What to Look for in any U.S. Withdrawal from the Paris Agreement

by Duncan Hollis

President Trump has indicated that he will announce a decision on future U.S. participation in the Paris Agreement later today at 3 pm.  Reports suggest that he has already made up his mind to withdraw.  That decision is likely to receive extensive attention (not to mention criticism) on the merits.  And certainly that attention is warranted.  But I believe an equally important issue will be how the Trump Administration pursues its withdrawal.

Contrary to popular opinion (and this erroneous NY Times Q&A), the Paris Agreement was never intended to be non-binding.  It is, on its face, pretty clearly a treaty in the international law sense of that term (see the standard definition in Art. 2(1)(a) of the 1969 Vienna Convention on the Law Treaties (VCLT)).  True, one key provision of the Paris Agreement (Article 4) contains language that does not evidence an intention to create legal rights or obligations (and the negotiation of which almost blew up the original deal).  But the rest of the agreement was clearly intended to create a treaty and the language used manifests such intentions.  For confirmation, one only has to look to the U.N. Treaty Office (which is home to some of the world’s leading experts on treaties) and note how it has always regarded the Paris Agreement as a treaty.

The United States formally joined the Paris Agreement on November 4, 2016, following its acceptance of that treaty on September 3, 2016.  It is true that the United States did so without seeking the U.S. Senate’s advice and consent under Art. 2, cl. 2, section 2, nor did Congress specifically authorize U.S. participation as it did for treaties like NAFTA or the WTO Agreement.  But U.S. treaty law and practice has long accommodated other means for the United States to enter into treaties in the international law sense, including through the President’s sole executive powers or where prior Congressional authorization supports U.S. participation.  In the case of Paris, the precise grounds for U.S. acceptance are contested (see Dan Bodansky and Peter Spiro’s impressive take on these issues here).

As far as international law is concerned, however, there is little question that the United States is currently bound by its acceptance.  The law of treaties is most famous for the foundational principle pacta sunt servanda, or as VCLT Art. 26 puts it, “Every treaty in force is binding upon the parties to it and must be performed by them in good faith.”  And for those less familiar with the VCLT, it is important to note that although the United States never joined the “treaty on treaties” every Administration since Richard Nixon’s has regarded almost all of its provisions as customary international law (the exceptions being provisions on signature and consultations in the event of breach). Thus, to withdraw or otherwise end its obligations under the Paris Agreement, the United States will have to look to the law of treaties.  Indeed, VCLT Article 42 provides that a treaty’s validity or a State’s consent can only be impeached through the VCLT’s application and, more pertinently, “[t]he termination of a treaty, its denunciation or the withdrawal of a party, may take place only as a result of the application of the provisions of the treaty or of the present Convention.”

So, how can the United States get out from the Paris Agreement?  I predict the Trump Administration will invoke one of four possible avenues for its exit later today.

(1) Invalidating U.S. Acceptance Due to a “Manifest Violation” of U.S. Domestic Processes? First, the United States could argue that the Obama Administration’s acceptance was never valid to begin with.  VCLT Art. 46 provides in relevant part

1. A State may not invoke the fact that its consent to be bound by a treaty has been expressed in violation of a provision of its internal law regarding competence to conclude treaties as invalidating its consent unless that violation was manifest and concerned a rule of its internal law of fundamental importance.

2. A violation is manifest if it would be objectively evident to any State con ducting itself in the matter in accordance with normal practice and in good faith.

Thus, the law of treaties allows a State to avoid treaty obligations by invoking a manifest violation of its domestic processes for treaty-making. The negative framing of this option suggests, however, that U.S. efforts to invoke it are not likely to be convincing.  Past attempts by States to invoke flaws in their domestic approval processes have been dismissed (see, most notably, paras 265-68 of the International Court of Justice’s opinion in the case of Land and Maritime Boundary between Cameroon and Nigeria).  What’s more, the United States has a long history of using at least four different methods for joining international treaties. As such, it would be hard to sustain a case that the Obama Administration’s use of the Executive’s constitutional authority alongside existing regulations and prior Congressional authorizations was nonetheless a “manifest” violation of extant U.S. processes for concluding treaty commitments.

(2) Withdrawal under the Paris Agreement’s Own Terms:  Second, under VCLT Article 54, “The termination of a treaty or the withdrawal of a party may take place: (a) In conformity with the provisions of the treaty.”  Where the treaty is silent on termination, the VCLT lays out certain default rules, but these are not relevant here since, in the case of the Paris Agreement, Article 28 lays out the conditions for a State’s withdrawal:

1. At any time after three years from the date on which this Agreement has entered into force for a Party, that Party may withdraw from this Agreement by giving written notification to the Depositary.

2. Any such withdrawal shall take effect upon expiry of one year from the date of receipt by the Depositary of the notification of withdrawal, or on such later date as may be specified in the notification of withdrawal.

3. Any Party that withdraws from the Convention shall be considered as also having withdrawn from this Agreement.

As noted above, the Paris Agreement entered into force for the United States on Nov. 4, 2016.  As such, the United States cannot give formal written notice of its withdrawal until more than three years have passed from the date of the Agreement’s entry into force, or Nov. 4, 2019.  And, even then, the actual legal effect of the withdrawal will not occur for an additional twelve months (Nov. 2020) barring a (very unlikely) U.S. willingness to specify a later date.

Does this mean the Trump Administration cannot withdraw tomorrow under the Paris Agreement’s terms?  I think the answer is, legally speaking, quite clearly yes.  On the other hand, there is probably nothing precluding the Trump Administration from announcing its “intention” to give the proper notifications at the earliest possible date.  Yet, the United States would remain bound to perform the Paris Agreements obligations (like Art. 16(7)’s information sharing commitment) until such time as it is no longer bound by them.

(3) Withdrawal under the Terms of the UNFCCC:  Paris Agreement Article 28(3) offers a third exit route.  It provides that if the United States exits the “Convention” (that is, the UN Framework Convention on Climate Change or “UNFCCC”) it can also bring to an end its obligations under the Paris Agreement.  The UNFCCC exit clause, Article 25, is similar to the Paris Agreement’s terms:


1. At any time after three years from the date on which the Convention has entered into force for a Party, that Party may withdraw from the Convention by giving written notification to the Depositary.

2. Any such withdrawal shall take effect upon expiry of one year from the date of receipt by the Depositary of the notification of withdrawal, or on such later date as may be specified in the notification of withdrawal.

3. Any Party that withdraws from the Convention shall be considered as also having withdrawn from any protocol to which it is a Party.

The difference here is that the three year period has long since passed (the United States has been bound by the UNFCCC since it entered into force on March 21, 1994).  Thus, if the Trump Administration is willing to walk away from the entire climate change framework, with all the extra signaling that would accompany such a move, it could effectuate its exit after the passage of only one year from its withdrawal notification (i.e., by June 2, 2018).  As with invoking Article 28 of the Paris Agreement, however, the United States would still be legally bound by both the UNFCCC and the Paris Agreement itself in the interim. [Update: André de Hoogh helpfully notes that paragraph (3) here only refers to Protocols (e.g., the Kyoto Protocol) and not “Agreements” like the Paris Agreement, challenging its availability in this case.  It’s an interesting argument that requires further study although my initial reaction is that I would find it unlikely that a Party could withdraw from the UNFCCC on one year’s notice, which the United States can clearly do, but still be a party to a clearly subsidiary agreement for 3 more years].

(4) A Breach of the Paris Agreement:  But if we’ve learned anything from the last several months of the Trump Administration it is its willingness to buck traditional conventions.  Thus, we might (and probably should) anticipate a fourth option tomorrow – a simple statement of U.S. withdrawal that purports to take effect immediately.  Now, if the United States does this, I believe it would most likely be in breach of its treaty commitments under the Paris Agreement itself.  Unfortunately, the VCLT is notorious for having limited remedies in the event of such a breach.  For starters, the VCLT only provides remedies for material breaches, which might lead the Trump Administration to question if an immediate withdrawal rises to such a level.  But, VCLT Art. 60(3)(a) makes clear that a material breach includes “a repudiation of the treaty not sanctioned by the present Convention.”   In such cases, however, the only remedies the VCLT affords is to give other affected parties the right to collectively or individually suspend or terminate their obligations under the relevant agreement.  Given that this whole current crisis turns on the fact that almost every other country in the world wants the Paris Agreement to continue rather than cease, this is a profoundly unsatisfactory response option.

There are, however, at least two other potential remedies States parties to the Paris Agreement might employ if they believe the Trump Administration has breached Article 28 (or UNFCCC Art. 25):  retorsion and countermeasures.  Under the so-called “law of State responsibility” States are always free to respond to internationally wrongful acts (a category that includes treaty breaches) with lawful, but unfriendly responses — i.e., cutting off diplomatic relations, reducing or eliminating financial assistance, suspending information sharing, etc.  The idea is to sanction the breaching party by ceasing or forgoing certain behavior even if such behavior was not legally required in the first place.  Countermeasures, in contrast, involve otherwise unlawful behavior by the victim State(s) that is legitimated by the prior unlawful acts of the breaching State.  For example, even though international law may prohibit States from taking certain trade sanctions, where there is a prior breach, that otherwise unlawful behavior may become lawful.  Now, there are a number of conditions on the use of counter-measures (for example, they must be reversible, proportional and temporary) and there are circumstances like force majeure that may preclude its very application. Still, countermeasures could be a powerful tool for States, whether acting individually or collectively, to respond if they interpret any Trump Administration withdrawal from the Paris Agreement as a violation of the United States’ treaty commitments.

In sum, the mechanics of treaty withdrawal are complex.  There are additional arguments and counter-arguments that might arise, but which I’ve not addressed out of concerns with time and space (for example, could the United States invoke other grounds for terminating its Paris Agreement commitments like a fundamental change of circumstances, impossibility or necessity). For now, I just wanted to lay out the most likely options if the predicted withdrawal comes to pass.  As always, I’m cognizant that many readers have deep expertise in the law of treaties.  Thus, I’d welcome further thoughts or comments on this issue.  I’m sure there are interesting additional angles worth considering in what will be the Trump Administration’s most serious foray into the treaty realm to date.



11 Responses

  1. A few comments:

    First, with respect to option (3), article 25 UNFCCC does not, on its face, appear to be applicable. Paragraph 3 of that provision stipulates that a withdrawal from the Convention will also be considered withdrawal from any protocol. However, the Paris Agreement has precisely not been styled a protocol, as I understand to allow the Obama Administration to argue that it would not be subject to the procedure that requires the Senate’s advice and consent (although I do not have reference handy for this). As a result, this would not allow withdrawal of the Paris Agreement on a one year’s notice.

    Second, following up on the previous comment, this entails that option (2) is the only viable one in terms of withdrawal according to the terms of the treaty concerned. Paragraph 1 of article 28 Paris Agreement stipulates that withdrawal will be possible three years after entry into force for a particular party, with a one year’s notice. However, paragraph 3 of that provision provides that withdrawal from the Convention also entails withdrawal from the Paris Agreement. The Convention contains the same rule as the Agreement, namely withdrawal three years after entry into force with a one year’s notice. As the Convention has already been in force for almost 25 years for the US, this might suggest the possibility to withdraw with a one year’s notice from both Convention and Agreement.

    Third, this raises the question of the relationship between paragraphs 1 and 3, since application of the former would yield a different result (withdrawal taking effect four years from 4 November 2016, assuming notification after three years) than that of the latter (withdrawal taking effect one year from the moment of notification). It makes sense to assume that the parties negotiating the Agreement were aware of the fact that the three years period after entry into force of the Convention had already lapsed for most if not all of the parties concerned. Nevertheless, they did not enter a caveat in paragraph 3, for instance by including the phrase ‘notwithstanding paragraph 1 of this provision’.

    In conclusion, this suggests that paragraph 1 ought to be considered a lex specialis in relation to paragraph 3, and that withdrawal from the Paris Agreement according to its terms will only be possible in accordance with paragraph 1 of article 28.

  2. I’m not sure I follow the above comment — apologies if I’m reading too hastily. As to the second point, why doesn’t Article 28 paragraph 1 (specially) address the circumstance in which a party withdraws from the Agreement, and paragraph 3 (specially) address the effect on party status under the Agreement for a party that *also* withdraws from the Convention? As to the first, why would not Article 28 paragraph 3 at least strongly (and specially) imply that the Agreement views itself as the kind of pact affected, per its own terms, by any withdrawal from the Convention, regardless of the general consequences sketched by the latter?

  3. As noted in an update by the author, para 3 of Art 28 reflects the effect on party status under the agreement, as a subsidiary agreement to the Convention. Read in the light of Art 20 of the agreement on entry into force, parties to the Convention alone can be parties to the agreement.

  4. Correction: Art 20 on signature, ratification, accession and so on.

  5. If the response was to me, my suggestion was that 28(3) of the Agreement means that withdrawal from the Convention, per the Convention’s terms for withdrawal from the Convention itself (that is, Article 25(1)-(2)), accomplishes withdrawal from the Agreement. That reading would not depend on Article 25(3) of the Convention, which operates independently to discharge obligations under any protocol … which need not, accordingly, have a provision comparable to 28(3) of the Agreement. But maybe that’s wrong.

  6. Agreed. That reading does not depend on Art 25(3) of the Convention.

    As for whether the agreement is a “protocol” that falls under Art 28(3) of the Convention, there is considerable debate. Less than the language of “protocol”, “treaty” or other, Obama negotiated for specific provisions in the agreement to remain non-binding to streamline ratification in US law, but this does not affect its status as a treaty (or protocol or agreement…)under international law. More relevant, perhaps, is that the parties disagreed as to whether the agreement should indeed be subordinate to the Convention, and the language choice was determine in part for that reason. See: Dan Bodansky: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2735252

  7. http://unfccc.int/paris_agreement/items/9485.php

    No communication from the US here or Press Release from the entity.

    Anyone seen the US formal notification of withdrawal?


  8. I’d be interested to see a post on the constitutionality of California, New York and Washington’s commitment to follow the Paris Agreement.

  9. non liquet – voluntary action by states to comply with said treaty has been seen since at least the Kyoto protocol. If our federalism gives them the space, they can do it. They are not treaty parties nor have a treaty obligation but are acting consistent with what that agreement requires parties to do. Also by cities like Burlington, Vermont.

Trackbacks and Pingbacks

  1. […] President Trump has indicated that he will announce a decision on future U.S. participation in the Paris Agreement later today at 3 pm.  Reports suggest that he has already made up his mind to withdraw.  That decision is likely to receive extensive attention (not to mention criticism) on the merits.  And certainly that attention is warranted.  But I believe an equally important issue will be how the Trump Administration pursues its withdrawal… lire […]

  2. […] ongedaan worden gemaakt door Trumps opvolger. Er zijn grofweg drie alternatieven, die onder andere door Hollis in dit blog verder worden verkend. Ten eerste kan Trump betogen dat omdat Obama het verdrag niet langs de Senaat heeft laten gaan, de […]