By Hannah Chang
The term “legally binding” has become a touchstone of sorts in international climate policy. The Copenhagen Accord taken note of by the fifteenth Conference of Parties (COP) under the United National Framework Convention on Climate Change in December 2009 is not legally binding. Heads of state and activists alike call for a legally binding agreement to be made at the next COP in Mexico at the end of 2010. Meanwhile, China and India have vociferously emphasized that the mitigation actions they submit under the Copenhagen Accord are “voluntary in nature and not in the nature of legally binding commitments.” But what does “legally binding” even mean?
Popular notions of a binding international agreement mistakenly assume that the binding nature of the obligation translates into actual implementation by states and therefore equates with effectiveness of the agreement. But in fact, there is no necessary connection between the legally binding nature of an agreement and its effectiveness.
I. The gap between internationally binding obligations and domestic action
U.S. law distinguishes between self-executing and non-self-executing international agreements. A self-executing treaty immediately takes effect as law of the United States upon ratification – supreme over state law and judicially enforceable – and does not require new legislation to enable the United States to carry out its international obligations. A non-self-executing treaty, on the other hand, requires the enactment of domestic legislation to give effect to its terms. Notably, international agreements are considered “binding” on parties when the agreement enters into force, regardless of the nature of execution at the domestic level. If the U.S. does not enact domestic legislation implementing the terms of a non-self-executing international agreement, then, the international obligation remains no less binding but the U.S. is simply in default of its international obligation.
This bifurcation between domestic lawmaking and international lawmaking in the form of non-self-execution is relatively unique to the U.S. legal system, but in a broader sense reflects a general reality of the international system: international laws take effect only through the actions of individual sovereign states. States can enter any number of “legally binding” international agreements, but it is the domestic actions of each state – in the form of enacted laws, policies, and regulations – that are necessary to meet those international obligations. In essence, the international classification of an agreement as “legally binding” or merely politically binding, has little to do with what actually must happen at the domestic level to make the agreement effective.
II. The reality of international enforcement and compliance
Confronted with the fact that state actions are what matter in achieving the goals of international agreements, the question arises: but can’t a “legally binding” agreement force states to take the necessary action by wielding sticks to ensure compliance?
The answer may be seen in compliance with the Kyoto Protocol, a legally binding international agreement that commits countries to specific emissions reduction targets. In a 2008 assessment of the progress of Annex I countries to the Protocol, some twenty countries, including Canada, New Zealand, and Japan, were not on track to meet their Kyoto obligations. This, not to mention that the U.S., the world’s largest greenhouse gas emitter at the time of the Protocol’s signing, was not even a party, and China, currently the world’s largest emitter, has no emissions reduction obligations under the Protocol.
The Kyoto Protocol’s Compliance Mechanism has been hailed as the “most robust ever adopted” for a multilateral environmental agreement. It consists of a facilitative branch that provides advice and assistance to parties to promote compliance and prevent non-compliance, as well as an enforcement branch that is empowered to take more “hard” enforcement actions. The enforcement branch determines three different forms of non-compliance – with the emission targets, with the methodological and reporting requirements, and with the eligibility requirements for participation in the Protocol’s flexibility mechanisms – and requires the non-complying state to take certain actions depending on the type of non-compliance.
In the case of non-compliance with emission targets, the offending Annex I party must make up the difference between its emissions and its target, including an additional 30% deduction, in the next compliance period. In other words, for every ton of emissions by which a party exceeds its Kyoto target, its emissions allocation for the subsequent compliance period will be lowered by an additional 1.3 tons as a penalty. Whether the specter of this penalty has any effect in deterring non-compliance is questionable, however.
For one thing, like an individual indefinitely passing on debt into the future, a country could simply borrow from one commitment period to the next and never meet its target. If the international community failed to “force” the country to meet its target in the first compliance period, it’s unclear how it might force the country to meet its target with the included penalty in the subsequent period. Second, each party negotiates its own target for each commitment period, so a non-complying party could simply negotiate a higher target in a second period to accommodate the 30% penalty, effectively negating any impact of the enforcement mechanism.
This second consideration raises an important point. States themselves have to consent to be enforced upon – just as a state can negotiate a higher target for itself, it can also refuse to sign onto an agreement. “Self-inflicted punishment” is bound to be lax while a too-stringent enforcement framework will simply discourage countries from signing on . . . unless there is strong incentive to participate.  The World Trade Organization is touted as an effective treaty arrangement because it simultaneously allows for the credible threat of non-self-inflicted punishment (that is, punishment inflicted by another state or international body) while also offering reduced trade barriers as a strong incentive for participation. In the face of such substantial benefits to participation, countries are willing to sign on to the prospect of stringent enforcement.
The same cannot be said of the Kyoto Protocol. Under Article 18 of the Kyoto Protocol, any compliance mechanism “entailing binding consequences” must be approved by amendment requiring ratification by at least three-quarters of the Kyoto parties, and even then, would be binding only on the parties that ratified it. The parties to the Protocol have not managed to pass such an amendment. So actually, idealized though it may be by those calling for a binding climate agreement, the Kyoto Protocol’s compliance mechanism itself is a political agreement that has no binding legal status.
So, in reverse: 1) international enforcement mechanisms are limited by the need to have states consent to enforcement, and 2) state actions are what make an international agreement effective. These realities do not seem to bode well for a vision of an international scheme in which all countries would sign on to the emissions targets necessary to prevent a dangerous rise in global temperatures and actually comply with their targets.
That said, states do not act at will within a vacuum. They are motivated to sign onto agreements and to take action to comply with those agreements for any number of reasons relating to self-interest, public pressure, reputation, horse-trading – in effect, political reasons. The “legally binding” nature of the obligation is simply not likely to be a significant one of those reasons.
 Letter from Prime Minister of India Manmohan Singh to the Prime Minister of Denmark Lars Rasmussen, dated Jan. 19, 2010.
 Xueman Wang & Glenn Wiser, The Implementation and Compliance Regimes under the Climate Change Convention and its Kyoto Protocol, 11 RECIEL 181, 198 (2002).
 Scott Barrett, Climate Treaties and the Imperative of Enforcement, 24 Oxford Rev. Econ. Pol 239, 246 (2008).
 Id. at 244.