Why People Travel to Different Places

8/6/2019 Why People Travel to Different Places

http://slidepdf.com/reader/full/why-people-travel-to-different-places 1/16

~pergamon

www.elscvier.corn/locate/aruur es

PH: SOI60·7383(00)00014,1

..1"""" ofTourism Reuanh, V o l. 2 1 1, N o. 1 ,1 '1 '. 1 · 6· 1 -1 7 9 , ZOOIT} 2000 EI.evier Science Lid. All ri~h(s reserved

Prin Id in Great Britain01 GO.7383!J)O / f20 . 00

WHY PEOPLE TRAVEL TODIFFERENT PLACES

Andreas Papatheodorou

University of Oxford, UK

Abstract: This study aim, at providing an economic explanation for the. observed variety inthe actual consumer choice of destina tions. Despite its contribution to tourism research , the

traditional demand theory is insufficient to justify comprehensively the direc tion of touristflows in 'pace and time, mainly because it cannot account lor the Importance of productdiffcrcntiarion and corporate power. To address these issues, the Gorman/Lancastercharacterist ics framework is applied to tourism and a comparative exercise is undertaken insix dillurcnt fields. The theoretical conclusions arc appealing as they match. demand andsupply, offering a holistic answer to the question o f tourist choice and a useful benchmark[or further research in the area. Keywords: demand, characteristics, dillcreut intion, indus ..trial organization. ,!) 2000 Elsevier Science Ltd. All rights ruscrvcd.

Resume: Pourquoi voyage ..(..on a des en drnits dif f erents? C etrc etude vise 1 1 donner un eexplication economique a la \'ariete dans lcs choix de desrlnations de Ia part desconsommateurs. Mulgre sa contriburion a la recherche en tourisrne , la theori« traditinnnellede la demande nc suint pas pour expliquer compleremenr Ia direction des flux touristiquesdans l'espacc C.t It: temps, surtout parer qu'clle nc ricnt pas compte de l'irnporrancc de ladiJ1crcntialion du produit ct du puuvoir d'entrcprisc. Pour aborder ces points, le cadre descaracreristiqucs de Gorman et Lancaster est applique au rourisrne, et on entreprend unexcrcicc cornparatif dans six domain". different,. Les conclusions thcoriqucs sontinteressante" car elles marienr la demandc avec l'offre, donnant une reponsu holistique 1 1 laquestion du cl io ix (ouristiquc ct un rcpctc utile pour la recherche future. Mots·c1~s:

dernande, carncnirisr iques, diJTerentiat ion, organisation indusrr ielle. [ ') 2000 Elsevier ScienceLtd. AI L rights reserved.

INTRODUCTION

Destination choice has always been a central issue in the tourism

management literature, In most cases research has sought to justify

the direction of the observed flows by relying rather scholastically

on the analytical framework provided by the traditional demand

theory. In particular and following the separability paradigm

(Deaton and Muellbauer 19S0b), a representative consumer is

assumed to allocate financial resources among the tourist and the

non-tourist products at a first stage, in a way that maximizes his/

her utility given the existing constraints; subsequently the quan-

Andreas Papatheodorou is a tourism researcher in the School or Geography at theUniversity of Oxford (Mansfield Road, Oxford. OX I 3TB, UK Email (undreas.papathcodor [email protected] .ac.uk)l. He has previously published on tourism demand and is currently working01 1 the Iinkagus between the industrial evolution in tourism and the spatial interactions in

resorts. His geographical area of interest covers predominantly the Mediterranean region.

164-

http://www.elscvier.corn/locate/aruur

http://www.elscvier.corn/locate/aruur



ANDREAS PAPATI-IEODOROU 1 65

tiries to consume are from each tourist good 011 a similar rational

basis.This line of thinking is most apparent in the empirical work on

tourism demand. The majority of econometric models follow a time-series, single-equation approach whereas in a more advanced con-text a number of demand systems is also estimated. In the firstcase, the dependent variable (usually the number of arrivals in aparticular destination) is regressed on the tourist's disposableincome and a group of cost factors for the examined area and itscompetitors, such as prices of local tourist products, exchange rates,and transportation costs (Archer 1976; Artus 1972; Gray 1966;Johnson and Ashworth 1990; Sheldon 1990; Witt and Martin 1987).In some occasions dummy variables are introduced to pick up theeffect of special events (Gunadhi and Boey 1986; Loeb 1982; Wittand Witt 1992). Moreover, the inclusion of lagged variables or of a

time trend (Martin and Witt 1988; Witt 1980) captures dynamic el-ements unless these are explicitly modeled (Syriopoulos 1995).The construction of demand systems is closer to the ideal or

microeconornic theory as complementarity or substitutability withother tourism areas is fully taken into account. Most of the litera-ture applies the Almost Ideal Demand System originally developedby Deaton and Muellbauer (1980a); the market shares of the desti-nations under consideration are regressed on the representativetourist's total expenditure; the price of each area, and a compositeprice index. Dummy variables or a trend are also introduced. Policyimplications are drawn on the basis of the estimated income andprice elasticities (O'Hagan and Harrison 1984; Papatheodorou 1999;Syriopoulos and Sinclair 1993; White 1985).Despite its contribution and prominence for tourism research,

however, the application of the traditional demand theory in tour-ism suffers from a number of serious drawbacks, as it ignores theparticularities or the product. First, the assumption of a representa-tive tourist who visits simultaneously all the destinations under COIl-

sideration is highly unrealistic. In fact, consumer heterogeneity is astylized fact and all the efforts of marketing aim at discovering andtargeting specific leisure groups. Moreover, tourism cannot be com-pared to supermarket shopping, as the former is a time-consumingactivity applied to different spatial entities.Second, the static nature of the traditional demand theory cannot

account for the evolutionary features of the tourism product,namely the emergence of new destinations and the withering ofothers (Butler 1980). More importantly, however, conventionalmicroeconomics assumes the existence of a homogeneous good and

does not consider disparities in the horizontal dimension. This is aserious drawback as variety is a central issue in leisure studies. Infact, the well-known mathematical model of horizontal differen-tiation introduced by Dixit and Stiglitz (1977) cannot be appliedsatisfactorily in the context or tourism. In addition to the previouspoints of criticism, this approach assumes symmetric consumer pre-ferences over the variety of goods. On the other hand, it is highly



16 6 WHY PEOPLE TRAVEL

unlikely that a tourist would ever treat the Greek tourism product

in a similar way to the Chinese or to the Mexican one. Although itcan be argued that symmetry may survive within a narrower groupof destinations (such as in the Eastern Mediterranean Region), thetraditional theory fails to explain clearly the foundations of such aclassification process. Same conclusions hold for disparities in thevertical dimension; mainstream microeconomics cannot account forquality differentiation which is also very crucial in tourism.Last but not least, the classical theory can only function within a

competitive environment where the producers act as pathetic pricetakers, who are incapable of coordinating their strategies or ofmanipulating tourist flows. However, if the trend towards globalmarket consolidation through mergers and acquisitions (The

Economis t 1998) is combined with the gradual emergence of dualindustrial structures in tourism (Papathcodorou 2000), it becomes

apparent that the suppliers are potentially able to reap the advan-tages of their oligopolistic and oligopsonistic power to the detrimentof consumers and destinations (Debbage 1990). As a result, thecompetitive framework is invalidated and a serious identificationproblem (Gujarati 1988) emerges in the empirical demand researchsince the latter ignores the importance of the tourism supply sidealtogether.For all the above reasons, it is believed that the mainstream

demand theory cannot justify satisfactorily the movement of touristflows in space and time. Therefore, in order to face the classicalcaveats and offer a holistic answer to this challenging problem, thepresent paper applies the Gorman/Lancaster characteristics frame-work (Gorman 1980; Lancaster 1966, 1971) in the context of tour-ism. The idea is definitely not new, as Rugg (1973) and Morley

(1992) have previously dealt with this topic. Interestingly, however,the analytical potential of this framework has not been fullyexplored and its appropriateness for tourism has not been clearlystressed. This paper introduces a discrete choice version of themodel and a comparative exercise is undertaken with graphical sup-port. The implications for policymaking and future research arealso discussed.

TOURISlvl AND CHARACTERISTICS THEORY

In contrast to the traditional consumer theory where the econ-omic agents derive utility directly from goods, Gorman andLancaster argue that utility is related to the consumption of theproducts' intrinsic properties, namely characteristics. Preferences

are assumed to be well behaved, and in addition to the budget con-straint, a system of equations representing the consumption tech-nology is introduced on the constraints side. This technology isuniversal, objective, linear, and additive and describes a transform-ation process with goods as inputs and characteristics as outputs.Formally, the collection of the ith characteristic Z i possessed bysome collection ( x ] , -"2 , ... , Xli') of goods} is given by the equation:



ANDREAS PAPATHEODOROU 16 7

N

Z i = I)ij.\jj=l

where b U is the consumption technology coefficient.

(1)

The Model

Rugg (1973) was the first to apply the characteristics approach totourism, by introducing the following model in a system of N desti-nations:

max U =(Ztour)

s.t. Ztour =GXtour

Y2'::jhourX tour +PtransC

T ~ c ' X tour +j)(ransc

(2)

Z, x,j), d 2 ' : : O Y , T ~ O

where U is the consumer's utility function, Ztotlr is the column vectorof tourist characteristics in each destination, G is the matrix of con-sumption technology coefficients, X1llur is the column vector of thecomposite tourism product defined as the number of days spent ineach destination, Pl11ur and jJ tratis are the row vectors of the compo-site prices and transport costs, respectively, dtratls is the row vectorof transport time between the origin and each of the available desti-nations, c is a column vector whose elements are all one, Y is avail-

able expenditure, and T is time available for tourism. All elementsare considered to be non-negative. The problem may be solved byusing non-linear programming methods.The explicit assignment of origin and destination tags in this

model, however, creates a paradox of location. In particular, thetime constraint encapsulates the distance factors between the originand the destinations; nothing is implied about the distance amongthe various destinations. Unless the tourist is assumed to travel to asingle destination, this problem may put the whole characteristicsspace mapping into question. Consequently, internal consistencyrequires the adoption of a discrete choice framework, where theconsumer travels solely to the resort that is associated with thehighest utility (Ben-Akiva and Lerman 1985; Morley 1994); after all,multi-destination tourism has only a very minor share in the real

world.Diagrammatically, the analysis is encapsulated by Figure I,where quantities of facilities and attractions are measured on thehorizontal and vertical axis, respectively. Visits to destinations A, B,

and C generate characteristics in the proportions a, b, and c, asshown by rays OA , OB , and ~C , respectively. These rt:!Ysqf character-

istics summarize the information contained in the consumption tech-





ANDREAS PAPATHEODOROU 1 6 9

marion and advertising, agglomeration, and, finally, the emergence

of new destinations.

T h e C om j) ar ati ve E xe rc is e

Expenditure and T ime C onstra ints. An exogenous increase(decrease) in available expenditure or time for tourism shiftsthe corresponding points outwards (inwards) in a parallel way.For example, a higher salary or an enlargement of the paid va-cation period boosts tourist activities. However, changes in thetime constraint can also be endogenous and affected by trans-port innovations, which, though universally applied, may favoronly a particular destination. For example, whereas up to themid-60s, the South was Britain's Cote d'Azur, competition-induced rapid technological innovations in the air transportindustry rendered the Mediterranean and later the CaribbeanRegion widely accessible to British tourists. Given the superior-ity of the latter regions in terms of sunlust characteristics, theBritish summer resorts entered gradually into a period of crisis.The time shrinkage of the world has also been accompanied by

transport price reductions in real terms; as a result, changes in thetime constraint affect the expenditure one as well. Two furtherexamples of interdependence come to mind. First, unless vacation ispaid, the well-known trade-off between remuneration and leisureaffects substantially the relative position of the two constraints. Asan illustration of the second, one may consider the choice betweentwo transport modes with costs (travel time) C l (t 1 ) and c 2 ( / 2 ) , re-spectively, where (1 < C2 but t 1 > i'l. In this case, the consumer con-siders the g en er al iz ed d iff er en tia l c os t of'mode 2, D2 (Quandt (970):

(3)

where (-a) is the rate of substitution between cost and time. Inother words, if the consumer enjoys a journey time decrease afJ.t,

the user is ready to pay for a cost increase equal to fJ.C in order toremain indifferent. This rate is assumed to be the same for all indi-viduals in the same income class; however, it is found to increase inits absolute value with the level of prosperity. Said differently, asone's income grows, the time available for tourism rises, since oneis willing and able to switch from low cost/slow transport modes, tohigh cost/fast ones. The secluded, albeit luxurious resorts of theCaribbean have benefited from this trend.

Prices. The price of the tourism productj)tuur

is given by theweighted sum of the prices in its constituent industries, mainly ac-commodation, entertainment, and catering. Similarly, the transportprice J1trans incorporates all transport modes used during the trip. Inboth cases, the ability of each industry for mark-up pricing can bequantified by the following mathematical formula, which relatesprice with marginal cost, elasticity of demand, and competitive con-



I() WHY PEOPLE TRAVEL

duct (Port er 1983);

P-MC

P=

o, P;?;ATC

e(4)

where P, AIC stand for price and marginal cost, respectively, e is theprice elasticity of demand, 0 is a measure of the industry's competi-tive conduct, and ATC represents average total cost; unless price ishigher than ATC, industrial sustainability is undermined.Price ranges between the levels determined by perfect (or

Bertrand) competition P =MC for e =0 and profit maximizingmonopoly P = eMC /(1 + e) for e = 1 . In other words, the competi-tive conduct measures the degree of coordination within the indus-try group; a large (close to one) value of 0 reflects the existence ofmarket cartelization. It is suggested that the conduct depends on

the prevailing industrial conditions (Scherer and Ross 1990):

e =g(C,B,X) (5)

where C measures industry concentration, B is a vector of entry bar-rier measures, and X represents a set of other industry character-istics. In addition, as a significant number of vacations is sold bytour operators, the competitive conduct of these intermediaries andthe resulting bargaining game with the tourism suppliers should beexplicitly taken into consideration. In other words, a more formalanalytical treatment requires that e in Equation (4) should bereplaced by 1 1 , where:

1 1 =h ( O , c p ) (6)

and q > is an aggregate measure of the competitive conduct in thetour operators industry.It should be noted, however, that although the relation between

concentration, entry barriers, and conduct is relatively clear in astatic context, the interdependence of these variables in a strategicdynamic environment complicates significantly the price forecasts.For example, in order to disfavor market entry the incumbents mayengage in limit pricing; otherwise, and following the Schumpeterianparadigm (1996), they may utilize their profits for research anddevelopment on product innovation and achieve an endogenousdecrease of their production costs.Having the above in mind, differences in prices among the var-

ious resorts may be attributed to a variety of reasons. First, theexisting spatial configurations may involve technologies associated

with different cost structure. For example, urbanization in resort Amay be based on small hotels with low fixed outlays but high unitcosts, whereas resort B may be characterized by large establish-ments which rely on scale economies. Similarly, A may be accessibleonly by ship, whereas B may offer a wider selection of transportmodes. Second, the demand sensitivity in price changes may varyacross resorts for reasons of uniqueness.



A. J '\ lDREA.S PAPATHEODOROU 1 71

Most importantly, however, one should explicitly consider the role

of competitive conduct. The market structure of the small hotels inresort A of the previous example may be dose to perfect compe-tition and hence vulnerable to tour operation oligopsonists. On theother hand, ship-owners in A may take advantage of their accessibil-ity monopoly and acquire super-profits. In general, the higher theprice of a resort, the lower the maximum purchasable number ofdays in it. Consequently, a price increase shifts the choice set pointsto the left, whereas a price decrease shifts them to the right. Forexample, the decision of suppliers in resort B to form a cartel maylead a considerable number of tourists to shift to C or A , c ete ris p ar i-bus. In the context of international tourism, differences in competi-tive conduct may be powerful enough to generate substantialdeviations from the law of one price, both for spatially fixed and fortradable tourism goods.

Con sume r P r ef er e nc es . These are given by the utility function, andthe curvature of Uo indicates the relative importance of the twocharacteristics in Figure 1. In Plog's (1973) terminology, the allo-centrics are expected to exhibit a greater interest for attractionsthan facilities; thus, their utility curve is relatively flat. The oppositeis expected for psychocentrics, whereas Figure 1 is closer to thedescription of a midcentric. If for some reason consumers becomemore facilities oriented, they may choose resort C instead of B,c e te r is par ibus . Within a dynamic framework, allocentric endogenouspreferences encourage dispersion or tourist flows, whereas psycho-centric ideology induces the fortification of prevailing concentrationpatterns among destinations. As argued later, spatial and environ-mental considerations are very important in this context.

Q Jla lif;)" I nfo rm a tio n, a nd A d ue rtisin g. Traditional economics hasabstained from vertical product differentiation issues as quality is avalue of the intrinsic properties of goods and consequently cannotbe analyzed within the standard framework. This notional gap iscovered suitably by the Gorman/Lancaster framework where a pro-duct is better qualified than another if it offers a larger bundle ofcharacteristics for the same quantity. Moreover, in the case of incre-mental differences, better quality is associated with higher price sothat all goods may remain on the efficient choice set. On the otherhand, as a result of drastic quality disparities and innovations, anumber of products may offer a dominated bundle of characteristicsand consequently drop out of the market. The latter result deviatessubstantially from traditional consumer theory, which cannotaccount for market exit.Quality disparities are apparent in tourism; hotel star classifi-

cation and partition of the airplane's body into different serviceclasses provide some good examples. Moreover whole resorts,regions, or even countries may be associated with a specific brandimage (Urry 1990). Quality innovations may be used as a strategicvariable in the competitive process; business magazines are full of



1 72 WHY PEOPLE I'M.VEL

advertisements about new amenities in five-star hotels and business

class air-seats. By providing these frills, the up-market global tour-ism suppliers try to justify their price premia and persuade consu-mers about their value [or money. As shown by Keane (1996), theinterpretation of high prices as a signal of quality is closely relatedto the experience nature of tourism and is most apparent in thecase of repeat-visiting, where reputation effects are strong.Not surprisingly, wealthier consumers are usually more willing to

pay for high quality than the poorer ones, as they have a lower mar-ginal utility of income and a higher taste parameter (Gabszewiczand Thisse 1979). As a result, and given the positive trend of worlddisposable income and leisure expenditure, up-market tourismareas are likely to prosper in the future to the detriment of regionswhich offer a cheaper but less sophisticated product. However, thishas nothing to do with the Veblen effect of conspicuous consump-

tion, which induces the tourist to pay a higher price for a function-ally equivalent good simply for reasons of fashion, image, andprestige (Bagwell and Bernheim 1996). For example, occupancyrates in Mykonos are always very high, despite its expensiveness incomparison with other equally beautiful islands in the Cyclades,Greece.Diagrammatically, quality issues may be treated similarly to in-

formation ones. The latter are particularly important within a time-series context, where it seems plausible for a consumer to acquireknowledge about the characteristics of a tourism product eitherthrough search and exposure to advertising or through personal ex-

A

o facilities

Figure 2. Tourism and Advertising




perience and repeat-visiting. In this case, utility becomes a function

of lfifictive tourist characteristics. Based on Auld (1974), Figure 2 il-lustrates some further points graphically.More specifically, one assumes for simplicity that the two con-

straints do not cross, so that the focus is on the tighter one only; F,G, and AI are the points of interest. Assume that, for some reason,quality in destination B improves or its marketing promotion is sue-cessful, One possible outcome is that the consumer becomes awareof receiving more of both characteristics from B in an unchangeableratio. In this case, one observes a movement out along the OB

characteristic ray, to a variable point called G'. If G' is between Gand D, then all three countries remain efficient. If G' is between Dand K, destination .4 becomes dominated by B and disappears,whereas if G' is to the right of K, then B is the sole place to be vis-ited. However, it is also possible that the new quality (or infor-

mation) elements affect consumption technology in such a way thatthe ratio of the two characteristics for destination B changes. In thiscase the relevant characteristic ray rotates. Consider, for example,the ray OB* and the point L, where B dominates .4, limiting choicebetween Band C.Both formal and informal advertising are expected to enhance

the popularity of tourism. For example, during the 50s, theMediterranean market was almost monopolized by Italy, as touristswere virtually unaware of any other country to visit in the region.However, gradually but steadily, the whole region was "discovered"and the market shares changed substantially (Papatheodorou 1999).In other words, information is an explicitly dynamic feature, whichis also affected by reputation and endogenous preferences.Furthermore, the experience nature of the product renders the in-formation and advertising channels controllable by the intermedi-aries to a significant degree. As a result, both the industrialorganization of tourism and its relation with the suppliers of resortservices are expected to influence the available information andbookings of packages.

Agglomeration. The assumptions of limitless linearity and additivitymay be suitable for an agricultural consumption technology, but areproblematic in the context of tourism, where space and environ-ment generate interdependence of characteristics. In this sense, it isimportant to consider issues of spatial fixity and plasticity (Clarkand Wrigley 1995). More specifically, while the ray slope may beconstant at a single point in time, it is expected to change dynarni-cally, because of the interaction between centripetal and centrifugalforces at an aggregate level. Persisting destination attractivenessinduces sunk investment and favors agglomeration. In fact, tourismurbanization is associated with a multiplication of the available fa-cilities for a given level of natural attractions. This may be pursuedto such a degree that the destination divorces from the geographicalenvironment (Wolfe 1952). As a result, the ray of characteristicsbecomes Hatter over time. This trend is reinforced by environmental



WHY PEOPLE TRAVEL

degradation in case the increased tourism revenue is not used for

the restoration and improvement of the attraction points. Tourismurbanization, however, is also associated with the creation of builtattractions, such as theme parks; this latter trend renders the over-all relation of facilities to attractions indeterminate. Complexityrises further, when one considers the implications of land rentincreases and social intolerance, which discourage tourism develop-ment, both in terms of facilities and built attractions.Again, industrial organization issues are important. Both within

an intra-resort and inter-resort framework planners and managersmay follow a product differentiation strategy in order to relax com-petition and counteract the effects of agglomeration shadows that amajor resort may have on a smaller one. In this sense, the bundleand proportions of available characteristics may he strategicallydetermined in a dynamically competitive framework.

Emergence qf a N ew D e stin atio n. The explicitly systemic approach ofthe characteristics framework is further enhanced by its ability toaccount for the emergence of new resorts. The associated impli-cations are interesting, particularly when the new destination domi-nates only partially, rather than fully, the existing ones. Figure 3illustrates graphically this point.Consider the characteristics ray OD for the new destination D

and assume that the expenditure (time) constraint is associatedwith point X (T). In this case, destination C becomes dominated byD with respect to the time constraint, but remains efficient againstthe expenditure one. As a result, all four destinations remain on the

A

D

facilities

Figure 3. Introduction of a New Destination, the Case of Partial Dominance



ANDRE/I.S PAPATHEQDOROU 17 5

efficient choice set consisting of bundles corresponding to points F,

G, H, and X. However, if the expenditure constraint shifts suffi-ciently upwards and to the right, so that there is no notional cross-ing between the two constraints, then C disappears completely fromthe market. Given the positive trend of tourism expenditure on theone hand, and the relative rigidity of the "natural" time constrainton the other, this outcome seems plausible and provides a furtherexplanation for the domination of the Mediterranean andCaribbean regions over the British sunlust resorts.Furthermore, it should be always borne in mind that the intro-

duction of a new destination is almost exclusively determined by theeffective knowledge about it. Whereas Butler's (1980) resort life-cycle argues that during the involvement stage, tourist activities inthe terra nova are controlled by the locals, resorts are not unusuallydiscovered and promoted by large tour operators from the origins or

simply by externals to the region who possess the know-how.Government planning and investment may also produce instantresorts, as in the case of Cancan (Gordon and Goodall 1992). Again,the supplier's identity and the underlying industrial structure playaprominent role.

CONCLUSION

Therefore, it seems that the characteristics approach can offer anexplicitly systemic framework, where actual destination choice isbased on a set of solid rnicrofoundations, Moreover, the drawbacksof the traditional demand theory are confronted in a neat and effi-

cient way; taste heterogeneity and discrete choice in space are expli-

citly recognized, horizontal and vertical differentiation becomeinherent features of the model, and a number of dynamic issuessuch as information, advertising, agglomeration, and the emergenceof new destinations is sufficiently addressed. Most importantly, theanalysis reveals the importance of corporate power and industrialorganization, whereas the graphical support offers an integratedmapping of the model and the comparative exercise. On thesegrounds, it is believed that the proposed framework provides a moreprofound, accurate, and holistic explanation of tourist flows thanthe traditional theory. As a result, it should be seriously taken intoconsideration both by researchers and policymakers.As to implications for research, the characteristics framework is

undoubtedly more complicated than the classical demand theory.Consequently, it is not surprising that it has not been applied exten-

sively in the empirical tourism research so far. Generally speaking,the closest econometric methodology to this alternative approach isoffered by the hedonic price analysis, where the price of a product isregressed on a set of characteristics (Lancaster 1971; Triplett 1975).The estimated coefficients measure the explanatory importance ofthese features for the observed value of the good. Moreover, theseshadow prices may be subsequently used for the construction of a



176 WHY PEOPLE TRAVEL

budget and/or time constraint and assess the competitiveness of the

particular product.In the context of tourism, a hedonic price study would regress the

cost of living in a particular destination (or the price of the respect-

ive holiday package in the case of inclusive tours) on a number ofsunlust, wanderlust, and infrastructure characteristics (as well as onthe various features of the vacation deal with the tour operator).The analysis could be facilitated by the construction of attractionindexes, similar to those used in location studies and in economicgeography. In fact, this line of methodology has been followed to alarge extent by the relatively scant empirical literature in the area,which aims at assessing the price competitiveness of package desti-nations and tour operators (Clewer, Pack and Sinclair 1992;Sinclair, Clewer and Pack 1990; Taylor 1995). It should be noted,

however, that all these studies assume the existence of a competi-tive framework (Rosen 1974), whereas this analysis has stressedthat tourism suppliers are potentially able to take advantage oftheir corporate power. The future research should take this issueseriously into account; one possible suggestion is the introduction ofa two-step hedonic price analysis, where some of the initially esti-matcd coefficients arc subsequently regressed on a number of vari-ables related to market structure and profitability.

Interestingly, the characteristics framework can provide usefuladvice to the tourism authorities and policymakers by stressing theparticularities and interdependencies that emerge in the case of theproduct. At a first level, it should be widely understood that pricingpolicies should not be pursued independently from quality, advertis-

ing, and urbanization strategies, as they all affect the tourism pro-duct in different, albeit related, dimensions. Needless to say,sustainability considerations should play a dominant role in allcases (World Commission on Environment and Development 1987).

Most importantly, however, and in conjunction with the TourismSatellite Accounts (World Travel and Tourism Council 1999), thecharacteristics approach contributes to the solution of the supply-side conundrum of tourism (Smith 1998). It should be apparentfrom the analysis of this paper that the actual destination choice is

at least subject to the prevailing conditions in the transport, accom-modation, and tour operation sectors. But in most regions orcountries these activities are controlled by different ministries ordepartments; the proposed policies are rarely coordinated or may beeven mutually invalidated (Wheatcroft 1994). Therefore, to facethis problem the policymakers should overcome the emerging politi-cal economy difficulties and adopt a holistic approach to the indus-tty by pursuing comprehensive policies which augment thebargaining power of their regions/countries and establish a morefair and stable business relation with the international tourism con-glomerates.1f



ANDREAS PAl'ATHEODOROU 1 77

REFERENCES

Archer, B. H.1976 Occasional Papers in Economics. Bangor: Univcrsiry of Wales Press.

Artus,]. R .

1972 An Econometric Analysis of International Travel. IMF Staff Papers19:579-614.

Auld, D. A. L.1974- Advertising and the Theory of Consumer Choice. Quarterly Journal ofEconomics 8S:4S0-487.

Bagwell, L. S., and B. D. Bernheim1996 Veblen Eifccts in a Theory of Conspicuous Consumption. AmericanEconomic Review B6:349-373.

Bcn-Akiva, M., and S. R. Lerman1985 Discrete Choice Analysis: Theory and Application to Travel Demand.Cambridge, l\1A: MIT Press.

Butler, R.1980 The Concept of a Tourist Area Cycle of Evolution: Implications for

Management of Resources. Canadian Geographer 14:5-12.

Clark, G. L., and N. Wrigley1995 Sunk Costs: A Framework Ior Economic Geography. Transactions mGNS 20:20'1-223.

Clower, A., A. Pack, and M. T. Sinclair1992 Price Competitiveness and Inclusive Tour Holidays in European Cities.In Choice and Demand in Tourism, P. Johnson and B. Thomas, cds., pp. 123-143. London: Mansell.

Deaton, A., and J. Mucllbauer1980a An Almost Ideal Demand Svstcm. American Economic Review 70:3(2-326. .

1980b Economics and Consumer Behavior. Cambridge: Cambridge UniversityPress.

Dcbbagc, K. G.1990 Oligopoly and the Resort Cycle: in the Bahamas. Annals of TourismResearch 17:513-527.

Dixit, A. K, and J . E. Stiglitz1977 Monopolistic Competition and Optimum Product Diversity. American

Economic Review 67:297-308.T h e E c on omi st

1998 Travel and Tourism Survey. The Economist 34·6 (january),Gabszcwicz, j., and J. F. Thissc1979 Price Competition, Quality and Income Disparities. Journal of EconomicTheory 20:340-359.

Gordon,!., and B . Goodall1992 Resort Cycles and Development Processes. Built Environment 18:41-56.

Gorman, W. M.1980 A Possible Procedure for Analyzing Quality Differentials in the EggMarket. Review of Economic Studies 4·7:843-!l56.

Grav, H. P.

1966 The Demand for International Travel by the United States and Canada.International Economic Review 7:83-92.

Gujarati, D. N.1988 Basic Econometrics (2nd cd). Singapore: McGraw-Hill.

Gunadhi,. H., and C. K. Boey

1986 Demand Elasticities of Tourism in Singapore .. Tourism Management7:239-253.

Johnson, P., andJ. Ashworth1.990 Modeling Tourism Demand: A Summary Review. Leisure Studies [1:145-160.

Keane, M.J.1996 Sustaining. Quality in Tourism Destinations. Applied Economics 2B:15'W-1553.






Syriopoulos, T. C., and M. T. Sinclair1993 An Econometric Study of Tourism Demand: The AIDS Model of US and

European Tourism in Mediterranean Countries. Applied Economics 25: 1541-

1552.

Taylor, P.1995 Measuring Changes in the Relative Competitiveness of Package TourDestinations. Tourism Economics I:169-182.

Triplett,]. E.1975 Consumer Demand and Characteristics or Consumption Goods. InHousehold Production and Consumption, N. Terlcckvi, cd., pr. 305-323.Washington, DC: National Bureau of Economic Research.

Urry,j.1990 The Tourist Gaze: Leisure and Travel in Contemporary Societies.Newbury Park, CA: Sage.

Whc.atcroft, S.1994- Aviation and Tourism Policies: Balancing the Benefits. London:Routledge.

White, K. j.19S5 An International Travel Demand Model: US Travel to Western Europe.

Annals of Tourism Research 12:529-545.Witt, S. F.19110 An Abstract Mode: Abstract (Destination) Node Model of ForeignHoliday Demand. Applied Economics 12:163-I!lO.

Witt, S. F., and C. A. Martin1987 Econometric Models for Forecasting International Tourism Demand.Journal of Travel Research 25(3):23-30.

Witt, S. r., and C. A. Witt1992 Modeling and Forecasting Demand in Tourism. London: Academic Press.

Wolfe, R . I.1952 Wasaga Beach: The Divorce I t ) ! " the Geographic Environment. CaJHldianGeographer 2:5i-66.

World Commission on Environment and Development1987 Our Common Future. Oxford: Oxford University Press.

World Travel and Tourism Council1999 Travel & Tourism Satellite Accounting Research Estimates andForecasts. W'lTG, htlp:/h\·ww.wttc.org.

Submit ted 6 October 1997. Resubmit ted 18 Dec embe r 1999. Accep ted 18 ja nu ary ' 2 00 0.R ef er ee d a n on ym a us ly . C o or di na ti ng E d it or : B o rn Vukunii

Documents

Why People Travel to Different Places