18.1 PP 447 484 The Scope of The SCM Agreement Specific Subsidies

15
The scope of the SCM Agreement: specific subsidies
A specific subsidy exists under the SCM Agreement when a government

makes a financial contribution or provides income or price support
that confers a benefit on a specific recipient. The case law developed
the ‘private market test’ to detect whether a financial contribution (or
income or price support) indeed benefits a specific recipient, though the
Appellate Body in the recent Canada – Renewable Energy / Feed-in Tariff
Program case partly deviated from this approach. The specificity test
filters out whether a specific recipient has received such a contribution
(or support) at better than market terms.
In general, three lines of concern have been articulated on this
‘specific subsidy’ definition as laid down in Articles 1 and 2 of the SCM
Agreement and further drawn upon in the case law. First, not all govern-
ment measures with similar trade-distorting effects are disciplined in
the same way. Second, subsidies with a corrective rather than distortive
effect on markets are disciplined similarly to those having no objective
of correcting market failures. This concern seems to have inspired
the Appellate Body in Canada – Renewable Energy / Feed-in Tariff
Program to deviate from the private market test. Third, by disregarding
the effect on the recipient’s competitive situation, measures having no
trade-distorting effect could be covered under the subsidy definition, and
privatization might extinguish subsidization even though the trade-
distorting effect is still present.
Before addressing these three concerns, recall that the SCM Agreement
does not outlaw the provision of specific subsidies as such. The definition
serves a double purpose: it acts as a gateway to the substantive disciplines on
subsidies but at the same time limits the measures against which members
could take CVD action. Part of the criticism related to the subsidy definition
might be better tackled under the disciplines imposed on such subsidies.
Other lines of concern seem to be unavoidable in the light of the subsidy
definition’s threshold function.
447
https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

448 normative analysis
15.1 The financial contribution element

15.1.1 The closed list of government interventions
Article 1 of the SCM Agreement offers an exhaustive list of government
interventions that could qualify as a ‘subsidy’ under the SCM Agreement
if they confer a benefit. In the words of the Appellate Body, the financial
contribution element ‘involves consideration of the nature of the trans-
action through which something of economic value is transferred by the
government’ and the panel in US – GOES considered that the focus of
‘income or price support’ should also be ‘on the nature of government
action, rather than upon the effects of such action’.1 Generally speaking,
the case law offered a rather expansive interpretation of the three types
of financial contribution (Article 1.1(a)(1) of the SCM Agreement),
whereas they endorsed a narrow reading of ‘income or price support’
(Article 1.1(b)(2) of the SCM Agreement). Governments could subsidize
by positive action when transferring monetary or non-monetary resour-
ces directly or indirectly through private actors. Alternatively, they could
subsidize by negative action when refraining from collecting revenue
otherwise due. Only two types of government intervention are explicitly
excluded from the subsidy definition: the provision of general infra-
structure, which in essence relates to the specificity element, and border
tax adjustments on indirect taxes and import duties.2
The closed list given in Article 1 implies that other government
measures generating similar effects are not covered under the SCM
Agreement.3 For instance, the panel in US – Export Restraints correctly
concluded that export restraints are not captured, even though such
measures could very well benefit domestic producers (in this case down-
stream producers) in a way similar to when a financial contribution is
offered.4,5 Moreover, the financial contribution element does also
1
Appellate Body Report, US – Softwood Lumber IV, para. 52 (emphasis added); panel
report, US – GOES, para. 7.85.
2
See Article 1.1(a)(1)(iii) of the SCM Agreement and footnote 1 of the SCM Agreement.
3
Recall that export subsidies included in the Illustrative List are prohibited, regardless of
whether they are covered under Article 1 of the SCM Agreement.
4
Panel Report, US – Export Restraints, para. 8.75.
5
Janow and Staiger offer an alternative reasoning on which ground the US claim could be
rejected. Applying the Lerner theorem, they indicate that an export tax is equivalent to an
alternative programme in which an export subsidy of the same magnitude is placed on every
other export good and an import tariff of the same magnitude is placed on each imported
good. On this basis, they argue that an export tax confers a subsidy to production in every

scope of the scm agreement: specific subsidies 449
exclude so-called ‘regulatory subsidies’ from the scope of the SCM

Agreement. This refers to the failure of a government to provide certain
levels of regulation on, for instance, environmental protection or labour
standards.6 Such negative action by the government cannot be labelled as
a subsidy because it falls outside the closed list.7 This implies that such
‘regulatory subsidies’ (so-called ‘social dumping’) cannot be counter-
vailed by importing countries.8 For example, developing countries could
implement the 2008 Growth Report’s suggestion to allow export-
oriented firms to recruit workers on easier terms (e.g., in export process-
ing zones) than those prevailing in the formal sector so as to overcome
labour market failures (i.e., surplus labour) without any risk of unilateral
or multilateral action by other members.9 Lastly, as is further explained
below, an undervalued exchange rate (so-called ‘exchange dumping’)
does not, according to most scholars, qualify as a specific subsidy under
the SCM Agreement, even though it has an effect similar to an export
subsidy across the board.10
Hence the relevance of the closed list included in the SCM Agreement
is its limiting of the scope of government measures that can be counter-
vailed and are disciplined under the SCM Agreement. Indeed, as stressed
other sector of the economy. Therefore the panel could have rejected the US claim on the
basis of the specificity test. Yet their argument seems inaccurate in legal terms because, even if
the export tax generates these non-specific benefits to all other sectors, it still confers a specific
benefit for downstream domestic producers, as they benefit from a reduced domestic price.
Hence the specificity test will still be passed. As Janow and Staiger acknowledge as well, their
equivalence argument ‘does not for example rule out the possibility than an export tax on logs
would have a large expansionary impact on the volume of exports of logs’. Sykes also
indicates that, from an economic point of view, Janow and Staiger’s view does not refute
the argument that an export tax benefits downstream producers. J. Janow and R. W. Staiger,
‘US – Export Restraints, United States – Measures Treating Export Restraints as Subsidies
(WT/DS194; DSR 201:XI, 5767)’, in H. Horn and P. Mavroidis (eds.), The American Law
Institute Reporters’ Studies on WTO Case Law – Legal and Economic Analysis (Cambridge
University Press, 2007), 214–48, at 242–4; A. O. Sykes, ‘The Questionable Case for Subsidies
Regulation: A Comparative Perspective’, 2:2 Journal of Legal Analysis (2010), 473–523.
6
See, e.g., Jackson, above Chapter 2 n. 57, at 296; Mavroidis, Messerlin, and Wauters,
above General introduction n. 3, at 303; M. Schlagenhof, ‘Trade Measures Based on
Environmental Processes and Production Methods’, 29:6 Journal of World Trade (1995),
123–55, at 145–6.
7
It could also not be considered as income or price support under Article 1.1(a)(2) of the
SCM Agreement.
8
During the initial GATT negotiations such ‘social dumping’ was already being discussed
(E/PC/T/C.II/48, 11 November 1946; E/PC/T/C.II/54, 16 November 1946).
9
Commission on Growth and Development, The Growth Report – Strategies for Sustained
Growth and Inclusive Development (Washington, DC: World Bank, 2008), at 45–8; see
also Torres, above Chapter 6 n. 20, at 218.
10
See below Chapter 17, section 17.3.4.

by the panel in US – Export Restraints, the ‘financial contribution’

requirement was precisely advocated by most countries to counter the
purely effect-based definition of the United States, under which any
government measure having the effect of benefiting domestic producers
could in theory be countervailed.11 By narrowly interpreting the scope of
‘income or price support’, the panel in US – Export Restraints also aimed
to foreclose an effect-based approach being introduced under this alter-
native for the subsidy definition.12 The focus is again on the nature of the
government intervention, and situations where the movement in prices
(or income) are only an indirect effect of another government interven-
tion (tariffs, quantitative restrictions) do not qualify as price (or income)
support. The panel’s narrow reading of the scope of ‘income or price
support’ fits the architecture and purpose of the subsidy definition better
than an expansive interpretation which reintroduces an effect-based
approach and reads the carefully drafted closed list of financial contri-
butions out of the Agreement.13 Thus the rejection of an effect-based
approach is legally sound. After all, if ‘all government measures capable
of conferring benefits would necessarily fall within Article 1.1(a)’, there
‘would be no need for Article 1.1(a), because all government measures
conferring benefits, per se, would be subsidies’.14 At the same time, such a
closed list carries the risk that members might shift to government
measures that do not qualify as a subsidy but that generate equivalent
effects.15
Yet the list as interpreted in the case law is sufficiently broad to capture
the most common forms of subsidies. Indeed, as underscored by the
Appellate Body, ‘a wide range of transactions falls within the meaning of
“financial contribution” in Article 1.1(a)(1)’.16 Next, one might wonder
whether those government actions falling outside its current scope really
ought to be disciplined under the SCM Agreement. First, it is important
11
Panel Report, US – Export Restraints, para. 8.75. During the Uruguay Round, the United
States proposed to define the term ‘(actionable) subsidy’ as ‘any government action or
combination of actions which confers a benefit on the recipient firm(s)’. See Submission
by the United States, Elements of the Framework for Negotiations (MTN.GNG/NG10/
W/29, 22 November 1989), section II.1(a). See also Appellate Body Report, US –
Softwood Lumber IV, para. 52, n. 35.
12
Panel Report, US – Export Restraints, paras. 8.63–8.72. See above Chapter 3, section 3.1.3.
13
An expansive, effect-based reading would not fit the minimal attention, if any, given by
the negotiators to the inclusion of ‘income or price support’.
14
Appellate Body Report, US – Softwood Lumber IV, para. 52, n. 35.
15
See, e.g., Rubini, above Chapter 3 n. 8, at 160.
16
Appellate Body Report, US – Softwood Lumber IV, para. 52.

to understand that incentive schemes not qualifying as a subsidy will

often fall within the ambit of the GATT. Article III of the GATT outlaws
internal measures that discriminate against imported like products, and
the carve-out in paragraph 8 will not be available.17 Export quotas are
even in principle outlawed under Article XI of the GATT, whereas export
taxes could be scheduled under Article II of the GATT and could thus be
subject to tariff negotiations.18 Moreover, as Sykes indicates, allowing
export restraints to be labelled as ‘subsidies’ would generate difficulties in
quantification and raise the question whether any regulation that lowers
the price of inputs would be treated similarly.19 Second, with regard to
the so-called ‘social dumping’ or ‘exchange dumping’: what level of
regulation or exchange rate would be appropriate (i.e., non-beneficial)
under the SCM Agreement? How would WTO adjudicating bodies or
members wishing to respect their WTO obligations define the appro-
priate benchmark in the absence of specific guidance?20 Rather than
bringing regulatory or exchange-rate action under the ambit of the
SCM Agreement, specific oversight preventing the risk of social or
exchange dumping would be better concentrated under other, more
specialized international organizations or agreements.
15.1.2 Subsidization by forgoing revenue otherwise due

The inclusion of subsidization by negative fiscal action recognizes that a
subsidy in its ordinary meaning is a synonym for a tax and that a negative
tax is thus simply equivalent to a subsidy. Subsidization by positive and
negative action generates similar effects, which explains why the latter is
also covered by the SCM Agreement.21
Although this inclusion is therefore solid on economic grounds,
articulating the appropriate benchmark for determining whether ‘rev-
enue is forgone’ is not a straightforward exercise. This is particularly
relevant for direct taxes (and social security charges), as no border tax
17
See also below Chapter 16, section 16.1.1.2.2.
18
On the regulation of export restraints see Mavroidis, above Chapter 4 n. 172, at 16–17,
42–62, 84–5. The US interest in qualifying export restrictions as subsidies was that CVD
action could be undertaken against subsidized downstream imports instead of bringing a
multilateral claim on the basis of the GATT. See MTN.GNG/NG10/W/4, 28 April 1987,
at 26–7.
19
Sykes, above n. 5, at 509–510.
20
See also R. E. Hudec, Essays on the Nature of International Trade Law (London:
Cameron May, 1999), at 263.
21
Appellate Body Report, US – Large Civil Aircraft, para. 811.

adjustments are allowed to level the playing field on either the import or
the export side. Countries predominantly relying on direct taxes, such
as the United States, hold that the economic rationale underpinning this
distinction between direct and indirect taxation is not convincing,
because direct taxes could very well be reflected in the final price.
Again, the criticism holds that the measure’s nature (i.e., indirect versus
direct taxation) rather than its potential effect determines whether it is
covered under the SCM Agreement. What is more, it has been argued
that the application of the relevant benchmark could even generate the
result that measures having the same nature (e.g., direct taxes) and
effect could, nonetheless, be disciplined differently. After all, the
relevant benchmark refers to the domestic fiscal system as relevant
benchmark. The concrete implications thereof are nicely illustrated by
the US defence in the US – FSC case. The United States argued that the
European territorial tax system has the same economic effect as the FSC
exemption from its world-wide tax system, because both tax systems
exempt the foreign-source income of exporters. After all, whereas under
a territorial tax system all foreign-source income is exempted from
taxation, a worldwide tax system in principle taxes all sources of income
and therefore needs an exemption to exclude foreign-source income
of exporters. The United States concluded that the WTO ‘should not
penalize a country using a world-wide system for incorporating
elements of a territorial system in order to obtain comparable tax
treatment for its exports’.22 Yet the EU objected that the United States
was responsible for having chosen a general tax system that puts its
exporters at a disadvantage, and this view was in essence followed by
the panel in US – FSC:
[T]he United States is free to maintain a world wide tax system, a territorial
tax system or any other type of system it sees fit. This is not the business of
the WTO. What it is not free to do is to establish a regime of direct taxation,
provide an exemption from direct taxes specifically related to exports, and
then claim that it is entitled to provide such an export subsidy because it is
necessary to eliminate a disadvantage to exporters created by the US tax
system itself. In our view, this is no different from imposing a corporate
income tax of, say, 75 percent, and then arguing that a special tax rate of 25
percent for exporters is necessary because the generally applicable corporate
tax rate in other Members is only 25 percent.23
22 23
Panel Report, US – FSC, para. 7.121. Ibid., para. 7.122 (original footnotes omitted).

Hudec commented that this ‘your own fault’ response by the panel
begs the following question: why should one government be allowed to
create subsidy-like tax effects in one case but not in another? In partic-
ular, why should the WTO reach a different result regarding a subsidy-
like tax exemption for exporters depending on whether the country also
exempts other foreign-source income?24 After all, the only difference
between both systems is that the territorial tax systems exclude all
foreign-source income, whereas the US FSC exemption merely excludes
foreign-source income of exporters. At first sight, there seems no eco-
nomic logic available to explain this different stance.25,26
Considered at a general theoretical level, however, such a different
outcome is inevitable, given that the benchmark is located in the domes-
tic legal order, which evidently varies among members. The example
given by the panel in US – FSC can serve as an illustration. Member A can
apply a generally applicable corporate tax rate of 75 per cent with an
exemption of 25 per cent for exporters, whereas member B might
provide a generally applicable corporate tax rate of 25 per cent.
Moreover, the domestic level is the only appropriate benchmark in the
absence of a level playing field. Indeed, the WTO adjudicating bodies
cannot rely on an agreed international benchmark (e.g., a common
corporate tax rate of 75 per cent), and the WTO itself is not a standard-
setting organization but is essentially a negative integration model. In the
words of the Appellate Body, ‘the SCM Agreement recognizes that WTO
Members are sovereign in determining the structure and rates of their
domestic tax regimes’.27 So the benchmark has to refer to the domestic
legal regime of the member in question. As a consequence, member B’s
tax system will probably fall outside the reach of the SCM disciplines,
whereas the tax exemption provided by member A might constitute a
prohibited (export) subsidy even if the economic effect upon exporters is
largely similar.28 It should be emphasized that the absence of an
24
Caution should be exercised when considering the territorial tax systems as ipso facto
SCM-compatible, given that the WTO did not yet have to decide on this (Panel Report,
US – FSC, para. 7.123).
25
See Hudec, above Chapter 4 n. 88, at 190.
26
A tax exemption confined to exporters might, however, be more trade distortive, because
its bears the risk of shifting resources from non-exporting domestic firms to exporting
firms. This is similar to the trade distortion rationale for the specificity test.
27
See Appellate Body Report, US – Large Civil Aircraft, para. 811.
28
Recall that a general tax cut implemented by a regional government is treated as non-
specific, whereas a similar regional tax cut installed by the national government would
be specific (above Chapter 3, section 3.3.3). Hence members A and B can impose an

international benchmark generates no similar effect with regard to

indirect taxes, because differences between such taxes could be adjusted
at the border in order for a level playing field to be created.
15.2 The benefit element

15.2.1 The market benchmark
The private market test relies on the market as the primary benchmark for
deciding whether a benefit is conferred. Two general lines of concern have
been articulated with regard to the application of this private market test,
which are discussed in turn. First, the private market test would disregard
whether the government pursues a legitimate policy objective by providing
financial contributions (or income/price support). This concern seems
to have inspired the Appellate Body to deviate from its established case
law on the private market test in the Canada – Renewable Energy / Feed-in
Tariff Program case. Second, the private market test looks at the financial
contribution in isolation from any other measure.
15.2.1.1 The Canada – Renewable Energy / Feed-in Tariff

Program case: feeding legitimate policy objectives into the
benefit analysis?
In Section 3.2.1.2 of Part I we explained the alternative approach to the
benefit analysis adopted by the Appellate Body in Canada – Renewable
Energy / Feed-in Tariff Program. The Appellate Body held that ‘Where a
government creates a market, it cannot be said that the government
intervention distorts the market, as there would not be a market if the
government had not created it’, and ‘the creation of markets by a govern-
ment does not in and of itself give rise to subsidies within the meaning of
the SCM Agreement’.29 Instead, when a market is created by the govern-
ment, the benchmark has to be found within the contours of this newly
created market and consists of ‘what a hypothetical market would yield’
within this market.30 In this section we reflect on the implications of this
decision for the benefit analysis under the SCM Agreement.
identical tax measure but whether this might be disciplined and countervailable depends
on their governance structure.
29
Appellate Body Report, Canada – Renewable Energy / Feed-in Tariff Program, para. 5.188
(emphasis in original).
30
Ibid., paras. 5.228, 5.233, 5.234.

15.2.1.1.1 The Appellate Body’s new approach to the benefit

analysis Deliberating in 2007, the panel in Japan – DRAMs (Korea)
observed that it was ‘now well established’ that the benefit concept is
‘defined by reference to the market, such that a financial contribution
confers a benefit when it is made available on terms that are more
favourable than the recipient could have obtained on the market’.31
This could be revealed by comparing the terms of the financial contri-
bution with the terms available to the recipient on the private market.
This comparison with a market benchmark perfectly fits the rationale
behind the benefit element, which is to act ‘as a screen to filter out
commercial conduct’.32 The benefit element is indeed included in the
SCM Agreement to exclude those financial contributions from the sub-
sidy definition that are made on commercial terms, in which case the
financial contribution does not alter the market outcome.
Two situations emerged in the case law where the market benchmark
for conducting the benefit analysis seems not readily applicable. First, in
several cases the Appellate Body explained that an alternative bench-
mark could be used when the market benchmark is distorted by the very
same financial contribution. This is an extension of its well-established
case law. Second, in Canada – Renewable Energy / Feed-in Tariff
Program, the Appellate Body found that when a government creates a
market, an alternative benchmark could also be used. This implied a
fundamental departure from its well-established case law.
The reason to allow recourse to an alternative benchmark is funda-
mentally different in both situations. On the one hand, when the govern-
ment intervention distorts the market benchmark, the application of the
traditional ‘market’ benchmark would lead to false negative findings: the
financial contribution is so dominant that it determines the market
benchmark (e.g., suppresses the price on the ‘private’ market), implying
that no subsidy would be present if the terms of the financial contribu-
tion were to be compared with this distorted market benchmark. On the
other hand, when the government intervention creates a market, the
application of the traditional ‘market’ benchmark would, according to
the Appellate Body’s argumentation, lead to false positive findings: the
creation of markets would in and of itself give rise to subsidies within the
meaning of the SCM Agreement – at least, if such creation is achieved
through a financial contribution (or income/price support).
31
Panel Report, Japan – DRAMs (Korea), para. 7.275.
32
Panel Report, Korea – Commercial Vessels, para. 7.28.

This illustrates why the first situation is an extension of the case law on
the benefit concept, while the second situation deviates from the ration-
ale underlying the benefit concept. The case law allowing an alternative
benchmark when market prices are distorted perfectly fits this rationale:
financial contributions should not be excluded from the subsidy defini-
tion if they are so dominant that they distort the market benchmark
itself. In this situation the government is not behaving as a commercial
actor, and an alternative ‘market’ benchmark could still be constructed
to establish subsidization. This alternative benchmark is still a ‘market’
benchmark: it serves to assess whether a financial contribution on
similar terms would be available in the absence of the financial contri-
bution. The analysis focuses on the market of the financial contribution,
which is conceptually the right focus for defining the market benchmark.
In contrast, the Appellate Body’s case law on ‘creating’ markets funda-
mentally departs from this rationale: here, the benefit analysis filters out,
not commercial conduct, but rather government support pursuing cer-
tain legitimate policy objectives that would, by definition, not be pursued
by commercial actors.
To achieve this outcome, the Appellate Body in Canada – Renewable
Energy / Feed-in Tariff Program developed the formalistic argument that
‘where a government creates a market, it cannot be said that the govern-
ment intervention distorts the market, as there would not be a market if
the government had not created it’, and, without any substantive argu-
ment, decided that ‘the creation of markets by government does not in
and of itself give rise to subsidies within the meaning of the SCM
Agreement’.33 Here, the Appellate Body seems to have been guided by
normative considerations (i.e., it wanted to avoid any such financial
contribution being characterized as a subsidy), rather than by legal
reasoning. In the light of the Appellate Body’s reasoning, it appears
that its analysis focuses on the market of the supported good, which is
conceptually not the right focus for distilling the benchmark for the
benefit analysis.34
33
Appellate Body Report, Canada – Renewable Energy / Feed-in Tariff Program, para. 5.188
34
As explained above (section 3.2.1.2), in Canada – Renewable Energy / Feed-in Tariff
Program, no distinction had to be made between the market of the supported good and
the market of the financial contribution, because this financial contribution involved the
purchase of the good. However, the Appellate Body’s reasoning reveals that its focus
seems to be on the market of the supported good.

Had the Appellate Body only been guided by legal reasoning and its
prior case law, it would have shared the dissenting panelist’s argumen-
tation. The dissent found that a subsidy is present if a government
provides financial contributions in order to make an uncompetitive
good (i.e., renewable electricity) competitive with existing goods (i.e.,
non-renewable electricity) on the market. The dissent argued that this
could be demonstrated by comparing the FIT price with the competitive
wholesale market for electricity that could exist in Ontario, but also
accepted that ‘facilitating the entry of certain technologies into the
market that does exist – such as it is – by way of a financial contribution
can itself be considered to confer a benefit’.35 By ‘bringing these high cost
and less efficient electricity producers into the wholesale electricity
market, when they would otherwise not be present’, a subsidy was
considered to be present.36 Thus the difference in production costs
from competing producers on the market (i.e., non-renewable electricity
producers) was considered as evidence of the presence of subsidization.
As explained, the Appellate Body took a very different approach by
turning towards supply-side substitutability: the difference in produc-
tion costs rather implied that no subsidy was demonstrated. Differences
in production costs might imply that two goods are not competing in the
same relevant market. If the government provides financial contribu-
tions to make the high-cost product competitive, through, for instance,
purchasing this product, it creates a separate market for this product.
No subsidy is present if the price for this product is not higher than
‘what a market would yield’, taking the creation of this market as a given.
As demonstrated in the Canada – Renewable Energy / Feed-in Tariff
Program case, constructing this benchmark is highly demanding. In
principle, it could consist of prices for these products under older pro-
grammes, prices in other provinces, or even prices in other countries. But
these prices have to be based on a mechanism that ensures a market
outcome, and have to be adapted to the conditions in the country in
question as well as to the government’s current definition of the supply-
side mix. Reading between the lines, it seems that the Appellate Body
suggests that the market would normally yield a price that equals pro-
duction costs plus a reasonable rate of return, although this assumes
competition among suppliers in this newly created market.
35
Panel Report, Canada – Renewable Energy / Feed-in Tariff Program, para. 9.3.
36
Ibid., para. 9.23.

The Appellate Body suggests that this benchmark is still a ‘market’

benchmark. The government first autonomously decides about creating
certain markets – that is, it defines the supply side-mix – and then
market forces decide on how this supply-side mix is met. A market
benchmark could, therefore, still be found within this newly created
market. But this is a very atypical ‘market’ benchmark, because it has
to be found within the contours of the market created by the very same
government’s financial contributions. This market is created by the
government to make the supported product competitive with products
existing in the market. The government’s intervention therefore inher-
ently alters the market outcome. A genuine market benchmark would
consist of the price for the competing products on the market, and not of
the high production costs which make the supported product uncompe-
titive without government intervention. Hence demand-side substitut-
ability should have guided the identification of the benchmark, because
this reveals what the market would be willing to pay for products on the
market considered to be similar.
The Appellate Body’s distinction between the pre-defined supply-side
mix and market forces operating within this supply-side mix seems
somewhat artificial. For instance, the Appellate Body held that ‘the
relevant question is whether windpower and solar PV electricity suppli-
ers would have entered the wind- and solar PV-generated electricity
markets absent the FIT Programme, not whether they would have
entered the blended wholesale electricity market’.37 But the FIT
Program itself created the wind- and solar PV-generated markets and,
in this way, defined the supply-side mix. Clearly, absent the FIT
Program, solar and wind-generated power could not have entered this
renewable energy market, because this market would not have existed.
15.2.1.1.2 The relevance of the pursued policy objective under the

Appellate Body’s new benefit analysis Prior to the Canada –
Renewable Energy / Feed-in Tariff Program case, it was generally under-
stood that the benefit analysis disregards the objective pursued by the
government. The only relevant question is whether the financial contri-
bution alters the recipient’s position, regardless of whether this is con-
sidered distortive (considered illegitimate) or corrective (considered
legitimate, as it corrects market failures) in nature. After all, the private
37
Appellate Body Report, Canada – Renewable Energy / Feed-in Tariff Program,
para. 5.199.

market test neglects whether the private benchmark price also equals the
socially optimal price. In the presence of market failures, the market
price does not reflect all benefits (i.e., positive externalities) or costs
(i.e., negative externalities) and the socially optimal price thus deviates
from the market price. Hence a subsidy would be present even if this only
lowers the price to the socially optimal price and thus internalizes a
positive externality. What is more, corrective government interventions
might precisely take the form of a beneficial financial contribution.
Governments playing their complementary role to that of the private
market make financial contributions on terms that are by definition not
available on the private market. In the discussion on export credit
support (see Part II), it was illustrated that governments effectively
fulfilling their complementary role to the private financial market
would precisely offer financial contributions on beneficial terms.
Conversely, a subsidy is not present where the government fails to
internalize negative externalities (i.e., the socially optimal price is higher
than the private market price). On this basis, Stiglitz’s call for CVDs to be
imposed against US products because they do not bear the cost of
environmental pollution seems legally flawed. He argued:
A subsidy means that a firm does not pay the full costs of production. Not
paying the cost of damage to the environment is a subsidy, just as not
paying the full costs of workers would be. In most of the developed
countries of the world today, firms are paying the cost of pollution to
the global environment, in the form of taxes imposed on coal, oil, and gas.
But American firms are being subsidized – and massively so.38
Yet not paying for produced negative externalities is not considered a

subsidy under the SCM Agreement. As elaborated above, low levels of
regulatory standards as well as general low levels of taxation fall outside
the scope of the subsidy definition.39,40 In sum, until Canada – Renewable
38
J. Stiglitz, ‘A New Agenda for Global Warming’, 3:7 Economists’ Voice (2006), 1–4, at 2.
39
As explained above, the domestic market forms the benchmark for assessing whether
revenue is forgone under a tax system.
40
See also Bhagwati and Mavroidis, above Chapter 4 n. 78, at 302–3; S. Z. Bigdeli, ‘Incentive
Schemes to Promote Renewables and the WTO Law of Subsidies’, in T. Cottier, O. Nartova,
and S. Z. Bigdeli (eds.), International Trade Regulation and the Mitigation of Climate Change
(Cambridge University Press, 2009), 155–99, at 157–76; J. Pauwelyn, ‘US Federal Climate
Policy and Competitiveness Concerns: The Limits and Options of International Trade Law’,
working paper, Duke University, 2007, at 14–15. However, Howse and Eliason agree with
Stiglitz that CVDs could be imposed in response to a lack of undertaking action under the
Kyoto Protocol. They suggest that the allowance for emitting carbon constitutes the

Energy / Feed-in Tariff Program, the case law suggested that the subsidy
definition disregards the question whether government interventions are
corrective rather than distortive in nature.41,42
The Appellate Body in Canada – Renewable Energy / Feed-in Tariff
Program seems to have been concerned about this outcome, particularly
because the SCM Agreement lacks an Article XX GATT type of
defence.43 This seems to have inspired the Appellate Body to take a
different approach to the benefit analysis. But what is the relevance of
the legitimate policy objectives in deciding when recourse to an alter-
native benchmark is allowed? The Appellate Body’s report raises this
question, but fails to give clear guidance on the answer. Overall, its report
suggests that an inquiry into the pursued objective is not needed to
define the relevant market and thus to decide on whether an alternative
benchmark could and should be used.
The Appellate Body firmly states that ‘introducing legitimate policy
considerations into the determination of benefit cannot be reconciled
with Article 1.1(b) of the SCM Agreement’, but subsequently discusses at
length the legitimate policy considerations for providing support to
provision of a good (e.g., a right to emit carbon is given by the government) which confers a
benefit, whereby an adjusted market price in a country that has installed an emissions trading
market (e.g., the EU) is used as benchmark. However, their creative interpretation seems to
be based on an overly broad reading of Articles 1.1(a)(1)(iii) and 14(d) of the SCM
Agreement and of the Appellate Body’s jurisprudence on the use of alternative benchmarks.
De Cendra, Hufbauer, et al. correctly disagreed that an allowance can be defined as a good
under the SCM Agreement. Regarding the benefit determination, the proposal to use the
price in countries having an emission trading system in place does not fit the Appellate
Body’s jurisprudence on alternative benchmarks (see above Chapter 3, section 3.2.1.1.1.1)
(e.g., their proposed alternative benchmark would also be distorted because governments,
according to their reasoning, are the sole providers of the right to emit carbon, implying that
no market benchmark could be constructed). See R. Howse and A. Eliason, ‘Domestic and
International Strategies to Address Climate Change: An Overview of the WTO Legal Issues’,
in T. Cottier, O. Nartova, and S. Z. Bigdeli (eds.), International Trade Regulation and the
Mitigation of Climate Change (Cambridge University Press, 2009), 48–93, at 73–6; J. de
Cendra, ‘Can Emissions Trading Schemes Be Coupled with Border Tax Adjustments? An
Analysis vis-à-vis WTO Law’, 15:2 Review of European Community and International
Environmental Law (2006), 131–45, at 137; G. C. Hufbauer, S. Charnovitz, and J. Kim,
Global Warming and the World Trading System (Washington, DC: Peterson Institute for
International Economics, 2009), at 61–2.
41
Whether the provision of goods serves a public objective is also irrelevant for the
determination whether it constitutes ‘general infrastructure’, which is excluded from
the subsidy definition (above Chapter 3, section 3.1.1.2).
42
Schwartz and Harper revealed a similar approach under the pre-SCM Agreement period.
W. Schwartz and E. W. Harper, ‘The Regulation of Subsidies Affecting International Trade’,
70:5 Michigan Law Review (1972), 831–58, at 833.
43
See above, section 4.4.2.

renewable energy in the case at hand.44 Referring to the presence of

externalities, the Appellate Body concludes that:
[A] comparison between renewable energy electricity generators and

conventional energy electricity generators requires consideration of
the full costs associated with the generation of electricity. In this
respect, if, on the one hand, higher prices for renewable electricity
have certain positive externalities, such as guaranteeing long-term
supply and addressing environmental concerns, on the other hand,
lower prices for non-renewable electricity generation have certain neg-
ative externalities, such as the adverse impact on human health and the
environment of fossil fuel energy emissions and nuclear waste disposal.
Considerations related to these externalities will often underlie a govern-
ment definition of the energy supply-mix and thus be the reason why
governments intervene to create markets for renewable electricity
generation.45
Hence considerations related to externalities often explain why govern-

ments intervene to create markets, and could therefore clear the way to
rely on the alternative market benchmark. In this way, the Appellate
Body’s jurisprudence gives policy space to correct market distortions by
creating new markets, but legitimate policy considerations do not seem,
as a matter of law, to be a prerequisite to having recourse to this
alternative benchmark. Apparently, the supply-side mix as defined by
the government should be taken as a given, regardless of the govern-
ment’s motivations therefor.
Conversely, legitimate policy considerations could also explain why
governments correct distortions in existing markets and thus do not
always result in the creation of new markets and the corresponding
alternative benefit benchmark.46 For instance, we explained above that
government support for certain industries in times of a financial crisis
would, under the new jurisprudence, likely still be characterized as a
subsidy. Hence the above-mentioned conclusion still seems to hold that
a subsidy is present when governments fulfil their complementary role to
the private financial market.
44
Appellate Body Report, Canada – Renewable Energy / Feed-in Tariff Program, paras.
5.185–5.189.
45
Ibid., para. 5.189 (emphasis added).
46
Recall that the traditional benchmark should be employed when governments support
certain players ‘in markets that already exist, or to correct market distortions therein’
(ibid., para. 5.188).

Further, the fact that two products created by the government are
similar in the light of the pursued legitimate objective also seems not to
be relevant to deciding whether both products fall within the same
relevant market. For instance, solar and wind electricity are both created
to achieve the same environmental objectives, but were still considered
as part of different markets, because the government had decided to
support both technologies (i.e., part of the supply-side mix that should
be taken as a given) and from the perspective of the supply-side they
were different (i.e., solar generated electricity is much more costly than
wind generated electricity).
Next, the Appellate Body’s jurisprudence implies that in the presence of
positive externalities (e.g., from renewable energy), government support to
make this product competitive to other products (generating no positive, or
even negative, externalities) does not ipso facto amount to subsidization
within the meaning of the SCM Agreement, despite the fact that, from an
economic perspective, this would be a textbook example of what a subsidy is
and when such subsidization is justified. However, it bears noting that the
Appellate Body’s jurisprudence does not ensure that this support is limited
to internalizing the positive externality generated by the environmentally
friendly product. Broadly speaking, such support would not amount to
subsidization as long as the price paid for the environmentally friendly
product is not above production costs plus a reasonable rate of return, even
if this is higher than what the market would be willing to pay plus the value
of the positive externality. This could again be illustrated on the basis of the
difference between wind generated and solar generated electricity: both
types of electricity are arguably similar in terms of positive externalities,
but very unequal in terms of production costs, as solar generated electricity
is much more costly to produce. Still, the price paid for solar generated
electricity could be substantially higher than the price paid for wind gen-
erated, without amounting to subsidization in the meaning of the SCM
Agreement.
Finally, we explained above that governments’ lack of action to inter-
nalize negative externalities should not be regarded as subsidization
within the meaning of the SCM Agreement. This conclusion still holds,
as it is not affected by the Canada – Renewable Energy / Feed-in Tariff
Program case law.
In conclusion, this case law suggests that legitimate policy consider-
ations for government support are neither a necessary nor a sufficient
condition for triggering the application of the alternative benefit bench-
mark. But it is clear that the Appellate Body aimed at excluding some

corrective interventions from the disciplines of the SCM Agreement by

providing for a carve-out for those government interventions creating a
separate market. The Appellate Body’s extensive discussion of the pres-
ence of legitimate objectives in the case at hand, and the fact that its
ruling departs from the established case law, suggests that this carve-out
should not be interpreted too broadly. It could try to achieve this out-
come by interpreting ‘creating’ a new market in a rather narrow sense, in
such a way that, de facto, it would largely capture those situations where
a government is driven by legitimate policy considerations. How this
could be achieved is discussed next.
15.2.1.1.3 Limiting the implications of the Appellate Body’s new

approach to the benefit analysis Given its open and broad formulation,
the Appellate Body’s jurisprudence in Canada – Renewable Energy / Feed-in
Tariff Program could be read in an overly expansive manner that would
allow recourse to the alternative benefit benchmark when support is given to
a domestic industry that would, without government support, simply be
uncompetitive with foreign producers because of higher domestic produc-
tion costs. This reading could hinge on the Appellate Body’s finding that two
products are not in the same market if there is low supply substitutability
(e.g., high difference in production costs), even if there is high demand
substitutability (i.e., consumers consider both products to be like).
Obviously, such a reading would eviscerate the very essence of the SCM
Agreement’s disciplines, as it would give leeway to governments to subsidize
uncompetitive domestic industries without any proper discipline. To avoid
this outcome, differences in production costs resulting from the origin of the
products should not be taken into account in deciding whether two products
(i.e., domestic and foreign products) are part of the same relevant market. In
Canada – Renewable Energy / Feed-in Tariff Program, the Appellate Body
referred to intrinsic differences in costs between renewable and conven-
tional electricity, and concluded that markets for renewable electricity ‘can
only come into existence as a matter of government regulation’.47 This
conclusion would not hold where differences in production costs relate
simply to the origin of the product. The market exists if it is (potentially)
supplied by imported products (without government support48), implying
that support to uncompetitive domestic producers is characterized as a
47
Appellate Body Report, Canada – Renewable Energy / Feed-in Tariff Program, para. 5.175.
48
This is a thorny issue which seems unavoidable under the Appellate Body’s
jurisprudence.

subsidy under the SCM Agreement. If foreign products are kept out of the
market by way of government regulation, the government again cannot
claim that the market does not exist and could be ‘created’ through govern-
ment support. This is not a market that ‘can only come into existence as a
matter of government regulation’, but rather requires less government
regulation regarding imported products. Governments might not be the
reason why markets do not exist, as this would mean that they could pave
the way for the alternative benefit approach. WTO case law is clear that
governments’ own conduct should be discounted in the analysis of whether
two products are substitutable.49
There are other ways to further limit situations where governments
are considered to ‘create’ new markets, resulting in recourse to the
alternative benefit benchmark. First, it is helpful that the alternative
market benchmark seems only available where differences in production
costs are so significant that the uncompetitive product would not have
entered the market at all without government intervention. Second, the
Appellate Body seems to suggest that recourse to the alternative bench-
mark could be limited in time, namely for as long as the differences in
production costs remain significant. Third, it would be helpful if future
panels or the Appellate Body were to confirm that this jurisprudence is
confined to the creation of the supported goods markets, and does not
extend to the creation of the market of any financial contribution (e.g.,
financial market). This would mean that the regular market benchmark
would be applied if the government, in time of a financial crisis, supports
its existing domestic industries with loans unavailable on the private
market. Fourth, and finally, a future panel or the Appellate Body might
be inclined to limit recourse to the alternative benchmark to situations
where the government purchases goods, and thus foreclose this option
for other types of financial contribution (e.g., transfer of funds).
However, this is not to be recommended, because it would make the
benefit analysis fundamentally different depending on the type of finan-
cial contribution, and governments could still find their way to the new
benefit approach by converting such financial contributions into direct
or indirect purchases of goods.
In sum, recourse to the alternative benchmark should be confined to
situations where the government ‘creates’ the market of the supported
goods, when and for as long as intrinsic and substantial differences in
49
See, inter alia, Appellate Body Report, Korea – Alcoholic Beverages, paras. 114–122.

production costs with competing goods preclude their presence in the

market without government support.
15.2.1.1.4 Was the Appellate Body’s new approach to the benefit

analysis warranted? As a matter of law, the Appellate Body got it
wrong. We explained above how the Appellate Body deviated from the
correct market-based approach to the benefit analysis. It adopted this
approach to craft policy for corrective subsidies, but we equally showed
that the presence of legitimate objectives seems neither needed nor
sufficient to rely on the alternative market benchmark. Regrettably, the
implications of its decision are highly unclear. When does a government
create a market instead of distorting or correcting existing markets? And
what exactly is the appropriate benchmark in those circumstances? As
our discussion illustrated, several aspects of these basic questions are still
open. This legal uncertainty is problematic not only for governments
considering implementing policy interventions in the light of their WTO
obligations, but also, more specifically, for their CVD investigating
authorities in deciding whether action is allowed against foreign alleged
subsidization.
From an economic perspective, the Appellate Body again got it wrong. It
excludes from the subsidy definition support measures that economists
would consider as subsidies par excellence. For economists, a FIT scheme
serves as a textbook example of a corrective subsidy.50 Needless to say, the
subsidy definition under the SCM Agreement could take a different
approach, but there was no need to interpret the benefit analysis in such a
way that it deviates from economic logic respected in all previous case law.
The fact that the Appellate Body carved out certain market-creating
government intervention from the subsidy definition implies that such
support is not disciplined under the SCM Agreement and could not be
countervailed. Conceptually, the subsidy definition is not the right place in
which to read legitimate policy considerations into the agreement.
First, as illustrated in the Canada – Renewable Energy / Feed-in Tariff
Program case, it implies that certain market-creating support contingent
on the use of domestic input is not a prohibited local content subsidy
50
In the words of Frontier Economics, ‘The public benefit to society from lower emissions
is entirely predicated on there being a private benefit to the producers of energy from
renewable sources.’ ‘FIT for Purpose? A Review of the Economics of a WTO Panel
Ruling on Feed-in Tariffs and Local Content Requirements’, client briefing by Frontier
Economics, March 2013.

under Article 3.1(b) of the SCM Agreement. As also demonstrated by the

same case, this has no substantive legal implications because such sup-
port is anyway inconsistent with the Illustrative List of the TRIMs
Agreement and Article III:4 of the GATT.51 Still, this jurisprudence
precludes the stricter procedural disciplines stipulated in the SCM
Agreement for local content requirements linked to support that could
be deemed subsidies par excellence. Certainly, the Appellate Body’s new
approach to the benefit analysis provides a strong disincentive to chal-
lenging local content subsidies under the SCM Agreement.
Second, this case law could also have repercussions for the second type
of prohibited subsidy, namely export subsidies. Strictly speaking, the
Appellate Body’s argumentation appears not to exclude invoking the
alternative market benchmark for market-creating support contingent
on exportation. For instance, a government might provide loans to create
environmentally friendly cars, but only if these cars are exported. On the
other hand, as this example illustrates, if the government is genuinely
aiming at creating a market for goods having positive spillover effects, it
will in practice have no incentive to make its support contingent on
exportation. Further, as a matter of law, a complainant might still argue
that such support is beneficial even under the alternative benchmark
articulated by the Appellate Body. Still, the export contingency element
itself does not appear to make such support ipso facto beneficial, given
that this element is only assessed under Article 3 of the SCM Agreement.
The level of support could also de facto be limited to what a market
would yield, implying that no subsidy would be present. In that scenario,
a complainant could try to argue that such support qualifies as a
prohibited export subsidy in the meaning of any of the items of the
Illustrative List. As explained above, the case law has allowed complai-
nants to bypass the subsidy definition under Article 1 and invoke directly
one of the listed items under the Illustrative List. This will raise the
thorny question whether the Appellate Body’s alternative benefit
approach will also be transposed to the Illustrative List.
Third, and most fundamentally, the Appellate Body’s jurisprudence
risks blurring the dividing line between the GATT and the SCM
51
The only substantive difference is that a violation under the GATT and TRIMs could in
principle still meet the conditions under Article XX of the GATT. But this difference
seems merely theoretical, because such local content measures are in principle not
necessary to achieve a legitimate objective. In Canada – Renewable Energy / Feed-in
Tariff Program, Canada did not even invoke Article XX of the GATT as a defence.

Agreement. Subsidies paid exclusively to domestic producers are exemp-

ted from the national treatment discipline by virtue of Article III:8 of the
GATT, but are disciplined under the SCM Agreement. However, the
Appellate Body’s jurisprudence implies that certain market-creating
support for domestic producers would not meet the criteria of the
subsidy definition under the SCM Agreement. This raises the question
whether such support would therefore also fall outside the scope of the
carve-out under Article III:8 of the GATT. As a matter of law, the
Appellate Body could still conclude that such support qualifies as a
‘subsidy’ within the meaning of the carve-out, even though it is not a
subsidy within the meaning of the SCM Agreement. Although this would
run counter to a harmonious reading of different WTO agreements, the
Appellate Body might be inclined to reach this conclusion if it were
confronted with this question, because its jurisprudence would uninten-
tionally have created less policy space for certain corrective subsidies if
this support were disciplined under Article III GATT. This is further
illustrated below in the case study on environmental subsidies.52
Beyond deviating from the benefit analysis, the Appellate Body’s
interpretation thus needlessly undermines the internal coherence of
the SCM Agreement and its harmonious coexistence with the GATT.
The Appellate Body failed to acknowledge the limited but important
threshold function of the subsidy definition under the SCM Agreement.
Instead of adopting its alternative approach, the Appellate Body should
have reiterated that the subsidy definition does not outlaw corrective
subsidization as such. Conceptually and systemically, it appears more
appropriate to introduce flexibility on corrective subsidies under the
applicable disciplines instead of excluding such interventions from the
subsidy definition in the first place. This approach was exactly imple-
mented by introducing ‘green light’ types of subsidization and likewise
partly explains the rationale for S&D treatment given to developing
countries. To a certain extent it was not up to the Appellate Body to
inject more policy space into the agreement, as this is the responsibility
and prerogative of the WTO membership. Still, as further elaborated on
below, with some creative – but still legal – interpretation, the Appellate
Body could instead have further developed the specificity test in such a
way that it excludes some corrective subsidies from subsidy disciplines
and CVD responses. Further, the likeness, relevant market, and trade
effects tests under Article 6 could also be operated in a manner sensitive
52
See below section 16.1.1.2.2.

to the objective pursued by the subsidy.53 These approaches would have

preserved the gatekeeping role of the subsidy definition in the SCM
Agreement (and coherence with the GATT 1994) and the market-
based approach to the benefit analysis in particular.
15.2.1.2The disregard of the broader regulatory

framework for the benefit analysis
Scholars have underlined the fact that the benefit analysis only detects
whether the financial contribution (or income or price support) in itself
is provided at better-than-market terms and disregards the broader
regulatory framework. Indeed, where a firm receives a loan at below
market rates, a benefit is conferred under the SCM Agreement without
any consideration of the various taxes or regulatory burdens imposed on
the firm in question.54 Because the financial contribution is isolated in
the benefit test, it is simply irrelevant whether or not it is provided to
compensate, for instance, for overvalued exchange rates, higher environ-
mental standards, or location in a disadvantaged region. To be precise,
the only government-imposed burdens that can be compensated are
indirect taxes or import duties, since these can be adjusted at the border.
By disregarding whether the benefit is offset by other taxes or regulatory
burdens, the subsidy definition fails, according to Sykes, to sort out the
net impact of the government on the firm’s competitive position.
Although recognizing that such an exercise would be ‘extraordinarily
complicated and fraught with error’, he concludes that ‘to ignore the
problem is to render the system unable to detect true subsidization of an
industry except by chance’.55 Hudec seemed to share this concern, but
rightly considered that singling out the net effect would simply be
impossible.56 While largely neglected under the subsidy definition, this
concern helps to explain, to some extent, why more flexibility was given
to certain forms of subsidy. The ‘green light’ status of subsidies intended
for implementing higher environmental standards or for firms to locate
53
See below Chapter 16 section 16.1.1.2.2.2.2.2.
54
A. O. Sykes, ‘Subsidies and Countervailing Measures’, in P. F. J. Macrory, A. E. Appleton,
and M. G. Plummer (eds.), The World Trade Organization: Legal, Economic and Political
Analysis: Volume II (Heidelberg: Springer Verlag, 2005), 83–107, at 86. See also Green
and Trebilcock, above Chapter 5 n. 36, at 131.
55
Sykes, above n. 54, at 100.
56
According to Hudec, ‘There are many differences between business conditions found in
one country and another [and] [w]e know that there is no way to measure the net
balance of advantage produced by such differences’. Hudec, above n. 20, at 263, 237–8.
Somewhat along the same lines see also Green, above Chapter 4 n. 448, at 405.

in disadvantaged regions was defended by the EU during the Uruguay

Round precisely because it compensated for higher costs.57 This line of
reasoning is thus based on what could be labelled the ‘level playing field’
argument: by offsetting other taxes and regulatory burdens (or even
market distortions), such subsidies would simply level the playing field
and would, therefore, not distort competition.
Even if one accepts that the benefit analysis narrowly focuses on the terms
of the financial contribution, the test fails to single out those contributions
that generate an impact on the recipient’s competitive position. This relates
to the criticism formulated by Grossman and Mavroidis on the benefit test
developed in the case law, which will be discussed next.
15.2.2 The impact of privatization and the distortive

effect on competition
Grossman and Mavroidis have criticized the ‘benefit’ definition devel-
oped in the case law, for the reason that it does not discern those
subsidies that effectively generate an effect on the market.58,59 Because
the SCM Agreement seems to protect producer welfare,60 the definition
57
Hence the EU argued that such subsidies did not affect competition (MTN.GNG/NG4/
W/36, 2 February 1990). This proposition is based on the unlikely assumption that the
subsidy merely affected the location decision of the firm within the subsidizing member.
A. O. Sykes, ‘Second-Best Countervailing Duty Policy: A Critique on the Entitlement
Approach’, 21 Law and Policy in International Business (1990), 699–721, at 717.
58
See G. M. Grossman and P. C. Mavroidis, ‘US – Lead and Bismuth II, United States
Imposition of Countervailing Duties on Certain Hot-Rolled and Bismuth Carbon Steel
Products Originating in the United Kingdom: Here Today, Gone Tomorrow?
Privatization and the Injury Caused by Non-Recurring Subsidies? (WT/DS138; DSR
2000:V, 2595; DSR 2000:VI, 5623)’, in H. Horn and P. Mavroidis (eds.), The American
Law Institute Reporters’ Studies on WTO Case Law – Legal and Economic Analysis
(Cambridge University Press, 2007), 183–213; G. M. Grossman and P. C. Mavroidis,
‘United States – Countervailing Measures Concerning Certain Products from the
European Communities (WTO Doc. WT/DS212/AB/R; DSR 2003:I, 5; DSR 2003:I,
73): Recurring Misunderstanding of Non-Recurring Subsidies’, in H. Horn and
P. Mavroidis (eds.), The American Law Institute Reporters’ Studies on WTO Case
Law – Legal and Economic Analysis (Cambridge University Press, 2007), 381–90; see
also H. Horn and P. C. Mavroidis, ‘United States – Preliminary Determinations with
Respect to Certain Softwood Lumber from Canada (WT/DS236; DSR 2002:XI, 3597):
What is a Subsidy?’, in H. Horn and P. Mavroidis (eds.), The American Law Institute
Reporters’ Studies on WTO Case Law – Legal and Economic Analysis (Cambridge
University Press, 2007), 523–49.
59
See also Sykes, above n. 5, at 510–511.
60
This conclusion is stated from the perspective of positive theory. Grossman and
Mavroidis, ‘Here Today, Gone Tomorrow?’, above n. 58, at 198–9.

of a subsidy must be one that ‘helps to identify policies that inflict such
harm’.61 Accordingly a benefit is conferred, in the view of these authors,
when a competitive advantage is offered to firms vis-à-vis other firms.
Hence the benefit test should not simply look at whether the recipient is
generally better off (e.g., in terms of wealth) than without the financial
contribution. Instead, it should examine whether its competitive position
is improved compared with ‘but for’ the financial contribution. Behind
their alternative approach seems to be the recognition that a precondi-
tion for causing harm to foreign producers is that domestic firms’ output
levels are increased as a result of a financial contribution,62 which in turn
principally assumes, at least in perfectly competitive markets, that it directly
or indirectly lowered their marginal costs.63,64 A profit-maximizing firm in
competitive markets sets output decisions in such a way that marginal costs,
which are a function of production costs, equate marginal revenue. Put
differently, a competitive advantage is present in cases where domestic
firms’ output level (or marginal costs)65 would have been lower (or higher)
‘but for’ the financial contribution.66
Hence the ‘benefit’ approach developed in the case law (and supported
by Diamond67) takes the market value as the benchmark to determine
the existence of subsidization: is a financial contribution made at better-
than-market terms? In contrast, the approach suggested by Mavroidis
61
Ibid., at 199–200.
62
See, e.g., R. Diamond, ‘A Search for Economic and Financial Principles in the
Administration of United States Countervailing Duty Law’, 21 Law and Policy in
International Business (1990), 507–608, at 539.
63
The suggestion to focus on the effect on the firm’s competitive position somewhat
resembles the plea of proponents of the so-called entitlement rationale in the 1980s
(e.g., Goetz, Granet, Shwartz, and Diamond). These authors advocated that CVDs
should only be imposed to the extent marginal costs are lowered as a result of a subsidy.
An important difference seems to be that Diamond rejects the argument that the
competitive effect could be assessed under the benefit analysis as currently inscribed
in the SCM Agreement (see below Chapter 18, section 18.2). See C. J. Goetz, L. Granet,
and W. F. Schwartz, ‘The Meaning of “Subsidy” and “Injury” in the Countervailing Duty
Law’, 6 International Review of Law and Economics (1986), 17–32; R. Diamond,
‘Economic Foundations of Countervailing Duty Law’, 29 Virginia Journal of
International Law (1989), 767–812. See also Sykes, above Chapter 5 n. 250, at 99.
64
Sykes illustrates that there could be circumstances where marginal costs are not reduced
as a result of subsidization but where output levels have nonetheless increased. Most
cases of higher output levels do result from a reduction in marginal costs.
65
This was the focus of the proponents of the entitlement theory (see above n. 63).
66
Horn and Mavroidis, above n. 58, at 533; Goetz, Granet, and Schwartz, above n. 63, at 23.
67
Diamond holds that there is no interpretative basis for supporting Grossman and
Mavroidis’s suggestion. Diamond, above Chapter 5 n. 249.

and Grossman looks at the financial contribution’s impact on the com-

petitive position to determine the existence of subsidization: does a
financial contribution improve the competitive position of the recipient?
Both approaches have apparently resulted in a different answer to the
question whether privatization of a firm at arm’s length and for fair
market value extinguishes the continued existence of a benefit derived
from a prior non-recurring financial contribution.68 In short, does priva-
tization at a fair market price extinguish benefits resulting from prior non-
recurring subsidization? After briefly recapitulating the Appellate Body’s
response, we turn to the criticism by Grossman and Mavroidis and analyse
whether their criticism could be integrated under the current case law’s
reading of the benefit element. Finally, their broader argument – that the
SCM Agreement should single out only those subsidies affecting the
producers’ competitive position – is assessed.
15.2.2.1 The impact of privatization

Benefits of non-recurring subsidies are often allocated over the average
useful life of the acquired assets in the relevant industry.69 For instance, if
the government has financed on favourable terms the acquisition of a
machine with utility value of ten years, the benefit of such a subsidy
under the ‘useful life’ approach would be spread over ten years. The
Appellate Body considered such ‘useful life’ practice permissible under
the SCM Agreement, so long as the presumption is not irrebuttable.70
But is such allocation over the useful life interrupted by the firm’s
privatization at arm’s length and for fair market value? The Appellate
Body in US – Countervailing Duties on Certain EC Products answered
this question affirmatively in case of full privatization. Although it
acknowledged that the utility value of the equipment acquired on sub-
sidized terms does not extinguish because of privatization, it considered
this element irrelevant under the ‘benefit’ analysis, as this test looks at the
market value: if a fair market value is paid at the moment of privatiza-
tion, ‘the market value is redeemed’, and the benefit thus in principle
extinguishes.71 Relaxing its stance taken in US – Lead and Bismuth II, the
68
For definitions of ‘arm’s length’ and ‘fair market value’, see above Chapter 3 n. 342.
69
The benefits of recurring subsidies are usually considered to be fully absorbed in the year
of receipt.
70
Appellate Body Report, US – Countervailing Measures on Certain EC Products, para. 84;
Appellate Body Report, US – Lead and Bismuth II, para. 62.
71
Appellate Body Report, US – Countervailing Measures on Certain EC Products, para. 102

Appellate Body decided that such a fair market price does not necessarily
extinguish prior subsidization but only offers a presumption thereof.72
Grossman and Mavroidis have firmly criticized this case law that a fair
market price could extinguish the continued benefit of past non-
recurring subsidies.73 The change in ownership – at a fair market price
or not – simply has no bearing on the question of the continuation of
benefit, at least if the ‘benefit’ concept is understood correctly as an
amelioration of its competitive position and not in terms of economic
‘wealth’ as is understood by the Appellate Body. The price at which a
firm acquires assets should be qualified as a ‘sunk cost’, which does not
affect their future profit-maximizing output decisions inasmuch as mar-
ginal costs are left unaffected.74 The enhanced competitive position
resulting from the firm’s prior subsidization is not affected by the price
paid for the change in ownership and thus leaves the harm caused to
foreign firms untouched.75 In their words, ‘a change in ownership – at
fair market prices or otherwise – has no bearing on competitive con-
ditions in the world market’.76 Grossman and Mavroidis correctly draw a
parallel with non-recurring subsidies paid directly to private firms.77
Ownership shares in private firms frequently transfer in the private
capital market (by definition at a fair market price). Hence would such
common private-to-private sales also extinguish (parts of) the benefit of
the non-recurring subsidy in the Appellate Body’s view, given that the
new owner does not personally benefit (in wealth terms) from the
subsidy? As explained, the Appellate Body in EC – Large Civil Aircraft
was divided on this question: (only) two out of three members consid-
ered that no extinction would take place.78 Observe that one of the
72
Ibid., paras. 122–127.
73
Because the benefit element cannot be interpreted along the lines suggested by
Grossman and Mavroidis, Diamond concurs with the Appellate Body’s reasoning on
the effect of privatization. Diamond, above Chapter 5 n. 249, at 6–8, 21–6.
74
Grossman and Mavroidis, ‘Recurring Misunderstanding of Non-Recurring Subsidies’,
above n. 58, at 386–8; Diamond, above n. 62, at n. 105.
75
In the words of one of the Appellate Body members in EC – Large Civil Aircraft (para.
726), ‘The central point is that a sale of shares, whether or not it conveys control,
transfers rights in the shares to a new owner. The assets of the company, to which the
shares attach, do not change at all.’
76
Grossman and Mavroidis, ‘Here Today, Gone Tomorrow?’, above n. 58, at 201.
77
Palmeter is credited by Grossman and Mavroidis for articulating this parallel.
78
Appellate Body Report, EC – Large Civil Aircraft, para. 726. In my view, this could only be
rejected in a coherent way if the Appellate Body equally reverses its case law on full
privatization, given that the same mechanism and rationale for accepting extinction are at
stake. Hence the view expressed by one of the Appellate Body members (as well as the panel

members refuting the extinction theory seemed to formulate precisely

the criticism advanced by Grossman and Mavroidis, and thus equally
disagreed with the previous case law that the sale at fair market price in
the case of full privatization could extinguish prior subsidization.79 The
panel in US – Large Civil Aircraft had decided somewhat earlier that the
privatization case law could not be extended to private-to-private trans-
actions, because the nature of both transactions would be dissimilar and
because this would ‘significantly undermine the disciplines of the SCM
Agreement’.80 Although it seems hard to grasp how their difference in
nature underpins this reasoning, the panel at least understood that if
private-to-private sales extinguished the benefit, then governments
would be free to provide publicly traded companies with subsidies
because the benefit would quickly extinguish as their shares are traded
on a daily basis.81,82
Essentially responding to the criticism advanced by Grossman and
Mavroidis,83 the Appellate Body in US – Countervailing Measures on
Certain EC Products refuted the argument that the costs and volume of
production remain the same regardless of the sale price of privatization,
‘since these costs include, as a necessary component, the cost of capital’:
For example, if a government makes a ‘financial contribution’ that

‘benefit[s]’ a state-owned enterprise, and then sells that enterprise for
less than its fair market price, would this not normally result in a ‘better
off’ return for the private capital newly invested in that enterprise?
Would that not suggest, as a consequence, that the under-priced enter-
prise may then attract more investment than it would have attracted
otherwise, if the government had sold it for fair market price? Why would
this government-induced additional investment not then reduce the
enterprise’s cost of raising capital (either by borrowing it from the bank
or from, say, shareholders) and, ultimately, reduce the firm’s overall costs
of production?84
in EC – Large Civil Aircraft) that the full privatization case law could not simply be
transposed in the case of partial privatization or private-to-private sales is not convincing.
79
The ‘subsidy continues to benefit the recipient, even if the ownership of the recipient’s
shares changes from one day to another’. Appellate Body Report, EC – Large Civil
Aircraft, para. 726(c).
80
Panel Report, US – Large Civil Aircraft, paras. 7.864–7.871.
81
Panel Report, US – Large Civil Aircraft, para. 7.870.
82
Likewise, the current case law opens the door for subsidization of government-owned
enterprises that will be fully privatized.
83
This was articulated by the United States. See Appellate Body Report, US –
Countervailing Measures on Certain EC Products, para. 103.
84
Ibid., para. 103 (original footnotes omitted).

The Appellate Body’s reasoning qualifies, but certainly does not refute,
the core of the argument articulated by Grossman and Mavroidis. First of
all, the core thesis that the artificially enhanced competitiveness gener-
ated by prior non-recurring subsidies will not be eliminated by a fair
market price is not addressed at all by the Appellate Body’s argumenta-
tion.85 The Appellate Body’s response simply shows that an additional
subsidy would be generated when privatization is done at below market
prices.86 The claim that the level of the sales price has no bearing what-
soever on subsequent, profit-maximizing behaviour might indeed be too
rigid, as a below the market price could improve the firm’s financial
position and thus generate a reduction in its cost of raising capital, which
could affect future investment decisions. To be sure, this reasoning is
based on the premise that the firm in question is constrained in its access
to capital. If not, an increase in cash flow resulting from this extra
subsidy (i.e., the sale below market price) would not affect this firm’s
investment behaviour and would thus not generate extra investments.87
In conclusion, under the present case law, there is a presumption that
privatization at a fair market price extinguishes the continued benefit of
prior non-recurring subsidies.88,89 This case law could be criticized, since
a change of ownership – regardless of the actual price – does not affect
the artificially generated competitive advantage. Privatization could
only generate an additional subsidy if made at below market terms
85
As illustrated above, the Appellate Body simply rejects this argument because the benefit
analysis looks at the market price.
86
The United States seemed to share this view as well: ‘the fact that the private owner pays
full market price for the enterprise indicates only that the private owner is not receiving a
new subsidy’. Appellate Body Report, US – Countervailing Measures on Certain EC
Products, para. 99 (emphasis in original).
87
As explained in US – Large Civil Aircraft, the Modigliani–Miller theorem implies that it
could be assumed that firms invest at optimal levels when they are unconstrained as to
capital. In that case, a subsidy would be passed on in its entirety to shareholders
(dividends) instead of higher investments. See Panel Report, US – Large Civil Aircraft,
paras. 2 and 8 of Appendix VI.F.2.
88
In order to implement the DSB’s recommendations in US – Countervailing Measures on
Certain EC Products, the United States has drafted a new privatization methodology (see
Panel Report, US – Countervailing Measures on Certain EC Products (Article 21.5 – EC),
n. 313). This new methodology was not as such challenged before the compliance panel
(ibid., n. 206 and para. 7.89).
89
The panel in US – Countervailing Measures on Certain EC Products (Article 21.5 – EC)
decided that the CVD administration has to determine in a sunset review whether
privatization has occurred at arm’s length and for fair market value, and cannot simply
assume that this has been the case. See Panel Report, US – Countervailing Measures on
Certain EC Products (Article 21.5 – EC), paras. 7.198–7.217.

(which could cause adverse effects if the firm is constrained in raising

capital), but could not neutralize the competitive effect of previous
subsidies.90 To be clear, it is sometimes suggested that the Appellate
Body in US – Countervailing Duties on Certain EC Products has partly
met this criticism from Grossman and Mavroidis by shifting from a
irrebuttable to a rebuttable presumption of the extinction of continued
benefit. But the somewhat peculiar relaxation adopted by the Appellate
Body seems not to be related at all to the criticism articulated by
Grossman and Mavroidis.91 Recapitulating the Appellate Body’s central
view as it still stands today with respect to full privatization: ‘once a fair
market price is paid for the equipment, its market value is redeemed,
regardless of the utility the firm may derive from the equipment’, and
only this market value is considered relevant under the benefit analysis.92
But the wide divergence of views among all three Appellate Body mem-
bers in EC – Large Civil Aircraft shows that this case law is still not
settled. What is more, it will be explained that the Appellate Body’s
demarcation in this case between the benefit analysis and adverse effect
analysis seems implicitly to entail that the criticism voiced by Grossman
and Mavroidis can and should be accepted when an actionable subsidy
claim is formulated. It will equally be explained that integrating their
criticism seems somewhat more difficult when the CVD option is
considered.
Grossman and Mavroidis have suggested that the impact of the finan-
cial contribution on the firm’s competitive position should be integrated
into the benefit analysis. Yet the case law is legally correct in focusing
exclusively on whether the recipient is made better off in wealth terms as
a result of the financial contribution when analysing the benefit ele-
ment.93 Still, their fundamental criticism of the privatization case law is
correct, but this could be integrated under the present SCM Agreement,
certainly with regard to actionable subsidy claims (multilateral track).
After all, the Appellate Body in EC – Large Civil Aircraft correctly
confirmed that Article 1 of the SCM Agreement does not mandate
90
This also seemed to be the view of the panel in EC – Large Civil Aircraft, para. 7.243.
91
Grossman and Mavroidis first articulated their viewpoint in a case note on US – Lead
and Bismuth II. See Grossman and Mavroidis, ‘Here Today, Gone Tomorrow?’,
above n. 58.
92
Appellate Body Report, US – Countervailing Measures on Certain EC Products, para. 102
(emphasis in the original).
93
Contextual support could be found in Article 14 of the SCM Agreement. See also
Diamond, above Chapter 5 n. 249.

demonstration of ‘continuing benefit’, but simply mandates that a bene-

fit is conferred at the moment the financial contribution is made.94 It is
not required to demonstrate that the current owner (after privatization
or private-to-private sales) still enjoys a benefit in wealth terms to
demonstrate the existence of a subsidy. The benefit in wealth terms
should be present at the moment of subsidization, whereas the question
whether this subsidy still continues to cause adverse effect should be
analysed under the causation standard of an actionable subsidy claim
(Articles 5 and 6 of the SCM Agreement).95 The Appellate Body consid-
ered that the performance of the subsidy over its finite life might be
relevant to the adverse effect test. As part of this adverse effect analysis,
‘how the subsidy is affected, both by the depreciation of the subsidy that
was projected ex ante and the “intervening events” referred to by a party
that may have occurred following its grant’ must be assessed.96 If the
previous case law is to be confirmed, full privatization would be an
intervening event such that it would (presumably) disrupt the continu-
ation of the benefit and thus bring the subsidy to an end. Nonetheless,
the Appellate Body plainly stated that subsidies could still cause adverse
effects, even if the benefit (or financial contribution) has expired. Only
present adverse effects and not present subsidization (and thus benefit)
has to be demonstrated under Articles 5 and 6 of the SCM Agreement.97
Recalling that the Appellate Body in US – Countervailing Measures on
Certain EC Products acknowledged that the utility value of subsidized
equipment is still present after privatization, the Appellate Body has to
accept that the effect of the subsidy is still in place after privatization. The
extra investment level induced by previous subsidization is not altered as
a result of privatization at fair market value. In refuting the need for a
continued benefit, the Appellate Body’s extinction theory of the benefit
as a result of privatization should not be decisive, because the effect of
this subsidy remains unchanged.
In contrast to an actionable subsidy claim, a successful CVD inves-
tigation requires demonstration of a ‘continued benefit’ during the
94
The benefit test ‘should focus on the relevant market benchmark at the time the financial
contribution is granted to the recipient’ and this is ‘an ex ante analysis that does not depend
on how the particular financial contribution actually performed after it was granted’.
Appellate Body Report, EC – Large Civil Aircraft, para. 706 (emphasis in the original).
95
Appellate Body Report, EC – Large Civil Aircraft, para. 709; see also Panel Report, EC –
Large Civil Aircraft, paras. 7.218, 7.221.
96
Appellate Body Report, EC – Large Civil Aircraft, para. 710 (emphasis in original).
97
See above Chapter 4, section 4.2.3.1.

reference period. Indeed, the panel in EC – Large Civil Aircraft con-

firmed that this obligation arises from Article VI:3 of the GATT and
Article 10 of the SCM Agreement, which require that CVDs should offset
no more than the ad valorem amount of the subsidy determined to have
been granted on the production of a product.98 Hence Part V of the SCM
Agreement gives ‘rise to a pragmatic need to link a subsidy to imported
products’.99 Incorporating the valid criticism of Mavroidis and
Grossman might thus be more difficult, given that the Appellate Body
considers that the benefit is assumed to be extinguished after privatiza-
tion. Their concern can only be integrated in the current language of the
SCM Agreement if it is accepted that the benefit still continues after
privatization. Apparently disagreeing with previous Appellate Body case
law, the panel in EC – Large Civil Aircraft reasoned that the benefit
definitely continued after privatization because:
the fact that ‘the owners’ investment in the privatized company will be
recouped through the privatized company providing its owners a market
return on the full amount of their investment’ suggests that the ‘benefit’
conferred by prior financial contributions is ultimately not in fact ‘paid
for’ by the new owner.100
So the panel argues that the benefit is not extinguished as a result of

privatization, since the new owner would still continue to benefit in
wealth terms. Indeed, even if the benefit in wealth terms is captured by
the old owner (as part of the market price, as the Appellate Body seems to
suggest), the fact remains that the products in questions are still subsi-
dized: this is precisely the benefit for which the new owner will have
paid.101 The benefit of the subsidy in wealth terms at the time the
financial contribution is made is allocated over the useful life of the of
98
Equally, Articles 19.4 and 21.1 of the SCM Agreement limit the imposition of CVDs to
the amount of the subsidy found to exist, calculated in terms of the subsidization per
unit of the subsidized and exported product. See Panel Report, EC – Large Civil Aircraft,
para. 7.220. The Appellate Body in EC – Large Civil Aircraft also argued that, in contrast
to prior privatization cases, it was not in a ‘Part V context where the question arises as to
the rate of subsidization present in the product that is countervailed’. Appellate Body
Report, EC – Large Civil Aircraft, para. 724.
99
Panel Report, EC – Large Civil Aircraft, para. 7.220 (emphasis in original).
100
Panel Report, US – Large Civil Aircraft, para. 7.244 (emphasis in original).
101
In this respect, I agree with the dissenting Appellate Body member who argued that ‘the
assets of the company, to which the shares attach, do not change at all. Nor could it be
otherwise, because the buyer would then not acquire the full benefit of the bargain: the
buyer would pay for an asset (the subsidy) that had in the very sales transaction been
“extinguished”’. Appellate Body Report, EC – Large Civil Aircraft, para. 726(c).

this subsidy, regardless of whether it is still the old owner or a new one
who collects this benefit over this period. Hence the products continue to
be subsidized as required by the obligation to limit CVDs to the ad
valorem subsidy level.
Thus the extinction theory of privatization should not be endorsed
under the existing SCM provisions even if the benefit concept is –
correctly – understood as an improvement in wealth terms that disre-
gards the competitive effect. First, considering the implications of the
Appellate Body in EC – Large Civil Aircraft, the effect of a prior subsi-
dization in wealth terms remains untouched after privatization, even if
one supports the view that the benefit has been extinguished. Such
benefit extinction does not preclude a positive adverse effect test under
Article 5 of the SCM Agreement as a ‘continued benefit’ should not be
shown. Alternatively, it could be argued that the products remain sub-
sidized even after privatization and that a continued benefit is thus
present. This would align the current obligations on the imposition of
CVDs with the economic logic that privatization does not affect the
firm’s competitive position. In the next section we return to the broader
question whether the SCM Agreement disciplines only those subsidies
effectively improving the competitive position of the firm.
15.2.2.2 The distortive effect on competition

The more general concern voiced by Grossman and Mavroidis (i.e., that
the SCM Agreement should only single out those subsidies that effec-
tively improve the firm’s competitive position) can and should be con-
sidered under the causation standard for actionable subsidy claims but
not under the benefit analysis. As argued, a benefit is present when a
financial contribution (or income or price support) is made at better
than market terms. Yet a positive trade effect test presumes that this
subsidy affected the firm’s competitive position, although an exact
quantification of the trade effect is not required. Only for export and
local content subsidies could their trade effect be ignored since they are
prohibited as such but the fact that they are contingent on, respectively,
exportation and local content, justifies their being assumed to improve
the firm’s competitive position.102 An actionable subsidy claim has to
pass the trade effects test, which is, as shown by the case law, far from
evident. The jurisprudence clearly illustrates how the nature of the
102
Unless they are used in an actionable subsidy claim, which was the case in US – Large
Civil Aircraft. See Panel Report, US – Large Civil Aircraft, paras. 7.1808–7.1810.

subsidy as well as the nature of the market affects this trade effect test.103
For instance, showing adverse trade effects in the case of subsidies tied to
production (and certainly when tied to exports) is certainly easier than
showing such effects from untied subsidies.104 Further, showing adverse
effects on competitors in a duopolistic market (e.g., the Boeing and
Airbus disputes) seems easier than showing them in competitive markets
(e.g., US – Upland Cotton case), because of the direct substitution effect
as well as the presence of market power (giving the ability and incentive
to use subsidies to ameliorate its competitive position) in the case of a
duopoly. The particularities of the duopolistic market also seems to
explain why a quantitative analysis was not a strict prerequisite. In
competitive markets, a quantitative analysis has to model the adverse
trade effect caused by subsidization.
Turning to the CVD context, the competitive effect would in theory
equally be singled out, since only specific subsidies causing injury to the
domestic industry should be countervailable. However, as explained above,
the effect of subsidization on production levels could be largely neglected in
CVD investigations because the case law only mandates the establishment of
a causal link between subsidized imports and injury instead of between
subsidies and injury.105 Hence CVD investigations do not detect only
those subsidies affecting a firm’s competitive position. This forms a defi-
ciency in the standard set for demonstrating the causal relationship require-
ment, rather than one in the standard set for the subsidy determination.
Further, as mentioned before, the risk that this would result in large CVDs
offsetting insignificant subsidization is tempered by the obligation that
CVDs are not to be imposed above the subsidy level. Yet it does not
neutralize the risk that large CVDs are imposed offsetting significant sub-
sidization that does not de facto cause injury.
103
See above Chapter 4, sections 4.2.3.5.3; 4.2.4.
104
The Appellate Body in US – Large Civil Aircraft (paras. 1336, 1348) accepted the
proposition that subsidies that are not directly contingent on production or sale can
still cause serious prejudice. However, even cumulation of their effect on other subsidies
for which adverse effects have been shown is not easily demonstrated. As explained,
such cumulation only requires demonstration of a ‘genuine’ (and not a ‘substantial’)
causal connection to the same adverse effects. Cumulation was not accepted regarding
those untied subsidies that were not somehow linked to the relevant sales campaign
(e.g., relocation subsidies). In US – Upland Cotton, the panel declined aggregation of
non-price subsidies to price contingent subsidies for its price suppression analysis. See
above Chapter 4, sections 4.2.3.5.4; 4.2.4.2.
105
See above Chapter 5, section 5.2.2.3.

15.3 The specificity element

15.3.1 The rationale for the specificity test
The legal analysis in Part I explained that only ‘specific’ subsidies are
challengeable and countervailable. It is not self-evident, however, why
this specificity test was included in Article 2 of the SCM Agreement.
Broadly speaking, three reasons could be distinguished.
First, subsidies are more likely to distort trade if they are targeted
on specific beneficiaries because their relative price effect will be greater
(i.e., trade distortion rationale).106 Scholars have correctly explained
that, in basic economic models, an effort to stimulate all industrial
activity uniformly would in the longer run ‘presumably have no real
effects at all, and would wash out following some combination of price
and exchange rate adjustments’.107 At the same time, some of the same
scholars acknowledge that a government measure will in practice almost
never be so uniform that it does not de facto favour one economic sector
over another.108 Hence such measures will de facto affect domestic
resource allocation. As Trebilcock and Howse point out, investing in
basic infrastructure or education alters a country’s comparative advant-
age (dynamic concept) and thus affects domestic resource allocation.
The fact that a subsidy will de facto almost always benefit some indus-
tries over others implies that this trade distortion rationale is not
watertight.109 Although the ‘specificity’ test is inspired on its own
CVD procedure,110 the United States had even reached the conclusion
during the Uruguay Round that such a test simply had ‘no economic
106
See World Trade Report 2006, above Chapter 1, n. 4, at 51, 198; see also Jackson, above
Chapter 2 n. 57, at 296–7.
107
Sykes, above n. 5, at 512; Jackson, above Chapter 2 n. 57, at 297; see also Hufbauer and
Shelton Erb, above Chapter 2 n. 12, at 55 (in the context of general tax measures);
Benitah, above Chapter 6 n. 108, at 258–60; Luengo, above Chapter 3 n. 135, at 129–30.
108
Sykes, above n. 5, at 512 (in the context of general tax measures); Trebilcock and
Fishbein, above Chapter 3 n. 367, at 22; Trebilcock and Howse also doubt that the
exchange rate adjustment assumption holds in a world where exchange rates are
increasingly determined by capital flows rather than trade flows. Trebilcock and
Howse, above Chapter 5 n. 1, at 289.
109
See also Mavroidis, Messerlin, and Wauters, above General introduction n. 3, at 350–1.
110
Anderson and Husisian, above Chapter 2 n. 80, at 331. EU state aid disciplines also only
targeted aid favouring ‘certain undertakings’. See Evtimov, above Chapter 3 n. 434, at
456; Luengo, above Chapter 3 n. 135, at 330.

justification’.111,112 This conclusion seems, however, too stark, as the

more specific a subsidy is, the greater its potential to divert resources
away from other, non-subsidized domestic enterprises.
Second, part of the rationale for excluding non-specific subsidies
seems to be related to the nature of the government action rather than
to the question whether it affects firms’ competitive position (i.e., the
market failure rationale). In Jackson’s view, the specificity test is also a
useful tool for excluding general activities by all governments (such as
police or fire protection, education, roads) ‘which really ought not to be
brought into a countervailing duty or other international process’.113
Hence the provision of non-specific goods or services with public goods
characteristics is considered non trade-distorting, not so much because it
might not affect trade but because it would do so in a corrective rather
than distortive way.114 In contrast, specific subsidies would give an
‘unfair’ advantage to the beneficiary.115 At first sight, the targeting
principle for correcting market failures might be partly at odds with
the specificity element. As Sykes indicates, in a case where a principled
justification for a subsidy exists, this ‘will likely arise narrowly and case-
by-case’, so that the intervention will be specific.116 Hence some targeted
corrective subsidies seem rather easily to pass the specificity threshold.
For example, production subsidies directly stimulating the development
of climate-friendly goods will be specific under the SCM Agreement.117
Clearly, by crafting a category of ‘green light’ subsidies, the negotiating
members showed their awareness that corrective subsidies could very
well be specific under Article 2 of the SCM Agreement, but that these
subsidies still (temporarily) deserved green light status.118 At the same
time, respecting the targeting principle might often call for subsidizing
111
See SCM/M/48, 21 December 1990, paras. 74, 76.
112
See also Jackson, above Chapter 2 n. 57, at 298.
113
Jackson, above Chapter 2 n. 57, at 297. See also D. Palmeter, ‘Safeguard, Anti-dumping,
and Countervailing Duty Disputes in the Transatlantic Partnership: How to Control
“Contingency Protection” More Effectively,’ in E.-U. Petersmann and M. A. Pollack
(eds.), Transatlantic Economic Disputes: The EU, the US, and the WTO (Oxford
University Press, 2003), 141–73, at 155; A. O. Sykes, ‘International Trade: Trade
Remedies’, in A. T. Guzman and A. O. Sykes (eds.), Research Handbook in
International Economic Law (Cheltenham, Edward Elgar, 2007), 62–112, at 103.
114
As Sykes observes, ‘basic expenditures on publication education and road construction,
for example, can have profound effects on costs’. Sykes, above n. 511, at 31–2.
115
Sykes, above n. 113, at 103. 116 Ibid., at 101; Sykes, above n. 5, at 513–514.
117
As Green explains, ‘not all subsidies that promote environmental concerns may be specific
but some valuable ones are targeted’. Green, above Chapter 4 n. 448, at 400–1, 405.
118
See Article 8.1(b) of the SCM Agreement.

the end consumer instead of producers (i.e., consumption subsidies) or

for subsidizing activities (i.e., functional subsidies), rather than specific
sectors (i.e., selective subsidies). As is further explained below, such
subsidies might under certain conditions qualify as non-specific under
the SCM Agreement.
The market failure rationale is to a certain extent related to a third
potential rationale indicated by Sykes. Specific subsidies might be con-
sidered more likely to be motivated by protectionist considerations, and
thus by political-economy rather than welfare motivations (i.e., the
political-economy rationale).119 Although this argument has some
merit, Sykes likewise considers it to be highly ‘speculative’ because the
degree of specificity of a subsidy programme might very well be based on
political rather than economic considerations, which is illustrated by the
widespread subsidization of the agricultural sector. Again, there seems,
however, to be some validity in this justification, as non-specific, func-
tional subsidies are generally less likely to be inspired by pure political-
economy motivations than specific, sectoral subsidies.120
The specificity test as set out in Article 2 of the SCM Agreement
reflects these different rationales. The tension between the application
of the ‘objective criteria’ exemption (Article 2.1(b)) and the de facto
specificity test (Article 2.1(c)) seems to result from the different ration-
ales underpinning the specificity test. To recall, paragraph (b) states that
a subsidy shall be non-specific in cases where objective eligibility criteria
are established, whereas paragraph (c) explains that such a subsidy could
still be specific where it is de facto specific to certain enterprises. The
‘objective criteria’ test (paragraph (b)) might implement the distinction
between selective and functional government interventions whereby the
former are generally considered more vulnerable to capture by private
interests and often less targeted to the origin of the market failure
(political economy and market failure rationales). For example, subsi-
dizing R&D across all sectors might in theory be more effective than
subsidizing only high-tech sectors. Yet if the essential purpose of the
specificity test is to isolate those subsidies having the potential to distort
trade levels (trade distortion rationale), the ultimate question is, rather,
whether subsidies are de facto specific, regardless of whether their
119
See also Sykes, above n. 54, at 100; see also Trebilcock and Howse, above Chapter 5
n. 1, at 289.
120
Further, the fact that subsidies such as agricultural subsidies are widely available does
not imply that they are non-specific. See above Chapter 3, section 3.3.2.

allocation is based on objective criteria. This is established under para-

graph (c) of Article 2.1 (i.e., the de facto specificity test).
15.3.2 The application of the specificity test

So far, the case law has not offered clear guidance on how it interpreted
the purpose of the specificity test. The panel in US – Large Civil Aircraft
explicitly refuted the trade distortion rationale previously endorsed by
the panel in US – Softwood Lumber IV:121
[I]f the purpose of Article 2 is as the panel suggested in US – Softwood

Lumber IV, to prevent trade distortion, it would seem that any express or
de facto limit on the availability of a subsidy, such that it is not universally
available, would create a distortion. This is because resources would be
diverted from those enterprises not benefiting from the subsidy to those
that are so benefiting . . . However, this is not the way in which Article 2
has been interpreted in prior disputes.122
Instead, this panel reiterated the panel’s observation in US – Upland

Cotton that the purpose of the specificity test is ‘to determine whether a
subsidy is sufficiently broadly available throughout an economy’ so as
not to benefit ‘certain enterprises’.123 Although this does not reveal the
rationale for this test, it suggests that the case law has not interpreted the
specificity test so strictly that it only excludes from the scope of the SCM
Agreement those subsidies that are non-trade distortive because they are
universally available.
The Appellate Body’s findings in US – Large Civil Aircraft also suggest
that the specificity test could be useful for excluding certain functional
subsidies from the scope of the SCM Agreement. This appears to give
leeway to the market failure and political-economy rationales for the
specificity test.
The Appellate Body achieved this result by insisting on a concurrent
application of the principles set out under the three paragraphs of Article
2.1, implying that the de facto specificity test does not simply trump the
appearance of non-specificity resulting from the application of the de
jure specificity and ‘objective criteria’ tests.124 More specifically, the
question whether a subsidy is ‘disproportionately’ given to certain
121
Panel Report, US – Softwood Lumber IV, para. 7.116.
122
Panel Report, US – Large Civil Aircraft, para. 7.763 (emphasis in original).
123
Ibid., paras. 7.765–7.766; Panel Report, US – Upland Cotton, para. 7.1142.
124
Chapter 3, section 3.3.2.3.

enterprises (i.e., de facto specific) seems to be determined in the light of

the conditions for allocation assessed under the de jure specificity and
‘objective criteria’ tests. Subsidy amounts that appear to be dispropor-
tionate when assessed in the light of all eligible enterprises, could still be
proportionate when assessed in the light of the activity giving access to
the subsidy.
Consider, for example, the case of R&D subsidies. Even if their
allocation is based on genuine objective criteria, they might de facto
predominantly benefit research-intensive industries, and would thus be
considered ‘specific’ in cases where the de facto specificity test was
applied in isolation.125 Yet the Appellate Body’s jurisprudence suggests
that such subsidies would not be specific because the reason why the
subsidy is predominantly used by research-intensive industries is
explained by the activity giving access to the subsidy (i.e., R&D). To be
sure, such R&D subsidies have to be truly horizontally available among
different industries, because an explicit compartmentalization of R&D
subsidies between different industries would make them de jure
specific.126
Finally, the fact that export subsidies and local content subsidies are
deemed to be specific under Article 2.3 of the SCM Agreement seems
justified in the light of the different rationales. First, export subsidies
have a more direct effect on trade than domestic subsidies and also
favour exporting industries at the expense of the other industries
(trade distortion rationale). Moreover, export subsidies are hard to
consider as legitimate general activities by all governments (market
failure rationale). Flexibility is better laid down with regard to the
disciplines imposed on export subsidies (e.g., S&D for developing coun-
tries). Second, local content subsidies distort trade in the input market
(trade distortion rationale) and are in principle not targeted on the
market failure, if any, in the input industry (market failure rationale),
because subsidies directly provided to the input industry are more
targeted than local content subsidies provided to the downstream
industry.
125
Likewise, Green gives the example of subsidies aimed at providing energy efficiency or
emission reduction payments across all sectors. These would be de facto specific if the
de facto subsidy test is read in isolation. Green, above Chapter 4 n. 448, at 405.
126
Chapter 3, section 3.3.2.1.

18.1 PP 447 484 The Scope of The SCM Agreement Specific Subsidies

Uploaded by

Document Information

Original Description:

Original Title

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

18.1 PP 447 484 The Scope of The SCM Agreement Specific Subsidies

Uploaded by

Copyright:

Available Formats

15

The scope of the SCM Agreement: speciﬁc subsidies

A speciﬁc subsidy exists under the SCM Agreement when a government

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

15.1 The ﬁnancial contribution element

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

exclude so-called ‘regulatory subsidies’ from the scope of the SCM

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

by the panel in US – Export Restraints, the ‘ﬁnancial contribution’

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

to understand that incentive schemes not qualifying as a subsidy will

15.1.2 Subsidization by forgoing revenue otherwise due

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

international benchmark generates no similar effect with regard to

15.2 The beneﬁt element

15.2.1.1 The Canada – Renewable Energy / Feed-in Tariff

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

15.2.1.1.1 The Appellate Body’s new approach to the beneﬁt

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

The Appellate Body suggests that this benchmark is still a ‘market’

15.2.1.1.2 The relevance of the pursued policy objective under the

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

Yet not paying for produced negative externalities is not considered a

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

renewable energy in the case at hand.44 Referring to the presence of

[A] comparison between renewable energy electricity generators and

Hence considerations related to externalities often explain why govern-

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

corrective interventions from the disciplines of the SCM Agreement by

15.2.1.1.3 Limiting the implications of the Appellate Body’s new

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

production costs with competing goods preclude their presence in the

15.2.1.1.4 Was the Appellate Body’s new approach to the beneﬁt

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

under Article 3.1(b) of the SCM Agreement. As also demonstrated by the

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

Agreement. Subsidies paid exclusively to domestic producers are exemp-

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

to the objective pursued by the subsidy.53 These approaches would have

15.2.1.2The disregard of the broader regulatory

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

in disadvantaged regions was defended by the EU during the Uruguay

15.2.2 The impact of privatization and the distortive

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

and Grossman looks at the ﬁnancial contribution’s impact on the com-

15.2.2.1 The impact of privatization

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

members refuting the extinction theory seemed to formulate precisely

For example, if a government makes a ‘ﬁnancial contribution’ that

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press

(which could cause adverse effects if the ﬁrm is constrained in raising

https://doi.org/10.1017/CBO9781139046589.021 Published online by Cambridge University Press