Prescribing for the tourism-induced Dutch disease: A DSGE analysis of subsidy policies

Abstract

Tourism-induced Dutch disease can be particularly detrimental to small open economies due to deindustrialization and potential long-term welfare loss. This study adopts an innovative macroeconomic modelling tool, dynamic stochastic general equilibrium modelling, to investigate the effects of external inbound tourism booms on the national economic account of a small open economy. The results confirm the existence of Dutch disease, but tourism booms also bring welfare gains to the destination country. We further model the effects of two strategies to mitigate the Dutch disease by assuming that the government can tax the tourism sector and subsidize the manufacturing sector in two ways: production subsidies and investment subsidies. The results show that the effectiveness of production subsidies is very modest, while investment subsidies can almost completely overturn the Dutch disease. In terms of welfare, investment subsidies lower welfare gains in the very short term, but the positive effects persist over the longer term, which is different from the production subsidy case. Last, practical implications are provided.

Keywords

Dutch disease dynamic stochastic general equilibrium modelling government subsidy tourism tax

Introduction

Along with the fast growth of the tourism and travel industry globally, the significant economic contribution of tourism to the local community and destination country has been largely recognized and applauded (Pablo-Romero and Molina, 2013). The tourism industry is touted as the most important earner of foreign exchange in many countries and regions (Balaguer and Cantavella Jordá, 2002). Tourism growth tends to create job opportunities, increase public and private investments in local infrastructures and extend the openness of foreign investment to various sectors (Lin et al., 2019), all of which are likely to contribute to the local economic prosperity. The development of tourism relies heavily on endowments of tourism resources as tourist attractions, and this provides valuable opportunities for many less developed regions with rich natural/cultural tourism resources to develop their tourism industries (Yang and Fik, 2014).

Despite the considerable economic contributions that tourism brings in, many negative consequences of tourism growth have been highlighted. In addition to the assorted negative impacts from social, cultural, environmental and ecological perspectives (Gössling and Peeters, 2015; Lee, 2016), some negative economic consequences were also recognized. For example, Lanza et al. (2003) suggest that an overdependence on tourism can impede long-term economic growth. In particular, the phenomenon of the Dutch disease regarding deindustrialization may be induced by a tourism boom (Chao et al., 2006), and the decline of the capital stock may cause a welfare loss in the long run (Nowak and Sahli, 2007).

Various types of macroeconomic models have been introduced to understand the impact of tourism on the local economy. Econometric models (e.g. the vector autoregression and dynamic panel data model) have been prevalent to understand the short- and long-term relationships between tourism expansion and economic growth (Iglesias et al., 2018), and the input-output (I-O) and computable general equilibrium (CGE) models have frequently been used to understand how an economy reacts to changes in tourism income and revenues (Meng and Pham, 2017). However, the dynamic stochastic general equilibrium (DSGE) model, which is a cutting-edge macroeconomic model, has not been applied to tourism economics so far.¹ Unlike the econometric analysis of data, a DSGE model provides a theoretical and structural empirical account of macroeconomic relationships; it is micro-founded, immune to Lucas critique (Lucas, 1976), enclosing agents’ intertemporal optimization under uncertain environments into the model. Unlike other static macroeconomic analysis tools such as I-O and CGE models, the DSGE model takes dynamic factors and stochastic terms into consideration under the general equilibrium theory (Hashimzade and Thornton, 2013). All the above features are absent in most I-O and CGE frameworks. While previous types of analysis are of decent benefits, more sophisticated techniques that are better able to represent a real economy, are available, and we take advantage of the advances in macroeconomic modelling by incorporating DSGE type of framework into tourism economics analysis.

While there is much literature on the economic impacts of tourism (Song et al., 2012), these issues have rarely been examined in a DSGE framework, which has been very popular in mainstream macroeconomics (Wickens, 2014). To fill the research gap, we propose a DSGE model to investigate the impacts of an external tourism boom on a small open economy with two sectors: tourism and manufacturing. The small open economy is calibrated based on the macro data of Thailand, a Southeast Asian economy that heavily depends on the tourism industry (Wattanakuljarus and Coxhead, 2008). The reasons why we pick up Thailand are threefold. First, the export of tourism services is a major driver of the Thai economy. The value of tourism services reached 50 billion US dollars in 2016, accounting for 12.5% of national gross domestic product (GDP; Bank of Thailand, 2016). Second, the Asia-Pacific region is the second most visited region in the world after Europe and has been the fastest growth in recent years (UNWTO/GTERC, 2017). Third, the rate of growth of Thailand’s share of inbound tourists to the Asia-Pacific region is increasing recently, which is companied by faster growth in the tourism sector than the regional and global average (Bank of Thailand, 2016).

Our postulated hypotheses are as follows: (1) the inbound tourism demand shock would cause Dutch disease, (2) the induced Dutch disease would lead to welfare loss and (3) governmental policy can help alleviate the tourism-induced Dutch disease. In particular, we are interested in the occurrence of tourism-induced Dutch disease and how to alleviate the Dutch disease with government tax and subsidy policies. By looking into the dynamics of the effects, we are able to see the dynamic trajectories of how the inbound tourism demand impacts the two sectors of the domestic economy and how governmental tax and subsidy policies counteract the Dutch disease effects. Therefore, this study contributes to the current knowledge of the Dutch disease in at least two major ways. First and foremost, our study represents one of the pioneering research efforts applying DSGE to tourism economics (Su, 2017). Using a DSGE model, we are able to capture the dynamics of the structural relationship among different variables and understand how Dutch disease takes place from inbound tourism shock via various channels. Second, we propose potential prescriptions for treating Dutch disease from a rigorous quantitative perspective instead of qualitative measure mainly considered in the literature, and the effectiveness of production subsidies and investment subsidies in the manufacturing sector is tested and compared under the DSGE framework to highlight the policy trade-off between the two subsidies.

The main results of our study can be summarized as follows. Our analysis confirms the existence of Dutch disease, but tourism booms also bring welfare gains to the destination country. In terms of prescription of Dutch disease, we compare the effectiveness of production subsidies and investment subsidies. The results show that the efficacy of production subsidies is very modest, while investment subsidies can almost completely overturn the Dutch disease. Regarding welfare, investment subsidies lower welfare gains in the very short term, but the positive effects persist over the longer term, which is different from the production subsidy case. The rest of this article is structured as follows. The second section briefly reviews the literature. The third section constructs the small open economy DSGE model for our analysis. The fourth section calibrates the parameter values and provides the empirical results. Concluding remarks are given in the last section.

Literature review

In recent years, there has been a growing interest in academics, international policy institutions and central banks in developing small-to-medium, even large-scale, open economy macroeconomic models called a DSGE framework (Fernández Villaverde et al., 2016). The term ‘DSGE’ was initially introduced by Kydland and Prescott (1982) in their seminal contribution on the real business cycle (RBC) model, which is based on a neoclassical framework with micro-founded optimization behaviour of economic agents with flexible prices. Soon the model was extended into small open economies to account for the stylized facts observed in typical candidate countries (Mendoza, 1991). The use of DSGE models used to be criticized due to their inability to fit the data (Pagan, 2003), but the new generation of DSGE models developed by Christiano et al. (2005) among others has shown great promise of improving the empirical fit in both the closed economy (Smets and Wouters, 2004) and open economy (Adolfson et al., 2007) scenarios. Although DSGE models have become popular for studying macroeconomic fluctuations and analysing quantitative policy issues, the application of DSGE models in the tourism field is still profoundly underdeveloped (Liu et al., 2018).

Turning to the tourism issue, for a tourism-oriented economy, an expansion in tourism sector is naturally welcomed because more visitors can bring more income to the local economy (Hazari and A-Ng, 1993). International tourism is considered as a major source of foreign exchange and an important engine of local economic growth (Hazari and Sgro, 1995). The positive effect of tourism on long-run economic growth can work through several channels. First, the foreign exchange earned from inbound tourists can be used to pay for imported capital or basic inputs used in the production process (Nowak et al., 2007). Second, tourism plays an important role in spurring local investment in new infrastructure like transportation facilities (Chou, 2013). Third, tourism contributes to generate employment and to increase income for relatively unskilled labour (Inchausti Sintes, 2015). Fourth, tourism is an important factor of diffusion of technical knowledge, simulation of research and development and accumulation of human capital (Sihabutr, 2012). Based on these, tourism is believed to bring positive impact to local economy, and tourism policies are becoming major concerns for many countries.

Apart from the positive economic effects, the influx of foreign tourists tends to also generate undesirable consequences to residents of the host country, such as congestion in the tourism related consumption and infrastructure and the degradation of environmental quality (Chang et al., 2011). Additionally, a tourism boom tends to boost the demand for these non-traded goods and thus moves the productive factors away from the traded (manufacturing) sector to the non-traded sector. An expansion in the non-traded sector is coupled with a contraction in the traded sector, thereby leading to the emergence of the Dutch disease (Copeland, 1991). Dutch disease is traditionally related to the effects of an expansion of a booming sector on the rest of the economy, where the main concern is with the possibility of deindustrialization (Corden and Neary, 1982). In the tourism context, one major factor contributing to the Dutch disease is the price effect from the tourism boom; selective studies of tourism-related Dutch disease among others are summarized in Table 1. Chao et al. (2006) show that an expansion of tourism increases the relative price of non-traded goods, which leads to a gain in income via the secondary terms of the trade effect. This price effect results in a decrease in the demand for capital and therefore a reduction in output by the traded good sector. Additionally, Nowak and Sahli (2007) demonstrated that the tourism-induced Dutch disease partly results from labour losses and intense tourism-related land use.

Table 1.

Brief summary of previous studies of tourism-induced Dutch disease.

Studies	Countries	Model used	Conclusions
Copeland (1991)	Generic	SGE model	Deindustrialization exists, but not necessarily undesirable, tax policy can be used to extract the rent from tourists
Chao et al. (2006)	Generic	DGE model	Tourism boom induces deindustrialization, causing decline of the capital stock and welfare loss in the long run
Capó et al. (2007)	Spanish islands	Empirical evidence from descriptive statistics	Rapid expansion in tourism industries would result in Dutch disease, urging the government to pay more attention for a more balanced economy
Nowak and Sahli (2007)	Small islands	GE framework	Inbound tourism boom would cause Dutch disease, but the welfare effect is ambiguous
Chang et al. (2011)	Generic	Dynamic optimizing macro model	Dutch disease can be treated effectively if the government tax tourism provides public services for the manufacturing sector, but ineffectively if the tax revenue is rebated to local residents
Sheng (2011)	Generic	Partial and general equilibrium models	Taxing tourism and subsidizing the non-tourism will not increase local welfare, but further empirical studies are needed
Inchasti Sintes (2015)	Spain	Dynamic CGE model	Dutch disease occurs, leading to economic boom in the short run, but jeopardizing future productivity gains and economic growth in the long term
Pham et al. (2015)	Australia	CGE model	Mining boom in Australia diverts resources away from tourism sector and investment in accommodation and aviation can treat the Dutch disease to tourism sector effectively
Chen et al. (2016)	Generic	DGE model	Tourism-induced Dutch disease will not impose welfare cost if international borrowing is allowed, even result in welfare improvement if tourism good is capital intensive and international financial market is closed

Note: SGE: static general equilibrium; DGE: dynamic general equilibrium; CGE: computable general equilibrium.

Several studies have empirically or mathematically confirmed the tourism-induced Dutch disease (Capó et al., 2007; Forsyth et al., 2014; Inchausti Sintes, 2015; Pham et al., 2015), albeit with some exceptions (Holzner, 2011). Copeland (1991) used a static trade model to study how a tourism boom would affect the welfare and pattern of production in the destination country and concluded that the Dutch disease will occur, whereas welfare is improving when there is no distortion. Chao et al. (2006) constructed a specific-factor model and found that a tourism boom leads to deindustrialization improved welfare when there is no capital externality for manufacturing production. Chang et al. (2011) further developed a dynamic optimizing macroeconomic model with touristic congestion externalities and found deindustrialization in the manufacturing sector led by tourism expansion. In another study, Chen et al. (2016) considered a general two-sector model in which a tourism good also needs capital to be produced and indicated that the Dutch disease occurs after tourism expansion but welfare will be affected only when there is an imperfect international financial market in which the destination country cannot borrow freely.

A handful of studies have discussed the possible ways to treat the Dutch disease. In their discussion of the general tourism taxation policies, Sheng and Tsui (2009) and Sheng (2017) suggested that taxing tourism can be a viable way to alleviate the symptoms of the Dutch disease. In another study, Sheng (2011) specifically investigated the policies to combat the Dutch disease. Using a general and partial equilibrium analysis, they demonstrated that a combination policy using taxes and subsidies is not appropriate due to the adverse effects on local welfare. More recently, Su (2017) confirmed that a tourist subsidy that is financed by government debt has a small effect on helping the tourism industry. By and large, the above studies related to Dutch disease in tourism and its cure are mainly using qualitative tools (focusing on the direction of change in some variables as related to change of some other variables), ignoring the quantitative aspects of the problem, which provide valuable information for policy implementation. The DSGE framework with calibrated parameter values has been well-recognized for its merits in quantitative analysis, and we can also enclose uncertainties (as stochastic terms) in a dynamic context to match more closely to the reality, on the ground of treating Dutch disease in tourism perspective in a more advanced economic modelling environment.

DSGE model of Dutch disease

Following the convention in the tourism economics literature (Chao et al., 2006; Copeland, 1991; Su, 2017), we consider a small open economy with two types of final goods: a tourism good $y_{t}^{T}$ ,² which is non-tradable, and a manufacturing good $y_{t}^{M}$ , which is tradable. This dichotomy provides the most common categories of tourist consumption, which has merit for analysing the resource allocation between sectors. A tourism good is non-tradable in the sense that it can only be consumed within the destination country’s territory. Although non-tradable, tourist consumption can bring foreign currency earnings to the destination country when consumers are foreign visitors.

In our model, we assume that there are five groups of decision-makers in the national economy: domestic households, foreign tourists, tourism firms (in the tourism sector), manufacturing firms (in the manufacturing sector) and a government. Domestic households derive utility from consumption and leisure (and disutility from working). Subject to their budgetary constraints, the households seek to maximize their lifetime utility, defined as the present value of current and future instantaneous utility that would be stated shortly after. They consume both types of goods, tourism goods $y_{t}^{T}$ , which are produced exclusively by tourism firms, and manufacturing products $y_{t}^{M}$ , which are produced exclusively by manufacturing firms. Unlike members of domestic households, foreign tourists consume only the tourism good $y_{t}^{T}$ . In both the tourism and manufacturing sectors, firms maximize their profits by producing goods through Cobb–Douglas technology, which is a common setup in the tourism economic literature (Kostakis and Theodoropoulou, 2017). The government balances its budget in each period; it provides either lump-sum transfers to domestic residents or subsidies to manufacturing firms by levying taxes on the tourism consumption of foreign tourists (Candela et al., 2015). Figure 1 summarizes the circular flow of the model economy.

Figure 1.

Circular flow of the economy. Note: Solid arrows mean definite flows, whereas dashed ones capture either cases (government transfers to households or subsidizes manufacturing firms).

Household

The economy is inhabited by a unit measure of identical and infinitely lived domestic households, which derive utility from consumption and disutility from working (Shen et al., 2018). The disutility of working is the reverse of the utility of leisure. The household’s lifetime utility function is given by

E_{0} \sum_{t = 0}^{\infty} β^{t} {\frac{{[c_{t}^{d}]}^{1 - σ}}{1 - σ} - ψ \frac{{[l_{t}^{d}]}^{1 + θ}}{1 + θ}}

where $c_{t}^{d}$ and $l_{t}^{d}$ are the consumption and labour supply of domestic households, respectively. Instantaneous utility is positively associated with consumption and negatively associated with labour. Moreover, σ measures the elasticity of the intertemporal substitution for consumption, and θ measures the Frisch elasticity of the labour supply.³ $c_{t}^{d}$ is the composite consumption of the tourism good $c_{t}^{T,d}$ and the manufacturing good $c_{t}^{M,d}$ and is given by constant elasticity of substitution aggregation as follows

c_{t}^{d} = {[{ϕ^{c}}^{\frac{1}{χ^{c}}} {(c_{t}^{T,d})}^{\frac{χ^{c} - 1}{χ^{c}}} + {(1 - ϕ^{c})}^{\frac{1}{χ^{c}}} {(c_{t}^{M,d})}^{\frac{χ^{c} - 1}{χ^{c}}}]}^{\frac{χ^{c}}{χ^{c} - 1}}

where $ϕ^{c}$ is the share of tourism consumption and $χ^{c}$ is the elasticity of substitution between tourism and manufacturing consumption. The setup implies imperfect substitutability between the two sectoral consumptions. $l_{t}^{d}$ is the composite labour service of the tourism sector $l_{t}^{T,d}$ and the manufacturing sector $l_{t}^{M,d}$ and is given by

l_{t}^{d} = {[{(ϕ^{l})}^{- \frac{1}{χ^{l}}} {(l_{t}^{T,d})}^{\frac{1 + χ^{l}}{χ^{l}}} + {(1 - ϕ^{l})}^{- \frac{1}{χ^{l}}} {(l_{t}^{M,d})}^{\frac{1 + χ^{l}}{χ^{l}}}]}^{\frac{χ^{l}}{1 + χ^{l}}}

where $ϕ^{l}$ is the share of tourism labour and $χ^{l}$ is the elasticity of substitution between tourism and manufacturing labour. The setup implies imperfect substitutability between the labour in the two sectors, suggesting that the job nature is at least partially heterogeneous in the two sectors and that there are some costs/difficulties when the labour in one sector shifts.

The households face the following intertemporal budget constraint

\begin{array}{l} c_{t}^{d} + [k_{t}^{T,d} - (1 - δ) k_{t - 1}^{T,d}] + [k_{t}^{M,d} - (1 - δ) k_{t - 1}^{M,d}] \\ = w_{t} l_{t}^{d} + r_{t}^{T} k_{t - 1}^{T,d} + r_{t}^{M} k_{t - 1}^{M,d} + Π_{t} + T_{t} \end{array}

where w_t is the composite wage index.⁴ $k_{t}^{T,d}$ and $k_{t}^{M,d}$ are the capital in the tourism and manufacturing sectors, respectively, and $r_{t}^{T}$ and $r_{t}^{M}$ are the respective returns from investing in the two sectors. Households receive $Π_{t}$ as dividends from firms and a lump-sum transfer of $T_{t}$ from the government, which will be defined later. Finally, δ is the depreciation rate of capital, which is assumed to be identical in both sectors. The left-hand side stands for the expenditures, whereas the right-hand side represents the sources of income of the household, who will choose $c_{t}^{d}$ , $l_{t}^{d}$ , $k_{t}^{T,d}$ and $k_{t}^{M,d}$ to maximize their lifetime utility (1) subject to the budgetary constraint (2).

Foreign tourists

In line with previous literature on tourism (e.g. Chang et al., 2011; Chao et al., 2006; Chen et al., 2016), we assume that besides the domestic consumers, there are also foreign tourists in the economy. By assumption, they do not work or invest but only demand the tourism goods and pay consumption taxes to the government of the destination country. The demand of foreign tourists $c_{t}^{f}$ takes the following functional form without the loss of generality

c_{t}^{f} = {[(1 + τ_{f}) p_{t}^{T}]}^{- η} z_{t}^{f^{ξ}}

where $p_{t}^{T}$ is the relative price of the tourism good (as a bundle) and $τ_{f}$ is the rate of the consumption tax levied on foreign tourists. In this equation, $c_{t}^{f}$ is assumed to be decreasing in the after-tax rate of price $(1 + τ_{f}) p_{t}^{T}$ to satisfy the law of demand, and η measures the price elasticity of tourism demand. $z_{t}^{f}$ represents the exogenous factor of tourist activity that captures the boom and bust of the tourism demand of foreigners, and ξ captures the elasticity of tourism demand to exogenous factors such as foreign tourists’ incomes. Note that we assume an autocorrelation of tourism demand to capture the persistence of tourism demand (Balaguer and Cantavella Jordá, 2002) and that $z_{t}^{f}$ follows the $AR (1)$ process in logarithmic form as follows

ln z_{t}^{f} = ρ ln z_{t - 1}^{f} + ε_{t}^{f}

where $ε_{t}^{f}$ is the shock/disturbance to the tourism demand of foreigners. Equation (3) states that foreign tourists’ demand is negatively associated with the tax on and relative price of tourism goods.

Firms

The setting of the production environment largely follows the literature (Chen et al., 2016). Both the tourism and manufacturing sectors are assumed to be perfectly competitive for simplicity.⁵ Aggregate production is the sum of tourism and manufacturing production given by

y_{t} = p_{t}^{T} y_{t}^{T} + p_{t}^{M} y_{t}^{M}

where $p_{t}^{T}$ and $p_{t}^{M}$ are the relative prices of tourism and manufacturing goods, respectively.

Tourism sector

Given the empirical results and the common specification in the tourism literature (Chen et al., 2016), we assume that the tourism firms leverage labour $l_{t}^{T}$ and capital $k_{t}^{T}$ as inputs to produce the tourism good $y_{t}^{T}$ . Capital can be thought of as including endowed tourism resources (e.g. national parks, world heritage sites, historical sites) and accommodation facilities (e.g. hotels, recreation and amusement centres). Labour services refer to the performance of specific tasks required to meet the needs of tourists. For instance, a resort needs labour inputs from managers, front desk receptionists and housekeeping and maintenance staffs. The production function of a tourism good producer is in the Cobb–Douglas form given by⁶

y_{t}^{T} = z^{T} {(k_{t - 1}^{T})}^{1 - α^{T}} {(l_{t}^{T})}^{α^{T}}

where $z^{T}$ is an exogenous technological factor in the tourism sector, such as innovation in hotel management (Mattsson and Orfila Sintes, 2014) or transportation services (Yang et al., 2019). $α^{T}$ is the share of labour services in this sector. The tourism good producer faces the following optimization problem:

max_{{k_{t - 1}^{T}, l_{t}^{T}}} p_{t}^{T} y_{t}^{T} - r_{t}^{T} k_{t - 1}^{T} - w_{t}^{T} l_{t}^{T}

where $r_{t}^{T}$ is the rental rate of tourism sector capital and $w_{t}^{T}$ is the wage rate of labour services. Tourism firms will choose the amount of rented capital and hired labour to maximize the profits defined in equation (6) subject to equation (5).

Manufacturing sector

The setting of the manufacturing sector’s production is standard. Each producer employs capital $k_{t - 1}^{M}$ and labour $l_{t}^{M}$ to produce good $y_{t}^{M}$ using Cobb–Douglas technology as follows:

y_{t}^{M} = z^{M} {(k_{t - 1}^{M})}^{1 - α^{M}} {(l_{t}^{M})}^{α^{M}}

where $z^{M}$ is the exogenous technological factor for the manufacturing sector and $α^{M}$ is the share of labour services in this section. The forms of the production function in the two sectors are similar, but the tourism sector is relatively more labour intensive, whereas manufacturing sector is more capital intensive such that $α^{T} > α^{M}$ (Chen et al., 2016). The manufacturing firm faces the following optimization problem:

max_{{k_{t - 1}^{M}, l_{t}^{M}}} p_{t}^{M} y_{t}^{M} - r_{t}^{M} k_{t - 1}^{M} - w_{t}^{M} l_{t}^{M}

where $r_{t}^{M}$ is the rental rate of capital and $w_{t}^{M}$ is the wage rate of labour services. Finally, as both sectors are perfectly competitive, free entry and exit guarantee zero profits for each producer when both factors of production are variable (Jehle and Reny, 2011).

Government

The government plays the role of redistributing the resources among different sectors in the economy. The only source of revenue is from the tax levied on foreign tourists’ consumption, whereas the flow of tax revenue can be redistributed either to households for enhancing overall welfare or to manufacturing firms for promoting sectoral production. Manufacturing subsidies can be conducted in two different ways: production subsidies and investment subsidies. Our purpose is to compare the economic outcomes of the different scenarios and make a policy prescription based on that.

Households transfer

In the benchmark scenario, the government levies taxes on the tourism consumption of inbound tourists $p_{t}^{T} c_{t}^{f}$ and redistributes these tax revenues to domestic residents as a transfer payment $T_{t}$ in a lump-sum manner. The government’s budget constraint can be expressed as follows:

T_{t} = τ_{t}^{f} p_{t}^{T} c_{t}^{f}

where $τ_{t}^{f}$ is the tax rate on foreign tourists’ consumption. The government balances its budget in any period by adjusting the lump-sum transfer $T_{t}$ . After receiving the transfer, domestic households can allocate the additional resources among consumption and saving based on their optimization.

Production subsidy

Concerning the possibility of deindustrialization associated with the Dutch disease after a tourism boom, the government may subsidize manufacturers instead of households. In this scenario, the government collects tax revenues from foreign tourists and subsidizes manufacturing production outputs. Then, equation (8) becomes

max_{{k_{t - 1}^{M}, l_{t}^{M}}} (1 + s_{t}^{Y}) p_{t}^{M} y_{t}^{M} - r_{t}^{M} k_{t - 1}^{M} - w_{t}^{M} l_{t}^{M}

8′

where $s_{t}^{Y}$ is the subsidy per unit value of manufacturing production and the total subsidy is financed by taxing foreign tourists as follows:

s_{t}^{Y} p_{t}^{M} y_{t}^{M} = τ_{ft} p_{t}^{T} c_{t}^{f}

With the production subsidy, the manufacturing firm’s marginal product is higher than without; hence, at the optimum, the manufacturing firm will potentially hire more labour and has more capital in production. This ultimately heightens the production outputs and increases wages and capital returns.

Investment subsidy

The second way to subsidize the manufacturing sector is investments in manufacturing capital. In this scenario, the government collects tax revenues from foreign tourists and subsidizes households who invest in manufacturing capital. Households’ budget constraint (2) becomes

\begin{array}{l} c_{t}^{d} + [k_{t}^{T,d} - (1 - δ) k_{t - 1}^{T,d}] + (1 - s_{t}^{M}) [k_{t}^{M,d} - (1 - δ) k_{t - 1}^{M,d}] \\ = w_{t} l_{t}^{d} + r_{t}^{T} k_{t - 1}^{T,d} + r_{t}^{M} k_{t - 1}^{M,d} + Π_{t} \end{array}

2′

where $s_{t}^{M}$ is the subsidy per unit of manufacturing capital return and the total subsidy that is financed by taxing foreign tourists is as follows:

s_{t}^{M} [k_{t}^{M,d} - (1 - δ) k_{t - 1}^{M, d}] = τ_{t}^{f} p_{t}^{T} c_{t}^{f}

With investment subsidies for the manufacturing sector, households will find it attractive to postpone consumption for investment, particularly in the manufacturing sector. What we expect is that the consumption of manufacturing goods drops, whereas the investment in the manufacturing sector rises. When the total consumption is lower, the investment becomes higher. Higher investments result in more capital, thus leading to a higher marginal product of labour and higher wages.

By combining the households’ budget constraint (2), the definitions of both the tourism and manufacturing firms’ profit maximization (6) and (8) and the government’s budget constraint in three different scenarios (9), (10) and (11), respectively, the economy-wide resource constraint is given by

p_{t}^{T} c_{t}^{T} + p_{t}^{M} c_{t}^{M} + i_{t}^{T} + i_{t}^{M} = p_{t}^{T} y_{t}^{T} + p_{t}^{M} y_{t}^{M} + τ_{t}^{f} p_{t}^{T} c_{t}^{f}

where $i_{t}^{T}$ and $i_{t}^{M}$ are the investments in the tourism and manufacturing sectors, respectively, where $i_{t}^{T} = k_{t}^{T} - (1 - δ) k_{t - 1}^{T}, i_{t}^{M} = k_{t}^{M} - (1 - α) k_{t - 1}^{M}$ .

It is worth noting that in departing from the common aggregate resource constraint of a closed economy, the revenue from inbound tourism taxation $τ_{t}^{f} p_{t}^{T} c_{t}^{f}$ enters the aggregate resource constraint as an additional source of income. This addition will affect the relative price and the economy’s resource allocation, which will be discussed in detail when we present the results.

Equilibrium condition

The equilibrium in this model economy is defined by a sequence of prices ${p_{t}^{T}, p_{t}^{M}, w_{t}^{T}, w_{t}^{M}, r_{t}^{T}, r_{t}^{M}}$ , real allocations ${y_{t}^{T}, y_{t}^{M}, c_{t}^{T}, c_{t}^{M}, c_{t}^{f}, i_{t}^{T}, i_{t}^{M}, k_{t}^{T}, k_{t}^{M}, l_{t}^{T}, l_{t}^{M}}$ and policy variables ${τ_{t}^{f}, s_{t}^{Y}, s_{t}^{M}}$ that satisfy the following: (i) the optimization problems of households, (ii) the optimization problems of tourism and manufacturing firms, (iii) the budgetary constraints of households and the government, (iv) the market-clearing condition for tourism goods and (v) the aggregate resource constraint.

Model parameterization and results

Due to the complexity of our specified model, we are not able to solve it analytically. Therefore, we turn to numerical simulations. It is widely known that Dynare is a powerful software platform for handling a broad class of economic models, especially DSGE models. We, therefore, resort to the Dynare toolbox for all numerical calculations for the specified model (Collard and Juillard, 2009).

Model calibration

The economy we are investigating in this study is Thailand, a major tourism-based country in the Southeastern Asian region. Other countries in this region like Malaysia and the Philippines can also serve our purpose, but the main reason we selected Thailand is due to the size of its tourism sector and its economic importance in the region. The calibration procedure is based on the statistical data and previous studies in general and for Thailand in particular, which is a standard practice to calibrate parameter values (Gomme and Rupert, 2007). The macro data for unknown parameters are mainly obtained from Aroonrueangaram (2013) covering the period 2002–2012. At the outset, we specify various parameter values throughout various equations proposed. These include the following: β, households’ subjective discount factor indicating the patience of households regarding the future; σ, the inverse of the elasticity of intertemporal substitution of consumption in different time periods; θ, the inverse Frisch elasticity of the labour supply; ψ, the weights of labour in the utility; $ϕ^{c}$ and $ϕ^{l}$ , the shares of consumption and labour in the tourism sector, respectively; $χ^{c}$ and $χ^{l}$ , the elasticities of substitution of domestic consumption and labour services, respectively; $α^{T}$ and $α^{M}$ , the labour shares in tourism and manufacturing goods’ productions, respectively; δ, the capital depreciation rate; η and ξ, the price and exogenous factor (income) elasticity of foreign tourism demand, respectively; and ρ, the $AR (1)$ coefficient of inbound tourism demand measuring the persistence of demand.

Since the model is specified on a quarterly basis, the discount factor β is set to 0.99, which corresponds to a steady-state real interest rate of 4% (Alp and Elekdag, 2012). We set $σ = 1.5$ , which is common in open economy RBC models (García Cicco et al., 2010), and $θ = 2$ , which is consistent with the empirical results that the Frisch elasticity of labour supply is 0.5 (Keane and Rogerson, 2015). We calibrate ψ so that the steady-state labour hours are approximately one-third of the time endowment, which is consistent with the 8-h daily working time. According to the World Travel & Tourism Council’s Global Economic Impact & Issues 2018 report, the average total value of the tourism sector relative to those countries’ GDP is approximately 20%, and the employment ratio is approximately 15%. Therefore, we set $ϕ^{c} = 0.127$ to account for the average consumption GDP ratio and set $ϕ^{l} = 0.15$ . The labour income share in the manufacturing sector is set, according to Alp and Elekdag (2012), at $α^{M} = 0.5$ . Considering that the tourism sector is service-based and more labour intensive than the manufacturing sector, after a few computational steps in the model’s steady states, we get $α^{T} = 0.52$ . Assuming that capital depreciates by 10% annually indicates that $δ = 0.025$ (García Cicco et al., 2010). For the elasticity of substitution between tourism consumption (labour) and manufacturing consumption, we have no exact prior information on the parameter values; however, we are certain that the elasticity should be low according to previous studies (Stockman and Tesar, 1995), so we set $χ^{c} = 0.5$ and $χ^{l} = 0.5$ . The previous estimates of the price elasticity of foreign tourism demand have a wide range of results in different countries, and we set $η = 0.9$ as an average value of those estimates (Untong et al., 2014). The elasticity of the exogenous factor is a free parameter to set. We assume that the exogenous factor mainly comes from the income of foreign tourists (Peng et al., 2015), which is exogenous to the modelled economy in the destination country and set at $ξ = 2$ , implying that international tourism is a luxury good. Finally, the persistence of foreign tourists’ demand shock is set to $ρ = 0.7$ as a benchmark, and we experiment with different values (Assaf et al., 2012). The parameter values are summarized in Table 2.

Table 2.

Model description and parameterization.

Parameter	Description	Value	Source
β	Households’ discount factor	0.9923	Alp and Elekdag (2012)
σ	Intertemporal elasticity of substitution	1.5	Garcia Cicco et al. (2010)
θ	Inverse of Frisch elasticity of labour supply	2	Keane and Rogerson (2015)
ψ	Weight on labour in utility	49	Calculated based on steady state
$ϕ^{c}$	Share of domestic tourism consumption	0.127	WTTC (2018)
$ϕ^{l}$	Share of labour in tourism sector	0.15	WTTC (2018)
$χ^{c}$	Elasticity of substitution in domestic consumption	0.5	Stockman and Tesar (1995)
$χ^{l}$	Elasticity of substitution in labour	0.5	Stockman and Tesar (1995)
$α^{T}$	Labour share in tourism good production	0.52	Calculated based on steady state
$α^{M}$	Labour share in manufacturing good production	0.5	Alp and Elekdag (2012)
δ	Capital depreciation rate	0.025	Alp and Elekdag (2012)
η	Price elasticity of foreign tourism demand	0.9	Untong et al. (2014)
ξ	Exogenous factor elasticity of foreign tourism demand	2	Untong et al. (2014)
ρ	Persistence of foreign tourists demand shock	0.7	Assaf et al. (2012)

Note: WTTC: World Travel & Tourism Council.

Numeric simulation

Equipped with all the parameter values, we are ready to numerically solve and simulate the model when the system is externally shocked by the tourism demand from foreign tourists $ε_{t}^{f}$ , which is assumed to be the only stochastic disturbance considered in this study. To see the effects of the shock to the whole system, we rely on impulse response functions (IRFs) to display the results. IRFs describe the evolution of the variables of interest in the system along with a specified time horizon after a shock in a given moment (Cogley and Nason, 1995). With IRFs, we can clearly unveil the dynamics of the variables of interests, thus confronting the boom in the tourism sector that resulted from an inflow of foreign tourists.

Benchmark results

In the DSGE model, we describe the mechanisms through which foreign tourists’ demand shock affects the dynamics of the destination country’s economy when the government transfers tax proceeds to domestic households. This shock may occur as a result of either an increase in foreigners’ income or domestic policies designed for attracting foreign visitors. Figure 2 provides the IRFs of the model’s main endogenous variables to a 1% positive shock to foreign tourists’ demand from the benchmark calibration; the ordinate shows the variables’ percentage deviations from their steady states. In the benchmark case, we find that a positive demand shock raises the consumption of foreign tourists ( $c_{t}^{f}$ ) and brings a boom to the tourism sector that the relative price $(p_{t}^{T})$ rises, and that ultimately, the production of tourism goods $(y_{t}^{T})$ increases. Higher production requires more labour $(l_{t}^{T})$ and capital $(k_{t}^{T})$ in the tourism sector, and thus, sectoral wages $(w_{t}^{T})$ and investments $(i_{t}^{T})$ increase. By contrast, the manufacturing sector is contracting. We observe some level of decline in sectoral production $(y_{t}^{M})$ , labour $(l_{t}^{M})$ , wages $(w_{t}^{M})$ and investments $(i_{t}^{M})$ . After being hit by the external tourism shock, there is a reallocation of resources from the manufacturing sector to tourism sector, thus causing deindustrialization as the typical symptom of the Dutch disease in the destination economy (Corden and Neary, 1982).

Figure 2.

Impulse responses of the system to external tourism demand shock (different shock persistence).

Surprisingly, the occurrence of the Dutch disease does not necessarily cause an automatic welfare loss.⁷ After the economy is hit by highly persistent shocks to foreign tourism demand, long-lasting foreign tourist arrivals bring huge and persistent amounts of foreign exchange to the destination country, which ultimately reaches households’ pockets through government transfers. The strong wealth effect increases domestic consumers’ consumption of goods and leisure, so the consumption of manufacturing goods $(c_{t}^{M})$ increases, whereas labour $(l_{t}^{M})$ contracts. The consumption of tourism goods from domestic consumers $(c_{t}^{T})$ declines because of the higher price of tourism goods, and apparently, the substitution effect dominates. However, the overall production $(y_{t})$ , investment $(i_{t})$ and capital $(k_{t})$ still increase, implying that the deindustrialization process only partially offsets the boom in the tourism sector and does not necessarily lead to the contraction of the overall economy. The overall higher consumption $(c_{t})$ and lower labour $(l_{t})$ over time make households better off.

Results with different shock persistence

Next, we conduct another simulation experiment with different persistence levels of foreign tourism demand. By persistence, we mean the time required for the shock to disappear entirely: the larger the persistence is, the longer the time needed for the shock to disappear altogether. In fact, when tourism firms and other stakeholders have a clear appreciation of the volatility in tourism demand, they may adopt strategies to reap the benefits of positive effects or avoid being victimized by a negative effect. By confronting the shock with different levels of persistence, the economy may respond in quite different ways. In our case, the parameter capturing the persistence of foreign tourists demand shock, ρ, can be empirically calibrated from the first-order autocorrelation coefficient in the time-series analysis of inbound tourism demand (George Assaf et al., 2010). Although important, the exact value of ρ is not without controversies. Assaf et al. (2012) find mixed degrees of persistence for tourists from different countries, but most of the sample countries have autocorrelation coefficients between 0.5 and 0.9. Barros et al. (2014) studied the persistence characteristics of tourist arrivals from four Nordic countries in Madeira and find similar results. Therefore, we raise the autocorrelation coefficient to $ρ = 0.9$ and lower it to $ρ = 0.5$ to see how the economy responds to different specifications. Figure 2 also plots the IRFs for the three specifications. All variables move in a very similar way, but the results with less persistent shocks (smaller ρ values) fade out faster over time. In general, the results indicate that the Dutch disease exists in all three cases in the sense that the resources shift away from the manufacturing sector to the tourism sector and that this relocation causes deindustrialization. However, the total output $(y_{t})$ , investment $(i_{t})$ and capital $(k_{t})$ all rise after the impulse as an external shock from inbound tourism, implying that the boom in the tourism sector is large enough to fully offset the bust in the manufacturing sector and that the overall economy activity is up from the steady state. The overall economy is in an expansion despite the occurrence of the Dutch disease.

Results with production subsidy

Because of the Dutch disease, the government is motivated to balance the development of the manufacturing and tourism sectors. For this purpose, the government may collect tax revenues from foreign tourists and provide subsidies to manufacturers instead of households as the first option. With production subsidies, a manufacturing firm’s marginal product would be higher than without, so at the optimum, the manufacturing firm hires more labour and uses more capital in the production. Hence, the output will be higher, and wages and capital returns will also rise. All these results are reflected in Figure 3, in which the IRFs of the benchmark (solid) and production subsidy cases (dashed) are plotted. From the IRFs, we can clearly see that the effect is very modest and that the main impact results from manufacturing labour $(l_{t}^{M})$ and output $(y_{t}^{M})$ , both of which are slightly higher than those in the benchmark scenario (responding slightly less negatively). Furthermore, manufacturing wages $(w_{t}^{M})$ and capital returns $(r_{t}^{M})$ respond in the direction that was expected. Most of the variables from the tourism sector (e.g. $y_{t}^{M}$ , $c_{t}^{M}$ and $k_{t}^{M}$ ) barely deviate from their counterparts in the benchmark scenario. Overall, the Dutch disease is quantitatively modestly mitigated by manufacturing production subsidies, but it still exists.

Figure 3.

Impulse responses of the system to external tourism demand shock (benchmark vs. subsidies).

Results with investment subsidy

Alternatively, the government has a second option to subsidize the manufacturing sector: by collecting tax revenues from foreign tourists and subsidizing households who invest in manufacturing capital. With investment subsidies for the manufacturing sector, households will find it attractive to postpone consumption for investment, particularly for goods in the manufacturing sector. Figure 3 plots the IRFs of the model’s main endogenous variables in the investment subsidy scenario (dotted). We can observe that the dynamics of the model economy are dramatically changed. From the manufacturing sector, the subsidy would postpone consumption $(c_{t}^{M})$ and strongly stimulate investment $(i_{t}^{M})$ even more than fully offsetting the impact of a tourism boom that switches resources out of the manufacturing sector. Along with the positive response of manufacturing investment, manufacturing capital $(k_{t}^{M})$ also increases after the shock. Nevertheless, manufacturing output $(y_{t}^{M})$ declines initially due to decreased labour $(l_{t}^{M})$ , but it quickly expands because of the higher capital stock. By contrast, the tourism sector is not as prosperous as in the benchmark scenario, and output $(y_{t}^{T})$ , investment $(i_{t}^{T})$ and capital $(k_{t}^{T})$ are all quite calm after the shock. Consequently, the interaction between the two sectors raises total consumption $(c_{t})$ and lowers investment $(i_{t})$ . Higher investment transfers to more capital $(k_{t})$ , thus leading to a higher marginal product of labour and higher wages $(w_{t})$ . To summarize, the Dutch disease symptoms are largely alleviated using the manufacturing investment subsidies.

Discussion about welfare

As we argued, if the government is concerned about the reallocation of resources between the tourism and manufacturing sectors, it can subsidize the manufacturing sector instead of making transfers to households when confronting the shock. The Dutch disease can potentially be mitigated in this way. The first option is production subsidies, which mainly work as an intra-temporal factor. This option can raise consumption and labour slightly compared to those in the benchmark scenario, but the overall effect is very modest. The Dutch disease weakens only marginally, and welfare improves marginally also. On the other hand, investment subsidies, as the second option, play a much stronger role intertemporally in postponing consumption for investment. With strong motives to invest, manufacturing firms would produce more while simultaneously hiring more labour. Therefore, immediately after the shock, consumption is lower and labour is higher compared to the benchmark scenario, and welfare is lower at the beginning. As time goes by, higher investment transfers to larger stocks of capital, and consumption is persistently above the steady state with the support of large amounts of capital. Welfare also strongly improves persistently. To sum up, investment subsidies lower the welfare at the beginning, but they keep it above the steady state for a longer period. If the government wants to take advantage of the external tourism boom and get the most out of it over the long term, it should turn to investment subsidies. Production subsidies are the choice for short-term purposes. Table 3 summarizes the results of the modelled economy confronting the external tourism boom. The Dutch disease occurs completely (output, investment and capital) in the first two scenarios, while investment subsidies cure the symptoms related to investment and capital, only leaving output for a very short period of time (2–3 quarters). Welfare gains can be obtained in all three scenarios, but there is a trade-off between production and investment subsidies. Production subsidies are better if short run gains are preferred, while investment subsidies outperform if the long term is the focus. The comparison is shown in Figure 4 with a longer time horizon to display variables’ tendencies to revert to the steady state.

Table 3.

Comparison of three different scenarios.

		Benchmark (w/o subsidy)	Production subsidy	Investment subsidy
Dutch disease	Existence	Yes	Yes	Yes, but only in production
Dutch disease	Magnitude*		Slightly weaker	Very short lived
Welfare gain	Existence	Yes	Yes	Yes
Welfare gain	Magnitude*		Slightly stronger	Weaker in short term and stronger over long term

* The magnitude comparison based on the benchmark results.

Figure 4.

Impulse responses to external tourism demand shock (welfare with longer horizon).

Conclusion

We developed a two-sector DSGE model, and we numerically solved and simulated the model to understand the tourism-induced Dutch disease in a small open economy like Thailand. Our results confirmed the existence of the tourism-induced Dutch disease highlighted in past studies (Capó et al., 2007; Forsyth et al., 2014; Inchausti Sintes, 2015; Pham et al., 2015). Through the dynamic analysis using IRFs and structure modelling of each agent in the economy, our results provided a more holistic view of the channels contributing to the tourism-induced Dutch disease as well as the dynamic effects of external tourism booms on a small open economy. Nevertheless, our results show that an externally driven tourism boom can actually improve welfare, albeit with the coexistence of the Dutch disease. When the government transfers tax proceeds to households, there is a wealth effect leading to higher consumption and lower labour over time after the shock hitting the economy, thereby making the households better off.

Sheng (2011) theoretically examined, in a general equilibrium framework, the impacts of combining tourism taxes and non-tourism subsidies on a small tourism-dependent economy. He found that the policy is not effective, but claimed that the issue needs to be further investigated through solid empirical studies. Based on the benchmark DSGE analysis, we further extended the model to understand the impacts of two government policies on alleviating the Dutch disease. To maintain resources within the manufacturing sector, the government can promote it using either production subsidies or investment subsidies. After comparison with model simulation, we found that production subsidies only work marginally, whereas investment subsidies can almost overturn the Dutch disease. In terms of welfare, production subsidies are mainly effective over the short term, while investment subsidies work better in the long run. For policy implication, there is a trade-off for the policymakers to cure tourism-induced Dutch disease and make tourism development sustainable over time. In order to contain the symptom, the short-term welfare has to be sacrificed, while an overlook of Dutch disease would ultimately harm the economy over the long term. The exact implementation of policy may need a more comprehensive calculation of the trade-off.

Our work provides the initial step to bring the DSGE framework into the study of the economics of tourism and its relationship with other sectors in the overall economy. As a preliminary research endeavour, this article did not consider a hybrid subsidy policy that includes a combination of investment and production subsidies. The ratio of the two subsidies can be endogenous based on the goal of the government. Our results show a great potential for studying both theoretical and empirical tourism issues by applying the DSGE methodology, and there is plenty of room for further extension. For instance, it will be particularly interesting to see how different policies that are used to promote the tourism sector perform over time, how the strengthened tourism sector can give a feedback effect to the rest of the economy and how the tourism sector behaves as part of the transmission mechanism absorbing or propagating other shocks that are hitting the economy, among other issues. We will leave these for future research endeavours.

Supplemental material

Supplemental Material, Appendix - Prescribing for the tourism-induced Dutch disease: A DSGE analysis of subsidy policies

Supplemental Material, Appendix for Prescribing for the tourism-induced Dutch disease: A DSGE analysis of subsidy policies by Hongru Zhang and Yang Yang 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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Social Science Foundation of China under Grant 17AJY030 and Jiangxi University of Finance and Economics Faculty Research Seed Fund under Grant K62192012.

ORCID iD

Yang Yang

Supplemental material

Supplemental material for this article is available online.

Notes

References

Adolfson

Lindé

Villani

(2007) Forecasting performance of an open economy DSGE model. Econometric Reviews 26: 289–328.

Alp

Elekdag

(2012) Shock therapy! What role for Thai monetary policy? IMF Working Paper 12269.

Aroonrueangaram

(2013) Inflation and asset returns in a new Keynesian model for Thailand, including the housing sector. Master Thesis, Thammasat University, Thailand.

Assaf

Gil Alana

Barros

(2012) Persistence characteristics of tourism arrivals to Australia. International Journal of Tourism Research 14: 165–176.

Balaguer

Cantavella Jordá

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

Bank of Thailand (2016) Yearly report on economic and monetary conditions in 2016. Available at: https://www.bot.or.th/English/MonetaryPolicy/EconomicConditions/AnnualReport/Pages/default.aspx

Barros

Gil Alana

Gonçalves

(2014) Persistence characteristics of Nordic tourist arrivals in Madeira and their forecasting. Scandinavian Journal of Hospitality and Tourism 13: 44–57.

Candela

Castellani

Mussoni

(2015) Keynesian policies for tourism: Taxation without coordination. Tourism Economics 21: 527–541.

Capó

Font

Nadal

(2007) Dutch disease in tourism economies: Evidence from the Balearics and the Canary Islands. Journal of Sustainable Tourism 15: 615–627.

10.

Chang

(2011) Congestion externalities of tourism, Dutch disease and optimal taxation: Macroeconomic implications*. Economic Record 87: 90–108.

11.

Chao

Hazari

Laffargue

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

12.

Chen

Lai

Chu

(2016) Welfare effects of tourism-driven Dutch disease: The roles of international borrowings and factor intensity. International Review of Economics and Finance 44: 381–394.

13.

Chou

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

14.

Christiano

Eichenbaum

Evans

(2005) Nominal rigidities and the dynamic effects of a shock to monetary policy. Journal of Political Economy 113: 1–45.

15.

Cogley

Nason

(1995) Output dynamics in real-business cycle models. American Economic Review 85: 492–511.

16.

Collard

Juillard

(2009) Stochastic simulations with Dynare: A practical guide. Dynare Tutorial. Available at: http://www.dynare.org/documentation-and-support/tutorial

17.

Copeland

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

18.

Corden

Neary

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

19.

Fernández Villaverde

Rubio Ramírez

Schorfheide

(2016) Solution and estimation methods for DSGE models In: Taylor

Uhlig

(eds) Handbook of Macroeconomics. Elsevier, pp. 527–724.

20.

Forsyth

Dwyer

Spurr

(2014) Is Australian tourism suffering Dutch disease? Annals of Tourism Research 46: 1–15.

21.

García Cicco

Pancrazi

Uribe

(2010) Real business cycles in emerging countries? American Economic Review 100: 2510–2531.

22.

George Assaf

Pestana Barros

Gil Alana

(2010) Persistence in the short- and long-term tourist arrivals to Australia. Journal of Travel Research 50: 213–229.

23.

Gomme

Rupert

(2007) Theory, measurement and calibration of macroeconomic models. Journal of Monetary Economics 54: 460–497.

24.

Gössling

Peeters

(2015) Assessing tourism’s global environmental impact 1900–2050. Journal of Sustainable Tourism 23: 639–659.

25.

Hashimzade

Thornton

(2013) Handbook of Research Methods and Applications in Empirical Macroeconomics. Cheltenham: Edward Elgar.

26.

Hazari

A-Ng (1993) An analysis of tourists’ consumption of non-traded goods and services on the welfare of the domestic consumers. International Review of Economics and Finance 2: 43–58.

27.

Hazari

Sgro

(1995) Tourism and growth in a dynamic model of trade. The Journal of International Trade and Economic Development 4: 243–252.

28.

Holzner

(2011) Tourism and economic development: The beach disease? Tourism Management 32: 922–933.

29.

Iglesias

Gegundez

Golpe

, et al. (2018) How do foreign income shocks affect the magnitude of Spanish tourism? Tourism Economics 24: 839–871.

30.

Inchausti Sintes

(2015) Tourism: Economic growth, employment and Dutch disease. Annals of Tourism Research 54: 172–189.

31.

Jehle

Reny

(2011) Advanced Microeconomic Theory, 3rd ed. Upper Saddle River: Prentice Hall.

32.

Keane

Rogerson

(2015) Reconciling micro and macro labor supply elasticities: A structural perspective. Annual Review of Economics 7: 89–117.

33.

Kostakis

Theodoropoulou

(2017) Spatial analysis of the nexus between tourism–human capital–economic growth: Evidence for the period 2000–2014 among NUTS II Southern European regions. Tourism Economics 23: 1523–1534.

34.

Kydland

Prescott

(1982) Time to build and aggregate fluctuations. Econometrica 50: 1345–1370.

35.

Lanza

Temple

Urga

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

36.

Lee

(2016) The socio-cultural impact of Growth-Pole Theory (GPT) tourism development. Anatolia 27: 268–270.

37.

Lin

Yang

(2019) Where can tourism-led growth and economy-driven tourism growth occur? Journal of Travel Research 58(5): 760–773.

38.

Liu

Song

Blake

(2018) Modelling productivity shocks and economic growth using the Bayesian dynamic stochastic general equilibrium approach. International Journal of Contemporary Hospitality Management. DOI: 10.1108/IJCHM-10-2017-0686.

39.

Lucas

(1976) Econometric policy evaluation: a critique. Carnegie-Rochester Conference Series on Public Policy 1: 19–46.

40.

Mattsson

Orfila Sintes

(2014) Hotel innovation and its effect on business performance. International Journal of Tourism Research 16: 388–398.

41.

Mendoza

(1991) Real business cycles in a small open economy. American Economic Review 81: 797–818.

42.

Meng

Pham

(2017) The impact of the Australian carbon tax on the tourism industry. Tourism Economics 23: 506–522.

43.

Nowak

Sahli

(2007) Coastal tourism and ‘dutch disease’ in a small island economy. Tourism Economics 13: 49–65.

44.

Nowak

Sahil

Cortes Jimenez

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

45.

Pablo-Romero

del

Molina

(2013) Tourism and economic growth: A review of empirical literature. Tourism Management Perspectives 8: 28–41.

46.

Pagan

(2003) Report on modelling and forecasting at the bank of England. Bank of England Quarterly Bulletin 43: 1–29.

47.

Peng

Song

Crouch

, et al. (2015) A meta-analysis of international tourism demand elasticities. Journal of Travel Research 54: 611–633.

48.

Pham

Jago

Spurr

, et al. (2015) The Dutch disease effects on tourism–the case of Australia. Tourism Management 46: 610–622.

49.

Shen

Yang

SCS

Zanna

(2018) Government spending effects in low-income countries. Journal of Development Economics 133: 201–219.

50.

Sheng

(2011) Taxing tourism and subsidizing non-tourism: A welfare-enhancing solution to “Dutch disease”? Tourism Management 32: 1223–1228.

51.

Sheng

(2017) Factors determining the success or failure of a tourism tax: A theoretical model. Tourism Review 72: 274–287.

52.

Sheng

Tsui

(2009) Taxing tourism: enhancing or reducing welfare? Journal of Sustainable Tourism 17: 627–635.

53.

Sihabutr

(2012) Technology and international specialisation in tourism. PhD Thesis, Université Toulouse le Mirail - Toulouse II.

54.

Smets

Wouters

(2004) Forecasting with a Bayesian DSGE model: An application to the Euro Area. Journal of Common Market Studies 42: 841–867.

55.

Song

Dwyer

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

56.

Stockman

Tesar

(1995) Tastes and technology in a two-country model of the business cycle: Explaining international comovements. American Economic Review 85: 168–185.

57.

(2017) The impact of tourism receipt fluctuations on Taiwan’s macroeconomy. Institute of Economics. Master Thesis, National Sun Yat-sen University.

58.

Untong

Ramos

Kaosa Ard

, et al. (2014) Thailand’s long-run tourism demand elasticities. Tourism Economics 20: 595–610.

59.

UNWTO/GTERC (2017) Annual report on Asia tourism trends. Available at: https://www.e-unwto.org/doi/book/10.18111/9789284419142

60.

Wattanakuljarus

Coxhead

(2008) Is tourism-based development good for the poor? Journal of Policy Modeling 30: 929–955.

61.

Wickens

(2014) How useful are DSGE macroeconomic models for forecasting? Open Economies Review 25: 171–193.

62.

WTTC (2018) Travel & Tourism economic impact 2018 Thailand. Available at: https://www.wttc.org/-/media/files/reports/economic-impact-research/countries-2018/thailand2018.pdf

63.

Yang

Fik

(2014) Spatial effects in regional tourism growth. Annals of Tourism Research 46: 144–162.

64.

Yang

(2019) Public transport connectivity and intercity tourist flows. Journal of Travel Research 58(1): 25–41.

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

0.36 MB