Tourism-induced financial development in Malaysia: New evidence from the tourism development index

Abstract

This article explores the relationship between tourism development and financial development by incorporating economic growth and the real effective exchange rate as additional determinants in the finance demand function of the Malaysian economy. Thus, the analysis relies on the bounds testing approach by accommodating structural breaks to examine the cointegration among the variables and investigates the causal direction between the variables by applying the Toda–Yamamoto Granger causality approach. The results show that all the variables are cointegrated over the period from 1975 to 2016; tourism development is positively related to financial development; economic growth is positively linked with financial development; and real exchange rate is negatively associated with financial development. The Granger causality analysis demonstrates the presence of bidirectional causality between tourism development and financial development as well as a unidirectional causal relationship running from financial development and tourism development to economic growth.

Keywords

demand function financial development Malaysia tourism development

Introduction

In the last 20 years, an increasing body of literature has extensively examined the relationship between tourism activity and economic growth in developing and developed countries. The fundamental motive for this increasing interest is the key role of tourism expenditure in stimulating economic growth in many countries. According to the statistics of the World Tourism Organization, international tourism expenditure¹ increased dramatically from $462 billion in 1995 to $1371 billion in 2015; during this same time period, the Malaysian tourism expenditure increased by approximately 400% to $10.5 billion in 2015. It is well established that tourism induces an increase in a country’s household income, employment creation, tax revenues, and a positive balance of payment (Durbarry, 2002; Dwyer et al., 2004).

Analogous to the tourism–growth nexus is the connection between financial development and tourism activity, which has received scant attention in the literature. Notably, several studies have demonstrated a positive link between financial development and economic growth. In line with this view, King and Levine (1993) observe that development in the banking sector leads to an increase in economic growth; this finding implies that a more efficient financial system increases the gross domestic product (GDP) of an economy. Ang (2008) examines the link between financial development and economic growth in Malaysia. The empirical findings indicate a positive connection between financial development and higher output growth in Malaysia. The generally increasing trend of tourism in Malaysia has been contributing toward stimulating new investment, increasing industrial production, augmenting foreign exchange reserves, and decreasing unemployment. Hence, tourism generates positive economic externalities. As a result, the development of the travel and tourism sector has often been considered to enhance financial development and promote economic growth, employment, and welfare (e.g. Chao et al., 2006, 2009; Lee and Kwon, 1995).

In 2016, the global travel and tourism sector generated US$7.6 trillion (equivalent to 9.8% of the global GDP) and supported 284 million jobs, corresponding to 1–10 jobs in the global economy.² In Malaysia, the tourism industry is a particularly important contributor to economic growth and employment, and in the last few years, this country has become an attractive tourist destination in the Asian region, with tourist arrivals increasing by 6.7% to 27.4 million in 2014. Since the adoption of the Third Industrial Master Plan 2006–2026, the development of the tourism sector has become a matter of prime importance.

The Malaysian Minister of Tourism and Culture, Dato’ Seri Mohamed Nazri Abdul Aziz, has defined an ambitious goal of attracting 36 million arrivals by 2020. To achieve this target, the Malaysian government has been implementing the best possible environmental conditions for tourism. Some collaborative efforts between private and public sectors have also been conducted to ensure that the tourism industry continues to thrive. In 2015, the total contribution³ of tourism activity to Malaysia achieved a record in generating employment and value added, that is, the value added generated by the tourism industry increased by 10.3% per year to reach RM 166 billion in 2015, compared with RM 101 billion in 2010. Concomitantly, the number of employees increased by 4.4% per year to reach 2.5 million, compared with 2 million in 2010 (The Malaysian Department of Statistics).

Various authors have already demonstrated that tourism stimulates the Malaysian economy. For example, Kadir et al. (2010) and Lau et al. (2009) have proved that the reinforcement of tourism development led to an increase in GDP in Malaysia. In the same vein, Kumar et al. (2015) find that tourism accounted for approximately 0.26% of the long-run growth of the Malaysian economy from 1975 to 2012. Their analysis clearly supports bidirectional Granger causality between tourism and capital per worker and implies that tourism and private and public investments mutually reinforce each other. Similarly, Shahbaz et al. (2017) investigate the tourism–growth nexus in Malaysia using an augmented Solow production function from 1975 to 2013. They incorporate trade openness and financial development into this function, and their major empirical findings indicate the existence of bidirectional pairwise Granger causality between tourism and output per capita.

Policy makers should implement relevant economic decisions to sustain efforts to spur growth in the tourism sector. An accurate understanding of the causal direction between tourism development, financial development, and economic growth is essential for the success of macroeconomic policies. Few empirical studies have investigated the long- and short-run relationships between tourism activity, financial development, and economic growth (e.g. Cannonier and Burke, 2017; Ohlan, 2017; Shahbaz et al., 2017). For instance, Cannonier and Burke (2017) find that an increase in tourism expenditure contributes to increasing the depth of the Caribbean financial system; this finding implies that tourism activity affects financial development positively. Another notable fact highlighted by Shahbaz et al. (2017) is that the promotion of private investments and private savings may improve tourism activity in Malaysia. More specifically, financial development positively affects economic growth and tourism development from 1975Q1 to 2013Q4. Additionally, the Granger causality analysis indicates a bidirectional effect between tourism demand and financial development, implying a mutually reinforcing effect between the variables. Therefore, the establishment of a healthy financial system seems crucial to the success of tourism development policies in Malaysia.

Although several empirical studies have investigated tourism–growth nexus, a study that focuses on the relationship between financial development and tourism development in Malaysia is necessary (Shahbaz et al., 2017). To the best of our knowledge, this study is the first to empirically examine the relationship between financial development (the dependent variable in the model) and some other crucial economic indicators in the context of Malaysia. Thus, the analysis investigates whether tourism development, economic growth, and the real effective exchange rate (REER) contribute effectively to the financial development of the Malaysian economy.

Several structural changes that affected Malaysia’s tourism sector are incorporated into our analysis for the period from 1975 to 2016. For instance, Malaysia was recently ranked 26th of 184 countries regarding the relative importance of the contribution of the country’s tourism to its output. This finding shows that the Malaysian tourism industry has achieved rapid growth and undergone development in the last few years. The Malaysian government has recently underlined that tourism is fundamental to improving their GDP. The tourism industry directly accounts for 6% and indirectly greater than 15% of Malaysian’s GDP.

This study makes four significant contributions to the literature. (i) The linkage between financial development and tourism development while controlling for the REER and economic growth is explored. (ii) A battery of relevant econometric tests that consider the characteristics of the economic and financial data is applied, and the long-run relationship is investigated by the autoregressive distributed lag (ARDL) cointegration analysis developed by Pesaran et al. (2001) by accommodating structural breaks in time series data. (iii) The causal linkages are tested by applying the Toda and Yamamoto (TY) (1995) Granger test. (iv) Finally, the robustness of the Granger causality’s results is investigated by applying the innovative accounting approach (IAA).

The empirical findings confirm a long-run relationship between financial development and its determinants. Specifically, tourism development leads to financial development. Economic growth and the REER are positively and negatively linked with financial development, respectively. The feedback effect is observed between financial development and tourism development, as confirmed by the TY Granger causality test. An improved understanding of the interactions between tourism activity and financial development can highlight the key role of the financial system in the promotion of economic growth and tourism development in Malaysia.

The remainder of this article is structured as follows. The second section presents the related literature, the third section describes the methodology, the fourth section discusses the empirical findings, and the fifth section concludes and provides some policy implications

Literature review

Tourism-growth nexus

An increasing body of literature has debated the potential effects of tourism development on the economy, especially on economic growth. These links have been investigated by considering four hypotheses (e.g. Balaguer and Cantavella-Jorda, 2002; Brida et al., 2008; Chou, 2013; Gunduz and Hatemi-J, 2005; Narayan et al., 2010). First, the economic growth-led tourism hypothesis specifies the existence of Granger causality running from economic growth to tourism development. In this situation, adequate investment in physical and human capital should support government policies to boost the tourism sector. Second, the tourism-led growth hypothesis asserts a Granger causality running from tourism development to economic growth. In this case, tourism is the main impetus to economic growth. Third, the feedback hypothesis identifies the existence of bidirectional Granger causality between tourism development and economic growth. Fourth, the neutrality hypothesis does not consider the existence of any Granger causality relationship between tourism development and economic growth. In summary, this literature has provided evidence that tourism increases foreign exchange earnings, household income, government revenues, taxes, and employment (e.g. Balaguer and Cantavella-Jorda, 2002; Dritsakis, 2004).

In line with these findings, many governments have formulated plans to promote tourism development. Nevertheless, results of these empirical studies are sparse and no consensus is observed, suggesting that the tourism–growth nexus may be country-specific (e.g. Arslanturk et al., 2011; Lee and Chang, 2008, Tang and Jang, 2009). Countries can be different in terms of size, trade openness (Kim et al., 2006), financial development, and the degree of specialization in tourism (Oh, 2005). Singh (2008) provides evidence that developing countries are likely to move their economies toward tourism development. Conversely, Salmani et al. (2014) argue that tourism activities increase economic growth in developing and developed countries, albeit possibility more prominent in developing countries.

The empirical findings have also been sensitive to the sample period and methodology employed to test for Granger causality between tourism activity and economic growth (Arslanturk et al., 2011), and several studies have confirmed that the tourism-led growth hypothesis holds (e.g. Brida et al., 2008, 2015; Ertugrul and Mangir, 2015; Jalil et al., 2013; Tang and Abosedra, 2016; Tang and Tan, 2015). For example, Balaguer and Cantavella-Jorda (2002) empirically examine the role of tourism development in explaining the economic growth in Spain, observe that tourism accounts for approximately 5.9% of GDP, and demonstrate that application of the Granger causality tests confirms the tourism-led growth hypothesis. These findings justify public policies focused on promoting tourism in Spain.

By applying an econometric methodology similar to that aforementioned, in Turkey, from 1963 to 2002, Gunduz and Hatemi-J (2005) observe a unidirectional Granger causality running from tourism development to economic growth. Similarly, in Chile, from1988 to 2008, Brida and Risso (2009) observe a unidirectional Granger causality running from tourism development and the REER to economic growth. In a recent study finding that tourism leads to economic growth, Tang and Abosedra (2016) empirically support the tourism-led growth hypothesis in the Lebanese economy.

The opposite Granger causality relationship, running from economic growth to tourism development, that is, growth-led tourism, has also been investigated in the literature. Among them, Payne and Mervar (2010) employ the TY Granger causality test to investigate the tourism–growth nexus for Croatia from 2000 to 2008 and identify a unidirectional causal link running from economic growth to tourism receipts. The same causal link is found by Lee (2012) and He and Zheng (2011) for Singapore and China, respectively. They suggest that this causal direction is partly related to the limited economic development of these countries.

The feedback effect between tourism and economic growth has also been reported. For instance, Dritsakis (2004) demonstrates bidirectional Granger causality between the Greek tourism and economic growth. Similarly, Lee and Chien (2008) and Ongan and Demiroz (2005) have reported the feedback effect between the variables in Taiwan’s and Turkey’s economies, respectively. Kim et al. (2006) undertake a similar study in Taiwan and observe a bidirectional Granger causal relationship, supporting the tourism-led hypothesis and its reciprocal hypothesis. This finding implies that these economies benefit from international tourism due to public policies oriented toward tourism development. Few studies have revealed a neutral effect between tourism and economic growth (Katircioglu, 2009; Tang and Jang, 2009). For example, Brida et al. (2011) investigate the relationship between tourism and economic growth and report that these variables are independent in Brazil’s economy from 1965 to 2007.

An alternative strand of empirical studies has considered the tourism–growth nexus in a multicountry analysis by examining the long- and short-run relationships between tourism and economic growth using panel data econometric tools. For example, Lee and Chang (2008) investigate the causal relationship between tourism development and economic growth for two panels of countries—OrganiSation for Economic Co-operation and Development (OECD) and non-OECD countries including the Asian, Latin American, and Sub-Saharan countries—from 1990 to 2002. Their findings suggest that tourism activity has a greater effect on the economic growth of non-OECD countries. Their empirical results also underscore a unidirectional Granger-causality running from tourism to growth for the OECD panel, whereas a feedback effect is found in non-OECD countries.

By employing Pedroni’s panel data cointegration, Narayan et al. (2010) investigate these links for a group of countries highly specialized in tourism, observe a significant long-run effect of tourism on economic growth for four Pacific Islands countries, and highlight factors that hamper tourism development in this region, such as the heavy dependence on food imports, natural disasters, and political instability. Chou (2013) investigates the link between tourism development and economic growth in 10 European transitional countries from 1988 to 2011. Chou’s (2013) empirical results are sparse and depend on the investigated transitional countries and support the neutrality hypothesis for three of these 10 transitional countries—Bulgaria, Romania, and Slovenia; the growth hypothesis for Cyprus, Latvia, and Slovakia; and a bidirectional causal effect for Estonia and Hungary.

Tourism–financial development nexus

Few empirical studies have investigated the tourism–growth nexus by adding relevant macroeconomic variables in the domestic production function. The inclusion of these variables might equip the model to limit the problem of omitted-variable bias. Although the tourism–growth relationship depends on a host of factors, financial development is a notable variable that can be examined.

In this context, Kumar and Kumar (2013) investigate the causal link between tourism and economic activity with other relevant variables, such as financial development and urbanization, in Fiji from 1981 to 2009 and observe that financial development is the largest contributor to tourism development. Similarly, Kumar (2014) demonstrates the feedback effect between tourism and output per worker in Vietnam from 1980 to 2010, and the empirical results further reveal a neutral effect between tourism and financial development. Başarir and Çakir (2015) observe the same causal link between tourist arrivals and financial development in the European Union from 1995 to 2010. Shahbaz et al. (2017) observe bidirectional Granger causality between tourism and output per capita in the Malaysia from 1975 to 2013, and their findings indicate that tourism development is central to enhance all sectors and improve the overall income level of Malaysians. Their Granger causality analysis reveals that financial development Granger causes tourism and vice versa. Ohlan (2017) investigates the tourism–growth nexus in India from 1964 to 2014 by considering financial development as an additional factor in the augmented production function. They find that tourism Granger causes economic growth. They also apply a Granger causality analysis and demonstrate that tourism and financial development are independent, implying a neutral effect for the Indian economy.

In the same vein, Cannonier and Burke (2017) examine the link between financial development and tourism activity in Caribbean countries from 1980 to 2013. Using a generalized method of moments (GMM) estimation, their results indicate that tourism expenditures have a positive and significant effect on financial development. Another important finding concerns the negative correlation between the size of the economy and financial development. Despite the emergence of academic literature regarding the causal relationship between tourism development and financial development, the results have been sparse and mixed.

Methodology

This article provides an accurate econometric modeling by investigating the long- and short-run effects of tourism development, economic growth, and the REER on the financial development of the Malaysian economy. Tourism development affects financial development through economic activity. Tourism increases economic activity and demand for financial services and affects financial development (Cannonier and Burke, 2017). Growth in income per capita increases demand for financial services and affects financial development (i.e. the growth-led financial hypothesis) (Levine, 2003; Shahbaz, 2009). The relationship between the REER and financial development is ambiguous, but Boyd et al. (2001) argue that exchange rate leads to inflation and affects the performance of financial markets and financial development. More recently, in 2009, Aghion et al. (2009) demonstrate that a high degree of REER variability leads to lower growth, especially in small open economies such as Malaysia, that is, REER volatility has a negative impact on long-run growth in countries with relatively low financial development.

Based on this theoretical background, the general finance demand function is modeled as follows

F_{t} = f ({TD}_{t}, Y_{t}, R_{t})

where F_t, TD_t, Y_t, and R_t denote financial development, tourism development, economic growth, and the REER, respectively. We use a log-linear transformation of the variables to reduce the effects of changing variability of the data and estimate the elasticity coefficients, except the REER. The estimable equation of the model can be expressed as follows

ln F_{t} = β_{1} + β_{2} ln {TD}_{t} + β_{3} ln Y_{t} + β_{4} R_{t} + μ_{t}

where $ln F_{t}$ is the natural log of financial development measured by real domestic credit to private sector per capita,⁴ $ln Y_{t}$ is the natural log of real GDP per capita, R_t is the REER, and $ln {TD}_{t}$ is natural log tourism development comprising tourist arrivals, tourism receipts, and tourism expenditures.⁵ $μ_{t}$ is residual term assumed to be normally distributed. We expect $β_{2} > 0$ if tourism promotes financial development, otherwise $β_{2} < 0$ , and $β_{3} > 0$ if economic growth has positive effect on financial development, otherwise $β_{3} < 0$ . The negative effect of the REER on financial development leads to $β_{4} < 0$ otherwise $β_{4} > 0$ .

This study focuses on the Malaysian economy from 1975 to 2016. The data on domestic credit to the private sector as share of GDP and real GDP (constant 2010 LCU) are from the World Development Indicators (CD-ROM, 2017). The data on tourist arrivals, tourism receipts,⁶ and tourism expenditures are from Tourism Statistics, Ministry of Tourism Malaysia (2017). The data on the REER for Malaysia are from the International Financial Statistics (CD-ROM, 2017).

We used total population to transform data into per capital units, except the REER, following Shahbaz et al. (2017). We employed a principal component analysis to generate a tourism development index (see the Online Supplemental Appendix A, for more details).

The estimation strategy has three stages. First, we test the stationary properties of financial development, tourism development, economic growth, and the REER. In doing so, we apply two traditional unit root tests, namely, the augmented Dickey–Fuller (ADF) developed by Dickey and Fuller (1981) and Phillips and Perron (PP) advanced by Phillips and Perron (1988). We also perform the Zivot and Andrews (ZA) (1992) unit root test to account for structural breaks into the time series investigated that were potentially caused by macroeconomic and political events during the considered period.

Second, we employ a cointegrating approach that fits the characteristics of our variables of interest. The ARDL model bounds approach testing developed by Pesaran et al. (2001) is employed to investigate the long-run relationship between financial development, tourism development, the REER, and economic growth to enhance the power of the cointegration tests. This cointegration technique accurately models the long-run linkages between our considered variables, that is, it provides better results for small sample data. In addition, this technique is more suitable when variables are integrated of order, I(0) or I(1). Traditional methods to test the cointegration require the series to be integrated at a unique order of integration (e.g. Engle and Granger, 1987; Johansen and Juselius, 1990; Phillips and Hansen, 1990). This approach investigates the short and long run. The unrestricted error correction model version of ARDL can be expressed as follows

\begin{array}{l} Δ ln F_{t} = β_{1} + β_{F} ln F_{t - 1} + β_{TD} ln {TD}_{t - 1} + β_{Y} ln Y_{t - 1} + β_{R} R_{t - 1} + \sum_{i = 1}^{p} β_{i} Δ ln F_{t - i} \\ + \sum_{j = 0}^{q} β_{j} Δ ln {TD}_{t - j} + \sum_{k = 0}^{r} β_{k} Δ ln Y_{t - k} + \sum_{l = 0}^{s} β_{l} Δ R_{t - l} + D_{t} + μ_{t} \end{array}

\begin{array}{l} Δ ln {TD}_{t} = δ_{1} + δ_{F} ln F_{t - 1} + δ_{TD} ln {TD}_{t - 1} + δ_{Y} ln Y_{t - 1} + δ_{R} R_{t - 1} + \sum_{i = 1}^{p} δ_{i} Δ ln {TD}_{t - i} \\ + \sum_{j = 0}^{q} δ_{j} Δ ln F_{t - j} + \sum_{k = 0}^{r} δ_{k} Δ ln Y_{t - k} + \sum_{l = 0}^{s} δ_{l} Δ R_{t - l} + D_{t} + μ_{t} \end{array}

\begin{array}{l} Δ ln Y_{t} = φ_{1} + φ_{F} ln F_{t - 1} + φ_{TD} ln {TD}_{t - 1} + φ_{Y} ln Y_{t - 1} + φ_{R} R_{t - 1} + \sum_{i = 1}^{p} φ_{i} Δ ln Y_{t - i} \\ + \sum_{j = 0}^{q} φ_{j} Δ ln F_{t - j} + \sum_{k = 0}^{r} φ_{k} Δ ln {TD}_{t - k} + \sum_{l = 0}^{s} φ_{l} Δ R_{t - l} + D_{t} + μ_{t} \end{array}

\begin{array}{l} Δ R_{t} = θ_{1} + θ_{F} ln F_{t - 1} + θ_{TD} ln {TD}_{t - 1} + θ_{Y} ln Y_{t - 1} + θ_{R} ln R_{t - 1} + \sum_{i = 1}^{p} θ_{i} Δ R_{t - i} \\ + \sum_{j = 0}^{q} θ_{j} Δ ln F_{t - j} + \sum_{k = 0}^{r} θ_{k} Δ ln {TD}_{t - k} + \sum_{l = 0}^{s} θ_{l} Δ ln Y_{t - l} + D_{t} + μ_{t} \end{array}

where Δ is the difference operator, μ_t is the residual term, and p, q, r, and s are lag orders selected by the Akaike information criterion (AIC). D_t is a dummy variable used to capture the effect of the structural break date (t), which is derived from the ZA unit root test applied on the time series. Estimation of the ARDL model allows testing the null of no cointegration between the variables in equation (1) if, for example, in equation (2), $H_{0} : β_{F} = β_{TD} = β_{Y} = β_{R}$ is against the alternative of no cointegration $H_{0} : β_{F} \neq β_{TD} \neq β_{Y} \neq β_{R}$ . Using the nonstandard F-bounds test, Pesaran et al. (2001) develop asymptotic critical values for the upper and lower critical bounds., the In particular bounds test computes lower bound critical values for a given confidence level by assuming that all variables are stationary. Next, the test computes an upper critical bound for the same confidence level by assuming that all variables are I(1). The decision rule is as follows: If the empirical F-statistic is less than the lower bound critical value, the test fails to reject the null of no cointegration while rejecting the null of no cointegration in the opposite case. Our computed F-statistics are $F_{F} (F / TD, Y, R)$ , $F_{TD} (TD / F, Y, R)$ , $F_{Y} (Y / F, TD, R)$ , and $F_{R} (R / F, TD, Y)$ for equations (3) to (6), respectively. Thus, we use the critical values collected by Narayan (2005), rather than those of Pesaran et al. (2001).

Residual tests are performed to assess the goodness of fit of the ARDL estimation and to check for serial correlation, functional form, heteroscedasticity, and normality of the residual term. The stability of the ARDL parameters is investigated using the cumulative sum (CUSUM) of residuals and the CUSUM of square (CUSUMsq). The Chow forecast test is also performed to test for the constancy of the ARDL coefficients. This battery of tests specifies that the long- and short-run parameters are consistent and stable for studying the linkages between our considered variables.

The third stage tests the direction of Granger-causality between financial development, economic growth, tourism development, and the REER. We use the Toda and Yamamoto (TY) (1995) approach to perform our Granger-causality links between our considered variables. This version of the so-called Granger causality test has been demonstrated to avoid the problem of invalid asymptotic critical values when all the variables are nonstationary series. This test is valid whether or not the variables are integrated at the same order. We run the TY test of Granger causality test by considering the presence of additional lags $k + 1, \dots, k + d$ into a vector autoregression (VAR) model as follows

ln F_{t} = β_{1} + \sum_{i = 1}^{k + d max} β_{2} ln F_{t - 1} + \sum_{i = 1}^{k + d max} β_{3} ln {TD}_{t - 1} + \sum_{i = 1}^{k + d max} β_{4} ln Y_{t - 1} + \sum_{i = 1}^{k + d max} β_{5} R_{t - 1} + ϖ_{t}

ln {TD}_{t} = δ_{1} + \sum_{i = 1}^{k + d max} δ_{2} ln {TD}_{t - 1} + \sum_{i = 1}^{k + d max} δ_{3} ln F_{t - 1} + \sum_{i = 1}^{k + d max} δ_{4} ln Y_{t - 1} + \sum_{i = 1}^{k + d max} δ_{5} R_{t - 1} + ϖ_{t}

ln Y_{t} = φ_{1} + \sum_{i = 1}^{k + d max} φ_{2} ln Y_{t - 1} + \sum_{i = 1}^{k + d max} φ_{3} ln F + \sum_{i = 1}^{k + d max} φ_{4} ln {TD}_{t - 1} + \sum_{i = 1}^{k + d max} φ_{5} R_{t - 1} + ϖ_{t}

R_{t} = θ_{1} + \sum_{i = 1}^{k + d max} θ_{2} ln R_{t - 1} + \sum_{i = 1}^{k + d max} θ_{3} ln F_{t - 1} + \sum_{i = 1}^{k + d max} θ_{4} ln {TD}_{t - 1} + \sum_{i = 1}^{k + d max} θ_{5} ln Y_{t - 1} + ϖ_{t}

where F_t, TD_t, Y_t, and R_t are financial development, tourism development, economic growth, and the REER, respectively. In empirical equations (7) to (10), each variable of interest is regressed on each other with a lag order starting from 1 to k + d max, where k is optimal lag order, d is the maximum order of integration for each variables of the models, and ϖ_t is the residual term.

Results, interpretations, and discussion

We report the descriptive statistics and pairwise correlations in Table 1. We observe that tourism development is highly volatile when compared with financial development. Economic growth volatility is less than the volatility that stems from the REER. The Jarque–Bera test analysis reveals that financial development, tourism development, economic growth, and the REER have normal distributions. The correlation analysis indicates that tourism development is positively correlated with financial development. A positive correlation exists between economic growth and financial development. The REER is negatively correlated with financial development, tourism development, and economic growth. Economic growth and tourism development are positively correlated.

Table 1.

Descriptive statistics and pairwise correlations.

Variables	$ln F_{t}$	$ln {TD}_{t}$	$ln Y_{t}$	R_t
Mean	8.5458	9.1080	8.6408	2.9949
Median	9.0055	9.2625	8.7511	2.7295
Maximum	9.5231	12.863	9.3082	4.1450
Minimum	6.7538	4.2508	7.8233	2.1818
Standard deviation	0.8185	2.7168	0.4381	0.6093
Skewness	−0.7872	−0.1892	−0.1975	0.3849
Kurtosis	2.3964	1.7493	1.7425	1.6456
Jarque–Bera	2.9759	2.9878	3.0402	4.2474
Probability	0.2330	0.2244	0.2186	0.1195
$ln F_{t}$	1.0000
$ln {TD}_{t}$	0.2106	1.0000
$ln Y_{t}$	0.2489	0.1604	1.0000
R_t	−0.5270	−0.02806	−0.3498	1.0000

To examine the stationarity properties of the variables, we apply the ADF, PP, and ZA structural break unit root tests. The results are reported in Table 2. We observe that tourism development is stationary at level, but financial development, economic growth, and the REER have a unit root problem at level with intercept and trend. We apply the PP unit root test to validate the results provided by the ADF unit root test. The PP unit root test reveals that all the variables are nonstationary at level but, after first taking the difference, financial development, tourism development, economic growth, and the REER are stationary. This result implies that all the variables have a unique order of integration, namely, I(1).

Table 2.

Unit root analysis.

Variable	ADF unit root test		PP unit root test
Variable	Level	First difference	Level	First difference
$ln F_{t}$	−2.5547 (1)	−5.5008 (4)*	−2.1148 (3)	−5.7471 (3)*
$ln {TD}_{t}$	−3.6486 (2)**	…	−2.9570 (3)	−14.7516 (3)*
$ln Y_{t}$	−2.1997 (1)	−4.3133 (2)*	−2.3501 (3)	−5.2022 (3)*
R_t	−2.4509 (3)	−4.1499 (3)**	−1.9180 (3)	−5.3973 (3)*
Variable	ZA test at level		ZA test at the first difference
Variable	T-statistic	Break year	T-statistic	Break year
$ln F_{t}$	−2.256 (2)	1980	−6.246 (3)*	1992
$ln {TD}_{t}$	−4.120 (1)	2007	−5.666 (2)*	1999
$ln Y_{t}$	−3.662 (3)	1991	−5.999 (1)*	1998
R_t	−4.104 (3)	1986	−6.415 (1)*	1985

Note: This table reports the findings of the three unit-root tests used in this study. ADF, PP, and ZA denote the augmented Dickey–Fuller, the Philips and Perron, and the Zivot and Andrews unit root tests, respectively.

*** Significance at 1% level.

** Significance at 5% level.

* Significance at 10% level.

The worth of traditional unit root tests, that is, ADF and PP, is questionable if a time series contains structural breaks; thus, we applied the ZA unit root test, which contains information regarding the single unknown structural break in the series. The results provided by the ZA unit root test are reported in Table 2 (lower segment). We observe that financial development, tourism development, economic growth, and the REER contain the unit root problem. The structural breaks 1980, 2007, 1991, and 1986 are for financial development, tourism development, economic growth, and the REER, respectively. The Malaysian government implemented financial and economic reforms in the mid-1980s to attract foreign direct investment, which affected the development of the financial sector. All the variables are stationary after first differenced, and this result implies that financial development, tourism development, economic growth, and the REER are integrated at I(1).

The unique order of integration of the variables leads us to apply the bounds testing approach to examine the cointegration among variables for the Malaysian economy from 1975 to 2016. Before applying the bounds testing approach, we chose the appropriate lag length using the AIC. The AIC has superior explanatory power. The bounds F-statistic is sensitive to the choice of the lag length. The empirical results provided by the bounds F-test would be ambiguous if the lag length selection is not appropriate. Therefore, we chose the lag length based on the AIC to attain consistent and efficient empirical results.

Next, we perform the bounds F-test estimation, and the results are reported in Table 3. We include a dummy variable capturing the structural break period of 1980–2016 in the financial development-based ZA unit root test, by following Shahbaz et al. (2017). Notably, the calculated bounds F-statistics are greater than the upper critical bounds at the 1% and 5% levels, respectively, because we treated financial development and tourism development as dependent variables. This result leads us to reject the null hypothesis of no cointegration between the variables. We may accept the null hypothesis of no cointegration between the variables because we considered economic growth and the REER as dependent variables.⁷ This implies the presence of the two cointegrating vectors and confirms the existence of financial development, tourism development, economic growth, and REER from 1975 to 2016 by accommodating structural breaks in the series. The application of CUSUM and CUSUMsq validates the stability of the ARDL cointegration analysis.

Table 3.

ARDL cointegration analysis.

Estimated models	Bounds testing to cointegration			Diagnostic tests		Stability test
Estimated models	Lag length	Structural break	F-statistics	$χ_{ARCH}^{2}$	$χ_{RESET}^{2}$	CUSUM	CUSUMsq
$F_{F} (F / TD, Y, R)$	1, 1, 1, 1	1980	7.995*	0.4006	1.2639	Stable	Stable
$F_{T} (TD / F, Y, R)$	1, 0, 1, 1	2007	9.265**	0.0119	0.9852	Stable	Stable
$F_{Y} (Y / F, TD, R)$	1, 1, 0, 1	1991	2.600	0.2745	1.2934	Stable	Stable
$F_{R} (R / F, TD, Y)$	1, 1, 1, 0	1986	4.222	1.1461	0.6195	Stable	Stable
Significant level	Critical values^#
Significant level	Lower bounds I(0)	Upper bounds I(1)
1% level	10.150	11.230
5% level	7.135	7.980
10% level	5.915	6.630

Note: ARDL: autoregressive distributed lag; CUSUM: cumulative sum; CUSUMsq: cumulative sum of square; AIC: Akaike information criterion. Optimal lag length is determined by AIC. I(.) is the order of diagnostic tests. ^#Critical values are collected from the study by Narayan (2005).

*** Significance at 1% level.

** Significance at 5% level.

* Significance at 10% level.

Cointegration between the variables suggests examining the impact of tourism development, economic growth, and the REER on financial development in the long and short run. The results are reported in Table 4. The results reveal that tourism development has a positive and significant impact on financial development at the 1% level. This result shows that tourism development contributes to financial development. A 1% increase in tourism development leads to a 0.1931% increase in financial development, keeping other things constant.

Table 4.

Long-run and short-run analyses.

Dependent Variable = $ln F_{t}$
Variables	Long run		Short run
Variables	Coefficient	T-statistic	Coefficient	T-statistic
Constant	−11.46506	−2.6464	0.086638	2.642471
$ln {TD}_{t}$	0.1931*	2.7090	0.1222**	2.0826
$ln Y_{t}$	0.1400*	10.4840	−0.1224**	−2.1926
R_t	−0.0665*	−3.1424	0.2526	0.6055
$D_{1980}$	0.5869*	4.8041	0.0974***	1.7569
${ECM}_{t - 1}$	…	…	−0.1865**	−2.0778
R ²	0.8732		0.2188
Adjusted R²	0.8653		0.1269
F-statistic	10.4623*		3.3609**
D-W test	1.7224		1.8756
Diagnostic tests
Test	F-statistic	Probability	F-statistic	Probability
$χ^{2} NORMAL$	1.7554	0.4157	0.2789	0.6511
$χ^{2} SERIAL$	0.2685	0.7572	0.8290	0.4456
$χ^{2} ARCH$	0.1969	0.8814	0.0052	0.9425
$χ^{2} RAMSAY$	1.3765	0.1505	1.0160	0.3185

*** Significance at 1% level.

** Significance at 5% level.

* Significance at 10% level.

This empirical evidence is consistent with Kumar (2014), who asserted that tourism development strengthens the nexus between economic growth and tourism development and leads to financial development in Vietnam. Similarly, Shahbaz et al. (2017) reported that tourism stimulates economic activity and contributes to financial development. Conversely, Ohlan (2017) posited that tourism does not contribute to financial development and financial development does not contribute to tourism. Furthermore, Ridderstaat and Croes (2015) used money supply as an indicator of financial development and posited that financial development affects tourism demand asymmetrically, depending on financial development stages.

The relationship between economic growth and financial development is positive and statistically significant at the 1% level. With other things constant, a 1% increase in economic growth contributes to a 0.14% increase in financial development. This result implies that economic growth contributes to financial development by increasing the demand for financial services associated with an increase in the per capita income level. Similarly, Law and Habibullah (2009) assert that economic growth is a significant determinant of financial sector development if institutional quality is strong. The impact of the exchange rate on financial development is negative and significant at the 1% level. A 1% increase in the exchange rate would lower the financial development by 0.0665%, keeping other things constant.

This negative long-run relationship can be explained by the decision of Malaysian policymakers in July 2005 to replace the exchange rate peg with the US dollar by applying a managed float system. The Bank Negara Malaysia must promote stability of the exchange rate to sustain low inflation, rapid economic growth, and a higher domestic interest rate. Notably, the Asian crisis in 2011 resulted in the depreciation of the Malaysian Ringgit (MYR) by greater than 6%. These depreciations affected households and led them to pay higher prices for their goods and services. Firms were similarly affected by a surge in the costs of imported goods. The Malaysian economy was severely affected by the increase in inflation. These instabilities explain the negative link between the real exchange rate and financial development in Malaysia (e.g. Boyd et al., 2001; Cannonier and Burke, 2017).

The effect of the dummy variable on financial development is positive and dominant at the 1% significance level. This result shows that the implementation of financial and economic reforms in 1980s played a significant role in promoting the financial sector in Malaysia. In the long-run model, the R² is 0.8732. This result implies that 87.32% of the financial development is explained by tourism development, economic growth, the REER, and the dummy variable. Autocorrelation is absent, as confirmed by the Durbin Watson statistic. Overall, the long-run model is statistically significant at the 1% level. This result shows the robustness of the long-run model. Additionally, diagnostic analysis reveals the absence of serial correlation and autoregressive conditional heteroscedasticity: no specification problem in the long-run model, and the error term has normal distribution.

In the short run, the results reported in Table 4 reveal that tourism development has a positive and significant effect on financial development at the 5% level. Economic growth is negatively and significantly linked with financial development. The relationship between the REER and financial development is positive but statistically nonsignificant. The impact of the dummy variable is positive and significant at the 10% level. The sign of the ${ECM}_{t - 1}$ estimate is negative and statistically significant at the 1% level. This result reveals that the short-run adjustment toward the long-run equilibrium path is corrected by 18.65% in the finance demand function. Furthermore, the statistical significance of the error correction term is also validated in the established long run.

The diagnostic tests indicate the absence of serial correlation and autoregressive conditional heteroscedasticity effects. The relevance of the short-run model specification is confirmed by the Ramsay reset test. The normal distribution of the error term is also validated. The CUSUM and CUSUMsq tests corroborate the stability of the short- and long-run estimates. The results of CUSUM and CUSUMsq tests are shown in Figures 1 and 2. We observe that the graphs of the CUSUM and CUSUMsq tests are between the critical bounds at the 5% level. This result corroborates the stability and reliability of the long- and short-run estimates.

Figure 1.

CUSUM. CUSUM: cumulative sum.

Figure 2.

CUSUMsq. CUSUMsq: cumulative sum of square.

We subsequently apply the Toda and Yamamoto (1995) Granger causality test to examine the direction of causal relationship between financial development and its determinants. We also included the dummy variable based on the ZA unit root test to capture the structural break effect. The results of the Toda and Yamamoto (1995) Granger causality test in the presence of the structural breaks in the series are given in Table 5. The Toda–Yamamoto Granger causality analysis indicates that tourism activity triggers financial development and, in turn, financial development triggers tourism activities; this finding indicates the existence of the feedback effect and implies that financial development and tourism development are complementary to each other. This empirical evidence is contrary to the findings of Ohlan (2017), who asserted that financial development does not cause tourism development and tourism development does not cause financial development, implying a neutral effect. Similarly, Shahbaz et al. (2017) reported that financial development triggers tourism development.

Table 5.

Toda–Yamamoto non-Granger causality analysis.

Variable	$ln F_{t}$	$ln {TD}_{t}$	$ln Y_{t}$	R_t	Break year
$ln F_{t}$	…	2.6808** [0.0361]	1.2210 [0.3219]	3.8878* [0.0096]	1980
$ln {TD}_{t}$	2.5323** [0.0391]	…	0.4708 [0.7052]	2.6404** [0.0198]	2007
$ln Y_{t}$	2.6708** [0.0305]	2.8388** [0.0293]	…	3.6194** [0.0263]	1991
R_t	3.0946** [0.0443]	0.8264 [0.4913]	1.9401 [0.1479]	…	1986

*** Significance at 1% level.

** Significance at 5% level.

* Significance at 10% level.

The bidirectional Granger causality is also found between financial development and the REER. In other words, the REER Granger causes the extension of domestic credit to the private sector, and, in turn, the extension of domestic credit to the private sector Granger causes the REER. Fluctuations in the REER affect the future prospects of investment and extension of domestic credit to the private sector. In particular, when the local currency appreciates, imports become relatively cheaper, compared with exports (and vice versa) for households and firms.

The unidirectional Granger causality is found from tourism development to economic growth and implies a tourism-led growth hypothesis. The tourism-led growth hypothesis reveals the important role of tourism development in stimulating economic growth. This empirical evidence is similar to Tang and Tan (2015, 2017), but contrary to Shahbaz et al. (2017), who reported that tourism development and economic growth are interdependent.

Financial development leads to economic growth validating the supply-side hypothesis. This empirical result is consistent with that of Majid (2007), who found that financial development leads to economic growth, and confirms the supply-side hypothesis. The REER leads to economic growth. Even if the empirical macroeconomic literature has been equally uncertain regarding the causal relationship between the REER and economic growth (Levy-Yeyati and Sturzenegger, 2003), some recent empirical studies, in line with the seminal work of Obstfeld and Rogoff (1998), have found that exchange variability affects economic growth (Alagidede and Ibrahim, 2017; Kandill, 2004). Hence, this finding is consistent with the manner in which the REER variability leads to real output growth. In particular, this connection depends on the degree of openness and level of financial development. In the Malaysian context, the successive depreciation of the MYR and variability of the REER have led to an increase in the inflation rate, which has affected economic performance (Alagidede and Ibrahim, 2017).

The IAA and vector error correction model (VECM) Granger causality method are complementary, that is, the IAA provides the extent of the causal relationship between the positive and negative directions of the Granger causality among the variables, and the VECM Granger causality approach provides empirical findings for the causal association among the variables within the selected sample period. The IAA is a relevant choice to detect the causal association among the variables ahead of the sample period. The IAA consists of the variance decomposition and impulse response function. Pesaran and Shin (1999) argue that the generalized forecast error variance decomposition method shows how an innovative shock stemming from one variable contributes proportionally to other variables. This approach is superior to the orthogonalized forecast error variance decomposition method. In IAA, the VAR system determines the unique order of the variables because the generalized forecast error variance decomposition method is insensitive to the ordering of the variables. This approach examines the simultaneous shock effects.

Engle and Granger (1987) and Ibrahim (2000) have asserted that the generalized forecast error variance decomposition method produces reliable empirical results, compared with the traditional approaches within the VAR framework. Table 6 provides the empirical results provided by the variance decomposition approach. We observe that a 52.54% portion of financial development is led by its own innovative shocks. An innovative shock stemming from tourism development contributes to 19.40% of financial development. Economic growth and the REER contribute 20.73% and 7.32% to financial development, respectively. The innovative shocks from financial development contribute to tourism development by 26.40%. An overall 49.30% of tourism development is contributed by its own innovative shocks, 14.85% is contributed by other variables, and 9.43% is contributed by shocks stemming from the economic growth. The contribution of financial development to economic growth is 34.10%. Tourism development contributes 20% to economic growth, and this result confirms the tourism-led growth hypothesis. The innovative shocks stemming from the REER explain the economic growth by 21.37%. Financial development and tourism development contribute to the REER by 39.22% and 24.82%, respectively. The contribution of economic growth to the REER is 12.94%, and the remainder is contributed by innovative shocks stemming from the REER: that is, 22.99%.

Table 6.

Variance decomposition analysis.

Horizon	Variance decomposition of $ln F_{t}$				Variance decomposition of $ln {TD}_{t}$
Horizon	$ln F_{t}$	$ln {TD}_{t}$	$ln Y_{t}$	R_t	$ln F_{t}$	$ln {TD}_{t}$	$ln Y_{t}$	R_t
1	100.0000	0.0000	0.0000	0.0000	8.7553	91.2446	0.0000	0.0000
2	92.7310	0.0087	5.2219	2.0382	6.4861	89.1494	4.1936	0.1706
3	82.1269	4.37413	11.3781	2.1208	6.4633	88.3957	4.5115	0.6293
4	70.5603	15.4690	11.8028	2.1678	5.5733	78.8081	14.7836	0.8349
5	62.9442	18.2885	16.4450	2.3222	5.0602	72.6912	21.0548	1.1936
6	59.0130	18.5433	19.2941	3.1494	5.1771	72.8247	20.8191	1.1789
7	57.5089	18.9604	18.8147	4.7158	6.5084	71.7743	20.5102	1.2069
8	55.5672	20.4144	18.2132	5.8050	11.5926	66.8362	19.1645	2.4065
9	54.6634	21.1883	18.0104	6.1378	17.3647	60.8962	17.4327	4.3062
10	54.1367	20.9817	18.2502	6.6312	20.5251	57.2309	16.8715	5.3723
11	52.7668	20.4231	19.5703	7.2396	22.4475	54.7624	16.9012	5.8886
12	52.0542	20.1007	20.4063	7.4387	24.4761	52.5207	16.2452	6.7578
13	51.8483	19.9850	20.7472	7.4193	25.6654	50.7405	15.6945	7.8995
14	51.9471	19.7492	20.9446	7.3589	26.0980	49.8771	15.2560	8.7687
15	52.5421	19.4068	20.7236	7.3274	26.4089	49.3039	14.8505	9.4365
Horizon	Variance decomposition of $ln Y_{t}$				Variance decomposition of R_t
Horizon	$ln F_{t}$	$ln {TD}_{t}$	$ln Y_{t}$	R_t	$ln F_{t}$	$ln {TD}_{t}$	$ln Y_{t}$	R_t
1	6.0571	20.0199	73.9229	0.0000	0.3194	8.8127	22.5635	68.3041
2	7.1559	30.0355	56.2981	6.5103	0.3805	25.3654	19.1619	55.0921
3	9.7417	26.9775	43.0065	20.2741	11.2928	32.5406	14.6639	41.5025
4	11.5189	27.6336	34.0568	26.7905	25.2139	28.3224	11.8753	34.5883
5	12.8746	26.9702	30.4977	29.6573	32.4275	25.3919	10.1285	32.0520
6	14.0120	25.9364	28.4515	31.6000	36.5532	24.9670	9.4890	28.9907
7	15.4456	24.9634	27.8055	31.7853	40.7871	23.1972	9.7126	26.3029
8	17.2374	23.6590	27.6440	31.4594	43.5081	21.9319	9.7828	24.7771
9	19.5585	22.7055	26.9875	30.7483	43.6616	22.1297	10.1870	24.0216
10	22.8156	21.4861	26.0690	29.6290	41.9951	23.5457	11.4871	22.9719
11	26.4123	20.4807	24.5564	28.5505	40.6212	24.3747	12.7601	22.2438
12	29.3002	20.0282	23.3109	27.3606	40.0235	24.5828	13.1358	22.2577
13	31.6512	19.6802	22.4404	26.2280	39.6189	24.6794	13.0629	22.6386
14	33.3248	19.5698	21.8321	25.2731	39.3675	24.7916	12.9641	22.8766
15	34.1011	19.8063	21.3772	24.7153	39.2255	24.8256	12.9496	22.9991

Overall, the empirical results indicate the bidirectional Granger causality between tourism development and financial development: that is, tourism development and financial development are interdependent. Financial development leads to economic growth and, in turn, economic growth triggers financial development. Tourism development and the REER lead to economic growth. The REER leads to financial development and economic growth.

The impulse response function (Figure 3) is an alternative to the variance decomposition method. The impulse response function explains how long and to what extent financial development reacts to a shock stemming from tourism development, economic growth, and the REER. The empirical results are reported in Figure 3, and we observe that financial development initially increases and subsequently starts to decline because of a forecast error from tourism development, indicating an inverted U-shaped relationship between tourism development and financial development. Financial development responds positively due to a forecast error in economic growth. The contribution of the REER to financial development is positive initially, but becomes negative after the eighth time horizon. Tourism development responds positively after the fourth time horizon as a result of the forecast error from financial development. Economic growth contributes positively to tourism development. The response of tourism development is negative due to a forecast error in the REER.

Figure 3.

Response to generalized 1 SD innovations ± 2 SE.

Conclusion and policy implications

This study investigates the relationship between financial and tourism development by employing the finance demand function and considering economic growth and the REER as additional determinants of financial and tourism development. For empirical purposes, we applied the ZA unit root test containing information on the single unknown structural break in a time series. The presence of cointegration among financial development, tourism development, economic growth, and the REER is investigated by applying the ARDL bounds testing approach, which accommodates structural breaks. The causal association between the variables is examined by applying the Toda–Yamamoto Granger causality. Additionally, the robustness of the causal relationship is investigated by applying the IAA.

The empirical results indicate the presence of cointegration between the variables in the presence of structural breaks in the series. Furthermore, tourism development adds to financial development. Economic growth positively affects financial development. The relationship between the REER and financial development is negative. The Granger causality analysis reveals the feedback effect between tourism development and financial development. Tourism development leads to economic growth, and vice versa. The bidirectional Granger causality is also observed between financial development and economic growth. The REER leads to tourism development, financial development, and economic growth.

The empirical evidence obtained in this study may have important implications for Malaysian policy makers. The findings underscore that the Malaysian government should pursue its economic transformation in two main directions. First, the Economic Transformation Program, launched on September 25, 2010, formulated a few measures in favor of financial services and tourism, based on the premise that these two economic sectors are interrelated. The primary goal of this program is to increase the national revenue per capita of $15,000 and create approximately 3 million new jobs by 2020. To achieve this goal, efficient financial services are required to ensure the following: (i) the development of tourism products and services and (ii) the diversion of new private and public investments and foreign direct investments toward infrastructure, technology, logistic, and management developments. By implementing these initiatives, the Malaysian government may ensure sustainable tourism management and the realization of enhanced economic growth and employment creation by 2020. Second, the quality of tourism offerings should be enhanced to attract, for example, tourists from China. Following several European countries’ examples (e.g. France, Italy, and Spain), the Malaysian Minister of Tourism and Culture should initiate a series of initiatives that promote alternative tourism packages, such as green tourism, gastronomy, heritage, and traditions, through the diversity of the local culture. Our results also confirm that the initiatives concerning the development of tourism and financial services proposed by the Malaysian government should help to ensure that the Malaysian economy achieves its targeted level of development planned for 2020.

Supplemental material

Appendix - Tourism-induced financial development in Malaysia: New evidence from the tourism development index

Appendix for Tourism-induced financial development in Malaysia: New evidence from the tourism development index by Muhammad Shahbaz, Ramzi Benkraiem, Anthony Miloudi and Aviral Kumar Tiwari in Tourism Economics

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental material

Supplemental material for this article is available online.

Notes

References

Aghion

Bacchetta

Rancière

. (2009) Exchange rate volatility and productivity growth: the role of financial development. Journal of Monetary Economics 56: 494–513.

Alagidede

Ibrahim

(2017) On the causes and effects of exchange rate volatility on economic growth: evidence from Ghana. Journal of African Business 18(2): 169–193.

Ang

(2008) What are the mechanisms linking financial development and economic growth in Malaysia? Economic Modelling 25: 38–53.

Arslanturk

Balcilar

Ozdemir

(2011) Time-varying linkages between tourism receipts and economic growth in a small open economy. Economic Modelling 28: 664–671.

Balaguer

Cantavella-Jorda

(2002) Tourism as a long-run economic growth factor: the Spanish case. Applied Economics 34(7): 877–884.

Başarir

Çakir

(2015) Causal interactions between CO₂ emissions, financial development, energy and tourism. Asian Economic and Financial Review 5(11): 1227–1238.

Boyd

Levine

Smith

(2001) The impact of inflation on financial sector performance. Journal of Monetary Economics 47: 221–248.

Brida

Risso

(2009) Tourism as a factor of long-run economic growth: an empirical analysis for Chile. European Journal of Tourism Research 2(2): 178–185.

Brida

Carrera

Risso

(2008) Tourism’s impact on long-run Mexican economic growth. Economics Bulletin 3: 1–8.

10.

Brida

Lanzilotta

Pereyra

. (2015) A nonlinear approach to the tourism-led growth hypothesis: the case of MERCOSUR. Current Issues in Tourism 18(7): 647–666.

11.

Brida

Punzo

Risso

(2011) Tourism as a factor of growth: the case of Brazil. Tourism Economics 17(6): 1375–1386.

12.

Cannonier

Burke

(2017) Tourism and financial development in small states: evidence from Caribbean countries. Tourism Economics 23: 1–9.

13.

Chao

Hazari

Laffargue

YES

(2009) A dynamic model of tourism, employment and welfare: the case of Hong Kong. Pacific Economic Review 14(2): 232–245.

14.

Chao

Hazari

Laffargue

. (2006) Tourism, Dutch disease and welfare in an open dynamic economy. The Japanese Economic Review 57(4): 501–515.

15.

Chou

(2013) Does tourism development promote economic growth in transitions countries? A panel data analysis. Economic Modelling 33: 226–232.

16.

Dickey

Fuller

(1981) Likelihood ratio statistics for autoregressive time series with a unit root. Econometrica 49: 1057–1072.

17.

Dritsakis

(2004) Tourism as a long-run economic growth factor: an empirical investigation for Greece using causality analysis. Tourism Economics 10(3): 305–316.

18.

Durbarry

(2002) The economic contribution of tourism in Mauritius. Annals of Tourism Research 29(3): 862–865.

19.

Dwyer

Forsyth

Spurr

(2004) Evaluating tourism’s economic effects: new and old approaches. Tourism Management 25: 307–317.

20.

Engle

Granger

CWJ

(1987) Co-integration and error correction: representation, estimation and testing. Econometrica 55: 251–276.

21.

Ertugrul

Mangir

(2015) The tourism-led growth hypothesis: empirical evidence from Turkey. Current Issues in Tourism 18(7): 633–646.

22.

Gunduz

Hatemi-J

(2005) Is the tourism-led growth hypothesis valid for Turkey? Applied Economics Letters 12(8): 499–504.

23.

Zheng

(2011) Empirical analysis on the relationship between tourism development and economic growth in Sichuan. Journal of Agricultural Science 3(1): 212–217.

24.

Ibrahim

(2000) Cointegration and Granger Causality Tests of Stock Price and Exchange Rate Interactions In Malaysia. ASEAN Economic Bulletin 17(1): 36–47.

25.

Jalil

Mahmood

Idrees

(2013) Tourism-growth nexus in Pakistan: evidence from ARDL bounds tests. Economic Modeling 35: 185–191.

26.

Johansen

Juselius

(1990) Maximum likelihood estimation and inference on cointegration, with applications to the demand for money. Oxford Bulletin of Economics and Statistics 52: 169–210.

27.

Kadir

Nayan

Abdullah

(2010) The causal relationship between tourism and economic growth in Malaysia: evidence from multivariate causality tests. Tourism and Management Studies 6: 16–24.

28.

Kandill

(2004) Exchange rate fluctuations and economic activity in developing countries: theory and evidence. Journal of Economic Development 29(1): 85–108.

29.

Katircioglu

(2009) Revisiting the tourism-led-growth hypothesis for Turkey using the bounds test and Johansen approach for cointegration. Tourism Management 30(1): 17–20.

30.

Kim

Chen

Jang

(2006) Tourism expansion and economic development: the case of Taiwan. Tourism Management 27: 925–933.

31.

King

Levine

(1993) Finance and growth: Schumpeter might be right. Quarterly Journal of Economics 108: 717–737.

32.

Kumar

(2014) Exploring the role of technology, tourism and financial development: an empirical study of Vietnam. Quality and Quantity 48(5): 2881–2898.

33.

Kumar

(2013) Exploring the developments in urbanization, aid dependency, sectorial shifts and services sector expansion in Fiji: a modern growth perspective. Global Business and Economics Review 15: 371–395.

34.

Kumar

Loganathan

Patel

. (2015) Nexus between tourism earnings and economic growth: a study of Malaysia. Quality and Quantity 49(3): 1101–1120.

35.

Lau

(2009) Tourist arrivals and economic growth in Sarawak. Sarawak Development Journal 9: 36–45.

36.

Law

Habibullah

(2009) The determinants of financial development: institutions, openness and financial liberalization. South African Journal of Economics 77: 45–58.

37.

Lee

Kwon

(1995) Importance of secondary impact of foreign tourism receipts on the South Korean economy. Journal of Travel Research 34: 50–54.

38.

Lee

(2012) Tourism, trade, and income: evidence from Singapore. Anatolia 23(3): 348–358.

39.

Lee

Chang

(2008) Tourism development and economic growth: a closer look at panels. Tourism Management 29: 180–192.

40.

Lee

Chien

(2008) Structural breaks, tourism development, and economic growth: evidence from Taiwan. Mathematics and Computers in Simulation 77: 358–368.

41.

Levine

(2003) More on finance and growth: more finance, more growth? Review, Federal Reserve Bank of Saint Louis 85(4): 31–46.

42.

Levy-Yeyati

Sturzenegger

(2003) To float or to fix: evidence on the impact of exchange rate regimes on growth. American Economic Review 93(4): 1173–1193.

43.

Majid

(2007) Does financial development cause economic growth in the ASEAN-4 countries? Savings and Development 31: 369–398.

44.

Narayan

(2005) The saving and investment nexus for China: evidence from cointegration tests. Applied Economics 37: 1979–1990.

45.

Narayan

Prasad

. (2010) Tourism, and economic growth: a panel data analysis for Pacific Island countries. Tourism Economics 16(1): 169–183.

46.

Obstfeld

Rogoff

(1998) Risk and exchange rates. NBER Research Working Papers Series, WP 6694.

47.

(2005) The contribution of tourism development to economic growth in the Korea economy. Tourism Management 26(1): 39–44.

48.

Ohlan

(2017) The relationship between tourism, financial development and economic growth in India. Future Business Journal 3: 9–22.

49.

Ongan

Demiroz

(2005) The contribution of tourism to the long-run Turkish economic growth. Journal of Economics 53: 880–894.

50.

Payne

Mervar

(2010) The tourism growth nexus in Croatia. Tourism Economics 16(4): 1089–1094.

51.

Pesaran

Shin

(1999) An autoregressive distributed lag modelling approach to cointegration analysis. In: Strom

(ed.), Econometrics and Economic Theory in the 20th Century: The Ragnar Frisch Centennial Symposium. Cambridge: Cambridge University Press, pp. 371–413.

52.

Pesaran

Shin

Smith

(2001) Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics 16: 289–326.

53.

Phillips

Hansen

(1990) Statistical inference in instrumental variables regression with i(1) processes. Review of Economic Studies 57: 99–125.

54.

Phillips

PCB

Perron

(1988) Testing for a unit root in time series regression. Biometrika 75: 335–346.

55.

Ridderstaat

Croes

(2015) The link between money supply and tourism demand cycles: a case study of two Caribbean destinations. Journal of Travel Research 14: 187–205.

56.

Salmani

Panahi

Razzaghi

(2014) Assessing the dynamic economic impact of tourism for OIC members. World Applied Sciences Journal 32(6): 1098–1105.

57.

Shahbaz

(2009) A reassessment of finance-growth nexus for Pakistan: under the investigation of FMOLS and DOLS techniques. The IUP Journal of Applied Economics 8(1): 65–80.

58.

Shahbaz

Kumar

Ivanov

. (2017) The nexus between tourism demand and output per capita, with the relative importance of trade openness and financial development: a study of Malaysia. Tourism Economics 23(1): 168–186.

59.

Singh

(2008) Small island developing states (SIDS): tourism and economic development. Tourism Analysis 13(5-6): 629–636.

60.

Tang

Abosedra

(2016) Tourism and growth in Lebanon: new evidence from bootstrap simulation and rolling causality approaches. Empirical Economics 50(2): 679–696.

61.

Tang

Tan

(2015) Does tourism effectively stimulate Malaysia’s economic growth? Tourism Management 46: 158–163.

62.

Tang

Tan

(2017) The relevance of the tourism-led growth hypothesis to Malaysia: A view through rolling-samples and disaggregated tourism markets in Tourism and Opportunities for Economic Development in Asia 188-202, IGI Global USA.

63.

Tang

Jang

(2009) The tourism–economy causality in the United States: a sub industry level examination. Tourism Management 30(4): 553–558.

64.

Toda

Yamamoto

(1995) Statistical inference in vector autoregressions with possible integrated processes. Journal of Econometrics 66: 225–250.

65.

Zivot

Andrews

(1992) Further evidence on the great crash, the oil price shock, and the unit root hypothesis. Journal of Business and Economic Statistics 10: 251–270.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.15 MB