 2003; Durbarry, 2004; Narayan, 2004, 2006; Lee and Chien, 2008 and Lee and Chang, 2008). Epirically these authors argued that that touris has prootes high econoic perforance, job creation and generating revenue for the governent. In addition, Nath (998) found that touris is the leading source of foreign exchange in Mauritius. As one of the incoe earner, it attracts great interest fro a policy perspective for continuous econoic growth in a country/state. With this otivation, this study intends to investigate the relationship between tourist arrival and econoic growth within the scope of Sarawak during the period 972 to In Sarawak, the touris industry is one of econoic sector that contributes greatly to the state s Gross Doestic Product (GDP) 2. With its colorful and ultiracial cultures with diverse festivals have attracted tourists fro different countries to Sarawak where the attractions were the cultural festival and celebrations of the different races and religions. For exaple, Cultural Village and the Sarawak Museu provide a different experience the ulti-ethnic tribes lifestyle in Sarawak. Besides the natural heritage, Sarawak has a treasure trove of flora and fauna and the largest rainforest than any states in Malaysia. 2 The Sarawak Touris Board that establishes in July 995 with the ai to proote Sarawak as a natural touris destination. They work closely with the Touris Malaysia and State Ministry of Touris in aking Sarawak as ain destinations for tourists worldwide. 2

4 Figure : Nuber of Tourist Arrivals to Sarawak fro year Tourists Arrivals in Sarawak Tourists Arrivals Source: Iigration Departent of Sarawak Year Figure plots the tourist arrivals to Sarawak fro 972 until By looking at the figure, one can identify that for the ost part of the tie frae, the arrival of tourists is in increasing anner except for the year 980, 983 and 984. For exaple, tourist increased fro in 972 to 9788 in 978, an increase about 44 percent. Over the decades of 980s, the aount increased fro to in 989. In 990s, when governent launch the first and second Visit Malaysia Year capaign, the arrival of tourists to Sarawak increase fro to peoples. With the success of the earlier Visit Malaysia Year capaign, the figures of tourists keep increasing throughout the years. The introduction of third Visit Malaysia capaign in 2007 aied to attract the foreign tourists to all the states in Malaysia including Sarawak. Also, the governent provides onetary incentives in the yearly budgets especially in prooting the growth and developent of the touris sector in Malaysia. 3

5 The paper is organized as follows. Section 2 provides the epirical approach adopted in the paper. Section 3 reports the epirical findings while concluding rearks is in Section ECONOMETRIC APPROACH 2. Univariate Unit Root Testing Procedures The standard ADF (see Said and Dickey, 984) and DFGLS (see, Elliott et al., 996) testing principles share the sae null hypothesis of a unit root. Their difference however centered on the way the latter specified the alternative hypothesis and treats the presence of the deterinistic coponents in a variable s data generating process (DGP). Specifically the DFGLS procedure relies on locally deeaning and/or detrending a series prior to the ipleentation of the usual auxiliary ADF regression. The use of the DFGLS tests statistics is likely to iniize the danger of erroneous inferences eerging when the series under investigation has a ean and/or linear trend in its DGP. This is so because these statistics have been shown to achieve a significant gain in power over their conventional ADF counterparts (Elliott et al., 996). The DFGLS ean (µ) and trend (τ) stationarity under a local alternative will be denoted by τµ and ττ respectively where they are constructed by estiating the following auxiliary regression of x n = 0 xt + β j xt j + ε t j= β () where x is the locally deeaned and/or detrended process obtained fro xt = xt β zt. Under this condition, z t = for the case of τµ while z t = ( t) for the case of τ τ and β is the regression coefficient of ~ xt on z~ t for which ( ~ x ~ ~, x2,..., xt ) = [ x ( ρl) x2 ),..., ( ρ L) xt )], ~ z ~ ~, z2,..., zt ) = [ z,( ρ L) z,...( ρl) zt ] under the local alternative of ρ = + ( c / T). The τ µ ( τ τ ) ( 2 test statistic is given by the usual t statistic for testing β 0 = 0 auxiliary regression for the appropriate x t in the associated ADF type variables shown in Equation (5). In addition, this 4

6 procedure requires the choice of the local to unity paraeter c through ρ = + ( c / T) are set to -7 in the case of τ µ and 3.5 in the case of τ τ (see Elliott et al., 996 for details). In contrast, the KPSS (Kwiatkowski et al., 992) sei-paraetric procedure tests for level (η µ ) or trend stationarity (η τ ) against the alternative of a unit root. The KPSS test statistic for level (trend) stationary is T 2 S 2 2 t s ( k) T t= η ) = (2) µ ( η τ where S t = t i= u i, ut are the residuals fro the regression of t X on a constant (a constant and trend) for the level (trend) stationarity, s 2 ( k ) is the non-paraetric estiate of the long run variance of ut while k stands for the lag truncation paraeter. In this sense, the KPSS principles involve different aintained hypothesis fro the ADF and DFGLS unit root tests. 2.2 Cointegration Procedure The syste-based cointegration procedure developed by Johansen and Juselius (990) to test the absence or presence of long run equilibriu is adopted in this paper. One advantage of this approach is that the estiation procedure does not depend on the choice of noralization and it is uch ore robust than Engle-Granger test (see Gonzalo, 994). Phillips (99) also docuented the desirability of this technique in ters of syetry, unbiasedness and efficiency. Their test utilizes two likelihood ratio (LR) test statistics for the nuber of cointegrating vectors: naely the trace test and the axiu eigenvalue test. The Johansen procedure is well known in the tie series literature and the detail explanation are not presented here. 5

7 2.3 Granger Causality Tests If cointegration is detected, then the Granger causality ust be conducted in vector error correction odel (VECM) to avoid probles of isspecification (see Granger, 988). Otherwise, the analyses ay be conducted as a standard first difference vector autoregressive (VAR) odel. VECM is a special case of VAR that iposes cointegration on its variables where it allows us to distinguish between short run and long run Granger causality. The relevant error correction ters (ECTs) ust be included in the VAR to avoid isspecification and oission of the iportant constraints. The existence of a cointegrated relationship in the long run indicates that the residuals fro the cointegration equation can be used as an ECT as follows: n GDPt = α 0 + β, i GDPt i + β 2, i TOURt i + µ ECTt + ζ t i= i= (3) n TOURt = δ 0 + φ, i GDPt i + φ2, i GDPt i + µ 2ECTt + ζ 2t i= i= (4) where is the lag operator, α 0, δ 0, β ' s and φ' s are the estiated coefficients, and n are the optial lags of the series Gross Doestic Product (GDP) and tourist arrivals (TOUR), ζ it s are the serially uncorrelated rando error ters whileµ and µ 2 easure a single period response of the GDP (TOUR) to a departure fro equilibriu. To test whether TOUR does not Granger cause oveent in GDP, H 0 : φ2, i = 0 for all i andµ 2 =0 in Equation (3) 3. The rejection iplies that TOUR causes GDP. Siilar analogous restrictions and testing procedure can be applied in testing the hypothesis that GDP does not Granger cause oveent in TOUR where the null hypothesis H 0 : β 2, i = 0 for all i andµ = 0 in Equation (4). In the case where cointegration is absence, the standard first difference vector 3 The F-test or Wald χ 2 of the explanatory variables (in first differences) indicates the short run causal effects ( φ2, i = 0 for all i ) while the long run causal ( µ 2 =0) relationship is iplied through the significance of the lagged ECT which contains the long run inforation. 6

8 autoregressive (VAR) odel is adopted. This sipler alternative of causality is feasible through the eliination of ECT fro both equations above. In other words, it only contains the short run causality inforation. 3. EMPIRICAL RESULTS 3. Data Sources Annual data of tourist arrivals and noinal Gross Doestic Product over the period for Sarawak were obtained fro Iigration Departent of Sarawak and Federal Accounting Unit Sarawak, Departent Statistic of Sarawak respectively. 3.2 Non-stationary and Stationary Test Overwhelingly, the ADF, DFGLS and KPSS tests suggest the existence of unit root or nonstationarity in level or I() for the two variables during the period of The findings that all the variables have the sae order of integration allowed us to proceed with the Johansen cointegration analysis Cointegration Results Based on the axiu eigenvalue and trace statistics, the results iply the existence of one cointegrating vector (see Table ). On the basis of these test results, we can interpret that a unique cointegrating relationship eerged for GDP and arrival of tourists in Sarawak. Rejecting the null hypothesis of no cointegration iplies that the two variables do not drift apart and share at least a coon stochastic trend in the long run. 4 The results are not presented here to conserve space. 7

9 Table : Cointegration Test Results k=2 r= Null Alternative Trace 95%C.V λ ax 95%C.V r = 0 r = * * 5.67 r r = Notes: The k is the lag length and the r is the cointergrating vector(s). Chosen r: nuber of cointergrating vectors that are significant under both tests. Asterisk (*) denotes statistically significant at 5 percent level. 3.4 Granger Causality Analysis With cointegration detected in saple period (Table ), we precede the causality analysis in the environent of Vector Error Correction Model (VECM). Results in Table 2 ay be suarized as follow. First, both GDP and TOUR do not cause each other in the short run. This is clearly shown with the insignificant x 2 -statistic and the p-values. Second, the errorcorrection ter (ECT) is significant and the burden of short run adjustent is beared by GDP carries the correct sign (negative). The adjustent easured by ECT stands at 20 per cent in a year which requires an estiate of 5 years to derive towards long run equilibriu due to the short run adjustents. In this sense, any disequilibriu to the syste, Sarawak needs approxiately 5 years to reach long run equilibriu or the ipact are fully absorbed in the econoy. Third, the significance of the ECT indicates the long run causality fro TOUR GDP. Table 2: Granger Causality Dependent Variables GDP TOUR ECT x 2 -statistic (p-value) Coefficient t-ratio GDP (0.568) * TOUR (0.34) The x 2 -statistic tests the joint significant of the lagged values of the independent variables while one error correction ter in included in the analysis. Asterisk (*) indicates statistically significant at 5 percent level. 8

10 4. CONCLUSION General consensus fro the vast aount of research has eerged that touris developent it not only increases foreign exchange incoe, but also creates eployent opportunities, stiulates the growth of the touris industry and by virtue of this, triggers overall econoic growth (Lee and Chang, 2008). As such, touris developent has becoe an iportant target for ost governents worldwide. In this regard, this study was conducted to contribute to the body of literature in respect to Sarawak, also known as the land of Hornbills or Bui Kenyalang ; largest state in Malaysia covering the land size of 24,450 sq. k. As in Figure, the nuber of arrivals has increased significantly since 970s into Sarawak. The epirical results indicate that long run cooveent relationship exists between tourist arrival and econoic growth in Sarawak. Causality experient point out that a continuous touris developent leads to expansion in the econoic growth for Sarawak. This supports the general consensus that touris developent acts as an engine of econoic growth for Sarawak during the estiation period. Therefore, touris-led econoic growth in Sarawak. Besides that, the developent of touris sector will also lead to the developent of other related industries. With the encourageent and financial support fro the governent agency, the arrival of tourists (doestic and foreign) into Sarawak would increase in a consistent anner each year. On the local front, Sarawak Touris Board work closely with players of the touris industry to intensify their prootions as well as developent of their products and services in eeting the expected influx of tourists into Sarawak. 9

11 REFERENCES Akis, S. (998) A copact econoetric odel of touris deand for Turkey. Touris Manageent, 9, Durbarry, R. (2004) Touris and econoic growth: The case of Mauritius. Touris Econoics, 0, Elliott, G., Rothenberg, T.J. and Stock, J.H. (996) Efficient tests for an autoregressive unit root. Econoetrica, 64, pp Federal Accounting Unit Sarawak, Departent Statistic of Sarawak, various issues, Kuching, Sarawak. Gonzalo, J. (994) Five Alternative Methods of Estiating Long Run Equilibriu Relationships. Journal of Econoetrics, 60, Iigration Departent of Sarawak, various issues, Kuching, Sarawak. Johansen, S. and Juselius, K. (990) Maxiu likelihood estiation and inference on cointegrating with applications for the deand for oney. Oxford Bulletin of Econoics and Statistics, 52, pp Kwiatkowski, D., Phillips, P.C.B., Schidt, P. and Shin, Y. (992) Testing the null hypothesis of stationarity against the alternative of a unit root. How sure are we that econoic tie series have a unit root? Journal of Econoetrics, 54, pp Lanza, A., Teple, P. and Urga, G. (2003) The iplications of touris specialization in the long run: An econoetric analysis for 3 OECD econoies. Touris Manageent, 24, Lea, J. (988) Touris and Developent in the Third World, London: Routledge. Lee, C. C. and Chang, C. P. (2008) Touris developent and econoic growth: A closer look at panels. Touris Manageent, 29, Lee, C. C. and Chien, M. S. (2008) Structure breaks, touris developent, and econoic growth: Evidence fro Taiwan. Matheatics and Coputers in Siulation, 77, Narayan, P. K. (2004) Econoic ipact of touris on Fiji s econoy: Epirical evidence fro the coputable general equilibriu odel. Touris Econoics, 0, Narayan, P. K. (2006) Are Australia s touris arkets converging? Applied Econoics, 38, Nath, S. (998) Environent resources, international touris and taxation: The case of Mauritius. Social Science & Huanities and Law & Manageent Research Journal,, Phillips, P. C.B. (99) Optial Inference in Cointegrated Systes, Econoetrica, 59, Said, E.S. and Dickey, D.A. (984) Testing for unit roots in autoregressive-oving average odels of unknown order. Bioetrika, 7, pp Sinclair, T. (998) Touris and econoic developent: A survey. Journal of Developent Studies, 34, -5. World Touris Organization. (2007). About United Nation World Trade Organization. Touris. World Wide Web: 0