Multi-layered differentiation in the climate regime: the gradual path from Rio to Paris

ABSTRACT

According to a commonly held view, states have fundamentally re-organized the differentiation between developed and developing countries in the climate regime in the 2015 Paris Agreement. In this view, the Paris Agreement replaces the ‘rigid’, ‘static’, and ‘dichotomous’ system of differentiation based on Annexes I and II to UN Framework Convention on Climate Change with a more ‘flexible’, ‘dynamic’, and ‘subtle’ solution. I argue that this view is incomplete. In fact, the early climate regime included additional layers of differentiation that go beyond the binary distinction between Annex and non-Annex parties. Through a discussion of three episodes in which states adjusted the system to the ‘special circumstances’ of regime members, I show how informalization and individualization – two hallmarks of differentiation in the Paris Agreement – had become central well before COP-21 and that the international climate regime thus developed a lot more gradually than is often assumed.

KEYWORDS:

• Climate politics
• differential treatment
• Kyoto Protocol
• Paris Agreement
• UNFCCC

1. Introduction

Differential treatment is a key element of many international environmental agreements. In international climate change law, it is institutionalized as the principle of ‘common but differentiated responsibilities and respective capabilities’ (CBDR-RC). In line with this principle, the 1992 UN Framework Convention on Climate Change asked countries listed in Annex I to take the lead in the fight against global warming. This annex included the OECD members at the time plus the Eastern European countries with ‘economies in transition’ (EITs). State parties confirmed and further specified this list-based approach in the 1995 Berlin Mandate and in the 1997 Kyoto Protocol. Importantly, the latter specified quantified emission limitations for Annex I countries, but not for other parties.

In 2015, however, the Paris Agreement set the international climate regime on a new institutional path. Unlike the Kyoto Protocol, the ‘new logic of international climate politics’ (Falkner 2016) was no longer based on internationally agreed targets and timetables for industrialized countries. Instead, the Paris Agreement relies on a combination of two elements. First, it defines a collective temperature goal: Article 2 mentions the ambition of ‘holding the increase in the global average temperature to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C above pre-industrial levels’. Second, Article 4 requires all parties – including developing countries – to determine their own nationally determined contributions (NDCs), to regularly update these contributions, and to increase the ambition of their NDCs over time (see also Allan et al. 2021, pp. 3–6).

Like its predecessor, the Paris Agreement contains provisions for differential treatment. For example, it asks each party to determine its NDC ‘reflecting its common but differentiated responsibilities and respective capabilities, in the light of different national circumstances’ (Article 4.3). It further states that ‘developed country Parties should continue taking the lead by taking economy-wide absolute emissions reduction targets’ (Articles 4.4). Nevertheless, a widely shared perception is that the ‘new era in international climate politics’ (Falkner 2016, p. 1108) has been marked by a significant erosion of differential treatment. This view is largely based on the observation that Annex I no longer serves as the primary basis for distinguishing between parties that hold special obligations and parties that do not. As a result, the ‘dichotomous interpretation’ of CBDR-RC is said to have given way to the ‘subtle differentiation’ of the Paris Agreement (Pauw et al. 2019; see also Bodansky 2016, Savaresi 2016, Voigt and Ferreira 2016).

In the remainder of this article, I argue that the conventional view, which suggests a fundamental rupture between ‘Kyoto’ and ‘Paris’, is incomplete. I develop two specific claims. First, I show that the UN Framework Convention consists of multiple layers of differentiation that go well beyond the distinction between Annex I-countries and non-Annex I countries. Second, I argue that these additional layers of differentiation provided the foundation for key changes associated with the shift from the Kyoto Protocol to the Paris Agreement. Notably, they introduced an individualization as well as an informalization of differentiated commitments among state parties early on. In sum, the shift from the list-based approach to informal and individualized differentiation in the Paris Agreement appears less radical than it is commonly portrayed. My observations contribute to a strand of literature that rejects the narrative of abrupt institutional change based on diplomatic breakthroughs and emphasizes the gradual development of the international climate regime instead (Maljean-Dubois 2016, Allan 2019, Allan et al. 2021, Depledge 2022).

The argument proceeds in three steps. In the following section, I summarize how the evolution of differential treatment norms in the international climate regime is conventionally narrated. A further section then identifies and describes additional layers of differentiation in the international climate regime – those that go beyond Annex I – and the dynamics to which these layers have given rise. In the third step, I reconstruct three episodes that show how individualized and informal differentiation not only predated, but also contributed to the developments leading up to the Paris Agreement. The material on which I draw for these reconstructions includes the official documents from the UNFCCC process, negotiation reports provided by the Earth Negotiations Bulletin, and secondary literature. The final section concludes.

2. Differential treatment in the international climate regime

Reading descriptions of the international climate regime after 2015 can give the impression that there are two international climate regimes that share little in common. According to such descriptions, international climate policy was organized around the UN Framework Convention and the Kyoto Protocol from 1992 to 2012. During this period, climate policy-making was commonly – though perhaps falsely (Depledge 2022) – defined by a ‘top-down’ approach based on ‘targets and timetables’. Here, the notion of a top-down approach refers to the fact that states sought agreement on common targets at international conferences and were then asked to implement and enforce measures to achieve those targets at home. Moreover, the original regime was characterized by a strong form of differentiation based on the distinction between Annex I and non-Annex I parties. In the Kyoto Protocol, Annex I parties thus agreed to collectively reduce their greenhouse gas emissions during the period from 2008 to 2012 by five per cent compared to 1990. Developing countries, by contrast, remained exempt from comparable commitments. This stark differentiation between developed and developing countries had been confirmed at the first Conference of the Parties (COP) in Berlin in 1995 and was since then known as the Berlin ‘firewall’ (see Oberthür and Ott 1999, Gupta 2014, pp. 59–98).

The ‘new’ climate regime emerged between 2007 and 2015 and followed a different logic. Instead of requiring states to agree on emission cuts made during international climate negotiations, it was defined by a ‘bottom-up approach’ in which each party defined its ‘nationally determined contributions’ (NDCs) at the domestic level. Moreover, the Paris Agreement has replaced differentiation based on membership in Annex I with a new scheme of differentiation that is individualized as well as informal.

Differentiation is individualized because, even though the Paris Agreement still speaks of developed and developing countries, each party defines its own target based on a reflection of its common but differentiated responsibilities and respective capabilities and now, crucially, also ‘in the light of different national circumstances’. In addition, differentiation is informal because parties no longer formulate (and defend) their ambition level in view of legally binding targets, but rather in view of a broader set of social norms. As a result, there is considerable leeway in evaluating the NDCs of individual parties as either ‘adequate’ or ‘inadequate’, and the informal assessment of such adequacy becomes part of the political process itself. This is a key difference from the Kyoto Protocol approach, in which the question whether Canada or Croatia had fulfilled their respective commitments could be answered more easily and with greater precision.

Explanations for the alleged fundamental transformation of the climate regime often refer to the challenges the ‘firewall’ approach generated and to the difficulties which the need for consensus decision-making posed for efforts to add or remove countries to or from Annex I. Falkner (2016, p. 1111), for instance, has argued that ‘the rigid divide between Annex I and non-Annex I countries had made it difficult to deal with the rapidly rising emissions of emerging economies’. On the one hand, the emerging powers were happy with the status quo and ‘did not want to stifle their future economic development by imposing limits on future emissions’ (ibid.). On the other hand, the US Senate had already signalled its opposition to developing country exemptions in its Byrd-Hagel resolution that was passed unanimously in 1997. The resolution considered ‘the exemption for Developing Country Parties (…) inconsistent with the need for global action on climate change’. It further envisaged that ‘the disparity of treatment between Annex I Parties and Developing Countries and the level of required emission reductions, could result in serious harm to the United States economy’. And it categorically ruled out that the US would ratify an agreement that included new commitments for Annex I countries without also mandating ‘new specific scheduled commitments to limit or reduce greenhouse gas emissions for Developing Country Parties within the same compliance period’ (US Congress 1997).

Initially, not all Annex I parties shared this view. EU members, for example, accepted the differentiation they had agreed to in the Kyoto Protocol’s first commitment period as a fair deal, based as it was on the UN Framework Convention and its requirement on Annex I parties to take the lead. By the time a second commitment period had to be negotiated, however, things looked different. Since 2007, parallel negotiations had been opened on continuing the Kyoto Protocol (the ‘Kyoto Protocol track’, launched at CMP-1 in 2005) as well as on developing a new agreement under the UN Framework Convention (the ‘Convention track’, launched by the Bali Action Plan in 2007) (see, e.g., ENB 12/375 of 16 June 2008: 18).¹ The Kyoto Protocol track, however, faced a major setback when Canada, Japan, and Russia declared that they would not take part in a second commitment period, and Australia and New Zealand left their participation open. An important reason for the reluctance of these countries was the absence of the US from the Kyoto Protocol. Another reason was that the relative share of emissions from rising powers like Brazil, China, and India had increased as these countries had developed economically. This created strong pressures on rising powers – often labelled as ‘major emitters’ – to accept commitments in some form. At COP-15 (2009) in Copenhagen, several developing countries including China and India responded with a first set of voluntary pledges, thereby signaling a willingness to control their greenhouse gas emissions in the future.

Ultimately, the Convention track won, and states subjected the architecture of the climate regime to a major overhaul in the years leading up to 2015. In their comments on the negotiations that led up to the Paris Agreement, several observers described the implications this new architecture had for the existing system of differential treatment in stark terms. Writing on COP-17 (2011) in Durban, where states agreed on the mandate for new climate negotiations, Rajamani (2013, p. 163) thus described the outcome of negotiations as ‘the final blow to differentiation in the climate regime’. This claim was largely based on the observation that states had decided to negotiate a new agreement that would no longer focus on the emissions of developed countries, but instead be ‘applicable to all parties’. Other observers concurred that the decisions taken in Durban moved the international climate regime towards greater ‘symmetry between mitigation commitments and actions taken, respectively, by developed and certain developing countries’ (Deleuil 2012, p. 280), and that the absence of an explicit reference to the CBDR principle was ‘indicative of the diminishing ability of the developing world, especially emerging powers, to maintain their case for “differential” treatment on the subject of climate change’ (Hurrell and Sengupta 2012, p. 472). Finally, developing country observers themselves interpreted the outcome as ‘blurring the distinctions between developed and developing countries in their respective and qualitatively different types and levels of commitments and responsibilities’ and therefore a weakening of the CBDR principle (Khor 2020, p. 71).

As I will argue in the following, this view is correct but incomplete. It is correct in pointing out that differentiation among parties is organized differently in the Paris Agreement. At the same time, however, it overlooks how the negotiation of the Paris Agreement builds on key elements and decisions in the early climate regime. In brief, the argument I develop in the following thus seeks a middle ground. On the one hand, I acknowledge key differences between the old and new systems of differentiation in the climate regime. On the other hand, I seek to show how the seeds for individualized and informalized differentiation had already been planted in the original climate regime. Because the latter argument is closely tied to my argument that the UN Framework Convention contains multiple layers of differentiation, I begin by describing these layers.

3. Beyond the annexes: further layers of differentiation

How does the climate regime differentiate between states? As outlined above, the most central distinction is between Annex I and non-Annex I parties. Beyond this distinction, however, the international climate regime comprises further instruments of differentiation that are often overlooked. Some of the additional layers I sketch below include elements that apply to all parties; others allow for a further differentiation within the groups of Annex I countries and among non-Annex I parties, respectively.

Linguistic qualifiers as a generalized tool for differentiation. Regarding the first category, the UN Framework Convention includes linguistic qualifiers that go beyond the question of whether or not a state is listed in Annex I to the UN Framework Convention. Notably, as Rajamani (2006, p. 163) has observed, ‘implicit norms of differentiation, by their use of language such as “if necessary”, “as appropriate”, “in so far as possible”, and “all practicable steps”, grant states considerable room for manoeuvre’. For example, Article 12 of the UN Framework Convention regulates which information parties are required to communicate to the Conference of the Parties. This information includes national inventories of greenhouse gas emissions or the steps a country has taken or plans to take to implement the Convention. As a qualifier, however, Article 12.1 stipulates that each state party to the Convention shall provide information on the national inventory and use ‘comparable methodologies’ parties have agreed upon ‘to the extent its capacities permit’. Similarly, material relevant for the calculation of global emission trends is to be included ‘if feasible’. For developed countries, in turn, the notion ‘as appropriate’ frequently serves as a qualifier. For instance, Article 4.5 (emphasis added) stipulates that,

the developed country Parties and other developed Parties included in Annex II shall take all practicable steps to promote, facilitate and finance, as appropriate, the transfer of, or access to, environmentally sound technologies and know-how to other Parties, particularly developing country Parties, to enable them to implement the provisions of the Convention.

Seen from this angle, the division in Annex I non-Annex I parties serves as a minimum standard. It informs parties about their general place in the international climate regime as well as about the different rights and obligations associated with that place. However, heterogeneity among the parties within each category remains large. To accommodate this heterogeneity, qualifiers such as ‘if possible’, ‘if feasible’ or ‘as appropriate’ allow states to make and defend legitimate claims for more specific kinds and levels of differentiation. Such qualifiers are ubiquitous in international law. Nonetheless, the fuzziness the qualifiers insert into the text introduces elements of individualization and informalization. Provisions which, like Article 4.10 of the UN Framework Convention, invoke ‘the situation of Parties’ as a yardstick for assessing the latter’s implementation efforts, allow states to make highly contextualized claims (individualization). Moreover, they turn the acceptance of such claims into a political rather than a legal question (informalization).

Further differentiation within Annex I. Once we look at differentiation among Annex I countries, Annex II of the UN Framework Convention adds another well-known layer. Unlike Annex I, which lists states that have special obligations in view of climate change mitigation, Annex II includes all countries that have special obligations to contribute to international climate finance and to the transfer of environmental technologies. In terms of composition, states that were OECD members at the time the UN Framework Convention was concluded are listed on both annexes. By contrast, the Eastern European countries with economies in transition (EITs) are only included in Annex I, but not in Annex II.

The group of EITs is also the object of a further differentiation recognized in Article 4.6 of the Convention. This article grants ‘a certain degree of flexibility’ in the implementation of their commitments to all those Parties that are included in Annex I but are, at the same time, ‘undergoing the process of transition to a market economy’. Notably, the flexibility includes the choice of the reference year in relation to which Annex I parties are expected to stabilize their emissions. Reiterating these implementation flexibilities, Article 3.6 of the Kyoto Protocol confirms this differentiation involving ‘economies in transition’.

Further differentiation within the group of non-Annex I parties. Finally, the Convention recognizes a diverse bundle of reasons that allow certain developing countries to invoke special circumstances. Article 4.9, for example, recognizes the specific needs of the least developed countries and calls upon other states to ‘take full account’ of these needs, notably ‘in their actions with regard to funding and transfer of technology’. As a result, the UN Framework Convention allowed LDCs to submit their first national communication on implementation ‘at their discretion’ (Article 12.5). Moreover, the recognition of LDCs as a separate category has informed institutional developments like the creation of an LDC expert group or the LDC fund. Overall, it thereby introduces a clear form of differentiation among non-Annex I parties.

Beyond the LDC category, Article 4.8 of the UN Framework Convention is of interest. This article addresses ‘the specific needs and concerns of developing country Parties’ that may arise from the adverse effects of climate change itself and/or from the impacts which the implementation of response measures may have on them, and it calls upon parties to ‘give full consideration’ to such needs and concerns. From the perspective of differentiation, the list of country categories which Article 4.8 introduces is particularly intriguing. It includes no less than nine additional categories of state parties that may claim ‘specific needs and concerns’, namely:

(a) Small island countries; (b) Countries with low-lying coastal areas; (c) Countries with arid and semi-arid areas, forested areas and areas liable to forest decay; (d) Countries with areas prone to natural disasters; (e) Countries with areas liable to drought and desertification; (f) Countries with areas of high urban atmospheric pollution; (g) Countries with areas with fragile ecosystems, including mountainous ecosystems; (h) Countries whose economies are highly dependent on income generated from the production, processing and export, and/or on consumption of fossil fuels and associated energy-intensive products; and (i) landlocked and transit countries.

Like the provisions discussed above, Article 4.8 recognizes heterogeneity. In formal terms, the legal domain to which the article applies may be relatively narrow: it is limited to assistance for developing countries’ adaptation to climate change or to the consequences of other countries’ mitigation activities. However, the general thrust of the article is an acknowledgment of difference in national circumstances, even though that term is carefully avoided.

As an underlying social norm rather than an agreed legal norm, this acknowledgement invites three kinds of claims. First, it lends itself to the argument that, if differences in national circumstances inform the climate regime in the field of assistance, the same logic may also be transferred to other regime areas. Second, it invites the claim that, if differences in national circumstances of developing countries are explicitly acknowledged, then a comparable logic may also be extended to developed countries that differ in these regards. Third, it raises the question whether the categories listed in Article 4.8 – plus maybe those of EITs and LDCs – are exhaustive or if, following the underlying logic of the ‘relevant differences’ which informs the list, additional categories may need to be considered. The final sentence of Article 4.8 explicitly addresses this latter question when it stipulates that ‘the Conference of the Parties may take actions, as appropriate, with respect to this paragraph’. Practically, moreover, at least one of the episodes I discuss below revolves around – and eventually led to – the creation of a further category of state parties that gained the right to differential treatment after the Convention had been adopted. Finally, it is worth noting that none of the categories listed in Article 4.8 are defined in the Convention text itself. In contrast to the list approach of Annex I (and Annex II) and to the LDC category for which a list of LDCs maintained and regularly reviewed by the United Nations provides the legal foundation, membership in the categories mentioned in Article 4.8 is thus based on the self-declaration of parties and subject to the (often tacit) acceptance of such declarations by other parties.

4. Multi-layered differentiation and the shift towards individualization and informalization in the Paris agreement

To what extent and through which mechanisms have these additional layers of differentiation contributed to the path from the alleged ‘rigid divide’ of the original climate regime to the more individualized and more informalized differentiation of the Paris Agreement? In the following, I will discuss three specific episodes in the history of the early climate regime to address this question. In brief, these episodes include:

1. The conflict over the placement of Turkey on the two annexes to the Convention in which the Conference of Parties eventually agreed to recognize the ‘special circumstances’ of Turkey and thereby acknowledged the possibility of dealing with countries on their own specific terms.

2. Iceland’s successful call for a special regulation focusing on ‘small economies’, in response to which parties accepted further exemptions. This episode illustrates best that the categories listed in Article 4.8 were not exhaustive, that the idea of further differentiation was not limited to non-Annex I parties, and that the notion of ‘special circumstances’ followed an inherently expansive logic.

3. Finally, the adoption of individualized targets for each Annex I party in the Kyoto Protocol added to the existing increasing ‘anomalies’ that resulted from the binary distinction between Annex I and non-Annex I parties. In the long term, it also undermined the idea that developing countries could be treated as a homogenous block and hence lent support for an individualization of their commitments, as well.

Taken together, these episodes provide an account of the development of the climate regime that, while acknowledging key differences between ‘Kyoto’ and ‘Paris’, also recognizes that individualized and informal differentiation was inherent to – and thus had always existed in – the climate regime. Importantly, moreover, the three episodes I discuss do not constitute an exhaustive list. Another development includes, for instance, the Republic of Korea’s acceptance of an implicit obligation to ‘graduate’ and step up its climate policy ambition in the context of its OECD accession in 1996 and despite retaining its non-Annex I status in the climate regime. Similarly, the group of Small Island Developing States (SIDS) successfully requested a seat in the COP Bureau (FCCC/CP/1996/2 of 22 May 1996: Rule 22), and parties also recognized concerns of Croatia as well as Italy over their baselines in relation to specific regime provisions regarding Land Use, Land Use Change and Forestry (LULUFC) under the Convention and Kyoto Protocol, respectively.² Taken together, these cases demonstrate that individualization and informalization are not new to the climate regime.

4.1 The placement of Turkey on annex I and annex II

The placement of Turkey illustrates the intricacies of differentiation associated with the list approach (see also Boyte 2010: 86–87; Depledge 2009: 180–181; Rajamani 2006: 169–170). An OECD member since 1961, the UN Framework Convention initially placed Turkey on Annex I as well as Annex II. The Turkish government, however, argued that Turkey’s placement ‘did not conform to the country’s economic circumstances’ and objected to being placed on both annexes (ENB 12/97 of 16 November 1998: 10; see also Rajamani 2006, p. 169). While some countries supported the removal of Turkey from Annex I and Annex II, others rejected the idea because of the ‘precedent it would set’ (ENB 12/123 of 8 November 1999: 7). The issue was temporarily settled at COP-7 (2001) when Turkey was removed from Annex II but not from Annex I (ENB 12/189 of 12 November 2001: 8). This exempted Turkey from the obligation to provide financial and technological assistance to developing countries, but not from its special obligations in the field of climate change mitigation.

In the academic literature, the conflict over the placement of Turkey usually serves as an illustration for the way in which the relatively crude distinction and rigid architecture of differentiated based on the annexes impeded the further development of the original climate regime. This view is based on three specific observations: first, that the rules for amending the annexes to the UN Framework Convention made adding or removing countries overly difficult; second and related, that it took nine years to negotiate a compromise on the status of Turkey under the Convention; and third, that Turkey continued to press for additional exemptions after that compromise had been agreed.

On the first point, the rules for amending the annexes are laid down in Articles 15 and 16 of the UN Framework Convention. The relevant provisions stipulate that amendments to the annexes follow the same procedure as the adoption of new annexes, that new annexes are adopted in the same way as amendments of the Convention itself, and that ‘Parties shall make every effort to reach agreement on any proposed amendment to the Convention by consensus’. Only if all such efforts fail, amendments may also ‘be adopted by a three-fourths majority vote of the Parties present and voting at the meeting’ – and then only become binding for those parties that have accepted the amendment in question. Effectively, a single state party could thus slow down the adoption of amendments significantly, and a group of countries representing one quarter of all parties to the Convention was able to block amendments altogether. In the history of the climate regime, this has not prevented parties from amending Annex I. Notably, parties have added Croatia, Cyprus, Liechtenstein, Malta, Monaco, and Slovenia to Annex I and replaced Czechoslovakia with its successor states, the Czech Republic and Slovakia. However, the rules made amendments difficult when the political stakes – and hence also the political controversy – were high.

The Turkish case illustrates this dynamic. The stakes were relatively high since the country was a large emitter as well as a candidate for joining the EU. In addition, the stakes further increased with the negotiation of a protocol to the Convention that was envisaged to include quantified emission reduction targets for Annex I parties. At COP-3 (1997) in Kyoto, Azerbaijan and Pakistan presented a proposal to delete Turkey from Annex I and Annex II based on the idea that Turkey was ‘a medium developed country’ and that its emissions were ‘fractional (…) compared to the Annex I average’ (ENB 12/66 of 3 November 1997: 11).³ If accepted, this proposal would have exempted Turkey from a quantified emissions target under the Kyoto Protocol.

Iran and Kuwait supported the proposal, but the EU and Australia objected. They argued that ‘Turkey should indicate willingness to undertake Protocol obligations under Article 10 before its deletion from the Annexes’ (ENB 12/76 of 13 December 1997: 5).⁴ The EU further demanded that ‘all OECD members should adopt commitments under a protocol’, thereby addressing not only Turkey but also the new OECD members South Korea and Mexico. ‘Pending a possible special regime’ for these three countries, the EU thus objected to the proposed amendment. Korea, in turn, insisted on the distinction ‘between the status it shares with Mexico as a non-Annex I Party and that of Turkey’. Likewise, Mexico insisted on the legal validity of the existing annexes and ‘rejected attempts to link membership in any organization with Convention obligations’ (ENB 12/66 of 3 November 1997: 11).

At subsequent conferences, the conflict dragged on. While Turkey reiterated that its status as an Annex I country ‘was an anomaly that delays its ratification of the Convention’ (ENB 12/97 of 16 November 1998: 6), developed countries asked for ‘information on Turkey’s approach to responsibilities’ (ENB 12/86 of 15 June 1998: 7). Agreement appeared elusive for some years, with states deferring the issue of Turkey’s status from one conference to the next. It took until 2001 for the Subsidiary Body for Implementation (SBI) to present a solution on which states could agree. The compromise formula stipulated that Turkey would be deleted from Annex II, but not from Annex I. This allowed Turkey to ratify the Convention – as the 189th party – in 2004 and to join the Kyoto Protocol in 2009. As a non-party at the time of negotiations, however, Turkey’s name had not been included in Annex B to the Kyoto Protocol which specified the quantified emission targets for other Annex I parties. In subsequent years, Turkey stated its intent to also retain its status ‘as an Annex I country that is not included in Protocol Annex B’ in a second commitment period (ENB 12/447 of 9 November 2009: 8). Moreover, it continued to ask for the deletion of its name from Annex I well after conclusion of the Paris Agreement in 2015.

In many ways, the conflict around Turkey’s status thus illustrates the difficulty of amending the annexes to the major international climate treaties in the absence of consensus. Yet, the solution that parties found in decision 26/CP.7 also entailed an important innovation. For the first time, the Conference of the Parties (COP) thus recognized the ‘special circumstances’ of a single party to the climate regime. More precisely, the COP decision not only ‘decides to amend the list in Annex II to the Convention by deleting the name of Turkey’, but also ‘invites the Parties to recognize the special circumstances of Turkey, which place Turkey, after becoming a Party, in a situation different from that of other Parties included in Annex I to the Convention’ (FCCC/CP/2001/13/Add.4 of 21 January 2002: Decision 26/CP.7, emphasis added).⁵

Practically, one effect of the decision was that the special circumstances clause allowed Turkey to ‘[receive] funding from the Global Environment Facility for the preparation of its first national communication’ (Depledge 2009, p. 281). For the present article, however, an indirect consequence may seem more relevant. Because the idea that only one of more than 190 parties to the UN Framework Convention could claim ‘special circumstances’ was not very persuasive, the logic of ‘special circumstances’ which Decision 26/CP.7 invoked was expansive almost by definition. Following this logic, Turkey regularly referred to its ‘special circumstances’ in a more generalized way in negotiations of a post-Kyoto regime. At COP-15, for example, it ‘proposed a reference to special circumstances of parties, including countries with economies in transition’ (ENB 12/459 of 22 December 2009: 16). At COP-21, it expected the COP Presidency ‘to work towards finding a solution to issues raised by Turkey, related to parties with special circumstances that have been recognized by the COP having access to support under the agreement’ (ENB 12/663 of 15 December 2015: 12). The tension between a general ‘special circumstances’ clause and the singular recognition of Turkey under that clause maybe became most evident in Turkey’s request ‘to be considered as a group for consultations on negotiations’ at COP-21 (ENB 12/663 of 15 December 2015: 9, emphasis added).

Substantially, moreover, Turkey also became a strong proponent of differentiation based on national circumstances. It emphasized a need for ‘flexibility in a future regime to account for the dynamic nature of national circumstances’ (ENB 12/383 of 30 August 2008: 4). It sought a new regime that would ‘[evaluate] national circumstances of parties in an objective manner’ (ENB 12/595 of 17 March 2014: 4). And it asked that a new agreement should ‘take into account socioeconomic realities and be guided by science’ (ENB 12/626 of 16 February 2015: 9). Overall, the conflict over the status of Turkey thus generates two main insights. On the one hand, the amendment procedures made formal change difficult. On the other hand, however, the solutions states invented to address this impasse paved the way for a recognition of the ‘special circumstances’ of parties. Once such circumstances were recognized for one party, however, it became difficult to limit them to that party alone.

4.2 Iceland’s claim for special rights for ‘small economies’

Another episode in which the state parties to the UN Framework Convention found a creative solution to the ‘special circumstances’ of one of its members concerned Iceland’s request to recognize the specific needs of ‘small economies’. Based on a specific paragraph included in the COP decision on the adoption of the Kyoto Protocol (Decision 1/CP.3, para. 5(d)), Iceland argued that for small economies, single industry projects could have a high proportional effect on national emissions. At COP-4 (1998) in Buenos Aires, it therefore proposed that industry projects that had come into operation after 1990 and contributed more than five per cent to the total greenhouse gas emissions of a country over the first commitment period of the Kyoto Protocol should not be counted to the party’s national total allocated for that period. According to the proposal, this exemption should be limited to ‘small economies’ defined as those countries whose total emissions remained below 0.05% of Annex I emissions in 1990. Moreover, it should only apply to projects that followed best environmental practice to minimize emissions (FCCC/CP/1998/MISC.11 and Add.1 of 9 November 1998).

Discussions at COP-4 revealed only limited support for the proposal. The negotiation summary by the Earth Negotiation Bulletin lists opposition and critical statements from Antigua and Barbuda, Austria on behalf of the EU, Barbados, Brazil, Canada, the Marshall Islands, and Tuvalu; the small island countries specifically ‘cautioned against granting exemptions prior to the Protocol’s entry into force’ (ENB 12/97 of 16 November 1998: 4). Australia and the US lent some support to the general thrust of the proposal. But as several countries requested more time for consultations, the conference tasked the Subsidiary Body for Scientific and Technological Advice (SBSTA) with further discussion of the issue at its next session (Decision 16/CP.4 of 11 November 1998).

At COP-5 (1999) in Bonn, discussions continued in the SBSTA. Canada ‘expressed reservations on the issue and opposed the draft decision tabled by Iceland at COP-4 in its current form’ (ENB 12/123 of 8 November 1999: 10–11). In its submission, it further clarified that its concern was to ‘ensure that this exercise [does] not lead to the re-negotiation of Parties’ emissions commitments for the first commitment period, as set out in Annex B of the Kyoto Protocol’. Moreover, both Canada and Germany (on behalf of the EU) asked for additional information on how accepting Iceland’s proposal might influence industries to locate production in Iceland rather than in other countries (FCCC/SBSTA/1999/MISC.3 of 7 April 1999: 3–5 and 6–8). As no agreement could be reached, the SBSTA again deferred consideration to its next meeting (FCCC/SBSTA/1999/L.17 of 30 October 1999).

Subsequently, states developed an ‘almost agreed’ draft decision in 2000 (ENB 12/151 of 18 September 2000: 5; ENB 12/163 of 27 November 2000: 15), and they eventually adopted a decision at COP-7 (2001) in Marrakesh. In comparison to Iceland’s original proposal, the decision limited the total emissions a Party could ‘report separately’ – and hence exempt from their budgets in the first commitment period – to 1.6 million tons of CO2 per year on average (FCCC/CP/2001/13/Add.1 of 21 January 2002: Decision 14/CP.7).

The decision itself does not alter the contours of differentiation in the international climate regime in a fundamental way. In response to the questions from other parties, Iceland had earlier argued that ‘this draft decision only applies to three small Annex I Parties’ (FCCC/CP/1998/MISC.11 of 7 November 1998), and eventually only two countries – Iceland and Monaco – notified the UNFCCC Secretariat of their intent to apply this clause (FCCC/CP/2002/Misc.2 of 24 October 2002 and FCCC/CP/2002/7 of 28 March 2003: para. 147). Nonetheless, the episode indicates that the list of reasons that may count as ‘special circumstances’ was neither fix, nor limited to developing countries. The UN Framework Convention set certain boundary conditions. But in principle, whose ‘special circumstances’ counted was contingent on political agreement.

4.3 The recognition of heterogeneity in the Kyoto Protocol

Finally, a generic problem of the annexes was that they left what some observers have called ‘anomalies’ (Bodansky 1993; see also Boyte 2010, p. 86) or ‘fuzzy margins’ (Rajamani 2006: 169–171). This problem was widely known. Dan Bodansky (1993: 507, note 336), for example, noted that, as an Annex II party, ‘Turkey, as a “developed” country, would have an obligation to provide financial resources to Saudi Arabia, a “developing” country’. This constituted an ‘anomaly’ because Saudi Arabia’s GDP per capita in 1992 just about doubled that of Turkey. Similarly, he argued, ‘Israel and South Africa, nations that some consider industrially and economically “developed”, are not included in Annex I’ (ibid.). This, too, was difficult to justify because, at least for Israel, GDP per capita ranked above several countries included in Annex I. Further ‘anomalies’ included Singapore and OPEC members like Qatar, Kuwait or the United Arab Emirates. While Singapore was considered an industrialized country under the rules of the World Trade Organization (WTO), the OPEC members were relatively rich, scored high on the Human Development Index (HDI), and had high per capita greenhouse gas emissions. However, none of them were Annex I parties in the climate regime (Rajamani 2006, pp. 170–171, Boyte 2010, pp. 86–87). In brief, because the placement of countries on the annexes did not follow clear indicators of economic development, the classification of some countries was hard to justify based on generalized principles. Moreover, even though the additional elements discussed above added some differentiation within each category, the annexes themselves failed to do so.

In 1997, however, the negotiation of the Kyoto Protocol introduced numerical differentiation within country categories. Negotiated at COP-3, Annex B of that Protocol assigns each Annex I party a specific emissions target for the ‘first commitment period’ from 2008 to 2012. Croatia (5%), Canada, Hungary, Japan, and Poland (6%), the US (7%), several of the European economies in transition, and the European Community (8%) committed to reduce their emissions to different degrees. New Zealand, Russia, and the Ukraine accepted an obligation to maintain their emission levels. Finally, Norway, Australia, and Iceland agreed to limit the increase of their greenhouse gas emissions by 1, 8 and 10% compared to a 1990 baseline, respectively.

In many ways, this specific solution added further impetus to the path of individualized differentiation on which the Paris Agreement would eventually embark. Its proposition was simple: if emission targets could take the specific circumstances of Australia and New Zealand into account and allocate each party its own appropriate amount, then why should international climate law treat all ‘developing countries’ the same? This question became only more pertinent in view of the anomalies discussed above. If Canada, Iceland, and Norway had numerically differentiated commitments, then why should Saudi Arabia, Singapore, and South Korea share the same mitigation commitments as Barbados, Bolivia, and Botswana? Because the adoption of the Kyoto Protocol demonstrated that differentiation could be nuanced (Oberthür and Ott 1999), the combined approach of (i) maintaining a ‘firewall’ between an Annex I country like Portugal and a non-Annex I country like South Korea, (ii) individually differentiating between Denmark (−8%) and Iceland (+10%), but (iii) not differentiating between Singapore and Botswana seemed increasingly incoherent. Overall, developing countries may thus have been satisfied with the outcome of the negotiations at COP-3 because the Kyoto Protocol – at least in its first commitment period from 2008 to 2012 – exempted them from formal commitments to limit their greenhouse gas emissions. In the long run, however, the individualized commitments which the Kyoto Protocol introduced followed an expansive logic, as well.⁶

5. Conclusion

Taken together, the three episodes illustrate that individualized and informal differentiation is not an invention of the Paris Agreement or the negotiations immediately leading up to it. The conventional narrative tends to see a fundamental overhaul that began to take shape after the climate conference in Bali in 2007 and then culminated in the diplomatic breakthrough in Paris in 2015. By contrast, the episodes I have reconstructed above suggest that the system of differentiation that emerged between 2007 and 2015 builds on developments that had already taken place much earlier.

Taking this insight on board has consequences. First and foremost, it makes the shift from Kyoto to Paris look a lot more gradual, incremental, organic, and thereby helps us recognize that the theoretical idea that institutions do not change overnight also applies in this field (see also Allan et al. 2021). This recognition matters: it prevents us from drawing the wrong lessons from history and, for instance, from unduly betting on the outcomes of single climate summits and diplomatic leadership. At the same time, it also corrects an important myth about the allegedly ‘rigid nature’ of the old climate regime. It is true that this regime was relatively rigid when it came to amending Annex I. Yet, this specific constraint had major implications itself. First, it forced states to show flexibility in other areas. Second, the additional layers of differentiation states created in response to the rigidity of Annex I became more expansive over time. Taken together, these developments led to a system of differential treatment among parties that was complex, but also far more nuanced than the common focus on the binary distinction between Annex I and non-Annex I parties suggests.

Next, taking the idea of gradual regime development seriously has consequences for how we large or small we perceive the differences between the ‘old’ and ‘new’ climate regime as well as for what we take to have made the ‘success’ of Paris possible (see also Allan et al. 2021). To be sure, the logic of the Paris Agreement differs significantly from the logic of the Kyoto Protocol. Nonetheless, as Allan (2019, p. 4) has remarked, many of its key provisions ‘represent continuity with existing climate policy, not a break with the past’. An important part of the reason for this continuity lies in the dynamics I have sketched in this article. When the Kyoto Protocol was negotiated in 1997, states could be confident that a major revision would not occur before the end of the Protocol’s first commitment period in 2012. Because some of them – in fact, quite many – had reasons to be dissatisfied with the ways in which differentiation among parties was organized and in which the annexes failed to recognize the ‘special circumstances’ of parties, they began to amend the regime at the margins.

Finally, the episodes also tell us something about agency in international climate politics. Thus, the status of countries like Turkey and Iceland may not have been central to the functioning of the regime itself. Yet their actions mattered in how the regime developed. Like the individualized differentiation among Annex I countries in the Kyoto Protocol itself, the solutions states found for each case I have discussed in this article contributed to the normalization of individualized and informalized commitments that now informs the Paris Agreement. While the Paris Agreement is novel in some ways, it thus also codifies and institutionalizes major ideas and practices states have gradually developed since the adoption of the UN Framework Convention in 1992. In the absence of such practices, we might therefore speculate, the ‘new’ climate regime would have been unable to emerge.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

The work was supported by the Swiss National Science Foundation (SNSF) [100017L_175887].

Notes

1. The Earth Negotiations Bulletin (ENB) is an independent reporting service that covers international environmental negotiations and is hosted by the International Institute for Sustainable Development (IISD). I cite reports according to their volume (volume 12 covers negotiations under the UN Framework Convention on Climate Change), running issue number, and date of publication. Reports for the years from 1995 to 2019 can be accessed at https://enb.iisd.org/enb/vol12/ (last accessed 19 August 2022).

2. I thank an anonymous reviewer for pointing out some of the additional cases.

3. In 1997, the World Bank records rank Turkey 25^th of all countries in terms of its total greenhouse gas emissions. While accounting for less than one per cent of the global emissions, the ‘fractionality’ of Turkey’s overall emissions might thus be contested. By comparison, its per capita emissions of 3.2t in the same year put Turkey significantly below those of other Annex I parties (see https://data.worldbank.org/indicator/EN.ATM.CO2E.PC?locations=TR, last accessed 24 March 2023).

4. At the time of the statement, Article 10 was a draft article meant to enable non-Annex I parties to take on voluntary commitments under the Kyoto Protocol. The proposed article was dropped in the negotiations and did not become part of the text of the Kyoto Protocol.

5. I cite official UNFCCC documents according to their UN document number. The documents are available and can be searched based on their official document number at https://unfccc.int/documents (last accessed 06 May 2022).

6. As another example of individualized differentiation, an anonymous reviewer noted that Annex I parties that were parties to the Kyoto Protocol also agreed on individualized caps for forest management activities at COP-7 in Marrakesh (2001) (see appendices to decisions 11/CP.7 and 16/CMP.1). This observation is in line with the argument I present here.