Based on the articles you have read for Week 2 (see Modules – Week 2), please specify and discuss four points on how sustainable tourism may contribute to

Annals of Tourism Research, Vol. 39, No. 3, pp. 1653–1682, 2012 0160-7383/$ – see front matter � 2012 Elsevier Ltd. All rights reserved.

Printed in Great Britain

http://dx.doi.org/10.1016/j.annals.2012.05.023 www.elsevier.com/locate/atoures

Review Article

TOURISM ECONOMICS RESEARCH: A REVIEW AND ASSESSMENT

Haiyan Song The Hong Kong Polytechnic University, Hong Kong

Larry Dwyer University of New South Wales, Australia

Gang Li ZhengCao

University of Surrey, United Kingdom

Abstract: This paper aims to provide the most up-to-date survey of tourism economics research and to summarise the key trends in its recent development. Particular attention is paid to the research progress made over the last decade in respect of approaches, methodo- logical innovations, emerging topics, research gaps, and directions for future research. Remarkable but unbalanced developments have been observed across different sub-research areas in tourism economics. While neoclassical economics has contributed the most to the development of tourism economics, alternative schools of thought in economics have also emerged in advancing our understanding of tourism from different perspectives. As tourism studies are multi- and inter-disciplinary, integrating economics with other social sci- ence disciplines will further contribute to knowledge creation in tourism studies. Keywords: tourism economics, research integration, demand, supply, impact, econometric model. � 2012 Elsevier Ltd. All rights reserved.

INTRODUCTION

Tourism, despite the ongoing debates about its definition over the past decades, is commonly recognised as a human activity that defines the demand for and supply of its products and the usage of resources that may result in either positive or negative socioeconomic consequences at both national and international level. The signifi- cance of the economic approach and perspective to understanding this human activity is widely known. As far as both its demand and supply are concerned, tourism has distinct characteristics which set it apart

Haiyan Song, PhD, (<[email protected]>) is chair professor of tourism in the School of Hotel and Tourism Management, Hong Kong Polytechnic University. Larry Dwyer, PhD, is professor of travel and tourism economics, School of Marketing, University of New South Wales, NSW, Australia. Gang Li, PhD, is Reader in economics, School of Hospitality and Tourism Management, University of Surrey, UK. Zheng Cao is a PhD candidate in the School of Hospitality and Tourism Management, University of Surrey, UK.

1653

1654 H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682

from other economic activities (Stabler, Papatheodorou, & Sinclair, 2010). Studying the characteristics of tourism from the economic per- spective is a relatively new area of research pioneered by Guthrie (1961), Gerakis (1965), and Gray (1966). Propelled by the tremendous evolution of tourism as an economic activity over the past 50 years, there has also been a remarkable growth, in terms of number of pub- lications, in tourism economics research. This trend has been even more pronounced since the 1990s with the establishment of Tourism Economics, a scientific journal devoted entirely to the publications of re- search outputs in this field. More recently, the publication of key texts on the economics of tourism, such as Dwyer, Forsyth, and Dwyer (2010), Stabler et al. (2010), and Tribe (2011), has marked the matu- rity of tourism economics as a field of study comprising comprehensive bodies of knowledge and theoretical foundations in the context of tourism.

The dynamics of tourism, as an activity and as an industry, call for continuous efforts in seeking new approaches, tools, and perspectives in order to acquire new knowledge and a greater understanding of the discipline. Therefore, it is both necessary and useful to comprehen- sively review the development of the research field in terms of where we were, where we are, and where we should be. Very few such endeav- ours have been made in this regard. Eadington and Redman’s (1991) work represents the earliest attempt to provide an overview of the developments in tourism economics. Key research areas identified in his review include demand elasticities and their modelling techniques, market structure and ownership, economic impacts, and policies. His recommendations for further research were in such directions as in- ter-sectoral linkages, the integration of economic models and statistical techniques for demand analysis, and the development of national and regional input-output (I-O) models for economic impact assessment. Sinclair (1998) surveys the literature over a period of two decades, highlighting some new developments, such as the system-of-equation approach to demand analysis and computable general equilibrium (CGE) modelling for economic impact assessment. These develop- ments echo Eadington and Redman’s (1991) earlier recommenda- tions. Tremblay (1998) focuses his review on different perspectives on industrial organisation and recommends institutional and network approaches. In addition, Sinclair (1998) directs scholars’ attention to the environmental issues related to sustainable tourism development. She highlights the fact that impact analysis was limited to the use of I-O tables, with CGE models receiving little attention, and the neglect of taxation and regulatory policy in relation to environmental issues. In another review, Sinclair, Blake, and Sugiyarto (2003) argue that re- search in tourism economics has been dominated by demand analysis, while little attention has been paid to the determinants of tourism sup- ply, including different forms of tourism business integration. More re- cently, Dwyer, Forsyth, and Papatheodorou (2011) have provided an overview of the state of research and the key developments in tourism economics, including perspectives on the implications for research of the recent global financial crisis. In their reviews, Li, Song, and Witt

H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682 1655

(2005) and Song and Li (2008) mainly concentrate on the methodo- logical developments in tourism demand studies.

This paper aims to provide the most up-to-date survey of tourism eco- nomics research, highlighting recent developments and likely future directions. Drawing on the latest publications up to the end of 2011, mainly from key tourism journals such as Annals of Tourism Research, Tourism Economics, Tourism Management, and the Journal of Travel Re- search, this review tracks the historical developments in each of the key research areas, paying particular attention to the research progress made over the last decade in terms of economic approaches, method- ological innovations, emerging topics, and directions for future research.

DEMAND

Demand analysis has the longest history in tourism economics re- search and has undergone remarkable developments in terms of diver- sity of interests, depth of theoretical foundations, and advances in research methods (Li et al., 2005). Its dominant position, noted by Sin- clair et al. (2003), is still observable in the latest developments in tour- ism economics. Based on the latest empirical evidence, the following section focuses on the issues in demand analysis that have emerged since the publication of previous review (e.g., Li et al., 2005; Song & Li, 2008).

Demand and its Determinants

The theoretical argument of tourism demand under neoclassical economic theory usually assumes a multi-stage budgeting process. Two pillars of the assumption are the composite commodity theorem and the separability of preferences. The composite commodity theo- rem states that various commodities can be aggregated to broad bun- dles of products, provided that prices within a bundle move in parallel. The separability of preferences means preferences within one bundle can be described independently of those in another one (Smeral & Weber, 2000). In the tourism context, such a multi-stage process implies that a typical tourist will firstly allocate the total budget over several time periods, then separate the goods into leisure goods and other consumption goods, and further choose among domestic trips, international travels and other activities within the leisure goods bundle. In the last stage, the destination country/region is deter- mined. It is also noted by Smeral and Weber (2000) that the decision at each stage can be thought of as corresponding to a utility maximiza- tion problem of its own, where the income effect and price effect are implicated in empirical models.

Tourism demand is predominantly measured by the number of arriv- als and the level of tourist expenditure (receipts), along with their vari- ations, in per capita terms (Song, Li, Witt, & Fei, 2010; Song, Witt, & Li,

1656 H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682

2009). One alternative measure, the number of tourist nights (length of stay), has appeared in recent studies. For example, Gokovali, Bahar, and Kozak (2007), Martinez-Garcia and Raya (2008), and Barros and Machado (2010) adopt survival analysis (duration model) based on Lancaster’s characteristics framework to examine the determinants of this measure.

According to the demand theory, the pivotal factors shaping a tour- ist’s budget line are the income of the consumer and the price of the tourism product/service. Specifically, in empirical tourism demand studies, the income of origin country/region, the own price of a desti- nation, and the substitute prices of alternative destinations are the most commonly considered determinants (Song, Witt, & Li, 2009). De- mand elasticities are thus of particular significance, and these have been one of the focuses of the published studies on demand analysis. The latest development is to construct confidence intervals for demand elasticity estimates using the bias-corrected bootstrap method (Song, Kim, & Yang, 2010). This method overcomes the limitation of the tra- ditional point estimates, which neglect the degree of variability and are thus less informative.

An additional variable that affects tourists’ decisions (though not linked to the budget line) is the marketing expenditure of the tourism product/service provider (at both destination level and firm level). However, the difficulty in accessing the relevant marketing data hin- ders its application in most empirical studies (Kulendran & Dwyer, 2009; Zhang, Kulendran, & Song, 2010). The magnitude of the effect of marketing expenditure has been found to be as low as several hundredths.

Beyond the neoclassical theory, Lancaster’s (1966) characteristics framework explores the individual’s consumption of specific fea- tures/attributes, through which he/she attains satisfaction and utility. Applications of the hedonic pricing approach in the context of tourism demand mainly focus on the prices of tour packages (e.g., Aguilo, Alegre, & Riera, 2001; Chen & Rothschild, 2010; Papatheodorou, 2002; Sinclair, Clewer, & Pack, 1990; Thrane, 2005). The public good components (e.g., cultural legacy, public safety, and public infrastruc- ture) embedded in tourism products have been considered in recent studies (e.g., Rigall-I-Torrent & Fluvia, 2007, 2011). The difficulties faced in using this approach include the selection of the appropriate explanatory variables and the potential multicollinearity problem among the variables (Chen & Rothschild, 2010; Sinclair et al., 1990).

Methodological Developments

Since the 1990s, demand modelling studies have shifted from the use of static regression models to a range of sophisticated dynamic specifi- cations. Dynamics in tourism demand are often accounted for by re- peat visits, word-of-mouth recommendations, time lags in implementing a decision, information asymmetry, supply rigidities, and long-term adjustments (Morley, 2009). A recent development in

H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682 1657

dynamic modelling is the integration of the time-varying-parameter (TVP) technique and the causal structural time series model (Song, Li, Witt, & Athanasopoulos, 2011), which combines the technical advantages of both methods and shows superior forecasting perfor- mance. Recently, a new cointegration-error correction method—the bound test of Pesaran, Shin, and Smith (2001)—has been applied to tourism (e.g., Halicioglu, 2010; Song, Lin, Witt, & Zhang, 2011). This is a test to detect the long-run co-integration relationship among vari- ables in a demand model. Its advantage lies in its ability to accommo- date variables with different integration orders.

Developments in system-of-equations approaches, such as the vector autoregressive model (VAR) and the almost ideal demand system (AIDS), have expanded the dimensions of investigations. These ap- proaches overcome the limitations of the single-equation methods by bringing solid theoretical foundations to tourism demand modelling and forecasting exercises. The AIDS model, designed to analyse the interdependence of budget allocations and different consumer goods/services, has received much attention over the past decade. The dynamic forms of AIDS, coupled with the error correction mechanism and TVP technique, represent the latest development of system-of-equation methods. Their applications shed new light on the substitution and complementary effects between destinations (e.g., Cortes-Jimenez, Durbarry, & Pulina, 2009; Li, Song, & Witt, 2006) or between consumption categories (Wu, Li, & Song, 2011, 2012) and destination competitiveness (Mangion, Durbarry, & Sinclair, 2005). VAR models have received relatively little attention in tourism until recently (e.g., Seetanah & Khadaroo, 2009; Song & Witt, 2006; Torraleja, Vazquez, & Franco, 2009). Panel data analysis techniques have not been widely applied in tourism demand research with only a few exceptions (e.g., Garin-Munoz, 2009; Ledesma-Rodriguez, Navarro-Ibanez, & Perez-Rodriguez, 2001; Naude & Saayman, 2005; Seetaram, 2010). Future studies should pay more attention to the dynamic version of panel data analysis and to more advanced estimation methods such as the fully modified Ordinary Least Square estimator (Pedroni, 2004).

In addition to selecting the best specified models for modelling and forecasting tourism demand, identifying the key economic determi- nants of tourism demand, calculating the demand elasticities, and eval- uating the forecasting performance of the demand models are the key research tasks in tourism demand studies. The conclusion based on the empirical evidence is that no single model can consistently outperform others on all occasions (Song & Li, 2008). Recent literature thus sug- gests combining the forecasts from different models with a view to improving forecasting accuracy (Shen, Li, & Song, 2011; Wong, Song, Witt, & Wu, 2007): Shen, Li, and Song (2011) use six linear combina- tion methods; Cang (2011) introduces the nonlinear alternatives; and Chan, Witt, Lee, and Song (2010) employ programming approaches to determine the weights of combination. These empirical studies gener- ally provide favourable evidence of forecast combination. In addition, Coshall and Charlesworth (2011) argue that many forecasting

1658 H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682

scenarios involve more than a single goal and advocate the use of goal programming, which can accommodate multiple criteria decision making.

The growth cycle of tourism demand Butler (1980) has received increasing research attention. One of its methodological developments is to employ Markov regime switching models to test the tourism lifecy- cle concept. According to this concept, a destination goes through six key stages: exploration, involvement, development, consolidation, stag- nation, and decline and/or rejuvenation. The three-stage Markov- switching process is particularly in line with the lifecycle concept, as it allows for a period of decline, slow growth, and rapid growth. In addi- tion, it implicitly allows for rejuvenation. This approach is applied by Moore and Whitehall (2005) to the context of inbound tourism in Barbados. Empirical evidence suggests that the lifecycle concept pro- vides an adequate explanation of the growth stages for each market.

Interdependence and Interrelation

Associated with globalization, market interdependence has become an emerging topic in the latest demand studies. Tourism demand in one destination tends to be affected by demand for alternative destina- tions due not only to cultural and environmental similarities and geo- graphic proximity, but also to similarity in the economic determinants that underpin destination choice. The interactions between tourist flows and their determinants at different destinations shape tourists’ behaviours when they decide where to travel.

In light of the turbulence in the world economy over the past dec- ade, efforts have been made to address the interdependence of tour- ism demands, although the number of published studies is still limited. These studies often firstly confirm the co-movements of tour- ism demand in different destinations using the co-integration tech- nique and then use VAR models to test for the cause-effect relationships among these demand variables of the destinations under consideration via the Granger causality test (Granger, 1969). Torraleja et al. (2009) and Seo, Park, and Boo (2010) identify the existence of causal relationships between the tourism demand variables across dif- ferent destinations. In particular, Seo, Park, and Yu (2009) recognise the time varying rather than the constant conditional correlations in their study and examine the determinants of conditional correlations among destinations using the VAR model. The results reveal that the industrial production index and the real exchange rate are the key determinants of tourism demand in all of the destinations studied.

One of the problems with the existing literature on the interdepen- dence of tourism demand is the lack of attention paid to economic foundations. The studies usually adopt the series of tourist arrivals itself and conduct analysis based only on its time series properties and omit the important economic indicators in the specifications of the models. Another issue is that the number of destinations under discussion is relatively small (usually four or five) and their selection tends to be

H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682 1659

ad hoc, which limits the scope of the analysis. The small destinations sys- tem may overlook the endogenous effects of variables that are not in- cluded in the models. Future studies should consider a theoretically justified demand system involving a large number of interactive desti- nations using appropriate econometric modelling techniques, such as the Global VAR modelling system (Pesaran, Schuermann, & Weiner, 2004).

FIRM, INDUSTRY, AND MARKET

Economic studies of tourism supply are complex and cover a diverse range of topics from the firm level to the industry and market level. Over the period 1970s-1990s, there was vigorous debate about whether tourism is an industry or a market when it is studied from a supply per- spective (e.g., Leiper, 1990, 1992; Smith, 1988, 1991). In the recent tourism economics literature, it has been commonly recognised that tourism is neither a single industry nor a single market (Dwyer et al., 2010; Stabler et al., 2010). Tourism is a composite product that involves a combination of a variety of goods and services provided by different sectors, such as transport, accommodation, tour operators, travel agen- cies, visitor attractions, and retailing. Moreover, tourism products are serviced and transacted in different markets. Therefore, tourism can be studied using both industry-based and market-based economic tools (Wilson, 1998).

Drawing its theoretical foundation from industrial economics or industrial organisation, the development of tourism supply research follows that of industrial economics. The neoclassical approach domi- nated the development of industrial economics until the late 1970s. The structure-conduct-performance (SCP) paradigm provides a useful framework for studying tourism supply from a market perspective. Ear- lier schools of thought, such as the Austrian school, evolutionary eco- nomics, and institutional economics, were developed to relax the restrictions of the neoclassical assumptions, such as rational prefer- ences, information symmetry, static equilibria, and profit maximisa- tion. In particular, the dynamic nature of the market and its institutional arrangements emphasised by these newer approaches is highly relevant to the operations of tourism businesses. By introducing game theory to the study of the firm and the market, new industrial economics has been developed, and this provides powerful tools for analysing firms’ strategic relationships, particularly in the context of a supply chain (Song, 2011; Stabler et al., 2010).

The Structure-Conduct-Performance (SCP) Paradigm

Based on the neoclassical economic theory, especially different mar- ket structure models, the SCP paradigm suggests that the type of the market structure within which a firm operates (e.g., monopolistic, monopolistically competitive, or oligopolistic) rigidly determines a

1660 H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682

firm’s conduct (e.g., output decisions, pricing behaviour, and innova- tion), which ultimately affects its overall performance (e.g., its efficiency, profitability, and growth) (Bain, 1956; Mason, 1939). In the further development of this paradigm, multiple feedback effects and causation flows were taken into account (Scherer & Ross, 1990; Shepherd, 1990). A number of empirical studies have tested the SCP paradigm in tourism, predominantly in the accommodation sector (e.g., Davies, 1999; Davies & Downward, 1996; Pan, 2005) but also in the restaurant sector (e.g., Jang, 2011). However, the findings are inconclusive, due largely to the different empirical settings and methods used. Davies (1999) and Pan (2005) both suggest that market structure directly influences the perfor- mance of a firm, with no clear intermediate effects between market structure and conduct. However, Cunill and Forteza (2010) find that a franchising strategy contributes to increasing market concentration by hotel chains. Tung, Lin, and Wang (2010) reveal a bidirectional causal relationship between the market structure and strategic behaviour based on a more advanced simultaneous equation model.

Although a number of tourism supply studies at the market or indus- try level do not exactly follow the SCP framework, their research fo- cuses fall into one of the following three groups.

Structure

Among all market structure models, oligopoly has attracted the most attention in the literature, such as the studies by Baum and Mudambi (1994) on the UK fully inclusive tour industry, Davies (1999) on the UK hotel industry, Baum and Mudambi (1995) on the resort hotel industry in Bermuda, Bresson and Logossah (2011) on the cruising sec- tor in the Caribbean, and Ciliberto and Tamer (2009) on the American airline industry. In these studies, the common characteristics of oligop- oly are identified in tourism markets in which a small number of large firms dominate the markets, which leads to both high market concen- tration and fixed costs due to entry barriers.

Conduct

Firms’ conduct, particularly pricing behaviour, has been well studied under certain market structures, especially oligopoly. For instance, Vila and Córcoles (2011) investigate such pricing strategies as dynamic pric- ing and price discrimination between flag carriers and low-cost airlines. Abrate, Capriello, and Fraquelli (2011) examine the effects of quality signals on price setting in the hotel industry based on the hedonic pric- ing approach. Poater and Garriga (2009) reveal that the price discrim- ination and peak-load pricing settings that are often exercised in the airline industry are also evident in some European hotels. Based on firm-level time-series data, Malighetti, Paleari, and Redondi (2010) find that the overall intensity of Ryanair’s dynamic pricing has decreased. The discussions of firms’ pricing behaviour are often related to yield

H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682 1661

management issues. Other research on conduct includes various growth strategies such as mergers and acquisitions (e.g., Vogel, 2009), innovation investment (e.g., Fernandez, Cala, & Domecq, 2011), and diversification (Andreu, Claver, & Quer, 2009). In their re- cent review article, Williams and Shaw (2011) stress the importance of globalisation and innovation strategies and their relationships.

Performance

Firm performance, particularly measured by productivity and effi- ciency, has been a long-standing topic in tourism supply studies. The hotel sector has attracted the most attention, followed by travel agents and restaurants. Research developments in this direction are mostly related to methodological advancement. The empirical literature on tourism firm efficiency has been dominated by a non-parametric approach—the data envelopment analysis (DEA) (e.g., Barros & Alves, 2004; Chiang, Tsai, & Wang, 2004). The main advantages of DEA com- pared to the standard econometric technique are that it (1) does not require any form of functional specification and (2) is able to handle multiple inputs and outputs readily in any (in)efficiency theoretical par- adigm (Bernini & Guizzardi, 2010). Based on the input and output data from DEA, a Malmquist Index can be constructed to measure produc- tivity change. The criticism of the DEA method is related to its potential statistical shortcomings. A further development of this method is to use the bootstrap approach to obtain statistical properties. This method has been applied to tourism firms by Assaf, Barros, and Machado (2011).

Another well-developed method is the stochastic frontier approach (a parametric approach). Its principal advantage lies in the decompo- sition of deviations from the efficiency levels between noise (stochastic error) and pure efficiency; however, it faces the challenge of determin- ing the appropriate functional forms (Barros & Dieke, 2008). Recently, a semi-parametric method which combines non-parametric and para- metric approaches was applied to tourism firms by Bernini, Freo, and Gardini (2004). Furthermore, Assaf (2010) employs a Bayesian pa- nel stochastic frontier model to study the cost efficiency of Australian airports.

In addition to efficiency and productivity, firms’ long-term growth is also used to measure firm performance. Based on the production func- tion (i.e., input-output), significant input factors are identified to ex- plain the growth (e.g., Smeral, 2009a). This line of research departs from the SCP paradigm and does not seek explanations of firm growth from market structure or conduct (e.g., Cunill & Forteza, 2010; Jang, 2011).

Game Theory and Supply Chain

To overcome the limitations of the original static form of the SCP paradigm, game theory provides a more powerful tool and a dynamic

1662 H. Song et al./Annals of Tourism Research 39 (2012) 1653–1682

approach to analysing situations in which the decisions of multiple eco- nomic actors affect each other’s payoff. As such, game theory deals with economic actors’ interactive optimisation problems and inter-firm rela- tionships (Cachon & Netessine, 2004). The game-theoretical approach is useful for analysing the strategic decisions of firms within the same industry, particularly in an oligopolistic market, such as tour operators at a destination (e.g. Zhang, Heung, & Yan, 2009). Recent studies emphasize inter-firm strategic interactions in the context of tourism supply chains (Song, 2011; Zhang, Song, & Huang, 2009). The game- theoretical approach has been employed to study the interactions among tourism supply chain members and their strategic options, such as price competition and coordination between a theme park and a tour operator (Song, Yang, & Huang, 2009); relationships among a theme park, hotels, and tour operators in a context of package holiday supplies (Huang, Chen, Song, & Zhang, 2010); and cooperation and competition between two supply chains (Yang, Huang, Song, & Liang, 2009).

Although a range of games have been developed, their applications to tourism supply are mostly restricted to non-cooperative determinis- tic games and involve simplistic strategy options in abstract experimen- tal situations instead of actual industries. Further applications should consider other types of games, such as cooperative, repeated, differen- tial, signalling and screening, and Bayesian games, especially in their dynamic forms (Cachon & Netessine, 2004). More useful managerial implications will be drawn if more realistic market environments are considered in the developed models.

Institutional Approach to the Behaviour of Firms

Moving away from the neoclassical perspective, the institutional ap- proach (such as transaction-cost and agency theories) regards the firm as a governance structure instead of a production entity (Coase, 1960; Jensen & Meckling, 1976). Agency theory recognises conflicts of inter- est between different economic actors and deals with the problems resulting from the principal-agent relationship, such as adverse selec- tion and moral hazards (Stabler et al., 2010). Transaction cost econom- ics adopts a contractual approach to the existence of the firm and focuses on the efficiency of making transactions internally compared to the cost of making such transactions through the market mechanism (Williamson, 1975).

Despite its usefulness, this approach has not been widely used in ana- lysing the behaviour of tourism fi

Collepals.com Plagiarism Free Papers

Are you looking for custom essay writing service or even dissertation writing services? Just request for our write my paper service, and we'll match you with the best essay writer in your subject! With an exceptional team of professional academic experts in a wide range of subjects, we can guarantee you an unrivaled quality of custom-written papers.

Get ZERO PLAGIARISM, HUMAN WRITTEN ESSAYS

Why Hire Collepals.com writers to do your paper?

Quality- We are experienced and have access to ample research materials.

We write plagiarism Free Content

Confidential- We never share or sell your personal information to third parties.

Support-Chat with us today! We are always waiting to answer all your questions.