Question 1: What explains FATF’s expansion and what does this mean about the future?
Looking back it is easy to assume that the current AML regime was a logical and likely outcome of FATF. Seen in historical context, however, the regime’s expansion and FATF’s evolution into a major financial actor today is unexpected. In 1998, Susan Strange  predicted that FATF was an empty effort by the G7 to be seen as doing something to respond to the expanding illicit narcotics trade in the 1980s. Thus, a primary theme—arguably the primary theme—of scholarship on FATF has been explaining the regime’s expansion.
Scholars primarily argue that FATF is a tool that powerful countries—the EU Member States and/or the United States, in particular—use to force their preferences on other jurisdictions. Simmons [26, 27] argues that it reflects the preferences, soft and ill-formed, of the United States alone. Drezner  argues FATF’s core agenda reflects consensus among the US and EU member states, while the flexibility inherent in the recommendations reflect enduring disagreements on the details. From a more strategic constructivist angle, Sharman [23, 24] argues that the US and EU use FATF to paint non-compliant jurisdictions as rogue, unreliable players, thereby scaring off would-be investors. Jakobi  uses network analysis to map the influence of the United States.
Articles in this special issue move this debate forward, albeit in different ways. Inês Sophia De Oliveira suggests that we might be in a new phase of rule-making in FATF. She argues that while powerful states historically determined the regime’s direction, today the private sector plays a much larger role, even challenging the primacy of the state. The private sector—in particular the financial services industry and its high-level representatives—is becoming a “non-great power influencer” in FATF. She writes that the shift away from a rule-based system and toward a risk-based one is a strong and substantively important example of private sector influence and away from the state domination of earlier years. The changes are an evolution, however, not a paradigm shift.
My own contribution to this debate attempts to provide a novel interpretation of FATF. Emphasizing the evolution of FATF over time, I argue that FATF has been most influential when it has operated in line with the principles of “experimentalist governance” . The ideal-typical experimentalist governance model emphasizes flexible and revisable standards over fixed and universal rules; broadly participatory networks over state centric quasi-hierarchy; and dynamic problem-solving over rule enforcement . Social learning that extends from peer review and implementation is a more important tool of change than threats of blacklisting or financial punishment. Thus, FATF has been most effective at extending the AML regime when members have used it in line with the principles of experimentalist governance. States and professions subject to FATF’s work have pushed back the most when material power became a primary tool of change. This interpretation challenges the existing common understanding of the importance of blacklisting within the regime and suggests a very different set of reforms that FATF members should consider going forward. Put most simply, bolstering enforcement within FATF is likely to be counter-productive, while reinforcing social learning and peer review are more likely to generate greater returns.
Question 2: Is FATF a financial or security network? To what effect?
Especially as FATF expanded into counter-terrorism financing in 2001, it created an important and as-yet unresolved tension. Is anti-money laundering and the work of FATF aimed at shoring up the financial system and protecting those invested in the system from abuse? Or has FATF become fundamentally a tool of security actors, aimed more at national security and, to a lesser degree, criminality? Two authors here provide two different interpretations, but both emphasize that balancing the two sides has been a challenge.
Anthony Amicelle argues that the AML must be understood as a “configuration,” or a social space characterized by a “tension-balance” between interdependent actors. As a result of AML, banking and law enforcement officials regularly must work together, but they bring with them different frames of reference. Security actors tend to see their jobs as ensuring national security and defending against societal risks over any other end, while financial actors see themselves as ensuring financial integrity and defending against institutional risks. Despite that, the interactions generate a “productive misunderstanding” that allows them to cooperate as if they agreed. Ultimately, however, this productive misunderstanding is embodied in the person of the compliance officer, and those officers work for the financial side, not the security side. So while security and financial officials in the UK and Canada, Amicelle’s two cases, cooperate in both formal and informal ways “to an extent inconceivable ten years ago,” the balance tips in favor of the financial actors. FATF and AML have led to the securitization of the financial sector, but the greater change, argues Amicelle, has been the financialization of security.
Anja Jakobi continues her work in comparing the governance of different transnational crimes. In this special issue she merges theoretical concepts from criminology—which often focuses on the domestic, policing side of crime—with insights from International Relations—which focuses more on the interaction of actors at or above the level of the nation-state. From this she builds an analytical framework that highlights five dimensions of security governance: a comprehensive security concept, multi-purpose rationalization, public-private cooperation, multi-nodal governance, and transnational security spaces. Using this framework, she points out that FATF members were successful in advancing the AML agenda in all of these aspects of security governance. This explains why the AML regime moved forward more than many observers initially expected. At the same time, it also explains variation within FATF: why some initiatives moved forward more than others. Those initiatives with a tie to security developed more than those that focused only on financial integrity. So while analysts often present FATF as a financial regulatory body, in fact FATF is better considered an instrument of security governance. AML, she writes, “is still a weak governance instrument for regulating financial crimes such as tax evasion or corruption, but it’s a strong one for security-related crimes.” This conceptualization provides other norm or organizational entrepreneurs with a roadmap toward greater prominence within the regime.
Question 3: What role does, and should, the private sector play?
Scholars have not researched the role of the private sector in FATF and the AML regime as thoroughly as one might expect, although there are important exceptions. This is perhaps because of the prominence of state-based explanations cited above, but those only raise more compelling questions. If state interests drive FATF and the AML regime, how have the interests of private financial sector played into those state preferences? Drezner  argues that financial instability in the 1990s left wealthy U.S. and European investors eager to re-regulate finance in exchange for financial stability. That may have bolstered the regime’s growth, but it ignores the fact that the regime got its start (in 1989) before the financial crises of the 1990s could shape investor preferences. Sharman’s work suggests that state preferences break along the lines of the emphasis of the domestic financial sector; states with financial sectors built around stability preferred stronger AML rules, while those built around secrecy preferred less scrutiny. Little work, however, has systematically traced the empirical footprints of those arguments to connect them to outcomes. The scholarship of some of the authors in this special issue provides a corrective to this long-standing gap.
Eleni Tsingou’s article here highlights the evolution of the compliance industry, in particular. She writes that compliance professionals not only are the “foot soldiers in the fight against money laundering,” but have evolved from being rule takers to rule makers. While previous work has looked at national cases of the compliance industry [31,30,31,34], Tsingou draws on interviews and participant observation in making the global compliance industry her unit of analysis. Its professionalization, she writes here, “has led to an extension of governance functions, from implementation, to active interpretation of rules, to shaping the content of governance through regulatory creep.” This transformation, and empowerment, of the compliance profession has happened at the highest levels because of the revolving door between regulators and banks. It has happened in the small numbers of firms that provide IT and other technical capacity to financial institutions. It has even filtered down to the junior- and mid-level officer level, where the development of internationally recognized qualifications, an international conference and workshop conference, and north-to-south capacity building efforts drive convergence. The result is a deeply transformed aspect of the AML regime, which feeds into FATF and the standard-setting process. An important implication of her argument is that we will not be able to understand FATF and the AML regime going forward without giving the compliance industry in particular primacy of place in our models.
Karin Svedberg Helgesson and Ulrike Mörth have worked to understand the role of “frontline” private sector actors outside the compliance industry. In their contribution to this volume, they draw on interviews with lawyers in Sweden to understand how FATF’s rules addressing the legal profession trickle down to the day-to-day practice of law and how lawyers respond. The legal profession has been skeptical of AML from the beginning. Lawyers fear that AML regulations impinge on attorney-client privilege and inhibit their ability to provide their clients with the best possible defense.Footnote 9 AML experts, however, fear that attorney-client privilege allows lawyers to help their clients launder money and hide behind privilege. Helgesson and Mörth emphasize that the new risk-based approach entails important assumptions about the nature of AML and the role these actors can and will play. It places increased responsibility on private actors, assumes that the risks of AML are knowable to those actors, and implies that those actors can and will act in line with the intentions of the standard-setter. In the Swedish case, law firms have developed “practices of separation” that allow them to navigate the shoals of FATF standards, on one side, and professional and business obligations and goals on the other. The result is a business-oriented pragmatism on the part of the legal profession, rather than a more productive engagement as we see even in some banking sectors. The authors argue that FATF members have not contextualized standards sufficiently to generate rules that appropriately distinguish between the banking and non-banking sectors, an important question for practitioners going forward.
The contribution by Peter Romaniuk and Tom Keatinge reveals a different path for private-sector involvement, but generates another call for sector-based contextualization. They provide a fascinating look at how the non-profit sector ended up involved in AML. Following the 2001 terrorist attacks in New York City, some FATF members, led by the United States, pushed hard to include a special emphasis on counter-terrorism financing. In service of that goal, members added eight so-called “Special Recommendations” (SR) in October 2001, before adding a ninth in 2004. It was arguably at this point in history that FATF and its newly named “40 + 9 Recommendations” became best known. SR8 addressed concerns that charities were serving as laundering agents for terrorist groups. With time, however, it became clear that states’ efforts to comply with SR8 were threatening the viability of the “third sector,” which was so vital to development and human security around the world. Romaniuk and Keatinge document how a “transnational advocacy network”  formed and worked successfully to re-frame the issue in terms of rights and proportionality. FATF members revised and re-issued SR8 in 2016. They also created a consultative forum for NPOs like that created for the private sector some years earlier. The outcomes and the path to success, they argue, provides a playbook for other sectors affected by FATF’s expanding activity. The case also holds an important lesson for FATF about the need for consultation and contextualization before issuing potentially harmful recommendations.
Question 4: How responsive is FATF and the AML regime?
Most of the articles cited thus far stress how FATF has changed over time. But how responsive is FATF to the changing world around it? It is a pressing question. The 2008 crisis confirmed that financial innovation can outpace our understanding of the implications of those innovations . This is particularly true in the adaptive environment of illicit markets. As with the other themes, many of the papers address this question indirectly or briefly. Two papers take on the question more squarely.
One way to think about responsiveness is the ability of FATF recommendations to apply and be effective in different contexts. I argue that the experimentalism of FATF should make it more adaptable and flexible, but William Vlcek in this volume suggests it may not go far enough. AML stretched more deeply into informal economies, he argues, because of displacement; AML drove money laundering out of formal economies, which made informal economies more attractive. But FATF’s reliance on typologies, or the development of AML guidance in response to specific patterns of money laundering, renders the system less adaptive and more rules-based than some analyses suggest. De-risking only exacerbates the problem, as it forces still more economic activity into the informal, less observable realm of the market. As an operable piece of advice, he writes that calls for formalizing informal economic actors are often “beyond the means and ability of the state in question.” There may be technological fixes that minimize reliance on human resource, but he also reminds would-be enforcers that those fixes “require resources and institutional capacity to maintain.” Capacity-building assistance—beyond sharing ideas and stretching into funding—might be an area that FATF members should give greater priority.
A second vector of adaptability addresses FATF’s capacity to respond to new means of money laundering. Malcolm Campbell-Verduyn considers whether FATF and the AML regime have responded appropriately to “crypto-coins,” such as Bitcoin. Many major financial institutions remain leery of them as standard currencies. For money laundering and illicit economies, however, crypto-currencies remain an important new frontier. “Altcoins” are “verified through decentralized peer-to-peer networks and then broadcast on public ledger that encode the transaction histories of each individual crypto-coin.” As such, they effectively by-pass most AML actors. They are quasi-anonymous, which frustrates the Know-Your-Customer principles at the heart of the AML regime. Monitoring their use is also difficult, which makes it difficult for investors and regulators to identify normal versus abnormal patterns. If FATF’s sole goal with crypto-currencies is to prevent money laundering, he argues that FATF’s guidance on the question falls short. It relies too heavily on private industry to regulate itself, promotes a technological arms race around anonymity between regulators and users, and emphasizes high-risk exchanges while overlooking the “middlemen” down the chain that have major financial incentives to verify suspicious transactions. That said, these governance gaps may not be fatal, because quasi-anonymity is only one of two blockchain pillars. The other is that every transaction with a “coin” is visible. Quasi-anonymity might prove a loss, but the development of currencies that carry their own logbook of transactions might prove a gain. There is a widening field of companies competing to provide crypto-currency forensics to police. “Permissioned blockchains” actually build in user identification, making them especially good for meeting Know-Your-Customer standards, while saving customers and financial institutions resources. The Isle of Man appears to be developing a comparative advantage in hosting AML-compliant, more reputable crypto-currency companies. Especially in relation to better established, “lower-tech” forms of money laundering, argues Campbell-Verduyn, the blockchain technologies of crypto-currencies likely provide a greater advantage to AML in the long-run. An important question moving forward is whether FATF and the AML regime can adapt to take advantage of this technology.
Question 5: Is FATF effective? How can we know?
Finally, the most important question that the articles here raise is the question of effectiveness. Every article in this special issue has something important to say about the question. It is also the question that FATF members have wrestled with the most in the past five years and, hopefully, the one they will be wrestling with in the years to come.
Readers of this journal are aware that the question of effectiveness is always difficult in social science research. Part of this has to do with the complexity of social phenomena and the slippery nature of causality in such settings. In International Relations, a debate relevant to this special issue focused on the impact of international institutions and why states tend to comply with their international obligations . For some, compliance was an indicator that international institutions (understood in the field to include most any agreement among states) mattered [38,37,40]. International institutions and the process of building them were capable of shaping participants’ identity in ways that brought their preferences in line with the goals of the institution. Institutions could facilitate persuasion and socialization amongst their members . Critics argued that compliance stemmed from the fact that most international institutions were merely codifications of pre-aligned preferences. Institutions did not cause convergence, they only reflected it . These “enforcement school” scholars argued that any agreement requiring states to veer very far from their top preferences also would require strong enforcement by leading states. Do institutions “cause” preferences to align, or do aligned preferences “cause” institutions? Counterfactual reasoning is also important: would the problem have been worse without the institution and how do we know?
The questions map well onto FATF. They may even be more difficult. If the fundamental problem of causality is complex in all social science inquiry, the challenge becomes greater still in studying illicit activity. Numbers are notoriously unreliable. The collection of papers in Andreas and Greenhill ‘s Sex, Drugs, and Body Counts vividly illustrate the problem.
A second problem is that scholars often use the word “effectiveness” to signify different concepts. Here the important divide is between what I will describe as endogenous or exogenous conceptualizations. Endogenous understandings define effectiveness according to the more immediate goals of the institution. Exogenous understandings focus instead on the impact an institution has on larger problem members understand the institution as addressing.
In the case of FATF, an endogenous approach might measure the diffusion of AML rules or the prioritization of AML. Are states getting the correct laws on the books? Are they implementing those laws? The exogenous conceptualization, in theory, is more straightforward: Is the AML regime slowing the amount of money being laundered?
Until the most recent round of mutual evaluations, FATF members used an endogenous definition. This approach is easy to critique. FATF might be perfectly effective at diffusing perfectly ineffective rules. Consider, however, the counterfactual: if FATF had the perfect rules for a perfect AML system but was unable to diffuse those rules, the AML regime would still be ineffective. Diffusing standards is a necessary, although insufficient, condition for an effective AML regime. Most of the themes above can be thought of as endogenous indicators of effectiveness. I briefly highlight three especially prevalent ones, re-treading some ground covered above to place the questions in the context of effectiveness.
The first is participation. Who should participate in FATF decision-making and what role should they play? Romaniuk and Keatinge show the dangers of simply imposing rules without consultation, but Tsingou’s work suggests that the AML regime may be susceptible to regulatory capture by the compliance industry. Nonetheless, many of the authors also make it clear that the private sector, both for-profit and non-profit, has a role to play. One measure of FATF’s effectiveness, therefore, would be to gauge that role. Does FATF consult them without bending to them? Does the private sector generally accept the outcomes of the governance process, even when the rules promulgated cut against their preferences? Are only service providers included, or are consumer advocates at the table, too? What about private sector representatives from all levels of markets, from developing to developed? What role is there for private sector security actors?
The second important theme regarding endogenous effectiveness sits somewhat in tension with the first, namely the ability to adapt to new challenges. With its focus on financial integrity, how effective is FATF at identifying and responding to new forms of money laundering?Footnote 10 The tension with a participatory understanding of effectiveness lies in the numbers. On one hand, increased participation brings greater experience, wisdom, and insight from the “front line” that would help identify and respond to challenges. At the same time, too many voices create a din from which it is hard to create order. The balance that FATF members must strike is between front-line experience and an efficient, independent decision-making process.
A third aspect of effectiveness is FATF’s adaptability to different contexts, which nearly every article addresses in some way. If one goal of FATF is to promote financial integrity while still encouraging financial innovation and growth, that no doubt calls for ensuring that jurisdictions adapt FATF’s 40 Recommendations to their local context. The risk-based approach (RBA) to AML in theory helps insulate the regime against one-size-fits-all approaches, but the RBA is far from settled practice in FATF. Members therefore could incorporate the implementation of a risk-based approach to AML as a partial marker of broader adaptability. Highlighting the interconnectivity among these angles of effectiveness, one way to promote adaptability of FATF’s standards is to ensure that participants in all FATF fora represent a broad array of development levels. In this regard the inclusion of the FATF-style Regional Bodies as associate members is a step in the right direction. Ultimately, it may do more to highlight their second-class status than to ensure that they are able to contribute to a more adaptable, more effective FATF. As the network grows, however, there is a danger that the Recommendations become top-down regulations. That cuts directly against the idea of contextualized adaptation.
While the exogenous definition of effectiveness is conceptually more straightforward—whether or not money laundering is less common—there are several serious obstacles to measuring that variable. The best example is a metric used by FATF itself: the number of money laundering convictions in a jurisdiction. FATF is critical of jurisdictions that do not show an increasing number of convictions. The problem with that measure is that a jurisdiction with the perfect AML system and a jurisdiction with no AML system would yield the same number of convictions: zero. In more nuanced terms, a strong AML system should generate a decreasing number of convictions as deterrence takes effect, but so would a weakening AML system, as fewer launderers are prosecuted. In reality, the number of convictions matters only relative to the number of times someone tries to launder money, but we do not have that baseline measurement. If we do not know the starting line, it is difficult to know how far we have come.
That said, reliable data is a problem in most any field, especially in the early phases of inquiry. The key is to begin gathering better data. That seems especially true in a field like international finance, which in so many ways is about the creative construction, production, and utilization of data. What the field needs—both the academic field and the field of practice—is a more concerted effort to solve the data problem. There are good examples of partial measures. Robert Kudrle , for example, has examined whether blacklisting leads to any demonstrable financial impact. He finds it has not. Importantly, that does not mean that FATF is in effective overall, but that blacklisting may not be a primary mechanism of effectiveness. Jason Sharman  conducts a bold experiment to see how easily he can open anonymous shell companies, in direct contradiction of FATF standards. He finds that it is disturbingly easy. These two works are excellent examples of the kind of research and oversight that the AML regime needs if we are interested in an exogenous understanding of effectiveness. In this same vein, the best measure is probably to track the price of money laundering. What premium do institutions charge for breaking AML regulations? Research and analysis that focuses on direct measures, however, remains the exception to the rule.
It is on this point that Michael Levi, Peter Reuter, and Terence Halliday take FATF to task in the final article of this special issue. They provide one of the first assessments of FATF’s efforts to measure its own impact on money laundering. They begin with the 3rd round of mutual evaluations, which included a provisional attempt to prompt thinking on AML effectiveness. The reports, they write, are inconsistent and incomparable. In reporting on the efficacy of Suspicious Activity Reports (SARs) systems, for example, Levi, et al., note that the evaluations ignore that some states have a threshold-based automatic reporting system, while others rely on human analysis to identify odd-seeming transactions. The result is that analysts can do little with the data on the numbers of SARs. Nor do the reports provide any qualitative assessment of the legal follow-ups to the SARs, which the authors note is probably more important than the numbers of reports generated. A lack of comparable data on outcomes makes learning even more difficult.
The 4th round evaluations, which have just started, do not promise to be much better. Levi et al., consider the National Risk Assessments, which are to establish the kinds of money laundering that are most likely in a state and serve as a guide in future evaluations. The reports require substantial cross-sectoral collaboration within a country, which is positive, but the content of the reports to date is not promising. The reports vary in quality and format. Most seemed to be simple aggregations of widely available data in acritical and irregular forms. As Levi et al., write: “The NRA, admittedly in its initial implementation, suggests how weakly FATF has articulated the role of data and data analysis in the assessment process and/or how modest have been the attempts to implement that requirement.” The NRAs to date seem to be a shaky foundation upon which to build an effort to judge FATF’s effectiveness going forward.
FATF leadership has been frank about the fact that measuring effectiveness will be difficult. No one disagrees with the point, but it verges on becoming a “common sense” truth that escapes scrutiny. The difficulty of gathering data on illicit finance should not stand as an excuse to stop trying. Like weather forecasting, we can only begin to identify meaningful patterns within the data after we have enough data points to consider.Footnote 11 Recall the problem with using the number of money laundering convictions as a marker of effectiveness. If we begin gathering data on that question (including historical data), we can begin to establish common patterns. Do they go steadily up as AML implementation advances? Is it an arch that sees numbers go up, then down? Is it a concave, but upward sloping curve, that sees an overall increase, but with a declining rate of increase after a certain point? Or is there no change? Comparisons with other crime rate analysis also would seem to be a productive project. Building the database that is required to better measure the effectiveness of FATF and the AML regime is no doubt a massive, multifaceted undertaking. It requires theorization about the variables that are most likely to matter and actually gathering the data. It is bound to be an inefficient, pragmatic process. It is also where the private sector might be compelled to participate more directly. Acknowledging the challenges of trade secrecy and client privacy, relevant financial and non-financial institutions should be sharing data, as should software companies that develop tools for AML compliance (and which therefore have data about patterns of financial flows). It also is a question that should see much greater engagement between academic analysts and practitioners.
Ultimately, the best measure of effectiveness is one that includes both endogenous and exogenous measures. The importance of legitimacy in promoting compliance is perhaps the best example that the two conceptualizations are different sides of the same coin . Just as neither GDP growth nor unemployment reflect the health of the economy, no single metric will fully measure the relative efficacy of FATF and the AML regime. Nearly three decades into FATF’s life, however, we are long overdue for a sustained dialogue on developing meaningful, multifaceted measures of effectiveness.