at SciVerse ScienceDirect

Tourism Management 36 (2013) 435e440

Contents lists available

Tourism Management

journal homepage: www.elsevier .com/locate/ tourman

Prices and brand diversity in touristic areas supermarketsq

Javier Campos, Juan-Luis Jiménez*, Ancor Suárez-AlemánUniversity of Las Palmas de Gran Canaria (ULPGC), Canary Islands, Spain

h i g h l i g h t s

< We compose a database of supermarkets in Gran Canaria.< We analyse the effects of tourism on the destination retail markets.< Price and brand equations are estimated using GIS techniques.< Prices in touristic areas are higher than those in other areas.< The number of brand varieties is lower than in non-touristic areas.

a r t i c l e i n f o

Article history:Received 16 April 2012Accepted 29 September 2012

Keywords:Tourism effectsPricesBrand diversitySupermarketsCanary Islands

JEL codes:L83L13

q Although the usual disclaimers remain, we acknofrom Jordi Perdiguero, Aday Hernández, Carmen Greferees. We appreciate GIS analysis by José Domíngfunding by the ACIISI Research Program (PROID20100* Corresponding author. Universidad de Las Palmas

Economía, Empresa y Turismo, Campus de Tafira, 3501928 458 191.

E-mail address: jljimenez@daea.ulpgc.es (J.-L. Jimé

0261-5177/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.tourman.2012.09.016

a b s t r a c t

Using a dataset of consumption patterns in the island of Gran Canaria collected by the authors, this paperattempts to quantify some non-positive effects of tourism on destinations retail markets for goods andservices. In particular, we empirically prove, controlling by factors such as population, size of super-markets or number of competitors, two main effects: first, that supermarkets located in touristic areascharge higher prices than those in non-touristic areas; and second, that brand diversity is lower in thesame stores, particularly in the case of smaller ones. These results confirm that local population do notalways benefit from living in a touristic city and possibly provide a more balanced view on the positiveand negative sides of tourism.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Many foreigners often see living in a touristic destination withenvy. During a large part of the year, and without suffering thenuisances of packed travelling or inconvenient accommodations,local residents enjoy at home the benefits from a benign climate,beautiful surroundings, ande sometimes e a dynamic society withplenty of cosmopolitan atmosphere. Although this paper does notnegate this evidence, it intends to show that there also exists anextra cost in living in such paradisiacal places.

wledge valuable suggestionsarcía, and three anonymousuez. The authors also thank209).de Gran Canaria, Facultad de7 Las Palmas, Spain. Tel.: þ34

nez).

All rights reserved.

With some notable exceptions and purely due to geographicalreasons, most popular sun-and-beach destinations tend to belocated in countries or regions with lower GDP per head than theplaces where touristic flows originate. According to Eurostat (2011),more than 150 million people from the UK, Ireland, Germany andthe Scandinavian countries fly southbound every year to theMediterranean shores of Spain, Greece or North Africa or to theAtlantic beaches in Portugal or the Canary Islands.

One of the most widely studied positive effects of thisphenomenon is the revitalization of local economic activitybrought by higher income visitors. When arriving at their desti-nation, tourists buy goods and services. Most surveys show that thelonger their stay abroad the higher tends to be their spending perhead. For example, the Spanish Tourism Expenditure Survey 2010(available online at www.iet.tourspain.es) confirms that theirexpenditure also increases when the difference with locals in termsof purchasing power parity (as compared to their prices and wagesat home) is larger. When tourists stay at non-hotel accommoda-tions (apartments or privately rented houses) or travel by

J. Campos et al. / Tourism Management 36 (2013) 435e440436

themselves (instead of booking holidays packages or all-inclusiveprograms) their spending at local stores is generally larger andmore frequent.

Of course, managers and local retailers see this wealthierdemand segment as an opportunity tomake profits. Although somegoods and services providers (crafts or souvenirs sellers, touristicrestaurants and bars, etc.) may decide to specialize on this partic-ular clientele, others (groceries, supermarkets, bookshops, etc.) willsell both to tourists and locals and, since price discrimination seemsunfair (and barely legal), it can be expected that (large) tourisminflows on certain destination areas will induce e as a result ofa simple income effect e higher (average) prices in most typicalconsumption baskets.

However, price is not the only decision variable that consumerscare about. In horizontally differentiated markets most retailersoffer a number of brand varieties for the same product in order toattract consumers with different tastes or preferences. Productdifferentiation is then supported with the help of advertising(either in place or via the media), attractive packaging or specificpromotion policies and discounts. But all these resources havea weaker effect on tourists, whose command of local language islimited. Therefore, and particularly in smaller shops e whereselling space is more valuable ewe can expect that brand diversityin stores of touristic areas will be lower than in non-touristic ones.1

Are these expected negative effects relevant enough? Shouldthey be included in any balanced review on the effects of tourismfrom now on? After a review of the related literature in Section 2,this paper addresses these two questions from an empirical view-point by providing evidence from a 2010 Canary Islands paneldataset. The example of the Canary Islands seems particularlyappropriate to test the claims made in this paper because thisarchipelago, located 1500 km southwest of mainland Spain,receives regularly every year more than 12 million European visi-tors, whereas the local population is about 2.1 million. On average,the ratio visitors/locals is above 6, although in some touristicmunicipalities these figures are closer to 10e12 (ISTAC, 2010).

As explained in Section 3, our source includes very detailedinformation on prices and brand varieties for a wide subset ofcommodities in a representative sample of all the supermarkets ofthe island of Gran Canaria. An additional relevant feature of our datais that stores have been exactly located using GIS techniques, whichallows a precise (but flexible) definition of geographic markets inconnection with the influence areas of touristic flows. We thenestimate in Section 4, several price and brand variety equations inorder to test the impactof tourismoneach supermarket according toits location (or not)within a touristicmunicipality.Wecontrol by thesize of the local market, the size of the stores and the existence (ornot) of nearby competitors. By using clustering techniques that takeinto account potential differences in variance amongmunicipalities,we finally produce estimates that confirm our expected results,which are analysed and discussed in Section 5.

2. The impact of tourism on prices and brands: previousresearch

The existing literature on the negative impacts of tourism overthe host community has traditionally classified them into threebroad categories: environmental, social effects, and purely

1 Although it is beyond the scope of this work, a theoretical support for theseideas can be found on many papers that model rivalry in price and variety amongsupermarket retailers (see Richards & Hamilton, 2006, for example), which usesa nested constant elasticity of substitution framework. There are not specificexamples for touristic supermarkets.

economic impacts. The first of these research lines is the mostextensive (see for example, Krippendorf, 1982; Lindberg & Johnson,1997; Mihalic, 2000; Orams, 1995; Romeril, 1989, among others). Ithas mainly focused on the relationship between tourists and resi-dents in terms of conflicting preferences over environmentalconservation (see Bujosa & Rosselló, 2007), or the alternative usesof existing natural resources (Concu & Atzeni, 2012). The secondcategory identifies the disruption of social relations (also Lindberg& Johnson, 1997; Thyne, Lawson, & Todd, 2006), or the changes inresidents’ attitudes and perceptions about foreigners (Butler, 1980;Diedrich & García-Buades, 2009; Lawson, Williams, Young, &Cossens, 1998; Mason & Cheyne, 2000; Ross, 1992, among manyothers) as the main social negative impacts of tourism.

The third category of negative effects has been much less studiedso far and, in particular, there are few studies on how the destinationmarkets for goods and services are affected by touristic flows.Harcombe (1999) and Mason (2008), for example, follow a macro-economic approach. They include as negative economic conse-quences of tourism both the opportunity costs for a society (ofdeveloping the tourism industry rather than other economic activi-ties, with the subsequent risk associated to sectorial over-dependence) and the tourism-driven inflation instability (caused byan extra and often fluctuating demand on local services), but do notquantify these effects.

Following a different approach, Sharpley and Telfer (2002)develop a theoretical analysis of the consequences of tourism onprices. They show that tourism may result in demand-triggeredinflation at destinations when visitors bring additional financialresources into host communities where the supply of goods andservices is not fast enough to adapt to the new demand. Sancho,García, and Rozo (2007) also explicitly consider tourism asa source of inflation, not only for commodities and basic products,but also in housing and land prices.

From an empirical point of view, Lawson et al. (1998) providesome evidence about the idea that tourism inflates the cost of livingfor locals. In their study for New Zealand, they find that priceincreases in touristic places may be so high that they even excludesomeNewZealanders. Horn and Simmons (2002) argue that in largecities benefited by tourism flows prices have fallen due to thebuilding of large malls and shopping centres, but the opposite hashappened in small communities. Another empirical study is Garcíaand Sancho (2000), who quantify how local population in fourtouristic Spanish regions perceived the causes of increased localprices. Torres (2003) argues that touristsnormallyenjoy their leisureactivities in places with prices lower than their home-cities, andshows that their demand induces a price increase at destinations.

It is not easy to find other studies on the impact of tourism onother microeconomic market mechanisms (in terms, for example,of product differentiation, location, entry or consumption patterns).Similarly, none of the most widely cited empirical papers on pricingin supermarkets that consider different consumer groups makea special consideration for tourism. Blinkey and Connor (1998), forexample, shows that a lower concentration level implies cheaperprices when consumers’ income is heterogeneous. Aalto-Setälä(2002) states that supermarket chains with larger market shareenjoy higher mark-ups, whereas Griffith and Harmgart (2008)conclude that barriers to entry increase equilibrium prices, inboth cases with two groups of consumers.

However, as explicitly pointed out by Richards and Hamilton(2006), it is quite clear that firms’ marketing policies considerboth price and variety as central elements. Although there areseveral other examples in this literature that explicitly consider theeffect of brand diversity and prices for foreigners (see for examplethe recent survey by Winsor, Mak, & Hsu, 2010), there still existsa gap in the empirical literature that studies the negative economic

Table 1Overall size distribution of supermarkets and sample size.

Sizea Number ofsupermarkets

Sample Percentage ofsampledsupermarkets

Size 1: less than 120 m2 341 40 12%Size 2: between 120 and 399 m2 208 24 12%Size 3: between 400 and 999 m2 68 6 9%Size 4: more than 1000 m2 51 9 17%Total 668 79 12%

a Supermarket size categories were defined according to tax criteria. Source: Ownelaboration based on the Official Business Census of the Regional Government.

Table 2Definition of touristic municipalities.

Municipalitya Population Number oftouristicbeds

Touristic bedsper 1000inhabitants

Is it a touristicarea?

Agüimes 29,431 68 2.31 NoArucas 36,745 41 1.12 NoGáldar 24,473 66 2.70 NoIngenio 29,640 34 1.15 NoMogán 22,638 36,419 1608.76 YesLas Palmas de Gran

Canaria383,308 7298 19.04 No

San Bartolomé deTirajana

53,288 92,417 1734.29 Yes

Santa Brígida 19,135 194 10.14 NoSanta Lucía 64,845 525 8.10 NoTelde 100,900 128 1.27 No

a The table only includes the 10 municipalities with sampled supermarkets(population > 15,000). Source: Own elaboration based on the Regional Governmentstatistical data (ISTAC, 2010). Touristic municipalities in bold.

J. Campos et al. / Tourism Management 36 (2013) 435e440 437

consequences of tourism on destination markets. That is where ourcontribution should be located.

3. Data and variables

In 2010 the island of Gran Canaria, in the Canary Islands, had838,397 inhabitants, which constitutes approximately 40% of thepopulation of the archipelago. The island is divided in 21 munici-palities and receives every year about 2.2 million visitors, particu-larly in the peaks of the winter and Easter seasons.

The empirical analysis carried out in this paper is based upona dataset collected by the authors in January and April 2010 whichincludes information on the prices and brand varieties for a widesubset of commodities sold at local supermarkets. It is a fairlyrepresentative sample since it is built on all the stores located inmunicipalities with at least 15,000 inhabitants. This represents93.2% of all the island supermarkets (688 out of a total of 738,according to the Regional Government Business Census; ISTAC,2010). A stratified random procedure by size was used in thesampling design. Table 1 shows the overall size distribution ofsupermarkets and the sample considered for each category.

The second step in our research was to distinguish betweentouristic and non-touristic supermarkets. Although the entireisland of Gran Canaria is a touristic destination for many Europeancountries, most of them stay during their visit at hotels andapartments located in the southern part of the island, where mostbeaches and touristic resorts are located. In order to develop a ruleto separate between touristic and non-touristic municipalities, weconsidered standard geographic criteria and built up a ratio of thenumber of touristic beds (both in hotels and apartments) perinhabitant as a proxy of the potential impact of tourism on thedestination markets as compared to the local population.

Table 2 shows that only two municipalities, San Bartolomé deTirajana and Mogán, concentrate the 94 per cent of tourism supply(they even have more beds than inhabitants) and can be separatelyconsidered as touristic areas. In fact, according to the SpanishNationalStatistical Office (INE, 2011) both municipalities had the largestoccupation index, 77.07% and 76.60% respectively in the island in2010. Therefore, they also concentratedmost of the touristic demand.

Table 3 finally presents the detailed size distribution of sampledsupermarkets in each municipality. Once each supermarket wasidentified and precisely located within each municipality, a pollstervisited it twice, in January and April 2010, and collected informa-tion on prices, product packaging and number and brands of closersubstitutes for a selected basket of 30 products, representative ofa typical consumption basket. The products included in the studywere rice, cornflakes, spaghetti, noodles, gofio,2 white bread,

2 Gofio is the name given in the Canary Islands to toasted flour made from wheator corn. It is a basic ingredient in the local inhabitants’ diet and, since it is seldombought by foreigners, allows us to consider (and discard) differentiated price effectsbetween touristic and non-touristic products and, therefore, effects on local demand.

chicken breast, fillet, ham, canned tuna, eggs, milk, yoghurt,banana, olive oil, water, lentils, potatoes, beer, cola, coffee, rum,chocolate, sugar, salt, toothpaste, mop, and detergent. To allowcomparisons, the definition of each product was homogenized bysize and presentation, i.e., we gathered the price of a box of whitemedium grain rice (no basmati rice, or other varieties) of 1 kg, andthe number of this type of rice that each supermarket offered.

Apart from price (Price) and the number of varieties per brand(NVarieties) as dependent variables, our empirical strategy e

whose results are summarized in next section e made use of thefollowing explanatory variables:

� SameXmetersjc. This variable includes the number of super-markets of the same chain located close to sampled super-market j at municipality c in a radius of X meters. It has beenconstructed using GIS techniques for all the supermarkets inGran Canaria and establishes a flexible hypotheticalcustomers’ attraction circle around each sampled super-market of X meters, between 50 and 1500, as usual in theliterature on supermarket analysis (see Gómez-Lobo, Jiménez,& Perdiguero, 2011). Since same-chain supermarkets do notact as competitors, we expect the sign of the estimatedparameter for this variable to be positive with respect toprices and negative for brand varieties.

� RivXmetersjc. This variable represents the number of super-markets of different chains (competitors) located close tosampled supermarket j at municipality c in a radius of Xmeters.Its construction procedure is similar to the previous one, butexpected signs are just the opposite.

� PopulationXmetersjc. This variable is the local populationsurrounding the supermarket j in municipality c (that is thepotential number of customers).3 It captures the effect ofmarket size on the supermarkets’ behaviour. A priori, it shouldbe positive in prices and in brand varieties.

� Touristicjc. This is a binary variable directly built from Table 2.It takes value 1 if the supermarket j is located at a touristic area(that is, the municipalities of San Bartolomé de Tirajana orMogán), and 0 elsewhere. This is the main variable in ourmodel: a significant coefficient would confirm a differentbehaviour explained by tourism.

3 All distances obtained are Euclidean ones. They have been calculated usingMatlab codes. Population was analysed assuming a uniform distribution withinmunicipalities. In fact, we used detailed micro data on population units smallerthan municipalities aggregating them with ArcGis software.

Table 3Distribution of sampled supermarkets by municipality and size.

Municipality Is it a touristic area? No. of sampled supermarkets By supermarket sizea

Size 1 Size 2 Size 3 Size 4

Agüimes No 2 (21) 2 (12) 0 (6) 0 (1) 0 (2)Arucas No 3 (22) 2 (14) 1 (4) 0 (3) 0 (1)Gáldar No 4 (27) 3 (16) 1 (7) 0 (3) 0 (1)Ingenio No 5 (40) 3 (28) 1 (6) 0 (2) 1 (4)Mogán Yes 15 (88) 9 (50) 4 (25) 2 (12) 0 (1)Las Palmas de Gran Canaria No 15 (203) 6 (98) 3 (54) 1 (25) 5 (26)San Bartolomé de Tirajana Yes 22 (141) 9 (63) 10 (63) 2 (11) 1 (4)Santa Brígida No 2 (12) 1 (7) 1 (4) 0 (1) 0 (0)Santa Lucía NO 5 (53) 2 (23) 1 (19) 1 (6) 1 (5)Telde No 6 (61) 3 (30) 2 (20) 0 (4) 1 (7)Total e 79 40 24 6 9

Note: Total supermarkets among brackets. Touristic municipalities in bold.a Supermarket size categories are the same as in Table 1. Source: Own elaboration based on the Official Business Census of the Regional Government.

J. Campos et al. / Tourism Management 36 (2013) 435e440438

� Supersizej. This variable controls the category size of super-market j, as described in Table 3. Indirectly, it captures scale andother size economies, that could yield to lower prices whensize increase, and to a higher number of brand varieties.

� BedsNumberc. This is a variable that takes into account thenumber of beds (not only in hotels, but also in apartments andrented houses) located in municipality c. Since many touristsstaying at hotels do not tend to buy at local supermarkets weintended to control by any potential distortion in foreigners’consumption patterns. In the estimations we also consideredthe ratio tourists/population by municipality and a touristicindex published by “La Caixa Studies”. Both variables showeda high correlation with “number of hotels” (0.88 and 0.97,respectively) confirming BedsNumber as a good proxy for thesize of the market.

We also included a binary variable to control the seasonaldifferences (Seasont), a binary variable to differentiate brandedfrom unbranded (white-label) products (Unbrandedi) and others toidentify fixed effects of supermarket chain (Chainsuperj) and typeof product (Producti).

Some descriptive statistics are presented in Table 4, dis-tinguishing between touristic and non-touristic areas. Supermar-kets in the first group show an average price of 2.44 euros, while innon-touristic is 2.45 euros (note the different number of observa-tions). Considering (as an example of X), a 250 m radius, touristicareas are more concentrated than the rest of the municipalities:

Table 4Descriptive statistics.

Variable Observations Average S.D. Minimum Maximum

Touristic areasPrice 1156 2.44 4.2 0.19 39.75Same in 250 m 2970 0.21 0.5 0 2Rivals in 250 m 2970 3.68 4.0 0 16Population in 250 m 2970 439.1 506.2 4 1910Supermarket

category size2970 1.63 0.77 1 4

Unbranded 2970 0.33 0.47 0 1Number of beds 2970 68,660 27,680.4 36,419 92,417Non-touristic areasPrice 1885 2.45 4.7 0.19 92Same in 250 m 3511 0.03 0.15 0 1Rivals in 250 m 3511 1.59 1.4 0 5Population in 250 m 3420 1335 1206.2 2 4800Supermarket

category size3511 1.9 1.23 1 4

Unbranded 3511 0.33 0.47 0 1Number of beds 3511 2709 3436.5 34 7298

S.D. is Standard Deviation. Source: Own elaboration.

each sampled supermarket has 0.21 supermarkets of the samechain within this radius, while a retailer located at a non-touristicarea has 0.03. The number of rivals follows a similar pattern (3.68versus 1.59).

On average, the surrounding population is equal to 439 peoplein a radius of 250 m around each store in touristic areas, while thisfigure is 1335 in non-touristic ones. With regard to supermarketsize, those located in touristic areas tend to be smaller. Finally, asnaturally expected, the number of total beds is larger in touristicmunicipalities.

4. Results

Tables 5 and 6 summarize the main results of our estimations.We have first considered an empirical model that explained theprice of each product i at supermarket j located at municipality c attime t as a function of being located at a touristic area whilesimultaneously controlling for other factors that could explain thedemand and the degree of competition in the market. In particular,the price equation

PRICEijct ¼ b0 þ b1SameXmetersjc þ b2RivXmetersjcþ b3PopulationXmetersjc þ b4Touristicjc

þ b5Supersizej þ b6InteractionT � S

þ b7BedsNumberc þ b8Periodt þ b9Unbrandedi

þX

bnChainsuperj þX

bnProducti þ εijct

has been estimated using alternative definitions of the X-radius(from 50 to 1500 m), since the literature considers that whenprecise data on the demand for each particular supermarket ismissing a safe way to approach it is by attraction circles (see forexample Abe & Kawaguchi, 2010, for the case of Japan; or Gómez-Lobo et al., 2011, precisely for the Gran Canaria market). Tocontrol for potential problems of spatial correlation, we haveincluded the clustering option in the estimations, which considerthe specific characteristics of each municipality.

The results in Table 5 clearly support the first hypothesis testedin this paper. For a wide subset of commodities in a typicalconsumption basket, supermarkets located in touristic areas chargeon average higher prices, as compared vis-à-vis with equivalentsupermarkets at non-touristic areas. Since competition factors havebeen controlled for, the explanation could lie in a standard incomeeffect on the consumers’ side. The parameters are positive andhighly significant for alternative definitions of the market size andtheir values remain very similar when the X-radius grows. The

Table 5Estimation results of the price models.

Variable Model 1(X ¼ 50 m)

Model 2(X ¼ 100 m)

Model 3(X ¼ 250 m)

Model 4(X ¼ 500 m)

Model 5(X ¼ 750 m)

Model 6(X ¼ 1000 m)

Model 7(X ¼ 1250 m)

Model 8(X ¼ 1500 m)

Same chain supermarkets in X meters �0.44 �0.20 0.05 0.06 0.007 �0.005 0.01 0.01Rival chain supermarkets in X meters 0.08** 0.05* �0.001 �0.01 �0.007 �0.008 �0.009 �0.009*

Population in X meters 0.001 4e-4 1e-4** 0.00003** 1e-5** 1e-5** 8e-6** 6e-6**

Touristic 0.45** 0.42** 0.49** 0.45** 0.43* 0.40* 0.38* 0.34**

Interaction touristic-supermarket size �0.1 �0.1 �0.1 �0.09 �0.09 �0.06 �0.06 �0.05Supermarket size �0.02 �0.009 0.019 0.01 0.001 �0.01 �0.02 �0.03Unbranded �0.84*** �0.84*** �0.84*** �0.84*** �0.84*** �0.84*** �0.84*** �0.84***

Number of beds �3e-7 4e-7 3e-7 1e-6 2e-6 3e-6 3e-6 5e-6Season 0.10 0.10 0.11 0.11 0.11 0.11 0.11 0.11Fixed effects by supermarket chain Yes Yes Yes Yes Yes Yes Yes YesFixed effects by product Yes Yes Yes Yes Yes Yes Yes Yes

Number of observations 2922 2922 2922 2922 2922 2922 2922 2922R2 0.58 0.58 0.58 0.58 0.58 0.58 0.58 0.58

Note: ***1%, **5%, *10% significance test. Touristic dummy in bold.

J. Campos et al. / Tourism Management 36 (2013) 435e440 439

other control variables seem less relevant, although the populationhas a small positive effect for X > 250 m and the presence ofunbranded products seems to increase the level of competitionamong retailers (by reducing prices, as expected).

Our second estimation, summarized in Table 6, was an empiricalmodel that explained the number of varieties of each product i atsupermarket j located at municipality c again as a function of beinglocated at a touristic areawhile simultaneously controlling by otherfactors. In this case, the brand equation

NVarietiesijc ¼ b0 þ b1SameXmetersjc þ b2RivXmetersjcþ b3PopulationXmetersjc þ b4Touristicjc

þ b5Supersizej þ b6InteractionT � S

þ b7Numberbedsc þX

bnChainsuperj

þX

bnProducti þ εijc

additionally includes an interaction variable (InteractionT-S) thatattempts to capture the specific effects of supermarket size attouristic areas. Different models using alternative definitions of theX-radius were estimated and the results also endorse our hypoth-esis: the estimated coefficients for the Touristic dummy variableare highly significant but negative. At this time the explanation liesin the size effect, as confirmed by the estimated coefficients of thesupermarket size and the interaction variables. This idea is sup-ported by Kim, Allenby, and Rossi (2002): ‘.given limited shelfspace, only a subset of the possible varieties can be displayed for

Table 6Estimation results of the brand variety models.

Variable Model 1(X ¼ 50 m)

Model 2(X ¼ 100 m)

Model 3(X ¼ 250 m)

Same chain supermarkets in X meters �0.46* �0.37* �0.19Rival chain supermarkets in X meters 0.03 0.03 0.04Population in X meters �0.004* �0.001* �0.0001**

Touristic L1.22** L1.22** L1.16**

Interaction touristic-supermarket size 0.11 0.09 0.11Supermarket size 0.34* 0.33* 0.34*

Number of beds 9e-6*** 9e-6*** 8e-6**

Fixed effects by supermarket chain Yes Yes YesFixed effects by product Yes Yes Yes

Number of observations 1577 1577 1577R2 0.58 0.58 0.58

Note: ***1%, **5%, *10% significance test. The SEASON effect was not included in this seconbetween January and April 2010. That explains the use of fewer observations. Touristic d

purchase at any one time. If consumers value variety, then a retailerwith lower variety must compensate the consumers in some way,such as a lower price level’. The relevance of different culturalpatterns in consumer preferences is also pointed out in Rozin,Fischler, Shields, and Masson (2006).

5. Discussion and conclusions

The thesis discussed in this paper is that in areas where tourisminflows outnumber or represent a large proportion as compared tolocal inhabitants the functioning of the markets for goods andservices may be affected by tourists’ consumption patterns. Inparticular we argue that prices and brand varieties found insupermarkets of touristic areas are significantly different fromthose found in their counterparts outside these areas.

Using a representative dataset from supermarkets in the islandof Gran Canaria, our estimations of different price and brand varietyequations seem to confirm the thesis that prices at touristicmunicipalities are higher (by 20.4% on average) and the number ofvarieties is smaller (by 36.9% on average) when compared to non-touristic municipalities in the same island, once other factors thatmight explain these differences are controlled for. The reason thatexplains the first result is a simple income effect, whereas thesecond one lies in the fact that touristic supermarkets do notbenefit from offering a wide range of brands to customers (tourists)who do not appreciate the differences due to the lack of locallanguage skills.

Model 4(X ¼ 500 m)

Model 5(X ¼ 750 m)

Model 6(X ¼ 1000 m)

Model 7(X ¼ 1250 m)

Model 8(X ¼ 1500 m)

�0.14 �0.05 �0.02 �0.01 �0.0050.02 0.01* 0.009 0.009 0.006�5e-5** �2e-4** �1e-5** �1e-5** �1e-5**

L1.07** L1.07** L1.04** L1.02** L0.98**

0.07 0.07 0.01 �0.002 �0.0020.37* 0.36* 0.39** 0.41** 0.43***

7e-6** 6e-6** 7e-6** 6e-6** 7e-6Yes Yes Yes Yes YesYes Yes Yes Yes Yes

1577 1577 1577 1577 15770.58 0.58 0.58 0.58 0.58

d equation because in most products the number of brand varieties did not changeummy in bold.

J. Campos et al. / Tourism Management 36 (2013) 435e440440

Arguably, these effects are not only negative. The upward shift inthe demand curve does not only increase prices but also quantities(and, indirectly, the level of economic activity, including realincome and employment) at the host community. This effect will behigher the more elastic the supply curve.

In any case, the point in discussing these two effects of tourismon local markets’ prices and brand diversity is not to questionwhether they exist or not, but to what extent are they relevant ascompared to other e more widely studied e impacts of tourism(both positive and negative). We think that only by attempting toprovide actual estimates, a more balanced cost-benefit of the mostrelevant consequences of tourism can be performed, and themicroeconomic policies in local retail markets e on competition,entry, opening hours, and so on e should take into account thedifferentiated characteristics of touristic areas.

Appendix A. Supplementary material

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.tourman.2012.09.016.

References

Aalto-Setälä, V. (2002). The effect of concentration and market power on foodprices: evidence from Finland. Journal of Retailing, 78, 207e216.

Abe, N., & Kawaguchi, D. (2010). Incumbent’s price response to new entry: the caseof Japanese supermarkets. Journal of the Japanese and International Economics,24, 196e212.

Blinkey, J. K., & Connor, J. (1998). Grocery market pricing and the new competitiveenvironment. Journal of Retailing, 74, 273e294.

Bujosa, A., & Rosselló, J. (2007). Modelling environmental attitudes towardstourism. Tourism Management, 28(3), 688e695.

Butler, R. W. (1980). The concept of a tourist area cycle of evolution: implications formanagement of resources. Canadian Geographer, 24(1), 5e12.

Concu, N., & Atzeni, G. (2012). Conflicting preferences among tourists and residents.Tourism Management, 33(6), 1293e1300.

Diedrich, A., & García-Buades, E. (2009). Local perceptions of tourism as indicatorsof destination decline. Tourism Management, 30, 512e521.

Eurostat. (2011). Europe in figures. Eurostat yearbook 2011. Brussels.García, G., & Sancho, A. (2000). A research of the impacts generated by tourism on

local population. In 5th International forum on tourism statistics. Glasgow.Gómez-Lobo, A., Jiménez, J. L., & Perdiguero, J. (2011). Entry under quality uncer-

tainty: Lessons from supermarkets. Cátedra Pasqual Maragall Working Paper 01/2011.

Griffith, R., & Harmgart, H. (2008). Supermarkets and planning regulation. CEPRDiscussion Paper No 6713.

Harcombe, D. (1999). The economic impacts of tourism. ABAC Journal, 19(2),Assumption University.

Horn, C., & Simmons, D. (2002). Community adaptation to tourism: comparisonsbetween Rotorua and Kaikorura, New Zealand. TourismManagement, 23, 133e143.

INE. (2011). Encuesta de Ocupación Hotelera 2010 [Spanish hotel occupation survey].Available at http://www.ine.es.

ISTAC. (2010). Anuario Estadístico de Canarias. Canary Islands: Government’s OfficialStatistics, Available at http://www.gobiernodecanarias.org/istac.

Kim, J., Allenby, G. M., & Rossi, P. (2002). Modelling consumer demand for variety.Marketing Science, 21(3), 229e250.

Krippendorf, J. (1982). Towards new tourism policies: the importance of environ-mental and sociocultural factors. Tourism Management, 3(3), 135e148.

Lawson, R. W., Williams, J., Young, T., & Cossens, J. (1998). A comparison of residents’attitudes towards tourism in 10 New Zealand destinations. Tourism Manage-ment, 19(3), 247e256.

Lindberg, K., & Johnson, R. (1997). Modelling resident attitudes toward tourism.Annals of Tourism Research, 24(2), 402e424.

Mason, P. (2008). Tourism impacts, planning and management, ISBN 978-0-7506-8492-7.

Mason, P., & Cheyne, J. (2000). Residents’ attitudes to proposed tourism develop-ment. Annals of Tourism Research, 27, 391e411.

Mihalic, T. (2000). Environmental management of a tourist destination: a factor oftourism competitiveness. Tourism Management, 21(1), 65e78.

Orams, M. B. (1995). Towards a more desirable form of ecotourism. TourismManagement, 16(1), 3e8.

Richards, T. J., & Hamilton, S. F. (2006). Rivalry in price and variety among super-market retailers. American Journal of Agricultural Economics, 88(3), 710e726.

Romeril, M. (1989). Tourism and the environment e accord or discord? TourismManagement, 10(3), 204e208.

Ross, G. F. (1992). Resident perceptions of the impact of tourism on an Australiancity. Journal of Travel Research, 30(3), 13e17.

Rozin, P., Fischler, C., Shields, C., & Masson, E. (2006). Attitudes towards largenumbers of choices in the food domain: a cross-cultural study of five countriesin Europe and the USA. Appetite, 46(3), 304e308.

Sancho, A., García, G., & Rozo, E. (2007). Comparativa de indicadores de sostenibi-lidad para destinos desarrollados y con poblaciones vulnerables. Annals ofTourism Research (in Spanish), 9(1), 150e167.

Sharpley, R., & Telfer, D. J. (2002). Tourism development: Concept and issuesIn Aspectsof tourism, . Channel View Publications. Butterworth Heinemann, ISBN 1-873150-35-0.

Thyne, M., Lawson, R., & Todd, S. (2006). The use of conjoint analysis to assess theimpact of the cross-cultural exchange between hosts and guests. TourismManagement, 27, 201e213.

Torres, E. (2003). El turismo residenciado y sus efectos en los destinos turísticos.Estudios Turísticos45e70, 155e156.

Winsor, R., Mak, L., & Hsu, M. (2010). A cross-cultural survey of shopping behaviorin U.S. supermarkets. In Proceedings of the annual meeting of the association ofcollegiate marketing educators. Advances in marketing. Dallas, TX. March 2e6,2010.

Javier Campos is an Associate Professor in Economics atthe University of Las Palmas de Gran Canaria, Spain. He isMSc in Economics by the London School of Economics andgot his PhD in Economics at CEMFI in Madrid. He hasworked as a consultant for The World Bank in infrastruc-ture and transport projects and also undertaken similarworks for the European Commission and the SpanishMinistry of Public Works. He has published in TransportPolicy, Review of Network Economics, Journal of Industry,Competition and Trade, among other journals.

Juan-Luis Jiménez holds a PhD in Economics and isAssistant Professor at the University de Las Palmas de GranCanaria, Spain. His main research fields are competitionpolicy and regulation. He has participated in projects forEuropean Commission or the Chilean Government, amongothers. He has published in the Review of Industrial Orga-nization, Renewable and Sustainable Energy Reviews or theJournal of Transport Geography, among others journals.

Ancor Suárez-Alemán has a degree in Economics anda MSc in Tourism Management. He is also a member of theEconomics of Infrastructure and Transport Research Groupof the University of Las Palmas de Gran Canaria, Spain. Hehas participated in several research projects and iscurrently finishing his PhD dissertation under Prof. Cam-pos’ supervision. His areas of specialization include mari-time transport, cost-benefit analysis, tourism economics,competition and regulation.