Few dispute that the global security environment has evolved significantly since the end of the Cold War. This evolution is reflected in the discourse employed by the variety of actors involved in security debates today. Buzzwords like ‘dirty bomb’, ‘asymmetric threat’, ‘brain drain’, and ‘illicit proliferation network’, are now commonplace in the vernacular and policy makers at the highest level are beginning to respond in kind. Take, for example, the changing threat associated with nuclear proliferation. In April 2010, US President Barack Obama stated, ‘The greatest threat to US and global security is no longer a nuclear exchange between nations, but nuclear terrorism by violent extremists and nuclear proliferation to an increasing number of states.’ That same month, the United States hosted 46 states for an unprecedented summit on nuclear security, reflecting the recognition not just of a common threat, but also of a global regulatory environment characterised by unprecedented fluidity, complexity, and uncertainty.Footnote 1
What implications does this evolving environment have for multilateral security arrangements? Existing scholarship on this question can be divided into two camps. One is an outgrowth of the longstanding debate over the role of institutions in mitigating impediments to cooperation such as mistrust and a fear of cheating. This camp largely focuses on questions of compliance and the mechanisms best suited to promoting it in the post-Cold War era. A second camp integrates new developments in the global governance literature with the empirical study of security. These scholars explore recent developments like the privatisation of security, the impact of the information age, and the rise of alternative mechanisms of regulation in response to a complex global environment.
While both of these camps offer valuable insights into the evolution of multilateral security arrangements, they largely overlook signs of what some in US and EU legal circles call ‘new’ or ‘experimentalist’ governance. This is true in part because this prospective ‘experimentalist turn’ is still inchoate and in part because experimentalist work has only engaged security as an area of application on a very limited basis.Footnote 2 In its pure form, experimentalist governance utilises standards and recommendations, an iterated standard-setting process, increased participation at multiple societal levels, and experimentation to generate new knowledge about the challenges stakeholders face.Footnote 3 After reviewing prevailing approaches to security governance and with an eye toward teasing out the core features and functions of experimentalist governance, the first section of this article identifies and expounds upon three core shifts that characterise this potential turn: from state-centric to network-based governance models; from fixed and uniform rules to flexible and revisable standards; and from rule enforcement to problem-solving.
The second part of the article considers a potential turn toward experimentalism by examining two separate, but complementary cases that serve as key tools in efforts to counter terrorism and nuclear proliferation: the Financial Action Task Force on Money Laundering (FATF), which seeks to bolster states' systems on counter-financing of terrorism (CFT), and United Nations Security Council Resolution 1540 and the 1540 Committee, which together guide efforts to fill governance gaps in the nuclear non-proliferation regime. Although both cases contain ‘traditional’ features such as monitoring and enforcement, we argue in practice that they rely increasingly on experimentalism.
Our analysis suggests that while traditional modes of governance continue to play important roles, they are operating within a more complex and diverse system of decision-making and policy formulation, largely in response to the transitory regulatory environment. Consequently, stakeholders have reason to rely increasingly on experimentalism, which could have important implications for the form and function of regulatory arrangements, even in the high stakes realm of security. The article concludes by considering the implications of these findings for security governance and for scholars seeking to explain its function and evolution over the years to come.Footnote 4
Beyond compliance?
A range of studies show how perceptions of security have drastically evolved over the past quarter century.Footnote 5 In keeping pace with these threats, actors face a global regulatory environment that is more complex than ever before with greater uncertainty concerning both the nature of, and the potential solutions to, an ever-mutating set of problems. Consequently, the demand for innovative global governance has never been higher.
Traditional approaches to security governance
Despite that demand, much of the security literature remains rooted in traditional debates over the role of institutions, which are heavily influenced by the logic of Jervis's ‘security dilemma’. Jervis argued that security institutions, while extremely valuable, are especially difficult to build and maintain because of the potential costs of individualistic action.Footnote 6 Consequently, to the extent that scholars focus on governance, they stress the need for rigorous verification and enforcement to secure compliance. Realists argue that the great powers fulfil those functions,Footnote 7 rendering security institutions ‘epiphenomenal’, while traditional liberal institutionalist approaches stress how institutions exert an independent influence by mitigating the effects of anarchy through the provision of informational benefits (for example, monitoring and verification) and selective incentives (for example, enforcement).Footnote 8 Thus, although the role of institutions is debated, conventional International Relations (IR) approaches are driven by a common question: to what extent can we mitigate the security dilemma? Both sides see verification and enforcement as key.
After the Cold War, the ‘management-enforcement debate’ revitalised discussion over the relative importance of verification and enforcement. The ‘Managerial School’ contends that states generally comply with agreements and seldom violate regimes intentionally, rendering enforcement mechanisms less important.Footnote 9 Instead, increasing transparency, improving dispute resolution, and intensifying capacity-building efforts represent more effective means. Despite these considerable insights, managerialists may underestimate how often states strategically exploit treaty ambiguities or capacity limitations. They also largely overlook the varied ways that enforcement can affect governance, whether through a ‘shadow of hierarchy’,Footnote 10 or through a ‘penalty default’ that incentivises meaningful participation in cooperative governance.Footnote 11 Finally, they could go further in exploring the social processes and institutional mechanisms that may be critical to promoting compliance and maintaining institutional relevance by adapting to meet an evolving set of challenges.
In contrast, enforcement scholars suggest that compliance may be the wrong metric because compliance is endogenous to the institution; states will not design and ratify treaties that will either expect much of members or be continually violated.Footnote 12 They judge institutions' impact by the ‘depth of agreement’, or ‘the extent to which an agreement captures the collective benefits that are available through perfect cooperation in a particular policy area’, and the ‘depth of cooperation’, or ‘the extent to which it requires states to depart from what they would have done’ in the absence of an agreement.Footnote 13 If deep cooperation is to occur, enforcement measures will be necessary to assuage fears of asymmetric losses.
While the ‘depth of cooperation’ argument is compelling, it underestimates the difficulty of verifying noncompliance and overestimates the likelihood that enforcement alone can address the problem. Even under the most intrusive monitoring arrangements, concrete evidence of noncompliance is extremely difficult to detect and verify.Footnote 14 The ambiguity of international treaties, cultural differences between states, and the burden of gathering sufficient information to interpret behaviour authoritatively all make that determination more difficult.Footnote 15 Nor does detection of a violation guarantee an effective response. Finally, when sanctions are applied, they do not necessarily have the coercive power required to bring about compliance.Footnote 16
More generally, in today's security environment, a governance model focused on securing state compliance with uniform rules may no longer be sufficient.Footnote 17 While legal arrangements often seek some uniformity across borders, they also must confront the variation in the social systems, legal orders, and actors involved in today's regulatory arena.Footnote 18 The result is much greater uncertainty, not just about the relative capabilities and intentions of potential competitors, but also about optimal solutions to security challenges and even about what constitutes a threat in the first place. Rules and norms need to be updated continuously in order to adapt, as compliance with irrelevant rules means little. Such changes raise the question of whether and how security governance has changed in response.
‘Global governance’ and security cooperation
A second camp of scholarship draws from the broader literature on global governance to investigate the implications of a changing security environment. At the last turn of the century, the phenomenon of legalisation and institutional design took centre stage in International Relations, advancing important, primarily rationalist frameworks to think more systematically about governance issues. By showing how institutional design can vary in important ways (based on, for example, the number of actors involved, informational availability, distributive concerns, and enforcement and monitoring problems), this literature helped push the field past debates about whether international institutions matter and onto questions of the conditions under which they matter and how. Authors discuss how ‘softer’ forms of governance are useful in lowering contracting and sovereignty costs, increasing flexibility and adaptability to react to change and diversity, enhancing transparency, and potentially serving as a precursor to ‘hard’ law.Footnote 19
While those approaches are not without their critics, they highlight effectively the important variation in the institutional arrangements that characterise international cooperation. Scholars since have begun to offer a set of approaches that focus more intently on grappling with the proliferation and interconnectedness of global governance arrangements, emphasising the important governance roles played by public-private partnerships,Footnote 20 club standards,Footnote 21 transnational networks,Footnote 22 and regime complexes.Footnote 23 Schaffer and Pollack, for example, emphasise that the integration of soft and hard law makes governance more responsive to a changing environment.Footnote 24 This work all points to a new level of complexity in governance, as overlapping institutions with differing memberships, mandates, and operational mechanisms generate their own challenges for effective cooperation.
A smaller literature has begun to apply these insights more explicitly to the security realm, emphasising the explanatory lens of ‘governance’ or ‘the existence of multiple actors who interact and contribute in the provision of security’.Footnote 25 This literature in general highlights three significant changes in security cooperation today.Footnote 26 The first is a remarkable proliferation of actors, including especially the increased role that private actors play in both challenging and providing security.Footnote 27 A second theme focuses on the informality of interactions, a development related to the proliferation of non-state actors in a legal system written for a world of states.Footnote 28 Finally, these scholars also emphasise that the concept of security today now entails new kinds of issues that require coordination from the local to the international level.Footnote 29
Caparini succinctly ties the three together: ‘What we are seeing, in other words, is the pluralisation of security, both within states and transnationally … That transformation is linked to the gradual shift from government to governance, or the diffusion or fragmentation of political authority among various public and private actors at the local, state and international levels. Governance encompasses shifting networks or constellations of actors who may interact formally or informally and at multiple levels.’Footnote 30 Moving beyond describing changes in form, Daase and Friesendorf argue that the operational effect is potentially positive: ‘By pooling the strength and expertise of states, international organizations, and private actors, security governance is seen to provide more effective and efficient means to cope with today's security risks.’Footnote 31 A more sceptical interpretation might predict the ‘fragmentation’ of governance.Footnote 32
This ‘governance camp’ raises important questions: if more actors are interacting in different ways to address increasingly complex problems that require more coordination across more levels of decision-making, how are they doing it? What form is this more variegated cooperation taking and how might it help the actors better understand how the environment and calibrate their policies in response? What are the implications for more traditional security governance arrangements? As we lay out in the following section, experimentalist governance constitutes one response to this changed environment and one that theorists are only beginning to explore more systematically in security.
The road to experimentalist governance
Traditionalist and governance approaches to security cooperation comprise a valuable body of literature. Each emphasises different means by which security institutions respond to uncertainty and change, whether through management or enforcement strategies, ‘hard’ or ‘soft’ legal mechanisms, or others. Those debates have advanced our understanding of the relative operational importance of certain governance mechanisms in those arrangements.Footnote 33 Nevertheless, important gaps in our knowledge remain. The dichotomous language of hard versus soft law in particular may mask a broader spectrum in which different forms of law interact with, and act upon, one another.Footnote 34 Partly as a result, less work directly considers whether there might be a more fundamental shift in how the design of governance arrangements is evolving to meet contemporary challenges and whether those arrangements follow an alternative logic.
An examination of security institutions through the theoretical lens of experimentalist governance offers some analytical purchase in this regard. Experimentalist approaches are most often contrasted with traditional, command-and-control forms of governance, which experimentalist theory criticises as exclusive, poorly equipped to address societal complexity, ill-prepared to adapt to changing circumstances, and poorly suited to produce the new knowledge needed to solve evolving problems.Footnote 35 Not every facet of experimentalism is new, however. Reflecting rationalist assumptions, experimentalist governance represents an attempt to design institutions that respond to the particular challenges of contemporary cooperation.Footnote 36 More in line with constructivist work, experimentalist arrangements do not reflect an equilibrium outcome to a particular ‘game’, but rather are designed to evolve in conjunction with the growing diversity, complexity, and uncertainty of today's global regulatory environment in part by promoting preference change among the actors. When taken as a whole, however, experimentalist theory expects a new institutional logic to develop.Footnote 37 De Búrca and Scott describe it in the following terms:
[T]he idea of new or experimental governance approaches places considerable emphasis upon the accommodation and promotion of diversity, on the importance of provisionality and revisability – both in terms of problem definition and anticipated solutions, and on the goal of policy learning. [Experimentalist] processes generally encourage or involve the participation of affected actors (stakeholders) rather than merely representative actors, and emphasize transparency (openness as a means to information-sharing and learning), as well as ongoing evaluation and review. Rather than operating through a hierarchical structure of governmental authority, the ‘centre’ (of a network, a regime, or other governance arrangement) may be charged with facilitating the emergence of the governance infrastructure, and with ensuring co-ordination or exchange as between constituent parts.Footnote 38
In a more distilled form, the experimentalist governance model rests on the idea of collective problem-solving in complex, multi-level arenas. It involves three fundamental, although not absolute, shifts in emphasis, which together suggest an alternative logic to the design and function of institutions.
(1) From state-centric to network-based governance models. Traditional approaches to international cooperation, especially in security, are often state-centric; cooperative arrangements are designed to regulate the activity of states and rely primarily on states to develop and implement the rules.Footnote 39 Newer work that aims to understand transnational regulatory networks also often views states as the principal actors or orchestrators of this process.Footnote 40
Experimentalist governance arrangements acknowledge the enduring role of states, but also are characterised by their intent to accommodate all relevant stakeholders into the process, including non-governmental organisations, multiple levels of government, or private entities.Footnote 41 Experimentalist institutions should therefore be less hierarchical than their traditional counterparts and should involve a much wider range of stakeholders, though states likely will remain important and even critical actors. This is more than the pluralisation of actors that many governance approaches justifiably emphasise. In its ideal form it is a systematic and substantive inclusion of all relevant stakeholders, ranging from the most local to the international levels of action. Experimentalist theory suggests that such inclusiveness could improve implementation of standards or regulations, as implementation often is a local action. Furthermore, broadening stakeholder access taps into the experiences and expertise of a more diverse set of actors to capitalise on available knowledge, thereby facilitating learning and broadening the procedural basis for institutional legitimacy. This shift also may increase the likelihood of a more thorough diffusion of norms across the many actors involved in implementing those norms, a diffusion that in turn makes actors more accepting of diversity within those common standards.
(2) From fixed and uniform rules to flexible and revisable standards. Traditional governance arrangements apply rules that are rigid across space and fixed over time. Such rules may aid monitoring because compliance assessment is based on a dichotomous formula (compliant/non-compliant) and a universal standard. Experimentalist governance, by contrast, entails a more concrete emphasis on flexibility across space and revisability over time, where the former allows for adaptation to local circumstances and the latter enhances responsiveness to change.Footnote 42 This is more than just ‘soft law’. In fact, much experimentalist scholarship consciously eschews the now standard division between ‘hard law’ and ‘soft law’. The emphasis instead is on the conscious effort to institutionalise revision and adaptation, a process driven by an iterative cycle in which participants establish standards, implement those standards, report on compliance with and effectiveness of the standards, and then integrate any lessons learned into new ones.Footnote 43
In this regard, experimentalism should be identifiable by an active acceptance of diverse implementations of standards and a regular or even routinised updating of strategies to meet standards, as well as of the standards themselves. This conscious construction of a ‘policy laboratory’ is a distinguishing characteristic. It also alters the logic of monitoring and verification, moving them from a punitive purpose and toward a diagnostic one. As a result, the broader body of participants may be more likely to engage the forum in a mindset of deliberation, rendering them more open to persuasion. That ‘thicker’ information from a more diverse set of views can allow actors to better understand each other's ways of thinking and promote common understandings and purpose through continued interaction, contestation, and justification of action.Footnote 44 Experimentalist theory also suggests that governance arrangements driven by policy learning are more likely to demonstrate a ‘ratcheting up’ dynamic over time, as actors capitalise on new knowledge to revise and strengthen standards.Footnote 45 This dynamic relates closely to a final shift.
(3) From rule enforcement to problem-solving. Traditional approaches to governance stress the importance of rigorous verification and material enforcement of compliance, especially in the context of security where mistrust is particularly pronounced. Yet even if a violation is detected, it can be very difficult to generate the political consensus to enforce the rules.
While enforcement and other coercive mechanisms may continue to play an important role, the experimentalist model assumes that the complexity of today's policymaking environment often increases uncertainty about the very nature of many problems and their potential solutions, which complicates conventional notions of verification and enforcement. Consequently, institutions must be concerned not only with ensuring compliance with existing standards, but also with identifying new challenges as they arise and crafting new, problem-specific solutions.Footnote 46 Greater emphasis may therefore be placed on creating an institutional environment that generates and disseminates new knowledge about the causes and effects of problems.Footnote 47
Thus, experimentalist governance might also be identifiable by efforts internal to the process designed to identify coming challenges and by an emphasis on generating new information about the causes, effects, and solutions associated with them. Such mechanisms include the use of benchmarks and iterative standard-setting.Footnote 48 In cases where the mechanisms laid out here fail to work, a number of alternatives are available. Blacklisting, or naming and shaming, takes advantage of the intersubjective knowledge promoted by experimentalist modes of governance to enhance peer pressure. Greater involvement of epistemic communities and deliberative mechanisms are, in principle, more likely to identify behaviour that violates the scope and purpose of standards and therefore ramp up social consensus that could help empower more coercive mechanisms. Some analysts argue that experimentalist governance may operate best in the ‘shadow of hierarchy’, in which actors are compelled to cooperate because a failure to do so will mean the imposition of legally binding rules from higher up.Footnote 49 Others vigorously reject the implied ‘correctness’ that shadow of hierarchy arguments entail, given the uncertainty surrounding the issues experimentalism addresses. Instead, there may exist a ‘penalty default’ in which policy makers create ‘rules sufficiently unpalatable to all parties that each is motivated’ to take advantage of the more flexible approach of experimentalism.Footnote 50
In sum, experimentalist governance aims to keep pace with the transitory environment through an increased reliance on ideational and deliberative mechanisms designed to accommodate diversity, tap into new knowledge that may not emerge at the interstate level, and ratchet up standards and implementation strategies in response. Monitoring, verification, and enforcement are likely to remain important features of governance, especially in security issues, but they alone may not be sufficient to address the regulatory challenges inherent in governing increasingly fluid and dense global interactions.
In recent years, scholars have identified experimentalist processes at work in a broad variety of fields. Experimentalism is especially common in the EU's complex, multi-level, and evolving governance context and it is in that setting that much of experimentalist theory has been developed, tested, and debated. Scholars have also begun identifying experimentalism beyond the EU, including governance of, for example, trade,Footnote 51 fisheries,Footnote 52 human rights,Footnote 53 and the environment.Footnote 54 The following section extends the area of application to consider the security realm.
Experimentalist governance in security cooperation: an application
To what extent might we be seeing a turn toward greater experimentalism in security governance? Here we examine two security arrangements that rely in varying degrees on experimentalism to address heightened concerns about WMD proliferation and terrorism after 9/11. The first case, the Financial Action Task Force on Money Laundering (FATF), had used experimentalism prior to 9/11 to establish an anti-money laundering regime built around recommendations. After 9/11, FATF adapted its mission to address terrorism financing and, despite calls for increased use of coercive mechanisms, ultimately reinforced the role of experimentalism in its operation. The second case, UN Security Council Resolution 1540 and the 1540 Committee, were created from scratch to fill a void in the non-proliferation regime made clear in part by the discovery of A.Q. Khan's illicit procurement network. The 1540 Committee is binding on all UN members, but also contains features that reflect a degree of experimentalism.
These cases offer insights into what policymakers have identified as top security concerns. And while a considerable amount of research examines the nature of the threats, comparatively little work analyses the governance response thereto. Most importantly, however, if one adopts a purely traditionalist perspective, preventing proliferation and terrorism through experimentalism runs counter to convention. Traditional approaches contend that when uncertainty is high and the stakes are great, states are most likely to insist on rigorous verification mechanisms with fixed rules and material enforcement. Viewed through this lens, we might expect to see either a least common denominator agreement, where any higher compliance is driven by coercion from a hegemonic power, or a unilateral or club standard that is forced upon others. Most approaches would not predict sustained efforts to promote ‘deeper’ cooperation that emphasise experimentalism. As such, if we find experimentalism at work in this unlikely setting, it lends credence to the argument that we might expect to see it at work elsewhere and in turn raises questions for future research.Footnote 55
The Financial Action Task Force on Money Laundering and counter-financing of terrorism
While US President George W. Bush promised to ‘starve terrorists of their funding’ in the wake of 9/11, the little intelligence available on terrorism financing suggested it would be difficult.Footnote 56 The sums involved were relatively small and traditional methods of money transfer, such as the person-to-person model of hwala, can keep funds out of the formal banking system entirely. The United States and its allies decided quickly to work through FATF.Footnote 57 Initially established in 1989 to help states combat drug trafficking through its ‘40 Recommendations’ for a better anti-money laundering (AML) system, counter-financing of terrorism (CFT) was added to FATF's mandate after 9/11. In October 2001, FATF members produced eight ‘special recommendations’ that focused on disrupting terrorist financial networks and added a ninth in 2004.Footnote 58
Existing understandings of FATF all highlight some corner of the network's activities. FATF standards began as a club standardFootnote 59 in which peer pressure was perhaps the most important tool.Footnote 60 Its technical nature means that experts exert substantial influence on its decisions.Footnote 61 In general, most work on the AML/CFT regime has suggested that coercion by the US and EU explains the diffusion and at times FATF also has moved toward a more coercive strategy, especially in CFT efforts.Footnote 62 An experimentalist lens, however, provides a more comprehensive understanding of the strengths and weaknesses of FATF's role in post-9/11 security and highlights how members, including the United States, ultimately turned back to FATF's experimentalist characteristics to tackle this new problem. Consider again the three shifts of experimentalism described above.
(a) From state-centric to network-based governance models
FATF is comprised of a multi-level, transnational network. National delegations to FATF include multiple domestic regulatory agencies and, despite scepticism, the private sector increasingly participates in FATF's now standard calls for public comment on new proposals. FATF has granted regional AML bodies, the ‘FATF-style Regional Bodies’, an increasing role in decision-making over time. FATF also works to ensure that major international organisations endorse and apply its standards, as we show below.Footnote 63 This wide, relatively flat organisational structure is reinforced by a tradition of consensus-based decision-making. The result is a network that stretches from individual banks, through the domestic regulatory system, through FATF and regional iterations of FATF, and up to IOs, an efficient network for diffusion.
We do not argue that all voices are heard equally. The capabilities of the financial powers, especially the US, give them additional influence in agenda-setting and beyond. In interviews with FATF participants, however, officials from ‘powerful’ and ‘weak’ states alike stressed that individuals within the network can strongly influence the outcome if they are seen as especially knowledgeable, regardless of their home jurisdiction.Footnote 64 Interviewees said the same of eager states; they can play a role that is larger than their share of the financial markets would imply they could.Footnote 65 Experimentalist theory and work on global governance more broadly alike has emphasised how this multi-level, participatory approach should narrow both the implementation and the legitimacy gap often associated with global governance, as those responsible for interpreting and implementing the standards influence their development.Footnote 66
(b) From fixed and uniform rules to flexible and revisable standards
FATF's ‘40+9 Recommendations’ are widely accepted as the global standard for AML and CFT. Consistent with an experimentalist approach, the recommendations are outcome-based, meaning states have flexibility in enacting legislation that achieves those outcomes. This design aims to enhance effectiveness by ensuring that laws fit within a state's specific institutional context. It also turns the states’ different systems into an aggregated policy laboratory that promotes learning regarding the best solutions to problems because those new ideas, in turn, can be integrated into the Recommendations as the standards are revised.Footnote 67 FATF reviews the standards annually for necessary changes and has undertaken comprehensive reviews three times. Where practice shows misunderstandings, FATF issues interpretative notes to clarify the intention of recommendations and regularly issues ‘best practices’ to diffuse effective solutions.Footnote 68 This systematised updating grants new members input on the evolution of standards, reinforcing the relatively flat organisation and identifying it as an experimentalist process.
(c) From rule enforcement to problem-solving
The question of enforcement is a critical one in the debate over FATF and bears discussing in greater detail. As others have noted, the enforcement provisions of FATF are strikingly weak.Footnote 69 Most of the work is left to financial institutions, which are not obliged to follow FATF's standards unless states pass laws to that effect. In the wording used early on, states were expected to ‘be able to apply appropriate countermeasures’ against non-compliant jurisdictions. FATF has never called for a full application of those countermeasures against any of its members and only twice prior to 2012 have FATF members threatened some measure of enforcement against one of their own. In the first case, Turkey had failed to even criminalise money laundering several years after joining. In the second case, Austrian regulations continued to allow for anonymous passbooks, which make money laundering simple. In both cases, the targets had been involved in the process leading to the threats and the threat to invoke enforcement reflected wide normative consensus by FATF members. Once made, the threats were credible enough to convince the states to quickly change domestic laws they had long defended.
In dealing with non-compliant non-members, however, FATF chose a much more coercive stance. For those targets, FATF initiated a blacklisting known as the Non-Cooperative Countries and Territories (NCCT) process.Footnote 70 Members were encouraged to apply all available countermeasures to the listed states and, as others have noted, most of the targets quickly changed their laws. FATF itself argued that the process had been successful.Footnote 71 Shortly after, however, FATF eliminated the NCCT process and replaced it with a new tool, the International Cooperation Review Group (ICRG). Evidence on the new ICRG is still lacking. Not surprisingly, public statements from its chair imply a new modus operandi. More tellingly, in interviews, one official from a national delegation confirmed that the process has been qualitatively different, particularly in that targeted states are more much involved from the beginning. Also, both members and non-members are subject to the same process. An experimentalist framework and FATF's own history suggests that if FATF is to work best, the ICRG should seek to rely even more on persuasion, learning, diffusion, and peer pressure, rather than coercive blacklisting.
The decision to eliminate the NCCT, a seemingly effective tool of enforcement, is difficult to explain from a traditional security studies approach. Sharman's more nuanced approach, which correctly stresses both a constitutive and constraining effect of blacklists, provides a more complete picture of the dynamics potentially at play in blacklists, but also has a hard time explaining the decision to abandon the process, since it effectively diffused standards.Footnote 72 Within an experimentalist governance framework, on the other hand, that move seems more logical, as the goal is more than securing minimal compliance. FATF did away with the NCCT, first, because many members were never fully comfortable with its coercive nature. It had been rejected as a strategy repeatedly in the past and opposition re-emerged once it was applied. Non-members also objected.Footnote 73 The World Bank and IMF, for example, made their endorsement of FATF contingent upon the cessation of the programme, a deal which FATF accepted.Footnote 74 Second, an experimentalist framework highlights that the NCCT process was only minimally effective. Most jurisdictions complied with the lowest expectations.Footnote 75 There is little evidence that the targeted states are fully committed to the AML-CFT mission, and thus most are not contributing new knowledge to the network. While blacklisting was largely effective from a compliance perspective because it provided targets with the incentive to comply with the minimum standards, from a problem-solving perspective its performance was weak and its abandonment is not surprising.
Regarding CFT in particular, after 9/11, the US pushed hard to use FATF for that end, a goal that members accepted. According to interviews, however, the US approached FATF members much more aggressively than it had in the past to promote its preferences. As one US delegation representative described it: ‘We went in hard. We called people out. We bloodied some noses.’Footnote 76 Ultimately, however, other actors pushed back in a way that re-established the experimentalist nature of FATF. The same participant continued: ‘But then we learned, that isn't how FATF works. You have to convince people you're right. After that we changed our approach.’Footnote 77 In the case of FATF, then, efforts to move away from experimentalism and toward coercion were explicitly rejected by members on various occasions in favour of returning to the iterated standard-setting and knowledge creation that had long been the focus of FATF activities.
In addition to members’ decisions to reject a more coercive FATF, an experimentalist framework also highlights the multi-faceted monitoring process within FATF as a source of peer pressure and new knowledge. Self-evaluations, which catalogues states’ laws ‘on the books’, are one aspect of that monitoring. The effectiveness of states’ systems is judged using a second process, the mutual evaluations, which includes on-site visits and interviews by a team of outside experts. A third monitoring mechanism is the mutual evaluations that other organisations, such as the regional AML-CFT organisations and the IFIs, carry out, all using a common evaluation methodology. The states are then subjected to a ‘peer review’ in the plenary in which states are pressed to explain gaps in their regulatory systems.
Finally, the ‘typologies exercises’ comprise a fourth, and particularly notable, means of monitoring. Working in part to monitor the AML/CFT system as a whole, the exercises bring together money-laundering experts to exchange information about new solutions to old problems and new trends in money laundering and terrorism financing. They generate and diffuse knowledge about what works and does not work, and the new knowledge gained from these exercises is used in revising the recommendations. They also constitute a forum in which the epistemic community of anti-money laundering officials can apply peer pressure.
Together these mechanisms comprise a monitoring system that extends beyond what most traditional security arrangements achieve even on paper. In interviews, FATF officials attribute this access to two factors.Footnote 78 FATF was (and technically remains) a temporary task force, although its mandate has been lengthened with each renewal.Footnote 79 More importantly, interviewees from ‘powerful’ and ‘weak’ states alike stressed that they understand the monitoring as an opportunity to explain their systems more fully and to receive useful feedback.Footnote 80 When detached from heavy enforcement mechanisms, verification becomes less punitive and more diagnostic, which would appear to corroborate experimentalist theory. It is less about revealing cheating and more about facilitating learning. Monitoring and verification are altered and strengthened by FATF's experimentalism. As an illustrative example of the creative potential of that system, and especially the exercises, it was at one such meeting where the notion of using FATF as a tool against terrorism financing was first raised. Surprisingly, that occurred before 9/11, in December 2000, and before terrorism was a prominent meme in security policy.Footnote 81
FATF's history contains important warnings about experimentalist institutions. Particularly if there are significant differences in capabilities among participants, powerful states may be tempted to forego more labour intensive and slow moving ideational or social mechanisms in favour of coercion.Footnote 82 But that option exists in most institutions where power asymmetries are significant. That said, FATF's history also suggests that experimentalist governance can be a powerful alternative to more coercive strategies for change.
UNSCR 1540 and the 1540 Committee
While CFT efforts through FATF represented an expansion of pre-existing experimentalist processes, 9/11 and, more immediately, A.Q. Khan's proliferation network, also revealed that longstanding, ‘traditional’ institutions designed to prevent WMD proliferation, such as the Nuclear Nonproliferation Treaty (NPT), had limited capacity to deal with non-state actors and ‘private’ procurement networks. In previous cases of member-state non-compliance, counter-proliferation efforts had focused on the creation of a more stringent monitoring apparatus (for example, the IAEA Additional Protocol) and strengthened export controls from supplier countries through arrangements such as the Nuclear Suppliers Group. Although important, those adaptations were less capable of addressing the respective challenges posed by terrorists and Khan's proliferation network. On one hand, meeting these challenges requires broader and deeper cooperation at the national and subnational levels and among the broad range of domestic agencies that oversee strategic trade control policy. On the other, the list of actors that should be involved in regulation varies across states and even the items to be regulated, which makes standardisation counterproductive.
In response, in April 2004, the UN Security Council acted under its Chapter VII authority to pass Resolution 1540 by unanimous vote. UNSCR 1540 aims to ‘enhance coordination of efforts on national, sub-national, regional and international levels in order to strengthen a global response’ to the threats posed by the proliferation of weapons of mass destruction to non-state actors.Footnote 83 Mirroring the vague language used in FATF, the resolution requires all states to: (1) ‘Adopt and enforce appropriate effective laws which prohibit any non-State actor to manufacture, acquire, possess, develop, transport, transfer or use nuclear, chemical or biological weapons and their means of delivery, in particular for terrorist purposes’; and (2) ‘Take and enforce effective measures to establish domestic controls to prevent the proliferation of nuclear, chemical, or biological weapons and their means of delivery, including by establishing appropriate controls over related materials.’Footnote 84
1540 was controversial for many reasons. Its passage under Chapter VII made it binding on UN members. It reached far into states’ domestic regulatory systems by obliging states to pass domestic legislation in accordance with 1540. In the highly politicised post-9/11 context, some saw 1540 as imposing the cost of security for wealthy, Western states on smaller, poorer states with different security priorities.Footnote 85 However, while 1540 appears on paper to deal with this significant cooperation problem through enforceable, ‘hard’ law, when viewed through an experimentalist lens, 1540 looks very different.Footnote 86 Consider again the three shifts advanced above.
(a) From state centric to network-based governance models
The work behind 1540 is coordinated through the aptly named ‘1540 Committee’. The Committee is comprised of states, but relies heavily on outside experts and national regulatory agencies to conduct its work. These experts are central to the Committee's outreach goals; in 2007–8, for example, 1540 Committee experts took part in just under 50 conferences around the world.Footnote 87 Critically, the experts also are responsible for reviewing and interpreting the reports gathered by the Committee.
The Council underscored the importance of a multi-level network approach in 2006 and again in 2008, when it urged the Committee to integrate states, as well as international, regional, and subregional organisations, in order to promote ‘experience-sharing and lessons learned’ regarding questions covered by the Resolution.Footnote 88 The Committee takes an active approach to networking. In its 2008 report, for example, the Committee relays ‘model laws and other measures’ that other organisations have developed to deal with common regulatory issues.Footnote 89 Reflecting goals of learning, mimetic diffusion, peer pressure, and deliberation, the Committee notes that it aims to serve as a ‘clearing house for information on the issue of assistance through formal and informal contact and dialogue with all States …’.Footnote 90 The ten-year mandate passed in April 2011 suggests that regional and subregional organisations should play a prominent role in coordinating implementation among their respective members.Footnote 91 Some have also argued that NGOs and regional organisations have an important role to play in applying pressure to states that are slow to implement the guidelines.Footnote 92
Again the 2011 Committee report confirms the importance of this shift, albeit with little detail. In the report's rather mathematical style, it notes that 83 states report having made ‘some effort to reach out to industry’, a slight increase since the last report in 2008.Footnote 93 More tellingly, the report emphasises a reliance on several other governmental and non-governmental organisations to provide model laws and regulations that would enhance 1540 compliance, including the IAEA, UN Office on Drugs and Crime, and the International Committee of the Red Cross.
(b) From fixed and uniform rules to revisable and flexible standards
Like FATF's recommendations, UNSCR 1540 establishes standards that states must meet but allows substantial flexibility in implementation. Also like FATF, the key provision requires all states to adopt and enforce ‘appropriate effective’ measures to prevent terrorists from acquiring WMD capabilities. It does not specify what constitutes ‘appropriate’ or ‘effective’ and states must determine what laws would meet that standard ‘in accordance with their national procedures.’
The 1540 Committee itself is comprised of representatives from each member state of the Security Council. The Committee's primary responsibility is to collect and assess comprehensive reports from state parties that provide details on the steps taken or to be taken toward implementation. It reviews reports, identifying gaps in regulatory systems and recommending changes. This procedure contains considerable potential as a means of diffusing new knowledge and promoting learning. Like FATF's peer review, it also allows would-be enforcers to develop a more nuanced understanding of a state's compliance status and react accordingly. In its 2011 activity report, the 1540 Committee suggested revising its compliance matrices ‘to better reflect progress towards implementation of the resolution, i.e., in term of assistance and cooperation, or lessons learned’, which indicates an emphasis on promoting learning and the potential for more fundamental revisability.Footnote 94
(c) From rule enforcement to problem-solving
While the resolution is binding, it appears to rely very little on traditional enforcement in practice. In contrast to FATF or the NPT, for example, 1540 contains no mention of any international enforcement and instead emphasises activities that promote the use of social mechanisms over material ones.Footnote 95 While monitoring is increasing (the ‘implementation matrix’ that experts have developed to ascertain compliance has blossomed into an extensive list of 382 elements),Footnote 96 to date monitoring has not been used as a basis for enforcement. Instead, the Committee has used the reports and experts’ analyses to clarify the intention of the original Resolution where misunderstandings exist, to identify weaknesses in states’ regulatory regimes, and to inform decisions about the provision of technical assistance. Reflecting that emphasis on assistance over enforcement, the 2011 report lays out a revised method of collecting requests for assistance and seeking the expertise needed to meet those requests.Footnote 97 Recent comprehensive reviews find that this softer approach is likely the best one, as the lack of capacity is the best predictor of low compliance with 1540's provisions.Footnote 98
UNSCR 1540 in practice exhibits several characteristics of experimentalism. It relies on self-reporting mechanisms, peer review, the involvement of multiple levels of governance and non-state actors, and social pressure to ensure that states submit reports and implement recommended changes. These mechanisms are designed to accommodate diversity and ramp up social pressure to strengthen export controls, but also to gather and share knowledge regarding the best ways to do so. This arrangement moves away from ‘cookie-cutter’ approaches and toward efforts to implement ‘appropriate effective’ measures that are tailored to a given state and infrastructure. This process might also help legitimise and make more effective any future attempts to enforce 1540.
Former 1540 Committee Chair and Slovakian ambassador Peter Burian signals a number of these dynamics in his discussion of 1540's impact. Burian notes that both proponents and opponents of 1540 have profited from the iterative nature of experimentalist governance.Footnote 99 The Committee, he notes, ‘now better understands the needs’ of developing states and has adjusted its approach accordingly, working to ensure that capacity building efforts are more in line with the target's development priorities. Simultaneously, the Committee works to alter targets’ understandings of their own security: ‘[W]e talk to the representatives of those countries and explain to them what is at stake. Their territories might be misused for purposes of trafficking or planning attacks against some other countries or hitting targets in those countries. Tasks connected with the implementation of 1540 might help them to address some other issues more efficiently through improved border controls and export controls, such as the issue of small arms and light weapons trafficking and drug trafficking.’Footnote 100 While experimentalism is not the only form of governance that promotes ideational convergence, its iterative nature and de-emphasis on coercion and enforcement may encourage these mechanisms more than other forms.
The 2011 report further confirms the experimentalist tendency of 1540 in practice. As suggested by experimentalist theory, the report in Annex XVI contains 11 pages of ‘relevant examples to which States may wish to refer in implementing resolution 1540’ as a way ‘to facilitate the sharing of experience…’. Those pages provide best practices for policies designed to meet 1540 standards. More generally, the recommendations for future efforts all stress knowledge creation and diffusion, inclusion of non-state actors, adaptation to local circumstances, and provision of assistance that new governance emphasises. Perhaps just as telling, given the sensitivity of the issue and continued recalcitrance by some states, the report clearly avoids suggesting a more confrontational approach to ensuring compliance that the resolution's legal basis would allow. The clear emphasis is on knowledge, deliberation, learning, and problem-solving, not enforcement.
Like FATF, 1540 is far from perfect. Some observers argue that it lacks legitimacy,Footnote 101 that its specific role in counter-terrorism is unclear,Footnote 102 that implementation will be uneven,Footnote 103 and that its expectations are vague.Footnote 104 In a comprehensive review, Finlay and Turpen argue that developed country support has been lacking when requested by developing countries, which only strengthens the perception that 1540 prioritises Northern security over Southern priorities.Footnote 105 Nonetheless, Finlay and Turpen see 1540 as a potential platform on which to build more comprehensive security efforts.Footnote 106
In sum, the 1540 case shows that experimentalist governance plays a significant role in efforts to supplement and adapt the current non-proliferation regime to address evolving threats. The overall move away from coercion and toward knowledge creation and learning; the inclusion of a broader network of actors; and the semi-regular updating of the standards as promoted by the regular issuing of new reports that involve new procedures and new models for action all reflect an experimentalist logic. Time will tell whether the shift continues or reverses course, but as the conclusion suggests, there is reason to think that the momentum toward experimentalism may continue to grow in the future.
Conclusion: an experimentalist turn in security governance?
In a complex and transitory security environment, scholars and practitioners alike are seeking new ways to keep pace with emerging and protean threats. Cooperative governance arrangements constitute an important tool in this effort. Classic treatments of these arrangements remain important. Yet, the enduring focus on rigorous verification and enforcement of fixed rules as reflected by ‘trust but verify’ dictum, a legacy of the Cold War, may obscure more recent developments in security cooperation. While broader research in global governance offers valuable insights as to why informal legal mechanisms and more participatory designs are being deployed with greater frequency, it offers fewer details on some of the more experimental dynamics emphasised here. This article began by asking whether we may, in fact, be beginning to see an ‘experimentalist turn’ in the design and operation of governance arrangements. By way of conclusion, let us revisit this larger question in light of the application of the experimentalist lens to assess two prominent multilateral efforts to address the threat of nuclear terror: FATF and the UNSC 1540 Committee. Four particular sets of observations stand out.
First, in both cases experimentalist mechanisms appear to be performing a great deal of the operational work, albeit at very different stages of development. Consider again some of the key observable implications of experimentalist governance. Reflecting the shift from hierarchical, top-down regulatory processes and to broader, flatter networks, both cases consciously expand participation in the process, thereby engaging more directly the range of actors required to solve complex problems. FATF, for example, now issues calls for public comment on its proposals and hosts regular conferences with private sector parties most affected. The 1540 Committee is less developed along these lines, but does seek outside experts to perform consultative roles. Both cases also reflect the shift from fixed and uniform rules as the primary means of promoting cooperation to more flexible and revisable standards. In FATF's case, efforts to turn a pre-existing experimentalist institution into a more coercive tool ultimately were rejected by participants as not being in line with ‘how FATF works’. In the case of 1540, while the resolution is legally binding and enforceable, the Committee appears to recognise the need for diversity within a common standard and has shifted its emphasis accordingly over time. It encourages capacity-building efforts, not threats of enforcement, as the primary means to help states build domestic regulatory regimes. And it acknowledges that those regimes should be designed to align with pre-existing legal institutions, not to comply with a universal and unchanging international law. The above shifts also reinforce the third shift of experimentalism, from enforcing compliance to promoting problem solving. Despite the fact that the extraordinary circumstances under which it was created means it is a legally binding and enforceable law, both in word and in practice the Committee has pushed its actions much more toward knowledge creation and learning. This is further reflected in the language of the continuing mandates, which talk increasingly about learning and knowledge creation and less about enforcement. FATF, too, largely withstood efforts to turn it into a coercive institution after 9/11. Instead, as indicated in the interviews cited above, participants ultimately reinforced and further developed its experimentalist qualities through regular production of best practices, the application of peer pressure through peer review, and the transformation of monitoring into a diagnostic tool instead of a punitive one.
Second, neither case suggests that experimentalist mechanisms and enforcement mechanisms are necessarily at odds with one another. Existing approaches that rely on dichotomous frames such as ‘hard and soft law’ may mask a broader spectrum in which different forms of law interact with, and act upon, one another. If Sabel and Zeitlin are correct in arguing that experimentalist processes most often emerge from traditional governance arrangements, understanding the interaction of the two will be critical to understanding experimentalism.Footnote 107 1540 and FATF may provide some additional insight in this regard.
By creating a more participatory forum that relies on an iterative process of norm development and a more flexible implementation of those norms, coercion becomes more of a last resort than a primary tool. Monitoring and verification serves first to facilitate knowledge creation and diffusion of best practices. Moreover, states are likely to be more accepting of extensive monitoring if it is not attached to strict material enforcement and is tied instead to assistance. Interviewees explained FATF's unusually extensive monitoring processes in precisely those terms.Footnote 108 Likewise, 1540 addresses strategic trade control issues that are highly sensitive for both security and economic reasons, making transparency a challenge. Tying monitoring to enforcement probably would make states still less likely to engage the process in an open, non-strategic way. At the same time, the potential for enforcement may help actors remain engaged in the process in order to avoid activating the potentially more costly and less sensitive coercive machinery: what Sabel and Zeitlin have called a ‘penalty default’.Footnote 109
In the event that the last resort becomes necessary, the ‘thick information’ environment of experimentalist governance can help stoke the political will necessary to overcome the collective action problem associated with enforcement, making threats more credible. The history of enforcement in FATF illustrates this logic. The rare cases of enforcement against members have been met with little resistance, even from the members themselves. Actions against non-members, however, were non-participatory, purely coercive, challenged by both members and non-members, of questionable impact, and ultimately abandoned in favour a new process. Thus, while enforcement still matters, it may be evolving in terms of how it interacts with emerging forms of governance. Sharman's work on blacklisting in FATF recognises the significance of the interaction between enforcement and social or discursive mechanisms, but recognising FATF's experimentalist designs helps explains how the balance or dynamic between material and social mechanisms can change the logic of an institution. In other words, our preliminary findings suggest that experimentalism may complement or even enhance the effectiveness of coercion, but while simultaneously making it less likely. Much more research is required on this critical point.
Third, given that experimentalist governance arises in response to a certain set of conditions, there is reason to suggest that this potential turn could have more general application. The broader experimentalist literature suggests that increased complexity and uncertainty in the regulatory environment places more of a premium on mechanisms that afford actors greater autonomy in interpretation without the constraints of uniform rules or threat of sanction. In principle, this greater flexibility enables the governed to experiment and tailor solutions to their specific problems, provides feedback mechanisms to share and build knowledge, and allows actors to establish minimum levels of adherence while routinising a ratcheting up of standards. While some studies have identified these mechanisms at work in the EU, international trade, and the environment, not much work has been done in the security realm.Footnote 110
The foregoing analysis suggests that both FATF and 1540 are responding to similar conditions in the international security environment. Are other security governance arrangements also relying more on what might be considered experimentalist mechanisms? Answering this question requires more empirical research that explores security arrangements in an evolutionary context, drawing from an experimentalist framework to assess whether the governance mechanisms doing the most operational work have changed over time. The discussion above provides identifying characteristics of experimentalism: a systematic evaluation and comparison of implementation strategies and results, a broader and more meaningful inclusion of stakeholders, a routinised updating of standards, and a drive to identify and address new threats, rather than aiming only to enforce pre-existing rules or promote increased dialogue.
While a detailed analysis is beyond the scope of this article, a number of arrangements would be interesting to explore through this lens.Footnote 111 For example, the Organization for Security Cooperation (OSCE) in Europe has long aimed to include non-state actors and engage in debates over goals of security cooperation.Footnote 112 However, further examination is necessary to analyse the degree to which the additional dynamics of experimentalist governance outlined above may be taking hold.
Another example worthy of closer scrutiny might be the Wassenaar Arrangement, which seeks to regulate the spread of conventional arms and dual-use technologies, and its predecessor, the Coordinating Committee for Multilateral Export Controls (COCOM). COCOM involved a network of experts that met regularly to review implementation and performance and even updated their standards somewhat regularly, but the system was strongly state-centric, a fact reinforced by the veto power that each member state held. Once the common threat of the Soviet Union fell away, COCOM was unable to adapt and quickly disbanded. As one analyst describes the process: ‘The accelerated page of political and technological change began to overwhelm the control capacity of the COCOM regime.’Footnote 113 That institutional plasticity is what experimentalism aims to avoid, in part by promoting normative convergence within a flexible set of standards. COCOM's replacement institution, the Wassenaar Arrangement, remains state-centric, but at a glance displays more experimentalist characteristics, including a more flexible implementation structure, increased transparency, and a drive to identify and respond to new challenges.Footnote 114 Other security examples might include the Nuclear Suppliers Group and Proliferation Security Initiative.Footnote 115 Ultimately, these shifts are best understood as spectra, not as dichotomous variables. That said, more developed experimentalist processes will display substantive movement regarding the three shifts indicated above, not minor changes and not along just one of the axes.
Finally, it is important not to conflate any experimentalist turn with effectiveness. Even if, as our analysis suggests, experimentalist mechanisms are playing a greater role in transnational regulation, it is another exercise entirely to gauge effectiveness. We must be careful not mistake potential for actual results and a comprehensive analysis of the cases presented here falls beyond the scope of this article. Such an exercise would, among other things, involve a systematic exploration of how widely the norms generated by such processes are diffused and an assessment of whether the standards diffused generate changes in actor behaviour that achieve the desired outcome.
Our research does, however, lend some initial support for the idea that experimentalist mechanisms can help improve effectiveness. The regular updating of standards in FATF, for example, or the issuance of best practices to clarify points of confusion revealed by the monitoring of implementation are examples. In the 1540 context, the 2011 report notes that 166 states now have measures in place to ‘penalize the involvement of non-state actors in the prohibited activities’, up from 63 just three years prior.Footnote 116
However, these insights are preliminary and we need a more rigorous examination of the emergence, problem-solving capacity, and normative implications of experimentalist mechanisms in security and beyond. Any such examination should also be mindful of the exercise of power. Power does not go away in experimentalist governance and these emerging regulatory constellations challenge analysts to widen their conceptualisations of power in a given setting.Footnote 117 It may very well be that the powerful will exercise disproportionate influence over experimental processes, at times to the detriment of the governed.
The defining question of international security policy today is not whether to cooperate, but how to do so. While it is too early to declare definitively that experimentalism will be the answer to that question, one thing seems clear: governance arrangements of any type will have to adapt to changing conditions or be rendered obsolete. As an analytical tool, experimentalist governance offers important insights regarding how regulatory arrangements may be evolving to meet the threats of the twenty-first century.
