The link between tourism activities and economic growth: Evidence from China’s provinces

Abstract

This study explores the causal relationship between international tourism receipts and economic growth in China’s eight central provinces (Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan) by analyzing these provinces for the period from 1995 to 2014, accounting for both dependency and heterogeneity across provinces. The empirical results of this study support evidence for the growth hypothesis in Hunan Province. A reverse relationship supports evidence on the conservation hypothesis for the provinces, such as Shanxi, Jiangxi, and Henan. A reciprocal causal relationship was found for Jilin, Anhui, and Hubei, while the result of a neutrality hypothesis supported only one of the provinces in Heilongjiang. The empirical findings of this study provide important policy implications for China’s eight central provinces.

Keywords

China’s central provinces dependency and heterogeneity economic growth international tourism receipts panel causality test

Introduction

Arslanturk et al. (2011) indicate that, over the past several decades, international tourism has been increasingly becoming important, and the tourism industry has begun to play an important role in the economies of many countries. Tourism is perceived as increasing overall economic activities, and this increase in the activity is generally considered to be desirable; namely, the positive impact of tourism on economic activities is frequently described. Tourism is recognized to have a positive effect on long-run economic growth through different channels. First, tourism is a type of foreign exchange earner in terms of inbound tourism that attracts foreign tourists who consume the products of local markets and businesses, allowing the country to pay for imported capital goods or basic inputs used in the production process. Second, it motivates governments and firms to invest in the new infrastructure and business environment, fostering the ability of local firms to compete with firms in other tourist countries. Third, it stimulates other economic industries through the direct, indirect, induced, and spillover effects. Fourth, it contributes to the employment generation and income increase. Fifth, it can result in the positive exploitation of economies of scales in national firms. Finally, it is an important factor in the diffusion of knowledge, stimulating learning, and the accumulation of human capital. Also, tourism has become a common developmental focus in many countries (see Andriotis, 2002; Fagance, 1999; Lin and Liu, 2000; Schubert et al., 2011).

In general, tourism development has been considered to be a positive contributor to economic growth. However, there is an unverified question of whether tourism development actually causes economic growth or, alternatively, whether economic expansion substantially contributes to growth in tourism. This study represents our attempt to revisit this issue using the data from 31 major provinces in China. As we know that an outward-oriented growth strategy is common among Asian countries, international trade, especially net commodity exports, has been viewed as the “engine of growth”. A centerpiece of the neoliberal strategy of outward-oriented development in many Asian countries has been considered to be the promotion of new growth sectors such as tourism or nontraditional exports (Theobald, 2001). The Asia-Pacific Region has become a rapidly growing tourism destination and has even surpassed the United States to become the world’s second largest tourist-receiving region since 2001 (Brida et al., 2016; Caglayan et al., 2012; Chen and Chiou-Wei, 2009; Katircioglu, 2010; Lee and Chien, 2008; Surugiu and Surugiu, 2013; Tang and Tan, 2015). It should be noted that although there has generally been an upward trend in international tourism in recent years, not all Asian regions and countries which have shared economic growth are equal. Tourism has been promoted in many Asian countries as part of the solution to their economic problems. Also, it has been seen as an important source of foreign exchange earnings, the employment of domestic labor and a contributor to economic growth.

As noted by Oh (2005), the recognition of a causal relationship between international tourism and economic growth will have important implications for the development of different tourism marketing and policy decisions in Asian countries. For instance, if there is an unambiguously unidirectional causality from tourism development to economic growth, tourism-led economic growth would be practical. If the results indicate the opposite causal direction, economic development may be necessary for the expansion of the tourism industry. Next, if the causative process is bi-directional, and tourism growth and economic growth have a reciprocal causal relationship, then efforts in both areas would be beneficial. Finally, if there is no causal relationship between tourism growth and economic development, such as the enthusiastic promotion of tourism which may not be as effective as what tourism managers and decision-makers currently believe.

Moreover, analyzing the relationship between economic growth and tourism development has been a popular topic in the recent tourism literature (Arslanturk et al., 2011; Biagi et al., 2015; Brida et al., 2016; Caglayan et al., 2012; Chen and Chiou-Wei, 2009; Corteś-Jimeńez, 2008; Katircioglu, 2010; Kim et al., 2006; Lee and Chien, 2008; Nunkoo, 2015; Santana-Gallego et al., 2010; Surugiu and Surugiu, 2013; Tang and Tan, 2015). Most previous studies have concentrated on one specific country or area. Because of different methods and data periods, it is difficult to compare the results and findings between those studies. For this reason, this study adopts longitudinal data of multi-countries in order to fill in the gap. We employ recently developed panel causality methods to offer compatible findings in those chosen countries or areas.

It is important for policy-makers to identify the nature of causal linkages between economic growth and tourism consumption. From a theoretical perspective (Balaguer and Cantavella-Jorda, 2002; Balassa, 1978; Brida and Risso, 2010; Copeland, 1991; Corden and Neary, 1982; Dissart et al., 2015; Figini and Vici, 2010; Narayan, 2004; Nowak et al., 2007; Song et al., 2012; Tang, 2013; Tang and Tan, 2015), it has been argued that in the existing literature, there are four hypotheses to exhibit the relationship between tourism receipts and economic growth, including growth hypothesis, conservation hypothesis, feedback hypothesis and neutrality hypothesis. Before making the proper government policies, authorities need to identify the relationship between tourism and economic growth to make an appropriate strategy. The growth hypothesis means that the tourism activities will positively contribute to economic growth. In this situation, the government should put more efforts into refining the infrastructure, such as transportation, public security, and accommodation.

More tourism resources investment is needed in order to attract local and foreign tourists. On the contrary, the conservation hypothesis means that people will spend more money on tourism when their economy becomes better. In this way, the government should promote economic growth in terms of the first priority policy. Economic growth then has a positive effect on the tourism industry. The feedback hypothesis means that there is a reciprocal effect between tourism consumption and economic growth. In this case, the government focusing on both tourism and economic development will have reciprocal benefits. The neutrality hypothesis means that there are no spillover effects between tourism and economic development. Policy-makers should find other strategies to promote economic growth (Lee and Chien, 2008; Oh, 2005).

There are several reasons why the experience of Chinese international tourism receipts (ITR) is of interest. First, in recent years it has witnessed strong growth of international tourist arrivals in the Asia-Pacific Region, specifically in China. In 2014, China ranked third in inbound tourism arrivals and fourth in inbound tourism expenditure proposed by the United Nations World Tourism Organization (UNWTO) (2014). Given the importance of inbound tourism in China, it is important to estimate its potential contribution to China’s economy. Second, China’s regional economies have become widely diversified owing to various socio-economic, economic geographical and political factors. Given the diversified structure of regional economies within which the Chinese tourism industry operates, it serves as an ideal case study to test whether ITR has the same impact on economic growth in regions that differ significantly in their degree of tourism dependence. Third, most previous studies have concentrated on one specific country or area. Because of different methods and data periods, it is difficult to compare the results and findings between those studies. For this reason, this study adopts panel data of multi provinces in order to fill in the gap. We employ recently developed panel causality methods to offer compatible findings in China’s eight central provinces.

As indicated by Oh (2005), this work investigates the causal relationship between tourism activities and economic expansion in China’s eight central provinces using the province-specific causality test developed by Kónya (2006) to determine the dynamic and causal relationship between growth in the tourism sector and the economic growth. This procedure will undoubtedly allow the effects of specific provinces to be more readily uncovered. We examine whether there is any causal relationship between growth in the tourism activity and economic growth using a bootstrap panel Granger causality test for a sample of China’s eight provinces over the period from 1995 to 2014. This is the first study that employs a bootstrap panel Granger causality test to study the relationship between ITR and economic growth in China’s eight central provinces. This study is expected to fill the gap in the current literature on growth in the tourism sector and economic growth.

Literature review

It is an important issue to study the casual relationship between tourism activities and economic growth over the past few years (Arslanturk et al., 2011; Kim et al., 2006). The policy-makers desiring to create the wealth in a country are concerned not only on the overall industrial development but also the appropriate resource.

When the tourism industry booms, it will create employment opportunities and increase civil and government income. Many researchers argue that the positive relationship exists between tourism and economic growth (Khan et al., 1995; Lee and Kwon, 1995). Previous studies have not offered a consistent response to the explanation of the relationship in different countries. In this study, we explore the causality between tourism and economic growth. There are discrepancy explanations in the tourism industry leading economic growth or economic growth leading tourism industry in the previous studies, and both of them have some theoretical explanations.

Tourism-led growth means that tourism plays an active role in an economy. By attracting more tourists to consume in a region, tourism causes catalytic reactions and spillover effects in economic growth. Because of tourism development, more jobs have been created and many people engaging in tourism activities have more income to consume. Furthermore, tourism income tax increases government income which affords government expenditure. Both civil consumption and government expenditure create the multiplier effect in economic growth. In sum, the tourism industry has a positive effect on economic growth (Khan et al., 1995; Lee and Kwon, 1995; Oh, 2005).

From a macroeconomics formula: y = c + i + g+ (x − m), all civil consumption, business investment, government expenditure, and net export may increase when the tourism industry is booming in a country. Subsequently, tourism development contributes to the national income. As a result, the economic growth increases. Balaguer and Cantavella-Jorda (2002) examine the relationship between tourism and long-term economic growth in Spain. The results indicate that tourism influences the economic growth. Gunduz and Hatemi (2005) find the one-way causality from the tourism industry to economic growth. Dritsakis (2004), Durbarry (2004) and Oh (2005) support the aforementioned results.

On the contrary, economic development positively drives tourism growth in a country. According to Engle’s Law, as income rises, the proportion of income spent on food falls, even if actual expenditure on food rises. More income urges people to increase the percentage of consumption in entertainment, cultural, and education. Oh (2005) finds that because of strong economic growth in South Korea, many foreign businessmen travel and work in South Korea. Based on the inductive logic, Oh (2005) argues that this result supports the economic growth which will influence the tourism industry. The research findings of Lanza et al. (2003) also support this perspective. Oh (2005) indicates that when the ratio of gross domestic product (GDP) in tourism is low, more probability will exist within the economic-driven tourism growth.

Several researchers have found that the reciprocal relationship exists between tourism and economic growth. Kim et al. (2006) indicate that there is a long-run equilibrium between the tourism and economic growth in Taiwan. Tourism and economic growth reciprocally affect each other. Chen and Chiou-Wei (2009) apply the Taiwan and South Korea data, finding that Taiwan is tourism-led economic growth and Korea is a reciprocal relationship between tourism and economic growth. Dritsakis (2004) and Durbarry (2004) find a reciprocal relationship in Greece and Mauritius. Furthermore, Arslanturk et al. (2011) indicate that there is no relationship between tourism and economic growth. In addition, they employed the Granger causality test from 1963 to 2006, revealing that the relationship between tourism and economic growth did not exist.

Oh (2005) argues that so far there are not enough research findings to clearly identify the relationship between tourism activities and economic growth. It may be because of the limitations of the data and methodology. Most of previous researchers have focused on one or few countries in a study. Even focusing on the same country, the studied data period may be different. Furthermore, different studies employ different statistical methods to research the relationship between tourism and economic growth. It is very difficult to compare different results of previous studies. In order to fill the gap in this study, this work applies one specific statistical method. Namely, the Granger panel causality approach examines both cross-state interrelations and province-specific heterogeneity. The modeling approach applied in this study can be considered to be a systematic way to detect causal linkages in the panel data framework. In this way, some light could be shed on this interesting issue of tourism activities and economic growth.

Methodology

Testing cross-sectional dependence

An important issue in panel causality analysis is to account for possible cross-sectional dependence across countries. This issue is noted because high levels of globalization, international trade, and financial integration enable a country to be sensitive to economic shocks in other countries. Importantly, the results of cross-sectional dependency in substantial bias and size distortions, implying that testing cross-sectional dependence is a crucial step in a panel data analysis, are ignored (Pesaran, 2006).

To test cross-sectional dependency, the Lagrange Multiplier (LM hereafter) developed by the Breusch and Pagan (1980) test has been extensively used in empirical studies. The procedure to compute the LM test requires the estimation of the following panel data model

\begin{matrix} y_{it} = α_{i} + β'_{i} x_{it} + u_{it} for i = 1, 2, \dots, N; \\ t = 1, 2, \dots, T \end{matrix}

(1)

In equation (1), $y_{it}$ is real GDP (RGDP), i is the cross-sectional dimension, t is the time dimension, $x_{it}$ is $k \times 1$ vector of explanatory variables (such as RGDP and ITR), and $α_{i}$ and $β_{i}$ are the individual intercepts and slope coefficients that are allowed to vary across provinces. In the LM test, the null hypothesis of no-cross sectional dependence, $H_{0} : Cov (u_{it}, u_{jt}) = 0$ for all t and $i \neq j$ is tested against the alternative hypothesis of cross-sectional dependence, $H_{1} : Cov (u_{it}, u_{jt}) \neq 0$ , for at least one pair of $i \neq j$ . To test the null hypothesis, Breusch and Pagan (1980) develop the LM test as follows

LM = T \sum_{i = 1}^{N - 1} \sum_{j = i + 1}^{N} {\hat{ρ}}_{ij}^{2}

(2)

where

{\hat{ρ}}_{ij}

is the sample estimate of the pair-wise correlation of the residuals from an Ordinary Least Squares (OLS) estimation of equation (1) for each i. Under the null hypothesis, the LM statistic has an asymptotic chi-square with

N (N - 1) / 2

degrees of freedom. It is important to note that the LM test is valid when N is relatively small and T is sufficiently large.

Pesaran (2004) proposes a cross-sectional dependency test (called the CD test) for panel data models where T→∞ and N→∞ in any order. However, the CD test is subject to decreasing power in certain situations where the population average pair-wise correlations are zero although the underlying individual population pair-wise correlations are non-zero (Pesaran et al., 2008). Furthermore, in stationary dynamic panel data models the CD test fails to reject the null hypothesis when the factor loadings have zero mean in the cross-sectional dimension (Sarafidis et al., 2009). To address these problems, Pesaran et al. (2008) propose a bias-adjusted test, which is a modified version of the LM test, using the exact mean and variance of the LM statistic. The bias-adjusted LM test is as follows

{LM}_{adj} = \sqrt{(\frac{2 T}{N (N - 1)})} \sum_{i = 1}^{N - 1} \sum_{j = i + 1}^{N} {\hat{ρ}}_{ij} \frac{(T - k) ρ_{ij}^{2} - μ_{Tij}}{\sqrt{ν_{Tij}^{2}}}

(3)

where

μ_{Tij}

and

ν_{Tij}^{2}

are the exact mean and variance of

(T - k) ρ_{ij}^{2}

, which are provided by Pesaran et al. (2008). Under the null hypothesis where first T→∞ and then N→∞, the

{LM}_{adj}

test is asymptotic standard normal distribution.

Testing slope homogeneity

The second issue in panel data analysis is to decide whether the slope coefficients are homogeneous. Causality running from one variable to another due to the imposition of a joint restriction for the whole panel is the strong null hypothesis (Granger, 2003). Moreover, the homogeneity assumption for the parameters is unable to capture heterogeneity due to country-specific characteristics (Breitung, 2005).

The most common method to test the null hypothesis of slope homogeneity— $H_{0} : β_{i} = β$ for all i-against the hypothesis of heterogeneity— $H_{1} : β_{i} \neq β_{j}$ for a non-zero fraction of pair-wise slopes for $i \neq j$ − is to apply the standard F test. The F test is valid for cases where the cross-sectional dimension (N) is relatively small, and the time dimension (T) of the panel is large, the explanatory variables are strictly exogenous, and the error variances are homoscedastic. Through the relaxation of the homoscedasticity assumption in the F test, Swamy (1970) develops the slope homogeneity test on the dispersion of individual slope estimates from a suitable pooled estimator. However, both the F and Swamy tests require panel data models where N is small relative to T (Pesaran and Yamagata, 2008). Pesaran and Yamagata (2008) propose a standardized version of the Swamy test (called $\tilde{Δ}$ test) for testing slope homogeneity in large panels. The $\tilde{Δ}$ test is valid as $(N, T) \to \infty$ without any restrictions on the relative expansion rates of N and T when the error terms are normally distributed. In the $\tilde{Δ}$ test approach, the first step is to compute the following modified version of the Swamy test

\tilde{S} = \sum_{i = 1}^{N} (β_{i} - {\tilde{β}}_{WFE})' \frac{x'_{i} M_{τ} x_{i}}{{\tilde{σ}}_{i}^{2}} (β_{i} - {\tilde{β}}_{WFE})

(4)

where

β_{i}

is the pooled OLS estimator,

{\tilde{β}}_{WFE}

is the weighted fixed effect pooled estimator,

M_{τ}

is an identity matrix while the

{\tilde{σ}}_{i}^{2}

is the estimator of

σ_{i}^{2}

. Thus, the standardized dispersion statistic is developed as follows

\tilde{Δ} = \sqrt{N} (\frac{N^{- 1} \tilde{S} - k}{\sqrt{2 k}})

(5)

Under the null hypothesis with the condition of $(N, T) \to \infty$ , as long as $\sqrt{N} / T \to \infty$ and the error terms are normally distributed, the $\tilde{Δ}$ test has asymptotic standard normal distribution. The small sample properties of $\tilde{Δ}$ test can be improved under the normally distributed errors using the following bias-adjusted version

{\tilde{Δ}}_{adj} = \sqrt{N} (\frac{N^{- 1} \tilde{S} - E ({\tilde{z}}_{it})}{\sqrt{var ({\tilde{z}}_{it})}})

(6)

where the mean is

E ({\tilde{z}}_{it}) = k

while the variance is

var ({\tilde{z}}_{it}) = 2 k (T - k - 1) / T + 1

Panel Granger causality analysis

The presence of both cross-sectional dependency and heterogeneity across China’s eight central provinces requires a panel Granger causality method that can account for such dynamics. The bootstrap panel Granger causality approach proposed by Kónya (2006) is able to account for both cross-sectional dependence and provinces-specific heterogeneity. This approach is based on the seemingly unrelated regression (SUR) estimation of the set of equations and Wald tests with individual, country-specific bootstrap critical values.

The system to be estimated in the bootstrap panel Granger causality approach can be expressed as follows

\begin{matrix} \begin{matrix} y_{1, t} = α_{1, 1} + \sum_{i = 1}^{{ly}_{1}} β_{1, 1, i} y_{1, t - i} + \sum_{i = 1}^{{lx}_{1}} δ_{1, 1, i} x_{1, t - i} + ɛ_{1, 1, t} \\ y_{2, t} = α_{1, 2} + \sum_{i = 1}^{{ly}_{1}} β_{1, 2, i} y_{2, t - i} + \sum_{i = 1}^{{lx}_{1}} δ_{1, 2, i} x_{2, t - i} + ɛ_{1, 2, t} \\ : \\ y_{N, t} = α_{1, N} + \sum_{i = 1}^{{ly}_{1}} β_{1, N, i} y_{N, t - i} + \sum_{i = 1}^{{lx}_{1}} δ_{1, N, i} x_{1, N, t - i} + ɛ_{1, N, t} \end{matrix} \end{matrix}

(7)

\begin{matrix} \begin{matrix} x_{1, t} = α_{2, 1} + \sum_{i = 1}^{{ly}_{2}} β_{2, 1, i} y_{1, t - i} + \sum_{i = 1}^{{lx}_{2}} δ_{2, 1, i} x_{1, t - i} + ɛ_{2, 1, t} \\ x_{2, t} = α_{2, 2} + \sum_{i = 1}^{{ly}_{2}} β_{2, 2, i} y_{2, t - i} + \sum_{i = 1}^{{lx}_{2}} δ_{2, 2, i} x_{2, t - i} + ɛ_{2, 2, t} \\ : \\ x_{N, t} = α_{2, N} + \sum_{i = 1}^{{ly}_{2}} β_{2, N, i} y_{N, t - i} + \sum_{i = 1}^{{lx}_{2}} δ_{2, N, i} x_{N, t - i} + ɛ_{2, N, t} \end{matrix} \end{matrix}

(8)

Where y denotes RGDP, x refers to ITR, and l is the lag length. As each equation in this system has different predetermined variables while the error terms may be contemporaneously correlated (i.e. cross-sectional dependency), these sets of equations derive from the SUR system.

In the bootstrap panel causality approach, there are alternative causal linkages for a country in the system such that (i) there is one-way Granger causality from x to y, if not all $δ_{1, i}$ are zero, but all $β_{2, i}$ are zero; (ii) there is one-way Granger causality running from y to x if all $δ_{1, i}$ are zero, but not all $β_{2, i}$ are zero; (iii) there is two-way Granger causality between x and y, if neither $δ_{1, i}$ nor $β_{2, i}$ are zero; and (iv) there is no Granger causality between x and y, if all $δ_{1, i}$ and $β_{2, i}$ are zero.

Data collection

Most of previous studies have used the government’s tourism and economic statistical data which accurately analyze the specific purpose. Authoritativeness and correctness are the main advantages of using the government’s data. However, the disadvantage of using the data is that the definitions of the variables may be different in different countries. It may mislead the result when researchers try to compare different empirical results in different provinces. In this study, we used the tourism statistics of National Bureau of Statistics (NBS)¹ of the People’s Republic of China. Annual data of ITR² index and RGDP index in China’s eight central provinces from 1995 to 2014 were adopted from the official website of NBS. NBS has collected and announced the tourism statistics since 1995. An obvious advantage of using the NBS database in our study is that all variables derive from the same origin and all of them have the same operational definition. NBS’s long period and multi-provincial data also allowed researchers to employ longitudinal analysis methods. The annual data employed in this study included the period from 1995 to 2014 for China’s eight regions or provinces (i.e. Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan).

The variables in this study include ITR and RGDP. Foreign travel and tourism receipts, including the receipts of residents for both business and leisure trips within a national region, are expressed in terms of millions of US dollars. The data derived from the NBS for the period from 1995 to 2014. RGDP is measured in constant 2005 US dollars, and the source of the data is the NBS Database. Figure 1 plots the ITR index versus RGDP index across China’s eight central provinces while the descriptive statistics are provided in Tables 1 and 2. In these provinces, Anhui and Hubei had the highest and lowest mean RGDP values: US$9,678.374 and US$5,212.402 in these provinces, respectively. The Jarque–Bera (J.B.) test for normality results indicated that the RGDP datasets for these provinces were normally distributed. Table 2 summarizes the descriptive statistics of ITR. Shanxi and Hubei had the highest and lowest mean ITR values: US$476.9745 and US$191.507, respectively. The J.B. test for normality results indicate that the ITR datasets for China’s eight provinces are normally distributed with the exception of Shanxi and Hunan.

Figure 1.

International tourism receipts (ITR) vs. real gross domestic product (RGDP) across eight central provinces.

Table 1.

A summary of statistics of RGDP for central provinces.^a

Provinces	Mean	Max.	Min.	Std. Dev.	Skew.	Kurt.	J.B.
Shanxi	7,617.762	20,848.8	1,810.66	6,138.019	0.971412	2.563185	3.304479
Jilin	6,672.608	15,039.4	1,991.4	4,379.79	0.747981	2.14	2.478489
Heilongjiang	13,685.59	34,938.2	2,988.37	1,0583.61	0.752312	2.15	2.483652
Anhui	9,678.374	27,379.2	2,109.38	8,091.061	0.988033	2.60	3.386413
Jiangxi	9,666.913	27,037.3	2,132.13	8,045.163	0.961038	2.535312	3.258592
Henan	5,705.598	15,714.6	1,169.73	4,717.633	0.911175	2.452702	3.017081
Hubei	5,212.402	13,803.1	1,137.23	4,229.045	0.873912	2.336803	2.912264
Hunan	5,373.562	12,761.5	1,076.03	4,178.251	0.678359	1.973513	2.411968

RGDP: real gross domestic product.

The sample period is from 1995 to 2014.

Table 2.

A summary of statistics of ITR for central provinces (unit: US$ million).^a

Provinces	Mean	Max.	Min.	Std. Dev.	Skew.	Kurt.	J.B.
Shanxi	476.9745	1,840.26	31.39	595.406	1.318819	3.225897	5.840136*
Jilin	423.1155	917.62	60.62	286.0205	0.350111	1.73	1.75279
Heilongjiang	277.082	659.98	60.2	203.8274	0.587538	1.85	2.260971
Anhui	441.708	1,238.51	73.17	402.9363	1.066325	2.65	3.890359
Jiangxi	455.319	1,014.34	45.77	320.0017	0.343175	1.695008	1.811734
Henan	191.8775	556.87	24.99	181.1075	0.889113	2.302021	3.041053
Hubei	191.507	583.9	38	179.0717	1.134062	2.903411	4.29476
Hunan	225.181	822.68	20.62	245.6069	1.2112	3.288991	4.959613*

ITR: International tourism receipts.

The sample period is from 1995 to 2014.

Indicate significance at the 0.1 level.

Findings and implications

Empirical results

As previously outlined, testing cross-sectional dependency and slope homogeneity in a panel Granger causality study is crucial for selecting the appropriate estimator. Accounting for cross-sectional dependency and province-specific heterogeneity in the empirical analysis is crucial, as the provinces are highly integrated, and their economic relations are characterized by a high degree of globalization. Therefore, the empirical study of this study begins to determine whether cross-sectional dependency and heterogeneity exist across the provinces concerned. In order to investigate the existence of cross-sectional dependence, we conducted four different tests (LM,

{CD}_{lm}

CD

{LM}_{adj}

) and presented the results in Table 3. It is clear that the null of no cross-sectional dependency across the provinces is strongly rejected at conventional significance levels, implying that the SUR method is more appropriate than province-by-province OLS estimation (Zellner, 1962). This finding implies that a shock that occurs in one of these provinces seems to have been transmitted to the other provinces. Table 3 also reports the results from the two slope homogeneity tests (

\tilde{Δ}

and

{\tilde{Δ}}_{adj}

). Both tests reject the null hypothesis of slope homogeneity, supporting province-specific heterogeneity. The rejection of slope homogeneity implies that a panel Granger causality analysis imposes a homogeneity restriction on the variable of interest results in misleading inferences. In this respect, a panel Granger causality analysis based on the estimation of a panel vector autoregression and/or panel vector error correction model using a generalized method of moments or pooled OLS estimator is not an appropriate approach to detecting causal linkages between ITR and economic growth in China’s eight central provinces.

Table 3.

Cross-sectional dependence and homogeneous tests of central provinces.

Test	Test. Stat.
Cross-sectional dependence test
${CD}_{BP}$	112.054***
${CD}_{lm}$	11.232***
$CD$	8.672***
${LM}_{adj}$	18.1842***
Homogeneous test
$\tilde{Δ}$	9.1079***
${\tilde{Δ}}_{adj}$	10.1829***
Swamy Shat	44.4314***

***

Indicate significance at the 0.01 level, respectively.

${CD}_{BP}$ , ${CD}_{lm}$ , $CD$ and ${LM}_{adj}$ are the cross-sectional dependence tests of Breusch and Pagan (1980), Pesaran (2004), and Pesaran et al. (2008), respectively.

$\tilde{Δ}$ and ${\tilde{Δ}}_{adj}$ are slope homogeneity tests of Pesaran and Yamagata (2008).

Swamy (1970) develops the slope homogeneity test that allows cross-sectional heteroscedasticity (Pesaran and Yamagata, 2008).

The results from the bootstrap panel Granger causality analysis are reported in Table 4. In terms of Granger causality between ITR and economic growth for China’s eight central provinces, the results indicate that Hunan has significant one-way Granger causality from ITR to economic growth (see Table 4). Also, the study’s result indicates that Shanxi, Jiangxi and Henan have significant one-way Granger causality from economic growth to ITR. There is reliable evidence to show that Jilin, Anhui and Hubei have a reciprocal causal relationship between ITR and economic growth while the other one province (i.e. Heilongjiang) shows no significant relationship between ITR and economic growth.

Table 4.

Granger causality between ITR and RGDP for central provinces.

		$H_{0} :$ ITR does not Granger Cause RGDP				$H_{0} :$ RGDP does not Granger Cause ITR
		Bootstrap critical value^a				Bootstrap critical value^a
Provinces	Wald statistics	1%	5%	10%	Wald statistics	1%	5%	10%
Shanxi	3.490441	23.5853	11.66094	7.22484	23.68619***	22.76376	12.38154	8.4912
Jilin	56.35824***	25.23411	13.21309	8.81998	81.03322***	44.8009	24.0316	15.97239
Heilongjiang	0.365901	52.08318	19.97895	13.71549	0.01191	26.28486	12.58246	8.19491
Anhui	57.41219***	35.89962	17.06836	12.63784	47.17884***	43.44052	22.983	18.09505
Jiangxi	0.050997	23.53036	14.07998	10.78514	18.66128**	25.34395	16.47482	12.13266
Henan	1.380042	26.84781	13.7473	9.47855	17.78511**	24.50577	17.58072	10.95707
Hubei	30.22531**	32.30095	21.6939	13.43433	47.12947***	38.18236	23.54739	15.82333
Hunan	14.89646*	37.19235	20.93612	13.36004	5.722043	17.69233	11.64305	7.64572

ITR: international tourism receipts; RGDP: real gross domestic product.

Bootstrap critical values are obtained from 10,000 replications.

Significance at the 0.1 level.

Significance at the 0.05 level.

***

Significance at the 0.01 level.

To have a clear picture, the different results for ITR and economic growth of the sample are shown in Table 5. A comparison for China’s eight central provinces of Granger causality between ITR and economic growth is made. First, the empirical results of this study indicate that in the provinces, such as Hunan Province, tourism receipts stimulate economic growth, and economic growth is thus dependent on ITR, implying that negative travel shocks and travel reduction policies may depress economic growth. The governments of these provinces need to emphasize the tourism industry to increase economic growth. The results also provide evidence for the tourism-led-growth hypothesis for Hunan Province. Tourism may stimulate other economic industries by direct, indirect, and induced effects. An increase in tourism expenditure may lead to additional activities in related industries, and the overall variation connected with it will be greater than the initial injection in spending. If this effect is activated, one of the best ways to enhance economic benefits is to integrate tourism into the national economy by establishing strong linkages between tourism and other economic sectors such as agriculture, fisheries, manufacturing, construction, and other service industries (Cernat and Gourdon, 2012). If the tourism sector makes use of products and services produced within the local economy, it will strengthen these other sectors and provide additional income. In economics, this process is known as the multiplier effect of the tourism sector on the overall economy which is empirically calculated either through the Tourism Satellite Account (Spurr, 2009) or through computable general equilibrium models (which use input–ouput tables; see e.g. Blake et al., 2006; Dwyer et al., 2004). These methods also allow one to estimate these leakages. As explained by Cernat and Gourdon (2012), internal linkages not only relate to imports, but they may also refer to repatriation of salaries of foreign staff and interest paid on earning duties in the local tourism sector. It is important before making proper government policies, authorities need to identify the relationship between tourism and economic growth to make an appropriate strategy. For example, the growth hypothesis means that the tourism activities will positively contribute to economic growth. In this situation, the government should put more efforts into refining the infrastructure, such as transportation, public security, and accommodation in these provinces.

Table 5.

A summary of causality test between ITR versus RGDP in central provinces.

Provinces	ITR→RGDP	RGDP→ITR	Effect
Shanxi	X	△	Conservation causal relationship hypothesis
Jilin	△	△	Reciprocal causal relationship hypothesis
Heilongjiang	X	X	Neutrality causal relationship hypothesis
Anhui	△	△	Reciprocal causal relationship hypothesis
Jiangxi	X	△	Conservation causal relationship hypothesis
Henan	X	△	Conservation causal relationship hypothesis
Hubei	△	△	Reciprocal causal relationship hypothesis
Hunan	△	X	Growth causal relationship hypothesis

ITR: international tourism receipts; RGDP: real gross domestic product.

“△” denotes Granger cause from the left-hand side variable to the right-hand side variable.

“X” denotes no Granger cause from the right-hand side and left-hand side variable.

Second, a relationship between economic growths to ITR was found in the provinces, such as Shanxi, Jiangxi and Henan, indicating that economic growth can increase the demand for ITR and then lead to the development of tourism sectors in these provinces. The results provide evidence for the conservation hypothesis for these provinces. In these provinces, governmental policies should focus attention on the stability and transparency of its political institutions and continue to foster adequate investment in both physical and human capital that promotes growth for the overall economy. As a result, the tourism sector should reap the benefits, as additional resources generated from such growth will develop the province’s tourism infrastructure further, as well as signal the province’s stability to international tourists.

Third, if the causative process is bi-directional, and ITR and economic growth have a reciprocal causal relationship, then efforts in both areas would be beneficial. A reciprocal causal relationship between ITR and economic growth implies that excessive travel protection and reduced travel consumption may result in pressure on economic activities. We found a reciprocal hypothesis for Jilin, Anhui, and Hubei. This result suggests that for Jilin, Anhui, and Hubei, tourism development and economic growth are endogenous, indicating that these two factors mutually influence each other, and that this reinforcement may have important implications for the conduct of economic or tourism development policies in these provinces. The study’s result indicates that tourism expansion and economic growth have effects on each other, and economic growth affect tourism expansion in these provinces. As indicated by the results of Granger causalities, there is feedback between ITR and economic growth both in this study period. Therefore, it is fair to conclude that sample evidence is supporting reciprocal causal hypotheses for these provinces. Since the relationship between tourism expansion and economic growth is bidirectional, the policies designed to improve one variable inevitably affect the other. Economic growth can be improved by strategic planning of the tourism industry and vice versa. According to the empirical results of this study, in order to develop the tourism industry and sustain a steady economic growth, each policy decision has to consider a holistic economic view incorporating tourism. Thus, it can be possible to increase and stabilize the tourism industry as well as the whole economy. As tourism is dependent on economic decisions, public support in tourism by infrastructure, investments, incentives, promotion activities, and supervision is necessary for the industry to expand.

Fourth, the neutrality hypothesis is only found in Heilongjiang, indicating that neither ITR nor economic growth is sensitive to the other. The neutrality between ITR and economic growth suggests that travel conservation policies do not exert an adverse impact on economic growth, and that an international tourism receipt is not affected by economic performance. The neutrality between ITR and economic growth is attributed to a relatively small contribution of ITR to overall output. Thus, ITR may have little or no impact on economic growth in these provinces. China’s eight central provinces are investigated in this study. Only one of China’s eight provinces shows that there is no direct relationship between tourism and economic growth. The study’s result may be shocked or puzzled. Many previous studies have indicated the positive relationship between ITR and economic growth. It is dangerous for a province to make a policy using the positive relationship to formulate industrial strategies.

This study indicates that the direction of causality between ITR and economic growth may be determined by various factors. We speculate that the size of the province’s economy and its levels of openness and travel restrictions are plausible factors generating the differences among China’s eight provinces. In addition to these factors, the degree of dependence on tourism and the level of economic development may be considered to be additional determinants. The time series approaches overlook cross-sectional dependency across countries in the causality test; therefore, they may result in misleading inferences regarding the nature of causality between domestic tourism spending and economic growth. We find strong evidence for the existence of cross-sectional dependence among these provinces. It may be concluded that policy implications driven from the causality approach account for cross-sectional dependency and seem to be more appropriate. Furthermore, we also detect cross-provinces heterogeneity in the panels of China’s eight provinces, implying that each province develops its tourism policies.

Conclusions and research limitations

Most of previous studies have shown that some kinds of causal linkages exist between tourism activities and economic growth. In this work, we stress that previous studies have focused on one or few regions. In this study, we employ the bootstrap panel causality approach which is adapted from Granger causality test (Granger, 2003). A longitudinal panel data used in this study allow us to offer a comparable sample and result to simultaneously identify the relationships between ITR and economic growth among China’s eight provinces.

This study applies the bootstrap panel Granger causality approach to examine whether ITR and economic growth use the data from China’s eight major provinces over the period from 1995 to 2014. First, the empirical results of this study provide evidence for the growth causal relationship in Hunan. This finding supports growth hypothesis proposed by Deng et al. (2013, 2014). Second, a reverse relationship supports evidence on the conservation hypothesis for Shanxi, Jiangxi, and Henan. This result supports the conservation hypothesis proposed by He and Zheng (2011) and Othman et al. (2012). Third, a reciprocal causal relationship was found for Jilin, Anhui, and Hubei. This result supports feedback hypothesis proposed by Katircioglu (2009) and Wang and Xia (2013). Fourth, the result of a neutrality hypothesis supports one of these eight provinces (i.e. Heilongjiang). The study’s result supports the proposition of Tang (2011) and Tang and Tan (2015) that there are no spillover effects between tourism and economic development.

The empirical findings of this study provide important policy implications for China’s eight central provinces. This study is limited to the given sample period and the selected variables. In the future research, employing different economic methods such as vector autoregression, time–frequency domain, bootstrap rolling-window or wavelet re-examines the causal relationship between international tourism receipts and economic growth in China’s regions. It is recommended that related variables should consider such as current account deficits or numbers of tourist arrivals, cost of accommodation or interesting rates as a control variable. In this work, however, we cannot find out the data in these regions in terms of the data constraints. In the future study, we will consider exports, consumption, local government expenditure, terms of trade, investment, education, or imports among others as a control variable.

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.

Notes

Author Biographies

Tsung-Pao Wu is an Associate Professor at the School of Accounting and Finance in Beijing Institute of Technology-Zhuhai, China. He holds a PhD in Finance and has published in Applied Economics, Applied Economics Letters, Applied Financial Economics, Romanian Journal of Economic Forecasting, Empirical Economics, Macroeconomics and Finance in Emerging Market Economies, Economic Modelling, International Journal of Economic Policy in Emerging Economies, Japan and the World Economy, and International Economic Journal, among others. His major research field is applied econometrics time series analysis.

Hung-Che Wu is an Associate Professor at Business School in Nanfang College of Sun Yat-sen University, China. He holds a PhD in hospitality and tourism marketing and has published in Journal of the Knowledge Economy, Industrial Management & Data System, Journal of Organizational Change Management, Innovative Marketing, Tourism Analysis: An Interdisciplinary Tourism & Hospitality Journal, Journal of Tourism, Hospitality & Culinary Arts, British Food Journal, Journal of Quality Assurance in Hospitality & Tourism, Asia Pacific Journal of Marketing and Logistics, Event Management: An International Journal, Journal of Foodservice Business Research, Asia Pacific Journal of Tourism Research, International Journal of Tourism Research, International Journal of Tourism Sciences, International Journal of Contemporary Hospitality Management, Journal of Travel & Tourism Marketing, Journal of Hospitality Marketing & Management, International Journal of Tourism & Hospitality Administration, Journal of Hospitality and Tourism Management, Journal of China Tourism Research, Tourism and Hospitality Research, and Journal of Hospitality & Tourism Research, among others. He is currently working as one of the editorial board members for the Journal of Hospitality and Tourism Research, International Journal of Marketing Studies, Science Journal of Business and Management, and Business and Management Studies. His research interests include hospitality management, service quality, customer satisfaction, and consumer behavior. He is the corresponding author and can be contacted at: wuhungche66@gmail.com

References

Andriotis

(2002) Scale of hospitality firms and local economic development evidence from Crete. Tourism Management 23(4): 333–341.

Arslanturk

Balcilar

Ozdemir

(2011) Time-varying linkages between tourism receipts and economic growth in a small open economy. Economic Modelling 28(1): 664–671.

Balaguer

Cantavella-Jorda

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

Balassa

(1978) Exports and economic growth: Further evidence. Journal of Development Economics 5: 181–189.

Biagi

Brandano

Lambiri

(2015) Does tourism affect house prices? Evidence from Italy. Growth and Change 46(3): 501–528.

Blake

Sinclair

Campos Soria

(2006) Tourism productivity: Evidence from the United Kingdom. Annals of Tourism Research 33(4): 1099–1120.

Breitung

(2005) A parametric approach to the estimation of cointegration vectors in panel data. Econometric Reviews 24(2): 151–173.

Breusch

Pagan

(1980) The LM test and its application to model specification in econometrics. Review of Economic Studies 47(1): 239–254.

Brida

Cortes-Jimenez

Pulina

(2016) Has the tourism-led growth hypothesis been validated? A literature review. Current Issues in Tourism 19(5): 394–430.

10.

Brida

Risso

(2010) Tourism as a determinant of long-run economic growth. Journal of Policy Research in Tourism, Leisure and Events 2(1): 14–28.

11.

Caglayan

Sak

Karymshakov

(2012) Relationship between tourism and economic growth: A panel Granger causality approach. Asian Economic and Financial Review 2(5): 591–602.

12.

Cernat

Gourdon

(2012) Paths to success: Benchmarking cross-country sustainable tourism. Tourism Management 33(5): 1044–1056.

13.

Chen

Chiou-Wei

(2009) Tourism expansion, tourism uncertainty and economic growth: New evidence from Taiwan and Korea. Tourism Management 30(6): 812–818.

14.

Copeland

(1991) Tourism, welfare, and de-industrialization in a small open economy. Economica 58(232): 515–529.

15.

Corden

Neary

(1982) Booming sector and de-industrialisation in a small open economy. The Economic Journal 92(368): 825–848.

16.

Corte’s-Jimeńez

(2008) Which type of tourism matters to the regional economic growth? The cases of Spain and Italy. International Journal of Tourism Research 10(2): 127–139.

17.

Deng

Cao

(2013) Tourism resource development and long-term economic growth: A resource curse hypothesis approach. Tourism Economics 20(5): 923–938.

18.

Deng

Shao

(2014) Has international tourism promoted economic growth in China? A panel threshold regression approach. Tourism Economics 20(4): 911–977.

19.

Dissart

Dehex

Marsat

(2015) Tourism, Recreation and Regional Development: Perspectives from France and Abroad, London, UK: Ashgate Publishing Limited.

20.

Dritsakis

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

21.

Durbarry

(2004) Tourism and economic growth: The case of Mauritius. Tourism Economics 10(4): 389–401.

22.

Dwyer

Forsyth

Spurr

(2004) Evaluating tourism’s economic effects: New and old approaches. Tourism Management 25(3): 307–317.

23.

Fagance

(1999) Tourism as a feasible option for sustainable development in small island developing states (SIDS): Nauru as a case study. Pacific Tourism Review 3(2): 133–142.

24.

Figini

Vici

(2010) Tourism and growth in a cross section of countries. Tourism Economics 16(4): 789–805.

25.

Granger

CWJ

(2003) Some aspects of causal relationships. Journal of Econometrics 112(1): 69–71.

26.

Gunduz

Hatemi

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

27.

Zheng

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

28.

Katircioglu

(2009) Tourism, trade and growth: The case of Cyprus. Applied Economics 41(21): 2741–2750.

29.

Katircioglu

(2010) Testing the tourism-led growth hypothesis for Singapore: An empirical investigation from bounds test to cointegration and Granger causality tests. Tourism Economics 16(4): 1095–1101.

30.

Khan

Phang

Toh

(1995) The multiplier effect: Singapore’s hospitality industry. Cornell Hospitality Quarterly 36(1): 64–69.

31.

Kim

Chen

Jang

(2006) Tourism expansion and economic development: The case of Taiwan. Tourism Management 27(5): 925–933.

32.

Kónya

(2006) Exports and growth: Granger causality analysis on OECD countries with a panel data approach. Economic Modelling 23(6): 978–992.

33.

Lanza

Temple

Urga

(2003) The implications of tourism specialization in the long run: An econometric analysis for 13 OECD economies. Tourism Management 24(3): 315–321.

34.

Lee

Chien

(2008) Structural breaks, tourism development, and economic growth: Evidence from Taiwan. Mathematics and Computers in Simulation 77(4): 358–368.

35.

Lee

Kwon

(1995) Importance of secondary impact of foreign tourism receipts on the South Korean economy. Journal of Travel Research 34(2): 50–54.

36.

Lin

Liu

(2000) A study of economies of scale and economies of scope in Taiwan international tourist hotels. Asia Pacific Journal of Tourism Research 5(2): 21–28.

37.

Narayan

(2004) Fiji’s tourism demand: The ARDL approach to cointegration. Tourism Economics 10(2): 193–206.

38.

Nowak

Sahli

Cortés-Jiménez

(2007) Tourism, capital goods imports and economic growth: Theory and evidence for Spain. Tourism Economics 13(4): 515–536.

39.

Nunkoo

(2015) Tourism development and trust in local government. Tourism Management 46: 623–634.

40.

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

41.

Othman

Salleh

NHM

Sarmidi

(2012) Analysis of causal relationship between tourism development, economic growth and foreign direct investment: An ARDL approach. Journal of Applied Sciences 12(12): 1245–1254.

42.

Pesaran MH (2004) General Diagnostic Tests for Cross Section Dependence in Panels. Cambridge Working Papers in Economics (No. 0435). Faculty of Economics, University of Cambridge, Cambridge.

43.

Pesaran

(2006) Estimation and inference in large heterogeneous panels with multifactor error structure. Econometrica 74(4): 967–1012.

44.

Pesaran

Ullah

Yamagata

(2008) A bias-adjusted LM test of error cross-section independence. The Econometrics Journal 11(1): 105–127.

45.

Pesaran

Yamagata

(2008) Testing slope homogeneity in large panels. Journal of Econometrics 142(1): 50–93.

46.

Santana-Gallego

Ledesma-Rodriguez

Perez-Rodriguez

et al. (2010) Does common currency promote countries’ growth via trade and tourism? The World Economy 33(12): 1811–1835.

47.

Sarafidis V, Yamagata T and Robertson D (2009) A test of cross section dependence for a linear dynamic panel model with regressors. Journal of Econometrics 148(2): 149–161.

48.

Schubert

Brida

Risso

(2011) The impacts of international tourism demand on economic growth of small economies dependent on tourism. Tourism Management 32(2): 377–385.

49.

Song

Dwyer

et al. (2012) Tourism economics research: A review and assessment. Annals of Tourism Research 39(3): 1653–1682.

50.

Spurr

(2009) Tourism satellite accounts. In: Dwyer

Forsyth

(eds) International Handbook on the Economics of Tourism, Cheltenham, MA: Edward Elgar, pp. 283–299.

51.

Surugiu

(2013) Is the tourism sector supportive of economic growth? Empirical evidence on Romanian tourism. Tourism Economics 19(1): 115–132.

52.

Swamy

PAVB

(1970) Efficient inference in a random coefficient regression model. Econometrica: Journal of the Econometric Society 38(2): 311–323.

53.

Tang CF (2011) Is the tourism-led growth hypothesis valid for malaysia? a view from disaggregated tourism markets. International Journal of Tourism Research 13(1): 97–101.

54.

Tang

(2013) Temporal Granger causality and the dynamic relationship between real tourism receipts, real income, real exchange rates in Malaysia. International Journal of Tourism Research 15(3): 272–284.

55.

Tang

Tan

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

56.

Theobald

(2001) Global Tourism, 2nd ed. Oxford, UK: Butterworth & Heinemann.

57.

United Nations World Tourism Organization (UNWTO) (2014) UNWTO World Tourism Barometer. Available at: http://www.unwto.org (accessed 3 July 2016).

58.

Wang

Xia

(2013) A study on the relationship between tourism industry and regional economic growth – A case study of Jiangsu Gaochun district. Modern Economy 4(7): 482–488.

59.

Zellner

(1962) An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. Journal of the American Statistical Association 57(298): 348–368.