Andrew Serdy*
Many readers of this
journal will, like, the present author, have been eagerly
awaiting the formal entry into force of the Agreement for the
Implementation of the Provisions of the United Nations
Convention on the Law of the Sea of 10 December 1982 Relating to
the Conservation and Management of Straddling Fish Stocks and
Highly Migratory Fish Stocks (the Agreement), and perhaps paying
periodic visits to the admirable website maintained by the UN
Secretariat’s Division of Ocean Affairs and the Law of the Sea
(DOALOS) to check on its disappointingly slow progress. This is
not the place to sing the Agreement’s praises; suffice it to
say that, if UNCLOS is a constitution for the oceans, then the
Agreement is its bill of rights for sustainable fisheries.
The DOALOS site contains
amongst a wealth of other information a status list for the
Agreement, periodically updated (most recently, at the time of
writing, on 24 September 2001), at www.un.org/Depts/los/reference_files/status2001.pdf.
This tells us that the number of ratifications and accessions
achieved as at that date is 29 (the most recent being Costa Rica’s
on 18 June 2001). The top row of the table informs us that the
Agreement is not yet in force, while a footnote to the heading
of the relevant column points out that the Agreement is
expressed to enter into force “30 days after the date of
deposit of the thirtieth instrument of ratification or accession”
(Article 40, paragraph 1). The 29 States listed as having
ratified are: Australia, Bahamas, Barbados, Brazil, Canada, Cook
Islands, Costa Rica, Fiji, Iceland, Iran, Maldives, Mauritius,
Micronesia, Monaco, Namibia, Nauru, New Zealand, Norway, Papua
New Guinea, Russian Federation, St Lucia, Samoa, Senegal,
Seychelles, Solomon Islands, Sri Lanka, Tonga, USA and Uruguay.
On its face, therefore, there is nothing untoward about this
information; the visitor in a hurry is likely to conclude that
we must be patient and wait for one further ratification or
accession to trigger the Agreement’s entry into force.
Sed quaere. Look
more closely, and against the names of two States that are
listed as having signed the Agreement but not ratified it are a
pair of revealing footnotes. One is about Italy:
On 4 June 1999, the
Government of Italy informed the Secretary-General that “Italy
intends to withdraw the instrument of ratification it
deposited on 4 March 1999, in order to proceed subsequently to
complete that formalilty [sic] in conjuction [sic] with all
the States members of the European Union.”
while the other relates to
Luxembourg. On 21 December 2000, the Government of Luxembourg
informed the Secretary-General of the following:
The Permanent Mission of
the Grand Duchy of Luxembourg had indeed received instructions
to deposit the instrument of ratification of the
above-mentioned Agreement with the Secretary-General of the
United Nations; this was done on 5 October 2000. It turned
out, however, that deposit on that date was premature since,
in accordance with decision 98/414/CE of the Council of the
European Union, of 8 June 1998, the instrument was to be
deposited simultaneously with the instruments of ratification
of all States members of the European Union.Accordingly, I
should be grateful if you would note that Luxembourg wishes to
withdraw the instrument of ratification deposited on
5 October 2000. A simultaneous deposit of the instruments
of the Community and of all member States is to take place
subsequently.
Is it correct to exclude
these two purportedly withdrawn ratifications from the tally, or
ought we to have celebrated the Agreement’s entry into force
some months ago? After all, Article 40 speaks not of 30 parties
but of the deposit of 30 ratifications or accessions – and
whatever the current status of Italy and Luxembourg, there have
been 31 instruments of ratification or accession deposited.
However this question is
answered, there is another route by which entry into force may
already have occurred. For a further twist to the tale lies in
store for those who are able to venture (for a price, entry
being by subscription) into another part of the United Nations
website, this one created by the Secretariat’s Treaty Section,
to look at a second status list for this Agreement, at
untreaty.un.org/ENGLISH/bible/englishinternetbible/partI/chapterXXI/treaty8.asp
(visited on 17 October 2001). Here one finds a third footnote,
this time concerning the United Kingdom. Most of it is not
material for present purposes, but one of its paragraphs reports
that:
On 3 December 1999, the
Government of the United Kingdom of Great Britain and Northern
Ireland informed the Secretary-General that the Agreement was
being ratified on behalf of the Pitcairn, Henderson, Ducie and
Oeno Islands, Falkland Islands, South Georgia and South
Sandwich Islands, Bermuda, Turks and Caicos Islands, British
Indian Ocean Territory, British Virgin Islands and Anguilla,
with the following declarations: [declarations not reproduced]
Unlike the other two, this
ratification has not been withdrawn.
This opens up an
intriguing range of possibilities. The correct number of
instruments deposited may be as low as 29 or as high as 32. For
no better reason than that it is simpler, let us take the UK’s
purported ratification first.
The difficulty here is
that the UN Secretariat has not made public its reasoning. One
might have expected to find it in another document on the
treaties database reserved for subscribers, the Secretariat’s
“Summary of Practice of the Secretary-General as Depositary of
Multilateral Treaties” (UN document ST/LEG/7/Rev.1,
untreaty.un.org/ENGLISH/Summary.asp,
visited on 22 October 2001, hereinafter “Secretary-General’s
Practice”). The only unusual feature of the United Kingdom’s
ratification that may have prompted the Secretariat to treat it
differently from those of other States is that it is expressed
to apply only in respect of certain of the UK’s dependent
territories. Conceivably this could have brought into play
Article 29 of the Vienna Convention on the Law of Treaties (“Unless
a different intention appears from the treaty or is otherwise
established, a treaty is binding upon each party in respect of
its entire territory.”) The applicability of Article 29 is,
however, far from clear-cut. Given that the “or” makes “appears
from the treaty” and “is otherwise established”
alternatives to each other, it is not necessary for the “establishing”
to be done within the immediate context of the text of the
treaty, such as from its travaux préparatoires. In other words,
Article 29 is expressed in such a way as not to rule out a party
unilaterally “establishing” its consent to be bound in
respect of less than the whole of its territory where the text
of the treaty is silent on the matter. And it is at least
arguable that there is nothing in the nature of the Agreement
itself that precludes this. Moreover, the mere fact that the
territories in question are dependent is surely irrelevant for
the purposes of Article 29.
Curiously, however, the
Secretary-General’s Practice appears to dictate a different
course from the one the Secretariat has taken here. At paragraph
277 it is stated that, “[w]hen neither the nature of the
treaty nor other special circumstances (e.g., the fact that the
treaty is the constitutive act of an international organization)
mandate the non-acceptance of the instrument containing a
declaration as to the limited application or non-application of
a treaty to Territories, the Secretary-General has…accepted
instruments containing reservations as to the limited
application or non-application to Territories, leaving it to the
other parties to draw the legal consequences of such declaration
that they may see fit.” Indeed paragraph 285 appears expressly
to endorse the reasoning above:
In all of the above-mentioned
circumstances…the Secretary-General, as depositary, has felt
that he was not [able] to pass judgement on the admissibility of
such declarations and he has duly circulated them…This
position would not appear inconsistent with the provisions of
article 29 of the Vienna Convention on the Law of Treaties since
it may be considered that the constant practice of certain
States (which still comprise “non-metropolitan” Territories)
in respect of territorial application and the general absence of
objections to such practices have “established a different
intention” within the meaning of article 29.
If the United Kingdom’s
ratification is effective, this alone raises the tally to 30 and
the Agreement will have been in force since 18 July 2001 even if
the deposit of Italy’s and Luxembourg’s ratifications are
disregarded. If not, then it is possible to agree with the
Secretariat that what the UK deposited was not an “instrument
of ratification”, even though that was what it purported to
be. So we are faced with a straight choice between accepting or
not the United Kingdom's ratification in respect of part of its
territory and adjusting the tally accordingly.
For his own part, the
author is at a loss to understand why the Secretariat did not in
this instance follow its own stated policy, which seems
perfectly sound, and respectfully suggests that it ought to have
done so. It may, however, be conceded that a further
ratification by the United Kingdom in respect of the remainder of
its territory should not increase the tally for the purposes of
Article 40 of the Agreement but instead be taken as merely
enlarging the territorial scope of the original
ratification.
The position in respect of
Italy and Luxembourg is more complex. Is there, then, any basis
on which the deposit of their ratifications can be disregarded
for the purposes of Article 40? This is really two questions:
(1) Can a ratification or accession of a multilateral treaty be
withdrawn before its entry into force generally? (2) If so, must
the deposit of the relevant instrument be treated for all
purposes as a nullity, i.e. as though it had never occurred, or
does the fact of its occurrence nonetheless continue to have
consequences, in terms of being counted towards a number of
expressions of consent to be bound that triggers entry into
force of the treaty for other States that maintain their consent
or subsequently express it? That is, in the instant case, can
the Agreement enter into force with fewer than 30 parties?
As to the first of these
questions, if the treaty itself provides an answer that is of
course what will be decisive. There is, however, no provision
in the Agreement that sheds light on it one way or the
other, so it is to the Vienna Convention on the Law of Treaties
that we must look for guidance. The Vienna Convention allows for
the expression of consent of a State to be bound to be vitiated
if it was procured by fraud (Article 49), corruption or coercion
of the representative of a State (Articles 50 and 51
respectively) or coercion of the State itself (Article 52), but
not merely if it is the result of an administrative error (see
the very restrictive rules on error generally at Article 48), as
the footnotes to the Status List indicate was the case for both
Italy and Luxembourg.
Although this points
strongly to a conclusion that both deposits ought to have been
counted, this time a clue is available as to the UN Secretariat’s
reasoning to the contrary. It lies in the Secretary-General’s
Practice. At paragraph 157, following the observation that the
question of withdrawals before entry into force was not
discussed at the Diplomatic Conference at which the Vienna
Convention was negotiated, the reason is given: “[T]he
practice of the Secretary-General has been to allow such a
withdrawal until the entry into force of the treaty, on the
understanding that, until that time, States are not definitely
bound by the treaty”. Paragraph 158 gives two precedents of
States being permitted to withdraw ratifications before the
relevant treaty entered into force according to its terms. It is
submitted, however, that these precedents are not persuasive,
and not merely because international law lacks a principle of
stare decisis. Rather, the precedents are both distinguishable
as instances of a State not withdrawing altogether its consent
to be bound, but simply substituting an expression of consent to
be bound subject to a reservation allowed by the treaty in
question for the earlier expression without reservation. As
paragraph 158 makes clear, this was necessary in order to comply
with the rule that reservations must be made at the time of
deposit of the instrument. Thus the situation of the Agreement
is not analogous, and that may leave the Secretariat’s action
resting on nothing more than a policy choice.
Is this a good thing? As
its title implies, the Secretary-General’s Practice does not
purport to have the force of law and the circulation without
comment of instruments deposited as quoted above bespeaks an
entirely proper desire on the part of the Secretariat not to
exceed its mandate. Yet, in the course of administering the
treaties of which the Secretary-General is depositary, the
Secretariat will from time to time, as on this occasion, be
unable to avoid making choices between two or more views of the
law of treaties. It is quite legitimate for the Secretariat to
make this choice on the basis of legal policy and, desirable
though it would be for the Secretariat to state that basis
openly, it is understandable that it does not do so more often.
On the question of whether
withdrawal of an expression to consent to be bound by a treaty
before entry into force of the treaty ought for policy reasons
to be permissible, there seem to be two competing
considerations: deference to States’ wishes, and the
undesirability of detracting from the reliance that may be
placed on the unqualified expression of a State’s consent to
be bound if its peremptory withdrawal is permitted. The former
is not without merit. If the aim is to encourage early entry
into force of multilateral treaties, then paradoxically this
should be assisted by a rule that leaves them revocable for a
time, as States are likely to be less hesitant about depositing
them in the first place. On the other hand, as long as the
treaty contains a provision allowing States to withdraw from it
(as is the case here under Article 46 of the Agreement, on a
year’s notice), the added incentive is marginal at best.
Yet the Secretariat does
not, as far as can be discerned, seem to have been following a
policy of deference. If it had, then it would have accepted not
only Italy’s and Luxembourg’s withdrawals but also the UK’s
ratification, particularly given that no State has called the
latter into question on treaty law grounds. Rather, the only
policy that can consistently with the evidence be imputed to the
Secretariat is one of resolving any doubt about a ratification
or accession in favour of the status quo, which necessarily
retards entry into force of the treaty. Ordinarily one would
expect those States that have already ratified or acceded to the
Agreement to wish others to do so in numbers sufficient to bring
it into force sooner rather than later. But it is not necessary
to presume this. The resolutions of successive sessions of the
General Assembly in the years since the Agreement was opened for
signature, calling as they do for precisely that (see paragraph
8 of Resolution 55/8, adopted on 30 October 2000, which has
almost identical antecedents in the resolutions of previous
sessions since 1995 listed in its first preambular paragraph),
leave no room for doubt that such a policy would be contrary to
the wishes of UN Member States as a whole.
Ultimately, however,
nothing in the relevant paragraphs of the Secretary-General’s
Practice goes to the question whether an instrument has actually
been deposited, much less whether “deposited” is required to
be interpreted as “deposited and not subsequently withdrawn”.
Paragraph 14 of the Secretary-General’s Practice lists “the
provisions of the treaty” uppermost among the factors guiding
the Secretary-General in the performance of his duties as
depositary. Accordingly, it is submitted with some hesitation,
the correct course would have been to follow the terms of
Article 40 paragraph 1 and count Italy’s and Luxembourg’s
ratifications at least for the purpose of triggering the entry
into force of the Agreement; this is consonant with the
commonsense conception of ratification as an event rather than a
process.
On this view, there is no
longer room for doubt as to whether the Agreement has come into
force but only whether it did so on 4 November 2000 – 30 days
after Luxembourg’s ratification on 5 October 2000 – or, if
the foregoing argument as to the effectiveness of the United
Kingdom’s ratification is incorrect, on 18 May 2001, 30 days
after New Zealand’s ratification on 18 April 2001, either with
28 parties or with 30. Alternatively, if the UK ratification is
effective, the question does not arise because there are now a
minimum of 30 parties whatever view one takes of the point
relating to Italy and Luxembourg.
The author’s hesitation
in reaching the above conclusion is because this rule,
preferable though it is on balance, itself gives rise to an
anomaly if it is not followed all the way through to
disallowance of withdrawals before entry into force. This
emerges from the Secretariat’s implicit answer in paragraph
159 of the Secretary-General’s Practice to question (2) above.
For here it is stated that, if by the terms of the treaty entry
into force occurs a given number of days after a certain number
of ratifications or accessions is reached, then, although
withdrawal of its expression of consent by a State in the
intervening period will be effective to prevent it from becoming
bound, that will not stop the treaty from coming into force for
the remaining parties on the expiry of the requisite number of
days since the triggering ratification or accession. Thus the
Secretariat appears to accept that it would be possible for this
Agreement to enter into force with fewer than 30 parties, but
only if Italy had withdrawn its ratification between 5 October
and 4 November 2000 (or, if Article 29 makes the UK’s
ratification ineffective, Italy and Luxembourg had both
withdrawn theirs between 18 April and 18 May 2001). This
condition creates a highly artificial solution not contemplated
by the terms of the Agreement and has the peculiar effect of
binding other ratifying and acceding States as against each
other while Italy and Luxembourg themselves escape these
obligations.
Indeed under a somewhat
different scenario the peculiarity is even more pointed. Suppose
the UK ratification is effective, Italy had not withdrawn its
ratification but Luxembourg had withdrawn its within 30 days of
depositing it. Then Luxembourg’s ratification would have been
the 30th by the Secretariat’s count. Generalising, in any
situation where only one more ratification or accession is
required to trigger it, a State can precipitate entry into force
without any intention of itself being bound, simply by
withdrawing its ratification or accession before entry into
force actually takes place. It is submitted that a rule allowing
such a perverse outcome cannot be good legal policy. On the
other hand, if expressions of consent to be bound are revocable
at any time before entry into force, which strikes the author as
more defensible in terms of both principle and the Agreement if
they are to be revocable at all, the peculiarity is unaffected.
The only way of avoiding the problem seems to be to disallow
withdrawals before entry into force.
Where does this leave us?
The table below sets out the possibilities.
