Stuck in the middle with you (and not much time left): The third intergovernmental conference on biodiversity beyond national jurisdiction

The third of four scheduled Inter-Governmental Conferences on the conservation and sustainable use of marine biodiversity beyond national jurisdiction met in New York in August 2019. This article tracks the progress made in the negotiations, focusing on the four key themes the draft treaty is addressing: (1) marine genetic resources, (2) area-based management tools, including marine protected areas, (3) environmental impact assessments, and (4) capacity building and transfer of marine technology. Drawing on process tracing (i.e. observations, interviews, and literature analysis), we have observed several critical issues in the emerging institutional design of a future agreement for ‘Biodiversity Beyond National Jurisdiction’ (BBNJ). These include the continued ideological polarization between existing ocean governance principles (‘freedom of the seas’ and ‘common heritage of mankind’), disagreements about the delegation of authority to existing or created institutions, uneven participation of scientific and industry stakeholders, and the challenge of formulating a legal instrument that relies on inchoate or inconsistently used concepts. The conclusion looks ahead to the fourth Inter-Governmental Conference, and assesses the potential of reaching an effective agreement before the negotiations are scheduled to conclude in April 2020.


Introduction
"The new instrument must end governments' grabs in the high seas" -Malawi on behalf of Least Developed Countries, opening statement 8/19/19 1

fn1
Covering three fourths of the earth's surface area, the ocean is the world's largest ecosystem. Areas beyond national jurisdiction (ABNJ), incorporating the high seas and the international seabed, comprise more than sixty percent of it. Nearly ten million tonnes of fish are harvested in this region annually, with an estimated landing value of US$16 billion, or 15% of total global marine landed value [1]. The legal framework for ocean governance in ABNJ is not operating in a void, however it is largely fragmented and uncoordinated, resulting in a patchwork of regulatory schemes covering issue areas from the protection of migratory birds, to deep sea mining, to the dumping of illegal wastes from ships, to pollution from land-based sources. There are at least 190 multiand bi-lateral agreements addressing a range of issue areas that affect the ocean, not including other forms of global governance, such as customary international law, working practice, or informal rules [2]. As a result, biodiversity protection has somewhat 'slipped through the cracks' of ocean governance, especially for the ABNJ [3], and there are also concerns about effective and equitable conservation on the high seas, as evidenced by the quote from Malawi, above.
In September 2018, following over a decade of informal efforts, the international community began a formal process of negotiating a new international legally-binding and over-arching instrument to address the conservation and sustainable use of marine biodiversity beyond national ☆ This reference is to the song "Stuck in the middle with you" by Stealers Wheel (1973).
jurisdiction (referred to here as the BBNJ negotiations), in line with United Nations General Assembly's (UNGA) Resolution 72/249. 2 The BBNJ negotiations are structured around four main sets of issues: (i) marine genetic resources, (ii) area based management tools, including marine protected areas, (iii) environmental impact assessments, and (iv) capacity building and the transfer of marine technology. These issues represent the basic mandate of the BBNJ negotiations, which is to produce a single agreement that covers them all. This mandate is restricted by the proviso that any new agreement "should not undermine" relevant existing frameworks, bodies, and instruments [4]. UNGA Resolution 72/249 stressed the need for the widest possible participation and the use of consensus-based decision-making.
The third of four scheduled Inter-Governmental Conferences (IGC-3) on BBNJ took place at the United Nations headquarters from August 19-30, 2019. Unlike the previous two conferences, the IGC-3 negotiations focused around draft text produced by the President of the negotiations, Rena Lee of Singapore, released to delegations and the general public on July 25th, a few weeks in advance of the meeting. This draft text was publicly praised by nearly all delegations during the start of the third meeting, and it substantially changed the tenor, pace, and detail of interventions compared with the first two meetings (see Refs. [4,5]). At the meeting, many delegations circulated text-based proposals, which were updated and re-circulated throughout the conference. In the discussion below, these are referred to as Conference Room Papers (CRPs), which were made available electronically to delegates. Despite a lack of significant progress in IGC-3, and privately-expressed concerns about the timeline, there was no formal discussion about whether additional IGCs would need to be scheduled beyond the fourth and final IGC in March 2020.
IGC-3 also differed from the previous two conferences with regard to format. Whereas all issues were discussed in plenary informal working groups in IGC-1 and IGC-2, with full access for intergovernmental and nongovernmental organizations (IGOs and NGOs), and which were webcast so that those not at the meeting could follow along, IGC-3 included fourteen "informal informals," which took place in a smaller, more closed session. In general, more contentious issues were scheduled to be discussed in the informal format. These informal informals were not webcast, press were not allowed, and the attendance by NGOs and IGOs was limited to ten total seats. In some small ways, this reduced accessibility extended to state delegations as well. For example, meetings often ran over time, meaning that interpretation services were discontinued and discussion proceeded only in English. And for the first time at these negotiations, the advent of informal informals and ad hoc changes in scheduling resulted in four sessions overlapping, which greatly disadvantages smaller delegations that may not have enough people to be present at multiple simultaneous sessions [6]. At the same time however, parallel sessions, especially informal discussion-oriented arenas, can foster compromise and dialogue in a way that is less possible in more formal arenas.
While the informal informals intended to move the negotiations forward by providing a lower-stakes atmosphere for working out compromises, and they were fairly transparent in allowing observers to be present, in practice they seemed very similar to the plenary informal working groups with regard to content. Discussions about specific terms like "established," "designated," and "existing" took up significant amounts of time, while polarized debates about the applicability of principles such as 'freedom of the seas' and 'common heritage of mankind' continued. The most notable effect, and unintended consequence, of the informal format may in fact have been to reduce the visibility and influence of NGOs and IGOs, who were not invited to speak, and who had to jockey amongst themselves for a limited number of seats. In fact, during IGC-3 we recorded NGOs and IGOs speaking a total of only 37 occasions (during plenary working groups). In contrast, during IGC-1 and IGC-2 they spoke a total of 69 and 85 times, respectively.
This article is the third in a series that seek to identify important variables shaping the negotiations, describe developing trends with regard to consensus building, and comment on the obstacles and challenges facing delegates. Our overall research question is examining the factors that can explain the prospects for and design of the final BBNJ agreement. Our analysis of IGC-1 explored whether and how the desire to maintain and expand national jurisdiction shapes the emerging BBNJ instrument [5]. Our analysis of IGC-2 considered whether and how the pre-existing ocean governance regime constrains or enables the nascent BBNJ agreement [4]. In this analysis, we explore how the negotiations seem to have reverted back to the dichotomy between the common heritage of mankind and the freedom of the seas, despite attempts to set these principles aside in favor of less polarized alternatives, and how we therefore appear to be stuck in the middle, with not much time left before the BBNJ instrument is due to be completed.

Methods
The findings presented in this analysis are part of a larger, on-going project addressing the governance of BBNJ. Our analysis draws on semistructured interviews, participant observation at the conference (including working group interventions and side events) and related "process tracing", which aims to identify sequences and patterns that support the development of theoretical ideas. Our overall goal is to construct an explanatory narrative that sheds light on the BBNJ process and outcomes, in particular the factors that explain the final outcome of the negotiations, placing them within the larger literature on regime creation and effectiveness [7,8]. We also conducted a desktop analysis, reviewing documents connected to the meeting, e.g. statements, official documents, and CRPs circulated at the conference, containing delegates' draft language changes and amendments, as well as Earth Negotiations Bulletin reporting and other reports from IGOs, NGOs, and the media. Quotations provided in this paper are from a dataset we developed and have been building at the IGCs, and have been verified with written statements and video, where available. At the request of the BBNJ conference leadership, this analysis does not draw directly from our observations of the informal informals. These sessions were attended as much as possible by the authors, however, in order to follow overall trends in the topic discussions.
The processes we are tracing represent influences on the emerging institutional design of the BBNJ agreement. The nature and content of regime design is critically important to achieving the effective governance of marine biodiversity, therefore tracing the factors that influence the design process can be useful for producing explanations about why particular regimes succeed or fail. Although the BBNJ regime is still coming into being, the framework for analyzing regime features provided by Ref. [9] provides a useful guide for categorizing and characterizing various proposals and trends. They argue that international agreements can be viewed as 'hard' or 'soft' depending on their degree of precision, obligation, and delegation. Without prejudging the suitability of hard or soft institutional forms for BBNJ governance, we use this framework to support our analysis of the emerging BBNJ regime design. In general, we found that debates over the level of precision required and the degree of obligation and delegation that states found acceptable pervaded the negotiations.

Overarching issues
"We must keep in mind not to … create any new obstacles to fishing or fisheries" Overall, some interesting trends were observed at IGC-3, including continued disagreement over certain issues, as well as new and emerging issues and players. IGC-3 revisited core themes from the first two IGCs, emphasizing the need for an "effective, practicable and future-proofed implementing agreement for the conservation and sustainable use of marine biodiversity in areas beyond national jurisdiction" (European Union delegation's opening statement, 8/19). This recognition of the need for future-proofing was echoed by Tuvalu on behalf of the Pacific Small Island Developing States (PSIDS), who noted that the agreement should be neither too prescriptive nor too broad, leaving too much to be determined by subsequent Conferences of Parties (COPs), which could delay implementation. The United States also warned against deferring too much decision-making to the COP or any new institution, but for a different reason: to ensure that member states would not be bound by any decision not made in the context of these negotiations.
In addition, the continued restriction that this agreement "should not undermine" existing approaches, which has run throughout the IGCs to date, means that the BBNJ treaty will likely not have a hierarchical relationship with other instruments [4,10,11]. In the revised draft text for IGC-4 3 , the positive framing of "promotes coherence and coordination" was eliminated in favor of the simple "does not undermine" framing (Article 4(3)). However, if the new treaty does not address the myriad problems that have emerged since the United Nations Convention on the Law of the Sea (UNCLOS) was negotiated in the 1970s and 80s, it may not succeed. In May 2019, a few months before IGC-3, the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services released a landmark assessment on the state of global biodiversity [12]. The report painted a stark picture of global biodiversity loss, indicating that a million species face extinction, many within decades. Shortly after IGC-3, in September 2019 the Intergovernmental Panel on Climate Change released a "Special Report on the Ocean and Cryosphere in a Changing Climate" [13] which highlighted the dramatic effect that climate has already had on the oceans, including ocean acidification, stratification, and oxygen loss. The importance of these global assessments was not lost at the BBNJ negotiations; several delegates, including the President, referred directly to them. However, whether these larger threats are helping bring state parties together to draft a strong agreement remains to be seen.
The visibility of individual delegations and coalitions also shifted somewhat in IGC-3. Unlike in previous sessions, the representative of the Holy See rarely intervened. The apparent reason for this shift was, in their own words, a feeling that their past interventions, which were intended to provide innovative middle ground solutions, were not picked up by other delegations or reflected in the draft text. The two largest coalitions -G77/China and the African Group -also expressed fewer, and less detailed, consensus positions than in previous IGCs, despite the dedicated efforts of their coalition leaders. The like-minded Latin American states, newly rebranded as Core Latin American countries (CLAM) at this IGC, emerged as a more forceful and focused group compared to previous IGCs. A variety of states spoke on behalf of CLAM, including Colombia, Brazil, Costa Rica, and El Salvador, and in a side event, Honduras noted that the CLAM group was one of the most diverse coalitions at the negotiations, in terms of country type. In their opening statements, CLAM and the persistently vocal CARICOM (Caribbean Community) noted the degree to which they shared positions. Among the coalition groups representing small island states, the PSIDS (Pacific Small Island Developing States) remained the most active, with AOSIS (Alliance of Small Island States) and PIF (Pacific Island Forum) rarely intervening. Five different member states spoke on behalf of the PSIDS this time, and the coalition intervened on every major issue. Their positions -which emphasize the delegation of authority to new bodies, the rights of adjacent coastal states, and special recognition of SIDS status (small island developing states) -seemed increasingly isolated or distinct from the interventions of other states and some other coalitions.
The polarization between supporters of the principles of common heritage of mankind (CHM) and freedom of the seas picked up again during IGC-3, after a lull in IGC-2. As before, this ideological dichotomy was particularly vocalized in the discussions around MGRs and echoed many of the same challenges discussed during the negotiations of Part XI of UNCLOS [34], circling the divergent goals of developing and developed nations [14][15][16]. Although participants at the UNCLOS negotiations had already committed to the CHM principle to govern the resources of the international seabed (the Area), it took many years of heated debate to agree on what exactly that meant for access to and exploitation of seabed resources [17]. Eventually, a compromise was reached for UNCLOS wherein all states could access and exploit the Area, subject to a management and benefit sharing regime administered by the newly-created International Seabed Authority (ISA). The compromise was not ideal for any of the countries, however, and it became a major issue for the United States in particular, even credited with being the reason for the United States' refusal to ratify UNCLOS [18,19]. Historically, maritime powers like the United States, United Kingdom, and Soviet Union preferred the application and expansion of the 'freedom of the seas' principle to ensure their continued ability to exploit ocean spaces and resources. These same general patterns of disagreement about the applicability and meaning of each principle persist in the BBNJ negotiations.
It appeared that the President hoped to side-step this debate, as the draft text did not include an explicit reference to either principle (although some articles did reflect parts of the CHM concept). Developing states emphasized its importance in their opening statements. Palestine, speaking on behalf of the 134 state members of the G77/ China, asserted that the overall goal of conservation and sustainable use of marine biodiversity in ABNJ "can only be achieved when guided by the bedrock principle of CHM" (G77/China opening statement, 8/19/19). Algeria, speaking on behalf of the African Group, said that without an explicit statement of the CHM principle, the agreement would be "like putting a ship in the water without a navigational instrument" African Group opening statement, 8/19/19. Malawi, speaking on behalf of the Least Developed Countries, concurred that the ABNJ as a whole must be recognized as the CHM. Despite these clear group statements, individual delegations also felt a need to stress the central importance of the CHM principle, including Cameroon, Egypt, Eritrea, India, Iran, Myanmar, Nicaragua, the Philippines, Thailand, and Sudan. This deep commitment to the CHM principle among developing states was evident throughout the issue-specific discussions, and the revised draft text for IGC-4 now includes an explicit reference to CHM in Article 5 on general principles and provisions.
Despite this disagreement about guiding principles, other issues were less contentious. For example, there was broad agreement across the package that a BBNJ agreement should only create obligations for state parties. For example, delegations roundly rejected the idea that proponents of a planned activity (which would often include private companies) should be responsible for determining whether an EIA is necessary. Avoiding provisions that directly obligate non-state actors is typical for international agreements, which overwhelmingly focus on international cooperation between states, but many delegates also expressed concern about the wisdom of letting companies self-regulate. The International Cable Protection Committee (ICPC) -the only clear industry voice represented during IGC-3 -repeatedly stressed the importance of involving industry in review or guidance functions, by proposing the addition of "sectoral stakeholders" and "sectoral expertise" in different parts of the text. Although the ICPC representatives remain actively engaged in direct informal talks with national delegates, these positions still lack specific state supporters in the formal interventions.

Elements of the BBNJ package
In addition to these overarching issues, IGC-3 delved further into the four key themes of the BBNJ. Working Groups and informal informals were scheduled to deal with the four agenda items separately, and also a fifth category of 'cross cutting issues' that included questions about institutional architecture. Unlike previous IGCs, where a number of contiguous days were assigned for each issue area, the IGC-3 schedule inter-mixed different issues within single days. These issue areas are treated as distinct in the draft text, and also in the analysis that follows. The topic of Marine Genetic Resources (MGRs), and specifically the creation of rules for access and benefit sharing, remains one of the most contentious areas of the BBNJ negotiations. The most disputed draft articles on MGRs were addressed during informal-informals: access, benefit sharing, monitoring, and intellectual property rights. In part, this is a result of obvious and on-going polarization between the positions of the G77/China coalition and a group of developed states that includes the United States, Japan, South Korea, and Russia. In general, the countries most likely to have nationals engaging in MGR collection and utilization in the near term are those which oppose additional burdens on that activity, and the countries least likely to engage in MGR collection and utilization tend to propose additional regulations and regulatory institutions. Some countries routinely took more moderate positions on MGR issues -such as Norway, Singapore, Canada, and the EU -although their interventions tended to be closer to the antiregulatory pole. Although G77/China remained united that the CHM principle should underlie the regime, and that benefit sharing should be meaningful and significant, the coalition expressed agreement on fewer details of the text compared to other issue areas. According to one knowledgeable interviewee, this relative lack of consensus on MGR topics is a result of both the rushed and changing schedule at IGC-3, and substantive disagreements within the coalition itself.
Interventions on MGR-related topics did change in notable ways compared to IGC-1 and IGC-2. The explicit clash between the principles of 'freedom of the seas' and CHM faded into the background, and was articulated less frequently, during the MGR discussions. This may have been the result of having a draft text for the first time, and a draft text which does not explicitly refer to either principle. But the overall clash of principles still suffused the debate, as the representative from Palestine regularly reminded delegates. The strongest and most explicit version of the 'freedom of the seas' principle came from Russia, which suggested that the treaty text should not include a provision regulating access to MGRs at all. Other states sought to apply the 'freedom of the seas' principle by arguing that the initial in situ collection of MGRs rarely has a commercial purpose, so that type of access is better understood as marine scientific research, a topic which has already been established by UNCLOS as an explicit -if undefined -freedom of the high seas (Article 87). In a debate that is very similar to the longstanding disagreements about freedom of navigation in the territorial sea, the idea of prior notification is seen as an unacceptable hindrance to some (such as Japan, Korea, the United States, and Russia) while others find a simple reporting procedure acceptable (such as the EU, Norway, Singapore, and CLAM) [20].
The clash between 'freedom of the seas' and CHM could also be discerned in the debate over the modalities of benefit sharing. Divergent preferences on the draft text of Article 11 in the CRPs show polarized positions. While the United States and South Korea accept voluntary, non-monetary benefit sharing, the G77/China, CARICOM, and the African Group support mandatory sharing of both monetary and nonmonetary benefits. Both Norway and the EU took middle positions, allowing for some mandatory sharing of non-monetary benefits. Each option -voluntary and/or mandatory, monetary and/or non-monetaryremains in the revised draft text for IGC-4, signifying a lack of movement on this topic in IGC-3. This debate is closely tied to both the topic of MGR access, which can itself be described as a benefit, and the CBTMT issue area, which also contains questions about mandatory and monetary transfers.
Another difference between IGC-3 and previous conferences was a lack of emphasis on the potential riches and rewards associated with commercialization of MGRs. Estimates of the monetary worth of MGRs are highly speculative, given the inherent uncertainties of the research and development process, and the continued lack of research on the distribution and potential value of MGRs in ABNJ and national jurisdictions [21]. This situation creates a major challenge for consensus building, because differing expectations about the potential benefits of MGRs seem to be shaping the cost/benefit analyses of actors with regard to the prospect of creating a heavy or rigid institutional architecture for regulating access and benefit sharing. If you believe that major profits are possible, especially in the near term, the risk of hampering scientific research is worth the potential reward of a well-functioning regulatory institution that ensures real benefit sharing. If you believe that a profitable scientific breakthrough from MGRs in ABNJ is unlikely or will take a very long time, the cost of impeding or delaying scientific research may seem unacceptably high. Although access and benefit sharing are addressed in separate draft provisions, these two regulatory areas are closely intertwined. Another example of this close connection concerns the debate over whether "access" includes the ability to use digital forms of genetic data, rather than just physical samples. Although regulating access to in situ or ex situ physical samples may be simpler from a definitional and practical perspective, the reality is that researchers and companies increasingly use digital information about genetic material that is often available in public databases [22].
Two communities that would be affected by the new rules for access to MGR and benefit sharing had limited presence at IGC-3: the scientific research community, and the industries that utilize genetic resources for product development. Instead, delegates from developed states and some NGOs would speak on behalf of the interests of these groups, typically in broad-strokes comments about the risks and costs of deterring academic and commercial research. Some delegates -including from developing states -referred to conversations with members of their domestic scientific community in order to provide support for their positions. Japan, for example, emphasized the need for scientists to research water quality right away after a major maritime accident, and 4 The MGR portion of the draft treaty (Part II) addresses overall objectives (Article 7), applications of the provisions (Article 8), activities with respect to MGRs of ABNJ (Article 9) collection of and access to MGRs of ABNJ (Article 10), the fair and equitable sharing of benefits (Article 11), intellectual property rights (Article 12), and monitoring (Article 13) concerns about the impact of prior notification requirements on their ability to collect important data. Some NGOs and IGOs reminded delegates that existing institutional resources within the scientific community could assist in the process of access and benefit sharing, a point which has been reiterated in the academic literature [23]. The limited formal voice of the scientific community in these negotiations risks the creation of MGR rules that are challenging to implement in practice. For example, the very idea that there is a discrete and definable category of 'marine genetic resources' is not reflected in scientific research about deep sea organisms, and neither is the distinction between ABNJ and national jurisdiction genetic material [24]. More direct involvement of the scientific community in the negotiations could mitigate the risk of rules that are out of touch with scientific practice and concepts [25]. Researchers can also make a positive impact by voluntary disclosure of the origins of genetic sequence data used in patents, to increase transparency and traceability for regulatory purposes, and to help reveal the links between collection, access, and exploitation [22,26].
The topic of Intellectual Property Rights (IPR) for MGRs is also difficult to resolve, for three main reasons. First, it is difficult to distinguish (legally or practically) marine scientific research from access to and utilization of MGRs. Until the commercialization stage, the activities associated with the collection, storage, and analysis of MGRs are essentially the same regardless of whether the intention is knowledge or profit. And commercially-valuable MGRs can be identified from samples that were not originally collected for that purpose [23]. Although marine scientific research is not defined by UNCLOS, it is an explicit freedom of the high seas (Article 87) and cannot constitute "the legal basis for any claim to any part of the marine environment or its resources" (Article 241). So, any system that regulates access and/or allows patent claims over MGRs must distinguish them from marine scientific research. Second, IPR regimes are typically designed to limit access to patented items and information, to protect the interests of investors and thereby encourage investment in research and development. This idea is in tension with the notion that a BBNJ agreement should provide access to and share benefits from commercial products based on MGRs. The major developing world coalitions strongly support inclusion of an Article on IPR in the agreement, to ensure that the IPR process for MGRs from ABNJ facilitates transparency, accountability, and compliance, especially with benefit sharing provisions. Third, the topic of genetic resources -how they are defined and how they can be patented -is the current subject of two non-BBNJ negotiations, in the World Intellectual Property Organization (WIPO) and the Convention on Biological Diversity (CBD). The discussions in CBD concern the roles of Digital Sequence Information in utilization and benefit sharing of genetic resources, while WIPO is hosting negotiations for an agreement about the relationship between traditional knowledge and IPR. 5 Although both forums explicitly limit the applicability of their agreements to ABNJ, the existence of other forums with related mandates, and the desire for uniformity across international agreements, allows some delegates to suggest that the BBNJ participants should delay, defer, or displace decisions about the IPR aspect a regime for MGR in ABNJ.

Area based management tools, including marine protected areas 6
"The Rio Declaration was in 1992 … colleagues we are in 2019 so I think we should move forward".
-Delegate from Switzerland, referring to the precautionary principle versus precautionary approach, Working Group on ABMTs 8/21 /19 Given this was the third time the international community had formally come together to discuss the role of area based management tools (ABMTs) including marine protected areas (MPAs) in the context of BBNJ, it was somewhat disappointing that delegates continued to struggle to define these tools and how they relate to one another. Most of the state delegates who offered draft language in the CRP documents proposed deleting portions of the definitions for ABMTs and MPAs related to "affording higher protection than that provided in the surrounding areas" for one or both of these tools in Article 1 on Use of Terms. Some also suggested removing "biodiversity" and/or "sustainable use" from the definitions, such that an MPA would simply aim to achieve "long-term conservation objectives." But even the "long term" was challenged, as Russia and others emphasized the need to re-evaluate any designated ABMT after a certain time period, in order to prevent the continuation of these tools after their goals have been achieved.
A related sticking point focused on the distinction between "establishing" versus "designating" MPAs, with Canada, CLAM countries, the G77/China, Iceland, the Philippines, and Turkey advocating for the former, while the EU and Maldives preferred the latter. The relative meaning of these terms in relation to one another within the context of the developing agreement was not clear, and the US suggested substituting both with "identifying" ABMTs and MPAs, while the IUCN offered "adoption and implementation" as alternative language with the clearest and least ambiguous meaning, which the international community would be wise to agree on. This debate over wording remains important, and points to the ongoing, wider need for internationallyrecognized definitions for MPAs [27]. It remains to be seen whether Conference President Rena Lee and/or the subject area facilitator will take the lead on suggesting and defining whatever terms might be chosen, or whether outside action from NGOs, IGOs, or other groups can contribute to a resolution.
The relationship between ABMTs and MPAs as categories also retained some ambiguity. Canada suggested incorporating language on other effective area-based conservation measures (OECMs) in several places, which would be in line with the CBD Aichi Biodiversity Targets. OECMs include multiple-use areas and fisheries closures, as well as private, local, community-managed, or other forms of informal and/or "de facto" protected areas [28]. Together with more strictly-protected areas, they can contribute to networks of ABMTs and MPAs, but an official definition of an OECM is lacking, further complicating the ABMT/MPA definition issue. The IUCN World Commission on Protected Areas has created a task force on the subject, which recently published a Technical Report for Recognising and Reporting OECMs. 7 Including these areas significantly helped boost Canada's progress towards attaining the Aichi Biodiversity Targets, in both terrestrial and marine 5   contexts [29], and as of 2019 they have surpassed the 10% target for marine protection, and are currently at 13.81%. 8 This category of ABMTs can complement MPAs but is not a substitution for long-term, effective conservation measures. The underlying criteria to be used to justify designating ABMTs and MPAs also continued to face disagreement, with some states hoping to include climate change and ocean acidification considerations outright (and even ocean noise pollution in a few cases) while others pushed back, arguing that criteria such as vulnerability, fragility, sensitivity, and slow recovery already take these pressures into account. A few delegates preferred not having criteria outlined here at all, but leaving it to the scientific/technical body to determine and then provide guidelines. Debates over whether to include socioeconomic factors were less prevalent than at the first two IGCs. As in previous IGCs, Eritrea gave an explicit defense of including socioeconomic factors, an issue which the delegate described as "close to our heart" (Working Group on ABMTs, 8/ 21/19). Other key issues that arose in the discussions included whether and where to include traditional knowledge from indigenous peoples and local communities, and whether ABMTs and MPAs should be timebound and adjustable; the latter considerations were favored by Russia, China and the USA.
Interestingly, as mentioned earlier, the ICPC emerged at IGC-3 as a clear voice for the interests of its industry. In the ABMT/MPA discussions, their contributions to CRPs requested inclusion of language acknowledging (1) the importance of "facilitation of international communications, in particular for small island developing states" (Article 14), (2) their role as "sectoral stakeholders, such as the owners and operators of existing and planned submarine cables" (Article 15), and (3) that proposals include "the coordinates of submarine cables", whether existing or planned (Article 17). These proposed amendments to the draft text reflect the general trend of UNCLOS as well, which recognizes the goal of facilitating international communication in the preamble, and which explicitly carves out rights for submarine cable laying in most maritime zones.
The revised draft text for IGC-4 has moved the criteria on identifying areas (Article 16) to an Annex, and it now includes language stipulating a time duration for a proposed area and measures in proposals (Article 17). It also more clearly points to the scientific/technical body's role in reviewing proposals (Article 18) but is less clear about whether the Conference of Parties will have the authority to take decisions on matters related to ABMTs (Article 19).
Moving forward towards IGC-4, critical questions remain about how to designate ABMTs and MPAs when relevant instruments/bodies already exist and/or where they are lacking, as well as how these instruments/bodies should coordinate with one another, e.g. the OSPAR network of MPAs and closures under regional fisheries management organizations [30]. The issue of adjacency also remains contentious, i.e. what role will coastal states have in decision-making regarding the adoption and implementation of ABMTs/MPAs adjacent to their maritime territories, and how/will their views be taken into account? Additionally, including some dynamic approaches to ABMTs and MPAs in ABNJ, such as mobile and adaptive sites, could be a thoughtful and practical way forward, which will continue be advocated by NGOs in the next IGC [31].
Hurdles also remain with regard to language incorporating the precautionary principle and ecosystem approach, which are both now mentioned in Article 5 of the revised draft text, on general principles and approaches. As evidenced by the quote at the start of this section from the Swiss delegate, the ABMT and MPA portion of the discussion at IGC-3 certainly highlighted the frustration felt by many that we should be further advanced at this stage.

Environmental impact assessments (EIAs) 9
"The EIA decision lies with the state partythe whole process of having an approval [by an international body] is not one that we support." -Delegate from the European Union, Working Group on EIAs 8/22/19 Divisions between states on the proper conduct of environmental impact assessments (EIAs) continued in the third round of negotiations. Much of the conversation revolved around two areas of disagreement: (1) what kind of authority a scientific/technical body would have compared to states parties, and (2) whether the stages of an EIA should be stipulated in the agreement or set out as (perhaps voluntary) guidelines. Neither of these issues were new to IGC-3, but the need to reconcile the disparate positions of states has become more acute as the negotiations head into their final scheduled meeting. Some observers suggested that this issue area has seen more progress towards consensus agreements, but it may be simply that the decisions that need to be made about the process are especially clear on this topic.
Most states were in agreement that states parties should be the ones to decide whether an EIA is needed. This, however, is one of the few points of general agreement. States could not agree on whether it was necessary to include in the BBNJ agreement a description of activity or ecosystem characteristics where an EIA would clearly be needed (this list will be prepared by the Conference of Parties as voluntary guidelines (Article 29)). Moreover, states were clearly divided on the process to be followed should a state determine that an EIA was not necessary. Many developing states, including the African Group, the G77/China, CLAM, and the PSIDS, argued that if a state decided an EIA was not necessary, they would need to provide evidence to support that decision. The US and Canada claimed that the deciding state merely had to make the information supporting that decision publicly available; CARICOM, PSIDS, and others stated that the decision to forego an EIA should be confirmed by the scientific/technical body; and Russia retained its opposition to the establishment of any kind of new decision-making body throughout the BBNJ agreement. Article 49 of the draft text discusses the scientific/technical body and highlights some of the possibilities under discussion for this institution. While states are in agreement that such a body should exist, the expertise necessary for its members and the overall role of the body are still up for debate. With such a range of positions and support, it is unclear how easily states will be able to reconcile into any sort of compromise in the upcoming IGC-4.
Likewise, there was controversy over how specific the agreement should be when it came to providing details for the conduct of EIAs. The first of these debates dealt with the issue of scoping. There was a split on whether states parties or the scientific/technical body should define the scope of EIAs. From there, the states moved into a discussion on whether socio-economic and cultural impacts should be included within the scope, with CLAM and PSIDS in favor and the EU and others arguing instead for a more general definition of scope. Similar issues arose on the second day of EIA discussion, in determining the specificity of guidelines 8  to be provided for EIAs. Some states, including the African Group, G77/ China, CARICOM, CLAM, and the EU preferred a list of mandatory information to be included in EIAs in the agreement. Russia called for an annex with an indicative list, and the Republic of Korea called for a detailed, voluntary list to be developed later. There was also a split on who should be responsible for the reviewing of reports, states parties or the scientific/technical body. This recurring debate about whether a new body should be created and empowered with a review and/or approval role sits at the heart of the EIA issue -who is in charge of overseeing new activities in ABNJ? This represents a basic question about institutional design: how much delegation of decision-making is necessary or acceptable?
Because EIAs would likely create a new burden or regulatory hurdle for private actors, the lack of diverse industry representation was notable. But once again, the ICPC advocated tweaks to the draft text in CRPs that would minimize obstruction of industry activities in general. The ICPC emphasized the need for minimizing "duplicative and inconsistent requirements" during discussions on the relationship of the draft treaty to other instruments (Article 23), as well as the "expected contribution to sustainable development" with respect to the scope of the agreement (Article 31). With respect to impact assessment and evaluation (Article 32), the ICPC suggested language indicating that the State Party (or proponent) "shall be authorized to define a planned activity as to include future contingencies, such as the maintenance and repair of submarine cables, which shall not require a separate environmental impact assessment absent exceptional circumstances." Such disparate and entrenched positions on the authority of states versus the authority of a scientific/technical body as yet to be created indicates that agreement in IGC-4 may be hard to come by. Discussions about the design of any scientific/technical body have been stymied by disagreement about what its proper functions would be. Nor is it clear if the working group will be able to come together on the important issue of what should be included in an EIA, and if a detailed or more general approach to such guidelines would be more favorable. While the general agreement for a scientific/technical body is there, there remain many details still to be worked out.

Capacity building and transfer of marine technology (CBTMT) 10
"Our group will put additional emphasis on topic of CBTMTthese fine words are enshrined in UNCLOS but their implementation has fallen far short of the expectations of developing countries." -Palestine on behalf of the G77/China, opening statement 8/19/19 At the previous IGC (IGC-2), the discussion about CBTMT had moved toward whether it would be monetary vs. non-monetary, and mandatory vs. voluntary, and delegates largely fell into traditional categories of developed vs. developing states, with Russia and the US both strongly favoring voluntary and non-monetary options [4]. During IGC-3, the special circumstances of SIDS and LDCs were emphasized again to answer these objections to putting money on the table, and the representative from the G77/China emphasized that "the idea here is to make sure that all countries who want to fulfill their rights and obligations under this instrument are able to and have the opportunity to have the capacity building and technology they need to do just that" (Working Group on CBTMT, 8/20/19). This is especially true, said a member of one of the groups, if the aim is to reduce the scientific and technological gap between developed and developing states. Tuvalu emphasized this in pointing out that for instance the PSIDS in combination shared the stewardship of 20% of the EEZs in the world -but only one of the countries has an oceanographic vessel and only one has government staff with expertise in oceanographic issues. This plea for a fulfillment of capacity building and technology transfer from developing countries was not something that developed countries necessarily were against during IGC-3 either, though they in general were clear on it still having to be voluntary and non-monetary. Especially Russia emphasized their known position, under which cooperation between nations should be strictly voluntary and without any kind of legal obligations. How much the oxymoron "voluntary commitment" could be expected in practice was not explicitly addressed in discussions, although many states emphasized past or on-going transfer programs as examples of this.
Another contentious issue was the reference to "developing middleincome countries" in the draft text. This reference and whether or not special attention should be given to those countries, was something that especially the US, supported by Canada, was against. They stated that "… special attention should be given to those most in need such as the LDCs -[we] do not believe middle income countries face similar challenges and that they should be singled out for special treatment in this regard" (Working Group on CBTMT, 8/20/19). This was also reflected in the CRPs on CBTT, where the US for example suggested striking out "… and developing middle income countries" all four times that it was mentioned. The EU and its member states similarly suggested striking it out, though only twice, and Canada wanted it struck out once. Sri Lanka, however, emphasized during the negotiations themselves that 73% of world's poor live in these countries and that they themselves would be one of these countries in the near future, but have very poor knowledge and technological ability to access and utilize MGRs. This was echoed by Iran, which stated that they could not accept a deletion of this term, and Palestine on behalf of the G77/China, who would not favor such deletion. Bangladesh and Togo also supported one other in retaining this reference, and none of those that gave input in CRPs wanted it struck out. One can speculate that a reason why the US opposed this, for example, could be due to the World Bank's classification system, which places inter alia China and Russia in this category (upper middle income) as well as India (lower middle income) [32].
The topic of CBTMT is closely connected to other aspects of the BBNJ agenda as well. For example, support for MGR access is sometimes described as a form of capacity building, while benefit sharing might include the transfer of marine technology related to MGR utilization. Developing states, especially SIDS, often argue that capacity building is critical to their ability to monitor and enforce ABMTs, or to evaluate the quality of EIAs. Despite these important connections, the CBTMT issue area has made perhaps the least amount of progress since IGC-1. The debate has not gotten more nuanced, and the issues have not been revealed as more complex than anticipated. Rather, countries and coalitions seem to be entrenching further into their existing positions, placing themselves on either side of the monetary vs. non-monetary and mandatory vs. voluntary figurative chasm.

Conclusion
The new BBNJ agreement is intended to connect and coordinate fragmented governance institutions to ensure the conservation and sustainable use of marine biodiversity in ABNJ [33]. The debate on cross-cutting issues, however, has revealed two major obstacles to achieving coherence and synthesis among fragmented institutions. First, the "should not undermine" commitment -detailed in our analysis of IGC-2 [4] -has been consistently deployed throughout the four issue areas to argue that a new BBNJ instrument should not be empowered with any oversight or coordination functions in its relationship with existing institutions. This means that biodiversity conservation in ABNJ must be achieved without the BBNJ treaty itself exerting any direct control over shipping or fishing activities. Second, the status of the BBNJ as an 'implementing agreement' has only two precedents to rely on: the Part XI agreement and the Fish Stocks Agreement, both referred to as 10 The CBTMT portion of the draft treaty (Part V) addresses overall objectives (Article 42), cooperation in CBTMT (Article 43), modalities for CBTMT (Article 44), additional modalities for TMT (Article 45), types of CBTMT (Article 46), and monitoring and review (Article 47) 'implementing agreements.' But the Part XI agreement effectively fused with UNCLOS, whereas the Fish Stocks Agreement is a freestanding treaty. The discussions about the modalities of a BBNJ instrument strongly suggest that delegates are anticipating an implementing agreement akin to the Fish Stocks Agreement, such that non-members of UNCLOS can fully participate. This represents a missed opportunity to enhance the unity and coherence of the ocean governance regime, and to promote universal participation by states .
Despite these structural limitations, a BBNJ instrument still has the potential to be a valuable tool in achieving the goals of conservation and sustainable use of biodiversity in ABNJ. The draft text provided prior to IGC-3 was an important step towards a treaty, as is the revised draft text released in November 2019. However, a substantial amount of intersessional work will be necessary prior to IGC-4 if the process is to finish on schedule. To support progress in the negotiations, any intersessional meetings and conversations will have to seek new compromises and build consensus across a broad, and geographically distributed, group of delegates. Several inter-sessional meetings that do not directly address the BBNJ agenda, but do so obliquely, or that bring together relevant actors, are already scheduled for the period before IGC-4 in March 2020. These include, inter alia, the Our Ocean Conference (October 2019, Norway), the Global Ocean Social Sciences meeting (November 2019, France), the Ocean Sciences meeting of the American Geophysical Union (November 2019, United States), and the World Biodiversity Forum (February 2020, Switzerland), as well as a number of less formal workshops organized by NGOs and IGOs. Bodies with overlapping work (e.g. the ISA and CBD) will hold various issue-specific meetings in the inter-sessional period, and the relevant WIPO committee is slated to continue its work on IPR for genetic resources, with the goal of submitting recommendations for a new instrument to the UN General Assembly in 2021. Additional meetings with a BBNJ-specific focus are currently being organized among small groups of actors, but whether and how these meetings might supper intra-and intercoalitional compromises remains to be seen.
The emerging BBNJ agreement seems "stuck in the middle" in several senses. While IGC-3 occurred just after the midpoint of scheduled negotiations, the discussions themselves do not feel "halfway" towards a final agreement. And many topics remain stuck between the two competing, and arguably opposite, principles of the Common Heritage of Mankind and the Freedom of the Seas, including the implied and hoped for monetary benefits that would emerge from the application of the former with reference to the potential exploitation of MGRs. Finally, the BBNJ agreement is intended to exist between the current suite of fragmented governance institutions in a way that allows it to enhance coordination and cooperation between them. But the commitment that a new agreement should "not undermine" existing agreements has been interpreted restrictively in many of the issue areas, increasing the likelihood that the BBNJ is boxed into the spaces between agreements as opposed to actively identifying and creating synergies between them. Indeed, although most major coalitions favored the positive framing of "promotes coherence and coordination" in Article 4(3), the revised draft text for IGC-4 has eliminated this text. Thus, at this stage, it seems as if the emergence of a new, effective, and consensus BBNJ agreement faces structural obstacles and persistent disagreements, leaving it stuck in the middle of the path between initiating and concluding this multi-year process.
Author statement "stuck in the middle" The authors are not depositing research data for this article. Though we do rely to some extent on data from a database that is under development (since the project period is not finished before the last IGC has been concluded for the BBNJ treaty), the article is a Viewpoint pieceand not a research article per se.
When the research team concludes the data acquisition period, hopefully by the end of April 2020, and publish a research article on the topic, the database will also be shared our. At that point, it will have been properly coded and made possible for external users to navigate so that it is both accessible and discoverable.