This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and

Transcription

1 Tis artice appeared in a journa puised y Esevier. Te attaced copy is furnised to te autor for interna non-commercia researc and education use, incuding for instruction at te autors institution and saring wit coeagues. Oter uses, incuding reproduction and distriution, or seing or icensing copies, or posting to persona, institutiona or tird party wesites are proiited. In most cases autors are permitted to post teir version of te artice (e.g. in Word or Tex form) to teir persona wesite or institutiona repository. Autors requiring furter information regarding Esevier s arciving and manuscript poicies are encouraged to visit: ttp://

2 Tourism Management 37 (2013) 86e98 Contents ists avaiae at SciVerse ScienceDirect Tourism Management journa omepage: An economic mode for tourism destinations: Product sopistication and price coordination Rainer Andergassen, Guido Candea, Paoo Figini * Department of Economics, University of Boogna, Piazza Scaravii 2, Boogna, Itay igigts < We identify te destination s key features: product sopistication and price coordination. < We mode te optima deveopment strategy of te tourism destination. < We state a Coordination teorem and a Love for Variety teorem. < We cassify destinations ased on type of coordination and type of resources. artice info astract Artice istory: Received 6 June 2012 Accepted 20 Octoer 2012 JEL Cassification: L83 O1 D11 Keywords: Tourism destination Product variety Natura resources Price coordination Deveopment strategy Te paper modes te optima deveopment strategy of a tourism destination y identifying and anayzing two key economic features: i) te ong-term coice of weter to invest in te enancing of natura and/or cutura resources (wic act as common goods in te destination) or to increase te degree of sopistication of te tourism product (ere intended as te variety of compementary services to accommodation tat are demanded y tourists); ii) te sort-term coice of weter or not to impement price coordination among oca firms, a proem stemming from te anticommon nature of te tourism product. We uid a two-stage mode for te tourism destination, tus identifying te optima degree of sopistication of te tourism product and te optima institutiona arrangement in terms of coordination. Tis approac eps sed igt on te rationae underying te deveopment pat taken y different destinations, tus overcoming some of te imits of existing iterature and providing a simpe taxonomy for te oserved diversity of rea-word destinations. Accordingy, we provide a cassification of destinations ased on te type of coordination and on weter te primary resource is natura, cutura or organizationa. Ó 2012 Esevier Ltd. A rigts reserved. 1. Introduction Te iterature in tourism studies as estaised te tourism destination (TD) as one of its key concepts. Indeed, many papers pivot around te organization, management, deveopment, and sustainaiity of tourism destinations. From te researcer s perspective, te TD emodies a te specific and proematic features of tourism, suc as its systemic nature, in wic space pays a fundamenta roe (Leiper, 1990). In fact, tourism suppy meets demand in te destination; environmenta and cutura resources, attractions and te ospitaity industry are a ocated in te destination; te demand for tourism is reveaed in te destination. In oter words, te TD is te conceptua ink etween te * Corresponding autor. E-mai address: paoo.figini@unio.it (P. Figini). compexity of te sector, te compementarity and sustitutaiity of te many goods and services of wic te tourism product consists, and te suppy of avaiae oca resources. Severa different definitions exist for te TD, ranging from management studies, were it is mainy interpreted as a product, to tourism geograpy were te destination is intended as te offer of te territory. In tis paper we consider te destination from te economics perspective, as a kind of (meta) economic agent: a territoria system suppying at east one tourism product ae to satisfy te compex requirements of te demand for tourism (Candea & Figini, 2012). Te term tourism product defines wat is commony known as a oiday, or wat tourism socioogy cas te tourist s experience. But, were is te economics of destinations mentioned in te iterature aout tourism? Atoug its specific features are indeed discussed y oter discipines, suc as geograpy, management, marketing and organizationa studies, study of te TD from te /$ e see front matter Ó 2012 Esevier Ltd. A rigts reserved. ttp://dx.doi.org/ /j.tourman

3 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 87 economics perspective is in its infancy. Indeed, neiter te micro nor te macroeconomic iterature currenty avaiae eps us fuy grasp te concept. In fact, most iterature considers te destination as a mere framework, noting more tan te astract concept of market, in wic to anayze te micro-eavior of specific industries, suc as te structure of te ote industry in a specific destination, or tourists (e.g. te determinants of te demand for tourism in a given destination). Simiary, macroeconomic iterature often identifies te destination wit an entire country, in a framework were, in order to attain te maximum growt rate, te economy can speciaize in tourism and no attention is paid to wat appens witin te destination itsef. Te many teoretica and empirica works testing te vaidity of te Tourism-Led-Growt ypotesis or weter tourism can e considered as an independent factor of economic growt (Brau, Lanza, & Pigiaru, 2007; Figini & Vici, 2010; Lanza & Pigiaru, 1995, 2000; Sequeira & Macas Nunes, 2008) can e recaed ere. However, ot micro and macroeconomic iterature ony scratces te surface of wat te specific and distinctive economic proem of te TD actuay is. Te most igy regarded economic mode specificay uit for destinations is proay te Tourism Area Life-Cyce (TALC) mode (Buter, 1980), a we-known appication of te product ife-cyce teory. Togeter wit Pog (1974), te psycograpic representation of te interreation etween types of tourists and te evoution of te destination, te TALC mode is sti today te main piar on wic te economic anaysis of te TD is ased (Buter, 2006). Regardess of its popuarity, owever, te TALC mode as significant imitations. Firsty, it is a purey descriptive mode, wit very itte expanatory power; secondy, it is a deterministic mode, and te TD is ound to pass troug te susequent pases of evoution given tat te mode is unae to fuy take different trajectories into account. Researc into te economics of tourism as ony recenty started investigating te TD more cosey, and two interesting strands of iterature can e identified. Te first provides a more teoreticay sound asis to te TALC mode. In particuar, Giannoni and Maupertuis (2007) anayze te dynamics etween te pattern of investment in tourism infrastructures, poicy coices and environmenta quaity, terey generating cycica pats in te numer of tourists osted (from te perspective of consumption waves teory, see aso Swann, 2010); simiary, Lozano, Gomez, and Rey- Maquieira (2008) uid a teoretica mode wose dynamics are consistent wit TALC; finay, Cerina (2007) investigates te reationsip etween growt dynamics and environmenta sustainaiity in a mode were tourism resources are interpreted as common goods, tus providing a teoretica asis for te concept of sustainae tourism. Neverteess, a tese modes sare te same caveat as tey are mainy macroeconomic growt modes were te TD competey overaps wit te economic system, i.e. te country speciaizes entirey in tourism, and were tere are no insigts into wat appens witin te TD. In oter words, te modes ave no micro-foundation. A second strand of iterature tries to take specific organizationa features and te economic caracteristics of te destination into account. In tis ine of researc, te first attempts to uid a compreensive economic mode for te TD were made y Huyers and Bennett (2003), Papateodorou (2003), Wacsman (2006), Candea, Figini, and Scorcu (2008), Candea and Figini (2010) and Andergassen and Candea (2012, 2013). Witin tis framework, te present paper argues tat te TD as specific and distinctive features wic ca for nove, origina economic anaysis. In particuar, te economic mode for te TD deveoped erein focuses on two specific aspects of te economics of tourism tat, in our opinion, are not propery addressed y existing iterature, i.e. te issue of coordination etween oca firms and te degree of sopistication of te tourism product. In addressing tem, we extend and integrate te works of Andergassen and Candea (2012), wo tacked te issue of sopistication, i.e. te suppy of a variety of different oca goods and services tat are aso demanded and purcased y tourists during teir stay, and Candea et a. (2008) and Candea and Figini (2010), wo addressed te issue of price coordination. Our approac foows Papateodorou (2003), wo was te first to formay anayze te issue of te compementarity and variety of services witin te tourism product, and Wacsman (2006), te first to formay anayze te proem of price coordination witin te destination (see aso Avarez-Aeo & Hernandez-Martin, 2009). Te novety of our paper is twofod. Firsty, we generaize te proem of coordination, tacking te main imitations in te resuts of Wacsman (2006) and Candea et a. (2008). Secondy, we jointy consider sopistication and coordination, tus uiding a unique economic mode to descrie te deveopment and te organizationa pattern for te TD. Our approac opens a new window troug wic to consider te economics of te destination, tus igigting important poicy impications for destination management and oca stakeoders. Te economic mode for te TD deveoped in tis paper stems from two intertwined perspectives, empirica and teoretica. From te empirica perspective, our mode aims at eing consistent wit te anecdota evidence of te great diversity of tourism destinations trougout te word, wic differ in teir istory, resources, organizationa structure, institutiona arrangement and speciaization. In tis respect, our mode depicts different trajectories for te TD, ence eing ae to overcome te deterministic ogistic sape of evoution descried in te TALC mode. Indeed, our set-up aows for mutipe equiiria. In te same way, we are ae to expain te reasons wy some destinations can e ocked into a certain stage of deveopment wie oters skip one or more stages competey. Tis ast proem is a key factor for potentia destinations (particuary in deveoping countries), in wic tourism as not (yet) deveoped, ut is seen as an opportunity, often eing considered a strategic pat for economic deveopment y ot poicy makers and oca stakeoders. As we wi igigt in te concusion, our specuative teoretica framework suggests future directions in empirica researc, e.g. testing weter te pattern of coordination and degree of sopistication in te tourism product are significant key factors in expaining te pat of deveopment of a given destination. From te teoretica perspective, te destination is a nove, interesting oject of study for economics. Wie some of te specific proems of te TD (te need to suppy puic goods and tacke externaities) are standard market faiures wic usuay ca for te intervention of te puic sector (atoug Huyers & Bennett, 2003, sow tat te puic intervention is not necessary if vountary cooperation among oca stakeoders for te management of common resources occurs), tere are two specific, distinctive caracteristics of te destination tat are under-investigated and are aso margina issues in te standard economic teory. 1. Te tourism product suppied and sod y (or witin) te destination can e defined as a unde composed of a set of eementary items. Suc goods and services (accommodation, transport, sopping, attractions, events) are demanded in a compementary or sustitutae way y tourists during teir oiday experience. Wie te concept of unde is a standard too in economics (it is commony used to uid price indices to e appied ot in teory and nationa accounting), wat is new in tourism economics is its roe as an oject of study. Te definition of te product as a unde of compementary and sustitutae goods opens up te issue of coordination and cooperation among oca firms suppying te individua components of te oiday. In tis paper, we focus mainy on te compementarity feature, wic is particuary reevant for

4 88 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 oidays and aows us to define te tourism product as an anticommon (Heer, 1998, 1999; Miceman, 1982). Accordingy, te oiday can ony take pace if te permission to stay is granted y a firms suppying compementary services to tourists. Even if one firm ony does not grant permission, te tourism activity in te destination cannot take pace. Stemming from te anticommon nature of te tourism product, te reevant questions to e addressed y an economic mode for te destination are: ow can te anticommon proem e tacked? Is tere any roe tat can e payed y te destination management? Wat is te optima pricing poicy for te tourism product as a woe? We wi present and discuss our mode s answers to tese questions in Section Te woe territory (intended as ot its resources and organizationa structure) enters te production function of te oiday as an input. Hence, te destination can e anayzed as a (meta) economic agent taking important decisions from te suppy-side at a eve tat is intermediate to te micro (firms and tourists) and macro-eves (te woe economic system, usuay te country). Te TD as sometimes een interpreted as a type of custer (Porter, 1998) since it sares some of te caracteristics of te industria district atoug it cannot e defined as suc (Candea & Figini, 2012; Micae, 2003). More specificay, in te industria district firms eiter produce sustitute goods (orizonta custer) or intermediate goods wic are ten assemed (vertica custer). On te contrary, in te TD, firms mainy produce compementary services tat are sod directy to consumers. Brandenurger and Naeuff (1997) ca tis a diagona custer, a concept tat finds a perfect appication in te TD and advocates wat tey ca co-opetition, tat is, te co-existence of ot competition and cooperation etween firms. At te same time, oca resources represent te main motivation for te trip, and terefore aso enter te utiity function of tourists. Suc a comination of an item tat is ot an input of production and an argument of utiity provides anoter nove economic proem to consider. Moreover, most of tese resources are freey avaiae (te andscape and te offer of territory in genera) and can e considered as common goods (Hardin, 1968), wie oter resources can e deveoped y te destination, e.g. an event or an amusement park. In tis respect, te reevant issues to e addressed y an economic mode for destinations are: wat is te process eading to te rise, te deveopment, te speciaization and te (environmenta and economic) sustainaiity of te TD? Wat are te key-factors eading to te deveopment of a tourism product ased on natura and/or cutura resources rater tan an institutiona arrangement promoting a sopisticated tourism product? We wi present and discuss te answers of our mode to tese questions in Section 5. Te remainder of te paper is structured as foows: Section 2 introduces and discusses te caracteristics of te tourism destination, tus providing te intuition eind te mode. Section 3 descries te structure and rationae of te mode, its assumptions and main imitations. Section 4 focuses on te issue of coordination and te optima institutiona set-up for te destination, wie te main resuts in terms of resource speciaization, variety and sopistication of te tourism product are descried in Section 5. Section 6 discusses te main resuts and caveats of te mode, tus paving te way for future researc, ot teoretica and empirica. 2. Te caracteristics of te tourism destination From te economics perspective, te TD does not necessariy coincide wit te destination management organization (DMO), one of te possie institutiona set-ups tat may prevai, or wit te oca poicy maker. Instead, te destination can e seen as a ypotetica (meta) agent, a territoria system wit a specific ojective function to maximize and suject to given constraints. In tis respect, te economics of destinations studies te reationsip etween demand (y different types of tourism osted in te destination) and suppy (y te mix of firms ocated in te territory) for te entire tourism product. In genera, a destination may offer different types of oiday for different types of tourism, eac possiy caracterized y a different mix of specific goods and services, incuding te consumption of oca resources, and accommodation. Terefore, te tourism product is made up of a te tourism destination specific and non-specific goods and services tat are demanded during one day of oiday; its quantity is measured troug te numer of overnigt stays and its vaue is te daiy price of te oiday. In te case of muti-tourism destinations, te woe tourism product can tus e interpreted as te weigted average of te many types of oiday offered in te destination. Note tat in te present context, were te demand function is known wit certainty, coosing te daiy price (wic coincides wit te weigted average price of te oiday) is equivaent to coosing tourism expenditure, i.e. te aggregate price of te oiday. It is we known tat many conceptions of vaue exist, eing particuary reevant in tourism were many nonmarket or semi-market goods (suc as environmenta and cutura resources) are demanded and used. As is typica to te economic approac, we ony focus on te market vaue of te oiday, as determined y its price, and we negect any impication regarding te socia and cutura vaue of tourism in te destination. Neverteess, we indirecty capture te cutura and environmenta vaues of te resources of te destination troug price and demand effects, as in our mode consumer demand canges wit te perception of environmenta quaity. Contrary to wat can e accompised in appied researc, in a teoretica mode we ave to rey on strong assumptions, ess descriptive of te rea-word compexity of destinations ut ae to unfod te core of te economic proem faced y te TD. We start y identifying two necessary conditions for te deveopment of a tourism destination, and one necessary condition for its ong-term sustainaiity (Andergassen & Candea, 2012). Firsty, a generic point of interest, e it natura or artificia, must exist in order for te destination to come into eing. In order to keep te mode as simpe as possie, wie maintaining its expanatory power, oca tourism resources are measured troug a quantitative index R, wic summarizes te overa endowment of te destination (its natura, cutura and organizationa resources, te accessiiity of its transport system, its infrastructures, etc.). R depends on ot exogenous (nature, istory and cuture) and endogenous factors, suc as te investment undertaken y te oca community e puic and private sectors e to adapt te endowment in order for it to e incuded successfuy in te tourism product, and to preserve and enance it y uiding amusement parks or conference venues, y organizing events, etc. In sort, any destination can e identified y its endowment of resources, te ony constraint eing tat R > 0. Given a certain quantity R, we ca z its quaity as it is perceived y tourists. We assume tat z depends (non-positivey) on te numer of overnigt stays, ence rougy accounting for and measuring crowding and congestion effects. Te sign of te reationsip etween parameter z and overnigt stays is a matter of discussion, and we soud aso consider te different case of a positive or, more generay, a non-inear reationsip. In tis paper we assume tat a tresod in te numer of overnigt stays exists, after wic te effects of congestion are at pay. Secondy, at east one variety of oca goods as to e suppied togeter wit ospitaity and te oca resource, oterwise tere is

5 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 89 no reason to visit te destination. We measure te variety of te tourism product troug n, te numer of differentiated tourism goods and services tat form te tourism unde; eac good/service is represented y te index i ¼ 1,., n. Te straigtforward interpretation is tat te iger te numer n 1, te greater te eve of sopistication (te variety) of te tourism product. In te imit case n ¼ 1 ony te asic service needed to access te main resource is provided in te destination. For te sake of simpicity, we assume tat tere are ony two firms in te destination, one ospitaity firm, suppying accommodation and te oter one producing a te differentiated goods/ services. Te assumption of just aving one firm suppying a te differentiated tourism goods/services greaty simpifies te exposition witout atering te quaity of te resuts (in footnote 2 we discuss ow resuts cange if tis assumption is reaxed). We consider identica tourists endowed wit a CES (Constant Easticity of Sustitution) utiity function, foowing a ong strand of iterature on product differentiation wic dates ack to te semina paper y Dixit and Stigitz (1977) and tat as aready een appied to tourism y Papateodorou (2003). Hence, from te point of view of demand, te consumer as to decide ow to aocate a given income among a set of generic consumption goods (y) and a tourism product (T), a unde in wic oca resource R is measured troug its overa quaity (z), ospitaity () and a variety of goods and services (x i ). Moreover, te suppy of different varieties i of te oca good x i and te compementary ospitaity service need to coordinate teir quantity, quaity and price, oterwise a suitae product ae to meet te tourism demand cannot e suppied. For te sake of simpicity, we ony consider coordination in price (assuming tat ot coordination in quantity and in quaity are aready in pace, see Candea & Figini, 2010; Wacsman, 2006) etween te ony ospitaity firm in te destination and te firm suppying differentiated tourism reated goods and services. However, te aove-mentioned conditions are insufficient to guarantee te surviva of te destination in te ong-term. In fact, ong-term economic profitaiity can ony e acieved if te overa tourism profits of te destination, U, net of te costs K orne y te destination for te investment undertaken to promote and foster te oca tourism sector, are at east as arge as U*, wic is an exogenousy defined tresod tat guarantees te ong-term surviva of te destination. Tis is te static equivaent of te dynamic principe in wic, in terms of endogenous growt teory, te growt rate attained troug speciaization in te tourism activity must e at east as arge as te one tat can e otained if te destination speciaizes in aternative activities (Lanza & Pigiaru, 1995, 2000). Concerning te market structure of te destination, we assume tat ot firms are monopoists. Tis can stem from te fact tat te destination operates in a monopoistic competition regime ecause of te pecuiarity of its oca endowment R tus transating into te market power of its ospitaity firm (sector) and of te firm producing differentiated tourism goods x i, since tese goods are inked to te oca resource and teir caracteristics differ across varieties (Candea, Figini, & Scorcu, 2009). We acknowedge tat in reaity many firms are aware of teir market power and terefore interact wit eac oter strategicay. Consequenty, atoug oigopoistic competition woud e a more appropriate setting for te anaysis, we wi eave tis issue for possie future researc. Tis modeing set up wi aow us to identify te conditions under wic a region can successfuy deveop into a tourism destination, eiter y foowing a poicy of enancing its resources R, y directy or indirecty infuencing te degree of sopistication n, or y seecting te strategic coordination of prices (p i, i ¼ 1,., n and p ) etween te producer of oca goods and te ospitaity sector. Tus, te impications for te tourism poicy can e investigated, providing answers to te many (peraps too many) opes of poicy makers and oca stakeoders wo see tourism as te key strategic sector for te economic take-off of teir region. Two positions can tus e mentioned. For some (Raffestin, 1986), tourism is ike te Peano curve, a space-fiing curve in te teory of fractas; ence any territory can ecome a tourism destination. For oters, tourism deveopment is ony triggered y te existence of an exogenous endowment and y te structure of preferences of te consumers, regardess of te investment in resource enancement, in artificia endowment or in te variety of oca goods. Witout suc prerequisites and conditions on demand, te territory woud never e ae to reac te tresod U* necessary to ecome a sustainae tourism destination. 3. Te mode In order to investigate te compex process eading to te rise and deveopment of te TD intuitivey descried in te previous section, we wi now deveop a forma mode. We represent te proem as a muti-stage one, te decision tree of wic is sown in Fig. 1. Firsty, te oca community as to decide weter to deveop a tourism destination or invest in oter economic activities, te former strategy eing cosen if te net economic return to te investment is at east as arge as its opportunity cost, U K U*, were U* is te destination s outside option. If tis inequaity ods, te optima amount of oca resources (ere intended as common goods), ospitaity and variety of te oca goods is determined. Te foowing stage of te proem tackes te issue of weter or not te tourism activity in te oca territory soud e coordinated. In tis pase, te equiirium prices for goods and services incuded in te tourism product (ere intended as an anticommon) are computed. Te decision on weter or not to coordinate depends on te sign of U C U NC, were U C and U NC are overa tourism profits of te destination wit and witout price coordination. In te case of coordination, tere are two possie soutions: coordination provided y te destination management, in wic tota profits are U DM, and coordination provided y a tour operator, in wic tota profits are U TO (see Fig. 1). Moreover, from te perspective of te destination it is aso important to distinguis etween oca and foreign tour operators as, in te atter case, te tour operator s profits are exported and eave te oca economy. Atoug we argue tat te soution wit te oca tour operator dominates te one wit a foreign tour operator in terms of tota oca profits, tis strategy is not aways feasie, particuary for deveoping countries. Indeed, tere may e a ack of te necessary professiona skis and competencies in many destinations or te oca tour operator migt ave iger production costs stemming from information asymmetry or weaker economies of scae and of scope: ence it is represented as a dotted ine in te decision tree in Fig. 1. An important assumption of te mode is tat te production costs of te two oca firms are ni (taking positive average production costs into account woud not ater te quaitative features ut ony compicate te exposition of our resuts) and consequenty tourism profits coincide wit revenues, and ence wit te tota expenditure of tourists in te destination. Tis as an interesting impication, since it aows us to reconcie te target of profit maximization (standard for economic teory) wit tat of maximization of tourism expenditure, te usua target for a oca tourism poicy. Hence, te mode can correcty descrie te reaword practice of ooking at overa tourism expenditure as te main indicator of tourism performance for te destination. We first descrie te demand side of te mode. Since overa tourism profits (U) are equa to tourism expenditure, wic is directy inked to te numer of overnigt stays, tey, in turn,

6 90 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 Ω Overa profit for te tourism sector in te destination Ω* Minimum acceptae tresod of profit for te destination Π TO profit exported y te foreign tour operator K cost of investment in resource enancement or in sopistication of te tourism product Ω = Ω NC No coordination Ω = Ω DM Tourism destination Destination management Ω - K Ω* Coordination Tour operator Loca Tour Operator Ω = Ω TO Ω - K<Ω* Foreign Tour Operator Ω = Ω TO Π TO Oter economic activity Fig. 1. Te decision tree of te tourism destination. depend on te caracteristics of te oca tourism product. We foow Papateodorou (2003) and Andergassen and Candea (2012) y assuming tat tourists demand depends on: i) te avaiaiity of natura and/or cutura resources (see aso Meian-Gonzaez & Garcia-Facon, 2003); ii) te avaiaiity of a variety of oca goods and services, suc as restaurants, recreationa activities, weness and sport faciities, etc., tat justify tourism in te destination, eyond te enjoyment of te main resource. In tis respect, we assume tat tourists sow a ove for variety in te tourism product, as defined y Dixit and Stigitz (1977). We consider a unit mass of identica tourists endowed wit a CES utiity function aving te foowing arguments: i) te engt of stay of te oiday at te destination; ii) te variety of n 1 differentiated tourism reated goods and services x i offered at te destination, wit i ¼ 1, 2,., n; iii) te index measuring te perceived quaity z of te destination s resource endowment R; iv) te consumption of a non-tourism product y (wic can aso e considered as te oiday in an aternative destination). Te tourism product is defined y a unde T incuding overnigt stays and te woe variety of oca goods, T ¼ (, {x i }). If we name wit U T,j te su-utiity of te consumer j as a tourist and wit U Y,j te su-utiity stemming from non-tourism consumption, tota utiity for te agent is U j ¼ U(y(j), z(j), (j), x i (j)) and can e written troug te compound CES utiity function (3) (compound CES utiity functions ave een used witin te context of industria organization and te economics of tourism y Jiandong (2003) and Andergassen and Candea (2012, 2013), respectivey): " U T;j ¼ z g ðjþþ Xn i ¼ 1 x a i ðjþ!g a #1 g U Y;j ¼ yðjþ (2) " U j ¼ U Y;j T;j1 þ U ¼ (y ðjþþz g ðjþþ Xn i¼1 x a i ðjþ!g a # g )1 (1) (3) In suc a mode, 0 < < 1 impies tat te non-tourism good y and te tourism product T are gross sustitutes (if ¼ 1, te two goods are perfect sustitutes); g < 1 impies tat overnigt stays and te consumption of oca goods x i are gross compements (if g / N te two goods are perfect compements). We aso assume tat a < 1, i.e. te degree of sustitutaiity etween tourism product T and non-tourism product Y is not greater tan te degree of sustitutaiity among oca tourism goods and services x i. Finay, z indicates te perceived quaity of te tourism resource, suc as eaces, mountains and/or te cutura eritage of te destination. Te resource quaity z as te O-ring property, tat is, it enters te utiity function (1) as a mutipicative factor since tourism exists if and ony if z > 0. Te price of te non-tourism good is taken as a numeraire, p y 1, p is te price of te overnigt accommodation in te ote, p i is te price of te i-t variety of te oca good. Te udget constraint of tourist j is ence: yðjþþp ðjþþ Xn i ¼ 1 p i x i ðjþ ¼I (4) were I is te tourist s overa income, wic is exogenous to our anaysis since tourists, y definition, are non-residents. Te tourism resource R, generay understood as te woe endowment of te territory, is considered a puic good of te destination and terefore does not appear in te udget constraint (4) (see Papateodorou, 2003). However, tis assumption does not od if te main resource of te destination is very specific and terefore semi-private, suc as an amusement park or a museum. We assume tat te perceived quaity of te resource, z, depends positivey on its quantity R, z R > 0, ut wit non-increasing returns, z RR 0, and non-positivey on H, te tota numer of overnigt stays in te destination, since te eve of satisfaction te tourist gets from visiting te resource is inversey inked to te crowding of te site, z H 0 wit wit z HH 0 and z HR 0. In particuar, we assume tat z is a continuous function of H and tat a tresod H for aggregate overnigt stays exists were z H ¼ 0 for H H and z H > 0 for H > H. Wit tis assumption we aim to mode

7 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 91 te case were aggregate overnigt stays reduce te quaity of tourism if tey exceed te tresod H (we wis to tank an anonymous referee for suggesting tis functiona form). Let te easticity of z wit respect to H e: ε z H ¼ vz H vh z 0; were ε z H ¼ 0 for H H and εz H > 0 for H > H. We first derive te individua demand function for accommodation and te differentiated goods and services, and ten cacuate aggregate demand functions. Given te symmetry on te suppy side of tourism products x i (y assumption average production costs of goods and services are a ni), we ave tat p i ¼ p for a i ¼ 1,., n. Since te mass of tourists is normaized to one, aggregate demand functions H, X i and Y are simpy te sum of te tourist s individua demand functions (j), x i (j) and y(j), for individuas j [0, 1]. More formay, H ¼ ðjþdj, X i ¼ x i ðjþdj Z 1 Z Z 1 and Y ¼ yðjþdj, were H represents tota overnigt stays in te 0 destination, X i tota consumption of te compementary oca good i and Y tota consumption of oter goods. We consider parameter vaues suc tat te price of overnigt stays at te equiirium is greater tan te price of differentiated tourism goods and services, i.e. p > p. Note tat te individua tourist is sma compared wit te overa mass of tourists in te destination and terefore takes te perceived tourism quaity as given wen making decisions aout ow to consume. At an aggregate eve, an individua s coices feed ack into tourism quaity wic, in turn, affects te individua and aggregate consumption eavior. In oter words, a fixed point proem as to e soved. Te foowing properties of aggregate demand functions can e estaised. Lemma 1. Te aggregate demand function for overnigt stays in te destination is H*(n, R, p, p ), were H* is increasing in n, and R and decreasing in p and p; te aggregate demand function for te singe tourism good is X*(n, R, p, p ), were X* is decreasing in n, increasing in R and decreasing in p and p. Proof. Te proof can e found in te Appendix A. Because of te compementarity etween overnigt stays and tourism reated goods and services, an increase in n eads to an increase in te demand for overnigt stays. On te oter and, an increase in n reduces te demand of eac singe tourism good X*. By fueing tourism quaity, an increase in R increases te demand of overnigt stays H* and te demand of tourism goods X. Canges in prices ave te standard effect on demand: te aw of demand ods; assuming tat te easticity of z wit respect to H* is not too arge in asoute vaues, te compementarity etween {x i } and impies tat an increase in p (p ) decreases te demand for H*(X*). 1 Given tat production costs are ni, te profits of te ospitaity sector are P (n, R, p, p ) ¼ p H*(n, R, p, p ) and tose of te firm producing te differentiated goods and services are np(n, R, p, p ), were P(n, R, p, p ) ¼ px*(n, R, p, p ). Overa tourism profits for te destination are: Uðn; R; p; p Þ¼nPðn; R; p; p ÞþP ðn; R; p; p Þ: 1 An increase in p unamiguousy reduces H*. An increase in p produces two effects on X*. Firsty, ecause of te compementarity etween {x i } and, it as a direct negative impact. Secondy, since it reduces H*, it increases tourism quaity z, terey producing an indirect positive effect on X*. If jε z Hj is not too arge, ten te former effect dominates te atter. Forma proof of tis resut is avaiae from te autors upon request. 4. Coordination etween firms in te destination and te optima pricing strategy Like a muti-stage proems, te mode as to e soved ackwards, and te equiirium prices of te different coordination aternatives (second stage of te proem) ave to e determined in order to otain te optima endowment of oca resources and sopistication of te tourism product (soution of te first stage proem). In tis section, we sove te second stage of te proem y moving to te suppy-side and assuming tat firms and destinations are price-makers. Tis ast ypotesis is coerent wit reaword tourism markets, wic are often non-competitive eiter ecause firms ave a monopoy or oigopoy position or, as is te case for destinations, ecause of te ig degree of differentiation of te tourism product at a goa eve (Candea et a., 2009). Hence, we investigate te caracteristics of optima pricing strategies for te TD for a given coice of n and R. Te daiy price of te oiday in te destination, v: v ¼ p þ np X* H * (5) consists of te price for accommodation services and te price of differentiated tourism goods/services mutipied y te quantity demanded per day (nx*/h*). Te equiirium quantity for te tourism product is identified y te numer of days spent at te destination (te numer of overnigt stays, H*). As aready mentioned, in tis set-up tourism in te destination can e interpreted as permission to stay granted y te firms suppying te compementary services demanded y tourists wie on oiday: te tourism activity cannot take pace if one of te two firms fais to grant permission. For instance, tourism demand woud e ni if restaurants were unavaiae in te destination, or if tere was no accommodation. Te catcy idea regarding te existence of a unique economic good, wose property is fragmented across different firms, is known as anticommon and is not new in Economics (see Heer, 1998, 1999; Miceman, 1982; Parisi, Depoorter, & Scutz, 2000; Parisi, Scutz, & Depoorter, 2004) atoug te concept as yet to e sufficienty expoited, particuary witin te fied of te economics of tourism were it neverteess finds perfect appication (Candea et a., 2008). Note tat anticommon is te exact opposite of te muc etter known common, a good wic is avaiae to everyone ut were property rigts are not wedefined (Hardin, 1968). In a genera perspective, firms in te destination ave to coordinate in quaity (in order to avoid a situation werey tourists staying in a uxury ote can ony find take-away restaurants, for exampe) and in quantity (to guarantee tere is no rationing in any of te services demanded). However, in tis paper we ave negected tese issues and simpy focused on price coordination. We introduce tree different cases: (a) no coordination, were eac firm maximizes its profits; () price coordination y means of destination management and (c) coordination provided y a tour operator suppying an a-incusive oiday No coordination Witout coordination, firms sove independent maximization proems. In particuar te maximization proem for te ospitaity firm is max p P. Te first order condition for tis proem is: vp vp ¼ H * þ p vh * vp ¼ 0 (6)

8 92 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 Te firm producing differentiated goods and services faces te maximization proem max p np. Te first order condition for tis proem is: vnp vp ¼ nx* þ pn vx* vp ¼ 0 (7) By considering te first order conditions (6) and (7), and assuming tat second order conditions are satisfied, we otain a system wose soution is, if existing, a Nas equiirium (see aso Wacsman, 2006) impicity defining te optima vaues of p and p: g ðn; R; p; p Þ¼0 gðn; R; p; p Þ¼0 For eac R and n, optima prices for te destination can e expressed as p NC (8) ¼ f NC ðn; RÞ and p ¼ fnc (n,r), firm profits are P NC ðn; RÞ and npnc (n, R), and tota tourism profits for te destination are U NC ðn; RÞ ¼P NC ðn; RÞþnPNC ðn; RÞ Coordination provided y te destination management Tis type of coordination, wic is externa to te market, takes pace if ot te ote and oca firm support te roe and ack te activity of a puic autority, namey, te destination management. We ten assume tat tis oca autority is ae to coordinate te oca tourism sector troug informationa and promotiona activities and se te tourism experience in te destination (oca goods and ospitaity) as if it were an a-incusive package. Formay, te anaytica proem ecomes te maximization of overa tourism profits (tourism expenditure) in te destination: max p;p vh ¼ Uðn; R; p; p Þ were y is te daiy price of te oiday in te destination (5). Assuming tat te second order conditions are satisfied, te first order conditions for tis proem are: vu vh ¼ H þ p vp þ np vx ¼ g vp vp ðn;r;p;p Þþnp vx ¼ 0 vp vu vp ¼ p vh þ nx þ npvx vp vp ¼ gðn;r;p;p vh Þþp vp ¼ 0 from wic we get for eac R and n te optima price for te destination p DM ¼ f ðn; RÞ and p DM ¼ f(n, R), firm profits P DM ðn; RÞ and npdm (n, R), and tota tourism profits U DM ðn; RÞ ¼P DM ðn; RÞþnPDM ðn; RÞ. Te foowing Proposition ods y comparing systems (8) and (9) and ecause of Lemma 1. Proposition 1. p DM < p NC and p DM < p NC and U DM > U NC. Proof. Te resut foows from te second order conditions of te proem and te comparison of systems (8) and (9). In particuar, suppose tat p ¼ p NC and p ¼ p NC, ten from (9) and Lemma 1 vu=vp ¼ npðvx=vp Þ < 0 and vu=vp ¼ p ðvh=vpþ < 0; and consequenty, ecause of concavity of U, p and p are too arge. (9) Te intuition eind tis resut is tat wen goods are compements, teir prices are too ig wen set individuay, since firms are unae to internaize te negative effects a price increase as on te oter firm s demand and consequenty its profits. Aternativey, te coordination of prices provided y te destination management makes it possie to set a more efficient daiy price for te tourism product, tus eading to an increase in overa tourism expenditure even toug individua prices are ower. However, profits for one of te two firms may e ower wen price coordination is in pace if te externaity is strongy asymmetric, i.e. te price of one good strongy affects demand for te oter, ut not te oter way around. For instance, consider te case were te price of overnigt stays as a strong negative impact on te demand for differentiated tourism goods and services, and tat te effect of a cange in te price of tese atter on te demand of overnigt stays is negigie. In tis case, te destination management woud internaize tis negative externaity and tus fix a ower price for overnigt stays, wie keeping te same price for tourism goods. As a consequence, te profits of te ospitaity service woud e ower, tose of te firm producing differentiated tourism goods iger and overa tourism profits woud aso e iger compared wit a situation in wic tere is no price coordination. In tis case, te destination management soud aso redistriute profits among its memers. Note tat Proposition 1 ods for a generic demand function as ong as vx/vp < 0 and vh/vp < 0 and second order conditions are satisfied Coordination provided y a tour operator Coordination can aso occur endogenousy wen te market itsef identifies a new type of firm for managing te anticommon proem. Suc a firm, known as tour operator in usiness practice, stipuates contracts wit otes and oca firms y anticipating a payment tat covers te market risk: te premium paid y firms for tis insurance activity is te discount granted on te fu market price (Casteani & Mussoni, 2007). Te tour operator ten promotes and ses te services witin an a-incusive oiday package, tus earing te risk of no sae. Te contract is accepted y te ote and te firm seing te differentiated goods if, despite te discount, teir profits increase (or at east do not decrease) compared wit te case of no coordination. Let us assume tat te tour operator offers a free-sae contract in order to uy services from oca firms in wic te discounted price is p d for te overnigt stay and p d for eac differentiated good, and were d and d respectivey, are te two discounts (to e interpreted as te insurance premium). Te economic goa of te tour operator is to maximize its own profits P TO (again, for te sake of simpicity, we assume tat average costs of te tour operator are ni) suject to te participation constraint of oca firms: tey accept te contract rater tan seing directy on te market if and ony if teir profits are at east as arge as te profits tey make witout coordination. Suc an optimization proem (for a different interpretation of te tour operator s activity in te coordination proem, see Avarez-Aeo & Hernandez-Martin, 2009), in a principa-agent setting is ence: max p ;pp TO ¼ p H þ npx ðp d ÞH nðp dþx (10) s:t: ðp d ÞH P NC and ðp dþx P NC (11) (11) are te participation constraints, were P NC and P NC are te firms profits otained in Section 4.1 and constitute teir outside options of not accepting te tour operator s contract. If we assume tat te tour operator (te principa) offers oca firms (te agents) te minimum revenues of acceptance (transforming inequaities (11) into equaities, and tus determining d and d ) and y repacing te inding participation constraints (11) into te tour operator s ojective function (10) we otain max p;p P TO ¼ p H þ npx P NC np NC (12) It is easy to verify tat te first order conditions of (12) are te same as tose in (9) and tus optima prices are p ¼ p TO and p ¼ p TO, wic are identica to tose of te destination management.

9 R. Andergassen et a. / Tourism Management 37 (2013) 86e Discussion We can state te foowing teorem y comparing te soution of no coordination wit tose in te case of exogenous coordination troug te destination management and endogenous coordination troug te tour operator. Teorem 1 (Te Coordination Teorem). Given te anticommon property of te tourism product, coordination among firms in te destination, wic can eiter e provided y te destination management or y a tour operator, increases profits from tourism. Hence, price coordination enaes te tourism activity in te destination to e more efficient. 2 Note tat tis is an exampe of te prisoner diemma were (price) coordination yieds a Pareto superior soution to non-coordination. In te case of coordination provided y te market as descried in Section 4.3, owever, tere is a distriutiona confict etween te tour operator and oca firms. Independenty of te way in wic te distriution is soved (wic depends on te argaining power of oca firms, te tour operator and, in a more genera setting, on te numer of tour operators competing for te destination, see Avarez-Aeo & Hernandez-Martin, 2009), it is crucia to assess weter or not te tour operator is a oca or foreign firm. In fact, if te tour operator is a oca firm, tota tourism profits are te same as is in te case of coordination y destination management, atoug te distriution of te overa profits among oca firms canges. More formay, tota profits in te presence of a tour operator are U TO ¼ P TO þ P NC þ np NC, and tus it is easy to see tat U TO ¼ U DM. On te contrary, if te tour operator is a foreign firm, its profits do not contriute to te destination s income and tus tota tourism profits of te destination are U TO ¼ P NC þ np NC, and tus U TO ¼ U NC < U DM, te difference U DM U NC eing te surpus generated y price coordination and tus forming te tour operator s profits. It is ten possie to state a coroary of te Coordination Teorem y focusing on te distriutiona consequences of endogenous coordination. Coroary 1. Wen coordination is provided y a foreign tour operator, oca profits are ower tan in te case of coordination provided y te destination management. Te type of coordination cosen in te destination, terefore, is not neutra wit respect to te distriution of profits. Ceary, te soution of a oca tour operator dominates tat of aforeignone(sinceinteformercasetetouroperator s profits remain in te oca economy), ut it is fundamenta to rememer tat tis coroary stems from te assumption of identica cost structures for ot tour operators, and tus may not od in a more genera setting. It is ikey tat for many destinations, particuary tose in deveoping countries,teocatouroperator migt ack te skis or te market conditions to produce at suc a (ow) cost. To simpify tis exposition, in te remaining part of te paper we focus soey on te case of a oca tour operator. 2 If we assume tat differentiated tourism goods and services are suppied y more tan one firm, eac wit some market power, ten in addition to te proem tat stems from te compementarity etween accommodation and differentiated tourism goods descried aove, tere is an additiona proem stemming from te sustitutaiity etween te individua differentiated goods/services, were it is we known tat firms carge ess tan sociay optima prices. In tis case, firms do not internaize te positive externa effect tat a price increase produces on te demand for anoter product. Tus, y considering togeter te compementarity and te sustitutaiity proem, a more genera coordination teorem sti ods were ot exogenousy or endogenousy negative (due to te compementarity) and positive (due to te sustitutaiity) externaities are internaized. 5. Te optima eve of sopistication of te tourism product in te destination In tis section, we move on to te first stage proem were te TD as to find te optima pattern of deveopment given te price soutions for te coordination proem defined in Section 4. We argue tat, foowing a ong-run strategy of deveopment, te oca poicy maker can engage in investments tat enric te destination s natura as we as cutura resource endowments R. Moreover, we argue tat it can directy or indirecty contro te degree of sopistication of te tourism product n. Tis is done, for exampe, troug eiter granting icenses to open sops or oter usiness activities or, in a more microfounded mode, y taxing or susidizing te set up cost for singe production/commercia faciities. In a fuy-fedged mode, one soud first cacuate te optima private degree of product differentiation foowed y te optima socia one, and ten find te optima poicy intervention suc tat te private one coincides wit te socia one (owever, tis is eyond te scope of te paper). More formay, te TD faces te foowing maximization proem max n;r U n; R; p C ; p C Kðn; RÞ (13) s:t: U K U * (14) were K(n, R) are te poicy maker s cost of enforcing a degree of sopistication n and endowing te destination wit resources R, were vk/vn > 0, vk/vr > 0, v 2 K/vn 2 > 0 and v 2 K/vR 2 > 0. A cange in n and R affects tourism profits directy and indirecty troug a cange in p and p. Since te soution of te second stage is eiter exogenous or endogenous price coordination wit equiirium prices ðp C ¼ p TO ¼ p DM ; p C ¼ pto ¼ p DM Þ, te indirect effect is of second order and can tus e negected ecause of te enveope teorem. Formay, te tota derivative of tourism profits wit respect to n is: du dn ¼ vu vn þ vu vp C vp C vn þ vu vp C vp C vn (15) were te first term is te direct effect of n on U and te ast two terms are te indirect effects troug p C and p C, respectivey. Because of te first order condition (9) te indirect effects are negigie, i.e. vu/vp C ¼ 0 and vu/vp C ¼ 0, and tus ony te direct effect matters, i.e. du/dn ¼ vu/vn. Te case of a cange in R gives symmetric resuts. Finay, (14) is a sustainae deveopment constraint requiring tat tourism profits must e sufficienty arge as to guarantee te surviva of te destination over time. Before caracterizing te soution to proem (13), we descrie te anaytica properties of U, wic is a specia case of te resuts in Andergassen and Candea (2012). Proposition 2. (a) U is increasing in R. () U is decreasing in te degree of tourism sopistication n if ε z H g > 1 and is increasing if ε z H g < 1, were g g=ð1 gþ ð 1; 0Þ. Proof. See te Appendix A. An increase in te destination s resource endowment increases te demand for overnigt stays as we as te one for tourism reated goods and services and tus overa tourism profits increase. Differenty, an increase in te degree of sopistication of te tourism product as an amiguous effect on tourism profits, depending on te degree of compementarity etween overnigt stays and tourism goods as we as on te easticity of tourism quaity wit respect to te size of te tourism activity. To

10 94 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 understand te intuition eind tis resut we ave to disentange te effects of te two opposing forces tat are at pay, namey, a ove for variety and a tourism depreciation effects. If overnigt stays and tourism goods are independent (tat is, for g / 0), ten an increase in n does not affect overnigt stays H. Consequenty, te tourism quaity does not decrease and tus te ony effect is tat tourists spend more in a given destination owing to teir ove for variety. Hence, tourism profits increase. In a simiar vein, profits increase unamiguousy if overnigt stays and tourism goods are compements, were an increase in n increases overnigt stays H, and te perceived tourism quaity does not decrease as H increases (i.e. for H H were z H ¼ 0). On te oter and, if z H < 0 (tat is, if H > H), ten an increase in H reduces tourism quaity, terey reducing tourists expenditure on te overa tourism product. If te reduction in tourism quaity is sufficienty strong (tat is, jε z H j > 1=j gj) ten tis atter effect more tan offsets te ove for variety effect and tus overa tourism expenditure, tat is, overa tourism profits, decrease. On te oter and, if te reduction in tourism quaity is sufficienty weak (tat is, jε z H j < 1=j gj), ten an increase in te degree of sopistication increases tourism profits. Assuming tat parameters are suc tat for some vaues of n, U n > 0 (if on te contrary U n < 0 for a n 1, ten tourism deveopment troug sopistication is not viae) te margina rate of sustitution etween R and n provides a teoretica answer to te question aout te conditions driving te creation of a tourism destination, its deveopment pattern and its sustainaiity. In fact, te optima degree of investment in speciaization (troug enancement of resources or sopistication) of te tourism destination is given y te maximization of tourism profits. Te first order conditions of te maximization proem (13) are: vu vr ¼ vk vr and vu vn ¼ vk vn and descrie te optima poicy mix of te destination, determined wit respect to te reative margina gain (te margina rate of sustitution etween R and n is MRS R;n ¼ U R =U n ) and to te reative margina costs of oca investments in R and n ðmrs R;n ¼ K R =K n Þ, respectivey. Moreover, y means of tecnica progress, te easticity ε z H affecting te properties of U descried in Proposition 2 migt cange. Te interpretation woud e tat te tourism sector is moving away from eing a ow tecnoogy sector, tus increasing its options of preserving, restoring and requaifying its resources. If te negative tourism quaity effect aways dominates te positive ove for variety effect (i.e. ε z H g > 1 for eac n 1 and R > 0), ten deveopment of te destination troug tourism sopistication is not feasie and can tus ony occur troug resource investments. Investment in resources may pave te way for tourism sopistication to ecome a viae deveopment strategy if it aeviates te negative tourism quaity effect (formay, if v z HnvR < 0) and tus te destination can twin resource investments wit tourism sopistication to foster tourism activity. Te different soutions make it possie to recognize different organizationa and deveopment patterns for rea-word destinations. For exampe, if we narrow te anaysis down to eac tourism in Itay, tere are destinations wit imited resources ut a igy sopisticated tourism product, ike Rimini for exampe; destinations wit extraordinary natura resources and no sopistication at a, wit oca suppy imited to ospitaity suc as some coasta areas of Sardinia; tere are destinations wit significant natura resources and a certain degree of variety of oca products ike te Costa Smerada. Tis as important teoretica and poitica impications as, according to te TALC mode, tese tree different types of destination are at different stages of teir evoution, wie 3 in te present mode tey are different optima equiiria stemming from aternative endowments of resources and speciaization patterns. Moreover, to compete te description of te soution to proem (13), we need to discuss te reevance of te sustainae deveopment constraint (14): U(n*, R*) K(n*, R*) U*. If tis condition ods we can affirm tat tourism deveopment is viae and sustainae. If instead U(n*, R*) K(n*, R*) < U*, ten tourism profits are eow te minimum tresod and te soution for a territory wiing to ecome a destination is not economicay viae. From a poicy perspective, te impication of tis section of te mode is tat destination management can trigger tourism deveopment in two ways: eiter y investing in te enancement, preservation and improvement of existing resources, or y supporting an increase in te variety of oca tourism goods and services, i.e. wat is known as te degree of sopistication of te tourism product. Wie te former strategy is usuay ound y te exogenous endowment of istorica sites, cutura eritage and natura environment, te atter can e impemented troug te interaction etween te private and te puic sectors and as te advantage of fueing forward and ackward inkages among tourism firms and etween tourism firms and oter sectors. Lasty, it is remarkae to note ow te increase in te degree of sopistication of te tourism product as a doue effect on tota expenditure: on te one and, it positivey affects profits troug an increase in te tota numer of overnigt stays; on te oter, it negativey affects it troug a perceived worsening of quaity due to te effects of congestion and crowding. As a particuar case, if te deveopment eve of a destination is suc tat it does not ear congestion effects, i.e. H < H were ε z H ¼ 0, or if tis effect is very sma, i.e. jε z H j < 1=j gj, tese properties define a Love for Variety Teorem for te destination, aowing tourism to take-off in te ong run. Variety in te tourism product can ten e a strategic asset. Teorem 2 (Love for Variety Teorem). As ong as te negative externaities on tourism quaity are sma, reorganization of te tourism destination toward increasing te variety of avaiae goods and services raises tourists wefare and teir wiingness to spend on tourism at te expense of non-tourism consumption, terey stimuating te economic deveopment of te destination. 6. Concusions In tis paper we ave deveoped an economic mode for te tourism destination y focusing on two specific aspects tat, in our opinion, cannot e studied propery wit te standard tooox of micro and macroeconomic teories and for wic te destination is an interesting oject of study from te economics perspective: i) te tourism product can e defined as a unde composed of a set of eementary items. Suc goods and services (accommodation, transport, sopping, attractions, events, etc.) are mainy demanded in a reationsip of compementarity y te tourist during te oiday experience; ii) te territory (its endowment of resources and its organizationa structure) is argument of ot production and utiity functions, and ence te destination can e anayzed as a (meta) economic agent taking important decisions from te suppy side at an intermediate eve to te micro (firms and tourists) and macro-eves (te country s entire economy). Two key issues ave een identified in order to understand te rise, speciaization, deveopment and institutiona arrangement of tourism destinations. Tese are i) te coice etween investing in te variety of te tourism product (its sopistication) or enancing oca resources; ii) te coordination of oca firms, stemming from te anticommon property of te tourism product.

11 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 95 Tese proems ave een jointy anayzed and tacked in tis paper for te first time, uiding on a recent strand of iterature (Andergassen & Candea, 2012, 2013; Candea & Figini, 2010; Candea et a., 2008; Huyers & Bennett, 2003; Papateodorou, 2003; Wacsman, 2006). Our teoretica set up aows us to state: i) a Love for Variety Teorem wic depicts aternative deveopment trajectories aowing te destination to reac its economic goa: from investing in enancing resources to te process of increasing te sopistication of te tourism product; ii) a Coordination Teorem, from wic different institutiona set-ups can e identified, ranging from oca destination management to te coordination provided y te tour operator. It is important to igigt tat, since in te rea-word eac destination as different cutura, natura and socio-economic caracteristics, and since stakeoders are often caed upon to make decisions witin a framework of ounded and imited rationaity, te mode presented in tis paper does not aim to reduce ad unicum and propose a unique and converging mode of destination deveopment and management. On te contrary, we wis to provide a teoretica asis for te puraity of rea-word soutions: in tis sense, we eieve tat our mode as more expanatory power tan te TALC mode. Indeed, as far as coordination is concerned, we can identify: * Individuay ased destinations, in wic tere is no coordination etween oca firms operating in te tourism sector. Given our assumption of no organizationa costs, or equivaenty identica organizationa costs, tis soution is aways dominated y (endogenous or exogenous) coordination. However, as we argue ater in tis section, tis may not aways e te case; * Community managed destinations, in wic oca firms are coordinated y a oca autority: te destination management (tat can e a puic ody, an association of oca firms, or an outsourced destination management organization e DMO); * Corporate ased destinations, in wic coordination is provided y a tour operator (and were it is important to distinguis weter te tour operator is oca or foreign). As far as tourism sopistication is concerned, we can identify: * Resource ased destinations, were te tourism product is ased on oca resources (eiter natura, cutura or artificia), wit a very imited variety of differentiated goods; * Sopistication ased destinations, were oca resources are very imited ut te tourism product is ased on a arge variety of oca goods and services; * Mixed ased destinations, were tere is a aance etween oca resources and a certain degree of sopistication of te tourism product. Overa, since te economic proem of te destination is identified in our mode y two dimensions (sopistication and coordination) and since we ist tree casses for eac dimension, we are teoreticay ae to propose a taxonomy of destinations in (at east) nine casses, to wic we ave to add: * Non-tourism destinations, regions in wic investing in tourism is neiter economicay viae nor convenient. We eieve tat our mode is a first step in jointy anayzing te two fundamenta features of tourism destinations (sopistication and coordination) wic were recenty introduced in iterature, and in providing a new perspective for tourism economics. We are aware of te many imitations of te mode, stemming, in particuar, from some over-simpifying assumptions tat make appying it to rea-word poicy panning difficut. However, te mode is aready sufficienty intricate in te present setting, and reaxation of some assumptions migt excessivey compicate its soution, at east at te present state of te art. In tis respect, te main imitation is tat te mode is a partia equiirium one, since te destination is anayzed in isoation. Te most important extension woud terefore e to move toward a genera equiirium framework wit inter-destination competition (y considering at east two competing destinations). A second important extension woud e to introduce a muti-destination payer into suc a genera setting ike, for exampe, an internationa ote cain wic is a monopoist in te sector and owns otes in a destinations. Tis is exacty wat Wacsman (2006) does in a simpified mode of coordination wit inear demand. It woud ten e very interesting to ceck te roustness of is resuts (tat te advantages of intra-destination coordination tend to disappear wen competition etween destinations is introduced and wen a monopoistic muti-payer appears in ot destinations) in our more genera framework. Simiary, an open question woud e to see wat appens if coordination is provided y te same tour operator in ot destinations. Future researc soud aso reax some oter, more tecnica assumptions. Firsty, accounting expicity for production and coordination costs, wic may reverse te resut tat price coordination is optima and migt affect in a non-trivia way investment and deveopment strategies in te first stage of te game. In fact, if one takes organizationa costs into account ten no-coordination may ecome te optima institutiona set-up, provided tat te reative organizationa costs of exogenous and endogenous coordination are sufficienty arge. Secondy, y assuming tat differentiated tourism reated goods and services are produced y more tan one firm wit some market power, tat is, aandoning te assumption of monopoistic competition and framing te mode in te context of oigopoistic competition, one coud expicity study te more genera price coordination proem were compementarity and sustitutaiity, te core eements of te tourism product, coexist. Tirdy, rea-word destinations often differ wit respect to te preferences of tourists to crowding. Hence te present assumption of tourists eing crowding-averse, wic is tantamount to assuming tat a sno-effect is at work, as to e extended to te aternative assumption of mass tourism, were a and-wagon effect woud e at work instead. In tis respect, Swann (2010) migt e an interesting approac to foow in a dynamic perspective. Fourty, anoter extension woud e to transform te mutipe-stage proem into a simutaneous equiirium in wic ot prices and types of investment are jointy determined. Lasty, dynamics coud e introduced expicity in te mode, to investigate te evoution of te destination and optima poicy intervention over time. Tis, and oter reated issues, are eft for furter teoretica researc. On a different perspective, tere are many interesting questions tat te mode s set-up and its concusions eave open to empirica researc. Firsty, tere is te issue of measuring sopistication and price coordination. Concerning sopistication, ot te numer of tourism usinesses and teir degree of diversification soud e taken into account. At te same time, a fundamenta issue regards ow to estimate te importance of oca resources, ot in terms of quantity R and quaity z. Concerning coordination, a carefu measure soud e ae to distinguis etween te different types of organizationa patterns. Secondy, wat are te factors determining te pattern of sopistication and coordination cosen y te destination? Is te empirica evidence coerent wit our teoretica mode? Tirdy, is te economic performance of te tourism destination and its evoution over time correated wit te degree of sopistication of its tourism product and wit te type of

12 96 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 coordination prevaiing among firms? Te answer to tis question is crucia in determining te vaidity of te mode, in wic te asence of catcing up and of a converging mode of deveopment for tourism destinations is a key resut. Wie te uge ody of iterature in destination management can provide us wit te state of te art on ow to approac tese empirica issues, we eave tese questions open for future researc. Acknowedgments Te autors wis to tank Mattias Fucs, Laurent Botti, Sauveur Giannoni, participants at te 3rd QATEM worksop, ed in Corte (France) on 13 Juy 2012 and at te 1st Internationa Conference on Tourism Management and Tourism Reated Issues, ed in Rimini (Itay) on 15e16 Septemer 2011 for teir usefu comments on previous versions of tis paper. Te usua discaimers appy. Appendix A wie from (17) and (16) we otain p ¼ z g þ n g ax g g 1 g 1. y 1 wic, using (19), reads as: y ¼ ðp Þ1 z þ n g 1 g 1 a a p p g g 1 g (20) For te foowing it is convenient to express parameters caracterizing te consumer s preferences in(1)e(3) using te definitions =ð1 Þ ð0; NÞ; g g=ð1 gþ ð 1; 0Þ; a ð1 aþ=a ð0; NÞ: We cacuate y sustituting (19) and (20) in te udget constraint (4) and otain: ðn;zþ¼ I p p 1þn ga 1þp p z 1þn ga g pp (21) were n > 0. Sustituting (21) ack into (19) and (20) one otains: Proof of Lemma 1. Since tere is a continuum of tourists, eac one as a negigie effect on tourism quaity z. Using Lagrange to sove te proem of maximizing (3) under te udget constraint (4), te first order conditions for te representative consumer read as foows: (y þ z " g þ Xn x a i i ¼ 1!g a # g )1 1 y 1 ¼ (16) I xðn; zþ ¼ p g p n a þ 1 p 2 n þ p z6 4 n 1 3 g þ n ga p 7 p 5 g ð22þ (y þz " (y þ z " g þ g þ X n x a i i ¼ 1 Xn x a i i¼1 Xn x a i i ¼ 1!g a # g )1 1 z " g þ Xn x a i i¼1!g a # g )1 1 z " g þ Xn x a i i ¼ 1!g a # g 1 g 1 ¼ p!g a # g 1 (17)!g a 1 x a 1 i ¼ p i ; ð18þ and, yðn; zþ ¼I 1 : (23) 1 þ p z 1 þ n pp ga g Since a tourists are identica, (j) ¼ and x(j) ¼ x, and consequenty H ¼ and since p i ¼ p it foows tat X i ¼ X ¼ x. Nextwe cacuate te aggregate demand function H(n, R), were te tourists aggregate coice H feeds ack into tourism quaity z. Using (21), we ave to sove te foowing fixed point proem: H ¼ f ðp ; p; n; zðh; RÞÞ I p p 1 þ n ga 1 þ p p z 1 þ n ga g pp (24) for i ¼ 1,., n, were is te Lagrange mutipier. Using te assumption tat a firms producing tourism reated goods are symmetric we ave p i ¼ p and ence otain x i ¼ x, for eac i ¼ 1,., n. From (17) and (18) we otain: 1 p x ¼ n 1 g g 1 a p (19) wic yieds te soution H ¼ Hðn; R; p; p Þ. In view of te assumption aout z, f H ðp ; p; n; zðh; RÞÞ 0, wit f(p, p, n, z(0, R)) > 0 and a unique H + soving f ðp ; p; n; zðh; RÞÞ ¼ H exists. Appying te impicit function teorem one otains vh + =vn ¼ f n =ð1 f H Þ > 0 and vh + =vr ¼ f R =ð1 f H Þ > 0; respectivey. After rearranging terms one otains:

13 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 97 vh + vn ¼ H +2p I p g a n ga 1 p 1 H +p I p z 1 þ p z 1 g 1 þ n pp ga g 1 þ n ga pp 1 g ε z H > 0 (25) and, vh + vr ¼ H +2p 1 þ n ga p I pp z z R z 1 þ n ga pp g > 0 (26) 1 H +p I p 1 z 1 þ n pp ga g ε z H After sustituting te expression for vh + =vn in te first ine of te derivative and rearranging terms we otain: p z 1 þ n pp ga g 1 p H +p a n ga 1 I gε z H 1 þ n ga 1 þ p pp z p 1 þ n pp ga g þ 1 1 H +p I p 1 z 1 þ n pp ga g ε z H Next consider te price effects. Oserve tat f H < 0, f p < 0 and tat: f p ¼ 1 H+2 þ I þ 1 þ g n g a 1 þ p pp z 1 þ n pp ga g < 0 ð27þ Appying te impicit function teorem to (24) we otain dh=dp ¼ f p =ð1 f H Þ < 0 and dh=dp ¼ f p =ð1 f H Þ < 0. By increasing z, an increase in R increases X. Since < a it foows tat a < 1 and tus an increase in n directy reduces X. Moreover, an increase in n increases H and tus reduces z, terey indirecty reducing X. Consider te price effects on X. It is easy to see tat te direct effect of p and p on X is negative. But since an increase in p or p reduces H, y increasing z, it increases X. It can e sown tat as ong as jε z Hj is not too arge, te former effect dominates te atter and tus an increase in p or p reduces X. Proof of Proposition 2. Proof of part (a) of te Proposition is in te text. Part (). Te derivative of te denominator of U wit respect to n is: p z z H z vh + 1 þ n ga g þ p vn pp z g 1 þ n ga pp g 1 g a n ga 1 p p Using te equiirium expression for H + (24) te derivative of U wit respect to n can e written as: U n ðr;nþ¼ U2 I p z 1þn pp ga g 1 a n ga 1 p p ε z 1 H +p I p 1 H g 1 z 1þn pp ga g ε z H wic estaises te resut. References Avarez-Aeo, C. D., & Hernandez-Martin, R. (2009). Te commons and anticommons proems in te tourism economy. Document de Trea XREAP, University of La Laguna. Andergassen, R., & Candea, G. (2012). Deveopment strategies for tourism destinations: tourism sopistication vs. resource investments. Economia Poitica e Journa of Anaytica and Institutiona Economics, 29, 173e191. Andergassen, R., & Candea, G. (2013). Less deveoped countries, tourism investments and oca economic deveopment. Review of Deveopment Economics, 17,16e33. Brandenurger, A. M., & Naeuff, B. J. (1997). Co-opetition. New York: Currency Doueday. Brau, R., Lanza, A., & Pigiaru, F. (2007). How fast are sma tourism countries growing? Evidence from te data for 1980e2003. Tourism Economics, 13, 603e 613. Buter, R. W. (1980). Te concept of a tourist area cyce of evoution: impications for management of resources. Canadian Geograper, 24, 5e12. Buter, R. W. (Ed.), (2006). Te tourist area ife cyce. Toronto: CanneView.

14 98 R. Andergassen et a. / Tourism Management 37 (2013) 86e98 Candea, G., & Figini, P. (2010). Destination unknown. Is tere any economics eyond tourism areas? Review of Economic Anaysis, 2, 256e271. Candea, G., & Figini, P. (2012). Te economics of tourism destinations. Berin and Heideerg: Springer. Candea, G., Figini, P., & Scorcu, A. E. (2008). Te economics of oca tourism systems. In R. Brau (Ed.), Tourism and sustainae economic deveopment: Macroeconomic modes and empirica metods. Cetenam: Edward Egar. Candea, G., Figini, P., & Scorcu, A. E. (2009). Destination management and tourists coice wit a two-part tariff price of te oiday. Rivista di Poitica Economica, 99, 107e125. Casteani, M., & Mussoni, M. (2007). An economic anaysis of te tourism contract: aotment and free sae. In A. Matias, & P. Nijkamp (Eds.), Advances in modern tourism researc. Heideerg: Pysica-Verag. Cerina, F. (2007). Tourism speciaization and environmenta sustainaiity in a dynamic economy. Tourism Economics, 13, 553e582. Dixit, A. K., & Stigitz, J. E. (1977). Monopoistic competition and optimum product diversity. American Economic Review, 67, 297e308. Figini, P., & Vici, L. (2010). Tourism and growt in a cross-section of countries. Tourism Economics, 16, 789e805. Giannoni, S., & Maupertuis, M. A. (2007). Environmenta quaity and optima investment in tourism infrastructures: a sma isand perspective. Tourism Economics, 13, 499e513. Hardin, G. (1968). Te tragedy of te commons. Science, 162, 1243e1248. Heer, M. A. (1998). Te tragedy of te anticommons: property in te transition from Marx to markets. Harvard Law Review, 111, 621. Heer, M. A. (1999). Te oundaries of private property. Yae Law Review, 108, 1163e1223. Huyers, T., & Bennett, J. (2003). Environmenta management and te competitiveness of nature-ased resource destinations. Environmenta and Resources Economics, 24, 213e233. Jiandong, J. (2003). Oigopoistic competition, tecnoogy innovation and mutiproduct firms. Review of Internationa Economics, 11, 346e359. Lanza, A., & Pigiaru, F. (1995). Te tourism sector in te open economy. In P. Nijkamp, & W. Coccossis (Eds.), Tourism and te environment. Adersot: Aveury. Lanza, A., & Pigiaru, F. (2000). Tourism and economic growt: does country s size matter? Rivista Internazionae di Scienze Economice e Commerciai, 47, 77e85. Leiper, N. (1990). Tourism systemsin Occasiona papers 2. Auckand: Massey University. Lozano, J., Gomez, C. M., & Rey-Maquieira, J. (2008). Te TALC ypotesis and economic growt teory. Tourism Economics, 14, 727e749. Meian-Gonzaez, A., & Garcia-Facon, J. M. (2003). Competitive potentia of tourism in destinations. Annas of Tourism Researc, 30, 720e740. Micae, E. J. (2003). Tourism micro-custers. Tourism Economics, 9, 133e145. Miceman, F. I. (1982). Etics, economics, and te aw of property. In J. R. Pennock, & J. W. Capman (Eds.), Nomos XXIV: Etics, economics and te aw. New York: New York University Press. Papateodorou, A. (2003). Modeing tourism deveopment: a syntetic approac. Tourism Economics, 9, 407e430. Parisi, F., Depoorter, B., & Scutz, N. (2000). Duaity in property: Commons and anticommonsin Law and economics researc paper series, N University of Virginia Scoo of Law. Parisi, F., Scutz, N., & Depoorter, B. (2004). Simutaneous and sequentia anticommons. European Journa of Law and Economics, 17, 175e190. Pog, S. C. (1974). Wy destination areas rise and fa in popuarity. Corne Hote and Restaurant Quartery, 14, 43e45. Porter, M. E. (1998). Custer and te new economic competition. Harvard Business Review, 77e90. Raffestin, C. (1986). Nature e cuture du ieu touristique. Méditerranée, 3. Sequeira, T. N., & Macas Nunes, P. (2008). Does tourism infuence economic growt? A dynamic pane data approac. Appied Economics, 40, 2431e2441. Swann, P. G. M. (2010). Te rise, fa and renaissance of te resort: a simpe economic mode. Tourism Economics, 16, 45e62. Wacsman, Y. (2006). Strategic interaction among firms in tourist destinations. Tourism Economics, 12, 531e541. Rainer Andergassen (orn 12 Septemer 1974, Bozano, Itay) is an Assistant Professor at te Department of Economics, University of Boogna, and is te Director of Economics and Market Poicy, a two-year Master degree offered at University of Boogna, Rimini Campus. He otained a aceor degree wit distinction from te University of Boogna (1997), a Master Degree in Economics wit distinction from te University Coege London (2000) and a P.D. in Economics from te University of Siena (2002). Among is current researc interests are te dynamics of tecnica progress and economic growt, executive compensation and tourism economics. He as autored or co-autored severa papers in refereed internationa journas, incuding Journa of Economic Dynamics and Contro, Journa of Economic Beavior and Organization and Economics Letters. His researc interests in tourism economics are focussed on tourism demand and oca deveopment. Guido Candea was orn in Boogna on January, te 28t, Ho otained is B.A. degree in In 1970 e was teacing assistant at te Scoo of Economics, University of Boogna. He ecame fu professor of Economic Poicy at te Scoo of Economics, University of Padua in 1976 and one year ater e moved to University of Boogna. In 2000 e moved to te Scoo of Economics (Rimini Campus) of te University of Boogna, were e acted as Dean of te Scoo from 2001 to Since 2010 e is Emeritus Professor. His researc interests are: i) teoretica economic poicy; ii) tourism economics; iii) economics of te arts. Paoo Figini is Associate Professor of Economic Poicy at te Department of Economics, University of Boogna since He ods a B.A. from University of Pavia (Itay) and a P.D. in Economics from Trinity Coege, Duin (Ireand). His main researc interests are: i) te economic poicy of goaisation and its socia consequences on poverty and inequaity; ii) socia economics (open source, sociay responsie consumption, etc.; iii) tourism economics. He is Director of te Master in Tourism Economics and Management (TEaM) at te Rimini Campus of University of Boogna.