likeness

"The existing (WTO) case-law on non-discrimination indicates that when defining likeness for the purpose of assessing regulation, one should rather focus on the interchangeability of the services by applying a cross-price elasticity test. In Japan – Alcoholic Beverages II, the Panel has stated explicitly that “the appropriate test to define whether two products are ‘like’

"likeness" Resources:

The following is from this very useful thesis

Starting at around Page 199.

GATS obligations based on discrimination
4.1 Introduction
483.
As has been stated already in Part II, the primary goal of the international trade
regime is to eliminate protectionism. 707 Defining ‘protectionism’ requires a number of tools or
‘proxies’ to avoid the adjudicator from getting entangled in the difficult process of defining
the ‘mens rea’ of a regulator. A major proxy for determining whether protectionism is at stake
is to define whether the measure at stake ‘discriminates’. Nevertheless, discrimination is a
proxy that is in many aspects flawed. Even if a measure has a discriminatory effect, it may not
necessarily be protectionist. Moreover, even measures that are not discriminatory may still
constitute an important barrier to cross-border trade.

To discipline protectionism in the area of services, GATS provisions prohibit WTO
Members to apply measures that are discriminatory on the basis of origin. 708 To the extent that

banking regulation gives rise to discrimination on the basis of the origin of banking services
or suppliers of banking services, there may be a violation of the GATS. Despite some
commonalities, the two main principles prohibiting discrimination: most-favoured-nation
treatment (MFN) and national treatment apply to a different extent in the GATS. They
involve two different situations of discrimination. The principle of MFN concerns
discrimination by a WTO Member between different trading partners. The principle of
national treatment addresses discrimination between domestic services or suppliers, on the
one hand, and foreign services or suppliers on the other hand. Moreover, besides the fact that
the provisions have a different content, they also do not apply to the same extent to the WTO
Members. The principle of MFN is included in Part II of the GATS, containing general
obligations and disciplines. This means that this MFN obligation applies immediately to all
WTO Members. In contrast, the national treatment obligation is included in Part III of the
GATS, titled ‘Specific Commitments’. Like the market access provision, the national
treatment obligation only applies as far as the WTO Member at stake has made a specific
commitment in a specific service sector to subject this sector to the national treatment
obligation.
485.
Before considering the individual particularities of the MFN and national treatment
provisions further in detail, two issues will be considered that are common to MFN and
national treatment: the concept of ‘likeness’ and the phrase ‘less favourable treatment’. It is
not always obvious whether a regulation at issue gives rise to a violation of the provisions
prohibiting discrimination or whether it creates a barrier to trade without being
discriminatory, in which case the provisions on non-discriminatory barriers apply (Article VI
GATS and the Disciplines on Domestic Regulation). 709 Still, because both sets of provisions
set out different tests of trade scrutiny, it is important to categorise banking regulation under
one of these provisions. It is through the interpretation of the concept of ‘likeness’ and the
phrase ‘less favourable’ that case-law and authors have tried to define this boundary.
486.
It must again be stressed that in interpreting the concepts ‘likeness’ and ‘less
favorable treatment’, the present research will stay as close as possible to the existing WTO
case-law. It will appear from the following discussion that the interpretation of these concepts
has been heavily criticized. What is more, the case-law relates mainly to trade in goods. It is
therefore quite likely that future disputes with regard to trade in services will give rise to
decisions that change earlier case-law with regard to goods. That said, the panels and
Appellate Body have stressed certain interpretations with regard to ‘likeness’ and ‘less
favorable treatment’ in GATT quite strongly and have repeated these statements in the few
cases relating to GATS. It is for this reason that the analysis will in the first place rely on the
existing case-law. However, it will also be indicated where alternative interpretations have
been suggested.
4.2 Likeness and less favobanking regulation gives rise to discrimination on the basis of the origin of banking services
or suppliers of banking services, there may be a violation of the GATS. Despite some
commonalities, the two main principles prohibiting discrimination: most-favoured-nation
treatment (MFN) and national treatment apply to a different extent in the GATS. They
involve two different situations of discrimination. The principle of MFN concerns
discrimination by a WTO Member between different trading partners. The principle of
national treatment addresses discrimination between domestic services or suppliers, on the
one hand, and foreign services or suppliers on the other hand. Moreover, besides the fact that
the provisions have a different content, they also do not apply to the same extent to the WTO
Members. The principle of MFN is included in Part II of the GATS, containing general
obligations and disciplines. This means that this MFN obligation applies immediately to all
WTO Members. In contrast, the national treatment obligation is included in Part III of the
GATS, titled ‘Specific Commitments’. Like the market access provision, the national
treatment obligation only applies as far as the WTO Member at stake has made a specific
commitment in a specific service sector to subject this sector to the national treatment
obligation.
485.
Before considering the individual particularities of the MFN and national treatment
provisions further in detail, two issues will be considered that are common to MFN and
national treatment: the concept of ‘likeness’ and the phrase ‘less favourable treatment’. It is
not always obvious whether a regulation at issue gives rise to a violation of the provisions
prohibiting discrimination or whether it creates a barrier to trade without being
discriminatory, in which case the provisions on non-discriminatory barriers apply (Article VI
GATS and the Disciplines on Domestic Regulation). 709 Still, because both sets of provisions
set out different tests of trade scrutiny, it is important to categorise banking regulation under
one of these provisions. It is through the interpretation of the concept of ‘likeness’ and the
phrase ‘less favourable’ that case-law and authors have tried to define this boundary.
486.
It must again be stressed that in interpreting the concepts ‘likeness’ and ‘less
favourable treatment’, the present research will stay as close as possible to the existing WTO
case-law. It will appear from the following discussion that the interpretation of these concepts
has been heavily criticised. What is more, the case-law relates mainly to trade in goods. It is
therefore quite likely that future disputes with regard to trade in services will give rise to
decisions that change earlier case-law with regard to goods. That said, the panels and
Appellate Body have stressed certain interpretations with regard to ‘likeness’ and ‘less
favourable treatment’ in GATT quite strongly and have repeated these statements in the few
cases relating to GATS. It is for this reason that the analysis will in the first place rely on the
existing case-law. However, it will also be indicated where alternative interpretations have
been suggested.


4.2 Likeness and less favourable treatment

487.
Since the principles of most-favoured-nation treatment and national treatment are
both expressions of the more general principle of non-discrimination, it is in both cases
necessary to perform a two-step analysis. The first step requires a consideration of the extent
to which the services or service suppliers are like. The second step involves a determination
to what extent certain services (or suppliers) are treated less favourable than others.

.........

C HAPTER III.2 L IMITATIONS ON THE RIGHT OF WTO M EMBERS TO REGULATE THE BANKING SECTOR
4.2.1 Likeness
488.
The concept of ‘likeness’ is one that has been interpreted extensively in the field of
trade of goods, 710 but it is also a cornerstone of the provisions that prohibit discrimination in
the GATS. Indeed, if one applies a non-discrimination standard, it is necessary to determine
whether like situations are treated in the same way and different situations in a different way.
Defining ‘likeness’ of the services or service suppliers is thus the first step in the analysis. To
the extent that the services or service suppliers are not considered ‘like’, there is no need for
further assessment on whether some are treated less favourable than others.
489.
It should preliminary be noted that the principles of most-favoured-nation treatment
and national treatment in the GATS concern the treatment of both services and service
suppliers. This stands in contrast with the GATT, where Article I and Article III GATT only
concern the treatment of goods. This may also be contrasted with the EU, where separate
provisions exist for the freedom of establishment of service suppliers and for the free
movement of services. 711 The simultaneous reference to services and service suppliers
seriously broadens the scope of the disciplines on discrimination in the GATS, but is
understandable given the way the service is often closely intertwined with the service
provider itself. This has also become clear in the determination of likeness of service
suppliers, where the Panel in Canada – Autos made a reference to the likeness of the services
they provide. The simultaneous reference is also linked to the fact that the GATS defines
‘trade in services’ on the basis of four modes of supply and not solely on the basis of whether
(long-term) establishment is created or not.
a) Likeness of Services
490.
The criteria that are used to determine the likeness of services are not much
elaborated upon in dispute settlement on trade in services. The Panel in EC – Bananas II

mentioned the nature and characteristics of the services as criteria. 712 This seems to suggest
that the same criteria are used as for goods, namely end-uses, consumer tastes and habits and
product characteristics (the so-called ‘Border Tax Adjustment criteria’). A further criterion
used for determining likeness of goods is the classification of the goods in the ‘Harmonized
Commodity Description and Coding System’, developed by the World Customs
Organization. 713 Nevertheless, several authors have noted that the criteria for goods are of
limited use for services. Indeed, it makes often no sense to define the physical characteristics
of something immaterial like a service. Neither is the classification of a service in the Central
Product Classification System (created by the United Nations) 714 of much use since this
System is much less developed than the Harmonised System for goods. 715
491.
The existing case-law on non-discrimination indicates that when defining likeness for
the purpose of assessing regulation, 716 one should rather focus on the interchangeability of the
services by applying a cross-price elasticity test. 717 In Japan – Alcoholic Beverages II, the
Panel has stated explicitly that “the appropriate test to define whether two products are ‘like’

or ‘directly competitive or substitutable products’ is the marketplace”. 718 This was cited with
approval by the Appellate Body 719 and repeated in later reports. 720 The marketplace is “the
forum where consumers choose between different products.” 721 While these statements related
to internal taxation in the framework of Article III:2 GATT, the Appellate Body in EC –
Asbestos noted that “a determination of "likeness" under Article III:4 [of GATT] is,
fundamentally, a determination about the nature and extent of a competitive relationship
between and among products.” 722 If it can be established, e.g. through consumer surveys, that
a sufficiently large amount of consumers will switch to a substitute if the price of the product
under consideration increases with a small but significant amount, this may show that the
consumers consider the services as interchangeable and thus as substitutes. Nevertheless, with
regard to trade in goods, the Appellate Body has stated that the cross-price elasticity-test is
not “the decisive criterion for assessment of whether products are directly competitive or
substitutable.” 723 It is merely one of the means of determining the competitive relationship of
goods. Still, as the traditional criteria that focus on physical characteristics of products are of
limited use for services, the cross-price elasticity test may be the first and primordial step in
the analysis of the likeness of services (or service suppliers).
492.
Nevertheless, one may question whether the finding of low cross-price elasticity
between two services is due to the fact that the services are indeed not interchangeable or
rather because the consumers only perceive the services as different because of years of
‘indoctrination’ through regulatory distinctions in government regulation and information
provision. 724 One may for instance think of situations where governments have for years
consistently warned of the dangers of certain foreign banking services. Furthermore,
consumers may have a bias against banks from certain countries merely because of their
origin, even if the regulatory framework in which the banks operate is largely similar.
Consumers may prefer (‘trust more’) banking services of a certain origin (especially
domestic) over services of another origin. If that is the case, different credit institutions will
be considered by consumers as ‘unlike’ and consequently, following the reasoning that
dissimilar situations can be treated different, the national regulator may discriminate among
them. In other words, this may result in ‘under-inclusiveness’ of the test to determine
discrimination. This shows that consumer perceptions may not always be an acceptable basis
for determining likeness.
493.
This argument was considered by the Appellate Body in Korea – Alcoholic
Beverages and EC – Asbestos, in the context of trade of goods. The Appellate Body stated
that evidence may be produced of latent consumer demand that is suppressed by regulatory
barriers. 725 The problem, however, is how to construct this ‘latent demand’. One will thus
need to construct a reasonable argument that there is, in fact, potential competition, even if  this is not immediately seen in present consumer behaviour. 726 Furthermore, the Appellate
Body noted that it is also possible to submit “evidence of substitutability from some relevant
third market”. 727 This would arguably lead to an objective and abstract test of determining
likeness from the perspective of a ‘reasonable consumer’. H OWSE and T ÜRK note that the
Appellate Body in EC – Asbestos applies a test from the perspective of consumer tastes and
habits in an idealised market place, where consumers have full information and negative
externalities have already to some extent been internalised. They argue that this in fact brings
the regulatory aims back in the picture, which were in the GATT 1947-cases considered under
the ‘aim and effects’-test. However, this approach was rejected in the case-law since the
WTO was established (below 496). The difference would be that here the perspective of
consumer behaviour is taken. 728 H OEKMAN and M AVROIDIS have argued that the Appellate
Body in EC – Asbestos implicitly applied a ‘reasonable consumer’-test, which had as
implication that it was not necessary to look for factual evidence of such consumer
perceptions in the market. 729
494.
Eventually, this test would lead to more services and service providers to be
considered as ‘like’ and thus stretches the scrutiny of the provisions that prohibit
discrimination in the GATS. However, since the WTO Agreements are concerned with
equality of competitive opportunities, the focus on a broader test of likeness, which includes
also potential competition, seems appropriate. If over-inclusiveness would follow from this
‘stretching’, this may be remedied through the second tier of the non-discrimination test,
where, as will be argued further, merely a detrimental effect on certain ‘like’ services or
service suppliers is not sufficient to find a violation. Moreover, in this second tier of the test,
the market will also be considered from the perspective of producers and thus attention will
be paid to supply substitutability.
495.
A related, but in fact opposite, problem with relying on consumer perceptions to
define the ‘likeness’ of services is that most consumers may not be aware of the risks linked
to a particular foreign service. For instance, with regard to a service provided by a foreign
bank, the consumer may not be aware of the quality of the regulatory framework in which this
bank is operating and the supervision performed in the home country. Hence, often they
cannot give due account to this fact in their decisions on the purchase of the banking service.
Financial services are a sector par excellence where information asymmetries play an
important role (see above 97-106). Hence, since consumers would consider the services
offered by e.g. a US and Chinese banking service supplier as ‘like’, the national authorities
would not be able to treat the suppliers as different, even if the authorities consider that they
give rise to different risks. 730 However, regulatory intervention draws distinctions precisely

because of the fact that some characteristics are not reflected in the market place and thus
seem not perceived by consumers as relevant. 731 Yet, when applying the aforementioned test
on what concerns a ‘like’ service or supplier, such regulatory distinction would risk not being
consistent with the principles that prohibit discrimination, since consumers consider the
banking service or service supplier as ‘like’. Consequently, WTO Members would not be
allowed to treat the ‘like’ services or suppliers in a dissimilar way. One could argue that the
result is an ‘over-inclusive’ outcome of the test: too many services would be considered ‘like’
by the consumers, even if there are, arguably, pertinent differences. 732
496.
It should be noted that during the GATT-era, a GATT Panel developed the so-called
‘aim and effects’ test, which implied that the aim and the effect of a measure would be
considered when comparing domestic and imported products. 733 Hence, two products would
not be considered to be ‘like’ when the regulatory aim underlying the distinction is taken into
account. Since the products would not be like, there could be no discrimination. Nevertheless,
once the WTO was established, the Appellate Body rejected the use of intent to determine
whether the products are different. 734 The main reason stated by the Appellate Body for the
rejection was that panels could not sort through all the subjective reasons a national regulator
may have. Moreover, it may be argued that by considering the aim of a measure when
assessing likeness, the regulatory distinctions may become part of the likeness criteria and
thus likeness is in fact determined by the State and not by the market. On the other hand, it is
clear that government intervention is precisely necessary where the market fails to make the
necessary distinctions. In any event, under present case-law, the non-protectionist regulatory
aims of measures are not taken into account to determine ‘likeness’. Also with regard to trade
in services, the Appellate Body in EC – Bananas III saw “no specific authority either in
Article II or in Article XVII of the GATS for the proposition that the "aims and effects" of a
measure are in any way relevant in determining whether that measure is inconsistent with those
provisions”. 735 The Appellate Body stated that the ‘aims and effects’ doctrine in GATT was
based on the phrase in Article III:1 GATT that national measures (taxation and regulation)
“should not be applied to imported or domestic products so as to afford protection to domestic
production”. According to the Appellate Body, the absence in the GATS of a reference to “so as
to afford protection”, made it clear that the aim of a measure should not be considered here.