Price parity on a par : modelling the effects of narrow price parity clauses combined with best price guarantees in the online hotel booking market

(1)

PRICE PARITY ON A

PAR

Modelling the effects of narrow price parity clauses combined with best price

guarantees in the online hotel booking market

Name: Francisca Wals

Student no.: 5970644

Supervisor: dr. Maarten Pieter Schinkel

Program: MSc Economics

(2)

1

2 INTRODUCTION

“Booking too powerful? Not at all.” It was this quote that figured as a headline for an interview with Booking.com CEO Gillian Tans (Bouma, 2016). The statement was a response to the much-heard complaint that the world’s market leader in online hotel bookings would abuse its dominant position by imposing unreasonable clauses on hotel companies and other accommodation owners, thereby distorting competition. According to Tans, this is “totally not” the case.

Numerous national competition authorities (NCAs) throughout Europe are or have been of quite a different opinion. Especially Booking.com’s so-called price parity clauses (PPCs) – which commits hotels1

to not charging a lower price for the same offer on any other online channel – have been the subject of extensive scrutiny and condemnation. This has led Booking.com to adjust its clauses to some extent, narrowing their scope. While these adjustments have been accepted by many NCAs, the Bundeskartellamt has taken a harsher stance. The German NCA decided to decline the amendments and issued a prohibition against Booking.com’s clauses. Also, in France a law has been passed which prohibits all price relation agreements (PRAs) in the online hotel booking sector.

These divergent approaches to Booking.com’s and also several other online booking platforms’2_practices,

have led to a legislative chaos across Europe. “Businesses and consumers expect legal certainty, predictability and a uniform application of the law”, as was recently stressed by the Directorate-General of Competition, Johannes Laitenberger (2016). Also European Commissioner for Competition Margarethe Vestager has expressed her concerns on the different treatments of Booking.com and stated that “[i]t is essential for our internal market that business can rely on consistent application of EU competition rules regardless of which authority is in the lead” (Vestager, 2015). “Greater coordination is needed.”

It is not only the existence of 28 separate and independent NCAs that hinders the achievement of a single approach; it is also the fact that a firm theoretical basis for European or even worldwide consistency regarding Booking.com’s price clauses is lacking. Most of the literature – both academic and applied – focuses on the effects of wide price parity clauses – which prohibit hotels from charging a lower price on both their own website and other online booking platforms. However, it is the narrow price parity clauses – which only prohibit hotels from charging a lower price on their own website – that currently raise issues.3

Where narrow price parity clauses are addressed, the analysis is either strictly qualitative or/and it misses key features of the way and the market in which Booking.com operates.

In this thesis, it is my aim to fill this blank and provide a more solid theoretical basis for evaluating the effects of narrow price parity clauses on consumer and total welfare and hotel and platform profits in the market for intermediary services of online hotel portals, with a special focus on Booking.com, that – with its 62 percent share of the European market – institutes a paradigmatic case. Doing so, I will take into account the fact that Booking.com not only imposes a narrow price parity clause on the hotels displayed

1_{Booking.com also mediates between suppliers that offer apartments or other accommodations. For brevity,}

however, I will just use ‘hotels’ also where I mean accommodations in a wider sense.

2_{Although Booking.com holds a very dominant position in Europe, there are some other, smaller online hotel booking}

platforms that have been scrutinized by the national competition authorities, such as HRS and Expedia.

(5)

4

on the platform; it also offers its customers a best price guarantee (BPG), granting them the opportunity of claiming a refund if they find the same offer for a better price elsewhere. This, as I will show by means of extending an existing game theoretic model, has a significant impact on competition. Eventually, I attempt to provide a solid theory of harm of the narrow PPC-BPG combination that Booking makes use of – thus contributing to the existing policy debate regarding the diverging NCA stances towards price parity clauses in the online hotel portal market.

This thesis is structured as follows. In chapter 3, I will discuss the way and the market in which online hotel booking platforms in general and Booking.com specifically operate. I will also devote attention to the proceedings against Booking.com and the legal framework in which this has been done. In chapter 4, I will discuss the relevant literature on price parity clauses in the online hotel booking sector and best price guarantees in general. In chapter 5, I will introduce and extend the model, with which I will show that the effects of imposing a BPG-narrow PPC on customer4_{surplus is significant.}

3 ONLINE HOTEL BOOKING PLATFORMS

In this section, I will elaborate on the operation of online hotel portals in general and Booking.com specifically. I will do this by recounting the way Booking.com operates – a business model that is quite general for online hotel booking platforms. I will then define the relevant market in which Booking.com operates, before discussing characteristics of this market and the services online hotel booking platforms provide in some detail, especially their potential of drastically reducing search costs. Next, I will discuss the price relation agreements online hotel booking platforms typically make use of, before I narrow focus again and elaborate on the proceedings against Booking.com and the legal framework in the context of which this has been done.

3.1

ON BOOKING

.

COM

Booking.com is currently the world’s largest online hotel portal, with almost 970,000 accommodations listed on the platform and more than a million bookings completed on it per day (Booking.com, 2016). Although exact figures are not available, Booking.com’s turnover and profit were estimated to be 4.2 billion and 1.6 billion respectively in 2014.5_{Founded in 1996 in the Netherlands, it was taken over by the}

American Priceline Group in 2005 for a mere 120 million euros. It was a deal that grew out to be the most profitable acquisition in the travel market in the 2000s, catapulting The Priceline Group’s value from around 100 million to 57 billion euros in ten years, with Booking.com being the main driver of its revenues (O'Neill, 2012). Booking.com is the biggest Google AdWords spender in the Travel and Tourism category

4_{In order to avoid confusion with other terminology, I use ‘customers’ instead of ‘consumers’ when I refer to}

(potential) hotel visitors.

5_{Booking.com is not very generous with data; these numbers are estimates based on reports from the Dutch}

(6)

5

and is the 101st_{most popular site worldwide}6_{. By comparison: its closest competitor, Expedia.com, holds}

a position of 430 (Alexa, 2016). The agency model

Like most European online hotel booking platforms, Booking.com operates according to the so-called agency model, as it acts as an ‘agent’, selling on behalf of the hotels that are listed on the platform. Importantly, Booking.com is not a ‘reseller’ that purchases from upstream suppliers in order to resell to downstream customers. Within the agency model, hotels set the final price and the platform receives a commission for each sale made.

Booking.com, in its turn, provides demand-enhancing services to hotels – such as search marketing7_,

advertising, after-sales support, guarantees and information in the form of reviews and orderly information in multiple languages – making them both (more) easily findable for customers searching online and also helping them achieve higher conversion rates8_{(Ezrachi, 2016). To customers, Booking.com}

provides services which can be summarized as ‘search, compare and book’ benefits (Bundeskartellamt, 2015 (a)), a term which captures the fact that customers’ search costs of finding an accommodation according to their budget and taste are greatly reduced, due to easy-access reviews and information on hotel room features, prices and terms in multiple languages. Booking.com also provides additional, transaction-related benefits, such as making payment more convenient9_{and offering after-sales}

services10_.

As mentioned, Booking.com charges hotels for its services by asking a certain percentage of the price for each realized booking on the platform. These commission fees are usually between 15 and 20 percent of the hotel room price, but can be as high as 30 to 50 percent, depending on for example the raking of the hotel in the search results of the platform (Bundeskartellamt, 2015 (a)). Typically, Booking.com doesn’t charge booking fees nor registration fees to its customers, on which I will elaborate in the next section. The fee that Booking.com charges to hotel companies, however, is passed through to customers to a greater or lesser extent, depending on the customers’ demand price elasticity (Edelman & Wright, 2015, p. 20), i.e.: how much their demand for hotel rooms decreases in response to a price increase, indirectly instigated by Booking.com increasing its fees.

3.2

THE RELEVANT MARKET

The market in which Booking.com operates has been strictly and clearly defined by the Bundeskartellamt in its extensive decision report on Booking.com (2015 (a), pp. 39-48), which refers back to the market definition in decision on HRS11_{(Bundeskartellamt, 2013 (a), pp. 23-40). Since this definition of the relevant}

6_{As estimated by the Alexa Traffic Ranks, based on pageviews and unique visitors (Alexa, 2016).}

7_{An important aspect of search marketing is the bidding on search words in order to achieve a high listing rank on}

search engines.

8_{‘Conversion rate’ refers to the ability to turn visitors into customers.}

9_{For example, with respect to standardized payment methods and automatic entering of customer information.} 10_{Such as the possibility to manage one’s bookings with an app or to receive updates and reminders.}

11_{Another important though much smaller online hotel booking platform, which forced to drop its price parity}

(7)

6

market has been endorsed by the Swedish, Italian and French competition authorities, I will take it as a guiding thread.

The market for intermediary services

The Bundeskartellamt has defined the market in which Booking.com and also other, smaller hotel platforms operate, as an offer-driven market for the online hotel booking platforms’12_intermediary

services, in which the platforms as suppliers stand opposite the hotels as consumers.13_{It is mainly from}

the perspective of the hotels as ‘consumers’14_{of the platform’s services that the relevant market is}

delineated – as consisting of those bundles of services that count as substitutes for one another, thus identifying competitors that are capable of constraining the behavior of the company in question (Bundeskartellamt, 2013 (a), p. 23).15_{As for end consumers, their behavior is relevant for defining the}

relevant market to the extent that their search, compare and book behavior influences the economic importance and thus the functional substitutability of hotel platforms’ services for hotels.

From the hotels’ perspective, Booking.com’s primary service is its intermediary service. Making rooms searchable online and letting customers compare them with other offers – a service bundle summarized as ‘search, compare and book’ functions – counts as a secondary service, in so far that a platform’s success in intermediating between hotels and their customers also to a large extent depends on providing customers with these secondary services (Bundeskartellamt, 2015 (a), p. 45). In other words: the way in which customers search, compare and book indirectly influences the functional substitutability of platforms’ intermediary services, from the perspective of hotels.

Offline sales

According to this market definition, offline sales channels, hotels’ own websites, metasearch engines and online travel agencies do not belong to the same market as Booking.com. To begin with, there exists a differentiation between online and offline sales (Bundeskartellamt, 2013 (a), pp. 27-29), as offline sales (via telephone, email16_{, brick-and-mortar travel agencies or in person) do not count as a substitute for}

online sales for the end consumer whose interest is in conducting a rapid, targeted search for the desired hotel room characteristics, with the possibility of comparing a wide range of offers.17

12_{The Bundeskartellamt refers to online hotel booking platforms as “hotel portals”. I chose a different terminology}

including “platform”, since this better corresponds to the literature.

13_{This is a remarkable reversion of the way in which the interaction between platforms and the companies that offer}

their products on it are usually depicted; namely, as platforms being the downstream players and suppliers operating upstream.

14_{This is the reason why I denote hotel visitors by ‘customers’ in stead of ‘consumers’.}

15_{The reason for this focus is that the Bundeskartellamt’s investigation concentrates on the contractual relationship}

between hotels and online hotel booking platforms.

16_{Also bookings via email count as an offline sales, as the booking is not processed through an online booking system.} 17_{“General experience suggests that a change from online to offline sales in any case does not take place with a}

simple hotel room search. According to a market survey of the European Commission travel customers who do not get the information from one online provider typically continue their search with another online provider, but they do not continue their search offline” (Bundeskartellamt, 2013 (a), p. 28).

(8)

7 Online sales

Within online sales, further distinctions are to be made between hotel platforms and other online channels. First, as argued by the Bundeskartellamt (2013 (a), p. 33), hotels’ own websites are not part of the same market as Booking.com, since customers can neither look for hotel rooms of other hotels on hotels’ own websites, nor can they compare hotel rooms and obtain information via customer reviews. as they only offer customers the possibility of booking a room in a specific hotel. Especially small and medium sized hotels that are hard to find without the help of a hotel platform or metasearch engine use their websites merely as an advertising site rather than as an additional sales channel which substitutes selling through a hotel platform. Also metasearch engines are not part of the same market (pp. 36-39), since they do not offer customers the opportunity to book a room on the engine; they merely direct (potential) hotel customers to the booking offers of hotel companies or booking platforms – their remuneration depending on the number of clicks generated. Hence, they operate on another level of the distribution chain and do not count as a substitute for hotel platforms. As for online travel agencies and tour operators (pp. 34-36) – that offer packaged tours often comprised of hotel rooms as well as flights, hire cars and tourist services – their target group is quite different from that of the hotel platforms, who focus on business and individual travelers who primarily wish to book hotel rooms on the hotel portal and look for further services, such as flights and hire cars, on other portals. Also, online travel agencies and tour operators generally only list the price of the total packaged tour, which prohibits the customer from comparing the prices of individual hotel rooms.

In conclusion, the market in which Booking.com operates is to be defined as the market for the intermediary services of online hotel booking platforms, in which hotels have contractual ties with the hotel platform and set the prices listed on the platform, on which potential hotel customers can enjoy the ‘search, compare and book’ service bundle.18

3.3

MARKET CHARACTERISTICS

Having defined the relevant market, I now shift focus towards the importance of the market for the services of hotel platforms in relation to other, complementary hotel room sales channels as well as Booking.com’s position in this market and other characteristics of the way in which this market functions. High market concentration

In Europe, 57 percent of the hotel room bookings take place online, of which around 45 percent through online hotel booking platforms. The market is highly concentrated; around 75 percent of the bookings through hotel platforms is generated by two players: 18 percent (of this 75 percent) by Expedia, and 82 percent by Booking.com. This comes down to a market share of 62 percent for Booking.com in Europe19

(Barthel & Perret, 2015). However, these numbers differ much per country. Booking has the highest market share in Greece, Norway, the Netherlands and Belgium at over 70 percent. In Hungary and France,

18_{The claim that other sales channels are no close substitutes of the intermediary service that online hotel booking}

platforms provide is endorsed by the way in which hotel companies responded to HRS (which has a market share of around 30 percent in Germany) increasing its commission by more than 15 percent and tightening the price parity clauses: they complained, but they didn’t withdraw from the platform (Bundeskartellamt, 2013 (a), p. 31).

19_{In the United States, the roles are reversed; there, Expedia is the dominant player with a market share of…}

(9)

8

however, Booking.com has a market share of just over 40 percent. In Germany, its market share is over 50 percent. The differences are also sizeable for Booking.com’s closes competitors, HRS and Expedia. In Italy, for example, Expedia has a market share of 15 percent, but only 2,5 percent in Finland. HRS – which was founded in Germany – has a market share of over 30 percent in Germany, but is only a minor player in most other European countries (Schegg, 2015). However much these figures differ, Booking.com is irrefutably by far the largest player in the relevant market in Europe.

A two-sided market

One of the drivers of this high market concentration is the fact that the market in which Booking.com and other online hotel booking platforms operate can be characterized as a two-sided market of which indirect network effects are a key characteristic. This means that users on one side of the market do not directly benefit from an increase in the number of users on their market side, but only indirectly: an increase in the number of users on their market side attracts more potential transaction partners on the other market side (Haucap & Stühmeier, 2015). Hence, if Booking.com generates a lot of traffic on its platform, this will attract more hotels offering their rooms on the platform, from which customers also benefit. Similarly, if many hotels decide to join the hotel platform, the more customers will be drawn to the platform, from which hotels then benefit. Because these indirect network effect imply that the more intensive use of a hotel platform by one side of the market triggers greater use of the platform also by the other side. Therefore, large platform sizes indispensable for achieving an efficient utilization of the platform, with both sides of the market coordinating on joining on the platform, maximizing the positive externalities of the indirect network effect. Hence, high market concentration is rather typical for this sort of two-sided online platform market. Another driver of the high concentration in this market is the fact that two-sided online markets experience increasing returns to scale, as the fixed costs of setting up and maintaining the technical infrastructure platform are relatively high, while variable costs (of additional bookings) are low (Haucap & Stühmeier, 2015). Hence, entry into the online hotel booking business is not easy, both due to the indirect network effect and the economy of scale.

Only hotels charged with per-transaction fees

A remarkable feature of most online hotel booking platforms is that it is the hotel-side of the market that is (directly) charged for the platforms’ intermediary services – and not the customers, neither through per-transaction fees nor registration fees for joining20_{. According to Wang & Wright (2016, pp. 31-32), a}

possible reason for this could be that in case consumers are explicitly charged for searching or booking on a platform, this could increase their incentive searching and/or booking elsewhere. In the case of sufficient platform competition, coordination on the platform that does not charge consumers any fees seems likely. Hence, the side of the market which is less sensitive to explicit fees.21_{Similarly, hotels are not charged any}

reservation fees, since they then might make a loss if consumer demand turns out to be less than expected, which may also lead hotel companies to coordinate on non-mediation or mediation by another platform.

20_{Of course, customers indirectly pay for these services, since the fees are passed through in the prices that hotel}

companies set on the platforms.

21_{‘Explicit’ fees, since, as mentioned, at least a part of the fees charged to hotel companies is passed through to hotel}

(10)

9 Search cost reduction

As mentioned, the reason customers choose to search through an online hotel booking platform, rather than searching offline or only searching on hotels’ own websites, is the ‘search, compare and book’ service bundle. As I will elaborate on in more detail in subsequent sections, it is the search element of this bundle that has a sizeable impact on both (end) consumer welfare and hotel profits.

Generally speaking, search costs are the cost incurred by a buyer (within this context, a hotel customer) to locate an appropriate seller (a hotel company) in order to purchase a product (a hotel room). These costs include the opportunity cost of time spent on search, as well as other expenditures such as telephone calls, driving somewhere to obtain information, et cetera (Yannis Bakos, 1997, p. 2). As the online hotel booking market is a market with differentiated products, search costs can be quite high without an appropriate means to easily compare not only prices, but also characteristics of the hotel room offers. Regarding its impact on (end) consumer welfare, it is in their process of price discovery and identification of appropriate hotel room offerings that the search-cost reducing potential of hotel platforms plays a crucial role. As is well known from microeconomic theory, commodity and service markets are closer to a competitive equilibrium the more that buyers are costlessly and fully informed about the prices and product characteristics. This is because with high search costs, consumers are more willing to settle for a less-than-ideal offer, in order to avoid the search costs. Therefore, as is the case in a standard differentiation model such as Hotelling’s model, sellers (hotel companies) have some price-setting power, which enables them to charge prices above-marginal costs (Yannis Bakos, 1997, p. 3) (Wang & Wright, 2016).22

Online (hotel) booking platforms provide customers with low-cost product and price information; this enables consumers to look at more or even all product offerings and thus purchase the one best serving their needs. In other words, customers hold out for a higher ‘match value’ (Wang & Wright, 2016, p. 12), resulting in a socially optimal allocation. Besides these allocational efficiencies, the search-cost reducing potential of online booking platforms makes consumers better off because of the increased competition – in both quality and price – among sellers due to the wider availability of information which allows customers to make informed comparisons between offers (Ezrachi, 2016, pp. 492-493) (Yannis Bakos, 1997, p. 3) (Wang & Wright, 2016, pp. 14-15). A third reason why reduced search costs benefit customers is that, even though customers tend to search more on a platform, the increased information flows tend to reduce net search costs. (Ezrachi, 2016, p. 493)

Sellers, however, are always made worse off by this reduction of search costs implied by the existence of online booking platforms. As mentioned, the higher the search costs, the less informed the customers, and the higher sellers’ market power, and the higher the prices they can charge. Because lower search costs forces sellers to price more competitively, their mark-up and thus their profits will be lower (Yannis Bakos, 1997, p. 3). Hence, sellers as a group have no incentive to join an online booking platform, because it will reduce their total profits. Collectively they would do better to pull out of the platform; individually,

22_{Yanis Bakos stresses that high search costs may even lead to market breakdown. This may be the case for small}

(11)

10

however, they prefer to join, since otherwise, the resulting lower demand for their offers will decrease their profits given that other sellers do in fact join the platform. (Wang & Wright, 2016, p. 16).

Price relation agreements: price parity clauses

As already briefly mentioned, various price relation agreements play a prominent role in the relationship between Booking.com and its suppliers, the hotel companies. These price relation agreements (PRAs) are rather typical in the online environment where sellers sell through a platform which operates according to the agency model and invests in demand-enhancing features (Ezrachi, 2016, p. 490).23

In particular, Booking.com – and also other major platforms such as HRS and Expedia – introduced price parity clauses24_{. A price parity clause (PPC) is a special type of price relationship agreement between a}

supplier (hotel) and a platform that determines the price the buyer (hotel customer) pays. These type of PRAs are called ‘third-party PRAs’ , since the effect of the agreement is realized on a party that does not partake in the agreement (Akman, 2015, p. 5).

There are two types of PPCs. The first type is a wide PPC, which Booking.com has been using up until it was forced by the Swedish, Italian and French competition authorities to make commitments, which I will discuss in more detail in the next section. Such a wide PPC obliges suppliers not to offer higher prices on the platform in question than on any other online or offline channel. That is, a hotel cannot offer a higher price on Booking.com than on other online hotel booking platforms, nor on its own online and offline direct selling channels. Thus, as hotels can no longer price differentiate between various sales channels, they base their pricing decisions on an average of the booking fees they face on several platforms. This incentivizes Booking.com to increase its fees, since there is no risk of other channels (either platforms or direct sales channels) offering a lower price. This implies that Booking.com expects to lose far fewer customers due to its fee increase than would be the case absent the wide PPC (Fletcher & Hviid, 2014, pp. 10-11). Hence, the wide PPC effectively removes all competitive restraints on commission fees (and, indirectly, hotel room prices) implied by horizontal between-platform competition and also vertical

23_{Fletcher & Hviid (2014, pp. 7-8) list four reasons why price parity clauses (‘retail price most favoured nation clauses’}

in their terminology) are primarily observed in online environments. The first reason is the legal treatment of price relation agreements – suchs as a price parity clause – is quite unclear, due to the fact that there is confusion about whether online platforms are to be viewed as ‘genuine agents’, which is a condition for the EU legislation on retail price maintenance to be applicable. A second reason is that online platforms have a strong negotiating position vis-à-vis the suppliers they deal with, due to network-related market concentration and the fact that the platforms serve as critical gateways to customers. This allows platforms to insist on PPCs, even though such clauses are not in the interest of suppliers. A third reason is that the enforcement of the clauses is far easier in the online than in the offline world, since suppliers’ prices can be permanently and almost costlessly tracked electronically. Finally, a fourth reason why we observe so many PPCs in the online world is that, due to low search and switching costs and minimum capacity constraints, online platforms are likely to face fierce price-based competition, which makes it especially attractive to adopt price relation agreements in order to limit or eliminate competition.

24_{Others have called these clauses ‘most-favored nation’ or ‘most-favored customer’ clauses. However, it is possible}

to price discriminate between customers with a price parity clause being in place. A PPC therefore doesn’t imply that a customer pays the best price available; it just implies that this specific customer will not find a lower price on the supplier’s direct sales channel (in this context, a hotel’s own website) nor – in case of a wide PPC – on any other platform. Also, the ‘customer’ terminology is confusing, since in the case of two-sided platforms, end consumers are usually denoted by ‘customers’, who are not part of the ‘most-favored customer’ contract. I therefore choose to use the ‘price parity clause’ terminology.

(12)

11

competition from hotels’ direct sales channels. Booking.com’s wide PPCs covered not only parity in prices, but also in room availability and terms in conditions; that is, hotel companies listed on Booking.com were not allowed to offer more rooms or better terms and condition on any other channel (Bundeskartellamt, 2015 (a), pp. 9-10).

The second type of PPC, a narrow PPC, is the agreement that the price the suppliers offer on the platform cannot be higher than the price they charge on their own direct selling channels. As mentioned, Booking.com implemented narrow PPC as of 1 July 2015, after several NCAs had issued complaints. By then, HRS was already forced to narrow the scope of its wide PPC and Expedia followed suit. This narrow PPC consists in the agreement that hotel companies are in fact allowed to price differentiate between Booking.com and other online booking platforms, whereas they cannot offer lower prices or terms on their own online channels. Hotel companies are allowed, however, to offer rooms for a better price on offline channels, provided that those prices are not published online (Bundeskartellamt, 2015 (a), p. 11). Formally, a narrow PPC leaves the competitive pressure implied by other booking platforms intact, while removing the competitive restraint implied by hotels’ direct sales channels.

Booking.com (and also other online booking platforms) defends its use of PPCs by means of a so-called ‘showrooming’ argument. The platform argues that both hotels and end customers are able to freeride on the services provided by Booking.com as long as hotel companies are placed in a position where they can offer a lower price on their own sales channels. This is because consumers have an incentive to use the platform to search for a suitable hotel, but then complete their reservation on the hotel’s own website if it offers lower prices and/or better terms, giving rise to a vertical externality. Another externality, a horizontal one, may emerge between platforms, since a lower-quality but also lower-price platform may freeride on a higher-quality, higher-price platform (Ezrachi, 2016, pp. 490-491). According to Booking.com, these types of freeriding behavior may eliminate or severely damage its ability to operate (Bundeskartellamt, 2015 (a), p. 30) and/or invest in benefits that do not tie the user to the platform (Ezrachi, 2016, p. 491).

Price relation agreements: best price guarantees

“We want you to pay the lowest price possible for your stay. Should you find your accommodation, with the same reservation conditions, at a lower rate on the Internet after you have made a reservation through us, we will match the difference between our rate and the lower rate.” It is with these words that Booking.com (2016) advertises its best price guarantee (BPG)25_{on its website. As such, the type of BPG}

Booking.com offers is a price-matching guarantee on advertised prices.26_{As is the case with price parity}

clauses, Booking.com isn’t unique in offering potential customers a best price guarantee; also HRS and Expedia have adopted them (Expedia, 2016) (HRS, 2016).

A best price guarantee is a special type of across-sellers agreement (Buccirossi et al. , 2012). Across-seller agreements encompass those agreements according to which the prices that sellers promise to charge depend on the price charged by competing sellers. In the present case, ‘sellers’ are not the actual sellers

25_{Best price guarantees are also known as ‘price matching’ or ‘price beating’ guarantees.}

26_{Arbatskaya (2001) defines best price guarantees (which she denotes by “low price guarantees”) as a vector of}

characteristics such as whether they are match or beat competitors’ prices, an whether they are based on advertised or effective prices – effective prices being the outcome of a recursive application of low price guarantees.

(13)

12

of hotel rooms, but their agents: online hotel booking platforms. Thus, we have the interesting construction of platforms offering best price guarantees on prices they do not set themselves.

Booking.com promises to match the price of a better deal bookable on any other online channel at the time of booking if the other deal is subject to the same booking condition a customer claiming a refund has chosen with Booking.com; i.e. the same property, the same room time, the same in and check-out dates and identical room conditions (breakfast, cancellation policy, etc.). These terms and conditions are almost identical to those attached to Expedia’s and HRS’s respective BPGs.

Summarizing, the two-sided market in which online hotel booking platforms – such as Booking.com – operates is the market for intermediary services offered to hotel companies, from the perspective of which benefits offered to customers are secondary services and into which entry is made difficult due to indirect network effects and economies of scale. From a welfare perspective, hotel platforms’ search cost reducing potential is a key factor, since lower search costs allow customers to find a better offer and increase competition between hotel companies. Trying to avoid customers freeriding on the non-transaction related benefits and search cost reduction, all major hotel platforms have adopted price parity clauses. To the other side of the market – potential customers – they offer best price guarantees.

3.4

PROCEEDINGS AGAINST BOOKING

.

COM

Having discussed characteristics – including commonly observed pricing strategies – of the market in which Booking.com and other hotel platforms operates in a more or less general way, I now narrow focus again and elaborate on Booking.com’s track record with several European (competition) authorities. But not before I briefly discussed the legal framework according to which Booking.com’s price parity clauses have been treated by the NCAs.

Legal framework

The legal provision under which online hotel booking platforms’ price parity clauses have been investigated and/or prohibited by several NCAs (see next section) is Article 101(1) of the Treaty on the Functioning of the European Union (TFEU) or its national equivalent27_{(Akman, 2015, p. 22). This article}

prohibits “all agreements between undertakings, decisions by associations of undertakings and concerted practices which may affect trade between Member States and which have as their object or effect the prevention, restriction or distortion of competition within the internal market”, especially if they “directly or indirectly fix purchase or selling prices or any other trading conditions” (EUR-Lex, 2008). However, an agreement or concerted practice violating Article 101(1) might be justified if it satisfies four conditions: 1) it must contribute to improving the production or distribution of goods or to promoting technical or economic progress; 2) consumers28_{must receive a fair share of the resulting benefits; 3) the imposed}

restrictions must be indispensable to the attainment of these objectives; and 4) afford such undertakings [that are part of the agreement] the possibility of eliminating competition in respect of a substantial part of the products in question (EUR-Lex, 2008). All of these conditions must be satisfied in order for the

27_{In principle, member states’ legislation on competition conform to those set out in the TFEU, especially articles}

101 and 102 (OECD, 2015, p. 3).

(14)

13

relevant agreement to be justified. The NCAs that have investigated or prohibited hotel platforms’ price parity clauses all deemed those clauses to be a restriction on competition by effect.29

Chronological overview

The trouble for Booking.com began in September 2010, when the British competition authority (OFT) opened an investigation on the ground of suspecting a breach of competition law “relating to discounting restrictions in arrangements between hotels and online travel agents” (Office of Fair Trading, 2015). Lengthy consultation periods followed resulting in only minor amendments on the part of the accused parties30_{(González-Díaz & Bennett, 2015). Other European competition authorities followed suit in}

opening up investigations, showing quite some more vigor. The German Bundeskartellamt reached a decision concerning HRS in December 2013, arguing that its wide price parity clause restricted the price-setting freedom of hotels, acted to soften competition between platforms and foreclosed entry (Bundeskartellamt , 2013 (b), pp. 5-6) – a decision which was subsequently upheld on appeal by the Düsseldorf Higher Regional Court. Early 2014, the French, Italian and Swedish NCAs were appointed by the European Competition Network to take the joint lead in the investigations concerning the wide price parity clauses in the online hotel booking sector, with special attention for market leader Booking.com (Chappatte & Townley, 2015). Their concerns were the same as the Bundeskartellamt’s concerning HRS: that Booking.com’s wide price parity clauses restricted price competition and hindered new market entry of other hotel platforms (Konkurrensverket, 2015) (Autorità Garante della Concorrenza e del Mercato, 2015) (Autorité de la concurrence, 2015).

In response to the NCAs’ concerns, Booking.com offered to make commitments, narrowing the price parity clauses which has been effective up until then in scope. The result was a package of commitments which consisted of allowing hotel companies to price differentiate between hotel platforms and also to charge lower prices offline, provided they are not published online. Also, the availability parity was changed into a ‘minimum availability’ agreement, according to which hotel companies commit themselves to making a minimum number of rooms available (Bundeskartellamt, 2015 (a), p. 12). On 21 April 2015, these commitments were accepted by the French, Italian and Swedish competition authorities, who argued that the proposed narrow price parity clause leaves between-platform completion intact – benefiting customers – while reducing the risk that hotel companies freeride on investments made by Booking.com, thereby eluding legal action based on Article 101(1) of the TFEU. The commitments were implemented by

29_{Akman (2015, pp. 23-32) argues that legal treatment of hotel platforms’ price parity clauses leads to inconsistencies}

and that the correct legal instrument to use is Article 102 of the TFEU prohibiting the abuse of a dominant position. She bases her position on her finding that since hotel platforms are agents of hotel companies, there is no real question of an ‘agreement’ between undertakings. Also, she argues, if the anticompetitive practice would be that of an anticompetitive agreement, addressing of the decisions to only one of the parties (that is, the hotel platform) is inconsistent: “[This] suggests that the theory of harm is based on unilateral conduct despite legal action being based on a provision prohibiting multilateral conduct through agreements. (…) [I]t creates a mismatch between the authorities’ theory of harm and their action, and thereby raising questions as to whether they are operating on the basis of the correct theory of harm” (p. 4). The Bundeskartellamt, however, claims that “[d]eclarations of intent are also considered agreements if one company is geven freedom to unilaterally design the contractual relationship, as is the case with general terms and conditions that are binding on the other company” (2015 (a), p. 48). I will leave the question as to whether the NCAs have adopted the correct legal approach to be answered elsewhere.

30_{The accused parties being Booking.com and Expedia entering into price parity agreements with InterContinental}

(15)

14

Booking.com throughout Europe on the first of July 2015. Hotel companies responded skeptically, arguing that also narrow PPC are anticompetitive and that also with narrow PPCs in place, their entrepreneurial freedom would remain severely restricted (HOTREC, 2015 (a)). The Dutch representative of hotels and restaurants urged the Dutch NCA to take action against the still “suffocating” clauses (Koninklijke Horeca Nederland, 2015). 24 European NCAs, however, reacted with approval3132_{. The NCAs’ objections nor the}

commitments mentioned anything about Booking.com’s best price guarantee, which it continued to offer, as managing director Europe, Middle East and Africa Peter Verhoeven stated in a response to the implementation of the commitments: “We will keep offering the [best price] guarantee, even if we have to pay it from our own pockets” (Van Noort, 2016).

The storm had seemed to have passed for a while. However, in December 2015, the Bundeskartellamt issued a decision concerning Booking.com’s narrow price parity clauses, ordering the platform to completely delete the clauses from its contracts, terms and conditions (Bundeskartellamt, 2015 (c)). The Bundeskartellamt opined that the complete ban on price parity clauses in the HRS case supported this decision (Bundeskartellamt, 2015 (a), p. 38), even though the 2013 decision report on HRS’s PPCs made no comment regarding narrow PPCs (p. 50). The extensive decision report on Booking.com, however, did include a thorough, qualitative assessment of the effects of Booking.com’s narrow PPCs as compared to a hypothetical situation in which such clauses are entirely absent (p. 55), the conclusion of which was that also narrow PPCs effectively restrain competition and that exemption under 101(3) is out of the question (pp. 25-26). Booking.com has announced to appeal the Bundeskartellamt’s decision33_{, yet removed the}

narrow PPCs from its contracts with German hotel companies, pending the outcome of the appeal (PRNewswire, 2015). Also in France and Italy Booking.com has ran into renewed trouble, this time coming from state intervention. In July 2015 the French Assemblée Nationale adopted a law (‘la loi Macron’) which prohibits the use of any type of price parity clause between hotel companies and online hotel booking platforms (HOTREC, 2015 (b)). Also the Italian parliament adopted a similar law by a huge majority in October 2016 (HOTREC, 2016).

Patchwork of approaches

These divergent approaches have led to a patchwork of legal approaches throughout Europe. Although the coordination between the French, Italian and Swedish NCAs that were appointed by the European Competition Network and the working group open to all interested NCAs was unprecedented (Chappatte

31_{The Dutch Autoriteit Consument en Markt (2015) for example, stated that it “believes this [the narrowing of}

Booking.com’s PPC] will result in more options for hotel owners to vary their offerings on various booking sites. For consumers, this means they will have more options to choose from. As a result, competition between the various hotel booking sites is expected to increase.” And: “ACM so far has refrained from launching an investigation of its own, as it expected that this pan-European approach would eliminate risks in the Dutch market.”

32_{That is, besides the NCAs that had led the investigations, all NCAs}

33_{In a press statement, Booking.com’s chief executive officer Gillian Tans stated that “We believe this decision is}

flawed because it does not recognize the immense benefits that online travel brands like Booking.com bring to both consumers and accommodations. Companies like ours bring transparency, choice and value to global travellers by aggregating information for hundreds of thousands of properties. We do not only save consumers time and money, we serve as a highly cost-efficient marketing channel for most hotels that could not otherwise afford to market their brand to domestic and international consumers” (PRNewswire, 2015).

(16)

15

& Townley, 2015, p. 1), the decentralized approach that was consciously pursued by the ECN34_has

eventually led to Booking.com and hotel companies being subject to different arrangements regarding its price parity clauses in different member states of the EU, either because of an NCA decision or state intervention. This inconsistency in approaches is remarkable, as similar conditions apply almost everywhere Europe, with Booking.com being the leading online hotel booking platform with two significant competitors35_{, similar terms and conditions and customers and hotel companies having similar}

demands (p. 3).

This inconsistent application of European competition law is concerning, as it leads to unfair outcomes with hotel companies in different member states being subject to different clauses, which also have their effect on consumer prices. It also leads to undesirable reductions in legal and business certainty, which is detrimental for investments and most definitely does not contribute to the European Commission’s commitment to develop a single European digital economy (European Commission, 2016). As Commissioner Margarethe Vestager noted, the Booking.com cases have been a “learning experience” and a “very good example” of how greater coordination is needed (Vestager, 2015).

How this greater coordination or a principled approach is to be achieved is still an open question. In theory, the European Commission could take jurisdiction from any of the national completion authorities under Article 11(6) of Regulation 1/2003, which states that NCAs should be “empowered to apply Community Law” (EUR-Lex, 2002) and was formulated for the express purpose of preventing conflicting decision by member states in exceptional circumstances (Chappatte & Townley, 2015, p. 3). If the European Commission was to take over the investigations concerning Booking.com’s and other hotel platforms’ clauses, the addressees of the resulting Commission decision would have the right to appeal to the EU General Court, to amend or decline the decision (European Commission, 2013), thus also removing the diverging state interventions concerning price parity clauses in the online hotel booking sector.

The most pressing problem, however, is that a firm theoretical basis for greater analytical alignment regarding Booking.com’s and other hotel platforms’ practices is lacking. Most of the existing literature on which NCAs have based their decisions concerns wide PPCs – and where narrow PPCs are explicitly taken into consideration, this is done so in a qualitative way. Furthermore, none of the three NCAs that led the investigations concerning Booking.com have taken into account the fact that Booking.com offers its customers a best price guarantee – a price-matching agreement that may have severe anticompetitive effects, especially in combination with a price parity clause. Even the Bundeskartellamt treated Booking.com’s BPG as a side issue. It is these theoretical lacunas and the consequences thereof that will become apparent in the next section discussing the existing literature on online booking platforms, price parity clauses and best price guarantees.

34_{The argument being that NCAs are in a better position to judge the competitive effects of certain practices on their}

own national markets.

(17)

16

4 RELATED LITERATURE

The formal analysis of this thesis is closely related to Wang & Wright (2016), whose approach I will follow to a great extent before adding to their model. In their recent paper Search Platforms: Showrooming and Price Parity Clauses Wang & Wright develop a theory of search on online platforms of which Booking.com is a paradigmatic example. They use this theory to explore the welfare implications of price parity clauses, both narrow and wide. In order to do so, they list the key features of online marketplaces such as Booking.com, but also Amazon.com, other booking and reservation systems as being that 1) firms set prices on the platforms; 2) customers search for firms on the platforms and may complete purchases through it; 3) when customers do complete a purchase through the platform, firms pay a commission fee to the platform; and 4) in most of the markets in which these platforms operate, customers can also purchase from the firms directly, at potentially different prices (p. 2). This opens up the possibility of ‘showrooming’; customers searching on the platform and enjoying its non-transaction related benefits, but switching to a direct sales channel in order to complete a purchase if direct prices are lower since firms do not have to pay the platform fees on their own direct sales channel. This “showrooming” behavior, Wang & Wright (p. 3) write, is a specific form of freeriding on a platform’s services which might – under certain conditions – undermine a platform’s ability to operate. From the platform’s perspective, the solution is a price parity clause – either narrow or wide – which eliminates firms’ possibility to charge a lower price on their own sales channel.

4.1 A

THEORY OF SEARCH

In considering the effects of the existence of online platforms that facilitate search, Wang & Wright’s analysis is very similar to Yanis Bakos’ 1997 paper on the implication for electronic marketplaces of buyer search cost reduction36_{. In this paper, Yanis Bakos expressly focuses not on transaction-oriented benefits}

that platforms may provide if an actual purchase is made on the platform, but on the “market making functionality of electronic marketplaces during the process of price discovery and identification of appropriate product offerings in a market” (p. 1)37_{, with respect to differentiated markets. In doing so,}

Yanis Bakos discerns three effect of the reduction in search costs as facilitated by online platforms38_which

correspond one on one with those resulting from Wang & Wright’s analysis, albeit formulated in a somewhat way. First of all, the net reduction in search costs find their counterpart in Wang & Wright as a higher “incremental expected benefit of one more search” (p. 11). Second, the allocational efficiencies that customers enjoy from being better informed about available products are summarized as customers holding out for “a higher match value” (p. 12) between their preferences and a firm’s offer, which reflects the additional surplus they enjoy from being able to search at a lower cost on the platform. Wang & Wright call this the “surplus differential” of the platform. Third, the lower prices that customers enjoy because of increased competition among sellers are denoted as the “markup differential” by Wang & Wright (p. 14), which reflects the lower margin that firms obtain from selling through the platform compared with the direct market. A key difference between the approaches is that in considering the welfare effects of online

36_{Wang & Wright do not explicitly mention this paper; I’m drawing the parallel.} 37_{The terms ‘platform’ and ‘marketplace’ can be used interchangeably in this context.} 38_{I already discussed these in §3.3.}

(18)

17

platforms, Yanis Bakos purposely fails to consider the additional benefits that platforms provide in assessing their impact on welfare. Wang & Wright, however, do take these transaction-related benefits into account.

4.2 P

RICE PARITY CLAUSES AND THEIR EFFECTS

Besides offering a theory of platform search, Wang & Wright develop a game theoretical model and solve it through backward induction in order answer the question whether the use of price parity clauses may be legitimate from a welfare point of view. They consider both the case with a monopoly platform and with competing platforms, with either none, a wide or a narrow PPC in place.39_{They find that – for certain}

benefit-cost ratios – narrow PPCs may be good for consumers, provided that there is no monopoly platform, that the platforms’ viability depends on the PPC and that sufficient platform competition is in place. Their approach differs from most existing literature on most-favored nation clauses – which were summarized by LEAR in a report prepared for the British Office of Fair Trading – in the sense that they focus on third-party price relationship agreements (PRAs) which affect both competition between sellers (hotel companies in the context of this thesis) and platforms, while determining the price paid by the end buyer, who is not a part of the agreement. Hence, their approach also differs from the literature on resale price maintenance (RPM), since the PRAs they focus on do not determine absolute price levels, but only relative ones (Buccirossi et al. , 2012, p. 87).

Within the distinct realm of price parity clauses in two-sided online markets, further distinctions can be drawn. Most notably, most of the existing literature in this domain – e.g. Boik & Corts (2014), Edelman & Wright (2015) and Fletcher & Hviid (2014) – only studies the effects of wide price parity clauses with competing platforms, whereas Wang & Wright devote extensive attention to narrow PPCs and also study the effect of PPCs in case of a monopoly platform. Moreover, Wang & Wright leave room for consumers switching between sales channels, and do not – as the abovementioned seminal papers do – take customer participation as given. Also in contrast to these papers, Wang & Wright assume the pass-through of platform fees into customer prices to be 100 percent and customer demand to be inelastic up until the point where consumers don’t expect any positive surplus from searching on a platform (p. 5).

An example of a paper where narrow PPCs are discussed is Ezrachi’s recent paper which bears the fitting title The Competitive Effects of Parity Clauses on Online Commerce. His findings predominantly correspond to those of Wang & Wright, in the sense that he identifies the same benefits stemming from the existence of online booking platforms as well as largely the same effects of wide and narrow PPCs respectively – although he discusses them in a qualitative way. A notable difference though, is that Ezrachi identifies additional, seller-related beneficial effects of online booking platforms, such as online platforms’ function of risk and cost mitigation associated with online advertising and commerce (p. 494). Sellers lack the economies of scale to achieve a high return on investment and a high conversion rate40_{since especially}

the small and medium-sized sellers generate little traffic in absence of the platform. As further additional benefits of online platforms Ezrachi mentions the safe environment in which customers may make

39_{In the case of a monopoly platform, there is no de facto difference between a narrow and a wide PPC, since there}

are no other platforms on which firms can offer their product.

(19)

18

purchases and the potential for seller entry, as online platforms enable smaller sellers to attract customers (p. 495), which is beneficial for the sellers in question, and also further increases competitive pressure.

4.3 R

EVERSE FREERIDING

Another point of disagreement between Wang & Wright’s analysis and other literature on narrow price parity clauses is the so-called “hub-and-spoke” effect. Both Wang & Wright (p. 28) an Ezrachi (p. 507) claim that with narrow price parity clauses in place and more than one platform operating, platform fees and therefore final prices are restrained by between-platform competition, as now platforms are incentivized lower their fees, since this will in turn incentivize sellers to lower their prices on that platform41_{. However, both the Bundeskartellamt and MAPP}42_{have argued that also with narrow PPCs in}

place a de facto price floor exists for hotel rooms, since hotel companies are very reluctant in practice to offer a lower price on a booking platform that charges lower fees. This is because if a hotel company wanted to set a price on another booking platform that was lower than the price on Booking.com, it would have to accept the fact that the price on its own website would have to be higher than on this other, cheaper hotel platform, since a narrow PPC still forces hotel companies to charge on their own websites a price no lower than on Booking.com (Bundeskartellamt, 2015 (a), pp. 57-58). This makes their own websites significantly less attractive from the perspective of customers, diverting online booking behavior towards hotel platforms – a situation which has been denoted as “reverse freeriding” (Bourguignon, 2015, p. 15) and which runs contrary to hotels’ economic interests as it makes them even more dependent on booking platforms through which their flexibility in changing prices and terms in response to especially strong or weak demand is very limited. Hence, as the Bundeskartellamt contends, “it must be expected that the hotel companies would be very reluctant in practice to make use of the possibility of differentiated pricing behavior” (2015 (a), p. 58) and therefore “the freedoms that the hotel companies were formally granted through the narrow best price clauses applicable since July 2015 in terms of price differentiation with different hotel portals are largely ineffective in practice” (p. 57). Accordingly, with Booking.com’s narrow PPC in place, there is hardly any serious incentive for hotel platforms to lower their commissions, as this will not result in lower hotel room prices being offered on their platform (p. 65).

4.4 B

EST PRICE GUARANTEES

Although I will follow Wang & Wright’s analysis in abstracting from the reverse freeriding effect and seller-related beneficial effects of online booking platforms, my analysis fundamentally differs from theirs as I will take into account that Booking.com (and its major competitors) offer a best price guarantee to customers. As will be seen, the presence of a BPG fundamentally alters the welfare effects of the adoption of narrow PPCs and is thus a fundamental feature of the way in which Booking.com operates and cannot be disregarded.

Brief mention of the relevance of BPGs is made in the Bundeskartellamt’s decision report on Booking.com (p. 66). However, it is treated as an ancillary factor, which merely enhances the anti-competitive effects of

41_{Remember that under a narrow PPC, sellers are in fact allowed to price discriminate between platforms.}

42_{MAPP (“Microéconomique Appliquée”) served as an economic consulting expert on behalf of the French NCA, the}

(20)

19

narrow PPCs – even though literature suggests that BPGs have severe anticompetitive consequences of their own. In neither the Bundeskartellamt’s decision, however, nor in any of the other NCAs’ decisions, it was suggested or impelled that BPGs should be banned. This serious anticompetitive potential of BPGs is in fact recognized by Akman (2015), who in her legal assessment of price parity clauses states that “at best, the recent intervention of the NCAs in this market [for online hotel booking services, FW] will have been without effect since they have not addressed the more disconcerting practice [of offering BPGs, FW] in their enforcement. At worst, they may have moved the industry closer to a potentially more anticompetitive equilibrium (…)” (p. 6). Nevertheless, Akman doesn’t give a formal analysis of the potential effect of BPGs whether or not in combination with narrow PPCs.

The academic literature on BPGs, the different forms they take on and their respective effects is extensive. Salop (1986) was a pioneer in discussing BPGs, though in a qualitative way. In his seminal paper, he presents his intuition that price-matching43_{– “meeting competition clauses” as he denotes them – by all}

competitors in a market leads to higher prices as firms do not have an incentive to undercut each other, since price cuts will be matched anyway. Hence, his argument goes, if all firms adopt BPGs and set the joint-profit maximizing price, no firm has an incentive to deviate (pp. 280-281). Following this same line of reasoning, Salop argues that BPGs can also serve as an entry-deterring mechanism, since given the incumbent’s threat of reducing its price in the event of entry by a lower-cost firm, “no actual entry would occur, because the entrant would anticipate earning insufficient profit at the lower post-entry price” (p. 282).

The intuition that BPGs lead to supra-marginal pricing was formalized by Belton (1987) and Corts (1995), who show that price-matching leads to monopoly prices. That is, in an equilibrium in which customers are fully informed, firms simultaneously set prices and choose whether to adopt a price-matching policy, the unique Pareto-efficient Nash equilibrium outcome is when all firms set the joint-profit maximizing price and adopt a price-matching policy. He also show that price-beating policies, in contrast, restore the competitive outcome, since now firms can profitably undercut each other and capture the entire market, a process which continues until the competitive outcome is reached.

Others have shown that BPGs have different anti-competitive effects under different assumptions. For example Corts (1996) argues that if one considers a market with both informed and uninformed (about the existence of some best price guarantee) customers, BPGs are better viewed as means of price discrimination, instead of a mechanism to alter competitors’ incentives. This is so, since only informed customers will claim the refund promised to them by the guarantee, while (a share of) the uninformed customers will purchase the offered product or service at a supra-competitive price. Corts shows that the informed customers are always better off under these price-matching policies. Yet other authors – e.g. Edlin (1997), Liu (2011) and Winter (2008) have considered BPGs as a facilitating practice, as BPGs allow deviations from a (tacit) collusive agreement to be quickly identified as customers signal them to the relevant firm, which removes the need for time and cost-intensive monitoring. Hence, they enable quick punishments, as a deviating price is immediately matched.

(21)

20

The entry-foreclosing potential of price-matching guarantees as described by Salop was challenged by Arbatskaya (2001), who claims that only price-beating guarantees are capable of deterring entry. The reasoning behind this claim is that only if an incumbent announces a price-beating guarantee on both effective and advertised prices44_{, an entrant cannot use any strategy to undercut the incumbent’s effective}

price. The entrant thus cannot hope to obtain any market share, and will therefore not enter. However, she argues, under special circumstances also a price-matching guarantee is capable of deterring entry: that is, if the product of the incumbent is of superior quality45_{. It under this assumption that a price-matching}

guarantee works in much the same way as a price-beating guarantee, since matching the price of an entrant who offers a lower-quality product will still result in zero demand for the entrant’s product. In constructing a model for assessing the welfare effects of a narrow PPC combined with a BPG, I will assume an incumbent online (hotel) booking platform that announces a price-matching guarantee on advertised prices and that offers superior transaction-related benefits. I will follow Belton (1987, p. 401) in assuming that consumer demand is unaffected by the mere presence of BPGs, which may not be entirely realistic, since the BPG may induce customers into believing that the firm’s posted price is in fact the lowest price. As I assume customers to be fully informed, and prices to be completely transparent, I abstract from the price-discriminating and collusion-facilitating effects of BPGs as identified by Corts (1996) and Edlin (1997) respectively. Since on the basis of these assumptions, a BPG can be part of both an entry-deterring strategy as well as a vehicle for softening-competition once entry has already occurred, I will argue that not having included a BPG in the analysis of the effects of Booking.com’s price relationship agreements is a fundamental deficiency of all models that claim to analyze the welfare effects of Booking.com’s price relationship agreements.

5 THE MODEL

Even though a lot has been written about price parity clauses, most of the literature focuses on wide price parity clauses, on the anticompetitive effects of which the lion’s share of analyses agree. However, in the present context it is the narrow price parity clauses that are of interest, as these raise discord both from an academic and a competition policy perspective. However, most studies of the effects of narrow price parity clauses are of a qualitative nature, and where they are quantitatively studied, the model in question fails to incorporate a key feature of Booking.com’s (and its competitors’) operation: the fact that it offers a best price guarantee. In this section, I will extend an existing model – Wang & Wright’s – to include the possibility of offering a best price guarantee, but not before I extensively review Wang & Wright’s model46_.

44_{The effective price being the outcome of a recursive application of low-price guarantees (Arbatskaya, 2001, p.}

1395)

45_{In Salop’s analysis, firms are not vertically differentiated.}

46_{That is, I follow Wang & Wright in their assumptions, timing and equilibrium selection. I take over the optimal}

search strategy they develop, as well as their analysis of different cases. I elaborate on their proofs in case I deem their steps are taken too quickly with my own analysis. The originality in my extension lies in my analysis and selection of the relevant cases as developed in Wang & Wright in case a BPG is imposed, as well as in the computation of profits and welfare.