Electronic copy available at: http://ssrn.com/abstract=1472196
1
The Nature of International Human Rights Obligations
Frédéric Mégret
Assistant-Professor, Faculty of Law, McGill University
Canada Research Chair in the Law of Human Rights and Legal Pluralism
1
The nature of human rights obligations .............................................................................. 1
I.
The “special character” of human rights ..................................................................... 4
II. Reservations ................................................................................................................ 8
III. State Succession .......................................................................................................... 18
IV. Implementation ........................................................................................................... 19
V. Incorporation ................................................................................................................ 22
VI. Limitations .................................................................................................................. 23
VII. “Margin of appreciation”........................................................................................... 25
VIII. Derogations .............................................................................................................. 28
IX. Withdrawal ................................................................................................................. 30
X. Remedies ...................................................................................................................... 34
XI. Enforcement ................................................................................................................ 35
XI. Conclusion .................................................................................................................. 39
What is the fundamental nature of human rights obligations? What does it mean to be
bound by international human rights law? Obligations can be defined in terms of their
duty-holders, and in terms of their beneficiaries. Some of these issues will be examined in
other chapters, but one can already posit that obligations under international human rights
law are typically (although not only) held by states and, most importantly, for the benefit
of individuals. In itself, this simple characteristic already tells us a lot about some of the
ways in which human rights obligations might be specific.
In trying to assess the “nature” of human rights obligations, this chapter will go beyond
assessing duty holders and beneficiaries in an attempt to understand what sort of
obligations are involved. For example, are human rights obligations fundamentally
unilateral or multilateral, reciprocal or non-reciprocal, temporal or eternal, conditional or
unconditional, relative or absolute? The main idea in this chapter is that human rights
obligations are in some ways radically distinct from other types of obligations existing
under general international law.
1
This is an advanced draft of a chapter to be eventually edited in Daniel Moeckli, Sandesh Sivakumaran,
Sangeeta Shah, and David Harris (eds.), International Human Rights Law, Oxford University Press (2010).
For proper citation and a reduced version with edits, please refer to the printed edition.
Electronic copy available at: http://ssrn.com/abstract=1472196
2
Why does this matter and why should general international law be the standard of
comparison? First, international human rights law is perhaps first and foremost a branch
of public international law. When human rights lawyers sought to internationalize the
human rights project in the middle of the 20
th
Century, they had no choice but to draw on
the body of law existing at the most global level, which had emerged several centuries
before it became acceptable to speak of human rights internationally.
2
Only public
international law could bind states “from above” and hopefully change domestic
practices. However, international law as a legal system is far from neutral: it comes with
its own legal assumptions, preferences and logic, some of which may, as will be seen, be
at odds with the very idea and purpose of human rights. It must be remembered that,
classically, international law was a law created by, between and for states. Its cardinal
concept was state sovereignty, and the typical source of obligation state voluntarism (i.e.:
states are only bound to that which they have consented to). International law was, in a
sense, dedicated to ensuring that these pillars would not be undermined. It would be
surprising if this origin did not lead to some tensions with the human rights project, a
project whose goal is very much geared towards redefining sovereignty.
3
The problem can be analyzed as one of tension between substance and form, where
ideally in any given legal system the two should be aligned. Substantively, the projects of
public international law and international human rights law could not be more at odds a
priori. Human rights law is the project of proclaiming and enforcing certain fundamental
guarantees for individuals against the state, and as such is both radical and revolutionary.
Contra this, the strict international protection of sovereignty is in principle largely
incompatible with a strong stance in favour of human rights, in that human rights deal
with issues that are largely domestic and traditionally considered to be part of states’
domaine réservé. Formally, human rights law as it emerged domestically, and
international law also differ very significantly. International law is the slow maturing of
ancient customary practices, often codified by treaty, emphasizing states’ obligations and
marked, aside from a few rules which are absolutely essential for the system to exist, by a
high degree of flexibility. Human rights law is the product of a particular liberal ideology
of rights whose preferred instrument is declarations, typically in rather grand and broad
brush fashion, and whose focus is the individual.
There inevitably arises a tension between the substance of human rights, and the form
they take internationally. International law has been seen historically as neutral about
ends, and essentially only offering states the means for a very basic interaction that
safeguarded their sovereignty by allowing them to coexist. International norms if
anything emerged from consent between equals; they were what might be called ‘inter-
subjective’. The model for them was the contract, i.e.: an obligation based on reciprocity.
One of the implications was that states could subscribe to whatever obligations they
2
See Chapter 1, on the history of international human rights.
3
Of course, in practice the dichotomy between a “public international law” and an “international
human rights law” is not so stark, largely because the latter has over time influenced and
transformed the former. But the paradigms remain at least theoretically, and in practice the
tensions they create are very much visible.
3
freely agreed to, modify the content of their obligations through consensus, and even
under certain conditions cease respecting certain obligations. Another implication was
that international legal instruments only created obligations between states, at the expense
of individuals who were considered mere “objects” of international law, at best
unintended and collateral beneficiaries of obligations undertaken by states. Although
international law diversified in due course, regulating ever more areas of international
life, it did and largely still does claim to lay the ground rules for all thematic “branches”
of international law. For example, the Vienna Convention on the Law of Treaties is not
interested in the particular content of treaties and whether they deal with trade, the
environment, or diplomatic immunities: the same basic rules apply to all.
The problem, of course, is that form – the particular expression international
commitments take legally - is not entirely neutral substantively, and this can create
substantial tensions when the project does not quite fit the form it is asked to take. The
point has been made eloquently by Mathew Craven in the context of human rights
treaties:
In human rights treaties the two elements of “form” and “function” [note:
substance] appear to be fundamentally at odds with one another. On the
one hand, the form in which the rights are expressed – the treaty –
supposes that human rights are merely the incidental subject of a
contractual bargain between states. Individual ‘right-holders’ are therefore
simply the fortuitous beneficiaries of a regime that is otherwise concerned
with promoting the rights and interests of states. The teleology of the
regime on the other hand – focused as it is on individual or group ‘human
rights’ – supposes that the treaties are quasi-constitutional in character.
4
Historically, this has created a very significant conceptual and practical challenge for
human rights lawyers: how could one avoid the human rights project being stultified and
even absorbed by the language of traditional international law? The answer has often lied
in redefining human rights obligations, particularly treaty obligations, as special in
nature, i.e. as dependent on, and yet also crucially distinct from, the way in which
international legal obligations typically operate. But this idea that human rights
obligations are somehow “special” needs to be grounded and explained, to understand
exactly why this is so, and what the implications are.
In this chapter I will introduce the basic idea of what it means for a legal obligation to be
described as “special” in nature in international law (I). I will then identify ten
consequences that can be said to flow from this character in terms of reservations (II),
state succession (III), implementation (IV), incorporation (V), limitations (VI), “margin
of appreciation” (VII), derogations (VIII), withdrawal (IX), and enforcement (X).
4
Matthew Craven, Legal Differentiation and the Concept of the Human Rights Treaty in
International Law, (2000) 11 Eur. J. Int’L 489.
4
I. The “special character” of human rights
Defining in general terms what defines the “special character” of human rights
obligations is a complex exercise. There are both intrinsic and extrinsic specificities,
which are intimately linked. Intrinsically, the special nature of human rights as a legal
obligation has to do with the nature of human rights as a transformative project that aims
to set out minimum standards applicable to individuals as against the powers that be
(typically, the state). This is done through the identification of “rights,” which are very
peculiar legal artefacts: not so much rules or even goals, as ways of thinking about the
justice of the relationship between the state and the individual, and even society and
communities. These “ways of thinking” prioritize the individual and his/her liberties over
collective or communal interests, and have a strong deontological, non-instrumental
focus. This does not prevent human rights, perhaps paradoxically, from being quite
dynamic and evolutionary in nature. Moreover, human rights obligations typically have a
strong axiological character, in that they are considered to be particularly highly placed in
relation to other norms, or otherwise would lose much of their specificity.
5
Although
human rights can undeniably benefit from being codified into positive law and this has
been their defenders’ ambition, there is a clear pre-positive to rights that suggests that
they exist even when they are denied under positive law. Finally, human rights are
unilateral obligations owed by the state (or the powers that be) to individuals, and which
do not impose reciprocal obligations on individuals (e.g.: the obligation to respect the
rights of states) that would restrict the enjoyment of their rights.
6
These intrinsic specificities, then, are carried over and imported into the international
realm where, in “becoming international”, human rights help create new (and renovate
old) international legal forms. It is useful in this respect to think in terms of the evolution
of international treaty law, given that so much of contemporary international human
rights law is now contained in treaties. In the 20
th
Century, treaties in general have
evolved substantially away from a pure contractual model, to one where treaties are
increasingly seen as “law making,” i.e.: creating obligations of a more general and
imperative nature. Instruments creating international organizations, such as the Charter of
the United Nations, are a typical example of this. Arguably, no branch of international
law has been more at the forefront of these developments than human rights law.
5
This is true both domestically and internationally, where human rights typically migrate towards
constitutional heights, even though in practice this is never an easy process and, particularly when
it comes to incorporating international obligations domestically may be resisted quite effectively.
6
There is of course a repeated tendency for states to consider that “criminals” or “terrorists” for
example have “forfeited” certain rights by their behaviour, but this is precisely where human
rights is insistent that all rights should be respected. Moreover, both domestically and
internationally, there has sometimes been a temptation to tie “human rights” and “human duties”.
This is particularly the case in authoritarian regimes. See in particular the African convention on
human rights, African Charter on Human and Peoples’ Rights (adopted 27 June 1981, entered
into force 21 October 1986) (1982) 21 ILM 58 (African Charter), and the Declaration on the
Rights and Duties of Man.(1789) France, Constitution of the Year III (1795). Although there is
nothing inherently wrong about proclaiming duties, it is the element of tying these duties to rights
that is perilous.
5
Of course, at a certain level international human rights law is hardly a rejection of all the
canons of public international law. As has been seen in the chapter on sources, for
example, there is no reason to think that the list of existing sources of that “branch” that
is international human rights law is entirely different from the “trunk” of public
international law (although some adaptations do occur, see chapter 4). International
human rights treaties borrow much from the traditional law of treaties, including the way
treaties come into existence, the simple idea that states are bound by treaties they ratify
(the pacta sunt servanda rule), or that they should discharge their obligations in good
faith. Many of the key aspects of state responsibility, moreover, are applicable in the field
of human rights. There is no reason to think that any of these rules apply in markedly
different ways simply because the subject matter of a treaty is human rights.
However, there are also clearly some ways in which human rights obligations are
different, even radically so. A traditional treaty is one that creates rights and obligations
between states, to the broad exclusion of any other actor, even if it may objectively
benefit from such arrangements. States typically exchange such rights and obligations,
with a view to securing an advantage commensurate with their costs. Treaties are said to
be synallagmatic, in that there is a quid pro quo involved. Moreover treaties are often
quite strictly relational and involve a strong element of intuitu personae (one only enters
into treaties with the states that one wants to enter into such treaties with). Finally, the
substance of classical treaties is quite indifferent, and international law does not see itself
as having a significant stake in what states commit themselves to (as opposed to the fact
that they should honour whatever commitments they make).
Human rights treaties superficially borrow the idea of an agreement between states as to
certain obligations. But they are also strikingly different. They are an agreement between
states whose beneficiaries are in fact a third-party, namely “populations within their
jurisdiction.” Indeed, the treaties create rights specifically for individuals who are not
normally subjects of international law generally or treaties specifically. To describe
human rights treaties as contractual, moreover, would be profoundly misleading, as it is
wrong to say that states commit to respecting certain human rights obligations “in
exchange” for other states doing so.
7
Finally, human rights lawyers certainly think that
human rights treaties have a substance that is inherently of a high normative worth, above
and beyond states’ consent to be bound by them.
All of these differences are very important, because they have an impact on the way
international human rights obligations are understood, administered and interpreted. In
order to capture them, human rights lawyers have forged the very unique and defining
7
Politically and in practice, joining a human rights treaty may be based on an expectation that
other states will do the same, and of course such an expectation is not negligible and also has
specific consequences, as will be seen. However, the crucial point is that from a normative point
of view, international human rights obligations are not at all as dependent on this idea of
reciprocity as most ordinary treaties, where the interest of incurring an obligation will very
quickly lapse if the other party or parties do not comply with theirs.
6
notion of the “special character” of human rights obligations.
8
At first, the meaning of the
idea of human rights’ “special character” was a little nebulous - one of these catch
phrases that merely seemed to imply that ordinary rules of international law do not apply
when it comes to human rights. However, there is no doubt that human rights bodies have
refined it over the years to reflect what is distinctive about human rights.
The first reference to the special character of a human rights treaty occurred when the
General Assembly sought an advisory opinion from the International Court of Justice on
reservations to the Genocide Convention. We will return to the practical questions raised
by the advisory opinion, but the important thing is that the Court distinguished between
“ordinary” treaties, and those of a humanitarian or human rights character. In the latter:
… the contracting States do not have any interests of their own; they
merely have, one and all, a common interest, namely, the accomplishment
of those high purposes which are the raison d'être of the convention.
Consequently, in a convention of this type one cannot speak of individual
advantages or disadvantages to States, or of the maintenance of a perfect
contractual balance between rights and duties.
9
The idea was thus very much that the Genocide Convention created something greater
than the sum of each party’s will. Moreover, there was a certain vertical element
involved: states are not so much bargaining between each other horizontally, as they are
solemnly committing themselves to a certain supranational standard of behaviour. Indeed,
there is almost something a bit circular about states becoming party to such treaties, in
that they are making a unilateral commitment to a multilateral goal not to engage (in the
case of the Genocide convention) in certain conduct.
A very similar idea was soon taken up by the European Commission of Human Rights in
1961, when it affirmed that:
The obligations undertaken by the High Contracting Parties in the
Convention are essentially of an objective character, being designed rather
to protect the fundamental rights of individual human beings from
infringements by any of the High Contracting Parties than to create
subjective and reciprocal rights for the High Contracting Parties
themselves.
10
The idea of the “objective character” of human rights obligations is a very characteristic
one under the European scheme. Objectivity here does not mean that the obligations in
question are absolute or that they are scientifically proven. Rather, it should be
understood in the legal meaning of “not strictly dependent on inter-subjective
8
This special character is most frequently adduced in the context of treaty obligations, but one
could make the case that it also has an impact on how obligations emerging from other sources
(e.g. custom) should be understood.
9
Reservations to the Genocide Convention (Advisory Opinion) 1951 ICJ Reports 15 at ¶23
10
Austria v Italy (Pfunders Case) (App 788/60) (1961) 4 Yearbook 116 (EComHR), ¶138.
7
commitments.” Again, the idea is that, despite appearances, there is something relatively
unilateral about states’ adherence to a treaty like the ECHR, even though some of the
traditional benefits of treaties – i.e. getting other states to behave in a particular manner –
retain a residual role.
11
States are better understood as either making a solemn “promise
to self”, to the “international community” or indeed to individuals within their
jurisdiction.
The Inter-American Court then followed suit, adding its own spin to the idea of the
special character. As the Court put it in one of its advisory opinions:
… modern human rights treaties in general, and the American Convention
in particular, are not multilateral treaties of the traditional type concluded
to accomplish the reciprocal exchange of rights for the mutual benefit of
the contracting States. Their object and purpose is the protection of the
basic rights of individual human beings irrespective of their nationality,
both against the State of their nationality and all other contracting States.
In concluding these human rights treaties, the States can be deemed to
submit themselves to a legal order within which they, for the common
good, assume various obligations, not in relation to other States, but
towards all individuals within their jurisdiction.
12
(emphasis added)
At a certain level, the “special character” of human rights suggests that even though
states may become parties to such obligations as a result of a very classical manifestation
of consent to be bound to other states, their ensuing human rights obligations are in some
ways independent of that consent. Human rights obligations have “a life of their own”
that takes over as soon as states have manifested their original commitment to be bound.
This also ties up interestingly with old claims about human rights pre-existing their legal
recognition, so that states are merely solemnly committing to something which they were
at least always morally or philosophically obliged to recognize. Another way of putting
it, using an old dichotomy in international law, would be to say that becoming a party to
human rights treaty is “declarative” of states’ obligations rather than “constitutive” of
them.
Subsequently, the ECHR has been described by the Court as a “constitutional instrument
of European public order.”
13
The reference to this constitutional character also underlines
11
For example, states may be aware that by joining certain international human rights treaties
they may contribute to “externalize” certain domestic preferences, “lead by example”, and
generally live in a broader area where principles they are committed to domestically will also be
honoured by “partner” states. Or they may think that having other states commit to these
standards, will inspire them to do so, or at least guard them against the temptation of reneging on
their commitments. None of these rationales for joining treaties should be overlooked, but they
are not primarily what is at stake when states join human rights treaties.
12
The Effect of Reservations on the Entry into Force of the American Convention (Arts 74 and
75), Advisory Opinion OC-2/82, Inter-American Court of Human Rights, Series A, no.2 (24
September 1982) at ¶29-30.
13
Loizidou v. Turkey (Preliminary Objections) (1995) E.C.H.R Series A no. 310, 27 at ¶75.
8
the extent to which international human rights treaties, perhaps especially regional ones,
can be seen as part of governance structures and, beyond that, as one of the normative
foundations of political community (in the case of the ECHR, that of the Council of
Europe). International human rights thus appear as the cement that binds groups of states
together in a collective project that is both domestic, transnational and supranational.
We will see other formulations of the “special character” in specific contexts where it has
been used. The notion might at first seem like merely a doctrinal construct, something of
more theoretical than practical relevance. What is remarkable, however, is the extent to
which concrete consequences have flown from this idea, some of which have clearly had
a considerable impact on the enjoyment of rights.
II. Reservations
Reservations are a particularly technical area of international law, but the study of this
rather dry subject in the context of international human rights law is enlightening. A
reservation is defined by Article 2 of the Vienna Convention as a “unilateral statement,
however phrased or named, made by a State, when signing, ratifying, accepting,
approving or acceding to a Treaty, whereby it purports to exclude or to modify the legal
effect of certain provisions of the Treaty in their application to that State”. The goal of
reservations is therefore clear: to limit states’ obligations under a particular treaty by
entering a sort of caveat. International law is not formalistic about what constitutes a
reservation: it does not matter whether a state presents it as a “general political statement”
or a “declaration of interpretation” - if it excludes or modifies the legal effect of certain
provisions, it will be considered a reservation.
Under classical international law, reservations would have been quite broadly tolerated.
This is for the reason that international law was not traditionally very invested in the
substance of treaties, so that it might as well allow for as much flexibility as possible. In a
sense, everything in a normal treaty is “up for grabs” so that any number of reservations
could theoretically be made. Of course, even under classical international law,
reservations created problems. In a voluntary framework, reservations have to be
accepted by other state parties to become applicable, or even for the treaty to enter into
force. But, in line with international law’s overall pragmatic attitude, it mattered less that
states might agree to a very different set of obligations, than that states were able to fine
tune the degree of their commitment.
In the international human rights context, states have availed themselves broadly of the
possibility of reservations, both quantitatively and qualitatively. For example, as of 8
August 2002, 37 of the 153 States parties to the International Covenant on Civil and
Political Rights had, between them, entered 173 reservations.
14
A typical reservation is
one whereby a state purports to interpret a particular internationally protected right only
14
ECOSOC (Sub-Committee on the Promotion and Protection of Human Rights),Working Paper
by Francoise Hampson, ‘Other Specific Human Rights Issues, Reservations to Human Rights
Treaties’ (8 Aug 2002) E/CN.4/Sub.2/2002/34.
9
in accordance with its domestic law, particularly constitutional or religious law. The US
has particularly illustrated itself by suggesting that its international human rights
commitments should be read in the light of constitutionally protected rights.
15
A typical
and problematic reservation from Muslim states, for example to the Convention on
Elimination of Discrimination Against Women has been that a state accedes to it “subject
to the general reservation that such accession cannot conflict with the laws on personal
status derived from Islamic Sharia”.
16
There are of course valid reasons why one might want to make reservations specifically
in the human rights context. A state might be generally committed to a human rights
treaty but have problems with one particular article and therefore want to phase in its
commitment. One might also want to inject a certain cultural understanding of a
particular obligation; or to guard against expansive interpretations by the implementing
body by preempting a particular article’s meaning. Or states might know in advance that
part of their domestic law is incompatible with an international human rights
commitment, yet nonetheless want to ratify, something that human rights lawyers will
want to encourage (in the sense that if states were to wait until their law was fully in
conformity with human rights obligations, it could be a long time before they ratified).
However, there is no doubt that the practice of reservations has become problematic for
international human rights obligations, even more so than they were in general
international law. Of course, one might consider that it was better than nothing and
reservations should be permitted to allow that state to at least become bound by some
obligations. But human rights are clearly more of a “package” than most international
normative instruments, and there is a very real tension between the idea of human rights
and the idea that one might, through reservations, pick and chose the obligations one
wants to be bound by. The very unity of human rights can be threatened, if they are made
to appear as little more than political preferences, a menu from which selected items can
be chosen. At a certain level, it is the very universality of human rights and the whole
point of international treaties which is threatened if states do not agree to honor the same
rights. Moreover, there is a sense that some states might “free ride” human rights treaties,
by obtaining the benefits broadly associated with being a party to them, whilst only
subscribing to an eviscerated set of obligations. Finally, reservations also create problems
of legal certainty, making it difficult for the intended beneficiaries (individuals) to
ascertain the exact scope of the rights they have been guaranteed.
Accordingly, human rights supervisory bodies have frequently deplored the abundance
and scope of reservations, and urged states to repeal them. This is in itself a first
departure from ordinary international treaties, where few bodies are authorized or would
see it fit to deplore something that is largely seen as falling within states’ prerogatives.
15
US on ICCPR: (3) That the United States considers itself bound by article 7 to the extent that
`cruel, inhuman or degrading treatment or punishment' means the cruel and unusual treatment or
punishment prohibited by the Fifth, Eighth, and-or Fourteenth Amendments to the Constitution of
the United States. See U.S. reservations, declarations, and understandings, International Covenant
on Civil and Political Rights, 138 Cong. Rec. S4781-01 (daily ed., April 2, 1992),
<http://www1.umn.edu/humanrts/usdocs/civilres.html>.
16
Such is the wording of the Libyan reservation to CEDAW.
10
International human rights law has also evolved specific notions of what reservations are
permissible (A), who may decide (B), and what consequences flow from reservations (C).
A. Permissible reservations
The general regime under the Vienna Convention is that states may formulate
reservations to a treaty especially when a treaty prescribes but not when a reservation is
prohibited by the treaty or, more problematically, when a reservation would be
incompatible with the object and purpose of the treaty.
17
Although Article 19 is presented
as authoritative,
18
and to provide “relevant guidance”,
19
and some states have insisted that
the general regime is sufficient,
20
human rights bodies have nonetheless developed quite
a specific approach to permissible reservations. This is particularly the case with the
ECHR which, despite paying lip service to the Vienna Convention, has developed an
original policy towards reservations based on the defence of human rights’ “objective”
character. Most human rights treaties neither allow
21
nor specifically exclude
reservations, so that most of the discussion has focused on whether certain reservations
could be considered compatible with the “object and purpose” of treaties. However, the
object and purpose has typically been understood more broadly than it would be for
ordinary treaties, so that almost all reservations except the most marginal have been held
incompatible with human rights obligations.
17
. Vienna Convention on the Law of Treaties, (adopted 27 Jan. 1980) 1155 U.N.T.S. 331 art 19.
18
American Declaration of the Rights and Duties of Man, OAS Res XXX adopted by the Ninth
International Conference of American States (1948) reprinted in Basic Documents Pertaining to
Human Rights in the Inter-American System OEA/Ser L V/II.82 Doc 6 Rev 1 at 25 (1992)
19
UNCHR, ‘General Comment 24 (52) on issues relating to reservations made upon ratification
or accession to the Covenant or the Optional Protocols thereto, or in relation to declarations under
article 41 of the Covenant’ (1994) UN Doc CCPR/C/21/Rev.1/Add.6 at ¶6 [GC 24].
20
See, for example, the UK declaration:
The United Kingdom does not (…) believe that rules different from those foreshadowed by the
International Court and in due course embodied in the Vienna Convention on the Law of Treaties
are required to enable the international community to cope with reservations to human rights
treaties. The correct approach is rather to apply the general rules relating to reservations laid
down in the Vienna Convention in a manner which takes full account of the particular
characteristics of the treaty in question. See Observations on General Comment No.24, 3 IHHR
(1996) 261, at ¶4.
21
There are a few exceptions. See Article 2(1) Second Optional Protocol to the International
Covenant on Civil and Political Rights, aiming at the abolition of the death penalty (adopted July
11, 1991) G.A. res. 44/128, annex, 44 U.N. GAOR Supp. (No. 49) at 207, U.N. Doc. A/44/49
(1989). Also Convention for the Protection of Human Rights and Fundamental Freedoms (entered
into force Sept. 3) 1953 213 U.N.T.S. 222, as amended by Protocols Nos 3, 5, 8, and 11 which
entered into force on 21 September 1970, 20 December 1971, 1 January 1990, and 1 November
1998, respectively, art 57.
11
One of the most important sources on the issue is the Human Rights Committee’s
General Comment n°24,
22
whose existence is in itself an indication of how seriously the
issue is taken by the UN’s supervisory bodies. According to the Committee:
In an instrument which articulates very many civil and political rights,
each of the many articles, and indeed their interplay, secures the objectives
of the Covenant. The object and purpose of the Covenant is to create
legally binding standards for human rights by defining certain civil and
political rights and placing them in a framework of obligations which are
legally binding for those States which ratify; and to provide an efficacious
supervisory machinery for the obligations undertaken.
23
(emphasis added)
As can be seen, the object and purpose is defined very broadly (creating legally binding
standards for human rights). Moreover, the sheer inter-dependence of rights suggests that
it will be difficult to make reservations to one right without altering the entire content of
the obligations contained in a treaty. The Committee has also specifically excluded a
number of reservations which it considers would be ipso facto incompatible with the
“object and purpose” of human rights treaties, such as those violating peremptory norms
(e.g.one could not make a reservation to the prohibition of slavery), or even those that
constitute customary international law.
24
And even though the Committee allows for the
possibility of reservations to such a relatively technical and detailed provision as Article
14 (fair trial guarantees), it also insists that “a general reservation to the right to a fair trial
would not be (acceptable)”.
25
Moreover, reservations for non-derogable rights are
excluded, which is highly understandable given the fact that human rights treaties already
provide some in-built flexibility (highlighting some rights as being derogable) in
situations of national emergency. It would defeat the purpose of that system if states were
allowed to create added space for manoeuvre through reservations. The same obviously
holds true for reservations that might relate to the definition of a situation of emergency,
something which the international community will be particularly insistent should be
defined in consistent ways.
Another interesting limit to reservations, according to the HRC, is that states may not
make reservations relating to “basic guarantees” associated with the rights protected, such
as the right to an “effective remedy” in case of violations. Defined as “supportive
guarantees”, these provisions are seen to provide “the necessary framework for securing
the rights in the Covenant and are thus essential to its object and purpose.”
26
In fact, they
“are an integral part of the structure of the Covenant and underpin its efficacy” so that to
22
General Comments are authoritative pronouncements by supervisory bodies on the treaties they
monitor. See Chapter 17 on UN organs.
23
GC 24, n19 at ¶7 .
24
The idea presumably is that since states are bound to honour such rights regardless of treaty
commitment (and since one cannot make a reservation to a customary obligation), then states
should not be allowed through treaty to derogate from such an obligation.
25
GC 24, n19 at ¶8.
26
Ibid, ¶11.
12
limit them would defeat the purpose of the ICCPR.
27
Finally, it almost goes without
saying that states could not make reservations to institutional arrangements designed to
supervise the implementation of treaties. For example, a state could not become a party to
one of the main treaties and make a reservation to the effect that it does not consider itself
bound by the obligation to submit periodic reports to its supervisory body.
This leaves very few areas where reservations might be valid. In fact, even if there are
such areas, the Committee adds a number of formal and substantive requirements to the
sort of reservation that can be made:
Reservations must be specific and transparent, so that the Committee,
those under the jurisdiction of the reserving State and other States parties
may be clear as to what obligations of human rights compliance have or
have not been undertaken. Reservations may thus not be general, but must
refer to a particular provision of the Covenant and indicate in precise terms
its scope in relation thereto. When considering the compatibility of
possible reservations with the object and purpose of the Covenant, States
should also take into consideration the overall effect of a group of
reservations, as well as the effect of each reservation on the integrity of the
Covenant, which remains an essential consideration. States should not
enter so many reservations that they are in effect accepting a limited
number of human rights obligations, and not the Covenant as such. So that
reservations do not lead to a perpetual non-attainment of international
human rights standards, reservations should not systematically reduce the
obligations undertaken only to the presently existing in less demanding
standards of domestic law. Nor should interpretative declarations or
reservations seek to remove an autonomous meaning to Covenant
obligations, by pronouncing them to be identical, or to be accepted only
insofar as they are identical, with existing provisions of domestic law.
States should not seek through reservations or interpretative declarations
to determine that the meaning of a provision of the Covenant is the same
as that given by an organ of any other international treaty body.
28
In other words, what is striking about the Human Rights Committee’s view of
reservations is that their usage is heavily constrained in advance by the numerous rules
that formulate the object and purpose of the Covenant. These rules, in addition, promote
an important integrated and holistic vision of the object and purpose of the Covenant
which, as a result, all but eliminates most reservations
29
27
Ibid.
28
Ibid, ¶19
29
The Committee does give a few examples of reservations that it says it would consider valid,
such as a reservation saying that violations can only be heard by the Committee after the entry
into force of the ICCPR for the state party, or that communications cannot be examined if they
are already being examined by a comparable international procedure. As can be seen, these
reservations are remarkably procedural rather than substantive in nature.
13
The ECtHR has by and large taken the same position through such leading cases as
Belilos v. Switzerland, where it found that reservations of a “general character,” which
are prohibited by Article 57 of the Convention, are “too vague or broad for it to be
possible to determine their exact meaning and scope”.
30
As a result, the Court struck
down Switzerland’s use of an interpretative declaration which attempted to limit
Switzerland’s obligations under Article 6(1) of the ECHR.
31
Perhaps even more
strikingly, the ECHR has occasionally considered certain reservations invalid almost
exclusively by invoking the Convention’s “special” nature. The leading case on this is
Loizidou v. Turkey, which involved the complaint of a Greek Cypriot women who had
lost property following the occupation of Northern Cyprus by Turkey and alleged an
interference with her property rights under Article 1 of Protocol 1 of ECHR. Turkey had
recognized the Court’s compulsory jurisdiction under Article 46 of the ECHR, but had
added a reservation to the effect that its recognition only applied to the territory of
metropolitan Turkey (i.e.not Cyprus). The Court rejected that reservation as invalid
insisting that it
… must bear in mind the special character of the Convention as an
instrument of European public order (ordre public) for the protection of
individual human beings and its mission, as set out in Article 19 (art. 19),
"to ensure the observance of the engagements undertaken by the High
Contracting Parties".
32
In effect, a reservation limiting the territorial reach of the recognition of the Court’s
jurisdiction was incompatible with the character of the European convention as a human
rights treaty. In particular, the Court stated that it would amount to excluding certain
populations from the benefit of the Convention (in itself something that seems
discriminatory) and relieving a state of some of its obligations. A state could not, from a
human rights perspective, only partly recognize the Court’s jurisdiction. The Court also
pointed out that the resulting “inequality between states” contradicted the Preamble’s
goal of “achieving greater unity in the maintenance and further realization of human
rights”. The collective guarantee that the Convention represents would, in other words, be
weakened by excessively particularistic recognitions of jurisdiction. Revealingly, the
Court rejected an analogy that Turkey had sought to make with the ICJ, pointing out that
this was a context quite distinct from that of human rights.
33
B. Responsibility for assessing reservations
Another aspect of the specificity of the regime of reservations to international human
rights treaties relates to who can raise the invalidity of a reservation. In a typical
international law scenario, the invalidity of reservations will be raised by other state
30
Belilos v. Switzerland, (1988) Series A no. 132, p. 28.
31
Ibid.
32
Loizidou v. Turkey (merits) (App no 40/1993/435/514)
ECHR 18 Dec 1996
33
Ibid, ¶84-85.
14
parties. This will be all the more likely to occur since treaties consist in the exchange of
promises between sovereigns, and thus states will be wary of other states committing
themselves to fewer obligations than they are already bound by. Indeed, state parties may
refuse to ratify a treaty if they fundamentally disagree with reservations made by other
states because they think that there is no real agreement. Moreover, under general
international law, a state is free to object to any reservation, even a perfectly valid one.
However, one can see how the dependence of this regime on states to object to
reservations is problematic from the point of view of human rights. For a start, states in
the human rights context are perhaps less likely to object to other states’ reservations, for
the simple reason that they are less directly impacted by these reservations.
34
States have
of course occasionally manifested their rejection of other states’ reservations, but this is a
haphazard process. In fact, the HRC considers that failure by other state parties to object
to a reservation is not conclusive as to its validity. As the Committee pointed out:
The absence of protest by States cannot imply that a reservation is either
compatible or incompatible with the object and purpose of the Covenant.
Objections have been occasional, made by some States but not others, and
on grounds not always specified; when an objection is made, it often does
not specify a legal consequence, or sometimes even indicates that the
objecting party nonetheless does not regard the Covenant as not in effect
as between the parties concerned. In short, the pattern is so unclear that it
is not safe to assume that a non-objecting State thinks that a particular
reservation is acceptable.
35
In fact, objections made by state parties are not conclusive of the invalidity of a
reservation, and states should not be able to object to valid reservations, thus potentially
delaying the entry into force of treaties.
These key differences have led the HRC to opine that “the operation of the classic rules
on reservations is inadequate for the Covenant.”
36
Consequently, a new rule has emerged
in the human rights context, whereby a purely consensual approach to reservations
(reservations are accepted or not) has been replaced by an increasingly axiological one
(reservations are valid or not). In other words, state parties, when pronouncing on
reservations made by other states are not asked to say whether they “agree” with them,
but to judge their validity, and object to them only if they find them invalid, for example
because they are incompatible with the object and purpose of the treaty. In other words,
states are asked to act as the guardians of the integrity of human rights treaties, rather
than simply the keepers of their own national interest.
Moreover, the limitations of a largely decentralized system of reservation manifestation
have led to an increasingly strong supra-national control by human rights bodies
34
Although of course, the whole point is that they are impacted a little since the collective
guarantee in which they thought they were investing has been cheapened.
35
GC 24, n19 at ¶17.
36
Ibid, ¶17.
15
themselves. Human rights bodies, whether jurisdictions or supervisory organs, have
keenly seen themselves as the defenders of the ordre public of international human rights
law and raised issues of invalid reservations even before they created actual disputes.
According to the Human Rights Committee, which has been at the forefront of
scrutinizing reservations:
It necessarily falls to the Committee to determine whether a specific
reservation is compatible with the object and purpose of the Covenant.
This is in part because (…) it is an inappropriate task for States parties in
relation to human rights treaties, and in part because it is a task that the
Committee cannot avoid in the performance of its functions. (…)) Because
of the special character of a human rights treaty, the compatibility of a
reservation with the object and purpose of the Covenant must be
established objectively, by reference to legal principles, and the
Committee is particularly well placed to perform this task.
37
(added
emphasis)
In other words, the particular interest that the international community has in the integrity
of human rights treaties necessitates that their legitimacy be maintained by a strong
system of international supervision.
C. Consequences
The consequences of reservations is a complex issue, especially if some of them turn out
to be invalid. This matter raises different problems for the reserving state, third states,
and the entire treaty. There is a fundamental tension at stake between, on the one hand,
wanting as many states to become bound by a treaty thereby allowing the treaty to enter
into force as quickly as possible, and, on the other, the risk of a treaty’s integrity being
compromised by too many reservations.
The problem arises first at the level of the entire treaty. The old, ultra-consensualist
solution, as once promoted by the League of Nations, was that a treaty and reservation
would only become valid if all states accepted the reservation. Needless to say that if
such a rule were still applicable, none of the principal international human rights
instruments would ever have come into force. Moreover, the whole idea that a treaty only
comes into force once all states have accepted all reservations portrays treaties as entirely
consensualist and purports that full agreement on every reservation is the essence of what
makes a treaty binding. This view, however, is quite remote from what international
human rights treaties in fact are. Clearly, state parties’ overall interest in human rights
treaties coming into force for themselves (and in a sense for their own benefit) outweighs
whatever qualms (even genuine ones) they may have about other states’ reservations.
Even in general international law, this rule has now been transcended by the Vienna
Convention which considers that only certain treaties require a reservation to be accepted
by all parties, namely those where it “appears from the limited number of the negotiating
37
Ibid, ¶18.
16
States and the object and purpose of a treaty that the application of the treaty in its
entirety between all the parties is an essential condition of the consent of each one to be
bound by the treaty”.
38
This would be typical of treaties between very few parties, but
international human rights treaties are the exact opposite, since their vocation is to
become universal. There is thus no question of reservations having to be accepted by all
states for a human rights treaty to enter into force, which explains why so many human
rights treaties have in fact entered into force despite considerable numbers of
reservations.
The second question is what happens to reserving states. The old rule was that they only
became party to a treaty once all parties had accepted their reservation. Under general
international law, the position has since evolved to consider that the reserving state
becomes bound if at least one other state accepts the reservation.
39
However, in the
context of treaties of a humanitarian or human rights character, the ICJ has found that
such rules miss the point, and that states can in fact become party to a treaty even if other
states object to its reservation:
The object and purpose of the Genocide Convention imply that it was the
intention of the General Assembly and of the States which adopted it that
as many States as possible should participate. The complete exclusion
from the Convention of one or more States would not only restrict the
scope of its application but would detract from the authority of the moral
and humanitarian principles which are its basis. It is inconceivable that the
contracting parties readily contemplated that an objection to a minor
reservation should produce such a result.
40
Not only is there an interest in states being bound as soon as possible, notwithstanding
minor incidents of consensualism, but the specific nature of human rights treaties
suggests that approval by other parties should really be seen as a secondary matter. This
was further clarified by the Inter-American Court. The Inter-American Commission had
asked the Court an advisory opinion on whether the Convention would enter into force as
soon as the requisite number of states had ratified it, or whether ratifications accompanied
by reservations should only be considered as complete when at least one state had
accepted them
41
(or, as the Vienna Convention stipulates, all state parties had failed to
object to them within a year of the ratification).
42
The Court found that because of the
special character of the Convention as “a multilateral legal instrument or framework
enabling States to make binding unilateral commitments not to violate the human rights
of individuals within their jurisdiction”, the Convention would enter into force regardless
38
United Nations, Vienna Convention on the Law of Treaties, 23 May 1969, United Nations,
Treaty Series, vol. 1155, p. 331 art. 20(2) [Vienna Convention]
39
Ibid, art.20(4)(c).
40
Reservations to the Genocide Convention (Advisory Opinion) n9 at ¶24.
41
The Effect of Reservations on the Entry Into Force of the American Convention on Human
Rights (Arts. 74 and 75), Advisory Opinion OC-2/82, September 24, 1982, Inter-Am. Ct. H.R.
(Ser. A) No. 2 (1982).
42
Vienna Convention, n38, art 20(4)(c).
17
of whether any other state accepted reservations appended to an instrument of ratification
(“no useful purpose would be served by such a delay”).
43
In other words, states became
bound the moment they deposited their instrument of ratification, notwithstanding the
fact that their reservations were not accepted and might even be invalid. This again
underlined the strictly non-relational character of human rights treaties.
One of the problems created by this emphasis put on human rights treaties entering into
force quickly and broadly is that the actual validity of reservations will only be examined
at a later date. This can create complex issues of retroactive application when a
reservation turns out to have been invalid all along. A state that ratified a treaty with a
reservation may insist that it ratified the treaty only on the basis of such a reservation, and
that if the latter is then found to be invalid it is either not bound by the provision to which
the reservation related, or even not bound by the treaty at all. The opposite stance is that
invalid reservations are “severable” from the commitment to be bound by a treaty so that
the reservation can be “quarantined” and the overall commitment to the treaty preserved.
This last proposition was spectacularly illustrated in the already mentioned Loizidou
case. As has been seen, Turkey’s reservation to the recognition of the Court’s compulsory
jurisdiction (excluding Northern Cyprus) was found invalid. This then required the Court
to ascertain whether Turkey’s recognition of jurisdiction was nonetheless still valid.
Under normal international law rules, the tendency would be to consider that the whole
nature of the consent to the treaty had been “contaminated” by the invalidity of the
reservation. Against Turkey’s arguments, however, the Court considered that the
reservation could be severed from the recognition of jurisdiction, so that Turkey would
still be considered to have accepted that jurisdiction.
44
This was all the more so since, the
Court argued, Turkey must have known at the time that its reservation was invalid,
especially in view of protests by several state parties. In other words, Turkey was made to
“pay the cost” of its invalid reservation, notwithstanding the fact that it had been made
very explicitly and had gone unchallenged judicially for several years before the issue
came to Court.
45
Finally, the situation of the third state is perhaps the simplest. Under normal international
law, a state that agrees to another state’s reservation is bound in its relations with the
other state by that reservation.
46
However, that is not a very helpful way of thinking of
human rights treaties, for which it should not make a difference to states’ commitment to
populations within their jurisdiction whether other states have sought to limit their
ratification in any way. In particular, the fact that a state accepts another state’s invalid
reservation certainly does not make that reservation applicable between them. Similarly,
a state that objects to another state’s reservation, even on grounds of invalidity, cannot
43
Effect of Reservations 41 at ¶33.
44
“The special character of the Convention regime (…), it must be said, militates in favour of the
severance of the impugned clauses since it is by this technique that the rights and freedoms set
out in the Convention may be ensured in all areas falling within Turkey's "jurisdiction" within the
meaning of Article 1 (art. 1) of the Convention”. Kefalas and Others, (1995) Series A no.38 at
¶96.
45
Lozidiou v Turkey, n13.
46
Vienna Convention, n38, article 20(4)(b)).
18
draw on this to then claim that it is not bound in its relations to the other state by the
relevant provision, or that it is not bound at all by the relevant provision, or that it is not
bound by the treaty at all. Needless to say it should again make no difference for the entry
into force of a human rights treaty that a state is objecting to a reservation, precisely
because such a reservation cannot in any significant way be said to “affect” the interests
of that state in the way normally understood in treaty law.
47
As the Human Rights
Committee put it:
Human rights treaties, and the Covenant specifically, are not a web of inter-State
exchanges of mutual obligations. They concern the endowment of individuals
with rights. The principle of inter-State reciprocity has no place. (…) Therefore a
State not released from his general obligation to abide by the International
Covenant, merely because it disapproves of a reservation made by another state.
48
III. State Succession
Another manifestation of the specificity of human rights obligations lies in their
propensity to pass over to successor states, an issue that may seem of marginal relevance
today but which was important in the context of decolonization and post-Soviet
transitions. The issue of state succession to obligations, whether arising out of treaties or
responsibility incurred by the previous regime, is an old one that pits different
international legal interests against each other. On the one hand, the international system
has an interest in the continuity of international obligations, if only for the sake of legal
certainty. On the other hand, respect for sovereignty suggests that an authentically new
state should not have to take on the obligations of its distinct predecessor.
It is this latter position that has historically gained most ground, especially in the context
of decolonization. The principal idea is that of the “clean slate” (sometimes described in
Latin as tabula rasa), namely that the new sovereign has the option to assume which
obligations of its predecessor it chooses. One can see, of course, how problematic this is
from the point of view of international human rights and how it could set back the rights
agenda of some of the states concerned by several decades. Perhaps most importantly,
there is a suspicion that many situations of state succession occur in the midst of state
breakups which often occasion political violence, and make human rights particularly
necessary. There is a keen interest on the part of human rights law of having the relevant
obligations shouldered by the new state “from day one”.
Again, it is the idea of the special character of human rights obligations that has allowed
human rights treaties to be treated differently from what would be the case under the
general regime of international law. The idea, borrowed from old minority protection
47
For example, the UK could not say that it is not bound by the ECHR or its recognition of the
Court’s jurisdiction simply because it objects to Turkey having limited its jurisdiction to its
continental territory. One can see from such examples how the application of normal principles of
international law to human rights treaties is almost nonsensical.
48
GC 24, n19, ¶17.
19
PCIJ cases
49
and early ILO practice,
50
has been to claim that rights are vested in
individuals (or, less plausibly, embedded in the territory) in the first place, so that no
change of sovereign could deprive individuals of their rights. This is what is sometimes
known as the theory of “acquired rights” and is a manifestation of the same sort of
reasoning that denied North Korea an option to withdraw.
51
It confirms a very strong bias
in favour of continuity and non-regression of human rights obligations, even as, out of
deference to sovereignty, states are otherwise free to decide what treaty obligations they
will assume.
The strategy was put to good use by the HRC in 1993 in relation to fall of USSR/ It was
used to ensure that “all the peoples within the territory of a former State party to the
Covenant remained entitled to the guarantees of the Covenant, and that, in particular,
Armenia, Georgia, Kazakhstan, Kyrgyzstan, Tajikistan, the former Yugoslav Republic of
Macedonia, Turkmenistan and Uzbekistan were bound by the obligations of the Covenant
as from the dates of their independence.”
52
It was used to similar effect to consider that
all successor states to the former-Yugoslavia became immediately bound by the
Genocide Convention upon becoming independent. Judge Weeramanty, in particular,
wrote a separate opinion in the ICJ genocide case emphasizing that whether successor
states were parties to the Genocide Convention was not simply a question for the ordinary
law of treaties, but had to be understood in relation to the “special characteristics of that
Convention.”
53
As a treaty of humanitarian character it “embodies a commitment of the
participating states to certain norms and values recognized by the international
community” which cannot be put into question by successor states.
54
IV. Implementation
Under traditional international law, the manner in which an obligation is supposed to be
discharged is typically not specified. States simply have to do what they commit
themselves to do, and considerable discretion is left as to the means chosen. International
obligations, whether conventional or customary, only create rights between states.
Because they do not create rights (even less human rights) for individuals, there is little to
be said about domestic implementation. Of course, if a state commits itself to a regime of
free trade, for example, it will in all likelihood have to adopt domestic laws to reduce
tariffs and other obstacles. However, that is essentially that state’s problem, and other
states have no particular say on how it goes about discharging its obligations.
49
Questions relating to Settlers of German Origin in Poland (Advisory Opinion) PCIJ Ser. B.,
No. 5, 1923.
50
It was the ILO’s practice to tell newly decolonized states that they would only become parties
to the ILO if they accepted all ILO treaty obligations incurred by their predecessors.
51
See S-G, C.N,467.1997.TREATIES10, 12 Nov. 1997; UNCHR, ‘General Comment no.26(61)’
(adopted by the Committee at its 1631
st
Meeting, 8 Dec 1997).
52
UHCHR, ‘Comprehensive implementation of and follow-up to the Vienna Declaration and
Programme of Action’ U.N. Doc. E/CN.4/1995/80 (1995) at ¶4.
53
Separate Opinion of Judge Weeramantry, Application of the Convention on the Prevention and
Punishment of the Crime of Genocide, ICJ Reports (1996) 595 at 645.
54
Indeed, probably by any state as a result of the customary character of the obligations involved.
20
When it comes to human rights, this traditional, laissez-faire approach would not do. This
is partly because human rights obligations are typically more complicated and diverse
than say, respecting diplomatic immunities. And it is partly because human rights, as a
much more intimate affair for the state, made the drive to make these obligations more
detailed much more necessary to counteract sovereign resistance to being told
internationally how to behave domestically.
International human rights law has thus come up with a complex and novel concept of
how obligations are to be discharged. An entire new vocabulary of obligation has
emerged. Typically, states are supposed to “ensure” the rights to all individuals”.
55
This
means that they must formally guarantee the rights, for example through appropriate
laws. However, what is at stake goes much further. Treaty bodies, confronted with claims
by states in their periodic reports
56
that “all was well in the best of worlds” simply
because certain rights were formally protected in the constitution, have insisted that the
duty to secure human rights includes an obligation to adopt more than just legislative
measures, but also “judicial, administrative and educative and other appropriate
measures”
57
and an obligation “to organize the structure of the state apparatus in a way
that ensures the full exercise of human rights”.
58
The duty to secure human rights is, perhaps first and foremost, a duty to “respect” human
rights. This means the fairly obvious idea that states should not consciously violate rights,
for example through their agents (civil servants, the police, the army). However, if that
“hands off” obligation were the only one, it would leave glaring holes in human rights
protection. Some civil and political rights, such as the right to vote, are meaningless if the
state does nothing to implement them. Even the more recognisably “negative” rights
(i.e.those that principally require states to abstain from doing certain things), in fact
require the state to engage in proactive action (for example, “not to torture” will often
mean all kinds of concrete measures to prevent and sanction torture). Economic and
social rights are perhaps a typical example where “respect” for these rights will entail
developing specific policies, and where “not doing anything” is not an option (which is
arguably not the same thing as simply having laissez-faire policies on the assumption that
these will be most conducive to ensuring economic and social rights).
In fact, increasingly respecting rights has been understood not only as not directly
violating them but doing everything reasonable to ensure that they are not violated in the
first place and that if they are, something is done about it. This involves both a preventive
and a reactive dimension. Failing to take action to remedy a human rights violation can
be considered tantamount to endorsing that violation and is in itself a violation of the
right initially violated. In the landmark case of Velasquez Rodriguez vs. Honduras,
59
55
ICCPR, nError! Bookmark not defined., art 2,
56
On periodic reports, see Section IV.
57
UNCHR, ‘General Comment No. 31 [80] Nature of the General Legal Obligation Imposed on
States Parties to the Covenant’ CCPR/C/21/Rev.1/Add.13 at ¶7.
58
59
Velasquez Rodriguez Case, Inter-Am.Ct.H.R. (Ser. C) No. 4 (1988)
21
which involved the forced disappearance of a student activist who was arrested and never
to be seen again, the Court insisted that:
The State is obligated to investigate every situation involving a violation
of the rights protected by the Convention. If the State apparatus acts in
such a way that the violation goes unpunished and the victim's full
enjoyment of such rights is not restored as soon as possible, the State has
failed to comply with its duty to ensure the free and full exercise of those
rights to the persons within its jurisdiction.
60
In addition, and this phrase is becoming ever more important, states should “protect”
individuals from human rights violations. This means that the state needs to proactively
ensure that persons within its jurisdiction do not suffer from human rights violations at
the hands of third parties.
61
This is much more akin to creating an environment in which
rights are enjoyed. Of course, the state does not become liable for every adverse
interference with individuals’ rights by other private actors. But, under the theory of
indirect horizontal effect, the state is liable for those failures that can be traced to its
shortcomings in protecting individuals from other individuals for example because it has
adopted a law that made the violation possible, or because it has failed to do something
when it should have which would have prevented the violation from happening.
62
Under
that latter scenario, states have been found liable for failing to protect demonstrators from
third parties,
63
or failing to protect an individual from murder despite the fact that the
police knew his life had been threatened.
64
The idea that the state should protect
individuals has been particularly used to protect women from domestic violence in cases
where the state failed to provide adequate structures or legal protection.
65
As the Inter-
60
Ibid, ¶176.
61
Note : this is analytically quite distinct from the issue of whether these third parties,
presumably mostly private actors, can be the subjects of human rights obligations, an issue dealt
with in Chapter 25.
62
See for example Case of X and Y v. The Netherlands (App No. 8978/80) ECHR, 26 March
1985. Also Young, James & Webster v. United Kingdom, 44 Eur.Ct.H.R. Ser. A (1981). The case
involved a “closed shop” agreement between British Rail and three trade unions, providing that
thenceforth membership of one of those unions was a condition of employment. The applicants,
who had been dismissed as a result of not belonging to a union, alleged a violation of article 11.
The Court understood “securing” rights under the Convention as not enacting laws that could lead
third parties to violate the rights of persons within their jurisdiction, so that the argument that it
was British Rail which had most immediately violated the right was irrelevant.
63
Plattform “Arzt für das Leben” v. Austria, (1988) Ser.A139.
64
Osman v. The UK (App No. 23452/94) ECHR,28 Oct. 1998; HRC, William Eduardo Delgado
Páez v. Colombia, Communication No. 195/1985, U.N. Doc.CCPR/C/39/D/195/1985 (1990).
65
CEDAW, ‘Views of the Committee on the Elimination of Discrimination against Women under
Article 7, paragraph 3, of the Optional Protocol to the Convention on the Elimination of All
Forms of Discrimination against Women’ Communication No.2/2003, Ms. A. T. v. Hungary
(Views adopted on 26 January 2005).
22
American Court put it, in the already mentioned Velasquez Rodriguez case,
66
in which the
identity of the abductors was not clear:
An illegal act which violates human rights and which is initially not
directly imputable to a State (for example, because it is the act of a private
person or because the person responsible has not been identified) can lead
to international responsibility of the State, not because of the act itself, but
because of the lack of due diligence to prevent the violation or to respond
to it as required by the Convention. (the state is liable when it) allows
private persons or groups to act freely and with impunity to the detriment
of the rights recognized by the Convention.
67
Finally, there is sometimes talk of a duty to “promote” human rights. Among other
things, this suggests that states should concern themselves with human rights beyond
their borders, and can be the foundation of a foreign policy based on human rights.
V. Incorporation
A particular problem that has arisen is whether the obligation to secure human rights
entails any consequence in terms of the domestic legal status of human rights obligations.
In international law generally, incorporation into domestic law is a secondary issue since
international treaties are not meant to create any domestic rights. But when it comes to
international human rights law, exactly the contrary is true, since the whole point of these
treaties it to create such rights.
As is well known, domestic legal systems vary in the degree and kind of receptivity to
international obligations. Monist states typically consider that international law becomes
applicable domestically without more being required. Although this can sometimes be a
bit abstract and it cannot harm a ratification that some of its provisions are “translated”
into domestic law, the advantage of human rights is that they are relatively “self
executing” and understandable from the point of view of domestic law. Dualist states,
among which the UK, US and most common law jurisdictions, consider that international
treaties only become applicable domestically to the extent they have been implemented
into domestic law (somewhat paradoxically, many dualist states also consider that
customary international law is the “law of the land”). Of course, there is nothing
implausible about adhering to a strict dualist orthodoxy even when it comes to human
rights. In the dualist mindset, states bind themselves internationally to certain standards
which they simultaneously refuse to grant domestic legal applicability, the idea being that
they remain liable for violations of their obligations internationally without those creating
specific legal relations domestically. It may in fact be that in some circumstances rights
will be protected, or at least not violated, despite not being protected as such under
domestic law.
66
Velasquez Rodriguez Case, n59.
67
Ibid, ¶172.
23
However, the dualist tradition has undeniably been strongly challenged by the emergence
of international human rights law. On the long term, a rigid dichotomy between domestic
and international law creates problems. First, it seems incompatible with the ambition of
international human rights of being domestically visible and operative so that they remain
as distant legal artefacts. This is especially the case since, as we have seen,
internationally guaranteed human rights require substantial modifications of domestic law
which one may think will stand a better chance of being implemented if human rights are
“closer to home”. Second, non-incorporation is an invitation to litigation by individuals
drawing attention on gaps between domestic law and the state’s international
commitments, that will sooner or later play against the state’s best interests
68
(so long at
least as the state is minimally committed to protecting human rights). Indeed, the very
nature of human rights recommends that they be applicable directly into domestic law,
and invocable before domestic courts. This is, again, because international human rights
guarantees are primarily for the benefit of individuals. It is individuals who will suffer
human rights violations, it is they again who are often best placed to complain about this
and it is, in an increasing number of cases, individuals who can bring petitions directly or
indirectly to international bodies. The adoption of the UK Human Rights Act, for
example, probably signalled the long term untenability of totally dissociating domestic
rights protection from the possibility of petitions before an international human rights
body. The adoption of the Human Rights Act in fact demonstrates the opinion that it is
far better if violations stand a chance of being corrected first and foremost under
domestic law.
That dualism which has undergone very significant changes is also underscored by the
fact that, even in countries that have not implemented international human rights
obligations domestically, there are a great many calls to interpret domestic law as far as
possible in ways that are compatible with those (for example, by presuming that the
legislator cannot have intended to contradict the state’s international obligations).
69
VI. Limitations
Limitations are not the same as reservations (where a state purports to exclude or
unilaterally reduce an obligation) or as derogations (where a state suspends certain
obligations). They operate within each obligation to qualify its exact scope. A specificity
of human rights obligations is a quite unique theory of limitations.
In general international law, few treaties come with broad limitation clauses allowing
states to limit the degree of their obligations because of social priorities. For example,
under the GATT states are supposed to reduce trade barriers, and cannot allege that
possible social cost implications should redefine what it means to reduce barriers. But
68
In some ways, the traditional concerns behind dualism – namely the need to better protect
sovereignty and democracy – can paradoxically be undermined in conditions where a right to
petition international bodies to engage the responsibility of the state exists.
69
See for example the Calcutta Declaration,
<http://www.pipfpd.org/index.php?option=com_content&view=article&id=43&Itemid=74>
24
human rights operate in more complex ways. They inevitably raise complex tensions
between individual and collective demands. To take an easy example, an individual who
is imprisoned for having committed a crime following a fair trial, cannot invoke a
violation of his right not to be detained. It is understood that the right to be free is limited
by the exigencies of living in society and public order. Moreover, it is at least primarily
for the state to define the framework of collective life that constrains certain rights, in
ways that other states will have little cause to complain about (unlike a normal
international treaty situation based on exact reciprocity).
Despite what is sometimes suggested, the vast majority of human rights are thus not
absolutes. One of the only exceptions is the right to be free from torture which has been
categorically excluded from any means-end calculus (despite attempts to bring that
equation back in). Sometimes, because a right is very sensitive (and all rights relating to
integrity and security of the person are) the limitations will be spelled out in detail. This
is the case for the right to life (which is limited by a residual possibility of imposing the
death penalty, war, and police work using reasonable force) or the right to be free from
detention (which is limited by the possibility of being sentenced to imprisonment for a
criminal offence, pursuant to a fair trial). However, it would be impractical to spell out in
advance the exact ways in which most other rights can be limited. For example, there is
certainly a right to privacy, but that right is susceptible to restrictions based on the pursuit
of certain legitimate objectives (e.g.the fight against crime or terrorism). Instead of trying
to specify all of the ways in which particular rights can be limited, international human
rights law relies on either specific (i.e.article by article) or general limitation clauses.
70
Typically limitations to rights have to be justified by the fact that they are (i) for a
legitimate aim, (ii) necessary and (iii) proportional.
According to the ECHR, a legitimate aim must be a “pressing social need” (one cannot
overturn human rights simply for idiosyncratic policy preferences). Typically pressing
social needs are identified by human rights treaties themselves as grounds for limitations
of rights that are prima facie valid. For example, the UDHR mentions “morality, public
order and the general welfare”.
71
These are then expanded through interpretation by
courts. The protection of democracy, pluralism, or secularism and a great many others
have all been presented at various times as being legitimate goals. It is interesting that
human rights law, which is often presented as being against the state, is also very much
about legitimizing a certain state role. Another particularly interesting limitation to rights
is the need to protect the rights of others.
Second, the measure has to be necessary. It does not suffice to simply allege a desirable
social policy and then claim to adopt a measure that is in fact not truly connected to that
goal. Many laws negatively constructing homosexuality, for example, which were
presented as in pursuant of the legitimate aim of maintaining good mores, were struck
down because they were not related to that goal at all, especially since they stigmatized
homosexuality in the private sphere.
70
See, for example, Universal Declaration of Human Rights (adopted 10 December 1948
UNGA Res 217 A(III) (UDHR) Art 29(2).
71
Ibid.
25
Third, the relevant measure must be proportional to the goal sought. Proportionality
involves a complex equation measuring the importance of the social aim, the importance
of the right guaranteed, and evaluating both in relation to each other, particularly by
looking at the degree of encroachment. Obviously certain rights will be very inelastic to
even the most important social goals (e.g.the right to be free from arbitrary detention will
be very hard to limit even for the sake of fighting terrorism), whilst others will be more
elastic (e.g.the right to demonstrate will be quite sensitive to states’ invocation of public
order or the need to protect the rights of others). It is quite frequent in the case law of the
ECHR for measures to be found disproportional. For example, in the Turkish Communist
Party case, whilst the Court was receptive to the threat to public order created by some
aspects of that party’s program, and the need to adopt some measures to deal with it, it
found that requiring the immediate and permanent dissolution of that party was
disproportionate so that the measure was in fact not necessary.
72
The onus to prove any limb of that test is typically on states, who cannot simply claim
that they “know better”. But by what standard are such limitations to be judged to be
legitimate? After all, what is necessary for some may be seen as perfectly abusive by
others. Several human rights instruments suggest that the key threshold is that of a
“democratic society”.
73
In other words, limitations are justified on all of the grounds that
have just been introduced, if they conform to what would be considered so in a
“democratic society”. This is certainly a standard that can be criticized as a little hazy
(not to mention the possible tensions between democracy and human rights), but at least
such limitation clauses acknowledge that human rights is not simply about upholding
absolutes but very often about the complex exercise of balancing individual rights against
legitimate social priorities. Obviously, human rights should still be “taken seriously” in
the words of Ronald Dworkin,
74
but the emphasis is on state justification, transparency
and accountability, where states might otherwise gladly invoke the priority of politics
over rights.
VII. “Margin of appreciation”
In general international law, all states are bound similarly by similar obligations. For
example, if states are obliged to respect diplomatic immunities, they are not free to define
the degree of that protection in any significant way to take into account the peculiarities
of their legal system (although they may choose the means most appropriate in the
context). The system would not work if it allowed this sort of discretion and it is in fact a
characteristic tenet of international law that states cannot invoke the specificities of their
domestic law to violate or even modify their internationally incurred obligations.
When it comes to human rights however, the fact that the range of obligations is much
broader threatens to modify long-held practices by requiring substantial domestic
72
United Communist Party of Turkey [TBKP] and others v. Turkey (App No.19392/92) ECHR 28
Oct. 1998.
73
For example see ECHR, nError! Bookmark not defined., art 8, 9, 10.
74
Ronald Dworkin, Taking Rights Seriously, (Harvard University Press, Cambridge, 1978).
26
modifications. Moreover, whereas the means and ends of many international obligations
are quite clear and spelled out in detail, human rights policy typically involves very
complex decisions that seek to translate broad goals into actually guaranteed rights.
75
At
the same time, states’ interest in ensuring that other states discharge their duties in
exactly the same way as they do is much relativized by the fact that no “exchange” of
rights and obligations is involved.
76
This explains that international human rights law has been at the forefront of efforts to
develop a geographically and culturally plural notion of implementation. International
human rights do not strive for absolute uniformity as much as to secure certain minimum
standards, while respecting the cultural, legal and political specificity of each state. One
rationale for this is that states and domestic courts are better suited to assess local
peculiarities and that there is simply too much uncertainty about how human rights are to
be implemented for international supervision to exercise more than relatively minimal
control. Issues of legitimacy also arise, as international bodies will probably provoke
political backlashes if they delve too deeply in matters that are seen as quite specific
culturally. Finally, deferring to states for the implementation of their obligations is also
recognition of the fact that human rights is not about replacing sovereignty as much as
limiting some of its excesses. As the ECHR put it:
the main purpose of the Convention is "to lay down certain international
standards to be observed by the Contracting States in their relations with
persons under their jurisdiction". This does not mean that absolute
uniformity is required and, indeed, since the Contracting States remain
free to choose the measures which they consider appropriate, the Court
cannot be oblivious of the substantive or procedural features of their
respective domestic laws.
77
At times, deferring to the local will be based on the particular ability of the state to
determine certain sensitive matters. The key example is that of national security issues,
particularly (and problematically) when situations of national emergency are involved.
As the ECHR put it in the case of Brannigan and MacBride:
It is not the Court's role to substitute its view as to what measures were
most appropriate or expedient at the relevant time in dealing with an
emergency situation for that of the Government which have direct
75
For instance, the right to be free from torture translates relatively simply into a “do not torture”
obligation, although even that requires a degree of implementation that is policy sensitive
(training, procedures, etc). However, a great many rights require considerable policy input and
adaptation to local realities. That is typically true of economic and social rights, for example,
which for historical reasons have built in a considerable neutrality as to means. It is also true of
most civil and political rights (e.g. freedom of expression, right to family life, the right to a fair
trial, etc).
76
In fact, many states may be keenly aware that their ability to exercise a modicum of discretion
in the implementation of human rights is dependent on their tolerating such discretion in others.
77
Sunday Times v. The UK (App No. 6538/74) ECHR 6 Nov. 1980.
27
responsibility for establishing the balance between the taking of effective
measures to combat terrorism on the one hand, and respecting individual
rights on the other (…).
78
This deference to a State party has been criticized as too broad, and it is probably the case
that the ECHR would today be more aggressive in challenging states’ assertions of
threats to national security as justifying derogations.
More generally deference to states’ evaluation is rampant in the international human
rights context. It often involves what has come to be known, particularly in European
circles, as the “margin of appreciation”. The margin of appreciation is the idea that
although states are bound by the same standards, they also have the unmistakable leeway
in assessing what these standards imply domestically. This is often related to the
jurisdictional notion that the “machinery of protection established by the Convention is
subsidiary to the national systems safeguarding human rights”.
79
The margin of
appreciation has been particularly relevant in matters of mores and morality where,
according to the Court, “… it is not possible to find in the domestic law of the various
Contracting States a uniform European conception of morals.”
80
This means that “by
reason of their direct and continuous contact with the vital forces of their countries, State
authorities are in principle in a better position than the international judge to give an
opinion”.
81
There are of course limits to the “margin of appreciation”; it has been
criticized as leading to an excessively relativistic application of human rights, that may
even contain the seeds of the project’s dissolution. Accordingly, the ECHR views the
margin as “go(ing) (…) hand in hand with a European supervision”.
82
The margin of appreciation has been particularly relevant to cases involving sexual
minorities,
83
corporal punishment
84
or the place of religion in society,
85
to mention only a
few examples. It contains both a temporal and spatial dimension. Time wise, the margin
of appreciation is something that can evolve over time in relation to a certain subject. For
example, the Council of Europe member states might have had a considerable margin of
appreciation 50 years ago when it came to criminalizing homosexuality, but the very
same treaty today is interpreted radically differently in a way that suggests that
criminalizing same sex relations between consenting adults is a violation of several
rights. There is also a spatial element to the assessment of what the margin allows in a
78
Brannigan and MacBridge v. The UK (App No.14553/89, 14554/89) ECHR 21 Feb 1992 at
¶59.
The Belgian Linguistic case, (Merits) (1968) Series A no. 6, p. 35, ¶10 in fine
80
Handyside v. United Kingdom (1976) 1 EHHR. 737 at ¶48
81
Ibid.
82
Kjeldsen, Busk Madsen and Pedersen v. Denmark, (Application no. 5095/71; 5920/72;
5926/72) ECHR 7 Dec 1976 at ¶49
83
Handyside v. United Kingdom, n80 ; Frette v. France (Application No. 36515/97) ECHR12
June 2001Goodwin v. The UK (App No. 28957/95) ECHR 11 July 2002.
84
Costello Roberts v. The UK (App No.13134/87) ECHR 25 March 1993.
85
Otto-Preminger-Institut v. Austria (App No13470/87) (1994) ECHR 26 Wingrove v. UK (App
No. 17419/90) (1996) 24 EHRR, at ¶57; Inter-American Court of Human Rights, Case of “The
Last Temptation of Christ” v. Chile (Olmedo Bustos et al.) (2001) Ser. C, No. 73, ¶64.
28
particular case, in that the Court will typically look at the degree of consensus (a
minimum common denominator) about a certain practice across member states. In the
absence of such consensus and more importantly, in the presence of divisions, the Court
will hesitate to impose on a minority (even less a majority) a particular understanding of
rights. However, the opposite is true and states can be “marginalized” by parallel
evolutions in a great many member states, that make their practices look “out of sync”
with evolving understandings about human rights.
One of the consequences of this is that the margin of appreciation militates in favour of a
conservative international assessment of rights. The ECHR, for example, has often been
less of a pioneer than a safe endorser of existing trends and developments. But on a more
positive note, the margin of appreciation also reinforces the sense of human rights being
rooted in a community of reference (which is also a community of interpretation),
dynamic in time, and a product of a constant interaction between rights and ideas about
society and justice.
VIII. Derogations
The law of ordinary state responsibility traditionally contemplates that a defence of
“necessity” is available to states in limited circumstances. According to the Articles on
State Responsibility drafted by the International Law Commission and adopted by the
General Assembly, such a defence is invocable when “an essential interest of a State is
faced with grave and imminent peril, the essential interest outweighs the interest of a
third-party State in non-breach, an no other means of safeguarding the essential interest is
available”.
86
These are stringent conditions when it comes to ordinary international legal
obligations. However, invocation of the ordinary regime of necessity under general
international law is not an option for human rights regimes, for at least three reasons.
First, the conception of a state of necessity under general international law is one that
requires balancing the essential interest of a derogating state with the “interest of a third-
party State in non-breach”, something which is clearly not applicable to international
human rights treaties, where the third-party state’s interest is negligible in relation to the
international community’s or that of human beneficiaries.
87
Second, human rights have
an axiological priority which suggests that they should be less elastic to the demands of
necessity than other obligations in that, quite simply, “more is at stake”. Indeed, one idea
is that, unlike “ordinary” international law obligations, human rights are most needed
precisely in times of necessity, which can cast doubt on states’ claims that human rights
need to be suspended to confront emergencies. Third and most significantly, international
human rights law already contains its own suspensive regime for situations of “national
emergency”, allowing states to derogate from certain provisions, as long as certain
conditions are complied with. This specific regime thus displaces the default international
86
Draft Articles on Responsibility of States for Internationally Wrongful Acts, with commentaries
(2001) at art 25, ¶11.
87
If anything, the relevant factor would be the interest of the population within the state’s
jurisdiction in non-breach versus that state’s essential interest.
29
law regime, and constitutes international human rights’ lex specialis (specific law, in
Latin). It would be incompatible with the object and purpose of human rights treaties,
which is already to create a regime for both ordinary times and times of exception, if an
added layer of “exceptionalism” could be claimed under general international law. In
other words, human rights treaties come with their own embedded situations of exception
which they regulate in a self-sufficient and exhaustive manner.
The consequence of recognizing states’ power to derogate from human rights treaties in
times of emergency, furthermore, is an increased regulation and supervision of such
derogations by international organs. The regime applicable in situations of emergency,
even though it borrows ideas from situations of necessity, is typically more constraining
than for ordinary international law obligations. First, a situation of national emergency
under international human rights law needs to be formally declared in advance (where a
state of necessity can merely be invoked ex post if confronted by another state claiming
an obligation towards it has not been fulfilled), and its existence (and abrogation)
formally
88
notified to relevant authorities. Second, a situation of national emergency
cannot be invoked, as under general international law, merely when “an essential interest
of a State is faced with grave and imminent peril” but rather when, in fact, there exists a
threat to “the life of the nation” itself.
89
This is clearly a higher threshold, and such a
threat cannot be inferred from “every disturbance or catastrophe.” Not even an armed
conflict will systematically justify the recognition of a public emergency.
90
The ECHR,
for example, has described a public emergency under Article 15 as “an exceptional
situation of crisis or emergency which affects the whole population and constitutes a
threat to the organized life of the community of which the State is composed.”
91
Third, in contrast to the ordinary regime of international law, “the restoration of a state of
normalcy where full respect for the Covenant can again be secured must be the
predominant objective of a State party derogating from the Covenant.” In other words,
human rights are only really suspended in order for the state to be able to guarantee them
as soon as the danger has passed, not to achieve something that is in itself alien to human
rights. Fourth, because situations of public emergency are regulated regimes rather than
simply a defence to a breach of an obligation, they come with various strict conditions
concerning their temporary, necessary and proportional nature. Derogations should
specify their territorial
92
and temporal
93
reach. A standard provision in several
88
Cyprus v. Turkey (App No. 8007/77) (4 Oct 1983) EComHR at ¶ 67.
89
International Covenant on Civil and Political Rights (adopted 16 December
1966, entered into force 23 March 1976) 999 UNTS 171 (ICCPR) art 4(3), Convention for the
Protection of Human Rights and Fundamental Freedoms (European Convention on Human
Rights, as amended) (ECHR) art15(2).
90
UNCHR, ‘General Comment 29, States of Emergency (article 4)’ U.N. Doc.
CCPR/C/21/Rev.1/Add.11 (2001)
91
Lawless v. Ireland (1979-80)1 EHRR 15, ¶28.
92
For example, as the Court recognized in the Sakik and others v. Turkey (App no
87/1996/706/898-903) (1997) , the fact that Turkey had declared a state of emergency in
Kurdistan did not mean that it was therefore in a similar state of exception in Northern Cyprus.
30
international human rights treaties, for example, is that emergencies should only be
proclaimed “to the extent strictly required by the exigencies of the situation, provided that
such measures are not inconsistent with their other obligations under international law
and do not involve discrimination solely on the ground of race, colour, sex, language,
religion or social origin”.
94
Fifth, whereas almost all ordinary international obligations
are susceptible to a “state of necessity” invocation, situations of national emergency can
only be invoked to suspend a limited set of derogable rights. For example, it may be
legitimate for a state confronting a war or a major natural catastrophe to suspend the right
to demonstrate or the right to vote. This means that conversely there are a number of
“non-derogable” rights that can never be suspended.
95
For example, one cannot
“suspend” the right to life in order to deal with a situation of emergency.
96
Sixth, where a
state of necessity might normally provide a sort of blanket defence against the non-
execution of international obligations, human rights state of emergencies require states to
show that “specific measures taken pursuant to the derogation must also be shown to be
required by the exigencies of the situation.”
97
This relatively intense regulation of situations of emergency follows from the “special
character” of human rights norms as “minimal” obligations, whose suspension should
only be allowed in the most extraordinary circumstances (unlike ordinary obligations for
which a higher threshold of tolerance exists). It can also be seen as linked to the
asymmetry of the human rights relationship between a state and individuals which
requires a high degree of regulation, especially where the invocation of states of necessity
between states does not as much.
IX. Withdrawal
In international law, withdrawal from a treaty under certain conditions is always an
option, and is generally know as the possibility of “denouncing” a treaty. This is
consonant with the fact that international law traditionally does not particularly “care”
about the substantive content of obligations, and is merely intent on making sure that
93
The latter is particularly important in light of the tendency of states of exception to prolong
themselves far beyond what is strictly necessary. “Sunset clauses”, for example, should indicate
the exact date or the occurrence of which event will bring an end to the public emergency
94
American Declaration, n18, art 27(1).
95
The list varies slightly with treaties, but the ICCPR lists the following rights as non-derogable :
articles 6 (life), 7 (torture), 8 (paragraphs I and 2) (slavery), 11 (imprisonment for failure to fulfil
a contractual obligation), 15 (nullum crimen sine lege), 16 (recognition as a person under the law)
and 18 (freedom of thought, conscience and religion).
96
This is especially the case since the right to life already comes with a number of exceptions,
some of which seem to deal with situations of emergency (i.e. the possibility of imposing the
death penalty for treason in war time).
97
HRC, General Comment 29, nError! Bookmark not defined., ¶4. For example, simply
because there is a situation of national emergency does not mean that the right to demonstrate has
to be restricted throughout the territory and in any way possible: you have to show that each and
every specific curtailment of the right is in fact something that is a useful response to a situation
of emergency.
31
states adhere to whatever commitments they have made. As long as other parties are not
prejudiced by such withdrawal (i.e. as long as what is involved is a withdrawal and not a
breach), then pulling out of a treaty is something relatively inconsequential and routine
from an international legal point of view.
The situation is obviously more fraught with complexity when it comes to international
human rights, even though governments occasionally tout the possibility and use it as a
threat. One can see how problematic such withdrawals are from the point of view of
human rights. In purporting to align human rights with the ordinary regime of public
international law, they make human rights obligations appear as little more than
transitory preferences to be repealed (more or less) at a whim. Concretely, withdrawals
may have the effect of leaving certain populations legally vulnerable (they are of course
always factually vulnerable) to violations, especially since supervisory or jurisdictional
machinery often comes attached to substantive human rights commitments.
The position of international human rights law on this matter is not entirely settled, but
there appear to be two possibilities, depending on whether renunciation is explicitly
anticipated by the relevant treaty or not. Some international human rights instruments
actually (and a little surprisingly, although this may be a way of encouraging states to
ratify) contain denunciation clauses. For example, Article 21 of CERD posits that “A
State Party may denounce this Convention by written notification to the Secretary-
General of the United Nations”.
98
The ECHR
99
and ACHR
100
also include a
“denunciation clauses”. It should be noted, nonetheless, that despite the existence of such
provisions, straightforward withdrawal is quite unlikely. As a legal strategy, it would be
extremely confrontational and would occasion obvious political shortfalls (states will
often simply violate human rights, even as they formally remain bound by the treaties).
Withdrawing from a human rights treaty, even though a state may do so for other alleged
reasons, will in all likelihood be seen as an admission that a state is up to no good. Only
two states have historically availed themselves successfully of the possibility of
denouncing a human rights treaty, namely Greece which denounced the ECHR from
1969 to 1974 following a military coup, and Trinidad and Tobago, which denounced both
Optional Protocol I to the ICCPR and the IACHR from 1998 to 1999 . In both cases, the
denunciations did not have retroactive effect, and therefore these states could at least not
escape liability for violations committed prior to it (even if these violations only became
known after the denunciation was deposited). The UK government has in recent years
publicly played with the idea of withdrawing from the ECHR in order to “re-ratify” with
a reservation, but this has been met with scepticism.
101
98
Interestingly, the statute of the International Criminal Court also provides for a possibility of
withdrawal. See Statute of the International Criminal Court, (1 July 2002) A/CONF. 183/9, Art
127.
99
ECHR, nError! Bookmark not defined., art 28.
100
American Declaration, n18, art 44.
101
Anneliese Baldaccini, ‘The Recent British Proposals on Asylum Policy: A Critical Evaluation’
(Lecture at Justice in London 13 Nov 2003)
<http://www.cicerofoundation.org/pdf/lecture_baldaccini.pdf>
32
In fact, the emerging view when it comes to human rights treaties that do not explicitly
allow for denunciation and the one arguably more consonant with their special character,
is that this is simply not an option. States are thus denied the comfort (if nothing else) of
being able to claim that they are “no longer bound,” despite their best efforts to do so.
There is only one case to date of the issue being decided internationally, but it is an
interesting and unusual one. In 1998,, North Korea sought to withdraw entirely from the
ICCPR, after becoming upset about the HRC’s criticism of its record.
102
The question
was referred to the Secretary General as depository of the treaty, who promptly returned
an opinion to the effect that one could not withdraw from the ICCPR. From a purely
interpretative point of view, the absence of an explicit denunciation clause was already a
telltale sign,
103
one that pointed to the a-temporal nature of the ICCPR, and its open-
endedness in time. The implicit idea was that no supervening event or change of
circumstances could possibly invalidate the guarantees afforded. More importantly
perhaps, the very nature of the treaty suggested that withdrawal was not an option.
According to the SG, “it is clear that the Covenant is not the type of treaty which, by its
nature, implies a right of denunciation. (…) The Covenant does not have a temporary
character typical of treaties where a right of denunciation is deemed to be admitted,
notwithstanding the absence of a specific provision to that effect”.
104
Moreover, in what
was quite a bold move, the Secretary General went on to note that:
the rights enshrined in the Covenant belong to the people living in the
territory of the State party. The Human Rights Committee has consistently
taken the view, as evidenced by its long-standing practice, that once the
people are accorded the protection of the rights under the Covenant, such
protection devolves with territory and continues to belong to them,
notwithstanding change in government of the State party (…).
105
In other words, and in striking contrast to normal international law treaties (where
indirect beneficiaries have no “acquired interest” in the maintenance of certain treaty
relations), rights once recognized (rather than granted) cannot subsequently be
withdrawn. This suggests a unique “stickiness” to international human rights
commitments, which cannot be undone even when states desire to do so formally. The
102
Although not exactly the same thing, the US did try to “unsign” the Rome Statute that created
the ICC, although that possibility was quite rightly open to question.
103
A contrario, the Secretary General pointed out that Article 41(2) of the Covenant does permit a
State party to withdraw its acceptance of the competence of the Committee to examine inter-State
communications by filing an appropriate notice to that effect. The absence of a more general
denunciation can thus not be attributed to an oversight.
104
UNCHR, ‘General Comment No. 26: Continuity of Obligations’
CCPR/C/21/Rev/1/Add/8/Rev.1, para.3.
105
Ibid, ¶4.
33
Secretary General did point out, rather wryly, that North Korea could withdraw from the
ICCPR if all party states agreed to let it do so, which was obviously very unlikely.
106
A less drastic move than seeking to withdraw from substantive human rights obligations
is the attempt by some states to withdraw their recognition of the jurisdiction of certain
international human rights bodies. Following several adverse rulings about its “anti-
terrorist” efforts, Peru sought to withdraw its recognition of the jurisdiction of the
IACHR in 1999.
107
Internationally, this would not normally be particularly problematic
and the ICJ, for example, anticipates that recognition of its compulsory jurisdiction can
be withdrawn. Although human rights bodies’ position on this is less strict than when it
comes to substantive commitments, there is certainly a reluctance to allow states to treat
their commitment to international monitoring quite as freely as would be the case under
general international law. The IACHR, in particular, found that, in the absence of a
specific clause allowing withdrawal of recognition of jurisdiction and considering the
object and purpose of the treaty, Peru would have had to denounce the convention as a
whole, which it had not (the issue was left open as to whether such a broad denunciation
would be valid).
108
In doing so, it followed the argument of the Commission which
insisted that:
The framers of the Convention established a unitarian system of rights and
obligations in the multilateral ambit, not a series of relationships among
States essentially contractual and reciprocal in nature. While the unilateral
withdrawal from obligations assumed conditionally in the sphere of
relations among the States of the latter category may be permissible under
certain circumstances (…) such an act has no legal basis in the special
regime for human rights, and is incompatible with the object and purpose
of the Convention.
109
It is also to be noted that, consonant with the idea that one cannot withdraw from a
human rights treaty, one cannot be excluded by others from one either. It is of course the
case that states can be excluded from certain international organizations as a result of
systematically violating human rights,
110
but being excluded from such institutional
arrangements is not the same thing as being released from the substantive commitments
that underlie them. Again, whilst exclusion makes sense for “ordinary” treaties, in which
a state is in violation of its obligations and where it should not be entitled to claim
106
In fact, one may wonder if theoretically it would even be possible for all states to allow North
Korea to withdraw from the ICCPR. At any rate, were something of the sort ever to happen, the
international human rights regime might as well have completely broken down.
107
See Peruvian Legislative Resolution No. 271532 in the Inter-American Court of Human
Rights press release CP2/01 at <http://corteidh.or.cr>
108
Inter-American Court on Human Right, Caso del Tribunal Constitucional, Compentencia,
Sentencia (24 Sep 1999) Series C No.55 and Inter-American Court of Human Rights, Ivcher-
Bronstein Case (Baruch Ivcher Bronstein vs. Peru), 6 Feb 2001.
109
Second Report on the Situation of Human Rights in Peru, Inter-American Court of Human
Rights, OEA/Ser.L/V/II.106, Doc. 59 rev. (2000).
110
The issue has arisen on several occasions in relation to Russia and the ECHR.
34
execution of their obligations from others, it is nonsensical for human rights treaties,
where the contravening state is not deriving any evident “reciprocal” benefit from being a
party to a human rights treaty with others. Moreover, exclusion could have disastrous
consequences (and it is hard to imagine that it could have positive ones) for the
populations that are the intended beneficiaries of human rights arrangements. The
international community would strangely be doing such states a “favour” in releasing
them from their obligations, where it should instead seek to put as much pressure on
them. The issue was, for example, always whether Russia should be excluded from the
Council of Europe or have its right to vote suspended, never whether Russia should
remain a party to the ECHR.
X. Remedies
A typical way in which international law obligations are enforced, traditionally, is
through termination, suspension or even counter-measures by the party that considers that
its legal interests have been violated.
111
Non-execution is a valid remedy to non-
execution, in a tit for tat that is very characteristic of fundamentally horizontal relations.
This is understandable in an inter-state world in which the dominant model is contractual,
and where one is really only bound to the extent that other parties respect their “share of
the deal”. It is also a consequence of the fact that quite often enforcement bodies will not
be available, so that each state has to decide for itself whether obligations owed to it have
been violated and whether one of these remedies is justified. This is what is sometimes
known as the “decentralized” enforcement of international law.
This was of course always problematic even under classical international law, but it is
arguably even more so when it comes to international human rights law. In fact, quite
strikingly, Article 60(5) of the Vienna Convention on counter-measures anticipates that
this possibility “does not apply to provisions relating to the protection of the human
person contained in treaties of a humanitarian character, in particular to provisions
prohibiting any form of reprisals against persons protected by such treaties”. The
reference to “treaties of a humanitarian character” was meant originally as a reference to
“laws of war” instruments, and to the long simmering debate on whether “belligerent
reprisals” are still legal.
112
However, the more general idea of provisions “relating to the
protection of the human person” clearly encompasses human rights treaties.
113
111
The possibility is anticipated by the Vienna Convention, n38, art 60 .
112
They undeniably once were. The idea of a reprisal is that one will engage in what would
otherwise constitute a violation of the law as “punishment” for another state’s breach of it.
Reprisals became heavily challenged as the goal of the laws of war gradually came to be
redefined not as protecting the interests of states, but as safeguarding non-combatants,
particularly civilians.
113
In fact one could argue that what is true for humanitarian treaties is a fortiori true for human
rights ones. At least in contexts of international armed conflict there is a clear interaction between
two parties which may create at least sociological expectations of a certain reciprocity. The same
is not true of human rights treaties.
35
The idea is simple enough: whereas under normal international law violations, it is
reasonable that the violating states suffer the consequences of its unwillingness or
inability to respect its obligations, the same cannot be said of human rights obligations,
which are not primarily secured for the benefit of other states. Several scenarios are
conceivable. For example, in a context where state A violated the human rights of
citizens of state B on its territory, one can imagine that state B could cease to honour its
human rights obligations vis-à-vis the citizens of state A. However, from a human rights
point of view, this would clearly run against the object and purpose of human rights
treaties, whose obligations are not primarily owed to other state parties (even when it
comes to their nationals). The populations in question cannot find themselves “taken
hostage” in an intestate dispute that is or should clearly not be their problem, and which
should be irrelevant to the fundamental teleology of human rights as a regime designed to
safeguard basic rights for all. The ICJ held that much when it affirmed that Bosnia could
not have invoked a violation of genocide convention by Yugoslavia, to do the same in
turn.
114
Even more nonsensical is the possibility that state A would cease to respect its
international human rights obligations vis-à-vis its own nationals as a counter-measure to
state B not respecting its own obligations vis-à-vis persons within its jurisdiction. This
would be the equivalent of a legal “shooting oneself in the foot” (the state would
essentially be punishing itself), which would have no effect on the offender state. It
would be perfectly absurd from a human rights point of view in that it would punish the
population of state A in order to safeguard the rights of persons under the jurisdiction of
state B. Non-execution is thus clearly not an adequate remedy in the case of human rights
violations. Instead of these highly decentralized modes of enforcement, international
human rights law has gradually evolved its own modes of enforcement and adjudication.
XI. Enforcement
Because human rights obligations are primarily incurred by states vis-à-vis persons
within their jurisdiction rather than between states, the inter-state element in human rights
enforcement has tended to recede. This has served to benefit the modes of enforcement
that emphasize the role of the international community and of individuals.
115
Inter-state procedures do exist for international human rights obligations. In theory,
nothing prevents states from bringing cases before the ICJ for violations of various
international human rights instruments. They have certainly done so in relation to the
Genocide convention in the Bosnia case;
116
or sometimes indirectly as in the cases
brought against the US concerning foreign nationals who had been sentenced to capital
114
Separate Opinion of Judge Weeramantry, n 53.
115
These matters are dealt with in more detail in Section IV on enforcement issue. They are only
mentioned here in relation to the main conceptual issue with which this chapter is concerned,
namely the nature of human rights obligations.
116
ICJ The Application of the Convention on the Prevention and Punishment of the Crime of
Genocide (Bosnia and Herzegovina v. Serbia and Montenegro), case 91, 26 February 2007, p.
108
36
punishment, but following violations of the Vienna Convention on consular relations.
117
Several human rights instruments specifically envisage interstate procedures, whether
before regional courts or treaty bodies.
118
Indeed there have been over half a dozen
interstate procedures before the ECHR.
119
However, it should be said that by and large resort to inter-state human rights procedures
has been very much the exception. Despite all of the universal human rights treaties
envisaging that possibility, including in quite sophisticated terms, it has never been used
at the UN.
120
Moreover, when interstate procedures have been used, it was in very exceptional
circumstances that often provided yet more evidence of the communal interest in the
enforcement of a human rights guarantee, rather than evidence of classical state interest.
This can be seen through the evolution of the notion of locus standi (the Latin expression
that designates standing to appear before a jurisdiction). Under traditional international
law, the only agents allowed to invoke a breach of an obligation are those that were
directly its beneficiaries and are therefore directly affected by its breach. Third parties do
not normally have a general interest in the law’s enforcement such that they could bring a
sort of actio popularis. The ICJ famously considered that Liberia and Ethiopia did not
have standing in the South-West African case because they could not adduce a violation
of a specific interest.
121
Subsequently, the ICJ did mention in passing that certain
obligations might have an erga omnes character.
122
What is a slightly mysterious obiter
dictum in the international law context, however, is fairly mainstream in the human rights
environment,
123
where it is understood that, precisely as a result of the special character
117
Case Concerning the Vienna Convention on Consular Relations (Germany v USA) (Request
for the Indication of Provisional Measures: Order) General List No 104 [1999] ICJ 1
118
For example see American Declaration, n18, art 44; ECHR, nError! Bookmark not defined.,
art24.
119
Greece v. The UK, (1960) Yearbook 2 of the ECHR, 174; Ireland v. The UK,(1978) E.H.C.R
Series A, no.25; , Cyprus v. Turkey, nError! Bookmark not defined.; Denmark, Norway,
Sweden and the Netherlands v. Greece (1984); Denmark, Norway, Sweden, Netherlands and
France v. Turkey (1970) Yearbook XIII.
120
This is in fact a little strange given the level of human rights animosity otherwise evidenced at
the interstate level.
121
South-West Africa Cases (Ethiopia v. South Africa; Liberia v. South Africa) (Second Phase)
[1966] ICJ,<
http://www.icj-cij.org/docket/files/46/4931.pdf>
122
Barcelona Traction, Light and Power Company, Limited (Belgium v. Spain) (Second Phase)
ICJ 5 February 1970, at ¶33. According to that famous case:
an essential distinction should be drawn between the obligations of a State towards the
international community as a whole, and those arising vis-à-vis another State in the field
of diplomatic protection. By their very nature the former are the concern of all States. In
view of the importance of the rights involved, all States can be held to have a legal
interest in their protection; they are obligations erga omnes.
123
In fact the Barcelona Traction quote seemed to be destined for human rights obligations. The
Court went on to say that “Such obligations derive, for example, in contemporary international
law, from the outlawing of … genocide, as also from the principles and rules concerning the
basic rights of the human person, including protection from slavery and racial discrimination.
37
of human rights obligations, each and every state (that is a party to the system) has an
interest in its maintenance. This means that even if the individual victim of a human
rights violation were incapable or even unwilling to bring a case, a third state might.
124
It
also means that states do not have to show that they were directly affected by a violation,
or rather that it is assumed that they are so affected because a violation by one state’s
obligations, even against its own citizens, is an attack in the collective interest of all in
the maintenance and prosperity of the system.
A case in point is Ireland v. the United Kingdom (1978). The Republic of Ireland had
brought an intestate case against the UK following the proclamation of a situation of
emergency in Northern Ireland (which had led to sweeping arrests in a context of
violence). The UK challenged the Court’s jurisdiction, arguing that Ireland had not
shown how its interests had been affected by the UK’s dealings in Ulster. Although here
a case could probably have been made that the Republic of Ireland was affected directly,
the Court found against the UK and argued that:
Unlike international treaties of the classic kind, the Convention comprises
more than mere reciprocal engagements between contracting States. It
creates, over and above a network of mutual, bilateral undertakings,
objective obligations which, in the words of the Preamble, benefit from a
"collective enforcement". By virtue of Article 24 (art. 24), the Convention
allows Contracting States to require the observance of those obligations
without having to justify an interest deriving, for example, from the fact
that a measure they complain of has prejudiced one of their own
nationals.
125
(emphasis added)
This confirmed the erga omnes character of human rights obligations, or at least, as
Theodor Meron pointed out, its erga omnes contractantes character. Another case in point
is the one brought by Sweden, Norway, Denmark and the Netherlands against dictatorial
Greece in the 1970s for violations of the ECHR. Again, none of the former countries had
suffered directly (albeit their nationals may have been marginally involved) but they took
a stand for what they saw as the collective interest of the ECHR, which was judged to be
receivable in Strasbourg.
Some of the corresponding rights of protection have entered into the body of general international
law ...; others are conferred by international instruments of a universal or quasi universal
character”. (emphasis added). The reference to international instruments of a universal character
is clearly in line with the way human rights treaties are generally conceived.
124
The system is not so laterally empowering that any individual can bring a sort of actio
popularis to enforce human rights obligations, but it might evolve this way. Already, the fact that
NGOs can routinely bring cases before the African Commission and Court, for example, suggests
that their petitions are part of a collective interest, which need not only be triggered by individual
complainants. Indeed, in many circumstances, because of security or financial problems, it will
not be open for individuals to petition human rights bodies. It may be that one day in those
regional human rights systems that still know of a dual structure (commission/court), human
rights commissions could bring cases more proactively, notwithstanding the absence of individual
complaints, merely as “general attorneys” or “ombudspersons”.
125
Ireland v. The UK, n119.
38
This erga omnes nature of human rights obligations, incidentally, has implications not
only for formal inter-state mechanisms but more generally for human rights foreign
policy:
126
by making human rights (regardless of where they may be violated) a matter of
concern for all, human rights law profoundly redefines the notion of sovereignty and the
internal affairs of states. Normatively this should make it more difficult for states to claim
that other states that criticize their human rights record are meddling in their internal
affairs (in practice and politically of course, it may not be that difficult to make such
claims).
The special character of human rights obligations also means that conditions for standing
before international jurisdictions will also often be opened broadly temporally. In the
Austria v. Italy case, in which Austria complained to the European Human Rights
Commission about criminal proceedings brought by Italy against six men of German
speaking origin in the South Tyrol, Italy objected to the Commission’s jurisdiction by
saying that Austria was not a party to the Convention at the time of the trials. The
Commission rejects these objections, arguing that Austria had full standing, on the basis
that:
the purpose of the High Contracting Parties in concluding the Convention
was not to concede to each other reciprocal rights and obligations in
pursuance of their individual national interests but to realise the aims and
ideals of the Council of Europe, as expressed in its Statute, and to establish
a common public order of the free democracies of Europe with the object
of safeguarding their common heritage of political traditions, freedom and
the rule of law.
127
Therefore, the fact that Austria was not a party at the time, -which, if it were any other
obligation than a human rights one, would definitely compromise its ability to bring a
case-, does not mean that it cannot complain about Italy’s violations of human rights.
128
However, perhaps most importantly then renovating the terms of inter-state procedures
(which remain marginal anyhow), international enforcement of human rights obligations
relies on a number of sui generis mechanisms that have been pioneered in this field and
are uniquely adapted to it. First are various mechanisms through which states are required
to provide periodic reports to international committees, outlining how they have complied
with their treaty obligations. Although such mechanisms are described in more detail in
section IV, it is interesting that they create a situation where states, in contrast to the
tradition in international law, have to report behaviour outside the occurrence of any
dispute or litigation. This undeniably reflects the higher interest of human rights in
126
For further discussions, see Chapter 4.
127
Austria v. Italy,n10, ¶139.
128
Interestingly, Italy could not have symmetrically brought a case against Austria for violations
occurring at the same time since the latter was not bound by the Convention yet. But since
reciprocity was not the motor of human rights obligations in the first place, Austria could bring an
inter-state petition against Italy, even if Italy could not have done the same.
39
preventing violations, rather than simply adjudicating upon them.
129
Second, and this
must surely count as international human rights’ greatest innovation in terms of
enforcement, many instruments create a right of petition for individuals before
international bodies (or at least the possibility of such a right if it is recognized by state
parties). As is well known, this is inconceivable in the inter-state context where the
possibility of individuals or even significant non-state actors intervening before the ICJ
for example was never seriously entertained. There are obviously fairly stringent
conditions to appearing directly before international bodies (such as the rule of
exhaustion of domestic remedies) but it is remarkably characteristic of international
human rights law that, in last analysis, it allows individuals to directly “sue” their state
internationally.
XI. Conclusion
This chapter has presented some key ways in which international human rights
obligations differ from mainstream international law obligations. The root of the tension
lies in the fact that human rights are intrinsically different in nature and belong to a very
different history and teleology than general international law. This initial difference then
manifests itself in a multiplicity of ways when the human rights project comes in contact
with public international law.
It is testimony to the flexibility of international law that it has adapted itself to this new
subject-matter, in the process changing many rules and practices. In return, human rights
law has also significantly shaped the evolution of public international law, and
contributed to redefine it less as a project of coexistence between equals, than as a
common effort to achieve certain universal minimal standards of treatment of the
individual.
129
It is worth noting that submitting oneself to such a supervision is generally considered
automatic and attached to the substantive treaty, rather than flowing from a specific recognition,
as in the case of most jurisdictional dispute settlement mechanisms. In other words, the
supervisory mechanism goes with the substantive obligations, and states could not join the latter
whilst opting out of the former.