A partnership with academia

Building knowledge for trade and development

Vi Digital Library - Text Preview

Emergency safeguard measures in the GATS: beyond feasible and desirable

Discussion Paper by Mario Marconini, 2005

Download original document (English)

What: An up-to-date resource on the issues arising on the context of WTO negotiations on an Emergency Safeguards Mechanisms (ESM) for services, it argues in favour of an ESM and special and differential treatment (SDT) for developing countries. Who: Could be of interest for anyone interested in current WTO negotiations. How: Detailed analysis on the need for an ESM for services and special and differential measures to implement services liberalization in developing countries.

9 March 2005




Mario Marconini*

Executive Summary

The paper gives an overview of the issues arising in the context of WTO negotiations on an emergency
safeguards mechanism (ESM) for services. After reviewing the history of these negotiations, the paper
makes the case for adopting an ESM in services. It explains specific aspects of such a mechanism (elements
include the safeguards situation, the definition of domestic industry, modal application, applicable
measures, compensation and S&D) and also provides options and recommendations in relation to each of
these issues. The paper addresses the arguments raised against the adoption of an ESM in services and
concludes by making suggestions on the modalities for applying emergency safeguards measures in

* The views expressed in this paper are those of the author and do not necessarily reflect the views of the
UNCTAD Secretariat.




INTRODUCTION ................................................................................................................. 3
A COMMENT ON THE NEGOTIATIONS................................................................................. 3
THE CASE FOR AN ESM IN TRADE IN SERVICES................................................................ 5
BETWEEN FEASIBLE AND DESIRABLE............................................................................... 8

Situations Justifying an Emergency Safeguards Measure ........................................... 8
Domestic Industry ...................................................................................................... 11
Modal Application ..................................................................................................... 13
Indicators and Criteria .............................................................................................. 15
Applicable Safeguard Measures ................................................................................ 17
Compensation ............................................................................................................ 18
Special and Differential Treatment............................................................................ 19
Notification, Consultation, Transparency, Surveillance Mechanism ........................ 20

BEYOND FEASIBLE AND DESIRABLE ............................................................................... 20
Negotiating Arguments .............................................................................................. 20
True Concerns............................................................................................................ 21

Fear of Abuse......................................................................................................... 21
The case of Mode 4 ............................................................................................ 22

Comfort Level......................................................................................................... 24
Burden of Proof...................................................................................................... 24
Avoiding a Monster................................................................................................ 25
Lack of Experience................................................................................................. 25
Equivalent Effect .................................................................................................... 26

Modalities .................................................................................................................. 27
Horizontal Approach ............................................................................................. 27
A Reference Document........................................................................................... 28
Schedule-based ESM.............................................................................................. 28

CONCLUSION................................................................................................................... 28



Much has been written and said about emergency safeguard measures and trade in
services. The negotiations on a possible emergency safeguards mechanism (ESM) have
taken place since the end of the Uruguay Round and have since missed their original
deadline of 1997 as well as various others subsequently. The present situation is that the
results of the negotiations will enter into effect on a date not later than the date of entry
into force of the results of the current round of services negotiations.1 There is so far very
little consensus on most of the issues on the negotiating table.

This paper does not intend to look at all technical details nor analyse all political
arguments advanced in the negotiations. Nevertheless, some repetition and some
revisiting will necessarily take place in an effort to gather all relevant elements in one
place and attempt to pass judgment, propose solutions - whenever possible. Some things
will be new, others old, others rearranged in a new context.

It should be noted that the predisposition of the author favours the inclusion of an ESM in
the GATS Agreement - not from a "protectionist" point of view, but rather because the
GATS is much too relevant an instrument not to be improved in its content and structure.
With appropriate special and differential treatment (S&D), an ESM would constitute a
useful instrument to encourage progressive liberalization, including through successive
bindings of evolving regulatory situations., while respecting national policy objectives.
An ESM could achieve this in two ways: it would help members to "sell" the logic of
GATS liberalization at home; and, it would do much to improve the overall functioning of
the Agreement.2 It is a fact that many members would appreciate the value of having an
ex post instrument such as an ESM as a means to convince reluctant sectors at home to
bind liberalization commitments. If the mechanism were adequate, it should help make
the GATS a more reliable and meaningful framework for liberalization.

One of the main premises of the present work is that without a significant change in the
manner in which the negotiations have been conducted there is not much likelihood of a
consensus by the agreed deadline. The path of endless technical details coupled with
conjectures about possible examples of emergency safeguards measures is bound to lead
the negotiations aimlessly to nowhere since it is indeed very difficult to visualize most of
the elements that some have deemed indispensable for a decision to be made. Two basic
conditions may seem to be especially necessary to get the negotiations back on track: a
collective effort to step back and look for simpler ways of resolving the most fundamental
issues at hand; and, the emergence of a significant group of countries that may support the
inclusion of an ESM on the basis of a common proposal. The final decision will, after all,
be in part political and depend, in that sense, on the objectivity of the proposal and the
support it may muster among members.

A Comment on the negotiations

The mandate for the negotiations on emergency safeguard measures in services emanates
from the text of the GATS itself which provides, in paragraph 1 of its Article X, that:

1 Fifth Decision on Negotiations on Emergency Safeguard Measures, S/L/159, 17 March 2004.
2 See also, Mashayekhi (2000).


“There shall be multilateral negotiations on the question of emergency
safeguard measures based on the principle of non-discrimination. The results
of such negotiations shall enter into effect on a date not later than three years
from the date of entry into force of the WTO Agreement.”

The difficulty with the negotiations has become explicit not only in light of the fact that
the deadline has been missed so far by several years; another still stronger factor is that,
after so much time and faced with a new approaching deadline, members continue to
disagree on the mandate itself, on what it means, stipulates or intends to do. While some
focus on the word “question” to argue that there was no original intention for a provision,
article or discipline on an ESM to be concluded by the agreed time-limit, others focus on
the phrase “results of such negotiations shall enter into effect…” to argue that the only
results that can enter into effect are precisely the ones relating to the application of a
provision, article or discipline on ESM.

The debate on the mandate of the negotiations already does much to reveal the difference
between those members that would favour the inclusion of an ESM within the scope of
the GATS and those that would rather conclude against such an inclusion. The
negotiations have, however, along the years revealed a considerable body of evidence as
to the distance from any basic consensus on the question. In fact, one could discern that
the two “camps” most involved in the deliberations have taken different approaches to
prove their point, with the paradoxical effect of extending the disagreement from the
macro-political, systemic issues all the way to micro-technical, applicability-related
matters concerning a hypothetical ESM.

In the negotiations, the burden of proof has clearly been on those that favour the inclusion
of an ESM in the GATS. Instead of assuming that it makes sense for a multilateral trading
system, particularly one that has hardly been “assimilated” by most national public and
private sectors, to have recourse to measures that can potentially increase the overall level
of confidence and security of most member countries, the course of the negotiations has
revealed that unless concrete proof is given of the need for such measures, some countries
will continue to oppose any agreement. That, in itself, represents an important departure
from the historical precedent of GATT’s Article XIX, for which most of the questioning
arose after the discipline was already fully integrated into the multilateral trading system –
and not before.

The fact is that if the burden of proof and the level of conceptual and technical
requirements imposed on the question of an emergency safeguards discipline in services
had been imposed on the question of an ESM discipline in goods, the multilateral trading
system might still today be undecided as to whether or not to venture into those waters. In
fact, if the existing discipline embodied in the Agreement on Safeguards (AS) were
subjected to the same level of questioning raised in respect of its potential counterpart in
services, it would also not withstand the test. There are still problems in the AS with data,
definitions, types of measures, applicability criteria, compensation and other issues. In
addition, one cannot say that Article XIX started when the world was already clearly on
the path of trade liberalization or that it was only acceptable because the level of
liberalization commitments was “acceptable” or sufficiently high in the eyes of some
trading partners. In fact, the level of commitments would increase with time and
safeguard measures were understood to be a facilitator of that process - and not the


It would be dishonest to pretend that the history of safeguard measures in the multilateral
trading system was itself free of problems or “unforeseen developments”. More
importantly, however, is that its mere existence was never questioned and that
improvements were always possible – as attested by the evolutions produced in the
Uruguay Round and incorporated into the AS itself. In fact, it is widely recognized that
the history of safeguards has been less traumatic to the system than its trade remedy
cousins – antidumping and countervailing duties. The level of discipline in safeguards has
traditionally been higher than those other remedies and its logic much more persuasive as

It is that logic which may contribute to dispersing the current stalemate in the safeguards
dossier of the Working Party on GATS Rules (WPGR). The general case for the inclusion
of safeguard measures to the still incipient services trade regime borrows from that logic –
as will be seen below.

The Case for an ESM in Trade in Services

If the multilateral trading system were intended to be as pure in its approach to free trade
and its benefits as is classical economic theory, it clearly would not have prospered and
expanded in the presence of various types of trade remedies. The case of safeguard
measures, the notion that in the face of a surge in imports that harm a particular domestic
industry countries may apply, albeit temporarily, trade restrictions that may help the
“adjustment” of that industry, is clearly counter to classical economic theory that is
predisposed to allowing comparative advantage to be realized based on differential factor
endowments across countries. For in that paradigm, if a country is having difficulty
competing in a particular sector, it would do better to leave the sector for foreign
producers who are more competitive and direct its resources to where it can have an
advantage – where it is more efficient in comparative terms. If the logic of the system is to
facilitate that sort of global adjustment, each country doing its part, why then should the
system permit its members to adopt measures that go in precisely the opposite direction –
postponing adjustment by applying the very restrictions the system is supposed to

Academically, the question is indeed very interesting except that there is nothing
academic about the real world in which trade takes place. The world trading system
understood the flaws of classical economic premises and functions on the basis of
economic and political realities – and not conjectures:

• Firstly, it is very clear that the adjustment costs of liberalization do exist,

particularly given the fact that in the real world there is no such thing as perfect
factor mobility – one of the tenets of economic classicism. In other words, to expect
that workers from an import-competing industry will immediately move to export-
promising industries, as if it were a matter of automatic deduction, is at best naïve.
It will indeed cost society, in some cases dearly, to replace, relocate and restructure,
and that is a reality well understood by all countries that have traditionally believed
in free markets. This is where safeguard measures have come in and facilitated the
promotion of free trade: by addressing that imperfection in factor mobility and in
markets in general, allowing countries to seize free trade, albeit gradually, so as to
render possible a new relocation of resources in time.


• Secondly, it is widely known that trade protection increases prices for the domestic
consumer and that fairly often the cost of adjustment is larger than the benefits a
country could obtain from the expansion of trade (in other words, often it would be
better, macro economically, to open up the economy than not to). However, even
then, countries tend to apply protection because the political economy of trade is
such that governments prefer to make concessions to vocal import-competing
producers than to silent import-consuming consumers. Producers tend to organize
and mobilize when faced with direct competition; consumers, on the other hand, do
not necessarily notice the benefits of trade, which are spread across society.
Safeguard measures, as other trade remedies, tend therefore to be favored by
governments as a means to appease potential “losers” while, hopefully, disciplining
them into restructuring within a particular window of time.

Is there any reason to suspect that trade in services is fundamentally different from trade
in goods in those respects? Not at all. There are in fact a number of additional reasons as
to why safeguard measures may be even more necessary in services than in goods trade.
Those reasons have to do with the specificity of trade in services. Thus, the fact that trade
in services is regulation-driven or intensive renders it much more delicate than trade in
goods in the following ways:

• It is very difficult, if not impossible, to have a high degree of clarity on what may

happen with one’s market once there is a regulatory change such as the one induced
by liberalization. Unlike trade in goods, where barriers tend to be at the border,
quantitative, and much simpler in their nature, barriers to trade in services (as
captured by GATS Articles XVI and XVII) are inherently complex (serving various
policy objectives and not only trade protection) and only experience can shed some
light on the real impact of their elimination in a liberalization process;

• It is natural that governments be cautious about liberalization in services, in the

presence of such uncertainty about its potential impacts. Even when countries have
already opened up their markets, much caution has to be exerted in order to monitor,
re-evaluate and perhaps re-regulate sectors depending on the assessment of evolving
market realities;

• The nature of trade in services therefore renders it especially prone to unforeseen

developments and countries could be assisted in their liberalization efforts by
having mechanisms in the agreement that manage to safeguard domestic sectors
from injury or total devastation;

• The liberalization of trade in services involves much more than the liberalization of

trade in goods because it touches on a myriad of regulatory aspects absent from
“visible” trade, in addition to potentially affecting a considerably greater cluster of
economic activities and related employment; once again, the sheer “size” of the
impact as well as the uncertainty about the precise nature of the impact down the
“services chain” render governments especially cautious even when liberalization
has already occurred;

• Unlike trade in goods, the liberalization of trade in services takes place through the

elimination of essentially domestic measures – truly the “field of operation” of any
government; the sensitivity of this aspect for members should not be


underestimated, especially in the context of the widespread questioning and
protesting that has become a fact of life since the Uruguay Round;

• In much the same way as the founders of GATT decided to opt for having a

mechanism that would assuage the domestic concerns of industries that felt
threatened by liberalization as a means to spearhead the cause of free trade while
minimizing opposition at home, the founders of GATS should accept that for most
countries the experience with services and services liberalization – within the
context of GATS or autonomously, is still a relatively novel affair, one that still
requires much caution and convincing of national public and private sectors.

It should be noted that the negotiations that led to the GATS itself demonstrated some of
the sensitivities of trade in services. The fact that the agreement is endowed with a
perceived high degree of flexibility in terms of its liberalization aims, objectives and
mechanisms reflects the caution with which the vast majority of countries faced those
negotiations. Scheduling via positive lists of sector or sub-sectors, specific exemptions
from the most-favored-nation principle, the possibility to leave specific modes of supply
unbound for scheduled sectors or sub-sectors, as well as the widespread recourse to vague
scheduling techniques such as mere references to economic needs tests in the absence of
criteria for their application – all these aspects of the original GATS agreement attest to
the fact that services are perceived as sensitive in their nature. Big and small countries
alike revealed vulnerabilities in the negotiations, some seeking flexibility in the
application of certain principles, others seeking the exclusion of certain sectors from the
application of the agreement as a whole.

The Uruguay Round was, of course, sui generis in the way that it accomplished
concomitantly the conclusion of a full-fledged agreement alongside the negotiation of
more than a hundred schedules of specific commitments on services for the first time in
history. It was to be expected that the schedules would be conservative in their scope
since for most countries the novelty and sensitivity of the theme was enough to warrant
full caution. It is interesting to observe, however, that if the premise is true that the GATS
is indeed a widely permissive agreement, perhaps the level of commitments could have
been more ambitious than it was.

Clearly, the comfort level perceived by most delegations was not high enough for that to
happen in those days. A safeguard discipline may be the element that was missing then
and that may make the difference now in order for countries to feel confident in binding
evolving, or already liberal, regulatory situations. After all, most of the instruments
available for members to use when binding commitments tend to be of an ex ante nature.
The ex post nature of an ESM whereby countries may have recourse to measures after the
commitments have been made and put into force may represent an important guarantee
for those who are reluctant to commit to higher levels of liberalization.

All arguments, whether in goods or services, with respect to safeguard measures seem to
point to one salient fact: that even though it would be convenient if markets were perfect
and perfectly readjustable, and services were not regulated in diversified and complex
ways, the most important contribution that an ESM could make has not much to do with
good economics, but rather with political expediency. It is an undeniable truth that
countries would feel better if they had a reliable mechanism with which to anchor their


commitment to liberalization in services. In that context, two preliminary conclusions
should underpin the rest of the paper:

• The issue should not be whether an ESM should be desirable or not; the issue

should be which ESM could be reliable and serve the purpose of complementing the
current GATS framework so as to provide for progressive liberalization in the
context of a higher (and not lower) level of predictability, transparency and
discipline for all members;

• One central purpose to be served by an ESM is political and should not be confused

with good or bad economics, or good or bad technicalities; the fact is that good or
bad economics (or technicalities) were never in question during the fifty years of
Article XIX/Agreement on Safeguards because the objective of having safeguard
measures was clear and unrelated to those matters: to provide a means to deflate
opposition to liberalization while perhaps making use of the opportunity to place
affected industries on the path of restructuring.

Between Feasible and Desirable

Even when faced with justifiable cases for the application of emergency safeguards
measures, many delegations have argued in the negotiations that technical aspects may
render impossible such an application. Much has been done regarding the vast array of
technical issues and little consensus has been achieved so far. The fact is that the less
feasible things seem to be, the more undesirable they also become in the eyes of many,
including a number of developing countries. Alternatively, proving that all or most of the
technical issues can be resolved still would not achieve the acceptance of reluctant
countries since in their eyes, even if feasible, safeguards in services may not be desirable
anyway. To prove that technical issues have solutions constitutes therefore a necessary
but not sufficient condition for an overall agreement on an ESM for trade in services.

In what follows, an overview of the main technical issues will be attempted.

Situations Justifying an Emergency Safeguards Measure

The objective should be the same as in trade in goods if and when one can define clearly
that injury to national services and/or services providers constitutes an emergency
situation that requires redressing. The approach adopted by the AS may be the most
straightforward for the GATS as well, focusing on three main conditions:

“…that such product is being imported into its territory in such increased
quantities, absolute or relative to domestic production, and under such
conditions as to cause or threaten to cause serious injury to the domestic
industry that produces like or directly competitive products.”

The conditions are therefore the increase in “quantities” (which admittedly would have to
be adapted to the case of trade in services – possibly “volumes” and “values”), the injury,
and the causal link between the increase and the injury. It is good to emulate the AS in
regard to the objective justifying the application of an emergency safeguards measure
because the language is clear, direct and avoids ambiguities while upholding the ultimate
aim of a safeguard. Objectives such as facilitating the development of services industries


in developing countries go beyond the scope of a “traditional” safeguards provision.
While deviating from the traditional objective would have the potential to complicate
things further,3 such an approach needs further consideration, including through additional
research and analysis.

The text of the AS has improved considerably on Article XIX. In addition to adding the
reference to “absolute or relative” quantities, historically a source of much confusion, the
AS eliminated two crucial elements from Article XIX which also have special relevance
to the discussions on safeguards in services – namely: the reference to “unforeseen
developments” as being the cause of an increase in imports; and, the linkage of
unforeseen developments and the increase in imports to “obligations” – “specific
commitment” in the language of the GATS.

In the WPGR, it has been generally understood that safeguard measures constitute
extraordinary remedies. Some members have taken the position that the concepts of
emergency and unforeseen developments are needed to somehow highlight the
extraordinary nature of an emergency safeguards measure. Others have great doubts about
whether the concept of unforeseen developments should be introduced as a condition for
the application of a safeguard, arguing that it would introduce ambiguity while
functioning effectively as a fourth condition (in addition to proving an import surge,
injury and the causal relationship between the surge and the injury) for the application of
a safeguards measure.

In the case of goods, the elimination of the reference to “unforeseen developments” was
indeed seen as an improvement precisely because of the ambiguity it often created.
Historically, countries would take liberties with the concept until it was deemed by most
experts to have become irrelevant as a criterion.4 If a GATS provision were to follow the
example of the AS, this would result in applying safeguard measures only in the case of
clear-cut surges that caused or threatened to cause injury to a domestic sector or provider.5
It would force members to focus on the three main conditions that should be met for the
application of an emergency safeguards measure. The improvements introduced in the AS
clarified that the drafters were not concerned with the fact that an increase in imports is a
natural occurrence in the wake of any liberalization effort and that safeguards would
potentially, in a sense, go in the opposite direction – an argument often put forward
against the introduction of an ESM in the GATS. The main focus of the drafters in the
case of AS was the fact that, regardless of the reasons for the increase in imports, in some
cases there will be injury due to the pace of liberalization, and countries need time to

There has also been a great deal of attention given to whether the surge and the injury
should be linked to a specific commitment. Once again, the AS has disposed of the
requirement that such a linkage be established. In services, this issue may be made simple

3 In a number of so-called “Europe Agreements” the European Union has taken an interesting approach as
regards the establishment of Community companies and nationals in partner countries. In fact, these
agreements include as justifications for safeguard measures the process of restructuring, serious social
problems, reduction of total market share of national companies or the advent of newly emerging industries.
4 The famous 1950-51 Working Party on “Withdrawal by the United States of a Tariff Concession under
Article XIX of the General Agreement” which interpreted a mere change in fashion as an “unforeseen
development” was perhaps the beginning of the end of the concept (Jackson, 1969).
5 Note that in the case of trade in goods, the concept of "unforeseen developments" is indeed absent from
the AS. It is, however a central element of Article XIX of the GATT.


in one aspect: countries that feel that there is an injury deriving from a surge of imports or
increase in the supply of foreign services will only seek an emergency safeguards measure
if a commitment has been made: only then would the measure potentially imply a change
in the schedule. In other words, safeguards measures would only be sought if
commitments had been made in schedules and a member felt that it needed to change it
anyway. If an increase in the supply of foreign services occurred in a sector not
committed in a particular schedule, the member in question could proceed without
complying with safeguard procedures since the measure would not necessitate a change in
the schedule.

If the GATS were to emulate the AS in this aspect as well, the most important effect
would be that countries would not have to prove that a particular commitment is indeed
the cause of a particular injury. An example could illustrate the worst case scenario in this
context: a country that in its schedule committed to a maximum of 51% of foreign capital
participation in mode 3 for a particular sector and which takes the opportunity when faced
with an increase in the relevant supply of services to decrease that limit to 30% even
though that may not have anything to do with the increase. In a case like this, it would
have been better for the country in question to have to show why the 51% ceiling had
anything to do with the injury. It should be noted, however, that the abuse in the
application of the emergency safeguards measure like this could still be avoided by
requiring that applying countries justify why a particular measure is being applied instead
of others.

Some countries have also proposed that safeguards measures should only be applicable to
future commitments and not to the ones already inscribed in schedules. This is clearly a
political position which links the negotiations on a systemic issue (an ESM) to the
negotiations on market access. At the same time, linking the negotiations on an ESM to
negotiations on market access might have broader implications. To prevent such results
turn from becoming problematic for developing countries, careful attention is required.

Absolute or relative
increase in quantities
under such conditions as
to cause or threaten to
cause serious injury

Most straightforward language; avoids

Follow AS


Following AS, might avoid ambiguities Consider both options, either following the AS
(dropping the unforeseen developments requirement) or
GATT Article XIX (containing the unforeseen
developments requirement).

Relation to specific

Relating emergency safeguards measures s to
specific commitments would serve to force
members to be more disciplined in the
application of such measures; alternatively,
disciplines on justifications as to why choosing
one over another measure would also resolve
the problem.

Follow AS: not relate only to specific commitments but
also require that the change in schedule would need to
be justified as to why that change helps in the injury
situation; in addition, one could envisage disciplines on
justifications as to what are the reasons of choosing one
over another measure.

Apply only to new

Merely a political position Apply to old and new specific commitments.


Domestic Industry

The definition of domestic industry has not been a major problem in the negotiations with
respect to situations where there is a movement of suppliers or consumers across borders
– primarily, therefore, modes 1, 2 and 4. The case of mode 3 is in fact sui generis for a
number of reasons. It could be said, however, that safeguard-related issues relating to the
movement of capital with a view to commercial presence (the pre-establishment phase)
tend to be less sensitive than the ones involving the presence itself in a foreign market
(post-establishment phase) since once there is a presence, acquired rights in the market of
the presence become an important consideration.

Essentially, the issue is whether in mode 3 the definition should be based on nationality or
residency of foreign established suppliers (FES). The nationality criterion is the one
reflected in Article XXVIII of the GATS, particularly if looked at as a package of
provisions (namely, XXVIII (k), (l), (m) and (n)). The residency criterion, whereby firms
“duly constituted or organized under the applicable law of the [host] country” are
considered domestic firms, is captured only by Article XXVIII (l). Additionally,
considerations relating to pre and post-establishment are also crucial as their implications
for the definition of domestic industry differ considerably.

From an economic standpoint, it does not make much sense to differentiate between
suppliers of national or foreign ownership and control since, once established, national or
FES alike add value to the national economy, engage in foreign trade, contribute to
employment, pay taxes, and so on. Economically, the question for the safeguards debate is
whether the aim of a measure should be to protect capital owners or the labour force,
which would correspond respectively to a definition based on nationality or residency.
Generally speaking in any case, the type of emergency a safeguard measure should be
addressing has more to do with labour relocation than with the protection of national
capital owners per se.

From a legal standpoint, however, things tend to be less straightforward as countries differ
considerably as to how they define national and foreign suppliers in their own regulatory
regime. Thus, for some countries, the nationality criterion applies domestically whereas in
others the law provides for full national treatment in post-establishment situations so that
they could not even discriminate against FES even if they wanted to. For these countries,
discriminating against FES would indeed be easily resolved against government
authorities in local courts – no need to resort to WTO dispute settlement procedures on
that account. This is indeed a major barrier to a common definition of domestic industry
in the context of an ESM in services since it would not be reasonable to make possible the
application of a measure only to a certain group of members and not another. It would be
unreasonable to require that a safeguard measure is only possible in mode 3 on the basis
of a definition of domestic industry that, albeit in conformity with the internal definition
adopted by some countries, refers only to a sub-set of the actual domestic industry of
other countries.

It would seem to be easier to visualize the application of a definition of domestic industry
on the basis of residency of the suppliers in the case of pre-establishment where the main
applicable measure would be restricting the entry of new suppliers. Most of the problem
seems to relate to the post-establishment phase where the grounds for discriminating in
the same market are not always obvious from an economic standpoint and creates


significant legal difficulties. Legally, in addition to the difficulty in discriminating among
resident firms for some countries, even when it is possible to discriminate, countries
would still need to be clear about the boundaries of acquired rights for resident firms. One
possible boundary in that context may be whether the supplier is allowed to expand its
presence – an approach which would preserve all other rights normally accruing to that
presence.6 Acquired rights is not an irresolvable issue but it is a fact that a common
definition in that context may be less desirable than a definition which is clearly grounded
and justified on the particular features of each case.

The complexity of the issue and the diversity of regulatory regimes applying to mode 3
would seem to warrant a cautious approach on the issue of domestic industry. An
important point, however, is that there is no overriding reason as to why a common
definition should be applied across all members if the members themselves differ in their
own definitions. A common definition in that sense would be “asking too much” of
members while it would not fulfill its main objective which would be to allow countries
recourse to “relief” from liberalization in order to adjust to new competitive realities,
thereby appeasing political opposition to liberalization moves.

It is here suggested that the burden of defining domestic industry be transferred to each
member seeking an emergency safeguards measure. Each member should have its own
reasons for applying a particular definition of domestic industry and, as long as it was
objective, comprehensive and transparent as to those reasons, it could have (or not have) a
good case. After all, reasonableness should also play its role. If a member, for example,
defines its domestic industry on the basis of nationality but that ultimately represents, say,
20 percent of all resident suppliers operating in the national market, that would simply not
withstand any test no matter what that Member’s regulatory regime foresees. The secret to
avoiding endless debates on the issue is accepting that members should have the right to
put forth their case; the rest of the solution to the issue lies in the disciplines that will
necessarily have to be included regarding the wholesomeness of the information provided
through petitions and subsequent documentation.

Finally, still in the context of defining a member’s domestic industry, there is the issue of
“like services”. If the intention were to resolve this issue before thinking up an ESM,
perhaps the working party should better give up on the task altogether. After all, in trade
in goods where “products” tend to be more visible than services, the definition of like-
product has never been very concrete. It is widely known that in antidumping cases, for
example, authorities tend to resort to eight- or more digit Harmonized System
classifications, which is clearly an inferior approach to checking the substitutability
between imported and locally-produced products. In services, where the definitions are
still rather imperfect (the Central Product Classification of the United Nations, the CPC, is
itself inadequate in many respects), substitutability could be coupled with the modes of
supply which at least go some way in determining likeness across members. Also, the
criterion of end use, whereby like services are those that have the same end use by
consumers, might be of relevance. It should be up to the members applying an emergency
safeguards measure to be specific about their definitions of like services. This is why in
this matter as in others involving definitions the best that could be expected is for a
mechanism to stipulate strong disciplines on the provision of information on why one
definition was preferred over others and how it fits into the overall structure of the case.

6 See Eugui (2002).


Modes 1, 2
and 4

Definitions based on the movement of
suppliers and consumers from one territory
to another

Domestic industry would be services suppliers that are resident in the
country of origin of the supplier or the consumer. Definition could be
common but depends on overall idea for the concept as a whole (see

Mode 3 –

The definition of Article XXVIII only
tracks broadly with the internal definition
of some, and not all, members

Leave it up to each member to decide whether to adopt the Article
XXVIII or any other definition.

Mode 3 –
nationality vs.

Members have widely differing regulatory
regimes in that respect.

Leave it up to each member to determine whether or not to adopt a
definition based on one or the other criterion, or whether to use its own

Mode 3 –

The core of the issue relates to the
application of emergency safeguards
measures to post-establishment on the
basis of a nationality criterion.

As the suggestion here is that each country can adopt either the Article
XXVIII or the nationality criterion or any other definition as the basis,
the problem would not present itself as a common concern. If the
nationality criterion were to be applied, members would need to
establish clearly the boundaries of what was to be considered acquired
rights in each specific case.


It is not possible to have a common
definition that all countries could accept
and that could provide for balance in the
rights and obligations that members would
have in the application of safeguards

Leave it up to each member to define its own domestic industry in each
case. The risk of abuse should be minimized through stringent
information requirements and justifications as to the definitions adopted.
Reasonableness should apply.

Like services Difficult in goods, even more difficult in
services; criteria may be substitutability,
modes of supply and end use by

Leave it up to each member to explain and justify its own definition of
like services.

Modal Application

The debate on the application of emergency safeguards measures to services has also
focussed on whether that application should extend to all modes, as well as on whether it
could be cross-modal in nature.

Not applying safeguard measures to all modes would ab initio create an imbalance since
the intention of the agreement was that it covered four modes of supply and not just some
of the four. The problem with applying to all four modes has more to do with the
feasibility than with the desirability of doing so. In other words, not applying a safeguards
measure to any of the modes should only be acceptable if it were indeed infeasible since
otherwise the Agreement itself would be creating different levels of rights and obligations
among the four modes – something not intended by the drafters.

It is true that each mode reveals slightly different sets of technical issues that may indeed
require differentiated treatment from a safeguards perspective. Mode 1, for example, may
always have an enforceability problem that is not particular to any other mode and could
require some attention to the issue of avoiding circumvention “within the mode” (for
example, circumventing a safeguard on the cross-border supply via courier services by
“migrating” to e-mail services to achieve the same previous commercial objective). Also,
the application to mode 1 should perhaps require some caution with indicators and criteria
that may have special relevance in cross-border transactions such as the linkage with other
services sectors, particularly the ones relating to infrastructure. Mode 2 would seem to
pose less problems, with the focus of any emergency safeguards measure being the
consumer him/herself. At the same time, such measures might, indeed, have an impact on
consumers, including by limiting the transfer of funds or by placing other restrictions on
consumption abroad.


Domestic industry in the case of mode 4 should also be interpreted to include both natural
and juridical persons since the movement of natural persons may take place independently
or as a result of a juridical person’s contracting, sending and/or transferring of those
persons. The competing industry, therefore, would be both independent natural persons as
well as resident/national firms involved in those activities. The fear that this mode may be
the easiest to use given previous practice and the sensitivity of the issue for all countries
should be addressed through stringent information and justification requirements (see
section “abuse of application” below) as well as through effective S&D mechanisms.

On the account of infeasibility, the main problem would, once again, relate to mode 3. It
is only mode 3 that exhibits the problems mentioned above in relation to
nationality/residency criteria, pre and post-establishment issues, including acquired rights,
and varying regulatory regimes across countries. If, for example, the definition were
adopted on the basis of the nationality criterion, applicable safeguard measures could
include those that discriminated against FES in the national market, in effect spanning the
whole spectrum of possible measures between pre and post-establishment situations. If,
on the other hand, the definition adopted were based on the residency criterion, safeguard
measures would not be applicable to post-establishment commercial presence (since FES
would be automatically part of the definition of domestic industry) so that only a portion
of mode 3 would effectively be under the scope of emergency safeguards measures (only
the pre-establishment portion). Were the suggestion made above to prevail, namely that
countries would define domestic industries by themselves (subject to stringent
information and justification requirements), safeguards measures in mode 3 would have a
variable application, depending on the merits of each case.

As to cross-modal application, the determining factor should be the origin of the injury
and the measure necessary to “correct” it. In that sense, the consideration of the issue is
already limited by the reality of the case. If, for example, a domestic sector is suffering
from an increased cross-border supply of a competitive foreign service it is hardly
imaginable that restricting supply via commercial presence could attenuate or eliminate
the injury: the urgency of the situation would dictate the application of an emergency
safeguards measure to the “injuring” mode. Even though the GATS would seem to favour
the interpretation that a specific firm is the same supplier across modes, it should be
desirable that safeguard measures address the specific importing surge that is at the origin
of the problem and not other transactions. When the mode causing the injury is clear, it, in
fact, seems to be a non-issue: by applying emergency safeguards measures to modes other
than the “surging” ones, a member would not only fail to diminish the injury but would
also run the risk of aggravating the situation since the originating flux would continue to

Admittedly, applying an emergency safeguards measure only to the mode where the
increase in foreign supply originates may leave open the possibility of circumvention
across modes. Members would always have the option to seek additional safeguards
measures to address the “migration” to other modes. An alternative solution, albeit more
difficult and energy-consuming, would be for the WG to agree on anti-circumvention
provisions. Given the current stage of the negotiations, it may be overly ambitious to take
on yet another complex task.


Application to all

The closest one can come to inferring the
original intention of drafters

Endeavor to apply to all modes in order to avoid imbalances

Mode 1 –
technical issues

Problems of enforceability and possible

Include it

Mode 2 –
technical issues

Incidence over consumers Include it

Mode 3 –
nationality vs.

Once again, application of emergency
safeguards measures to mode 3 would depend
on the definition adopted for domestic
industry. If the option is not to have a
definition, application to the mode would
vary according to each case.

As the suggestion here is to adopt either the Article XXVIII , or
the nationality or any other criterion as the basis, the problem
would not present itself as a common concern. If the nationality
criterion were to be applied, members would need to establish
clearly the boundaries of what was to be considered acquired
rights in each specific case.

Mode 4 –
technical issues

Domestic industry as both natural and
juridical persons, the competing industry
being, therefore, both independent natural
persons as well as resident/national firms
involved in those activities.

Adapt definition of domestic industry. In addition, elaborate
stringent information and justification requirements for the
application of emergency safeguards measures to this mode, as
well as effective S&D mechanisms


Focus should be on applying measures to
modes that are behind the increase of service
supply that causes injury

Be clear about applying emergency safeguards measures to where
the cause of the injury is.

application – anti-

It is very likely that members would
“migrate” from one mode to another when
faced with an emergency safeguards measure

Plan the negotiation of an anti-circumvention provision at a later
time, after some experience with whatever mechanism comes into

Indicators and Criteria

The question of which indicators and criteria to rely on when applying an emergency
safeguards measure is crucial in the case of services. Not only is trade in services more
complex than trade in goods in all respects, quantitative and qualitative, but the influences
impinging on the services market, defined according to the GATS as both its national and
international legs, tend to be much more diversified than in the case of goods. Public
interest issues such as consumer satisfaction tend to be much more readily relevant to
services transactions that often depend on direct client/supplier contact than to goods
production. In services, it would be more difficult not to take a number of non-trade-
related factors into consideration when determining the applicability of a safeguards
measure, given their nature and the nature of their commercialization.

There is wide consensus that, were the working party to agree on an ESM, it should have
a three-prong investigation procedure corresponding to the determination of an importing
surge, an injury or threat thereof, and a causal link. Clearly, some adaptations would need
to be made such as the requirement that increases in “imports” be reported not in
“quantities” but in volumes or values. What matters, for example, is not only the number
of professionals that entered the country during a particular period but also the value of
their work as reflected in wages and fees.

In the debate in the working party, the lack of reliable data has often been put forward as
one of the main obstacles to a reasonable application of an ESM in services. The fact is,
however, that data are available even though they may be more difficult to come by in
respect to an increase in the supply of relevant services than in respect to injury indicators.
It so happens that injury indicators are normally readily available with affected,
petitioning firms who normally have to produce data to co-substantiate their injury claims
– in the absence of which no safeguard measure can be taken anyway. There is in any case
a host of types of data that can corroborate a member’s claim that there has been an
increase in the supply of a particular service.. The important aspect to consider is that the


burden of proof will be all on the petitioning firm, which will have to make clear in its
petition to the trade remedy authorities that there was an increase in supply, an injury and
a causal relation.

As in the AS, an ESM in services should allow much room for authorities to look into
additional possible causes for injury other than increases in the supply of foreign-origin
services. This is clearly one of the areas, however, where a provision on safeguards in
services could endeavor to go further than has traditionally been possible in goods, with
the justification that any such new instrument needs to have a high level of discipline in
order to contribute to greater transparency and predictability in the GATS – and not to
greater ambiguity and discretion. By having a strong obligation in this respect, national
authorities would have to organize themselves better, perhaps promoting greater
coordination among ministries and agencies involved in economic affairs, so as to extract
the best possible scenario as to the real causes for injuries in the services sector. The
responsibility to produce an analysis of all other possible factors should be the trade
remedy authority’s; whether it should do it alone or in coordination with the rest of a
particular government’s relevant organs would be an internal decision.

Another area where the trade remedy authorities could coordinate and cooperate with the
rest of the government should be in what could be referred to as “an economic interest
test”. Once again, the special situation of services trade would warrant a broad analysis as
to whether it should be in a country’s wide-ranging economic interest (and not just in a
sector’s interest) to apply a particular emergency safeguards measure. This sort of
provision would allow, for example, finance ministries to have a say as to whether the
application of a safeguard measure would make sense in the context of an economic
stabilization plan aimed at promoting competition and controlling inflation (safeguards
measures do tend to go in exactly the opposite direction). The decision as to how to
coordinate and cooperate within a government will be internal but it would be crucial for
an ESM in the GATS to provide for a strong discipline on the basis of an economic
interest test.

The public interest provision that appears in Article 3 of the AS is also relevant for trade
in services but, once again, should be deepened in its level of disciplines relating to public
notices, public hearings, questionnaires to importers, exporters and other interested
parties, an obligation to take presented elements into consideration as well as the
production and publication of a detailed report on the findings. The importance of public
interest in the determination of the adequacy of applying an emergency safeguards
measure in services cannot be understated, especially as the agreement is still new and
requires tightening of its disciplines while providing for recourse to instruments that may
assist the liberalization effort.

It is hoped that tight disciplines in indicators and criteria, including an economic interest
and a public interest test as stringent conditions for the approval of the application of an
emergency safeguards measure, in addition to forcing governments to be comprehensive
in their analysis of their own real national interest, will also permit these governments to
say no to protectionism for protectionism’s sake. It should be noted also that with this
level of discipline on indicators, criteria and conditions for application, the issue of
allowing members to define their own domestic industry and other elements becomes
much less important. That is the main bias of the present proposal for an ESM in the
GATS: permit countries to come forward with definitions and arguments of their own


with regard to each case in services (which will always tend to be very specific with
respect to the various elements that impinge on trade in services) while obliging them to
be as comprehensive, disciplined and criterion-based in the information they base their
assessments on, as well as in the tests to which they submit each case before it can be

Three-prong investigation

Just like in goods: increase in quantities (which in the case of services
should be (volume or value), injury and causal link between the two

Specify the three-prong procedure
in the mechanism

Lack of reliable data More difficult in the case of increase of services or services supply
than with respect to injury indicators, which are usually gathered by
the petitioning firms

Detail type of data necessary for
both increase in supply and injury

Causes for injury other
than increases in the
supply of foreign-origin

GATS should go further than GATT or the AS in this respect Include specific language on the
need for special attention to this

Economic interest test This should go beyond GATT and the AS as well with a view to
limiting the application of emergency safeguards measures only to
cases that made an economic case; this allows the government as a
whole to determine whether a particular measure is good or not for
the national economy

Include explicit language on this

Public interest test Could borrow from GATT and the AS but should go further,
detailing procedures and requirements

Include explicit language on this

Applicable Safeguard Measures

There is an apparent consensus about the application of emergency safeguards measures
on an MFN basis and that it could take the form of a suspension of specific commitments.
In this case, it would not be necessary to draw up a list of permitted measures, the
schedules themselves providing a variety of possible safeguards measures. The argument
that additional commitments are different enough to warrant an exclusion from
application is not well founded since an injury could result from such commitments as
from any other “opening” provided for in the schedules. There is also merit in trying to
stimulate the application of quotas and subsidies.7

There seems to be a considerable agreement that measures would need to be the least-
trade-restrictive, the minimum necessary to effectively remedy the injury. It would seem
that engaging in the negotiation of criteria for the least-trade-restrictive options should not
be necessary, especially if a necessity test is applied to applying members whereby they
would have to prove why a particular measure was considered the least-trade-restrictive.
On the need for emergency safeguards measures to be easily undone, that could be best
achieved through quantitative restrictions and subsidies – both more amenable to quick
elimination as well as degressivity – or the progressive liberalization at regular intervals
of an such measure.

There is in fact wide agreement on the temporary nature of safeguards measures and
degressivity but not on the specific duration or the possibility of renewal. The duration
could be something similar to the AS (four years). Developing countries should have
longer time periods as well as a possibility for renewal. Such a renewal, which should not
result in an overall period of application exceeding between eight or ten years, could also
be subject to more stringent requirements.

7 See note prepared by Patrick A. Messerlin, “The Emergency Safeguard Measures in GATS”, (1997).


There is no consensus on the need for provisional measures, some pointing to the risk of
abuse, others to the difficulty in determining reimbursements in cases of wrong
application. The notion, common to goods, that it should not decrease the level of supply
below the average of a representative period also applies, although it may not be possible
to quantify volumes, etc., in some cases.

Application on the
basis of MFN

Tracks with AS and should be applied in

Apply emergency safeguards measures on MFN basis

List of permitted

There is no need to do that given the
measures already in the schedule and the
fact that countries should need to abide by
a necessity test: prove why a certain
measure was applied over another.

Introduce necessity test


An emergency safeguards measure may
also be the suspension of an additional
commitment since that commitment could
be at the root of the injury being caused

Provide for the possibility of application to additional

Least-trade-restrictive The emergency safeguards measure
should be least-trade-restrictive. One way
to ensure that is also through a necessity
test (simpler than drawing up a list of
permitted measures)

Introduce necessity test

Degressivity Difficult to implement but relevant as a
concept. Could be included in the
mechanism and members should be made
to demonstrate how it is being achieved

Include it in the ESM

Duration and renewal Could follow the AS example, allowing
for renewals (perhaps with more stringent

Include duration and maximum duration encompassing
extensions and provisional measures (if included), and make
renewal possible for developing countries only.

Provisional measures Could be applied to see if it works. Risk of
abuse is indeed rather big.

If included, should be only on a tentative basis (until the next
review, for example).

S&D Would ensure that safeguards measures
would not be abused in case of developing
country exports.

Options include the non-application of an ESM to mode 4
exports of developing countries, granting developing countries
longer time periods for the application of emergency safeguards
measures, as well as granting the possibility for renewal to
developing countries only.


The notion that members applying emergency safeguards measures should compensate
other affected members so as to “maintain a substantially equivalent level of concessions
and other obligations” (Article 8 of the AS) would have the same effect in services as it
has traditionally had in goods: to serve as an important incentive for countries not to apply
emergency safeguards measures to begin with. Governments frequently would in services,
as they have in goods, shy away from applying safeguards measures given the “cost” of
defending their application when it implies compensation on the part of other sectors of
the economy – sectors which in principle had nothing to do with the problems of the
sector being favored by the trade remedy.

The choice of including or not a compensation mechanism for safeguards measures has
therefore much to do with whether or not the intention would be to force the application
of such measures only in really unavoidable cases. If that were not the intention, a
compensation mechanism could be kept out of the GATS for the time being given that
there could be no other objective to be served by its inclusion.


It should be noted that, even if there were a wide agreement to include compensation in
the agreement, the problems with its application would be enormous. Not only are
services notorious for their incomplete and deficient data regime but their barriers, their
introduction or elimination, are also extremely difficult to quantify, so that terms such as
“substantially equivalent level of concessions” or even the one suggested by some
delegations in the working group, “general level of mutually advantageous
commitments”, can only give rise to endless negotiations and, perhaps, disagreements.
The same problem would, of course, apply to the recourse to retaliation which a
traditional compensation scheme normally foresees in cases where there is no agreement
between the applying and the affected parties.

Questions such as whether compensation should apply in the same mode of supply or the
same sector or sub-sector are easily resolved: they should just follow the goods precedent
and be open to where compensations are possible with no need for strict correspondence
in terms of modes or sectors. Whether it should be applied on an MFN basis also is easy:
it should not, since it refers to specific affected members (as in goods trade). The
difficulty is to decide whether the conditionality that compensation requirements represent
is too strong to make the application of emergency safeguards measures out of question
for some time – in which case, it may be better left outside the GATS.

The difficulty with compensation is that while the intention is not to make an ESM in
services too “prohibitive”, the level of discipline that it would add to the agreement would
do much to ensure that there would be no abuse in the application of safeguard measures.
Perhaps one solution would be to “trigger” compensation only after the first or second
year of continued application of an emergency safeguards measure. That way
compensation would still function as a significant disciplining factor by acting as a
deterrent against renewals or extensions of safeguard protection.

Objective of

To serve as an important incentive for countries not to
apply ESMs unless absolutely necessary

In the application of compensation, differentiate
between developing and developed countries.


In addition to bad statistics overall, the introduction or
elimination of barriers to services trade is difficult to

If one were to apply it, countries would have to
negotiate on the basis of the principle of “general
level of mutually advantageous commitments”

Modes and sectors Should compensation be thought of only in terms of the
same modes and sectors subject to an emergency
safeguards measure?

No, it should be as in the AS where compensation
is sought any place it can be had.

M.f.n. Should it be applied on an MFN basis? Application should not be on MFN basis but
should be undertaken instead in accordance with
negotiations with affected parties

Overall effect of

Compensation would introduce too stringent a condition
for countries to apply safeguards measures. That would be
good to avoid abuses but could prevent the application of
legitimate safeguards as well

Attenuate the impact of compensation by having it
apply only after some time of continued
application of a particular safeguards measure

Special and Differential Treatment

The non-application of emergency safeguards measures by developed countries against
imports originating in developing countries would be the best type of special and
differential treatment that could be had. Developing countries could seek to ensure that
the same type of provisions as in the AS apply in an ESM in services, while reflecting


upon additional and more effective possibilities for S&D. A minimal level of S&D
elements could include the following:


A de minimis clause,
differential percentages,
extension of investigation period
extension of the application of a safeguard measure without re-opening an investigation,
non-application of compensation
non-application of emergency safeguards measures to the least developed countries;
non-application of emergency safeguards measures to all developing countries in mode 4;
technical capacity-building cooperation for trade remedy authorities in developing countries.

Notification, Consultation, Transparency, Surveillance Mechanism

Article 12.2 of the AS should be enough regarding the obligation to specify the content of
notifications. More discipline than that would need to be justified. As in goods,
surveillance should be post-application and not go beyond goods in requiring any pre-
approval (multilateral or otherwise) of the use or content of an emergency safeguards

The main intention of notification, consultation, transparency and surveillance provisions
is to increase the level of discipline in the application of safeguards measures.

Beyond Feasible and Desirable

A solution to the present WPGR negotiations on safeguards may in fact need to transcend
the dichotomy created between the feasible and the desirable. One way to do this may be
to carefully examine – and ultimately rectify – those arguments that have been put forth as
a negotiating strategy. An additional, perhaps more effective way is to look at the real
concerns that negotiators have had in the negotiations. Drawing from elements reviewed
in the previous sections of the work, these main concerns will be addressed and proposals
will be made as to how to deal with them effectively through the provisions of a possible
ESM in the GATS.

Negotiating Arguments

There are a number of false arguments that have been put forward in the WPGR and that
need to be rectified. Most of them have to do with the confusion created by the
introduction of technical difficulties as a means to combat the logic of safeguard
application. Some of the more crucial ones are the following:

• If things are inapplicable, if data are unavailable, if countries cannot prove surges,

injuries or causal links, emergency safeguards measures will not withstand the test
to which they will be subjected at the WTO, and applying countries will pay a price
for that, losing panels, negotiating compensations, etc.

• If applying certain safeguards is bad for a country’s investment policy or overall

economic policy, the applying country will itself pay for that “policy cost”.


Countries should, after all, have the option of considering an emergency safeguards
measure and applying it or not on the basis its own national policy objectives.

• It is imperative for the WTO to supervise, monitor and demand stringent applying

conditions, but not to require more of trade in services than the system has ever
required of trade in goods.

True Concerns

It has become evident in the negotiations that many of the concerns expressed by
negotiators have to do with perceptions and projections as to what an ESM could or could
not do were it to be included in the GATS. The principal concerns are reviewed below.

Fear of Abuse

It could be said that much of the hesitation in accepting the need for an ESM in the GATS
has had to do with the perception that it could be the object of abuse in its application –
that countries, were the recourse available, would necessarily “overuse” it or, worse yet,
use it in improper (political) ways to simply bypass altogether the natural effects of
liberalization.8 Underlying that hesitation, therefore, are some very real concerns relating
to systemic matters, among which:

• The GATS would become “yet” more flexible by allowing, in addition to the

numerous ex ante options such as positive listing in schedules, binding of
restrictions such as economic needs tests without specified criteria for their
application, non-binding of specific modes of supply under scheduled sectors or
sub-sectors, etc., the recourse to a powerful ex post instrument – the ESM;

• The added flexibility could be a fatal blow to the already precarious credibility of

the GATS as a pro-liberalization tool at the multilateral level; this could have an
impact on the level of support certain countries may be able to garner internally for
multilateral trade disciplines in general; perhaps the “wrong” signals would be sent
to the various regional agreements under negotiation around the world;

• With an ESM the GATS could appear not only to do little for liberalization (already

the case in some quarters) but also to be actually moving backwards on it by
allowing countries to remove, albeit temporarily, commitments from their

• There could be a flood of emergency safeguards measures, which would pose a

practical problem; too many resources would need to focus on the application of
ESMs while there is still much else to be done to perfect the GATS.

8 One first clarification is that an ESM mechanism, as mentioned before, would indeed be intended to
address partly a political element – the fact that liberalization needs to be sold internally and that this task
may be facilitated in the presence of a “safety valve” like the possibility of applying safeguard measure.
Addressing a systemic issue of a political nature should not to be confused, however, with applying ESMs
to political cases, which should not be in the interest of any country that aims to have a serious trade policy


All these arguments can be rebuked and refuted at some level. The idea here, however, is
to take them as concerns that influence the acceptance of an ESM and deal with them
effectively by showing how the mechanism would address (and attenuate) these concerns.

In order to avoid abuse in the application of a possible ESM, the conditions should be
created for members to have to “think twice” before applying it. This is not to say that the
mechanism should become prohibitive or that there should only be a downside to its
application. As argued elsewhere in the present work, the notion that countries should
have access to ex post measures is supposed to be good for the system and in that sense
recourse to it should not be negated through too stringent conditions. Clearly, a middle
point needs to be found between the attractiveness and the repellency of a possible ESM.
Members should certainly not feel stimulated to resort to it but alternatively should not
feel prevented from doing so either.

The question is how to achieve a reasonable balance between “flexibilities” and
“rigidities” of a mechanism. On the “flexible” front, what has been suggested here is that
countries should be free to define by themselves a number of crucial concepts such as
domestic industry, modal application and cross-modal application, and be able to apply
emergency safeguards measures to both existing as well as future commitments. On the
“rigid” front, countries would have to comply with a high degree of discipline in terms of
indicators and criteria for the application of the measure, in particular in relation to the
following elements: matters relating to other factors that may be influencing a particular
situation in a sector, reasons as to why one type of measure is being applied over another,
the compliance with a economic interest test, the compliance with a public interest test.
The question of compensation may be the farthest one could go in the direction of
introducing stringent disciplines. This is why it is here suggested that compensation
would apply only after the first or second year of continued application of a safeguards

The advantage of a mechanism endowed with a high level of discipline in terms of
indicators, criteria and tests is that it allows governments to resist internally the
application of unjustified safeguards measures. The mechanism should provide elements
with which governments may build a more reliable internal system so as to deflect
unfounded cases. In addition, that should be important for governments because the cases
that would qualify for application would most likely be cases that would not encounter
many problems at the WTO, sparing them from having to defend the indefensible in WTO
committees or even dispute settlement panels.

The case of Mode 4

Mode 4 is often cited as the easiest “target” for the application of emergency safeguards
measures and, for that matter, potentially the easiest victim of abuse. That impression
comes from the fact that every country already has established and functioning
bureaucracies specialized in applying immigration and employment restrictions, that the
types of applicable measures are easier to fathom than in other modes of supply, that an
annex was necessary to carve out immigration, residency and employment measures from
the scope of the GATS and, finally, that despite all these things members still committed
very little in their schedules of commitments – a trend which has now been observed also
in the current related negotiations in the Doha Round.


The fact is that mode 4 is already widely “safeguarded” both in the body of the GATS via
its Annex on the Movement of Natural Persons as well as in country schedules of specific
commitments where the norm has been the binding of a series of conditions that need to
be met for a limited number of categories of natural persons. It is also a fact that most
countries already apply safeguard-type measures relating to immigration, residency and
employment and that the GATS already permits it through the carve-out of these
measures from the scope of the agreement. In other words, there is not much else that can
be done in the direction of limiting access through mode 4 – at least in the current
situation in the GATS and the Doha negotiations. The grounds for applying safeguards
measures to mode 4 are therefore also limited insofar as there is generally no significant
liberalization, or binding of liberal situations, in mode 4.

What is in fact missing from the current state of affairs is a way to avoid abuses, were
they effectively to occur some time in the future. The way matters stand now, were a
country to need to modify its schedule, it would seek a permanent modification through
Article XXI of the GATS - as opposed to a temporary suspension. In addition, the level of
security and predictability in the commitments already made in schedules is very low
given some of the ambiguities, or lack of clarity, as to what is actually meant by
“temporary” or “permanent” movement of natural persons and the fact that immigration
and employment policies in most countries do not establish such distinctions in practice.
In other words, even where commitments exist, it may not be sufficiently clear what can
be the effect of the immigration, residency and employment carve-outs in the annex; the
annex in fact goes as far as to stipulate that even in the case of temporary movement
countries are allowed to apply measures to “regulate the entry and stay” of natural
persons. It is true that this provision goes on to say that the advantages accruing from a
specific commitment should not be “nullified or impaired” but the relationship between
this provision and the carve-outs has not yet been tested. It should therefore be better for
countries that want to ensure the security and predictability of commitments made and to
be made on mode 4 to strive to fulfil the following objectives:

• Seek to limit changes in commitments in mode 4 to temporary suspensions when

the three primary conditions for the application of an emergency safeguards
measure are met: increase in the supply of mode 4-related services, determination of
injury, establishment of causal link between the increase and the injury;

• Ensure that the application of a safeguard be subjected to as high as possible level

of discipline in terms of indicators, criteria and economic and public interest tests
(as noted above);

• Seek an agreement that any change in scheduled commitments in mode 4 be

subjected to a “necessity test” whereby the member in question be required to
explain why the change should not be temporary and compliant with the provisions
of an ESM.

It would constitute an effective S&D instrument, if safeguards measures would not apply
to developing countries' mode 4 exports.


Comfort Level

The value of having an ESM incorporated into the GATS should not be underestimated in
terms of the comfort level it would represent for members in the new round of market
access negotiations. As has been explored elsewhere in the present work, it is an
undeniable truth that in order to commit more, and more deeply, members often face a
difficult political barrier internally and that the possibility of applying an ex post measure
such as an emergency safeguards measure could be crucial in attenuating that barrier.

Once again, the issue is how to provide for that comfort level without giving rise to other
problems – such as the possibility of abuse reviewed above. It should be noted that
comfort level also has a further interpretation: that members need to be comfortable that
commitments made by other members have security and predictability, particularly those
where there are potential exporting benefits as is the case for many developing countries
with respect to mode 4.

The solution to addressing comfort level from both the “importing” and the “exporting”
side has to do with the following elements:

• Agreeing that an ESM is necessary in order to advance in the liberalization process,

as many countries have internal difficulties in making or deepening commitments in
the absence of a mechanism that allows them to change the commitments ex post;

• Recognizing that the recourse to ex ante instruments available in the GATS can

only get worse in the absence of some ex post assurances such as an ESM;

• Adopting an ESM that can be evoked easily by members that have justifiable

situations and evoked only with difficulty by members that have unjustifiable

• Adopting an ESM whose focus will be on making sure, to the extent possible and

already at the national level, that only justifiable cases become the object of
emergency safeguards measures; it is the contention here that this may be
accomplished with stringent indicators, criteria and tests.

Burden of Proof

In a sense, there has been an inversion of the burden of proof in the current WPGR
negotiations on safeguards in services: the proof one should be seeking is why not to
apply an ESM to services given the interest in promoting further liberalization and
addressing the concern about the comfort level on the part of members. This inversion has
resulted in an even worse situation: those in favour are having to prove that a mechanism
is indeed feasible and, if so, desirable. This state of affairs loses sight of one important
and undeniable truth which requires no proof and should be the basis for the work of the
working party: that there is a political barrier to further liberalization in services in many
countries and that they need further means of appeasing it. This does not require proof and
should be a reality accepted by all. The burden of proof should not be on whether the
system should or not have an ESM that is feasible and desirable; it should rather be on
which ESM is sufficiently feasible and desirable to fit the aims and objectives of the
GATS agreement.


To prove, for example, that there are cases that may need safeguards is partly impossible
(there may always be questioning on aspects of any hypothetical example), partly
irrelevant, and partly senseless. After all, the whole point of an ESM is to address
emergency situations that cannot be anticipated beforehand – otherwise they would not be
an emergency. If the concept adopted is that of “unforeseen developments”, the logic
against conjecturing with hypotheses is even stronger. By definition unforeseen
developments will be unforeseen and may not be easy to detect even hypothetically.

Instead of trying to guess the unknown, the working party would do best to focus on
creating a mechanism that would be flexible enough to allow recourse to it but also
demanding enough to prevent abuse, assure the necessary comfort level, and transfer the
burden of proof of each case onto the evoking members themselves.

Avoiding a Monster

Another important concern in the current WPGR relates to the possible creation of an
overly burdensome ESM – something that is perceived as making the GATS
unnecessarily more complex than it already is. Several delegations insist on the need to
avoid too much prescriptiveness, which in addition to making things more complex could
fail to fulfil the purported objectives of an ESM.

It is here proposed that in order to avoid too many unnecessary complexities and achieve
agreement on a mechanism, the following recommendations should be pursued:

• Keep matters simple where they can be simple and focus on objective disciplines

that ensure that simplicity does not become a license for members to take liberties
and abuse the system;

• As suggested above, allow countries plenty of leeway to define concepts such as
domestic industry or like services and to be able to apply emergency safeguards
measures to both existing as well as future commitments while demanding stringent
indicators, criteria and tests as conditions for the application of a safeguard measure;

Lack of Experience

Another concern relating to an ESM in services is something that resembles the legendary
“chicken and the egg” problem. Some members have said that because there have been no
safeguard-like cases in services yet in the world, there should be no need to think up a
mechanism. Others would argue that the fact that nothing has yet supposedly justified the
application of a safeguard measure does not represent in any way a guarantee that it may
not happen in the future – especially given the complexity of services markets (e.g. the
recent situation in telecommunications) and the rapidity with which problems proliferate
and spread (e.g. the financial crises of the nineties). A parallel concern is that since the
system has not had to deal with any cases that justify emergency measures, it is not
prepared to devise a system that could “adequately” deal with them; however, without
creating a mechanism, albeit it provisional, it may remain unprepared to do so

The way to get out of this predicament would be to devise a mechanism, test it through
application of justifiable cases, set a time-limit for this “testing” phase, re-visit the issue


on the basis of the experience acquired and, possibly, review, revise, improve and add
necessary elements that could perfect it given the objectives of the GATS Agreement. It
should be in the interest of the system as a whole to set in place a mechanism, albeit
incomplete in some respects, that would permit countries to take measures in justifiable
situations, rather than not to do so and witness the perpetuation of bad practices through
bad scheduling, for example, as examined below.

Equivalent Effect

In addition to being flexible in its application in many respects, including the scheduling
of liberalization commitments, there is an important number of loopholes in the GATS
that allow members to avoid committing to the effective opening of their markets. Even
though emergency safeguards measures should have their own raison d’etre, which
relates admittedly to emergency situations that cause injury, the fact is that the flexibility
of the GATS along with its imperfections have rendered possible a peculiar situation: the
“safeguarding” of liberalization commitments through means other than a proper EMS.
The GATS regime applicable to mode 4, as seen above, is fully “safeguarded” in both the
text of the agreement and in member schedules. Perhaps the most important example of
how that has taken place refers to the so-called “economic needs tests” (ENTs) that have
been widely used by members (around 50% of all members have used ENTs in at least
one mode of supply and/or sector9).

ENTs take various forms, some of which have absolutely nothing to do with economic
needs.10 Some references involve a prohibition and others may exempt other trading
partners from its application. Most often, references point to a real test involving the
determination of whether there are economic needs in the domestic market that would
justify, for example, recourse to services or services suppliers from abroad or of foreign
origin. In the case of mode 4, also a highlight in the context of ENTs, most often the main
objective is to protect the local workforce through quantitative limitations. It would be
difficult to argue that an ENT that lays down a limit above which there will be a cut in the
permitted flow of services or services suppliers is very different from a safeguard
measure. In fact, the only difference is that such an ENT constitutes a safeguard without
any discipline as to its application. It constitutes a blank check for competent authorities
to decide whenever they well please, and without having to comply with a bare minimum
of indicators, criteria or tests of a broader-ranging nature, on the plain interruption
(possibly not even the suspension) of any flow of services or services suppliers. In that
sense an ENT is not a safeguard measure: it resembles it but is much worse than a proper
ESM could ever be.

The fact is that there is an “equivalent effect” between the manner in which ENTs have
been widely applied and a possible ESM. For members that have not resorted to ENTs in
their schedules but that continue to be concerned about the political, economic and
development cost of making further commitments there are two possible paths to follow,
both fairly unattractive: (1) wait for the ENT problem to resolve itself through some
agreement by members to clarify their schedules and to make them more transparent, for
example, through establishing specific criteria; or, (2) include their own ENTs in their
schedules when making commitments in the present round of negotiations. In other

9 UNCTAD (1999).
10 There are entries in the schedules that refer to authorizations being denied with a view to various non-
economic objectives such as the protection of areas of particular historic and artistic interest, for example.


words, the option for members that may be concerned with opening further their markets
in the absence of an ESM mechanism is to use, inter alia, vague ENTs as a proxy.

The question here is: is this kind of “solution” the best there is for the GATS system?
Shouldn’t the GATS be concerned with adding transparency, predictability and security of
commitments as opposed to permitting more lack of clarity, unpredictability and
insecurity? It is true that one should not try to resolve one problem by creating another. If
an ESM mechanism only added to the complexity and vagueness of some of GATS’
provisions, then it would indeed compound the problem and not help to resolve it. The
contention here is, however, that a good ESM mechanism could indeed help to move
forward on the issue of ENTs, for example, precisely because it would diminish the
pressure on governments to “reserve” their positions on binding of commitments. In the
presence of an ESM, the need to resort to ENTs would scale down considerably.

One corollary of what has just been exposed is that members that favour neither tighter
disciplines on the use of ENTs nor the inclusion of an ESM in the GATS may be sending
the message that it is better to achieve flexibility through a vague and imbalanced system
than through a solid body of disciplines that limit abuses.


In this section, a brief review of different modalities for the inclusion of an ESM will be

Horizontal Approach

This is the approach most favoured by the developing countries that are in turn most
favourable to the inclusion of an ESM in the GATS. The approach would be equivalent to
the one traditionally reflected in Article XIX of the GATT whereby an article gathers the
full discipline on emergency safeguards measures in one place. Alternatively, it could also
include the possibility of an annex that is an integral part of the framework agreement (in
much the same way as other annexes) – which amounts merely to a question of form and
not of substance.

Nothing in the various proposals included in the present work would prevent an ESM
from being fully embodied in one article (or annex) of the GATS Agreement. In fact, the
intention of most of the proposals was, in addition to pointing to as simple an instrument
as possible, to leave sectoral specificities to be dealt with on a case-by-case basis and on
the basis of the information and argumentation presented by applying governments.
Petitioning industries and applying governments would be responsible for introducing
relevant sectoral specificities in their application of safeguards measures, striving thereby
to prove their case in the most effective way. There would be no need for the GATS to
specify much in the first stages of application of an ESM. In any case, there would hardly
be sufficient time to start considering sectoral specificities at the current stage of the

At this early stage of GATS’ existence, it would be unwise to attempt to complicate
things by looking at sectoral specificities before coming up with an umbrella structure that
could gradually gather these specificities from concrete cases. It could be argued that after
some time, hopefully by the time a first review of the ESM occurred, the application of a


horizontal instrument would have already produced enough sectoral elements to warrant
(or not) the elaboration of some additional sectoral provisions.

A Reference Document

A more flexible approach to including ESM provisions in the GATS might be the
negotiation of an “understanding” or a reference paper on ESM in much the same way as
was done for financial and telecommunication services during the Uruguay Round. At the
time, a group of countries deemed that it was in their interest to seek a higher or broader
level of liberalization commitments through the acceptance of a common document that
countries members of the group would attach, individually, to their schedule of specific

The primary effect of such an approach was to permit some countries to move forward in
the level of liberalization commitments and to leave open the possibility for others to do
the same. The understanding and the reference document were useful and transparent
tools which spelled out the level of obligations relating to, in that case, financial and
telecommunication services. It was therefore more ambitious than some aspects of
liberalization under the GATS but was optional. This could also be applicable to an ESM
commonly agreed by a group of interested countries. The only downside, however, is that
it would amount to something which would be attached to national schedules – in which
case, it would still be subject to bilateral negotiations and depend on the bargaining power
of each individual member proposing it. One of the main differences with the proposal
below on schedule-based ESM has indeed to do with bargaining power: it may be easier
for reluctant members to accept that other members comply with a document attached to
their schedules if these other members mobilize and negotiate the attachment of the
document as a block.

Schedule-based ESM

The idea here is to permit members to include safeguards in the schedules of
commitments – by sector or sub-sector and by modes of supply, on the basis of a set of
specific rules such as the temporary nature of the measure and the need for objective,
identifiable criteria in the formulation of cases. The difficulty with this approach is that
the “right” to resort to safeguards measures will vary from member to member depending
on the specific interests involved and the bargaining power of the applying member vis-à-
vis its trading partners – in other words, the right to a safeguards measure will be a matter
of negotiations – and consequently bargaining leverage – and not an issue of principle.


With appropriate S&D, an ESM is a useful instrument to encourage progressive
liberalization, including through successive bindings of evolving regulatory situations,
while respecting national policy objectives. An ESM could achieve this in two ways: it
would help members to "sell" the logic of GATS liberalization at home; and, it would do
much to improve the overall functioning of the Agreement. It is a fact that many members
would appreciate the value of having an ex post instrument such as an ESM as a means to
convince reluctant sectors at home to bind liberalization commitments.