Multimarket contact, differentiation, and prices of chain hotels

Tourism Management 48 (2015) 305e315

Contents lists available at ScienceDirect

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

Multimarket contact, differentiation, and prices of chain hotels Rosario Silva* IE Business School, Alvarez de Baena 4, 28006 Madrid, Spain

h i g h l i g h t s
This study analyzes how two types of differentiation moderate the effect of multimarket contact on prices of chain hotels.
Hotels that belong to a chain can increase prices when they have more multimarket contact and are more differentiated.
Multimarket contact is a most effective mechanism to increase prices for hotels that do not follow a branding strategy.
Multimarket contact is a most effective mechanism to increase prices for hotels less differentiated in services.
Multimarket contact is a most effective mechanism to increase prices for 3-star chain hotels.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 22 October 2013 Accepted 18 November 2014 Available online

This study seeks to analyze the joint effect of multimarket contact and product differentiation on the intensity of price competition between chain hotels. The theory predicts that both variables reduce price rivalry, but little is known about their interaction effect. Multimarket literature establishes that when two firms meet one another in more than one market, the mutual recognition of their interdependencies will reduce the intensity of competition, leading to mutual forbearance. This article argues that differentiation moderates negatively the effect of multimarket contact on price because differentiation decreases the recognition of interdependence and thus, it makes difficult mutual forbearance. Using data from 1147 Spanish hotels that belong to a chain, this study shows that chain hotels with higher multimarket contact and higher differentiation (i.e. more stars, the Q certificate, common umbrella brand and a unique combination of services) charge higher prices. The evidence confirms the negative interaction but only for horizontal differentiation (i.e. different services and common brand). © 2014 Elsevier Ltd. All rights reserved.

Keywords: Multimarket contact Differentiation Prices Branding strategy Chain hotels Hotel industry

1. Introduction When two firms have multimarket contact, that is they meet one another in more than one market, the mutual recognition of their interdependencies will reduce the intensity of competition, leading to mutual forbearance (Bernheim & Whiston, 1990). The reason for such an effect is that if one firm is attacked in one market, it will have the ability to retaliate not only in the market in which the rival has initiated an aggressive action but also in all other markets in which it coincides with that rival. Therefore, as the markets of firms overlap, the rivalry intensity will diminish because of the risk of retaliatory actions in all of the markets in which they coincide (Gimeno & Woo, 1996). Given the relevance of multimarket contact to understand the firm behavior and performance,

* Tel.: þ34 91 568 9600. E-mail address: [email protected]. http://dx.doi.org/10.1016/j.tourman.2014.11.006 0261-5177/© 2014 Elsevier Ltd. All rights reserved.

there has been a proliferation of studies on multimarket competition research in the management literature in the last years (Yu & Cannella, 2013). As a result, there is strong evidence to support that multimarket contact affects the intensity of rivalry and firm performance in a variety of sectors, such as airlines (Gimeno & Woo, 1999), cement (Jans & Rosenbaum, 1997), personal computers (Kang, Bayus, & Balasubramanian, 2010), or hotels
ndez & Marín, 1998). (Ferna A stream of research within the multimarket literature has focused on the contingencies that moderate the effects of multimarket contact on mutual forbearance as for instance, firm size, internal coordination or strategic similarity respect to multimarket competitors (Yu & Cannella, 2013). However, the degree of product differentiation has received less attention. Economists distinguish between two types of differentiation: vertical and horizontal (Cremer & Thisse, 1991). Vertical differentiation makes the firm's product more attractive to all

306

R. Silva / Tourism Management 48 (2015) 305e315

customers of the market, whereas horizontal differentiation makes the firm's product more attractive to some customers and less attractive to others (Makadok, 2010). Yet it remains unclear whether product differentiation decreases the recognition of interdependence with multimarket rivals and thus it inhibits mutual forbearance. This study expands the literature on multimarket contact by analyzing how these two types of differentiation may influence the effectiveness of multimarket contact as a mechanism to restrict price rivalry: do vertical and horizontal differentiation strengthen (positive interaction) or weaken (negative interaction) the effect of multimarket contact on prices? Understanding the effects of the interaction between multimarket contact and differentiation on the intensity of rivalry experienced by hotel chains can be useful in predicting which particular competitive interactions will be more intense, thereby helping hotel chain managers to develop strategies to cope with situations of high rivalry. The setting of hotel chains is particularly relevant to test this research question. Hotel chains compete with each other in multiple geographical markets. Hotel chains may also vary in their degree of vertical and horizontal differentiation of each one of their hotels. For instance, a hotel chain may differentiate their hotels by their level of quality, by their brand or by their unique services. This paper proposes that differentiation weakens the effect of multimarket contact on prices because differentiation decreases the recognition of interdependence and consequently, it makes difficult mutual forbearance. This moderation effect is tested using handcoded data of 1147 hotels that belong to a chain. The database comprises detailed information about the characteristics of each one of the hotels, which allows including in the analysis several differentiation variables. The empirical evidence shows that multimarket contact and differentiation reduce price rivalry and that both mechanisms do not reinforce each other. Multimarket contact is an effective mechanism for reducing rivalry for chain hotels that are not horizontally differentiated (i.e. do not use a common brand and are similar in the level of services that they offer) as predicted. Additionally, the findings suggest that multimarket contact is more effective for three-star hotels than for hotels of higher or lower category. This paper contributes to existing literature in several ways. First, this study is one of the first studies that examine the moderator role of differentiation on the effect of multimarket contact on price rivalry. Second, the study suggests the dissimilar roles of the two types of differentiation in moderating the effect of multimarket contact on prices. Third, the negative effect on prices of multimarket contact for hotels that offer a unique combination of services and for lowest-category hotels point out that there can be some circumstances under which increasing the level of multimarket contact can be harmful. Fourth, the results can contribute to a better understanding on variables that determine the room prices of hotels that belong to a chain. The first section of this paper reviews the theoretical and empirical literature that analyzes the relationship among prices, multimarket contact, and the two types of differentiation. This review leads to the hypothesis section, which offers predictions regarding the interaction effects, followed by an empirical analysis. The final section discusses the main results and conclusions. 2. Literature review 2.1. The effect of multimarket contact on prices Multimarket contact between two firms is defined as the number of contacts in geographic markets (Gimeno & Woo, 1996). The multimarket literature establishes that as the degree of

multimarket contact between two firms increases, they are likely to become more interdependent, however this greater interdependence may not necessarily translate into higher intensity of competition (Jayachandran, Gimeno, & Varadarajan, 1999). On the contrary, the theory of multimarket competition suggests that the intensity of competition between firms with greater market overlap can be reduced due to “the mutual forbearance”. That is, firms that compete in many markets with other firms may be less motivated to compete vigorously against each other, because the gains in one market do not compensate for a general price war in all markets in which both firms coincide (Bernheim & Whinston, 1990). Empirical research supports the notion that higher multimarket contact is associated to mutual forbearance and can lead to a reduction in the intensity of competition (see Yu & Cannella, 2013 for a recent literature review). This intensity of competition has been conceptualized in different ways as prices, profits or market entry. Evidence shows that higher multimarket contact can lead to
ndez & Marín, 1998; higher prices (Evans & Kessides, 1994; Ferna Gimeno & Woo, 1999; Jans & Rosenbaum, 1997; Kang et al., 2010; € ller, 1997); higher profits (Hughes & Oughton, 1993; Parker & Ro Pilloff, 1999) and decreased rates of entry and exit (Baum & Korn,
mez, 2006). 1999; Fuentelsaz & Go The literature suggests that multimarket contact influences mutual forbearance because it enhances firms' interdependence trough increasing the familiarity with rivals' actions and increasing the ability of rival firms to deter each other (Jayachandran et al., 1999). In effect, multimarket contact facilitates mutual learning and increases the ability to prevent aggressive actions because a firm that is attacked in one market can counterattack in any other market in which both firms compete, creating serious financial loss for both sides. Research in multimarket competition has identified an array of firm and market factors, such as spheres of influence, strategic similarity, organizational structure, or market concentration that moderate the effect of multimarket contact on the intensity of rivalry (Jayachandran et al., 1999). These factors can increase or decrease the degree of familiarity and deterrence of firms that compete in multiple markets and consequently the likelihood of mutual forbearance occurring (Jayachandran et al., 1999). The literature on industrial organization recognizes that product homogeneity helps firms to reach collusive agreements, as they are thus able to acknowledge their interdependencies more easily (Chamberlin, 1929; Scherer & Ross, 1990; Stigler, 1987). A review of the empirical literature reveals that collusion occurs more frequently in homogeneous product markets than in differentiated product markets (Levenstein & Suslow, 2006). Just as product differentiation influences the likelihood of direct coordination or collusion, this article argues that product differentiation moderates the effect of multimarket contact and price, since mutual forbearance requires that firms recognize their interdependency.

2.2. The effect of vertical differentiation on prices Vertical differentiation occurs when consumers agree on the ranking of all of the variants of a given product line. Thus, in a situation in which prices are identical, consumers would all buy the same (presumably the best) variant of the product (Cremer & Thisse, 1991, p. 383). One method by which firms can achieve a vertical differentiation advantage is by offering superior quality relative to that offered by their rivals (Gabszewicz & Thisse, 1986; Shaked & Sutton, 1987). In this type of differentiation, consumers are assumed to desire higher quality, although they may differ in their willingness to pay for quality improvements. Because a

R. Silva / Tourism Management 48 (2015) 305e315

higher-quality product is more attractive, a firm that offers such a product can increase its prices and its economic value. The relationship between vertical differentiation and price is straightforward. Vertical differentiation creates value for consumers by offering attributes that they prefer, such as quality. Because consumers share the same ranking of preferences, the highest-quality product will be the most desirable and will thus command the highest price (Shaked & Sutton, 1982). Previous empirical studies have used hotel category and quality certifications as proxies for hotel quality. Evidence shows that the category (number of stars) is the attribute for which consumers are most willing to pay (Abrate, Capriello, & Fraquelli, 2011; Becerra,
, & Silva, 2013; Bull, 1994; Espinet, Saez, Coenders, & Santalo
, 2003; Rigall-I-Torrent & Fluvia , 2011). This variable, repreFluvia sented by the number of stars assigned by an official agency, may be considered a ranking of quality on which all consumers agree. Thus, a hotel is vertically differentiated when it offers a higher category than its rivals. Similarly, quality certifications can contribute to differentiate a hotel from its direct competitors since it is one of the strongest signals that a high-quality hotel can send to the market (Nicolau & Sellers, 2010).1 Evidence shows that quality certifications in hotels can increase customer satisfaction by improving the services offered (Nield & Kozac, 1999), the market value of hotel chains (Nicolau & Sellers, 2010), the performance of hotel chains (Alonso
n, & Rubio-Andrada, 2012), and the hotel Almeida, Rodríguez-Anto prices (Abrate et al., 2011). 2.3. The effect of horizontal differentiation on prices In the case of horizontal differentiation, consumers have different preferences regarding the attributes of a given firm or product; thus, even if products are offered at the same price, each will have a positive market share (Cremer & Thisse, 1991). Economic models of horizontal differentiation represent consumer preferences concerning factors such as locations within the product or geographic space (Hotelling, 1929). A firm is horizontally differentiated if it is located far from its competitors within either the geographic or product space, thus rendering the product more attractive to some consumers but less attractive to others. Therefore, horizontal differentiation renders products less substitutable, which limits the price competition between them. Examples of horizontal differentiation may be brand loyalty or distance from competitors within the geographic or product space (Makadok, 2010). Two firms can be horizontally differentiated when their customers are loyal to their brands. In the hotel industry, the use of a common umbrella brand fosters brand loyalty (O'Neill & Qu, 2006). As customers remain loyal to some brands, hotels become less likely to be direct substitutes for one another, thus decreasing the price rivalry between them. Firms can also be horizontally differentiated if they choose different positions in the geographic and product spaces in which consumers have preferences regarding the products that most closely match their ideals. For example, some customers will prefer a hotel located in the north of a city, whereas others will prefer a hotel located in the south. Likewise, a combination of services (i.e., services provided by a convention hotel) may be more attractive for certain customers (i.e., the business segment) and less attractive for others (i.e., the tourist segment). The hotel industry literature provides evidence suggesting that distinct segments of consumers value distinct attributes (Yavas & Babakus, 2003). Previous studies have identified the most

1

I thank an anonymous reviewer for this suggestion.

307

valuable attributes for the segment of business meetings (Renaghan & Kay, 1987), leisure (Chow, Garretson, & Kurtz, 1994; Lewis, 1985) and business (McCleary, Weaver, & Hutchinson, 1993; Taninecz, 1990). Hence, there is evidence that in the hotel industry horizontal differentiation in the service space may exist because clients have idiosyncratic preferences for some attributes. In consequence, hotels horizontally differentiated in services will charge higher prices, since they are offering a set of attributes unique and valuable for at least some group of customers.

3. Hypotheses 3.1. The moderation effect of vertical differentiation on the relationship between multimarket contact and prices The literature on multimarket contact has not directly addressed the relationship between vertical differentiation and multimarket contact. A vertically differentiated hotel is clearly better than other less differentiated rivals that compete in the same market. For instance, five-star hotels have resources that are superior to those of lower- and middle-ranked hotels. A five-star classification implies uniqueness in a hotel's architecture and exceptional interior design, large rooms, and state-of-the-art recreational facilities (Canina, Enz, & Harrison, 2005). The uniqueness of the resources is the basis upon which high-category hotels create superior value for their customers. Their differentiated positions allow them to raise prices above the competitive level. However, being in such positions may impede the ability of high-category hotels to perceive other hotels as competitors; for this reason, high-category hotels may have difficulty familiarizing themselves with the strategies of others, despite sharing the same markets. In addition, a high-category hotel may be less motivated to forbear from aggressive competition than a low- or mediumcategory hotel, as the superiority of resources within a highcategory hotel makes it unlikely for competitors to match their actions (Jayachandran et al., 1999). Moreover, it is unlikely that high-category hotels will respond aggressively to the actions of rivals, as aggressive price discounts can damage their reputation for quality. Consequently, regardless of multimarket contact, if one hotel offers the highest category, then aggressive retaliation by that hotel in other markets becomes less credible. By contrast, low- or medium-category hotels compete more intensely on prices (Becerra et al., 2013). This vulnerability could urge them to become more familiar with the strategies of other competitors. For example, a three star hotel will be more motivated to understand the strategy of other three star hotel located in the same city since, ceteris paribus, they compete with a similar offer and consequently, the actions of each competitor have a great impact on the performance of each other. However, a three star hotel can be also threatened by four-star hotels, than can lower their prices, or by one or two-star hotels that can improve their relationship price/quality. Therefore, it is more likely that low- or medium-category hotels view other hotels as significant competitors and for that reason, they try to become more familiar with their strategies. In addition, low- or medium-category hotels have the most to gain from mutual forbearance because such hotels are more susceptible to situations of high rivalry. That is, a three star hotel is likely that respond aggressively when other three star hotel initiates a price war since it can lose its entire market share if the competitor offers a lower price. Therefore, it will have the ability to deter the rival to initiate an aggressive action, since it hold a credible threat of retaliation in the markets in which both have presence. In summary, the ability of multimarket contact to reduce

308

R. Silva / Tourism Management 48 (2015) 305e315

the intensity of competition diminishes for low-and mediumcategory hotels. Hypothesis 1. The positive effect of multimarket contact on prices is weaker for high- category chain hotels than for low- and mediumcategory hotels. The same arguments can be applied to the situation in which vertical differentiation is achieved through a quality certification. Firms that own a quality certificate have improved their internal capabilities to provide a better customer service, which gives them the possibility to differentiate themselves from competitors (Nield & Kozac, 1999). This superior customer value allows firms to charge higher prices respect to direct competitors and earn higher profits (Abrate et al., 2011; Alonso-Almeida et al., 2012). Since the implementation of a certified quality system can be costly (Rubio
n, 2011), competitors Andrada, Alonso-Almeida, & Rodríguez-Anto cannot easily imitate the value created by a firm that owns a certification. This implies that even in a situation of price war between direct competitors, the quality certification would allow the firm to sustain higher prices. In consequence, as firms that own a quality certification are less affected by competitors' actions, they are less likely to become familiar with the strategies of competitors. Additionally, it is unlikely that hotels that are signaling their superior quality will respond aggressively to the actions of rivals, as aggressive price discounts can damage their reputation for quality. Consequently, regardless of multimarket contact, if one hotel offers the best quality, then aggressive retaliation becomes less credible. Hypothesis 2. The positive effect of multimarket contact on prices is weaker for chain hotels that own a quality certificate.

3.2. The moderation effect of horizontal differentiation on the relationship between multimarket contact and prices A stream of research within the multimarket literature has studied the effects of multimarket contact and strategic similarity (Chen, 1996; Evans & Kessides, 1994; Gimeno & Woo, 1996) and their interaction effect, both theoretically (Jayachandran et al.,
mez, 2006; Young, Smith, 1999) and empirically (Fuentelsaz & Go Grimm, & Simon, 2000). However, there is no consensus regarding the sign of the moderation effect. Jayanchandran et al. (1999) proposed that strategic distance weakens the effect of multimarket contact on mutual forbearance, whereas Fuentelsaz
mez (2006) and Young et al. (2000) argued that strategic and Go distance reinforces that effect. The first perspective proposes that resource similarity and multimarket contact are complementary mechanisms that reinforce one another. Firms with similar resources become mutually familiar and are more able to deter one another because they can more easily imitate the actions of their rivals. Therefore, the ability of multimarket contact to increase prices should be greater when firms have similar resources (Jayachandran et al., 1999). By contrast, the second perspective suggests that resource similarity and multimarket contact are substitute mechanisms by which firms obtain the information that is required to cooperate
mez, 2006; Young et al., 2000). Because either (Fuentelsaz & Go mechanism is sufficient to sustain cooperation, multimarket contact is less effective in avoiding rivalry when the strategic similarity between firms is greater. Empirical research supports this last
mez, 2006; Young et al., 2000). For perspective (Fuentelsaz & Go
mez (2006) showed that the negative instance, Fuentelsaz and Go effect of multimarket contact on entry rates was greater when the multimarket rivals were more dissimilar. In the same way, Young et al., (2000) found that the negative effect of multimarket

contact on the action frequency of a firm was stronger when the rivals' resources were more dissimilar. The moderation effect of similarity on the relationship between multimarket contact and intensity of rivalry has been empirically validated when rivalry is measured as entry rate (Fuentelsaz &
mez, 2006) and action frequency (Young et al., 2000). HowevGo er, it has not yet been tested when price is the variable that captures the level of intensity of rivalry. Jayachandran et al., (1999) argue that firms that compete with the same strategies may become more familiar with each other and may be able to deter each other compared with firms that follow different strategies. As indicated above, horizontally differentiated hotels may be attractive to some customers but not to others. For example, some travelers may have an intense loyalty to a particular brand, whereas others would never choose that brand, even at low prices. Likewise, business customers may prefer the combination of services offered by a hotel that specializes in business meetings rather than the services offered by a hotel that caters to the sunand-beach segment. Hence, horizontally differentiated hotels compete with different strategies because they seek to attract different segments of the customer pool with unique value propositions. Therefore, it is unlikely that horizontally differentiated hotels develop familiarity with the strategies of other hotels since they do not follow a similar strategy. In addition, more differentiated hotels may be in worse positions to deter one another, as their gains from mutual forbearance are not as high because their prices are less affected by an environment of high rivalry. In effect, a horizontally differentiated firm will not lose a large amount of market share when it is involved in a situation of aggressive competition, as it has a group of customers who consistently prefer its brand or its combination of services. Therefore, if a multimarket competitor initiates a price war it is unlikely that a horizontally differentiated firm may be motivated to retaliate aggressively. Because horizontal differentiation decreases the level of familiarity with respect to multimarket competitors as well as the threat of retaliation, it will negatively moderate the relationship between multimarket contact and prices. Hypothesis 3. The positive effect of multimarket contact on prices will be weaker for chain hotels that follow a branding strategy. Hypothesis 4. The positive effect of multimarket contact on prices will be weaker for chain hotels that are more differentiated in services.

4. Methodology 4.1. Data To build a database of hotels belonging to a chain in 2005, I collected data from the Official Hotel Guide, published by the Instituto de Estudios Turísticos. Among all of the hotels in Spain, I selected those found in the 135 locations for which the INE (National Institute of Statistics) gathers visitor information. This group accounts for 70% of visitors and 56% of existing hotels. From this sample of 3456 hotels, I selected those that belonged to chains. Thus, the initial sample consisted of 1269 hotels, a group that accounts for 70% of Spanish chain hotels. After discarding hotels with missing data (primarily because of a lack of price information), the final sample comprised 1147 hotels located in 110 tourist locations. From that sample of hotels for which price information was available, I collected data on double room prices for the peak and off-peak seasons. Thus, the final sample included 1866 prices.

R. Silva / Tourism Management 48 (2015) 305e315

The Official Hotel Guide contains information collected annually on every hotel in Spain, including each hotel's name, category (one to five stars), age, chain membership, location, prices for double and single rooms in the peak and off-peak seasons, the number of rooms, and approximately 40 other features (e.g., golf and tennis facilities, meeting rooms).

between them. The average distance from a given hotel to other chain hotels in the same city was computed using the measure of strategic distance proposed by Gimeno and Woo (1996) and
rrez (2006): applied to the hotel industry by Urtasun and Gutie

The dependent variable in the analysis was the price of a double room in the peak and off-peak seasons, as reported by the Guide. To check the validity of the pricing data in the Guide, I collected hotel prices advertised by two tour operators in a random sample of 70 hotels. I found a high correlation of .72 for peak season prices and a correlation of .68 for off-peak season prices, although the average prices of the Guide were 11% higher than the average prices of the tour operator. Since both set of prices are highly correlated and the differences among room prices across hotels are consistent in both sets, I can conclude that the pricing data from the Guide are reliable. The independent variables of the analysis were the following: 4.2.1. Vertical differentiation The number of stars (one to five stars) of each hotel as reported by the Official Guide. The number of stars is an official classification based on some measurable characteristics (i.e., hotel facilities, general services, etc.), therefore it is an objective indicator of the ~ ez-Serrano, Turrio
n, & Vela
zquez, 2014) which hotel's quality (Nún cannot coincide with the level of service quality perceived by each
pez & Serrano, 2004). customer (Lo Previous empirical studies have used the number of stars as a proxy for hotel quality based on tangible elements (Becerra et al.,
ndez & Marín, 1998; Nún ~ ez-Serrano et al., 2014). The 2013; Ferna hotel category, represented by the number of stars assigned by an official agency, may be considered a ranking of quality on which all consumers agree. Because hotels with more stars offer more facilities and larger rooms, every consumer would likely agree that a five-star hotel is better than a four-star, and so on. Therefore, a hotel is vertically differentiated when it offers a higher category score than its rivals. The Q certification as reported by the Official Hotel Guide. This dummy variable indicates if the hotel has the Q certification. Tourism market has many quality certifications, but the Q of the Spanish Tourism Quality Institute (Q certification) has been considered as the most recognized certification in the Spanish
n, & Alonso, 2010; Nicolau & tourism industry (Casadesús, Marimo Sellers, 2010). The Q certification establishes quality standards depending on the category, type of service and type of establishment, about the facilities, the final services delivered to customers, organizational processes of client services and tools for customer
n, & Alonso, 2010). Since the Q satisfaction (Casadesús, Marimo certification provides a signal for consumers about the high quality of the service offered by the hotel before buying it, a hotel will be vertically differentiated if it owns this certified quality system. Horizontal differentiation in the service space. The Guide provides information on 41 different services that a hotel may provide. This information was manually entered for each of the 1269 hotels belonging to chains to calculate the distance in services

Average Multimarket Contacti;l ¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2ffi PM   s s im jm j¼1 m¼1

Pn ServiceDisti ¼

4.2. Variables

309

n

where j is one of n hotels that belong to a chain located in the same city as hotel i and sim indicates whether hotel i offers (sim ¼ 1) or does not offer (sim ¼ 0) the service M, which represents the total number of different services offered by the established hotels. This measure captures the degree of horizontal differentiation, as hotels with distinct combinations of services attract different groups of customers. For instance, being close to a beach and offering sports facilities may be attractive to the “sun and beach” segment, whereas the availability of a conference center with smaller meeting rooms and facilities for simultaneous translation would be attractive to the business segment. Brand. This dummy variable indicates whether a hotel's brand
, NH, or Barcelo
, includes the parent company name, such as Melia or a common brand that it shares with other hotels of the same chain, such as Novotel or Tryp. The literature on brands within the hospitality industry identifies the diverse approaches taken by chains to manage the brand of their hotels (O'Neill & Mattila, 2010). Some chains attempt to identify and differentiate themselves in the minds of customers by following a branding strategy. For instance, some chains include their corporate name in all of their hotels, as do NH and AC Hotels, or in a group of hotels, as with Husa Hotels. Other chains follow a multibrand strategy to compete in different segments. For instance, Accor Hotels has developed the Ibis, Sofitel, and Novotel brands. By contrast, some chains do not follow a branding strategy; rather, they use individual brand names for each hotel, such as Sercotel Hotels. When chains follow a branding strategy, they normally appeal to different groups of customers. Since in experience goods there is an uncertainty with trying new products, once customers adopt a particular brand as its favorite they may have little incentive to switch to other brands (Ethiraj & Zhu, 2008). As customers remain loyal to particular brands, these brands are imperfect substitutes and in consequence, they are differentiated. Similar to service distance, a brand is a measure of horizontal differentiation because the value of each brand depends on individual preferences. Average multimarket contact. Following Gimeno and Woo (1996, 1999), I computed the average multimarket contact as the average number of multimarket contacts with a firm's competitors in market l. A multimarket contact occurs when hotel i (which competes in market l) and hotel j (which also competes in market l) meet in another market, p. Thus, the multimarket contact of firm i with competitor j is computed as the sum of multimarket contacts across all markets outside of market l:

Multimarket Contactij;l ¼

X

Multimarket contactij;lp

psl

The overall measure of multimarket contact is the average number of multimarket contacts with the competitors in market l, which is calculated as follows:

X 1 Multimarket Contactij;l Competitorsl isj

.10** .09** .15** .17** .14** .13** .03 .25** .07** .13** .04 .08** .05* .12** .01 .15** .43** .00 .05* .26** .10** .09** .09** .16** .02 .05* .12** .09** .35** .05* .27** .43** .09** .14** .12** .02 .13** .21** .03 .10** .18** .00 .00 .16** .17** .01 .20** .01 .03 .09** .05* .05* .05* .17** .16** .08** .09** .16** .08** .12** .14** .02 .00 .03 .18** .16** .12** .13** .31** .25** .23** .16** .45** .17** .08** .17** .05* .19** .05* .04 .08** .09** .07** .10** .05* .11** .15** .04 .12** .07** .11** .41** .19** .08** .14** .06** .24** .02 .25** .18** .24** .14** .08** .03 .05* .03 .15** .05* .03 .02 .06* .03 .06* .05* .09** .04 .02 .03 .01 .05* .12** .13** .23** .00 .01 .01 .31** .18** .35** .03 .13** .60** .07 ** .18** .08** .50 .49 .74 .85 .50 .46 .46 .17 .42 .32 .40 .31 3.37 .75 .30 .50 .33 4.72 .61 4.86 1.99 .47 .70 .31 .15 .23 .11 .20 .11 2.80 3.44 .10 .55 3.14 y p < .10; * p < .05 ** p <. 01.

2 I am grateful to an anonymous reviewer for suggesting me to double-check these results using SEM. The SEM results are similar to OLS results.

Table 1 Descriptives and correlations.

Ordinary least squares (OLS) regression analysis was performed to test the interaction effects of vertical and horizontal differentiation and multimarket contact on listed room prices.2 The first analysis included control variables and dummy variables for the 110 locations in the sample and the independent variables that measure hotel differentiation and multimarket contact. The second analysis included the hypothesized interactions. To avoid multicollinearity

N

4.3. Empirical analysis

1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866 1866

Mean

S.D.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

I used a large number of control variables in the study to isolate the possible effects of differences across hotels and locations on their prices. Local competition. The HerfindahleHirschman concentration index measures the level of local competition. This variable is computed using the proportion of hotel rooms to the total number of rooms of the same category in the same city. This variable includes both independent and chain hotels that compete in the same category and city. Fixed city effects. To control for any unobservable differences across the 110 cities in the sample, fixed city effects were included. I also controlled for certain characteristics of the specific site of the hotel within the city in which it is located to account for more subtle location differences among competing hotels. Those hotels located in specific areas within a city (such as downtown or closer to a beach, mountain, or renowned local tourist spot) may take advantage of their unique location, which would affect their prices. The following control variables capture these differences: Beach is a dummy variable that takes the value of one if the Guide indicates that the hotel is near a beach within the city. This variable is self-reported by hotels and exhibits intra-location variation because within a given city, some hotels are “beach hotels” while others, located farther inland, are not. Central location is a dummy variable that is coded as one when the Guide indicates that the hotel is centrally located in the city. Special is a dummy variable that takes the value of one when the Guide indicates that the hotel is located near a picturesque spot, such as Madrid's Museo del Prado or Barcelona's Sagrada Familia Cathedral. Finally, I included a number of control variables in the analysis to account for key attributes of the hotels within the context of the city, as such attributes are frequently used in the literature (Espinet, , 2003; Rigall-I-Torrent & Fluvia , 2011). Saez, Coenders, & Fluvia Some of these characteristics may be valuable to certain customers and may consequently increase hotel prices. Unless otherwise indicated, I obtained these data from the Official Hotel Guide. Hotel size reflects the total number of rooms in each hotel. As firm size is typically a significant variable in strategy research, hotel size may be expected to affect pricing policy (e.g., as a result of economies of scale). The log value of this variable is used because it is a skewed variable. The age of each hotel is also controlled to account for the possibly greater pressure on older hotels to lower their prices. The log value of this variable is also used because it is a skewed variable. Finally, different dummies were used to capture differences across hotels that may determine the prices that they charge. These control variables include Historic Building (for hotels located within a historic site), such as a palace or castle, Nursery (for hotels that provide nursery services), and Room service (for hotels with room service). The Convention variable indicates whether a hotel has an auditorium. Information regarding this last variable for 2005 was downloaded from the website of the Spanish Convention Bureau.

.00

R. Silva / Tourism Management 48 (2015) 305e315

1.Log-price 2.Peak Season 3.Number of rooms (log) 4.Age (log) 5.Beach 6.Centric 7.Special 8.Herfindahl index 9.Convention 10.Historic building 11.Room service 12.Nursery 13.Multimarket contact 14.Category 15.Quality certificate 16.Brand 17.Distance in services

310

R. Silva / Tourism Management 48 (2015) 305e315

among the interaction coefficients, the four independent variables were standardized before computing the interactions. 5. Results 5.1. Descriptive statistics Table 1 presents the descriptive statistics of the variables, including the means and standard deviations of the original variables and the correlation matrix. The average price of a double room is 112.17 Euros. The zero-order correlations show that higher prices are typically charged during the peak season and by larger and newer hotels, although not when they are located close to a beach. These control variables are significant and show the expected direction. Additionally, the correlation between price and the differentiation variables is positive and significant. As expected, a higher category rating, the Q certification, the use of a common umbrella brand, and higher service distance are all related to higher prices. Finally, the correlation between price and multimarket contact is also positive and significant. Table 2 shows the distribution of hotels by category, quality certificate, brand, and multimarket contact. The results show that 16% of the hotels that belong to a chain do not have any multimarket contact and that branding strategies are not widely used by Spanish chains. The statistics show that 44% of hotels belonging to a chain have an individual brand name, whereas the remaining 56% of these hotels follow a branding strategy. 5.2. Regression analysis Table 3 shows the results for the regression analysis of room price using dummies for the 110 locations. Model 1 includes the control variables, the dummies for the 110 locations, and the differentiation variables that explain 72% of the variance in price. As expected, the differentiation variables are strongly related to price. Hotels can raise their prices by 29% for each additional star and by 4% if they have the Q certificate. Additionally, a hotel that follows a branding strategy can raise its prices by 7% and an average difference of one service respect to the chain hotels that compete in the same city allows the hotel to charge a price premium of 6%. Regarding the effect of other variables, it should be noted that the number of multimarket contacts has a significant effect on price. Each new contact allows a 1% increase in price. These results show that -independent of the number of stars-hotels with a higher level of multimarket contact, with the Q certificate, with a common brand, or with a higher distance in services tend to charge higher prices. In addition, as expected, the hotel's location significantly

Table 2 Distribution of hotels by category, quality certificate, brand and level of multimarket contact. Number of Percentage hotels over total Category

1 Star 17 2 Stars 76 3 Stars 476 4 Stars 519 5 Stars 59 Q certificate 0 1,040 1 107 Multimarket contact 0 186 Greater than 0 961 Brand 0 505 1 642 Number of hotels in the final sample 1147

1.48% 6.63% 41.50% 45.25% 5.14% 90.67% 9.33% 16.22% 83.78% 44.03% 55.97%

311

influences room prices. Table 3 shows that hotels that are located in special places and near a beach charge higher prices. In Model 2, the added interaction between multimarket contact and category is not significant. Thus, this model does not support Hypothesis 1. Because the effect of the category of price can be nonlinear, I introduce a variable for each category in Model 3. This model replicates the same analysis as summarized in Model 1, but rather than the category variable, the model includes a variable for each star. The results show that the effect of the category is nonlinear. A four-star hotel can charge a price that is 27% higher than that charged by a three-star hotel, whereas a five-star hotel can charge a price that is 80% higher than that charged by a three-star hotel. Model 4 shows that the interaction of the Q certificate with multimarket contact is not significant and in consequence, it does not support Hypothesis 2. The interaction of brand with multimarket contact is included in Model 5. This model is significant and provides support for Hypothesis 3. An analogous effect is found for the interaction of multimarket contact with distance in services, shown in Model 6, which supports Hypothesis 4. The negative interaction of brand and distance in services with multimarket contact indicates that the effect of multimarket contact on price becomes less intense as hotels are more horizontally differentiated. Figs. 1 and 2 illustrate that the effect of multimarket contact is much stronger when a hotel does not follow a branding strategy or when a hotel offers services that are similar to those offered by its competitors. Fig. 2 indicates that multimarket contact has a different effect on prices depending on the hotel's distance in services. This figure shows a negative relationship between multimarket contact and prices for hotels that compete in markets where they offer a unique combination of services. In contrast, a positive relationship exists for hotels that compete in markets where rivals offer a similar combination of services. In the analysis presented in Model 1, the Category variable was forced to have a linear effect on price. However, a competitive strategy for low-quality hotels based on price is qualitatively different from competition among higher-category hotels, which is based primarily on service. To account for these differences, which are possibly nonlinear, I also grouped one- and two-star hotels into one category described by one dummy variable (Categories 1e2) and grouped four- and five-star hotels into another category described by another dummy variable (Categories 4e5), thus leaving three-star hotels as the reference group for comparison. The results for these two category dummies and their interaction with multimarket contact are shown in Model 7. The results for the hotel category dummies do not support Hypothesis 1, although they do provide new insight suggesting that their interaction with multimarket contact deserves to be discussed in greater detail. The negative interaction of the two category dummies with multimarket contact shown in Model 8 indicates that the effect of multimarket contact is more positive for middle-category hotels than for low-category (one- and two-star) and high-category (fourand five-star) hotels. With respect to room prices, multimarket contact affects the prices of three-star hotels more than the prices of four- and five-star hotels (b ¼ .01, p-value ¼ .01). This result is consistent with Hypothesis 1, which suggests that high-category hotels will be unable to take advantage of their level of multimarket contact because of their difficulty in becoming familiar with the strategies of other hotels and in forbearing from aggressive competition. However, contrary to expectations, an increase in the level of multimarket contact has a negative effect on prices for oneand two-star hotels (b ¼ .04, p-value < .01). This result indicates that for low-rated hotels, an increase in multimarket competition does not lead to mutual forbearance but rather leads to an increase in rivalry and consequently to lower prices. Fig. 3 illustrates the moderation effect of these category dummies on the relationship

312

Table 3 Regression Analysis Room prices. Model 1

Multimarket contact Category Quality Certificate Brand Distance in services Star 1 Star 2 Star 4 Star 5 Categories 1&2 Categories 4&5 Multimarket  Multimarket  ertificate Multimarket  Multimarket  in services Multimarket  Multimarket  Adj. R-squared

3.38 .47 .05 .01 .04 .01 .03 .09 .04 .07 .08 .01 .01 .29 .04 .07 .06

Category Quality C

** ** **

y

** ** * ** y ** *

(.11) (.01) (.01) (.01) (.03) (.02) (.02) (.06) (.02) (.02) (.02) (.02)

3.36 .48 .05 .01 .04 .01 .03 .10 .04 .07 .08 .01

(.00) (.01) (.02) (.01) (.03)

.02 .30 .04 .07 .05

.00

Model 3 ** ** **

y y y ** ** * ** ** *

(.11) (.01) (.01) (.01) (.03) (.02) (.02) (.06) (.02) (.02) (.02) (.02) (.01) (.01) (.02) (.01) (.03)

Model 4

4.32 .47 .04 .01 .03 .01 .03

** ** ** y

.04 .05 .05 .02

Model 5

(.10) (.01) (.01) (.01) (.02) (.02) (.02)

4.32 .47 .04 .01 .03 .01 .03

** ** ** y

y * *

(.02) (.02) (.02) (.02)

.04 .05 .05 .01

.01

**

(.00)

.04 .08 .02 .42 .28 .26 .79

y **

(.02) (.01) (.03) (.05) (.03) (.02) (.04)

y

** ** ** **

Model 6

(.10) (.01) (.01) (.01) (.02) (.02) (.02)

4.32 .47 .04 .01 .03 .01 .03

** ** ** y

y * *

(.02) (.02) (.02) (.02)

.04 .05 .05 .02

.01

**

(.00)

.04 .08 .03 .42 .28 .26 .79

y ** * ** ** ** **

(.02) (.01) (.03) (.05) (.03) (.02) (.04)

y y

Model 7

(.10) (.01) (.01) (.01) (.02) (.02) (.02)

4.18 .47 .04 .01 .04 .00 .03

y * *

(.02) (.02) (.02) (.02)

.04 .06 .05 .01

** *

(.02) (.02) (.02) (.02)

4.21 .48 .04 .02 .03 .01 .04 .36 .06 .09 .13 .00

.01

**

(.00)

.07

**

(.02)

.01

.04 .10 .02 .42 .28 .26 .80

y **

(.02) (.02) (.03) (.05) (.03) (.02) (.04)

.04 .08 .07 .43 .28 .26 .79

y ** * ** ** ** **

(.02) (.01) (.03) (.05) (.03) (.02) (.04)

y

** ** ** **

** ** ** y

(.12) (.01) (.01) (.01) (.02) (.02) (.02)

(.11) (.01) (.01) (.01) (.03) (.02) (.02) (.06) (.02) (.02) (.02) (.02)

4.19 .48 .04 .02 .03 .01 .04 .36 .07 .09 .13 .01

** ** ** *

*

(.00)

.01

**

(.00)

.05 .06 .05

* ** y

(.02) (.02) (.03)

.05 .07 .05

* ** y

(.02) (.02) (.03)

.35 .26

** **

(.03) (.02)

.28 .29

** **

(.03) (.02)

.04 .01 .6996

** **

(.01) (.00)

* ** * ** **

* ** ** ** **

(.11) (.01) (.01) (.01) (.03) (.02) (.02) (.06) (.02) (.02) (.02) (.02)

(.03) .01

Brand Distance

*

(.00) .02

**

(.01)

Cat1&2 Cat 4&5 .7210

** ** ** *

(.00) .02

.7209

Model 8

.7296

.7295

N ¼ 1866; yp < .10; *p < .05**p < .01; City Dummies included in all models, Standard Errors in Parentheses.

.7300

.7306

.6971

R. Silva / Tourism Management 48 (2015) 305e315

Intercept Peak Season Number of rooms (log) Age (log) Beach Centric Special Herfindahl Index Convention Historic Building Room service Nursery

Model 2

R. Silva / Tourism Management 48 (2015) 305e315

313

The goal of this study was to analyze the joint effect of multimarket contact and differentiation on the intensity of price competition between chain hotels. Empirically, the study presents evidence of the existence of two mechanisms that allow chain hotels to reduce rivalry (or increase prices) -multimarket contact and differentiation-as well as evidence of their interaction effect. According to the hypotheses, the results show that high multimarket contact leads to higher prices when chain hotels offer similar services respect to multimarket rivals that compete in the same city or when chain hotels do not follow a branding strategy (i.e. they do not use a common umbrella brand). On the contrary, the results do not support the hypothesis of the negative moderation effect of vertical differentiation. The evidence shows that multimarket contact is more effective for three-star hotels than for

hotels of higher or lower category. Finally, the results suggest that high multimarket contact leads to lower prices for one- and twostar hotels and for hotels differentiated in services. The first contribution of this study to the literature of multimarket contact is the finding that firm differentiation moderates the relationship between multimarket contact and prices. The negative interaction between multimarket contact and horizontal differentiation confirms that the lack of differentiation enhances mutual forbearance. This result is consistent with the theoretical prediction made by Jayachandran et al., (1999) regarding the interaction effect of multimarket contact and strategic distance, which had yet to be empirically validated in the context of price competition. The second contribution of this study is about the moderator role of vertical differentiation. The empirical results of this study suggest that the effectiveness of multimarket contact as a mechanism to reduce rivalry also depends on a hotel's category. The results show that three-star hotels benefit more from their high level of multimarket contact than do higher-category hotels. In contrast, multimarket contact is harmful for lower-category hotels. The findings show that an increase in multimarket contact has a negative effect on the prices of the lowest-category hotels, which implies that for one- and two-star hotels, multimarket contact does not lead to mutual forbearance but rather to an increase in price rivalry. As shown in the literature review, multimarket contact leads to mutual forbearance when there is familiarity and when firms have the ability to deter one another. In this sense, hotels with the lowest and highest category ratings do not appear to have the same ability to develop familiarity or to retaliate against rivals, as do mediumcategory hotels. A possible explanation for this result is that three-star hotels are more likely to develop familiarity with other hotels, as they may view both lower- and higher-category hotels that meet them in multiple markets as competitors capable of capturing part of their market share. High-category hotels may benefit from their differentiated position to attract customers from the middle segment, whereas low-category hotels may upgrade their services to compete more directly with middle-category hotels (Baum & Mezías, 1992). Because actions by all competitors can threaten the competitive position of middle-category hotels, these types of firms are more likely to become familiar with the strategies of hotels of different qualities and to recognize their interdependencies. Additionally, three-star hotels are more likely to compete aggressively if a competitor initiates a hostile action, as these hotels

Fig. 2. Moderation effect of Distance in Services on the relationship between multimarket contact and double room prices (log).

Fig. 3. Moderation effect of Hotel Category on the relationship between multimarket contact and double room prices (log).

Fig. 1. Moderation effect of Brand on the relationship between multimarket contact and double room prices (log).

between multimarket contact and prices. This figure shows a negative relationship between multimarket contact and prices for one- and two-star hotels. In contrast, a slight positive relationship exists for four- and five-star hotels. Finally, the positive effect is substantially stronger for three-star hotels than for four- and fivestar hotels. 6. Discussion

314

R. Silva / Tourism Management 48 (2015) 305e315

have more room to provide discounts than do lower-quality hotels. By contrast, low-category hotels may not pose a credible threat of retaliation in other markets. Hotels with one or two stars charge low rates throughout the year and are therefore not motivated to initiate aggressive actions because their low margins do not allow them to provide additional discounts under the list price. Consequently, when these hotels are attacked in one market, they may not be motivated to respond in other markets. With respect to highcategory hotels, their threat of retaliation is unlikely because aggressive price discounts can damage their reputation for quality. Thus, the ability to deter a rival's action can be stronger for threestar hotels than for high- and low-category hotels. The third contribution is to the research on the negative consequences of multimarket contact. The results show that, for hotels that are more differentiated in services and for low-category hotels, multimarket contact leads to more intense competition. The literature provides some possible explanations as to why multimarket contact can have negative consequences for a firm. For example, Thomas and Willig (2006) showed that asymmetric information regarding a rival's actions can lead to more intense rivalry. When firms do not accurately perceive the actions of rivals, there is a possibility of misguided retaliatory reactions. However, it appears improbable that such asymmetry of information would exist among these types of hotels because pricing information is rather transparent through hotel websites or online travel agencies. Other researchers suggest that the effect of multimarket contact may be different for younger and smaller firms than for larger and more mature firms (Yu & Cannella, 2013). Younger and smaller firms may experience difficulty in understanding how competitors can be influenced and their ability to influence them. For this reason, Yu and Cannella (2013, p. 101) claimed that “the theory of mutual forbearance may be inappropriate for young and small organizations.” The results obtained in this study contribute to this research stream by suggesting that the theory of mutual forbearance does not hold for low-category hotels and for hotels that offer a unique combination of services. However, more research is needed to examine the reasons that could explain why multimarket contact is harmful for these types of hotels. Finally, this study also contribute to the large literature on pricing in the hospitality industry. Although there is extensive research that has explored a large number of variables that determine hotel room prices (Espinet et al., 2003; Rigall-I-Torrent & , 2011), to the best of my knowledge, this is the first study that Fluvia specifically explore the variables that influence room prices of hotels that belong to a chain. 7. Managerial implications The results have relevant practical implications. First, they highlight the fact that multimarket contact and differentiation can have a positive effect on hotel's prices. Therefore, both mechanisms contribute to reduce price rivalry between chain hotels. Second, the negative interaction effect between multimarket contact and hotel's differentiation allows predicting which interactions will be more intense. These predictions can help chain managers to develop strategies to deal with high rivalry situations. The results show that the most intense rivalry is between no differentiated chain hotels with little multimarket contact (i.e. chain hotels that compete with similar services or without a common brand in markets with independent hotels or with local chains). Therefore, in these highly competitive markets managers should invest in differentiation (i.e. increasing quality, using a common brand or offering a unique combination of services). On the contrary, the results show that the least intense rivalry is between chain hotels similar in services and that do not use a

common brand with high multimarket contact. This implies that in the low-rivalry markets-where the hotel has high multimarket contact-managers should not invest in being differentiated respect to multimarket rivals. Finally, this study can have practical implications for chains' geographic expansion. Chains should target the same rivals' markets when the level of horizontal differentiation is low. Hotels with a low level of horizontal differentiation are in a better position to deter one another, as they can more easily match the actions of their rivals. Moreover, three-star hotel chains could increase their degree of overlap with rivals' markets because they are also more likely to hold a credible threat of retaliation against rivals. Lastly, low-category chains or chains with hotels that offer a unique combination of services should avoid entering the same rivals' markets because a higher degree of multimarket contact would increase the intensity of rivalry in those markets. 8. Limitations and future research This study focuses on understanding the joint effect of multimarket contact and vertical and horizontal differentiation on the intensity of price competition between chain hotels. The first possible limitation of the study is related to the dimensions of differentiation discussed in this paper. The existing literature on the hotel industry indicates that there are many attributes that influence a customer's choice of hotels, such as service quality, reputation, security, and cleanliness (Chu & Choi, 2000). In the present study, the overall differentiation of the hotels is measured by the number of stars, the Q certificate, the use of a common umbrella brand, and the distance in services; however, other sources of vertical and horizontal differentiation might also be considered. An additional limitation is that the empirical results are based on listed prices only. Although Avlonitis and Indounas (2007) reported that list pricing is the pricing policy most commonly used for service companies, future research should investigate whether the substitution effect between multimarket contact and horizontal differentiation holds with other pricing policies. Methodologically, the empirical approach adopted also has some limitations because it uses a single cross-section of the data and therefore is unable to control for individual firm fixed effects. Thus, this approach cannot completely eliminate the possibility that the results may be affected by unobserved hotel characteristics that are correlated with price, differentiation strategies, and multimarket contact. Finally, because this study focuses only on price, the next logical step is to test these hypotheses using the financial performance of firms as the dependent variable. This analysis would provide a better understanding of the joint effect of different profitgenerating mechanisms, as suggested by Makadok (2010; 2011). References Abrate, G., Capriello, A., & Fraquelli, G. (2011). When quality signals talk: evidence from the Turin hotel industry. Tourism Management, 32, 912e921.
n, J. M., & Rubio-Andrada, L. (2012). Reasons for Alonso-Almeida, M., Rodríguez-Anto implementing certified quality systems and impact on performance: an analysis of the hotel industry. The Services Industries Journal, 32(6), 919e936. Avlonitis, G., & Indounas, K. (2007). An empirical examination of the pricing policies and their antecedents in the services sector. European Journal of Marketing, 41(7/8), 740e764. Baum, J. C., & Korn, H. J. (1999). Dynamics of dyadic competitive interaction. Strategic Management Journal, 20(3), 251e278. Baum, J. A. C., & Mezias, S. J. (1992). Localized competition and organizational failure in the Manhattan hotel industry, 1898-1990. Administrative Science Quarterly, 37(4), 580e604.
, J., & Silva, R. (2013). Being better vs. being different: differBecerra, M., Santalo entiation, competition, and pricing strategies in the Spanish hotel industry. Tourism Management, 34, 71e79.

R. Silva / Tourism Management 48 (2015) 305e315 Bernheim, B. D., & Whiston, M. D. (1990). Multimarket contact and collusive behaviour. Rand Journal of Economics, 21(1), 1e26. Bull, A. O. (1994). Pricing a motel's location. International Journal of Contemporary Hospitality Management, 6(6), 10e15. Canina, L., Enz, C. A., & Harrison, J. S. (2005). Agglomeration effects and strategic orientations: evidence from the U.S. lodging industry. Academy of Management Journal, 48(4), 465e481. Casadesús, M., Marimon, F., & Alonso, M. (2010). The future of standardized quality management in tourism: evidence from the Spanish tourist sector. The Services Industries Journal, 30(14), 2457e2474. Chamberlin, E. H. (1929). Duopoly: value where sellers are few. Quarterly Journal of Economics., 44(1), 63e100. Chen, M. (1996). Competitor analysis and interfirm rivalry: toward a theoretical integration. Academy of Management Review, 21(1), 100e134. Chow, K., Garretson, J., & Kurtz, D. (1994). An exploratory study into the purchase process used by leisure travellers in hotel selection. Journal of Hospitality & Leisure Marketing, 2(4), 53e72. Chu, R. K. S., & Choi, T. (2000). An importance-performance analysis of hotel selection factors in the Hong Kong hotel industry: a comparison of business and leisure travelers. Tourism Management, 21(4), 363e377. Cremer, H., & Thisse, J. F. (1991). Location models of horizontal differentiation: a special case of vertical differentiation models. Journal of Industrial Economics, 39(4), 383e390. , M. (2003). Effect on prices of the Espinet, J. M., Saez, M., Coenders, G., & Fluvia attributes of holiday hotels: a hedonic prices approach. Tourism Economics, 9(2), 165e177. Ethiraj, S. K., & Zhu, D. H. (2008). Performance effects of imitative entry. Strategic Management Journal, 29(8), 797e817. Evans, W., & Kessides, I. (1994). Living by the ‘Golden Rule’: multimarket contact in the U.S. Airline Industry. Quarterly Journal of Economics, 109(2), 341e366.
ndez, N., & Marín, P. L. (1998). Market power and multimarket contact: some Ferna evidence from the Spanish hotel industry. Journal of Industrial Economics, 46(3), 301e315.
mez, J. (2006). Multipoint competition, strategic similarity and Fuentelsaz, L., & Go entry into geographic markets. Strategic Management Journal, 27(5), 477e499. Gabszewicz, J., & Thisse, J. (1986). On the nature of competition with differentiated products. Economic Journal, 96(381), 160e172. Gimeno, J., & Woo, C. Y. (1996). Hypercompetition in a multimarket environment: the role of strategic similarity and multimarket contact in competitive deescalation. Organization Science, 7(3), 322e342. Gimeno, J., & Woo, C. Y. (1999). Multimarket contact, economies of scope, and firm performance. Academy of Management Journal, 42(3), 239e259. Hotelling, H. (1929). Stability in competition. Economic Journal, 39, 41e57. Hughes, K., & Oughton, C. (1993). Diversification, multi-market contact and profitability. Economica, 60, 203e224. Jans, I., & Rosenbaum, D. (1997). Multimarket contact and pricing: evidence from the U.S. cement industry. International Journal of Industrial Organization, 15(3), 391e412. Jayachandran, S., Gimeno, J., & Varadarajan, P. R. (1999). The theory of multimarket competition: a synthesis and implications for marketing strategy. Journal of Marketing, 63(3), 49e66. Kang, W., Bayus, B., & Balasubramanian, S. (2010). The strategic effects of multimarket contact: mutual forbearance and competitive response in the personal computer industry. Journal of Marketing Research, 47(3), 415e427. Levenstein, M., & Suslow, V. (2006). What determines cartel success? Journal of Economic Literature, 44(1), 43e95. Lewis, R. C. (1985). The market position: mapping guests' perceptions of hotel operations. The Cornell Hotel and Restaurant Administration Quarterly, 26(2), 86e98.
pez, M. C., & Serrano, A. M. (2004). Is the hotel classification system a good indicator Lo of hotel quality? An application in Spain. Tourism Management, 25, 771e775. Makadok, R. (2010). The interaction effect of rivalry restraint and competitive advantage on profit: why the whole is less than the sum of the parts. Management Science, 56(2), 356e372. Makadok, R. (2011). The four theories of profit and their joint effects. Journal of Management, 37(5), 1316e1334.

315

McCleary, K. W., Weaver, P. A., & Hutchinson, J. C. (1993). Hotel selection factors as they relate to business travel situations. Journal of Travel Research, 32(2), 42e48. Nicolau, J. L., & Sellers, R. (2010). The quality of quality awards: diminishing information asymmetries in a hotel chain. Journal of Business Research, 63, 832e839. Nield, K., & Kozac, M. (1999). Quality certification in the hotel industry. The Cornell Hotel and Restaurant Administration Quarterly, 40(2), 40e52. ~ ez-Serrano, J. A., Turrio
n, J., & Vela
zquez, F. J. (2014). Are stars a good indicator Nún of hotel quality? Asymmetric information and regulatory heterogeneity in Spain. Tourism Management, 42, 77e87. O'Neill, J. W., & Mattila, A. S. (2010). Hotel brand strategy. Cornell Hospitality Quarterly, 51(1), 27e34. O'Neill, J. W., & Qu, X. (2006). The role of brand affiliation in Hotel market value. The Cornell Hotel and Restaurant Administration Quarterly, 47(3), 210e223. €ller, L. (1997). Collusive conduct in duopolies: multimarket contact Parker, P., & Ro and cross-ownership in the mobile telephone industry. RAND Journal of Economics, 28(2), 304e322. Pilloff, S. J. (1999). Multimarket contact in banking. Review of Industrial Organization, 14(2), 163e182. Renaghan, L. M., & Kay, M. Z. (1987). What meeting planners want: the conjointanalysis approach. The Cornell Hotel and Restaurant Administration Quarterly, 28(1), 66e76. , M. (2011). Managing tourism products and destinaRigall-I-Torrent, R., & Fluvia tions embedding public good components: a hedonic approach. Tourism Management, 32(2), 244e255.
n, J. M. (2011). Motivation and Rubio-Andrada, L., Alonso-Almeida, M., & Rodríguez-Anto impacts in the firm and stakeholders of quality certification: evidence from smallmedium-size service enterprises. Total Quality Management, 22(8), 833e852. Scherer, F. M., & Ross, D. (1990). Industrial market structure and economic performance (3rd ed.). New York: Houghton Mifflin Company. Shaked, A., & Sutton, J. (1982). Relaxing price competition through product differentiation. Review of Economic Studies, 49(155), 3e13. Shaked, A., & Sutton, J. (1987). Product differentiation and industrial structure. Journal of Industrial Economics, 36(2), 131e146. Stigler, G. J. (1987). The theory of price. Prentice Hall College Div. (4th ed.) Taninecz, G. (1990). 1990 Business traveler survey. Hotel and Motel Management, 205, 29e32. Thomas, C. J., & Willig, R. D. (2006). The risk of contagion from multimarket contact. International Journal of Industrial Organization, 24(6), 1157e1184.
rrez, I. (2006). Hotel location in tourism cities: Madrid 1936Urtasun, A., & Gutie 1998. Annals of Tourism Research, 33(2), 382e402. Yavas, U., & Babakus, E. (2003). What do guests look for in a hotel? A multi-attribute approach. Services Marketing Quarterly, 25(2), 1e9. Young, G., Smith, K., Grimm, C., & Simon, D. (2000). Multimarket contact and resource dissimilarity: a competitive dynamics perspective. Journal of Management, 26(6), 1217e1236. Yu, T., & Cannella, A. A. (2013). A comprehensive review of multimarket competition research. Journal of Management, 39(1), 76e109.

Rosario Silva is Professor in the Strategic Management Department at IE Business School (Alvarez de Baena, 4, 28006 Madrid, Spain; Email ). Her fields of interest include differentiation strategies and geographic agglomeration.