UCWG Vertical Restraints Project

Unilateral Conduct Working Group
Vertical Restraints Multi Year Project

2016-2019
April 2019
1
Overview – the vertical restraints project
1. In 2016 the Unilateral Conduct Working Group (UCWG) began a multi-year project
examining the effects on competition of vertical restraints. The project sought to promote
increased understanding, and where differences exist, to work towards convergence.
2. The vertical restraints project has three distinct stages
 Scoping phase: to help guide the development of the vertical restraints project.
The scoping work identified:
 an interest in a focus on online vertical restraints, but caution to not overlook

matters arising in offline markets
 support for theoretical analysis of issues arising with vertical restraints,

including theories of harm, market failures, and efficiencies
 support for analysis that focuses on particular forms of restraint, and
 support for the development of a selected case study resource.
 Development phase: design of hypothetical vertical restraint scenarios (parity

clause, and bans on online platform sales and online search advertising). The
scenarios were examined for their possible effect on competition and potential
resulting efficiencies.
A project group of ICN members and NGAs provided individual/jurisdictional-

specific responses for each of the hypothetical scenarios. The responses
provided by the project group members have allowed UCWG to develop reports
highlighting commonalities and divergent factors in the assessment of the vertical
restraints scenarios.
A summary of the hypotheticals is provided below.
 Completion phase: to be carried out in 2019-2020, the completion phase will

consider how the work products can be effectively implemented. The third phase
of the project is to be determined with the UCWG membership and may include:
 webinar discussions of the key points that emerged in the hypotheticals;
 adaption of the hypotheticals into training modules to be made available

through ICN Training on Demand; and
 developing a workbook chapter on vertical restraints.
Hypothetical One – online parity requirements
3. The first vertical restraint hypothetical considered the possible effect of online parity
requirements included in contractual arrangements between fictional Online Travel
Agents (OTAs) and fictional accommodation providers. The full report of this scenario
is at Attachment A to this paper.
Theories of harm – conduct of BestValueBed

4. It was alleged that BestValueBed, an OTA with a market share of 25%, had included in
its contracts with accommodation providers a ‘narrow’ room rate parity requirement. The
contract required accommodation providers to offer price parity between listings on
BestValueBed’s platform and any other online distribution channel controlled by the
Page 2
accommodation provider. It was also alleged that BestValueBed had required
accommodation providers to offer room inventory parity between listings on
BestValueBed’s platform and other online distribution channels used by the
accommodation provider (a ‘wide’ room inventory parity requirement).
5. The project group considered that parity requirements can have a detrimental impact on
competition. It was noted that parity requirements make selective discounting more
expensive and thereby reduce the frequency of such discounting (as such discounts
must, at a minimum, also be offered to the platform benefiting from the parity
requirement).
6. Members of the project group held different views on the potential competitive impact of
the parity requirements on accommodation providers.
 View 1: limited or no detrimental impact on competition
Narrow room rate parity requirements
The narrow price parity clause does not restrict price and quality competition among
accommodation providers – as it still allows accommodation providers to make
differentiated offers across non-proprietary distribution channels.
 View 2: detrimental impact on competition
The narrow parity requirements soften competition between accommodation
providers if they act as a disincentive for accommodation providers to differentiate
across distribution channels (e.g. by creating a price floor across some or all
distribution channels used by the accommodation provider).
It was considered that narrow parity requirements may not only restrict competition
between accommodation providers but may also restrict competition between OTAs.
Wide room inventory parity requirements
It was noted that further information/evidence would be required to establish the
existence of, terms, and potential effect of any wide room inventory parity
requirement.
It was recognised that wide parity clauses affecting price and non-price offerings can
soften competition between suppliers by requiring price and inclusion alignment
across all distribution channels (giving rise to a horizontal effect). A wide rate parity
requirement creates a ‘price floor’ below which an accommodation provider cannot
offer its rooms on a different distribution channel without breaching the parity
requirement, leading to price uniformity. The alleged wide inventory parity
requirement restricts price and non-price competition on room features and
inclusions.
7. Members of the project group also held different views on the potential competitive
impact of the parity requirements on distribution channels.
It was noted that the narrow parity requirement is a less intrusive requirement than
that of a wide parity requirement and does not affect the relationships between an
accommodation provider and other OTA platforms. The narrow price parity clause
would not restrict price and quality competition between non-proprietary distribution
channels.
Page 3
It was noted that narrow parity requirements applied by OTAs may restrict
competition between OTAs in cases where they produce equivalent effects to wide
parity requirements (see below). Since the narrow parity requirement prohibits the
accommodation provider from offering better room prices on its own website than on
the OTA which imposes the requirement, if the accommodation provider wishes its
own online offer to match a lower price offered on another OTA portal, it will be
obliged to reduce the price on the first OTA portal as well. Depending on the
particular facts of the relevant market, in particular the share of sales conducted
through direct and indirect channels, this price floor effect may reduce the incentive
for OTAs to compete on commission rates, by reducing the incentive for
accommodation providers to make use of any lower commission rates offered by
competing OTAs.
In addition, if accommodation providers and OTAs participate in the same market
then the narrow parity requirements would soften competition between the OTA and
the distribution channels controlled by the accommodation provider. It was
considered that by restricting an accommodation provider from growing direct
bookings through more favourable room rate and/or inventory offers, a narrow
clause would soften the competitive constraint that the accommodation provider’s
distribution channels may otherwise have imposed on non-proprietary distribution
channels.
It was noted that wide parity requirements can soften and may foreclose price and
non-price competition between distribution channels. Differentiated offers by
accommodation providers create incentives for distribution channels to compete for
their custom, for example by offering lower commission rates or other competitive
benefits as reward for better room rates or inventory being listed with their platform.
By requiring accommodation providers to offer them their best inventory the OTA
(BestValueBed) diminishes the competitive value of what may be offered to
consumers through other online distribution channels. This diminishes the incentives
of competing OTAs to offer lower commission rates/benefits to accommodation
providers and may reduce the competitive tension that would prevent BestValueBed
from introducing excessive commission rates. It was also noted that the parity
requirement may affect the take up of services for smaller or more marginal OTAs
and have a negative impact on their ability to increase scale or market share through
listing more competitive offers. A loss of competitive tension could result in a loss of,
or a reduction in, innovation in OTA services.
Arguments about pro-competitive effects that may be presented and factors relevant to
these arguments
8. Project group members noted that, in general, their jurisdiction does have regard to pro-
competitive effects, although there are differences in the manner in which such effects
can be taken into account across these jurisdictions.
 Parity requirements may prevent free-riding, by competing OTAs and

accommodation providers respectively, on the investments made by BestValueBeds.
OTAs promote inter-brand competition between accommodation providers through
increased price transparency. OTA platforms may also reduce consumer search
costs. The parity requirements may support an OTA’s business model by increasing
conversion rates (the ‘look to book’ ratio) by ensuring that the room rates and
inventory offered through their platform is superior.
Page 4
 The project group noted that further consideration would need to be given to whether
the parity requirement (wide or narrow) is reasonably necessary or indispensable to
obtain the benefits provided by BestValueBeds.
Areas of further inquiry by investigators
9. The project group identified a number of areas for further inquiry by investigators.
 Parity clauses: a range of matters including evidence of the scope, market

coverage, evidence of enforcement of the clauses, use of alternative business
models by OTAs.
 Market definition and other considerations: a range of matters including evidence
of consumer search and booking behaviour, evidence of the effect of the parity
requirements on OTAs, evidence of the effect of the parity requirements on online
distribution channels controlled by the accommodation provider.
 Other: a range of matters including features of hotel websites relative to OTA

websites.
Hypothetical Two –online selective distribution models
10. The second hypothetical scenario considered the potential impact of certain online sales
restrictions on selective distribution models of three fictional suppliers of infant strollers:
Babydream (20%), Wheelies (>30%) and Sport 2001 (>30%). The retail landscape is
populated by five large brick-and-mortar retailers, two large third-party platforms and
other small online/offline retailers. Online sales have gradually increased (from 20% in
2015 to 35% in 2017) together with online sales via third-party platforms (20% of the total
online sales in 2015 to 40% in 2017).
11. The hypothetical scenario considered two restrictions: a ban on selling through online
third-party platforms and a wide online advertising ban (incl. an online search advertising
ban). Similarly to hypothetical one, the project group has been invited to consider
possible theories of harm, arguments about pro-competitive effects and areas of further
inquiry. The full report on this hypothetical scenario can be found in Attachment B to this
report.
12. Some members did not envisage any competition concerns in relation to this hypothetical
scenario since in all distribution channels there are at least three competitors and the
retail landscape is populated by several players. In addition, by focusing on inter-brand
competition, these members noted that the restrictions imposed by the three
manufacturers might be reasonably necessary to maximize efficiency across all
distribution channels.
13. According to some other members, the restrictions in question can have detrimental
impact on consumer welfare by: (i) reducing intra-brand competition through the
protection of existing sales channels from competitive pressure of online channels; (ii)
softening inter-brand rivalry when considering the cumulative effects of above restrictions
under certain conditions (e.g., if third-party platforms are an important distribution
channel for the three manufacturers and their retailers); (iii) finally, in the long run by
foreclosing potential entrants in the downstream market for retail sale of baby strollers.
Babydream’s ban on selling to third-party platforms and wide ban on online advertising
14. Babydream prohibits all sales on third-party platforms and has a wide online advertising
ban, including on retailers’ own website and through the use of price comparison
websites; in particular Babydream retailers are prohibited from using search advertising
Page 5
(i.e., from using/bidding on search ads for Babydream brand name). Members of the
project group held different views on the potential competitive impact of the Babydream’s
restrictions.
Some members noted that Babydream’s two restrictions are unlikely to have material
consequences on competition and the viability of third-party platforms because
Babysdream’s market share is 20%. It was observed that, even when assuming that
Babydream’s restrictions may undermine the viability of third-party platforms, such
scenario is unlikely to materially harm inter-brand competition due to limited market
share of the third-party platforms channel (14% of total sales).
Some members carried out a two-step assessment which focused, as a first step, on
evaluating the proportionality and necessity of Babydream’s two restrictions within the
selective distribution system of Babydream, and, in a second step, on the analysis of
the potential restriction of competition. With respect to online sales restriction
through platforms, these members noted that such restriction serves a legitimate
aim such as maintaining Babydream’s brand image for high quality products and
protecting investments by its retailers in pre- and post-sale services. Also, the
restriction appears to be applied in a uniform manner and Babydream itself is not
selling to third-party platforms. For these reasons, the second step of the
assessment, that is, an analysis of the potential restriction of competition, was not
considered necessary to carry out by these members. One member, however, held a
different view stating that the online sales platform ban was not justified in light of the
non-luxury nature of the product (brand image and reputation could be achieved with
less restrictive means) and the growing relevance of online platforms as gateways to
end customers.
In relation to the wide online advertising ban, all members of the project group
carrying out an assessment on the necessity and proportionality of the restriction
noted that such advertising ban does not appear to pursue any legitimate objective of
promoting brand image or investing in safety demonstrations. On the contrary, it
appears to reduce the ability of Babydream’s authorized retailers to compete with
Babydream’s own online channel and avoid an increase in Babydream’s bidding
costs in the case of online search advertising. It would therefore likely amount to a
restriction of competition, on its own or in conjunction with the online sales platform
ban.
For some other members, an effects-based approach is needed to assess whether

the impact of Babydream’s two restrictions is detrimental on retail competition, on the
basis of one or more theories of harm indicated in paragraph 13 above.
Wheelies “equivalent” criteria required for selling through third-party platforms
15. Wheelies (>30%) permits authorised retailers to sell Wheelies’ products on third-party
platforms, provided that the platform meets the same qualitative criteria as the retailer or
equivalent ones.
 View: no impact on competition
All members of the project group considered that such a restriction is unlikely to raise
competition concerns and appears to be justifiable in light of the legitimate objectives
of protecting brand image and investments in pre- and post-sale services. In addition,
there is no indication that these criteria are applied in a discriminatory fashion or that
they are impossible to be met. In the latter event, according to one member, Wheelies
Page 6
effectively imposes the same restriction on sales on online platforms as the one
imposed by Babydream.
Sport 2001 ban on selling high-end products through third-party platforms
16. Sport 2001 retailers are allowed to sell only some of Sport 2001's models (the lower-end
models) on third-party platforms, with the exclusion of other models (e.g. the high-end
models). Members of the project group held different views on the potential competitive
impact of the Sport 2001 restrictions.
 View 1: no impact on competition
According to the members of the project group who did not envisage any competition
concerns generally (see paragraph 12), Sport 2001 ban is much less restrictive than
those of either Wheelies and Babydream and therefore they would not be
investigated.
Members of the project group assessing the necessity and the proportionality of the
restriction within the selective distribution system of Sport 2001 considered that such
restriction would not serve any legitimate objective such as the protection of brand
image. Relevant to this assessment were the fact that this restriction is not applied
uniformly since Sport 2001sells all its product ranges directly to third-party platforms
while forbidding the sales by its retailers of high-end strollers. In addition, less
restrictive options are available (like in Wheelies). Therefore such restriction has the
potential to reduce intra-brand competitive pressure for high-end Sport 2001
products, by limiting the ability of resellers to sell such products to customers,
especially if third-party platforms are an important sales channel.
For some other members, an effects-based approach is needed to assess whether

the impact of Sport 2001 restrictions is detrimental on retail competition, as indicated
in paragraph 13 above.
Arguments about pro-competitive effects that may be presented and factors relevant to
these arguments
17. Members of the project group stated that the above restrictions can contribute to the
promotion of competition on other factors than price:
 Brand image and brand positioning across the various distribution channels.
 Quality of pre- and post-sale services (e.g., safety demonstrations).
 Other factors such as the reliability of the platforms for delivery and payment.
Areas of further inquiry by investigators
18. The project group identified a number of areas for further inquiry by investigators
 Analysis of intra-brand competition and in particular the competitive constraints
posed by the different distribution channels: e.g., whether retail competition from
offline sources or non-platform online sources are good substitutes for retailing
through platforms.
 Inter-brand competition analysis: substitutability of Babydream baby strollers with
rival brands sold through platforms or through any other means, from a consumer
perspective.
 Analysis of the intent especially in the case of Babydream and the business
justifications.
Page 7
 The cumulative effects of the restrictions applied by the three manufacturers.
19. In conclusion, the hypothetical scenario has raised issues related to (i) the importance of
intra-brand competition versus inter-brand competition in the assessment of the
consumer welfare effects and (ii) the trade-off between price versus non-price
competition. Some jurisdictions focused their assessment on inter-brand competition and
therefore restrictions on online sales can be problematic only if there is market power at
the manufacturing level. Other jurisdictions also analysed intra-brand competition to
ascertain whether online sales restrictions at the distribution level affecting retailers
selling products of the same brand can raise competition concerns.
20. The scenario showed the interplay between price and non-price competition in the online
environment. Price competition is often the key parameter for (online) retailers, while
manufacturers are more concerned with other factors, such as brand image, quality and
innovation, which are potentially undermined by the free-riding problem. As a result,
manufacturers may consider (some) online sales restrictions as necessary to better
control for these other factors and preserve investments in high-level presale services,
promotion and advertising. Therefore, online sales may improve price competition while
reducing non-price competition at retail level (absent manufacturers restrictions) and the
former effect will not always be more important than the latter, both for firms and
consumers.
Page 8
Attachment A
Vertical Restraints Project
Hypothetical One – Online Parity Requirement
Introduction
1. A small group of volunteers from competition authorities and NGAs1 (the project group)
has considered a limited fact scenario addressing an online parity requirement (set out
in Attachment OPR-A to this report).
2. This report provides an outline of matters considered relevant by the project group in
assessing the hypothetical conduct. The report seeks to highlight approaches to
assessing such conduct, including areas of divergence.
3. Project group members have noted that significantly more information would be
required before a competition authority would reach a concluded view to prohibit or
otherwise take action against the conduct set out in the hypothetical scenario.
4. The project group has been invited to consider
1. possible theories of harm
2. arguments about pro-competitive effects that may be presented and factors

relevant to these arguments, and
3. areas of further inquiry by investigators.
5. The hypothetical scenario provided to project group members has four key actors
 BestValueBed (BVB) – an online travel agent (OTA), 25 per cent of online

accommodation bookings in this jurisdiction are made via its platform. BVB
has implemented a ‘narrow’ contractual parity requirement affecting online
room rates and is alleged to have implemented a ‘wide’ parity requirement
affecting online inventory. The parity requirements of BVB are set out at
paragraph 14.
 Find-A-Room (FAR) – a competing OTA, 40 per cent of online

accommodation bookings in this jurisdiction are made via its platform. FAR
has implemented a ‘narrow’ contractual parity requirement affecting online
and offline room rates and inventory. The parity requirements of FAR are set
out in footnote 3 to this paper.
The parity requirements of FAR are not the focus of this document.
 Online Hotel Beds (OHB) - a recently launched OTA platform, 10 per cent of
online accommodation bookings in this jurisdiction are made via its platform.
1
Please see Attachment OPR-B.
Page 9
Attachment A
 Accommodation providers – in the hypothetical scenario one accommodation

provider (Hoolten Hotels) has provided information to the competition
authority.
6. This document does not seek to identify the views of contributing members of the
project group. This report has benefited from a diversity of views. Not all members of
the project group agree that all of the factors and considerations are appropriate;
project group members may evaluate them in different ways.
Overview of online parity requirements

7. Parity requirements like those set out in the hypothetical scenario have been
characterised as ‘retail most favoured nation’ clauses (retail-MFN). A requirement that
a supplier ensure parity across different competing platforms is characterised as a
‘wide’ retail MFN. Where the parity requirement only applies to the direct sales
channels controlled by the supplier it is characterised as a ‘narrow’ retail-MFN. Retail-
MFNs can be distinguished from ‘wholesale-MFNs’. A wholesale-MFN obliges a
supplier to offer a wholesale price to a retailer that is no greater than the wholesale
price offered to any other retailer. Under a wholesale-MFN a retailer independently
sets the retail price.
8. Online markets have led to both an increase in price transparency and a reduction in
switching costs. It was noted by some members of the project group that retail-MFNs
appear to have become more common as platforms seek to offer a ‘best price
guarantee’ to consumers.
Legal frameworks
9. Different legal frameworks apply across ICN member jurisdictions, these frameworks
generally apply a form of effects test when assessing online parity provisions. An
effects-based approach requires analysis of the potential effects of a particular
behaviour. Attachment OPR-C provides a short non-exhaustive summary of
legislative frameworks applying in some ICN member jurisdictions.
Relevant market
10. Responses received from the project group noted that further information would be
required in order to formulate the relevant market definition (see Areas of further
inquiry, below). For the purpose of preliminary considerations the project group noted
that the relevant areas of competition may include:
 A market for the provision of online travel agency services to accommodation

providers and consumers located within the hypothetical jurisdiction. As with
other platform distribution channels, online travel agency services has
customer features consistent with a two sided market, linking:
 accommodation providers who wish to attract customers to their

properties
 consumers seeking to book accommodation.
Page 10
Attachment A
 The market for the provision of accommodation booking services, including by
providers of ‘traditional’ travel agency services and offline channels.
 Local markets in which accommodation providers supply various types of

accommodation to consumers.
11. The views of the project group differed on the question of whether accommodation
providers and OTAs were likely to participate in the same market. These differing
views can be summarised as:
 Separate markets: some members of the project group considered that

accommodation providers and OTAs offer different distribution services, with
the OTAs’ services being broader (search, comparison, and range of
properties on offer). It was noted that the differences in distribution services
offered by OTAs and accommodation providers impacted upon their likely
substitutability for consumers.
Some members also addressed meta search providers, noting that the more
limited services provided by these parties would place them outside of the
market in which OTAs participate.
It was also noted that in some jurisdictions a distinction is drawn between

online retail, in which the website sells its own products, and online
marketplaces that act as portals through which buyers and sellers may
interact.
On this basis some project group members considered that it was less likely
that OTAs and accommodation providers would operate in the same market. It
was noted that, for the purpose of analysing the parity requirements,
accommodation providers and the OTAs are in a vertical relationship which
could give rise to anti-competitive horizontal effects.
 Same market: other members of the project group noted that the websites of
accommodation providers may act as an important constraint on a
concentrated OTA sector – but that the strength of their demand-side
substitutability would need to be tested with consumers in considering
whether a broader market definition would be appropriate.
It was noted that further information from the OTAs, accommodation providers and
consumers would be required to resolve these questions. It was also noted that
online services are dynamic and present a range of challenges in determining the
appropriate market definition.
12. It is noted that market definition and consequently BVB’s share of that market are of
particular relevance when considering the application of safe harbours and legal
presumptions. Not all jurisdictions apply safe harbours and/or legal presumptions when
analysing the effect of conduct on the competitive process. To allow for a discussion of
possible anticompetitive effect and procompetitive benefit the project group has not
sought to apply safe harbours or presumptions of illegality in preparing this document.
Page 11
Attachment A
Theories of harm – conduct of BestValueBed

13. It was recognised by the project group that parity requirements can have a detrimental
impact on competition.
14. Project group members noted the information provided by market participants which
suggested that BVB has implemented a narrow room rate parity requirement and may
have implemented a wide room inventory parity requirement.
a) BVB has required accommodation providers to offer room rate parity between
listings on its platform and other online distribution channels controlled by the
accommodation provider (provided for by contractual arrangements).
b) An accommodation provider has alleged that BVB is requiring accommodation

providers to offer room inventory parity between listings on the BVB platform and
other online distribution channels (conduct falling outside of the contractual
arrangements between the parties (non-contractual))2.
It was noted that further evidence should be sought to establish whether a non-
contractual wide room inventory parity requirement is being enforced by BVB. For
the purposes of this report theories of harm relating to a wide inventory requirement
have been identified. These are presented below, taking in turn the effect of the
conduct on accommodation providers and then the effect on distribution channels.
Effect on accommodation providers

15. As noted the views of the project group differed on the question of whether
accommodation providers and OTAs were likely to participate in the same market. It
was noted that agreements between entities in a vertical relationship can give rise to
anti-competitive horizontal effects, such that the effect of the parity requirements on
accommodation providers would be considered regardless of whether a ‘same market’
or ‘separate market’ analysis was conducted.
16. Some members of the project group noted that the parity requirement would make
selective discounting by an accommodation provider more expensive and thereby
reduce the frequency of such discounting (as such discounts must, at a minimum, also
be offered to the platform benefiting from the parity requirement). A number of factors
were considered relevant in assessing the potential impact of a parity requirement: (i)
the volume of sales affected by the requirement, (ii) the difference between the
prevailing price and the potential discount, and (iii) the probability the parity clause will
be enforced.
2
Hoolten Hotel has alleged that BVB is seeking to achieve pricing and inventory parity across all platforms.
 In March 2016 Hoolten offered a “Family Staycation” promotion – offering rooms with free upgrades to consumers
who purchased packages through the FAR platform.
Hoolten advises that it had negotiated an advantageous commission rate and search listing with FAR as part of its
participation in the promotion on the FAR platform.
 The Family Staycation promotion was also available through Hoolten’s offline channels.
 Three days after the campaign commenced, Hoolten received an email from BVB which advised it had seen
Hoolten’s inventory package on the FAR platform and requested that Hoolten immediately offer the same inventory
on the BVB platform.
 Hoolten declined this request and claims that almost immediately it noticed a significant drop in the number of daily
bookings received through the BVB platform. Hoolten has claimed that this was the result of BVB ‘dimming’ its’
listing on the BVB platform – dropping the listing from the first page of search results to the sixth page of search
results. Hoolten claims that its bookings through BVB dropped around 40% during the promotional period but
returned to normal levels after the promotion ended, and the hotel returned to the first page of BVB’s search results.
Page 12
Attachment A
17. It was noted that if a parity requirement covers a large portion of the market and is
enforced, it generally will be more costly for the accommodation provider to discount.
If, however, the parity requirement covers a smaller portion of the market it will have a
smaller impact upon competition.
18. Members of the project group held different views on the potential competitive impact
of the parity requirements on accommodation providers. These views are summarised
below.
View 1: limited or no detrimental impact on competition

19. Some members of the project group noted that the narrow price parity clause would
not restrict price and quality competition among accommodation providers – as it still
allows accommodation providers to make differentiated offers across non-proprietary
distribution channels. In particular it was noted that the contractual parity requirements
form part of a vertical agreement, i.e. an agreement between two parties operating, for
the purposes of the agreement, on different levels of the supply chain. Independent
businesses active on different levels of the same distribution chain in general have a
common interest in increasing sales of the relevant products or services. It was
recognised that vertical agreements between businesses that are not competitors can
be pro-competitive by solving inefficiencies in the vertical distribution chain. Some
members of the project group considered the impact of the narrow parity requirement
on interbrand competition, noting that while the requirement affects intrabrand
competition, interbrand competition is maintained.

20. In discussing the conduct described at paragraph 14(b) (above) some project group
members noted that there were a range of commercial factors which may have
resulted in the reduced ranking of the accommodation provider on BVB’s platform –
including lower than normal sales on the BVB platform as consumers take up the
superior offer being made on FAR’s platform. It was noted that OTAs apply a range of
factors in determining search rankings on their sites – including volume of sales. It was
noted that further information would be required to test whether the alleged conduct is
widespread or otherwise represents a wide parity requirement.
View 2: detrimental impact on competition

21. Some members of the project group considered that narrow parity requirements would
soften competition between accommodation providers if they acted as a disincentive
for accommodation providers to differentiate across distribution channels (e.g. by
creating a price floor across some or all distribution channels used by the
accommodation provider). The potential for independently applied narrow price parity
requirements to have a cumulative effect in softening competition between
accommodation providers was also noted (the narrow price parity requirement of FAR
is set out below3). In this context it was noted that where narrow parity requirements
3 FAR’s contract with accommodation providers has the following parity requirements:
 accommodation providers must give equal or better rates to FAR as made available through all distribution
channels controlled by the accommodation provider (e.g. provider controlled websites, telephone inquiries,
loyalty program offerings, walk-ins).
Page 13
Attachment A
applied by OTAs produce equivalent effects the theories of harm more often
associated with wide parity requirements may arise. It was considered that narrow
parity requirements may not only restrict competition between accommodation
providers but may also restrict competition between OTAs. Since the accommodation
providers are still prohibited from offering better room prices on their own website than
on the OTA which imposes the narrow parity requirement on them, if they wish their
own online offer to match a lower price offered on another OTA portal, they will be
obliged to lower the price on the first portal as well. Accordingly, the incentive for
competition on low commission rates between OTAs may be lowered since the OTAs
are aware that the price floor created by the narrow parity requirements reduces the
incentive for accommodation providers to make use of a lower level of commissions
that any competing OTA could offer. It was noted that the question of equivalent effect
is of growing importance and that further information/evidence including empirical
would be required to assess the combined effect of the narrow parity clauses on
competition between accommodation providers (see Areas of further inquiry, below).
22. It was also noted that, if accommodation providers were found to participate in the
same relevant market for distribution services as OTAs, narrow price parity
requirements would be likely to soften the competitive constraint on OTAs that may
otherwise be offered by the accommodation provider’s website. As discussed at
paragraph 11 above, this issue of the definition of the relevant market was not settled
between members of the project group.

23. Greater concerns were identified in discussing the conduct summarised at paragraph
14(b) above. It was recognised that wide parity clauses affecting price and non-price
offerings can soften competition between suppliers by requiring price and inclusion
alignment across all distribution channels (giving rise to a horizontal effect). A wide
rate parity requirement creates a ‘price floor’ below which an accommodation provider
cannot offer its rooms on a different distribution channel without breaching the parity
requirement, leading to price uniformity. A wide inventory parity requirement such as
that alleged in paragraph 14(b) restricts price and non-price competition on room
features and inclusions. In assessing the nature and extent of this restriction on
competition the factors noted at paragraph 16 above would require modification to
allow proper consideration of the non-price restriction. More generally it was noted
that further information/evidence would be required to establish the existence of, terms
and potential effect of any non-contractual wide parity requirement that may have been
introduced by BVB (see Areas of further inquiry, below).
 accommodation providers must provide inventory equivalency, including packages, to FAR as made available
through online distribution channels controlled by the accommodation provider.
The project group noted that FAR’s parity clause is narrower than that alleged to have been implemented by BVB,
with the restriction affecting all distribution channels controlled by an accommodation provider (a narrow retail-MFN).
It was also noted that by affecting all distribution channels controlled by an accommodation provider the parity clause
introduced by FAR differs from the parity clause commitments applying in some ICN member jurisdictions.
Page 14
Attachment A
Effect on distribution channels
24. As with the effect on accommodation providers, project group members differed in
their views of the effect of the parity requirements on distribution channels. These
views are summarised below.
View 1: limited or no detrimental impact on competition

25. Some members of the project group considered that the narrow price parity clause
(paragraph 14(a)) would not restrict price and quality competition between non-
proprietary distribution channels. It was noted that the narrow parity requirement is a
less intrusive requirement than that of a wide parity requirement and does not affect
the relationships between an accommodation provider and other OTA platforms.
26. Under a narrow price parity requirement an accommodation provider can offer a
competing OTA a lower price or superior inventory in exchange for a better
commission rate or other commercial benefits. This interaction between an
accommodation provider and an OTA places competitive pressure on BVB to ensure
that its commission rates are not excessive. This competitive tension also ensures
that BVB (and other OTAs) are incentivised to maintain and improve the demand
enhancing features of their platform – supporting dynamic innovation.

27. As noted at paragraph 20 above, some members of the project group considered that
a range of commercial factors may have resulted in the reduced ranking of the
accommodation provider on BVB’s platform. It was considered that further information
would be required to test whether the alleged conduct is widespread or otherwise
represents a wide parity requirement.
View 2: detrimental impact on competition

28. As has been noted, the views of the project group differed on the question of whether
accommodation providers and OTAs were likely to participate in the same market.
Some members of the project group considered that if accommodation providers and
OTAs participate in the same market then the narrow parity requirements (set out in
paragraph 14(a), above) would soften competition between the OTA and the
distribution channels controlled by the accommodation provider. It was considered
that by restricting an accommodation provider from growing direct bookings through
more favourable room rate and/or inventory offers, a narrow clause would soften the
competitive constraint that the accommodation provider’s distribution channels may
otherwise have imposed on non-proprietary distribution channels.
29. It was noted that narrow parity requirements applied by OTAs may restrict competition
between OTAs in cases where they produce equivalent effects to wide parity
requirements. Since the narrow parity requirement prohibits the accommodation
provider from offering better room prices on its own website than on the OTA which
imposes the requirement, if the accommodation provider wishes its own online offer to
match a lower price offered on another OTA portal, it will be obliged to reduce the price
on the first OTA portal as well. Depending on the particular facts of the relevant
market, in particular the share of sales conducted through direct and indirect channels,
Page 15
Attachment A
this price floor effect may reduce the incentive for OTAs to compete on commission
rates, by reducing the incentive for accommodation providers to make use of any lower
commission rates offered by competing OTAs.

30. Some project group members noted that wide parity requirements (14(b), above) can
soften and may foreclose price and non-price competition between distribution
channels.
31. It was noted that the inventory parity conduct alleged to have been engaged in by BVB
would prevent an accommodation provider from offering better room inventory and
inclusions to other distribution channels. Project group members considered that
further information/evidence would be required in considering the existence of, terms
and potential effect of the alleged wide parity clause on competition between
distribution channels (see Areas of further inquiry, below).
32. It was noted that differentiated offers by accommodation providers create incentives
for distribution channels to compete for their custom, for example by offering lower
commission rates or other competitive benefits as reward for better room rates or
inventory being listed with their platform. By requiring accommodation providers to
offer them their best inventory BVB diminishes the competitive value of what may be
offered to consumers through other online distribution channels. This diminishes the
incentives of competing OTAs to offer lower commission rates/benefits to
accommodation providers and may reduce the competitive tension that would prevent
BVB from introducing excessive commission rates. It was also noted that the parity
requirement may affect the take up of services for smaller or more marginal OTAs and
have a negative impact on their ability to increase scale or market share through listing
more competitive offers.
33. In considering whether the wide inventory parity clause was likely to diminish
competitive tension between OTAs, some members of the project group also noted
that a loss of competitive tension could result in a loss of, or a reduction in, innovation
in OTA services.
34. As noted in paragraph 21 above, some project group members considered that regard
should be had to the cumulative effect of the narrow parity requirements of BVB and
FAR in assessing whether the theories of harm associated with wide parity
requirements are likely to arise. It was considered that further information/evidence
would be required to assess the combined effect of the narrow parity clauses on
competition between distribution channels (see Areas of further inquiry, below).
Other matters
35. Project group members differed on the question of whether dominance should be
considered in examining the conduct of BVB. For some members of the project group
the market share relativities of BVB and FAR (25 and 40 per cent respectively of total
online travel bookings made in this jurisdiction) made it less likely that dominance
would be considered in their jurisdiction. While some members of the project group
noted that BVB could be dominant, a factor relevant to their consideration was the
extent to which BVB is a vital channel for accommodation providers. It was considered
that a ‘must have’ platform will have a greater ability to act without constraint in its
dealings with accommodation providers. Balanced against this it was noted that
consumers in this market typically multi-home – that is they consider more than one
platform when searching for a suitable offer. Multi-homing by customers may reduce
Page 16
Attachment A
the ‘must have’ value of a platform that otherwise has high market share. Other
factors that would be considered included network effects, barriers to entry and the
impact of access to consumer data. Some members of the project group noted that if
dominance is substantiated further consideration should be given to whether the room
rate and inventory parity requirements could be an abuse of this dominant position.
Further information/evidence would be required in considering this (see Areas of
further inquiry, below).
36. It was also noted by a member of the project group that as a wide inventory parity
requirement, such as that alleged in paragraph 14(b), restricts price and non-price
competition on room features and inclusions it could, in some jurisdictions, be
assessed under the resale price maintenance (RPM) provisions. Views of the project
group differed on this, with some noting that the parity arrangements would not meet
the requirements for establishing RPM in their jurisdiction.
Pro-competitive effects and factors relevant to these

considerations
37. Project group members noted that, in general, their jurisdiction does have regard to
pro-competitive effects, although there are differences in the manner in which such
effects can be taken into account across these jurisdictions4. For some jurisdictions
pro-competitive effects would only be considered if competitive harm had been
established. As outlined above the views of project group members differed when
considering the potential competitive harm of BVB’s conduct.
38. It was noted that both wide and narrow parity requirements may prevent free-riding, by
competing OTAs and accommodation providers respectively, on the investments made
by BVB. It was noted that without the wide parity requirements, consumers may use
the websites and hotel portfolios of BVB to search for and compare accommodation
offers, but then book more cheaply through other distribution channels. Free riding
may deprive BVB of commission revenue and reduce or deter investment in services
(e.g., website development, search engine optimisation, online advertising). It was
suggested that investigators should consider whether free riding is empirically
significant. It was considered that evidence of free-riding conduct should be presented
by BVB in support of the parity requirements. It was noted that the pro-competitive
effects were more likely when considering narrow parity requirements.
39. It was noted that OTAs promote inter-brand competition between accommodation
providers through increased price transparency. OTA platforms also reduce consumer
search costs. The parity requirements may support an OTA’s business model by
increasing conversion rates (the ‘look to book’ ratio) by ensuring that the room rates
and inventory offered through their platform is superior. Displaying uncompetitive room
rates or inventory may harm BVB’s credibility with consumers.
40. The project group noted that further consideration would need to be given to whether
the parity requirement (wide or narrow) is reasonably necessary or indispensable to
obtain the benefits provided by BVB. Regarding the indispensability of the restraint
consideration should be given to the specific features of the restraint, the parties
adopting it, and the market in which the restraint will apply. In particular, consideration
of whether other approaches/business models could be used to obtain the same
benefits with a less competitively restrictive approach than a parity requirement. It was
noted that approaches could include charging hotels a small amount for display in
4
See Attachment OPR-C.
Page 17
Attachment A
each search request or for each ‘click through’ even if the consumer does not
ultimately book the room.
Areas of further inquiry

41. The project group identified a number of areas for further inquiry by investigators
assessing BVB’s conduct, including:
Parity clauses
a. Evidence of the scope of the parity clauses implemented by OTAs (e.g.
standard terms, individually negotiated agreements)
b. Market coverage of the parity clauses (e.g. are all accommodation providers
subject to such restrictions?)
c. Evidence of enforcement of the clauses (e.g. de-ranking, dimming, de-
listing?)
d. Evidence of the effect of the narrow parity requirements (e.g. do affected
accommodation providers offer differentiated rates or inventory between
OTAs? If not, is this because of the narrow parity requirement or for some
other reason?)
e. Evidence of other non-price factors affecting competition (e.g. reputation,
quality)
f. Use of alternative business models by OTAs (including in other jurisdictions)
g. Evidence of indispensability of parity requirements (e.g. operations of BVB,
FAR and other significant OTAs in other jurisdictions)
h. Evidence of the economic rationale for the parity clause (i.e. investment
protection, reputation and others)
Market definition and other considerations

a. Evidence of consumer search and booking behaviour (are offline sales
channels part of the same relevant market as online channels; do consumers
consider/search websites of accommodation providers as well as OTA
platforms; consumer views on the differences in functionality and ease of use
of different distribution channels)
b. Evidence of the effect of the parity requirements on OTAs (changes in shares
of bookings per channel; changes in OTA market shares, including
consideration of entrants; changes in OTA commission rates; changes in non-
price competition between OTAs)
c. Evidence of the effect of the parity requirements on online distribution
channels controlled by the accommodation provider
d. Evidence to enable a comparison of relevant distribution channels
(functionality; cost; efficiency; consumer willingness to use)
Other
a. Features of hotel websites relative to OTA websites (online visibility, instant
booking functionality etc.)
b. Impact of cross-channel room price and availability differentiation by hotels on
their conversion rate on particular OTAs
Page 18
Attachment A
c. Degree of commercial and financial risk undertaken independently by OTAs
(are they genuine agents/independent entities for the purposes of competition
law?)
Page 19
Attachment A
Attachment OPR-A –
Online Parity Requirement
Hypothetical scenario
Limited Fact Scenario – online parity requirements

Investigation of the effect on competition of parity provisions introduced by online travel
agent (OTA) BestValueBed Pty Ltd.
Background information
OTAs
 OTAs are intermediary platforms that bring together accommodation providers and
consumers who wish to book accommodation.
 OTAs provide an interactive search function, presenting results in an online list. An OTA
may display up to 50 results on each page and provides multiple pages of results. The
search results are often presented as a list of hotels recommended by the OTA, but the
consumer can also choose to have the search results presented according to number of
stars, location, price or reviews by previous customers. The order in which the hotels are
displayed or "ranked" in the search results is determined by various parameters decided
by the OTA. Research into consumer behaviour suggests that a significant number of
consumers choose accommodation from within the first section of search results.
 OTAs and accommodation providers negotiate a commission that the OTA will receive
for its services when a booking is made via its platform, a higher commission may be
agreed where an accommodation provider seeks preferential listing.
 Three OTAs are active in this jurisdiction:
o BestValueBed (BVB) – 25 per cent of total online accommodation bookings in
this jurisdiction are made via its platform.
o Find-A-Room (FAR) – 40 per cent of total online accommodation bookings in this
jurisdiction are made via its platform.
o Online Hotel Beds (OHB) - a recently launched OTA platform, 10 per cent of total
online accommodation bookings in this jurisdiction are made via its platform.
Parity Agreements
 Parity agreements can take a number of forms, in the context of OTAs they are a vertical
agreement between an OTA and an accommodation provider whereby the
accommodation provider agrees to offer the OTA at least the same number and type of
rooms, and/or at least as favourable a price, as it offers through any other booking
channel, including other OTAs and its own website. There are some variations on this
model.
Distribution and Marketing Channels

Accommodation is marketed and distributed in this jurisdiction via a number of online and
offline channels.
 Offline: Telephone bookings, ‘walk-in’ bookings, customer loyalty and reward programs
run by accommodation suppliers, bookings made through bricks and mortar travel
agents (other than through their online booking platforms) and wholesale tour
arrangements.
Page 20
Attachment A
Online:
 OTAs
 Bricks and mortar travel agencies with an online presence, the number and variety
of accommodation providers offered through such agencies is generally more
limited than offered by OTA platforms
 Individual hotel websites
 Metasearch websites5 and
 Purpose Specific and Package Online Agencies6.
Alleged conduct
Online Hotel Bookings
 Online Hotel Bookings (OHB) operates a locally-based online platform, it began trading
in 2015. OBH has approximately 10 per cent of total online accommodation bookings
made in this jurisdiction.
 OHB has complained that the parity requirements of BVB and FAR are preventing it from
competing effectively with BVB and FAR.
 OHB claims that it is offering a lower commission rate in order to attract accommodation
providers to its site. OHB considers that its lower cost service allows accommodation
providers to offer lower room rates via its site.
 OHB claims that it wants to grow its reputation as a competitive OTA with consumers
and the lower commission rate is a key element in its strategy. OHB claims that hotels
are willing to list rooms with it on the basis of the lower commissions but have stated that
they are unable to offer lower pricing as sought by OHB as the parity agreements in their
supply contracts with BVB and FAR prohibit them from doing so.
 OHB alleges that it is unable to gain market share because the room rates listed on its
site are the same as those listed with BVB and FAR. It has argued that this cannot be
effectively addressed through marketing or other strategies such as search optimisation.
Hoolten Hotels
 Hoolten Hotels (Hoolten) is a well-known, multinational hotel chain.

 It secures 40% of its bookings through OTA platforms; 15% through its offline loyalty
reward program; 10% through traditional travel agents; 15% through its own website;
and 20% through offline channels.
 Hoolten uses both BVB and FAR, the commission rates payable by Hoolten in 2014
were 14% and 20% for BVB and FAR respectively.
 In May 2015 FAR introduced the following parity requirements
5
Metasearch websites allow consumers to search for and compare offers from multiple OTAs and from the hotel itself for
the same accommodation product (same hotel, same room type, same dates). Metasearch sites generally do not provide
a booking service, but re-direct customers to an OTA or the hotel website to make their booking
6
Purpose Specific and Package Online Agencies (PSPOA), which offer search compare and online booking services for a
range of travel options, including flights, accommodation, car hire and package deals. In addition to independent
PSPOAs, this service can be provided through other online avenues such as airline websites
Page 21
Attachment A
9.2: The Supplier will offer FAR its rooms and packages at pricing levels
equal to or less than those offered directly by the Supplier to consumers
through the Supplier’s own online and offline channels, including (but not
limited to) the Supplier’s website, telephone, e-mail and walk-in bookings.
9.3: The Supplier will offer FAR the same number of rooms and
packages, and the same types of rooms and packages, as those offered
directly by the Supplier to consumers though the Supplier’s own online
and offline channels, including (but not limited to) the Supplier’s website,
telephone, email, and walk-in bookings.
 In November 2015 BVB introduced the following parity clause:
15.3.1: BVB requires the Supplier at all times to offer its rooms and
packages to BVB at pricing levels equal to or less than those offered by
the Supplier through online channels controlled by the Supplier.
 In accepting the 2015 agreement Hoolten negotiated a revised commission with BVB of
12%.
 Hoolten alleges that BVB is seeking to achieve pricing and inventory parity across all
platforms.
o In March 2016 Hoolten offered a “Family Staycation” promotion – offering rooms
with free upgrades to consumers who purchased packages through the FAR
platform.
Hoolten advises that it had negotiated an advantageous commission rate and
search listing with FAR as part of its participation in the promotion on the FAR
platform.
o The Family Staycation promotion was also available through Hoolten’s offline
channels.
o Three days after the campaign commenced, Hoolten received an email from BVB
which advised it had seen Hoolten’s inventory package on the FAR platform and
requested that Hoolten immediately offer the same inventory on the BVB
platform.
o Hoolten declined this request and claims that almost immediately it noticed a
significant drop in the number of daily bookings received through the BVB
platform. Hoolten has claimed that this was the result of BVB ‘dimming’ its’ listing
on the BVB platform – dropping the listing from the first page of search results to
the sixth page of search results. Hoolten claims that its bookings through BVB
dropped around 40% during the promotional period but returned to normal levels
after the promotion ended, and the hotel returned to the first page of BVB’s
search results.
Page 22
Attachment A
Inquiries of BVB
 BVB operates a popular online platform which was established in 2012. In 2016 BVB
had an annual turnover of $US300 million. BVB estimates that 25% of all online travel
bookings for accommodation in this jurisdiction are booked through its platform.
 BVB submitted that the primary aim of the parity clause was to address the risk of free
riding by accommodation providers on its investment. BVB argued that investment in its
platform would be sub-optimal in the absence of the parity clause leading to a loss of
consumer benefit.
 BVB argued that its platform provides consumers a valuable ‘one-stop shop’ where they
can search, compare and book accommodation. BVB maintained that without price
parity requirements, some of the benefits to consumers would be lost because
consumers would need to spend more time searching other platforms to be sure they
have found the lowest price. BVB considers that its platform promotes competition
between accommodation providers by increasing transparency about prices, quality and
availability. BVB maintains that these pro-competitive benefits would be lost if its
business model was undermined by the removal of parity requirements.
 In respect of Hoolten’s concern that it had been ‘dimmed’ as punishment for offering a
lower price on another platform, BVB maintained that this was the result of a drop off in
sales of Hoolten’s inventory and the ranking of its less competitive offer relative to other
accommodation providers on the platform at the same time.
Market Research
 Research suggests that consumers in this jurisdiction typically utilise more than one OTA
platform when searching for accommodation, switching between platforms to assess
price and non-price offerings. Research also suggests that hotels seek to join multiple
OTA platforms in order to gain access to each OTA’s customers.
 Generally, the proportion of bookings made by consumers directly with an
accommodation provider is greater for larger providers due to stronger brand-awareness
and repeat bookings driven by customer loyalty programs.
 Various factors can affect the share of accommodation bookings made through online
channels, including time of year, location of accommodation and type of accommodation.
 On an annual basis online platforms account for 60% of total accommodation
reservations within this jurisdiction. Bricks and mortar travel agencies, including those
with an online presence, account for 20% of total reservations, proprietary distribution
channels, including hotel websites, telephone bookings and ‘walk ins’, account for 20%
of total reservations.
Page 23
Attachment A
Attachment OPR-B
Vertical Restraints Project Group

The UCWG Co-chairs would like to thank the following members and NGAs for their contributions to
this paper:
Competition Agencies
 Administrative Council for Economic Defense – CADE, Brazil
 Autorité de la concurrence, France
 Authority for Consumers and Markets, Netherlands
 Bundeskartellamt
 Competition and Markets Authority, United Kingdom
 Competition Bureau, Canada

 Competition Commission, Hong Kong
 European Commission
 Federal Trade Commission, United States
 Japan Fair Trade Commission
 New Zealand Commerce Commission
 Swedish Competition Authority
NGAs
 Carolyn Oddie, Australia
 Daniel Andreoli, Brazil
 Martim Della Valle, Brazil
 Silvia Fagá, Brazil
 Bernardo Macedo, Brazil
 Joyce Midori Honda, Brazil
 Antonio Di Domenico, Canada
 Rachel Brandenburger – European Commission
 Rafael Allendesalzar , Spain
 Grant Murray, United Kingdom
 Daniel Sokol, United States of America
 Robert Friedman, United States of America
 Kitri Gupta, United States of America
 Elisa Kearney, United States of America
Page 24
Attachment A
Attachment OPR-C
Short summary of legal framework applying in ICN

member jurisdictions
Australia
In Australia the legal framework requires consideration of the effect of the parity
requirements on competition, a substantial lessening of competition will give rise to a
contravention. The prohibition in Australia does not have direct regard to offsetting
efficiencies or pro-competitive benefits. The Australian competition law provides for these
factors to be considered under an authorisation (exemption) regime.
Austria
MFN-clauses (price parity agreements) may be subject to Art 1 Austrian Cartel Act
("ACA") or Art 5 ACA (cartels or abuse of a dominant market position), as well as to Art
101 and 102 Treaty on the Functioning of the European Union.
As in many other EU member states there have been investigations into price parity
agreements of OTAs in Austria. Following the public discussions a per se prohibition of
price parity agreements used by online booking platforms was introduced into the
Austrian Unfair Competition Act (“UCA”). In detail, clause no. 32 of the annex regarding
aggressive commercial practices states that “the demand of an OTA against a lodging
enterprise not to offer more favourable prices or more favourable conditions as displayed
on the booking platform on any other distribution channel, including their own website, is
void“.
Besides the per se prohibition of the UCA for OTAs, MFN-clauses may be in general
assumed to restrict competition and will therefore in many cases fall within the scope of
the cartel prohibition as well as, if used by a dominant undertaking, to constitute an abuse
of a dominant position. However, a case-by-case analysis of the certain conduct in
question will be required regarding possibly applicable exemptions.
Brazil
Under Brazilian competition law a company is prohibited from unilaterally abusing its
dominant position in the market. A dominant position occurs when a company or group of
companies is capable of altering, in a unilateral and concerted manner, the market
conditions or when it controls twenty percent or more of the relevant market. In assessing
whether unilateral conduct is an abuse of a dominant position economic efficiencies will
be weighed against the potential anticompetitive effects.
Canada
In Canada, the Competition Act provides the legal framework for considering parity
clauses such as the ones described in this hypothetical. In particular, sections 78 and 79
establish the abuse of dominance provisions. Abuse of a dominant position occurs when
a dominant firm or a dominant group of firms in a market engages in a practice of anti-
competitive acts, with the result that competition has been, is, or is likely to be prevented
or lessened substantially. In order to assess this, the Competition Bureau will look to
various factors such as direct and indirect indicators of market power, the primary
purpose of the conduct, any pro-competitive or efficiency-enhancing rationale for the
conduct, and qualitative and/or quantitative evidence of competitive effects.
Page 25
Attachment A
Attachment OPR-C
European Union
Article 101 of the Treaty on the Functioning of the European Union ('TFEU') prohibits
agreements between undertakings which have the object or effect of restricting
competition and which may affect trade between EU Member States. The prohibition does
not apply to agreements which satisfy the four conditions of Article 101(3) TFEU, i.e. they
produce objective efficiencies; they allow consumers a fair share of the resulting benefits;
they do not impose restrictions which are not indispensable, and they do not substantially
eliminate competition. The burden of proving a restriction of competition lies with the
claimant or enforcing authority; the burden of establishing the Article 101(3) efficiency
defence lies with the parties to the agreement.
Retail MFN clauses, such as those in the Online Travel Agents hypothetical, are vertical
agreements, i.e. the parties operate at different levels of the supply chain for the purposes
of the agreement. Vertical agreements are exempted from Article 101 TFEU by the
Vertical Agreements Block Exemption Regulation ('VBER'), provided they do not contain
specified hardcore restraints (including retail price maintenance) and provided the market
shares of the supplier and the buyer under the agreement do not exceed 30%. Retail
MFN clauses are not currently treated as restrictions 'by object' and they are not hardcore
restraints under the VBER.
The presumption of legality conferred by the VBER may be withdrawn by a competition

authority where it finds that an individual agreement has effects which are incompatible
with Article 101(3) TFEU, for example because the agreement makes a significant
contribution to a cumulative effect produced by parallel networks of similar agreements.
Such withdrawal decisions only produce effects ex nunc.
Outside the VBER, retail MFN clauses are subject to a full effects analysis, taking into
account both actual and potential effects. The clauses are considered in their legal and
economic context, including the counterfactual scenario. Any potential restrictive effect
must be reasonably probable. Relevant factors include the nature and content of the
agreement, whether the parties have market power, and whether the agreement
contributes to the creation, maintenance, strengthening or exploitation of that market
power. This implies that the relevant product and geographic market(s) must be defined.
To fall within Article 101 TFEU, any restriction of competition must be appreciable.
Japan
In Japan the legal framework for considering parity requirements is provided for by the
Antimonopoly Act (AMA). Relevantly the AMA provides that a party must not trade with
another party on conditions which unjustly restrict any trade or other business activities of
that other party. In the context of the parity requirements this requires consideration of
whether the provisions have the effect of resulting in substantial restraint of competition,
or whether such requirements tend to impede fair competition. This analysis will consider
both anti-competitive and pro-competitive effects.
Turkey
The Turkish Competition Authority (TCA) amended of the Guidelines on Vertical
Agreements (Guidelines) including the assessment of most favored customer clauses
(MFC). The draft Guidelines concentrate on two main aspects of MFC practices; potential
Page 26
Attachment A
Attachment OPR-C
price rigidity problems associated with such practices and general approach to MFC
practices from the point of view of vertical agreements.
The TCA considers an MFC to be anticompetitive when the parties to the contract have
higher market shares compared to the competitors in the market, the concentration level
is high and the MFC beneficiary benefits from the clause at all times or the beneficiary
punishes the contracted seller by increasing its costs when the contracted seller provides
more favorable conditions to the competitors of the MFC beneficiary.
United States of America

Rule of reason - takes into account both the anti-competitive effects and the pro-
competitive benefits of the conduct. Common considerations under a rule of reason
analysis include the market power of the entities involved, the scope of the restraint, the
number of entities within the market adopting the restraint and the restraint’s source. The
rule of reason approach considers whether the restraint is reasonably necessary to
achieve the pro-competitive benefits.
Page 27
Attachment B
Vertical Restraints Project
Hypothetical Two – Online Sales Bans
Introduction
1. A small group of volunteers from competition authorities and NGAs (the “project group
members”, see Attachment OSB-A) has considered the second of the two hypothetical
scenarios prepared as part of the vertical restraints project. The second scenario
provided to the project group is set out at Attachment OSB-B.
2. The hypothetical considers ban on selling through online third-party platforms and
online advertising ban. The project group has been invited to consider:
4. possible theories of harm;
5. arguments about pro-competitive effects that may be presented and factors

relevant to these arguments; and,
6. areas of further inquiry by investigators.
3. This report provides an outline of matters considered relevant by the project group
members in assessing the hypothetical restrictions. The report seeks to highlight
approaches to assessing them, including areas of divergence, however without
identifying the views of contributing members of the project group.
4. A disclaimer applies: not all members of the project group agree that all of the factors
and considerations summarised below are appropriate because project group
members may evaluate them in different ways.
Overview of hypothetical scenario and restrictions

5. The hypothetical concerns various selective distribution agreements between three
major manufacturers of baby strollers and their authorised retailers (both brick and
mortar and online). By assumption, these agreements are generally considered legal.
However, complaints have been received alleging that additional restrictions imposed
by the three major manufacturers on their authorised retailers are contrary to
competition law. The restrictions include the following:
a) Babydream (20% market share) prohibits all sales on third-party platforms and
has a complete online advertising ban;
b) Wheelies (>30% market share) allows sales to third-party platforms if the

platforms comply with the same conditions imposed on the retailers themselves;
and
c) Sport 2001 (>30% market share) only allows lower/discounted goods to be sold
by retailers on third-party platforms but sells all products itself on third-party
platforms.
Page 28
Attachment B
6. Between 2015 and 2017, the proportion of online sales of total sales of baby strollers
has gradually increased (20% ⇒ 30% ⇒ 35%) and the proportion of online sales via
third-party platforms has increased as well (20% ⇒ 30% ⇒ 40%). The retail landscape
is populated by five large brick-and-mortar retailers, two large third-party platforms and
other small online/offline retailers.
Legal frameworks
7. While different legal frameworks apply across ICN member jurisdictions, these
frameworks generally include a form of effects test when assessing the above
mentioned restrictions. In some cases, the framework also envisages per se
prohibitions, the existence of presumptions of illegality and exemption for certain
categories of agreements under certain conditions.
8. The restrictions at stake have been analysed by project group members under the
general framework concerning agreements/restraints (in particular vertical). Some
other members also applied the framework of abuse of dominance.
Relevant market
9. Responses received from the project group members noted that further information
would be required in order to formulate the relevant product market definition while the
geographic scope is assumed to be national in the hypothetical scenario. For the
purpose of preliminary considerations the project group agreed that the market for the
retail sale of baby strollers is a differentiated product market where brand image and
reputation may play a role. In this regard, relevant issues to be further analysed in the
context of market definition are whether:
 discounted products are viewed to be substitutes to luxury baby strollers;
 brick-and-mortar retailers are in the same product market as the online

distribution channels; and,
 third-party online platforms are in the same product market as other online
distribution channels.
Theories of harm and pro-competitive effects

10. Some members of the project group commented that the above mentioned restrictions
can have a detrimental impact on competition. At a high level, members of the project
group have indicated the following possible theories of harm which would have to be
corroborated by the specific facts of the case at hand:
a) Reducing intra-brand competition: restricting online sales through third-party

platforms and online advertising can soften the pressure on retail prices and
protect existing sales channels from competitive pressure from online channels,
by directly limiting the choice of online consumers. This theory can be more
plausible if the manufacturer sells online, since these restrictions can protect the
manufacturer’s own online channel from competitive pressure by other online
retailers.
b) Softening inter-brand competition: if sales through third-party platforms is an

important means of competition among the three manufacturers and their
Page 29
Attachment B
respective retailers, the cumulative effects of above restrictions could limit
competition across manufacturers under certain conditions, e.g., if brands are in
the same relevant product market despite product differentiation and competition
from other online channels and offline sources do not provide sufficient
competitive constraints (see also paragraph 44).
c) Foreclosing existing or potential competition (in particular, through third-party

platforms but also other online channels) in the relevant market, especially in the
long run, and this concern was raised by one member adopting a framework
based on abuse of dominance. The general theory of harm suggested in this case
is whether manufacturers are leveraging their position in the upstream market, i.e.,
the manufacture and wholesale of baby strollers, to give themselves advantages
over their rivals in the downstream market, i.e., the retail sale of baby strollers to
final consumers.
d) Market partitioning where appropriate: in the context of the EU framework,

these restrictions may partition national markets according to national borders and
frustrate the EU objective of achieving an integrated EU single market for the
benefit of consumers.
11. However, some other members did not envisage any competition concerns. In
particular, they stated that the complaints do not suggest that there is a horizontal
agreement at any level of distribution, or that any of the manufacturers is engaging in
anticompetitive conduct (through vertical restraints or other means) that hinders the
ability of manufacturers to distribute their products or the ability of any retailers to
compete.
12. By focusing only on inter-brand competition, these members stated that the restrictions
imposed by the three manufacturers might be reasonably necessary to maximize
efficiency across all distribution channels, identified in the number of five: 1)
manufacturer websites, 2) large third-party online platforms, 3) smaller websites, 4)
large brick-and-mortar retailers, 5) small brick-and-mortar retailers. All identified
channels appear to be significant, with at least three competitors: therefore,
competition in stroller distribution appears to be structurally competitive. Moreover,
these members noted that the interests of retailers and manufacturers are never
completely aligned and they might diverge significantly when a large portion of a
manufacturer’s sales do not go through the retailers.
13. One member recognised that while intra-brand competition is reduced by such
restrictions, the information provided suggested that inter-brand competition could be
expected to constrain the effect of the restraint in the retail market. This member
considered that the effect on competition in the retail market for baby strollers was
unlikely to be of a sufficient magnitude to warrant intervention.
14. In terms of pro-competitive effects or business justifications, members of the

project group stated that the above restrictions can contribute to the promotion of
competition on other factors than price:
a) They can help protect brand image and the position of brand across the various
distribution channels (e.g., creation of an environment meeting consumer’s
expectations on product information, services, etc.).
b) They can help to ensure quality of pre- and post-sale services (e.g., safety
demonstrations). The services offered by third-party platforms are not controlled
by the manufacturers, which often do not have a contractual relationship with the
online marketplaces, allowing the former to enforce qualitative criteria on the
latter.
Page 30
Attachment B
c) Additional potential justification (which could be brought up by manufacturers)
could include the reliability of the platforms for delivery and payment.
15. While the above theories of harm and pro-competitive effects apply generally,
considerations with respect to the particular restrictions adopted by each of the three
manufacturers may differ as described in the subsequent sections. The key issues
raised by the hypothetical restrictions in question are related to the impact of e-
commerce on distribution networks and competition more generally:
a) The trade-off between price versus non-price competition. E-commerce has

the potential to intensify price competition through increasing the number of
participants in the market and easier price comparisons by consumers. While
price is often the key parameter for (online) retailers, manufacturers are also
concerned with other factors, such as brand image, quality and innovation, which
are potentially undermined by the free-riding problem. As a result, manufacturers
may consider (some) online sales restrictions as necessary to better control for
these other factors and preserve investments in high-level presale services,
promotion and advertising. In other words, online sales may improve price
competition while reducing non-price competition at retail level (absent
manufacturers restrictions) and the former effect will not always be more important
than the latter, both for firms and consumers.
b) The importance of intra-brand competition versus inter-brand competition in the

assessment of the consumer welfare effects. In jurisdictions where the focus of
the assessment is mainly on inter-brand competition, restrictions on online sales
can be problematic only if there is market power at the manufacturing level. In
jurisdictions focusing also on intra-brand competition, online sales restrictions at
the distribution level affecting retailers selling products of the same brand can
raise competition concerns.
Babydream’s ban on selling to third-party platforms and wide ban

on online advertising
16. Babydream prohibits all sales on third-party platforms and has a complete online
advertising ban, including on retailers’ own website7 and through the use of price
comparison websites; in particular Babydream retailers are prohibited from using
search advertising (i.e., from using/bidding on search ads for Babydream brand name).
Members of the project group held different views on the potential competitive impact
of the Babydream’s restrictions. These views are summarised below.
View 1: no impact on retail competition

17. Some members of the project group, whose competition law focuses on inter-brand
competition, noted that it is unlikely that these restrictions would violate competition
law, absent any harm to inter-brand competition. They noted that both restrictions on
online sales through third-party platforms and online advertising are unlikely to have
material consequences for the viability of third-party platforms because its market
share is 20%: the other two manufacturers, accounting more than 60%, allow the sale
of baby strollers via third-party platforms under certain conditions. It was noted that,
even when assuming that that Babydream’s restrictions may undermine the viability of
third-party platforms, such scenario is unlikely to materially harm inter-brand
7
That is, retailers can sell Babydream products through their webstore/app but only if consumers actively search
for them e.g. they are not able to promote them on their own webstore/app.
Page 31
Attachment B
competition due to limited market share of the third-party platforms (they account for
14% of total sales).
18. Some members of the project group recognised the importance of intra-brand
competition and, in relation to the ban on online sales via third-party platforms,
noted that the investigation would - based on the applicable case law in the jurisdiction
- in a first step require an assessment of the necessity and proportionality of the
restriction within the selective distribution system and only in the second step an
analysis of the potential restriction of competition. They stated that the ban on online
sales via third-party platforms is likely to be viewed as a legitimate requirement
justifying the reduction in price competition, such as maintaining Babydream’s brand
image for high quality products and to protect investments by its retailers in pre- and
post-sale services. According to the facts of the hypothetical, consumers attach a high
importance to brand image and safety of the product and third-party platforms may
offer less competition on these non-price dimensions. In addition, the ban is applied
uniformly across the distribution network and Babydream itself is not selling to third-
party platforms. Therefore, absent any contractual relationship between Babydream
and third-party platforms, it might be difficult for Babydream to enforce even less
restrictive qualitative criteria and monitor them on third-party platforms. Based on
these considerations, the ban would meet the first step requirement and a further
analysis of a potential restriction of competition would be unnecessary.
View 2: detrimental impact on retail competition

19. One member of the group noted that the total ban on online sales via third-party
platforms could be a restriction that is presumed to have severely anticompetitive
effects, irrespective of the market share of the firm concerned. In assessing the
necessity and proportionality of the restriction, this member considered that such
restriction in selective distribution systems could be justified only for luxury goods for
which brand image and reputation are a distinct feature, a view that was not shared by
other members (see paragraph above). It argued that in the market for the retail sale of
baby strollers, brand image and reputation could be achieved with less restrictive
means, e.g., quality requirements for the third-party platforms. In addition, it observed
that severity of such restriction has to be analysed in light of the growing relevance of
online platforms as gateways to end customers, especially for small and medium-sized
online retailers, as it can be derived from the hypothetical facts.
20. In relation to the wide ranging online advertising ban, some members of the project
group conducted a similar type of assessment suggested for the online sales platform
ban (see paragraph 18), i.e., an evaluation of necessity and proportionality of the
clause within the selective distribution system, concluding that such ban is likely to go
beyond what is necessary to achieve a legitimate aim. For these members, such a
wide online advertising ban does not appear to pursue any legitimate objective of
promoting brand image or investing in safety demonstrations. On the contrary, its main
purpose appears to be the reduction of (i) competitive pressure by authorized retailers
on Babydream’s own online retail activities and (ii) Babydream’s bidding costs in the
case of online search advertising. It would therefore likely amount to a restriction of
competition, on its own or in conjunction with the online sales platform ban.
21. According to these members, while Babydream retailers are in principle able to sell its
product on their websites and presumably their websites comply with Babydream’s
qualitative criteria, they are prevented from actively approaching customers via
targeted online advertising or by generally advertising for their website online. Thus,
retailers would not be able to generate any meaningful traffic on their websites, as the
ban undermines their ability to attract consumers to their websites: consumers
searching for Babydream’s baby strollers would be unable to locate the online offers of
Page 32
Attachment B
the retailers. Furthermore, it was noted that such restrictions may distort competition in
relation to the acquisition of online search terms, by preventing a retailer from
participating in an auction for search terms that are likely to capture the Babydream
brand name or its particular products. In conclusion, the advertising ban could limit
intra-brand and possibly inter-brand competition and, in the European context, partition
markets (by reducing the ability of retailers to advertise and sell to customers outside
their area of activity and cross border between Member States of the EU, using
internet).
22. One member noted that, if the relevant market is considered to be the online retailing
segment comprising manufacturers websites, retailers’ websites and third-party
platforms, all Babydream’s restrictions can be caught under the provisions prohibiting:
exclusive dealing with the ultimate objective of excluding online distribution channels
(third-party platforms) through input foreclosure; and/or, discriminatory practices. In
particular, under the exclusive dealing provision, such restrictions would be “per se”
illegal, i.e., without any further examination from the competition agency which would
use a rule of reason approach to analyse them under the discriminatory practices
framework.
23. Finally, for some members of the project group, an effects-based approach is
needed to assess whether the impact of Babydream restrictions is detrimental on retail
competition as described in paragraph 10 above, points a), b) and c).
24. One member applying the abuse of dominance framework highlighted that it would be
necessary to assess whether: i) Babydream has market power in the upstream market
according to the leveraging theory mentioned in paragraph 10 c) above; ii) the intent
behind Babydream’s restrictions is to exclude online third-party platforms in the
downstream markets; and, iii) the effects of the conduct.
25. In any event, the effects based approach would require further analysis on factors
listed in the section below.
Areas of further analysis

26. For a more comprehensive effects-based analysis on the impact of Babydream
restrictions on retail competition, project group members highlighted several factors
requiring further investigation which can be grouped as follows:
a) Intra-brand competition analysis: e.g., difference in prices between strollers sold

directly by Babydream and those sold by authorised retailers; analysis of
consumer ability to find the online offers of authorised distributors (e.g., how often
consumers use price comparison websites and their relevance); importance of
online advertising to intra-brand competition; more generally, analysis of sales
data to evaluate the effects of the restrictions, especially on online retailers who
do not have brick-and-mortar stores: for instance, the analysis of intra-brand
competition for Wheelies baby strollers (which are not constrained) could
represent a good counterfactual scenario to assess the effects of Babydream’s
restrictions; comparison with other geographic markets where restrictions do not
apply;
b) Competition analysis of the distribution channels: whether retail competition from

offline sources or non-platform online sources are good substitutes for retailing
through platforms.
c) Inter-brand competition analysis: substitutability of Babydream baby strollers with

rival brands sold through platforms or through any other means, from a consumer
Page 33
Attachment B
perspective(the weaker inter-brand competition the larger the anticompetitive
effects on intra-brand competition of the above restrictions); the closeness of
competition (analysis of prices and margins among manufacturers’ rival brands,
product positioning and reputation); evolution of market shares over time and
evidence of new entry or expansion of the small existing manufacturers.
d) Analysis of the intent: for example, an analysis of internal records to find whether
Babydream uses these restrictions as a means of better controlling prices and
discounts being offered in the online environment (and therefore reducing intra-
brand competition) with the result that consumers pay higher prices for the
products (as is alleged by some retailers).
e) Evaluation of the business justifications: for instance, if there are costs savings in
implementing these restrictions rather than having to police the conduct of each
online retailer to ensure that they have measures and policies in place to protect
the quality of each brand.
27. In addition to this analysis of the retail markets, the member of the project group
applying the framework for abuse of dominance also indicated several factors for the
analysis of dominance in the upstream market of the manufacturing, including the
analysis of the stability and durability of market shares, the entry conditions and the
countervailing buyer power of retailers (e.g., if there are retailers who refuse to carry
the products of Babydream due to the restrictions imposed or retailers who are large
enough to obtain more favourable terms in their distribution agreements).
Wheelies “equivalent” criteria required for selling through third-

party platforms
28. Wheelies permits authorised retailers to sell Wheelies’ products on third-party
platforms, provided that the platform meets the same qualitative criteria as the retailer
or equivalent ones. All members of the project group considered that such a restriction
is unlikely to raise competition concerns and appears to be justifiable in light of the
legitimate objectives of protecting brand image and investments in pre- and post-sale
services. In addition, there is no indication that these criteria are applied in a
discriminatory fashion.
29. However, it was noted that this restriction might soften intra-brand competition, by
limiting the ability to sell to customers, if third-party platforms that do not meet the
equivalent criteria are an important sales channel, and/or these criteria are applied in a
discriminatory manner.
30. Whether third-party platforms can and do comply is not stated in the hypothetical, but it
might be impossible for them to comply. In that event, according to one member,
Wheelies effectively imposes the same restriction on sales on online platforms as the
one imposed by Babydream.
31. Therefore, further analysis might be required on the definition by Wheelies of

“equivalent” conditions on third-party platforms and their implementation.
Page 34
Attachment B
Sport 2001 ban on selling high-end products through third-party
platforms
32. While retailers are allowed to sell some of Sport 2001's models (the lower-end models)
on third-party platforms, the sale of other models (e.g. the high-end models) is
contractually restricted. Members of the project group held different views on the
potential competitive impact of the Sport 2001 restrictions. These views are
summarised below.
View 1: no impact on competition

33. According to the members of the project group who did not envisage any competition
concerns generally (see paragraphs 11 - 13), Sport 2001 ban is much less restrictive
than those of either Wheelies and Babydream and therefore they would not be
investigated.
View 2: detrimental impact on retail competition

34. Some members of the project group considered there can be an impact on
competition. The theory of harm to be investigated by these members is whether such
restriction has the potential to reduce intra-brand competitive pressure for high-end
Sport 2001 products, by limiting the ability of resellers to sell high-end Sport 2001
products to customers, especially if third-party platforms are an important sales
channel. In a first step, the necessity and the proportionality of the restriction within the
selective distribution system of Sport 2001 would have to be assessed to ascertain
whether such restriction serves a legitimate objective which could be the maintenance
and protection of the brand image of high quality goods. In a second step, an analysis
would be required to determine whether the restrictions has the object or the effect of
restricting competition.
35. According to these members, it is unlikely that the restriction at stake serves the
objective of protecting brand image. Sport 2001’s argument - that it needs to control
the environment on third-party platforms by itself in order to protect its brand and that
retailers are unable to do that - appears to be inconsistent: because Sport 2001 is
already in a contractual relationship with the major third-party platforms, the
manufacturer may also have the means to ensure an environment that protects its
brand, thus raising the question as to why its retailers should not be allowed to
similarly sell the products in such an environment. In addition, the sales of lower-end
products by retailers on third-party platforms will affect the reputation of “all” Sport
2001 products, including the higher-end ones. Furthermore, these members noted that
Sport 2001’s restrictions are not being applied uniformly (Sport 2001 sells on third-
party platforms itself) and less restrictive options are available (like in Wheelies).
Finally, the protection of a certain level of quality of pre- and post-sale services and the
feedback on brand image cannot be considered a justification since Sport 2001
distributes its products on the major third-party platforms.
36. Given the above considerations, these members concluded that Sport 2001 ban may
violate their competition law as the ban may primarily serve the purpose and possibly
have the effect of better controlling prices and protecting Sport 2001's offer of high
margin products from price competition.
37. One member assessed this restriction under the exclusive dealing prohibition and
discrimination practices provision, offering a similar analysis to that one provided for
Babydream restrictions (see paragraph 22).
Page 35
Attachment B
38. For other members of the project group, an effects-based approach is needed to
assess whether the impact of Sport 2001 restrictions is detrimental on retail
competition as described in paragraph 10 above, points a), b) and c).
39. One member applying the abuse of dominance framework highlighted that it would be
necessary to assess whether: i) Sport 2001 has market power in the upstream market
according to the leveraging theory mentioned in paragraph 10 c) above; ii) the intent
behind Sport 2001 restrictions is to exclude online third-party platforms in the
downstream markets; and, iii) the effects of the conduct.
40. The effects-based approach would require further analysis on factors listed below at
paragraph 41.
Areas of further inquiry

41. For a more comprehensive effects-based analysis of the impact of Sport 2001
restriction on retail competition, project group members highlighted several factors
requiring further investigation which can be grouped as follows:
a) Analysis of the intra-brand competition for the Sport 2001 high-end products: the
competitive constraints posed by retailers’ own websites (which are free to
advertise without constraint); the extent to which consumers buying high end
products switch from third-party platforms to other online channels; the impact of
the ban on retail prices (e.g., to check whether the prices set by Sport 2001 on the
third-party platforms may possibly influence the prices offered on other sales
channel since it holds more than 30% of market shares); the level of
substitutability between high-end and discounted products of Sport 2001 (e.g., the
lower substitutability the larger the potential anticompetitive effects on intra-brand
competition).
b) Analysis of inter-brand competition and competition across distribution channels
as mentioned for Babydream (see paragraph 26 above).
c) Analysis of the intent and business justifications (see points c and d of paragraph
26).
42. In addition to this analysis of the retail markets, one member of the project group
applying the framework for abuse of dominance also indicated additional elements for
the analysis of dominance in the upstream market of the manufacturing, similarly to
those indicated for Babydream in paragraph 27.
Final remarks
43. Specific areas of further investigation have been already indicated in the previous
sections and include the definition of relevant markets (see paragraph 9) and the
assessment of potential and/or actual effects of the restrictions imposed by each
manufacturer.
44. At a more general level, one area of further investigation relates to the cumulative
effects on inter-brand competition of the restrictions put in place by the three
manufacturers. It is useful to recall that Babydream is not sold through platforms at all,
Wheelies is sold through platforms by both the manufacturer and retailer (as long as
retailer criteria are met), and only the Sport 2001 manufacturer sells all models of its
stroller through platforms while retailers cannot sell high-end models. In this context,
the following elements have been suggested:
Page 36
Attachment B
a) whether the baby strollers of Wheelies (whose sales are not restricted on third-
party platforms) provide a sufficient competitive constraint to the brands of the
other two manufacturers;
b) whether the costs to Babydream and Sport 2001 of foregoing or limiting sales
through third party platforms are outweighed by the benefits of reduced
competition among manufacturers;
c) to what extent the criteria for selecting authorised retailers within each distribution
system may differ across manufacturers and whether there are legitimate reasons
for any difference; and,
d) reasons why the assessments of whether such criteria can be met by the various
retailers may differ across manufacturers.
45. Finally, another area of further inquiry relates to the assessment of collective
dominance and this would require assessing whether a group of manufacturers holds
market power together and how vigorously they compete with each other.
Page 37
Attachment B
Attachment OSB – A
List of the project group members

This report is based on the assessments received from the following members and
NGAs:
Competition Agencies
 Administrative Council for Economic Defense – CADE, Brazil
 Australian Competition and Consumer Commission
 Bundeskartellamt, Germany
 Competition and Markets Authority, United Kingdom
 Competition Bureau, Canada
 Competition Commission, Hong Kong
 Department of Justice and Federal Trade Commission, United States
 European Commission
 Japan Fair Trade Commission
 Indonesian Competition Commission
 Swedish Competition Authority
 Turkish Competition Authority
NGAs
 Urška Petrovčič, NGA for the European Commission
The report has also benefited from additional comments from other agencies and
NGAs
 Office for the Protection of Competition, Czech Republic
 Grant Murray, Baker & McKenzie LLP, NGA UK
 Alvaro Ramos, Qualcomm, Spain NGA
 Rachel Brandenburger, EU NGA
 Silvia Fagá de Almeida, Director, LCA Consultores, Brazil NGA
Page 38
Attachment B
Attachment OSB – B
Online Sales Bans
Hypothetical Scenario
Hypothetical Scenario – online sales bans
Facts of the Hypothetical
- Baby strollers are an important product for consumers:

o Most consumers will purchase baby strollers only once or twice.
o Brand image plays a significant role in consumer decision-making.
o Safety is very important to consumers.
o Pre- and post-sales service are deemed important (e.g. safety, ease of use) by
consumers.
- Broad range of products on the market
o Each manufacturer has products in different price ranges, from discount to luxury
products.
o Branded and white label products (a white label or private label is manufactured
by one company and packaged and sold by other companies, e.g. retailers under
various brand names).
- Main manufacturers: Babydream, Wheelies & Sport2001
o Wheelies and Sport2001 each have a market share exceeding 30 % and
Babydream has a 20 % market share; other niche companies exist that have
market shares <5%.
o assume market is national.
o Manufacturers invest heavily in branding, advertising, online presence and
product development.
- Retail landscape
o 5 large national retail chains with strong brick and mortar presence .
o limited number of independent shops.
o Two large 3rd party online platforms (Tiber and eHarbor).
o and many more smaller online retailers.
o The main stroller manufacturers also have their own websites through which they
sell, as described below.
o brick & mortar stores are considered important by manufacturers for pre- and
post-sales services and building brand image.
- Sales
o Total offline sales account for 65% of total baby stroller turnover.
o Online stroller sales are increasing - for 2017 they accounted for 35% of turnover
(30% in 2016 and 20% in 2015).
o 40% of total online sales in this sector are made through 3rd party platforms and
60% through manufacturers and retailers own websites. (This was 30%-70% in
2016 and 20%-80% in 2015.).
Page 39
Attachment B
Online Sales Bans
Selective Distribution
- Main manufacturers Babydream, Wheelies & Sport2001 all sell directly to consumers
and through retailers. All three manufacturers sell through each of the five large retail
chains. However, they do not all sell through all of the independent retail shops (i.e.,
each independent shop may sell only one or two of the three manufacturers
products).
- They all use Selective Distribution, that is a distribution system whereby a
manufacturer enters into (vertical) agreements with a number of retailers, selected on
the basis of certain qualitative criteria set by the manufacturer. According to these
agreements, selected retailers are authorised to sell manufacturer’s products only
directly to final consumers or to other authorised retailers of the same manufacturer.
- All distribution agreements contain criteria which require retailers to take steps to
protect and promote the brand’s image, e.g. high quality presentation of the strollers,
and invest in safety demonstrations (i.e., provision of explanations of the features of
the brand and safe operation).
- All retailers employed by the three manufacturers are authorised distributors
according to their selective distribution agreements.
- Selective distribution does not involve exclusivity: all large retailers carry all three
manufacturers’ brands. Some receive better discounts from a particular manufacturer
for historical reasons, or because of their willingness to invest in safety
demonstrations. However, these distribution agreements do not contain any pricing
policy terms (e.g., minimum advertising price or retail maintenance price).
- Retailers are not required to have a brick & mortar store to meet the selective
distribution criteria, however they should meet equivalent qualitative criteria in an
online environment.
Babydream (20% market share) selective distribution agreements:

- Retailers are allowed to sell Babydream strollers online using their own
webstore/app.
- Retailers are not allowed to sell Babydream’s stroller on 3rd party platforms (e.g.
Tiber and eHarbor); similarly, they are not allowed to use price comparison websites.
- Babydream has its own online distribution channel but does not sell on 3rd party
platforms either.
- Additionally, retailers are not allowed to online advertise Babydream brand (including
on their own website8), and in particular they are prohibited from using search
advertising (i.e., from using/bidding on search ads for Babydream brand name).
- Babydream considers the above restrictions important to maintain its brand quality
and reputation, and protect its investment in pre- and post-sales services.
Wheelies (>30% market share) selective distribution agreements:
8
That is, retailiers can sell Babydream products through their webstore/app but only if consumers actively search for them e.g.
they are not able to promote them on their own webstore/app.
Page 40
Attachment B
Online Sales Bans
- Retailers are allowed to sell online using their own webstore/app.
- Retailers are allowed to sell Wheelies’ products on 3rd party platforms (e.g. Tiber and
eHarbor), but only if the 3rd platform meets the same qualitative criteria as the
retailer.
- Retailers are allowed to use price comparison websites.
- Wheelies has its own online distribution channel and sells on 3rd party platforms
under the same conditions as its retailers.
Sport 2001 (>30% market share) selective distribution agreements:

- Retailers are allowed to sell online in their own webstore/app.
- Retailers are allowed to sell some of Sport 2001’s models (mainly lower end products
or discontinued models) on 3rd party platforms.
- Retailers are allowed to use price comparison websites.
- Sport 2001 has a strong presence on the major 3rd party platforms and sells its entire
catalogue online, including on 3rd party platforms.
- Sport 2001 argues that it needs to control the environment on 3rd party platforms by
itself in order to protect its brand and that retailers are unable to do that.
Assume: all three selective distribution agreements are generally considered legal, except
for the questions regarding the platform ban clauses they contain. They are summarised
below.
Page 41
Attachment B
Online Sales Bans
Page 42
Attachment B
Online Sales Bans
Stroller Platform Ban Task Force
You have been chosen to head up the Stroller Platform Ban Task Force to investigate
complaints your authority received from various retailers:
- Some retailers complain that not being allowed to sell on 3rd party platforms prevents
them from competing online. Some prefer 3rd party platforms as it is more difficult and
expensive to attract consumers to their own websites, especially if online advertising
is also restricted.
- Several retailers of Babydream products complain that the absolute online
advertising ban (and in particular the search advertising ban) significantly limit their
visibility to potential purchasers in the online environment, thus compromising more
generally their ability to compete against the other retailers.
- Some retailers complain that it is unfair that Sport2001 can sell its entire catalogue
on 3rd party platforms but retailers are restricted to a smaller part of the catalogue
(generally lower margin products). They consider this cherry picking by their
manufacturer which doubles as a competitor as a retailer. They argue that Sport
2001’s intent is to better “control” prices and discounts being offered in the online
environment.
***
You are invited to assess the compatibility with the competition law of your jurisdiction
and in light of the facts of this hypothetical scenario of the 3rd party platform ban on
sales and the absolute ban on online advertising (including search advertising), by
highlighting in particular:
- their potential anti-competitive effects (theories of harm);
- their potential pro-competitive effects and/or business justifications, and
- areas of further inquiry by your Task Force.
You are encouraged to have regard to the ICN’s Unilateral Conduct Workbook as it
relates to their analysis.
A report presenting this analysis will be prepared for consideration by the Unilateral
Conduct Working Group.
Page 43

UCWG Vertical Restraints Project

Uploaded by

Copyright:

Available Formats

UCWG Vertical Restraints Project

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

UCWG Vertical Restraints Project

Uploaded by

Copyright:

Available Formats

Unilateral Conduct Working Group

Vertical Restraints Multi Year Project

2. The vertical restraints project has three distinct stages

 an interest in a focus on online vertical restraints, but caution to not overlook

 support for theoretical analysis of issues arising with vertical restraints,

 support for analysis that focuses on particular forms of restraint, and

 support for the development of a selected case study resource.

 Development phase: design of hypothetical vertical restraint scenarios (parity

A project group of ICN members and NGAs provided individual/jurisdictional-

A summary of the hypotheticals is provided below.

 Completion phase: to be carried out in 2019-2020, the completion phase will

 webinar discussions of the key points that emerged in the hypotheticals;

 adaption of the hypotheticals into training modules to be made available

 developing a workbook chapter on vertical restraints.

Hypothetical One – online parity requirements

Theories of harm – conduct of BestValueBed

 Parity requirements may prevent free-riding, by competing OTAs and

Areas of further inquiry by investigators

 Parity clauses: a range of matters including evidence of the scope, market

 Other: a range of matters including features of hotel websites relative to OTA

Hypothetical Two –online selective distribution models

For some other members, an effects-based approach is needed to assess whether

Wheelies “equivalent” criteria required for selling through third-party platforms

Sport 2001 ban on selling high-end products through third-party platforms

For some other members, an effects-based approach is needed to assess whether

Areas of further inquiry by investigators

Vertical Restraints Project

Hypothetical One – Online Parity Requirement

4. The project group has been invited to consider

1. possible theories of harm

2. arguments about pro-competitive effects that may be presented and factors

3. areas of further inquiry by investigators.

 BestValueBed (BVB) – an online travel agent (OTA), 25 per cent of online

 Find-A-Room (FAR) – a competing OTA, 40 per cent of online

 Accommodation providers – in the hypothetical scenario one accommodation

Overview of online parity requirements

 A market for the provision of online travel agency services to accommodation

 accommodation providers who wish to attract customers to their

 consumers seeking to book accommodation.

 Local markets in which accommodation providers supply various types of

 Separate markets: some members of the project group considered that

It was also noted that in some jurisdictions a distinction is drawn between

Theories of harm – conduct of BestValueBed

b) An accommodation provider has alleged that BVB is requiring accommodation

Effect on accommodation providers

View 1: limited or no detrimental impact on competition

Narrow room rate parity requirements

Wide room inventory parity requirements

View 2: detrimental impact on competition

Narrow room rate parity requirements

Wide room inventory parity requirements

View 1: limited or no detrimental impact on competition

Narrow room rate parity requirements

Wide room inventory parity requirements

View 2: detrimental impact on competition

Narrow room rate parity requirements

Wide room inventory parity requirements

Pro-competitive effects and factors relevant to these

Areas of further inquiry