Introduction

The international system has recently been characterised by rising institutional density. Particularly since the 1990s, the growth in the number of international institutions has given rise to the emergence of various types of interinstitutional orders such as nested, embedded or clustered regimes (Young 1996). A more recently identified phenomenon are regime complexes, ‘partially overlapping and non-hierarchical institutions governing a particular issue-area’ (Raustiala and Victor 2004: 279; see also Biermann et al. 2009; Orsini et al. 2013). Regime complexes have been identified in areas ranging from climate change (Keohane and Victor 2011) through refugee governance (Betts 2010), maritime piracy (Struett et al. 2013) and food security (Margulis 2013) to energy governance (Colgan et al. 2012), patents (Helfer 2009) and genetic resources (Raustiala and Victor 2004).

Research on regime complexes is characterised by a discrepancy between state behaviour and institutional outcomes. On one hand, complexes are said to develop relatively stable interinstitutional divisions of labour (Gehring and Faude 2014), to be path dependent (Zelli and van Asselt 2013: 9) or to display status quo biases (Oberthür and Stokke 2011: 311). On the other hand, scholars have identified a number of pertinent cross-institutional strategies which states adopt within regime complexes: switching processes to parallel institutions (forum shopping), shifting regulatory authority over a given issue area from one regime to another (regime shifting), or attempting to undermine a regime by deliberately creating contradictory rules elsewhere (strategic inconsistency; Raustiala and Victor 2004; Alter and Meunier 2009).

With some notable exceptions, the present literature falls short of providing a convincing account of how the two levels of analysis are linked. From an actor-centred perspective, it is far from clear how those potentially disruptive state strategies are reconcilable with institutional outcomes that tend towards stability. Only recently attempts have been made to link state agency systematically to institutional outcomes based on states being members in multiple elemental institutions of a complex having vested interests in institutional complementarity (Gehring and Faude 2014), or being incapable of creating new institutions even when dissatisfied with the status quo (Colgan et al. 2012). Although compatible with the approach pursued in this paper, I emphasise constraints over interests while attempting to specify the exact mechanisms which lead actors with reformist policy objectives to continue cooperation within institutional settings in which they can only achieve limited gains.

Not all regime complexes tend towards stable interinstitutional divisions of labour. While cases exist which are characterised by transformational institutional change and significant institutional proliferation (Rabitz 2014), the objective of this paper is to specify when and why regime complexes are subject to incremental change based on a well-established interinstitutional division of labour. In brief, I propose two explanatory variables: the existence of ‘critical’ actors that cannot effectively be excluded from cooperation because of their contribution to international public goods and/or negative externalities that would result from their exclusion; and those actors’ pursuit of conservative policy objectives relative to the institutional status quo. This leads to a double constraint on reformist actors in terms of employing cross-institutional strategies within a regime complex and cooperating through club settings outside of the complex. Under those conditions, change takes the form of institutional layering, whereby ‘original institutions are left in place, but new elements are added alongside the old system’ (Thelen 2009: 484). The viability of institutional layering results from the elaboration and specification of pre-existing rules, sidestepping the difficulties associated with the negotiation of new sets of rules among actors with strongly heterogeneous interests.

The next section makes the case that there is a mismatch in present scholarship on regime complexity regarding the relationship between state behaviour and institutional outcomes. With complexes frequently characterised by a stable interinstitutional division of labour, forum shopping, regime shifting and strategic inconsistency have, to differing degrees, potentially disruptive effects. This begs the question of how to link the two levels of analysis. In the subsequent section, I present an alternative approach which draws on the theory of public goods, negative externalities and institutional layering to explain when and why states with reformist policy objectives cooperate through institutional settings which, due to the presence of critical actors with conservative objectives relative to the institutional status quo, limit the prospects of goal achievement. The penultimate section applies this framework to the Genetic Resources (GR) regime complex, focusing on three processes of regime formation: a sectoral regime for Access and Benefit-Sharing (ABS) from PGRFA leading to the 2001 International Treaty under the United Nations Food and Agriculture Organization; an ABS compliance regime, culminating in the 2010 Nagoya Protocol on Access to GR and the Fair and Equitable Sharing of Benefits Arising from their Utilisation to the Convention on Biological Diversity; and a sectoral ABS regime for pathogens, giving rise to the 2011 Pandemic Influenza Preparedness Framework under the World Health Organization. The final section concludes this discussion.

Regime complexes, cross-institutional strategies, and institutional change

As institutions constitute stable patterns of interaction, institutional change is a perennial problem in social sciences. For regime complexes, sets of non-hierarchical and overlapping institutions jointly governing an issue area (Raustiala and Victor 2004), stability of interinstitutional relations has been identified as a common feature. Frequently, they are ‘characterized by a division of labour among the component institutions that reflects a specific configuration of these institutions within a given institutional complex’ (Gehring 2011: 228). Once in place, interinstitutional divisions of labour usually display a stability which is ‘remarkable in view of the paucity of strong instruments for joint or overarching interplay management’ (Oberthür and Stokke 2011: 322). Complexes are said to be ‘sticky’, as ‘[c]onstructing a coalition sufficient to impose large institutional changes, or to create major new institutions, is very difficult’ (Colgan et al. 2012: 120). Further, a frequent assertion is that complexes are path dependent (Oberthür and Pozarowska 2013: 114; Orsini et al. 2013: 37; Struett et al. 2013: 95; Zelli and van Asselt 2013: 9–10).

Accounts of interinstitutional stability are at odds with what scholars have identified as the major state strategies within regime complexes. Although there is disagreement about the differential effects of complexity on strong and weak actors, respectively, states are said to exploit systematically the overlapping institutional mandates (Alter and Meunier 2009; Helfer 2009). Three chief strategies have been identified in the literature, the conceptual boundaries between which are not necessarily clear-cut. Forum shopping has been defined as actors selecting ‘the international venues based on where they are best able to promote specific policy preferences’ (Alter and Meunier 2009: 16); regime shifting refers to states moving ‘from addressing problems through one regime to addressing those problems through an alternative parallel regime, possibly relocating the most relevant politics for a given issue-area from one regime to another’ (Betts 2010: 14; see also Helfer 2009); and strategic inconsistency results when states ‘attempt to force change by explicitly crafting rules in one elemental regime that are incompatible with those in another’ (Raustiala and Victor 2004: 301–02).

The degree to which states may use such strategies successfully may be limited, considering that institutional outcomes are frequently characterised by high stability of interinstitutional divisions of labour. The pursuit of narrow policy objectives through forum shopping has the most limited impact on the division of labour, as the distribution of regulatory authority between elemental institutions is not affected. Forum shopping also appears to be a relatively frequent strategy, with ample anecdotal evidence to be found in the literature (Braithwaite and Drahos 2000; Alter and Meunier 2006; Busch 2007; Murphy and Kellow 2013). There is less empirical evidence of regime shifting and, in particular, strategic inconsistency, and reasons to dispute their relevance on theoretical grounds. Redistributing authority in a complex via regime shifting aims at changing the existing division of labour. Although this does not appear to be a frequent phenomenon, individual instances have been identified (Helfer 2009). Finally, strategic inconsistency does not aim at redistributing regulatory authority from one institution to another, but rather seeks to contest a given institution’s authority for isolated issues or a larger issue area. Theoretically, this disrupts the division of labour by creating legal uncertainty and conflicts about the distribution of authority. Originally understood as one major consequence of regime complexity (Raustiala and Victor 2004; Alter and Meunier 2009), recent scholarship suggests that the degree to which states resort to strategic inconsistency may have been overrated (Oberthür and Stokke 2011). With strategic inconsistency likely to result in institutional dysfunction, states furthermore jeopardise the ‘collective goods produced by the elemental institutions to which they are a member’ (Gehring and Faude 2014: 7). Finally, we can observe that states go to great lengths to avoid rule inconsistency, with the interaction between multilateral environmental agreements and trade agreements being one salient example (Axelrod 2011).

Forum shopping, regime shifting and strategic inconsistency thus differ in terms of their respective intrusiveness into an interinstitutional division of labour, the first having only minimal impacts, the latter two being potentially disruptive. While this sits uneasily with the aforementioned arguments on path dependence and the status quo bias of regime complexes, the next section proposes scope conditions for when those cross-institutional strategies are subject to constraints, resulting in change in the form of institutional layering.

Regime complexes and critical actors

My conceptual framework revolves around two explanatory variables: the distinction between ‘conservative’ and ‘reformist’ policy objectives, designating the extent to which an actor prefers changes from the status quo (Meunier 2000); and the distinction between ‘critical’ and ‘non-critical’ actors, with the former being able to provide particularly large shares of international public goods and/or jeopardising the realisation of joint gains through negative externalities when excluded from cooperation. I argue that the existence of critical actors with conservative policy objectives constrains the scope for change from the institutional status quo within a regime complex in two ways: it eliminates the possibility of club cooperation among reformist actors through new institutions and it restricts the use of cross-institutional strategies by reformist actors. This leads the latter to pursue their objectives through institutional layering, that is, by gradually building upon and elaborating pre-existing rules while attempting to ensure the participation of critical actors in cooperative arrangements.

Critical actors and conservative policy objectives

I understand ‘critical actors’ as actors that cannot effectively be excluded from cooperation. Cooperation among subsets of actors can lead to pareto superior outcomes as soon as the number of participants reaches a minimum threshold (Schelling 1973). Nevertheless, asymmetrical resource endowments make viable coalitions depend less on how many, and more on which actors cooperate (Snidal 1985: 597–602). I consider an actor as ‘critical’ based on its contribution to international public goods (Barrett 2007) and/or the potential negative externalities its exclusion would generate for cooperating states. An actor cannot be effectively excluded if the realisation of pareto superior results from cooperation depends on its contribution to international public goods, under which I understand both goods that are non-excludable and non-rivalrous, as well as club goods characterised by varying degrees of excludability (Samuelson 1954; Buchanan 1999). Equally, an actor’s participation is critical if its absence imposes negative externalities on cooperating states that are sufficiently large to offset the latter’s joint gains. Although conceptually distinct, public goods and negative externalities are, in practice, frequently two sides of the same coin. That is, the exclusion of critical actors may either lead to joint gains from cooperation being insignificant or substantially reduced.

While critical actors enjoy large amounts of influence because of their indispensability for international cooperation, their constraining effect on institutional change addressed in this paper arises specifically from their pursuit of conservative policy objectives. ‘Policy objectives’ constitute a continuous variable with different actors preferring different degrees of change from the status quo. Actors with conservative objectives may prefer minor institutional adjustments or make concessions in order to realise gains on other issues as part of a package deal (Sebenius 1983). The term thus does not necessarily imply a strict preference for the status quo. Moreover, even actors with conservative policy objectives have stakes in sustaining international cooperation, frequently requiring some form of compromise. Where such actors cannot effectively be excluded from cooperation, they enjoy overproportional influence, though usually without being able to determine unilaterally the institutional outcomes.

Constraints on reformist actors

The presence of critical actors with conservative policy objectives constrains the range of strategies which reformist actors have available for effecting institutional change. Those constraints relate to the viability of exclusive cooperation among reformist actors, as well as to the use of cross-institutional strategies within the elemental institutions of a regime complex:

  • It is well-acknowledged that dissatisfied subsets of actors may opt to create new cooperative arrangements outside of the existing institutions (Urpelainen and van de Graaf 2015). Regime complexes are said to be stable over longer periods of time because of the often prohibitive transaction costs required for institutional innovation (Colgan et al. 2012). While transaction costs matter, an additional reason why dissatisfied actors may be unable to cooperate through alternative settings is their inability to realise significant gains from cooperation if critical actors are excluded. For instance, while Small Island Developing States have a particularly pronounced interest in mitigating global greenhouse gas emissions, even the most ambitious emissions targets within a club setting would be insufficient for reducing average temperature increases in the absence of efforts by major emitters. In contrast, a group of major emitters is in principle able to realise pareto superior outcomes through club settings and in the absence of mitigation efforts by minor ones; and use of Border Carbon Adjustments could ameliorate negative externalities from non-members (Weischer et al. 2012).

  • Where club cooperation is not feasible, reformist actors may attempt to exploit overlaps among those elemental institutions of a regime complex in which critical actors enjoy membership. In this case, the latter enjoy considerable influence over institutional outcomes. Such influence extends both to decision-making processes, as well as their decision whether or not to accede to new international treaties or to ratify amendments to the existing ones. For instance, as the fourth section shows, various possible venues exist for the creation of an ABS compliance regime. Such a regime requires the participation of actors with large numbers of commercial users (such as biotech companies) within their respective jurisdictions. While those actors cannot effectively be excluded from cooperation, they generally refuse to consent to ABS compliance linked to patent law. Accordingly, reformist actors’ attempts at regime shifting failed in the context of the WTO, various settings within the World Intellectual Property Organization and the Convention on Biological Diversity.

Taken together, the infeasibility of cooperation outside of pre-existing institutions, as well as the difficulties associated with forum shopping, regime shifting and strategic inconsistency within a complex reduce the scope which reformist actors have available for inducing institutional change. In principle, though, reformist actors may target elemental institutions which employ non-consensual decision-making procedures in order to outvote a minority of critical conservative actors. However, the latter may ‘invisibly weigh’ the decision-making process (Steinberg 2002) or simply avoid changing their behaviour in relation to decisions which are not legally binding. Most importantly, non-consensual decision-making procedures are irrelevant in cases where reformist actors require critical ones to enter into new legally binding agreements, or to amend the existing ones. Here, critical actors can simply refrain from ratification, thus avoiding being legally bound by decisions with which they do not concur.

The possibility of concluding package deals offers another possibility for reformist actors to alter the status quo substantially. Through linkages, critical actors may be enticed to consent to far-reaching changes even on those issue on which they pursue conservative policy objectives (Sebenius 1983). Whether such package deals are available or not depends on the extent to which an issue area allows for the construction of linkages (Keohane and Victor 2011: 9) as well as critical actors’ preference ranking. Where critical actors hold conservative views on a high-priority issue, reformist actors may be unable to offer sufficient concessions in other areas to effect abrupt changes to the institutional status quo. However, reformist actors may employ issue linkages effectively in order to attain incremental changes.

Institutional layering

When the existence of critical actors with conservative policy objectives constrains the strategies available to reformist actors, institutional change can take the form of cross-institutionally consistent layering. The concept of institutional layering has been defined in multiple ways and has been subject to varying interpretations (van der Heijden 2011). I understand it as ‘the introduction of new rules on top of or alongside existing ones’ (Mahoney and Thelen 2010: 15). The chief difference between how the concept of institutional layering has been used traditionally in comparative politics and how it may be used for the study of regime complexes lies in the existence of multiple, overlapping institutions. Instead of creating new layers on top of a singular institution, layering in regime complexes involves the creation of rules which build upon the rules of the overall complex in order to avoid, or at least minimise, ‘countermobilization by defenders of the status quo’ (Streeck and Thelen 2005: 23). With reformist actors being constrained in their ability to employ cross-institutional strategies within a complex while lacking viable alternatives to cooperation with critical actors, incremental rule development is more likely to meet with the approval of the latter, as new institutional layers build upon pre-existing commitments.

In order to distinguish whether the formation of a new regime within a complex constitutes an instance of incremental or abrupt change (Streeck and Thelen 2005: 9), I draw on the distinction between norms and principles on one hand, and rules and decision-making procedures on the other (Krasner 1982). I consider change to be ‘incremental’ when new regimes incorporate the norms and principles of earlier ones. As the fourth section shows, new regimes in the GR regime complex incorporate all of the relevant norms and principles of pre-existing ones, differing primarily in their specific ways of implementing them.

The concept of institutional layering is closely linked to path dependence (Pierson 2004). The latter concept is frequently applied in the literature on regime complexes, although scholars disagree on whether path dependence results from the entrenched interests of powerful actors (Zelli and van Asselt 2013: 9), from difficulties associated with institutional innovation (Colgan et al. 2012) or the negotiation of new rules within a dense institutional environment (Raustiala and Victor 2004: 279–80). Some scholars instead prefer the term ‘status quo bias’ (Oberthür and Stokke 2011: 311) or ‘interlocking governance structures’ limiting the potential for disruptive institutional interplay (Gehring 2011). While all those approaches refer essentially to the same idea, using the concept of institutional layering in the manner laid out above allows for systematically linking the micro and the macro levels (Gehring and Faude 2013) by specifying a causal mechanism which constrains reformist actors in changing the institutional status quo.

The concept of institutional layering can also elucidate how institutions, or regime complexes, may undergo transformative changes in the long run. While external shocks may lead to regime complexes exhibiting long periods of stability interspersed with rapid change (Colgan et al. 2012), incremental processes of change can contribute to gradual transformations emerging over time (Streeck and Thelen 2005: 8–9). Thus, besides linking institutional outcomes at the complex level to constraints at the unit level, institutional layering may also provide an explanation of transformative change as a result of endogenous factors playing out over time.

Institutional layering in the GR regime complex

The GR complex is ‘almost indisputably recognised as such in the literature’ (Morin and Orsini 2014: 304–05). Focusing on regime formation within the complex, I show that the presence of critical actors with conservative policy objectives has constrained reformist actors’ ability to employ cross-institutional strategies and to create alternative settings for club cooperation. At the core of the GR complex, there are international norms regarding Intellectual Property (IP) protection and the obligation of states to ensure that users of such materials share the benefits arising out of their utilisation ‘fairly and equitably’ with those countries exerting state sovereignty over them (Raustiala and Victor 2004; Oberthür and Pozarowska 2013). First, the 1995 Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) under the WTO and the International Convention for the Protection of New Varieties of Plants (UPOV Convention) define the minimum standards for member states’ domestic legislation on, respectively, the IP protection of inventions and plant varieties. Second, the 1992 Convention on Biological Diversity (CBD) provides a general framework for ABS from GR based on the principle of state sovereignty over biological materials. Below, I show how three regimes within the GR complex have been layered on top of pre-existing rules, how layering results from reformist actors being constrained by critical actors pursuing conservative policy objectives, and how this is contributing to the gradual transformation of ABS governance. In order to show that the mechanism postulated in the third section is operational, I focus on the relative inability of reformist actors to attain their respective policy objectives in those instances where it contravened the objectives of critical actors. I also employ counterfactual analysis (Fearon 1991) to explain why reformist actors dissatisfied with but unable to significantly change the status quo did not resort to club cooperation outside of the existing institutions.

The formation of a sectoral ABS regime for PGRFA

The Food and Agriculture Organization’s (FAO) 1983 International Undertaking had established a non-binding framework for unrestricted access to plant GR. Developing countries quickly came to regard the growing number of IP claims on seeds as limiting their access to breeding materials (Brand et al. 2008: 111–13). The revised 1991 UPOV Convention and the 1995 TRIPS agreement ratcheted up international standards for IP Rights . In 1992, the CBD established a system of state sovereignty under which ‘the authority to determine access to genetic resources rests with the national governments’ (Article 15.1). Users of GR would be obliged to obtain a provider country’s Prior Informed Consent before accessing a GR, while sharing with it monetary and/or non-monetary benefits on Mutually Agreed Terms, essentially a bilateral contract. As the CBD’s bilateral approach was incompatible with the nature of agricultural breeding, the International Undertaking required revision (Andersen 2008: 94–98), leading to the creation of a network of seed banks granting ‘facilitated’ access to certain Plant GR for Food and Agriculture (PGRFA) in 2001.

Critical actors

Capacities for the commercial utilisation of PGRFA, as well as the availability of materials in ex situ germplasm collections, vary strongly among countries. Whereas public sector plant breeding has been in worldwide decline since the 1980s, a wave of corporate mergers and acquisitions since the 1990s led to a handful of global players enjoying dominant market positions. As of 2011, the US-based companies alone hold 51.2 per cent of global market shares in seeds, with France, Germany and Switzerland jointly accounting for 20.9 per cent (ETC Group 2013: 6). Such companies are significant potential sources of benefits because of the scope of their commercial activities. The global distribution of breeding materials in ex situ germplasm collections is also highly asymmetric. European countries hold 26 per cent of accessions of PGRFA stored in national seed banks worldwide, with the United States and Canada jointly accounting for 10 per cent (FAO 2010: 56).Footnote 1 In the FAO’s Multilateral System (see below), industrialised countries presently account for 92 per cent of national contributions, with Germany, Canada and Italy alone contributing 21, 19 and 9 per cent, respectively.Footnote 2 The contrast to developing countries is stark. Here, breeding predominantly takes place at the farm level and through public research institutes; agricultural innovation often suffers from insufficient funds, technology and skilled personnel; and seed banks lack the technological sophistication found in industrialised countries.

Policy objectives

While all countries are highly interdependent regarding the exchange of breeding materials (Andersen 2008), their motivations for creating a new international regime differed. Seed companies from industrialised countries were concerned about access restrictions due to the CBD’s principle of state sovereignty. Conversely, developing countries regarded IP-related restrictions as the chief problem (Brand et al. 2008: 115–16), while also seeking to include benefit-sharing in the revised International Undertaking. Throughout most of the negotiation process leading to the conclusion of the International Treaty in 2001, the latter had argued for limitations to IP claims. In the FAO, it was proposed that no rights limiting facilitated access and/or utilisation shall be claimed (CGRFA 2000) or that IPRs should be non-applicable in the respective PGRFA’s country of origin (ENB 1998: 5). Within the WTO, an African Group proposal suggested amending TRIPS in order to recognise the ‘principles, objectives and measures’ of the CBD and the emerging FAO regime (WTO 1999: 5). For industrialised countries, limitations to IP claims were out of the question, with the United States, for instance, preferring an open access regime without any forms of obligations on the recipients of materials (CGRFA 1997: 48–52).

Institutional outcomes

The FAO’s 2001 International Treaty established a Multilateral System of ABS for crops listed in its Annex I which are ‘under the management and control of the Contracting Parties and in the public domain’ (Article 11.2). The system grants facilitated access to samples under a Standard Material Transfer Agreement. Where PGRFA received from the system have undergone innovation and are subsequently patented, recipients must pay 1.1 per cent of commercial profits to the System, to be subsequently shared with farmers, particularly in developing countries (Gerstetter et al. 2007). The regime thus does not restrict IP claims on Annex I materials beyond the pre-existing requirements under UPOV or TRIPS (Gerstetter et al. 2007), as the EU re-affirmed in its instrument of approval. Reformist actors succeeded, though, in layering the CBD’s benefit-sharing objective on top of the revised International Undertaking by conceding to industrialised countries’ demands for the patentability of enhanced PGRFA originating from the System (Brand et al. 2008: 120–21). In part, because of its institutional design, no payments have been received so far (Moeller and Stannard 2013).

Limits to club cooperation

Why did developing countries choose to cooperate with the industrialised ones when the prospects for fundamental institutional change were intrinsically limited? An exclusive ABS regime for PGRFA among developing countries could, in principle, have incorporated significant restrictions of IPRs, with TRIPS Article 27.3b allowing for legal consistency with the WTO rules. However, this would have foregone facilitated access to industrialised countries’ seed banks, as well as the potential sharing of any benefits generated by their domestic breeders. The infeasibility of club cooperation among developing countries, then, narrowed the scope for institutional change considerably, as critical actors with large ex situ collections and strong domestic seed companies necessarily had to be part of an eventual agreement. The latter, by conceding to benefit-sharing provisions limited in scope and depth, were thus able to ensure that Annex I PGRFA continue to be available for IP protection in accordance with UPOV and TRIPS.

The formation of an ABS compliance regime

The CBD aims to generate financial flows and other benefits from the private utilisation of GR and resulting patent-protected inventions. Yet, users have frequently failed to obtain the provider countries’ required Prior Informed Consent or share benefits on Mutually Agreed Terms (Robinson 2010). The fundamental problem is that the effective implementation of the benefit-sharing objective requires legislative changes within major user countries which have little economic incentives to oblige domestic industries to share parts of their profits or technologies with third countries.

Critical actors

Genetic raw materials and the capacities for their commercial utilisation are roughly divided among North–South lines. OECD countries outclass non-OECD ones in terms of the number of biotech companies, their size, R&D expenditures and sales volumes, with particularly the United States, Japan, Canada, France, Germany and Spain enjoying strong positions (van Beuzekom and Arundel 2009). Publicly-traded biotech companies concentrate overwhelmingly in the United States and Europe, with 64 and 24 per cent of the global total (as of 2008; Hugget et al. 2009: 716). From 2006 to 2010, the United States accounted for 37 per cent of worldwide patent applications in biotechnology, and 31 per cent in pharmaceuticals; Canada for 9 and 11 per cent and Germany for 9 and 8 per cent, respectively. With the exception of China, patent applications from non-OECD countries in those two areas are negligible (WIPO 2012a: 73).Footnote 3 Simultaneously, the highest levels of plant biodiversity are estimated to be in the Andean mountains, the Amazon basin, the Brazilian Atlantic coast, central and South Africa, Madagascar and Southeast Asia (Kleidon and Mooney 2000).

Policy objectives

ABS compliance entails cooperation between provider countries which seek to obtain benefits from their domestic GR, yet are unable to regulate effectively their utilisation, and user countries aiming to ensure domestic companies’ and researchers’ access to GR without having significant incentives to force them to share the benefits they generate. From the late 1990s, developing countries rich in biodiversity proposed numerous versions of a patent law-based ‘disclosure of origin’ requirement in a variety of international forums. Such a requirement would oblige patent applicants to provide evidence that they are utilising GR in compliance with Prior Informed Consent and Mutually Agreed Terms. Failure to do so could imply the non-processing of patent applications and, where fraudulent behaviour is discovered ex post, patents might be revoked (Hoare and Tarasofsky 2007). The US considered such a regime to interfere unduly with innovation (WTO 2004). The EU was initially willing to consider a global tracking system outside of patent law (WTO 2003), later offering a disclosure requirement leading to the non-processing of patent applications in case of non-compliance as part of a package deal incorporating Geographical Indications (WTO 2008).

Institutional outcomes

Developing countries proposed different versions of a disclosure requirement under the 2000 Patent Law Treaty; the (suspended) Substantive Patent Law Treaty; in the reform process for the Patent Cooperation Treaty; within the World Intellectual Property Organization’s (WIPO) Intergovernmental Committee on GR and IP, Traditional Knowledge and Folklore; and as an amendment to either Articles 27 or 29 of the TRIPS agreement (Carr 2008). Disclosure also figured prominently in the negotiation of the CBD’s Nagoya Protocol (ENB 2009). The Protocol, concluded in 2010, obliges contracting parties to ensure that GR utilised in their jurisdictions have been acquired in accordance with the laws and regulations of the respective provider country and that benefits are being shared (Wallbott et al. 2014). The Protocol contains provisions on monitoring, access to justice and legal recourse, on an international certificate of compliance as well as national ‘checkpoints’, which are to collect or receive user information in order to verify compliance. Notably, it does not require contracting parties to link ABS compliance to patent law in their implementing legislation, thus staying far behind the demands which developing countries had been voicing since the late 1990s. As the United States is not a party to the CBD, the EU had the only significant market for GR to be regulated by the Protocol, making its participation vital for an effective outcome.

In order to create an effective ABS compliance regime, providers of GR required the cooperation of those countries where benefits are generated. With (monetary) benefit-sharing being a zero sum game, the latter had little incentives to enter into such an agreement on its own merits. The disclosure requirement which the EU had proposed under the 2008 package deal in the WTO was more in line with developing countries’ preferences than the weaker provisions of the Nagoya Protocol, adopted as a package deal incorporating the post-2010 Strategic Plan for biodiversity (Wallbott et al. 2014). While demonstrating the relevance of issue linkages for enticing critical actors to make concessions on issues on which they hold conservative policy objectives, institutional outcomes also highlight the limited degree to which reformist actors are able to achieve their own objectives when effective cooperation hinges on the participation of the former.

Limits to club cooperation

Club cooperation among countries rich in biodiversity would have failed to secure the sharing of benefits without being able to prevent the misappropriation of GR effectively because of the inherent difficulty of monitoring illegal transboundary movements of biological materials or even transmissions of digitalised genetic sequence data (Rabitz 2015). The exclusion of critical actors would have failed to secure participation in the benefits generated by their domestic users, with negative externalities resulting from unregulated access to and utilisation of GR (Robinson 2010).

The formation of a sectoral ABS regime for pathogens

The risks posed by pandemic influenza captured the headlines in 2006/2007 with outbreaks of H5N1 ‘Avian Flu’ and H1N1 ‘Swine Flu’ in 2009/2010. Unlike seasonal varieties, pandemic influenza poses significant risks for global public health (Price-Smith 2009: 57–87). The WHO’s Global Influenza Surveillance Network had, since 1952, provided a system for unrestricted virus-sharing among member states, WHO-designated laboratories and vaccine manufacturers. Its purpose was to survey pandemic threats and allow for the rapid development of vaccines based on the genetic sequence data of viruses in the field. During the Avian Flu crisis, this system was challenged when Indonesia unilaterally ceased the sharing of H5N1 viral strains, arguing that patent-protected vaccines developed on the basis of Indonesian samples were priced at unaffordable levels (Hameiri 2014). This set in motion a negotiation process culminating, in 2011, in the adoption of the Pandemic Influenza Preparedness Framework by the World Health Assembly.

Critical actors

The transboundary nature of pandemic influenza requires international cooperation for surveillance and the rapid development of vaccines based on the genetic profile of the pathogen (Wilke 2013). The surveillance infrastructure clusters in industrialised countries, with five out of six WHO Collaborating Centres and all four Essential Regulatory Laboratories being located in Australia, Japan, the United Kingdom and the United States. Vaccine manufacturers from the United States, France and Switzerland jointly account for 80 per cent of global production (PHRMA 2013: 45). About 76 per cent of claimants of vaccine-related patents under the Patent Cooperation Treaty reside in OECD countries, 32 per cent in the United States alone, with Austria, Belgium, Canada, France, Germany, the Netherlands, Italy and the United Kingdom holding a combined share of 27 per cent (WIPO 2012b: 115). With limited global manufacturing capacities, access to patent-protected vaccines depends on purchasing power (Muzaka 2011: 31–37). Use of advance-purchase agreements, as well as vaccine stockpiling for reasons of national security (Elbe 2014), further impede access by developing countries in case of pandemic threats. With the latter lacking in innovation, manufacturing and surveillance capacities, an effective agreement required the cooperation of critical actors.

Policy objectives

In the wake of the Indonesian controversy, industrialised countries had stakes in maintaining the Global Influenza Surveillance Network of unrestricted access without limitations to IP claims. Particularly the EU and the United States initially opposed any types of linkage between the Network and the CBD. Conversely, developing countries sought to ‘obtain mandatory commitments from the recipients to share proprietary technology and know-how’ (SUNS 2011: 1) with domestic manufacturers. The WHO African Region proposed that providers would retain sovereign rights over pathogenic GR and recipients would not seek IPRs over the material or derived products (WHO 2008: 9–10). In parallel, efforts to bring pathogens under the scope of the CBD commenced in the final stages of the negotiation of the Nagoya Protocol (ENB 2010). By asserting state sovereignty over pathogenic GR, countries with insufficient access to vaccines sought to increase their leverage for attaining favourable benefit-sharing arrangements. Industrialised countries considered this a ‘potential deal-breaker’, undermining ‘the ability of the international community to effectively respond to public health concerns’ (Buck and Hamilton 2011: 58). As a European diplomat succinctly phrased it, ‘we need their virus, they need our vaccine, nobody needs this framework’ (quoted in GRAIN 2009).

Institutional outcomes

The 2011 Framework aims to share viruses with human pandemic potential as well as derived vaccines and other benefits (Kamradt-Scott and Lee 2011). Two SMTAs regulate, respectively, virus-sharing within the network of WHO-designated laboratories, and transfers out of the network, for instance to vaccine manufacturers. SMTA 2 obliges recipients to choose from a list of benefit-sharing options, such as donation of vaccines or antiretrovirals to the WHO, or their reservation ‘at affordable prices’ (Article 4c). Attempts to expand the regime’s scope to cover benefits resulting from the utilisation of viral samples were successful. As in the International Treaty, reformist actors failed to create linkages to the patent system, though: SMTA 1 incorporates only hortatory wording that neither provider nor recipient ‘should’ seek out IPRs (Article 6.1) while SMTA 2 does not contain any IP-related provisions.

Limits to club cooperation

The necessity to ensure the cooperation of critical actors with large vaccine manufacturing capacities and adequate infrastructure for influenza surveillance reduced the viability of club cooperation. The creation of stricter rules regarding patents, or more comprehensive benefit-sharing, would have failed to secure access to critical actors’ vaccines and surveillance capacities.

The gradual transformation of the GR complex

As an incremental process of change, institutional layering may lead to the gradual transformation of institutions (Streeck and Thelen 2005: 8–9). The above has shown that reformist actors’ attempts at shifting the scope of ABS regimes into IP law have failed in every single instance. The layering of rules regarding benefit-sharing on top of pre-existing institutions, on the other hand, was successful. While the depth of the benefit-sharing obligation under both the Framework (Kamradt-Scott and Lee 2011: 838–40) and the International Treaty (Moeller and Stannard 2013) is low, its extension to pathogenic GR and PGRFA, respectively, has two important consequences.

The first is increasing sectoralisation, a trend which continues with the ongoing process towards a regime for marine GR (Tvedt and Jørem 2013). While the Nagoya Protocol allows for specialised ABS regimes to take precedence, this increasingly supplants the overall framework of the CBD with more specific rules tailored for particular issue areas. Second, the extension of benefit-sharing to issue areas which do not allow for time-consuming authorisation procedures and negotiations has led to the use of SMTAs setting out non-negotiable conditions for access through private contracts. Although both the International Treaty and the Framework make reference to state sovereignty over biological materials, in practice, this implies management of public goods through private law and displaces monitoring and enforcement via national governments, as well as their authority to determine the conditions of access (CBD Article 15.1). While both regimes do not depart from the CBD’s principles, it is the specific implementation of those principles which leads to a gradual transformation of ABS governance. A possible consequence of institutional layering is thus a de facto scaling-back of the CBD’s functional scope as well as the central role it foresees for the regulation of ABS by public authorities.

Conclusions

This paper proposed a conceptual framework for explaining why and when regime complexes tend towards stability and incremental change, rather than transformational change as a result of actors’ cross-institutional strategies. I argued that change agents are constrained, first, by the existence of critical actors which cannot effectively be excluded from cooperation because of either their contributions to international public goods and/or the negative externalities they would generate on cooperating states; and second, because of those actors pursuing conservative policy objectives relative to the institutional status quo. As a consequence, actors with reformist policy objectives are constrained in their abilities to change the existing institutions and to resort to club cooperation, as well as their use of cross-institutional strategies. Instead, they may rely on institutional layering for gradually building upon pre-existing rules, thus minimising opposition from critical actors with conservative policy objectives. Over time, the accumulation of incremental changes can lead to ‘path-altering dynamics’ (Mahoney and Thelen 2010: 23) as the combined weight of new elements supplants the older ones.

The case of the GR complex highlights how reformist actors were unable to link ABS governance to patent law, yet succeeded in layering benefit-sharing-related rules on top of pre-existing institutions. The International Treaty, building on the CBD and the 1983 International Undertaking, does not preclude IP claims on Annex I materials received from its Multilateral System, yet establishes a private law regime for benefit-sharing. Equally, the WHO Framework, building on the earlier Global Influenza Surveillance Network while referencing the CBD’s principle of state sovereignty over GR, does not prohibit IP claims over pathogenic materials, save for the hortatory ‘should’ in its SMTA 1. Simultaneously, recipients are contractually obliged to select among various options for benefit-sharing. Rather than strictly preferring the status quo, critical actors in both of those cases sought to secure access to GR for their domestic users while precluding linkages to IPRs, as well as limiting the scope and depth of benefit-sharing obligations. The case of the Nagoya Protocol shows how critical actors had limited prospective gains from the stricter regulation of their domestic users. Yet, the possibility to strike a package deal at the Nagoya Summit, as well as the flexibility the Protocol leaves for domestic implementation, allowed for their consent. Finally, the increasing sectoralisation of ABS governance, as well as the growing reliance on private law, demonstrate how institutional layering can lead to ‘institutional discontinuity caused by incremental, “creeping” change’ (Streeck and Thelen 2005: 9).

While the difficulty of institutional innovation has been identified as a factor contributing to the long-term stability of regime complexes (Colgan et al. 2012), this paper has attempted to specify the precise nature of this difficulty as consisting of problems related to negative externalities and the provision of international public goods. A core argument of this text was that regime complexes are not stable per se. Stability results, in a trivial sense, when all actors involved share a common interest in upholding the existing institutional arrangements, or, as discussed above, when some actors are willing but unable to change the status quo.

Although a systematic generalisation is beyond the scope of this study, we should expect regime complexes to tend more strongly towards sudden, transformational institutional change when excluded actors do not generate negative externalities for cooperating states, when contributions to international public goods are distributed more symmetrically between actors and/or when critical actors have strong preferences for changing the status quo. In particular, we should expect the availability of border measures to be a necessary condition for sudden and transformational institutional change.

For instance, the availability of border measures could explain rapid institutional change in the global governance of hazardous waste, when reformist actors, considering the managerial approach of the Basel Convention on the Control of Transboundary Movements of Hazardous Waste as insufficient, entered into club settings based on the categorical prohibition of waste imports (Clapp 1994). The case of biosafety shows how a critical reformist actor (the EU), in an issue area where border measures are available, was central in creating a new regime (the Cartagena Protocol on Biosafety) based on a regulatory approach strongly diverging from pre-existing biosafety regimes (Oberthür and Gehring 2006). In both cases, processes of institutional change are abrupt, not incremental, as new regimes are based on principles fundamentally different from the earlier ones. This sets them apart from ABS governance, where new regimes incorporate the principles of earlier ones: while the three regimes discussed in the fourth section are continuous in relation to IPRs, gradual transformation results from the specific ways in which the International Treaty and the Framework implement the CBD’s principles, however, without directly breaking with them. The Nagoya Protocol acknowledges this transformation by recognising both as specialised regimes.

Finally, institutional outcomes should also be contingent on the number of critical actors. In the cases discussed above, the existence of only a few critical actors generally facilitated their collective action. Yet, the unsuccessful disclosure of origin/Geographical Indications package deal mentioned in the fourth section highlights the inherent difficulties in maintaining status quo-oriented coalitions when some members are more receptive to linkages than others. Accordingly, we should expect a larger scope for institutional change in issue areas with higher amounts of critical actors.