Informal organizations over time: Why do member states increase delegation?

Domestic factors affecting powerful countries

Domestic factors for explaining international cooperation outcomes have long been discussed in the IR literature. Scholars have examined how participating in IOs and signing treaties can signal credible commitments to a particular policy, the influence of domestic political institutions and their incentives, and/or the effect of the preferences of important domestic business groups (Schneider and Slantchev, 2018; Simmons, 2009; Vreeland, 2008)

This scholarship has also shown how, over the years, a given problem can rise in importance for member states due to new demands from domestic audiences (Tomz, 2007). For instance, exogenous changes or crises make certain topics more salient to domestic publics who demand that governments take these issues more seriously. Although a topic may have been a matter of national importance from the beginning, which is why a given state decided to dedicate resources to a particular organization, it can become further prioritized, resulting in a reconfiguration of preferences on the IO’s design.

I assume that this will be especially relevant if the country or group of countries affected by these new domestic demands are those that hold the most power within the IO, which are often better equipped to leverage their positions when bargaining and pushing for their preferred design outcomes (Fearon, 1998; Grieco, 1988). Furthermore, there may be cooperation problems for which this is particularly important, such as when governing policy areas where free-riding behaviors that were not anticipated in the initial design are hindering cooperation or when countries are dealing with a weakest-link problem where the provision of the good is conditioned by the behavior exhibited by the least-compliant contributor.

Due to the capacity of actors to identify and learn from their previous institutional experiences and incorporate that feedback into their calculations (Fioretos, 2017: 377), I expect that when states face the decision to reform an informal IO, they will seek to include design features that have worked well in previous IOs, including formal ones.

If governments need to send a signal to domestic constituencies that they are taking a matter seriously in the international arena, the literature has shown that powerful states have some incentives to formalize international cooperation when they want to use institutions as commitment devices (Ikenberry, 2001; Stone, 2011). One mechanism through which states can increase the formality of informal IOs without changing their complete design to use them as commitment devices is by increasing the amount of authority delegated to them (Rodriguez-Toribio, 2022). Secretariats and independent staff have shown an ability over the years to boost cooperation by operating as “neutral” actors through which information can be centralized and shared, which can help to solve cooperation disagreements among countries with different interests. This leads to the first theoretical expectation:

Through this strategic approach to the change in informal IOs, powerful states can still use informal IOs with relatively lower costs than their formal counterparts while enjoying some of the benefits of a formal centralization body for cooperation.

Functionalist factors

The complexity of dealing with a given cooperation problem can also increase over the years due to different endogenous and exogenous factors, including changes in IO member preferences and external shocks. The direction of this complexity may not have been foreseen by the initial designers of the IO and may risk overwhelming the design and capacities of the organization.

When this happens in formal IOs, states have often sought expertise from technical bureaucracies and agreed to delegate some agency and control over processes to them (Barnett and Finnemore, 1999; Franchino, 2004; Hawkins and Lake, 2006; Nielson and Tierney, 2003). For instance, even though the state representatives on the Executive Board are responsible for approving IMF programs, they mostly consider the data and reports presented to them by the IMF’s secretariat staff (Chwieroth, 2013). They also often have specific technical expertise, including regional knowledge, which can help to target policies to the on-the-ground realities in the countries expected to contribute to the solution.

The challenge with informal IOs is that, by design, they lack this pool of empowered bureaucrats that are able to maintain the vitality of an IO over time (Gray, 2018). This could be problematic if the IO becomes unable to perform its functions effectively over the years, as it could lose its role in an increasingly complex institutional context by being replaced by competing IOs.

However, as stated before, I assume that states are able to learn from previous experiences and based new decisions on successful institutional practices (Fioretos, 2017). Thus, increasing delegation to the IO and/or incorporating empowered bureaucrats into their design could allow informal IOs to address more effectively cooperation problems which complexity is increasing, serving as a compromising solution. It would mean that states lose some sovereignty, but in a controlled and limited way, in exchange for greater effectiveness of the IO and, hence, helping its vitality.

Although technical expertise can be “borrowed” from other IOs (Roger, 2022), state’s members in informal IOs may prefer to count on their own bureaucrats when they predict that they are going to be dealing with the problem for a long period of time instead of in an ad hoc situation. This would allow them to control the delegation process more tightly, while securing the vitality of the organization over time, as Gray (2018) shows in the context of formal IOs. This leads to the second theoretical expectation:

Alternative explanations

While the two theoretical expectations outlined above are indeed important mechanisms that I argue should be studied in more depth, there are also other alternative explanations that could be considered when examining the amount of authority delegated to informal IOs over time.

For instance, it could be argued that increasing the amount of delegation to an informal IO is a measure to boost accountability and legitimacy over time. Some authors have argued that while states are increasingly using informal institutions to govern cross-border problems, these designs have been conceived as the mechanism that states may use to evade political control and strengthen the power of bureaucrats as legislative actors (Lischewski, 2022). Recently, Roger (2022) has argued that this perception may be due to the study of procedures in isolation, without considering how informal IOs can also have accountability costs associated due to their linkages with formal institutions. It could as well be that redesigning parts of the IO so that it appears to have some independent or technical body such as a secretariat with hired professionals could help maintain the organization’s legitimacy.

Informal IOs could also be considered to have an “inferior” or “weaker” design compared with formal organizations. This would explain why, over the years, their more eager members will seek to include design features that are more formal than the ones they could have added during the initial negotiation stage. However, informal IOs are often designed to deal with issues that are already included in the mandates of large, well-established FIGOs and could be addressed by them, which deems this explanation alone an incomplete picture of what happens with all those IOs that overlap in mandate with formal organizations, of which there are many.

It could also be argued that informal organizations represent an initial cooperative stage, where trust can be established more easily. If this task is achieved, we may observe organizations becoming larger and with more delegation. We have observed this phenomenon in various European Union (EU) agencies, like Europol. While this explanation can partly clarify why states might opt for informal cooperation over formal arrangements in specific issue areas, it fails to explain why in some cases there is trust and no delegation and in others trust and delegation. In my theory, trust is conceptualized as part of the learning process, including their ability to discern over time what actors are reliable partners for cooperation.

Another approach would be to assign a more prominent role to the international environment and political context, suggesting that changes in these factors have prompted adaptations in informal IOs. While this approach is widely utilized in the literature, I view shocks and changes in the context as pivotal moments capable of influencing other variables, like countries’ preferences. In this regard, I emphasize the significance of contextual events. However, I argue that directly linking exogenous changes to the increase in authority delegated to an IO might overlook intermediate steps in the process.

Case selection

To the best of my knowledge, the institutionalist literature still lacks empirical analysis on how delegation of authority occurs in informal organizations, particularly informal intergovernmental organizations (IIGOs). One of the main data sets on IIGOs in the discipline, assembled by Felicity Vabulas and Duncan Snidal (2013, 2021), explicitly excludes informal IOs with secretariats, which are often used as proxies for delegated authority in the literature.

For instance, in their latest data set presented in the article “Cooperation under Autonomy: Building and Analyzing the Informal Intergovernmental Organizations 2.0 Dataset,” Vabulas and Snidal include 244 IOs that cannot be conceptualized as IIGOs according to their definition, but that they consider to be “close calls.” Out of these 244 organizations, at least 44, or approximately 18%, have some sort of formal secretariat and could have varying degrees of delegation granted to them.

Similarly, other data sets, such as the one assembled by Roger (2020) in his book or by Roger and Rowan (2022) in subsequent articles, do not collect specific information on this variable. More importantly, there is still not enough empirical exploration of how this variable may change over time.

This is why I have opted for a qualitative comparative design that allows us to study why informal organizations may see the authority delegated to them increased over time. I have selected the FATF, which is one of the IOs leading the global action on AML/CTF, founded in 1989 by the G7 as an informal IO without universal membership and without a secretariat or independent bureaucrats (FATF, 1990). The FATF is a good test bed to start testing the plausibility of theory-derived expectations about IO evolution over time because: (1) the IO has been operating for more than 30 years, which allows to study its development over a long period and (2) the amount of authority delegated to the IO has fluctuated over time.

Furthermore, the FATF meets the criteria of being both theoretically and empirically relevant. Theoretically, the FATF is what in process tracing is called a “positive case” in which my theory-based expectations could be analyzed (Van Evera, 1997). It is an illustration of a type of IO that is increasingly common in global governance (Roger, 2020) and the lessons learnt from it could help us move forward our knowledge about how states use these actors in contemporary international cooperation. Also, the lack of data and clear evidence are key handicaps of the broader literature on the AML/CTF global governance complex (GGC) (Deleanu, 2017; Levi et al., 2018; Yepes, 2011). This makes the results of this article relevant to scholars and practitioners focused on security and financial regulatory systems as it uncovers an important amount of novel data.

Once the case was selected, I had to decide how I will proceed with the comparison. I selected two shocks to structure the study of the design of the FATF more than 30 years. Even if reforms to institutional design can occur gradually in normal times (Carnegie and Clark, 2022), scholars studying IO reform have long highlighted that critical junctures like shocks that directly affect the countries involved in the IO and the issue at hand are likely to rapidly generate more notable changes (Colgan et al., 2011; Krasner, 1983). Collier and Collier’s (1991) research, published as Shaping the Political Arena, presents a complex form of the before-after design, which inspired the present research design. ³

I focus on two prominent shocks to the AML/CTF regime that raised the domestic salience of these issues and increased the need for greater participation and compliance with the regime’s rules from countries with substantial differences in their law, financial and enforcement systems: the terrorist attacks of 9/11 and the 2008 financial crises. This allows me to compare how the amount of delegated authority to the FATF in three different periods of time, while controlling for institutional context.

Data

The data were collected from different sources, including institutional documents of the FATF and other organizations, both formal (e.g., the IMF) and informal (e.g., the G20), and in-depth interviews with elites. The timeline of this research reaches from 1989 (when the FATF was established) to 2019 when the organization turned 30 years, just before the collection of data began. This timeframe allowed an in-depth analysis of the theorized expectations while controlling the point at which new data were being added. ⁴

As a starting point, annual reports and FATF reports to the G20 were selected from the broader population of documents for two main reasons. First, these documents, especially the annual reports, are where the FATF reflects on the changes to its design. ⁵ Second, the FATF reports to the G20 and its other types of Ministerial Declaration (which are always referred to in their respective annual report) were less descriptive and contained the official justifications for a particular reform.

From a population of 859 documents, I selected what the FATF calls “reports,” which in total comprised 37 documents (at the time of data collection 2020–2021), because of the high relevance of their content to the research question. Of these 37, 31 were Annual Reports (of between 30–70 pages each), and the rest were FATF reports to the G20.

The bias associated with each type of document was considered during the data collection and analysis stages to approach data from a reflective perspective. For instance, I acknowledge that the information available in the FATF documents is biased toward what its members want the public to read, and it may also even contain a bias toward the interests of its leading members (Gonzalez-Ocantos and LaPorte, 2022). This is why I also conducted elite interviews to triangulate my findings.

The informal nature of the FATF in combination with the limited prior research focusing on this organization made it necessary to gather data through interviews with elites to shed light on the causal mechanisms through which a dimension was further formalized or left informal. Intensive, in-depth interviews are especially useful when seeking to understand motivations and preferences (Rathbun, 2008). In fact, this method allows the gathering of data that would otherwise be extremely difficult to access (Berry, 2002).

After building a clear idea of the main changes to the design of the FATF over the years, the process of carrying out the elite interviews began. I decided to use semistructured interviews to adapt to the experience and background of my interviewees and obtain as much information as possible on their specialist areas of expertise. Also, sensitive topics such as money laundering and terrorism financing are easier to discuss one-to-one, as this format means confidentiality in a trusted environment and high discretion can be ensured.

A total of 21 interviews were carried out, and two of them off the record. I followed Kvale and Brinkmann’s (2008) approach which argues that we should “interview as many subjects as necessary to find out what you need to know” (Kvale and Brinkmann 2008: 113). When I had conducted these 21 interviews, I had already obtained the necessary data to answer my research questions as the interviews were carried out thoroughly, including several preparation hours (4 hours minimum each), were of long duration (1-hour minimum), and I ensured the careful management of transcripts and notes afterward (which took at least the duration of the interview multiplied by 2.5 and were personally carried out by the author).

The interviews were carried out between August and December 2020. The first one, which was treated as a pilot interview, was face-to-face and a relatively long interview. The rest of the interviews were carried out via either the Zoom or Microsoft Teams video chat platforms, with the researcher based in the United Kingdom.

The interview guide was developed using different sources, including the existing literature on the AML/CTF regime and the functioning of IGOs, data obtained from documents on the issue, and the experience the researcher had gained from personally participating in international forums. The interviews were not intended to assess or confirm the exact variation in levels of delegation, but to generate context on the process that had led to these changes. The interview guide addressed several general themes that were presented to all participants, with the focus depending on the experience of the person being interviewed.

In the case of interviews, I acknowledge that both personal and professional bias may be present. I approached each interviewee with an awareness of how their nationality or professional profile could be shaping their discourse.

I further triangulated these data and put into context by consulting secondary data sources, including academic literature, media outlets, public data sets, and reports from the private sector, the non-profit sector, and other IGOs. ⁶

Process tracing

Both documents and transcripts of interviews were process traced to analyze whether the theoretical explanations of interest had happened or not by observing “the sequence and timing of events and contemporary interpretations of those events” (Hall, 2013: 22).

To elaborate or assess theories, process tracing should have expectations introduced before starting the analysis so that the researcher can assess whether observed processes match the theorized causal mechanisms or not (Hall, 2003). However, at the same time, the in-depth search for evidence allows the researcher to be exposed to “serendipitous discovery,” as they might inductively uncover under-theorized factors that may be of relevance (Hall, 2003: 395). In other words, process tracing allows the researcher to study their topic both deductively and inductively. This allowed me to inductively obtain more precise measurements of levels of authority in the FATF at different points in time and to explore whether any crucial causal factors had gone untheorized.

In applying process tracing in this article, I followed what Bennett and Elman (2007: 183–184) consider to be the two crucial best practices of this method: (1) I paid explicit attention to alternative explanations in terms of “what else could be true” and (2) I undertook relentless empirical research into those hypothesized processes with a wide range of sources, always considering the potential motivations and biases of each source as highlighted above.

After building a clear image of the sequence of events, I classified the evidence I had gathered into three groups, following Blatter and Haverland (2012: 110–119). First were those observations that were necessary to provide a “coherent picture” of how the causal mechanism had worked, or not. Second, those pieces of evidence that could be considered “smoking guns,” which are a “coherent cluster of observations that closely link cause and effect in time and space.” Third, I used the term “confessions” to classify those pieces of evidence that provided me with “greater knowledge about the perceptions, motivations, and anticipations of important actors in crucial moments” (Blatter and Haverland 2012: 110).

Once the pieces of evidence were in sequential order and classified by importance, I could assess whether they were enough, or if I needed to consult further sources. When a satisfactory number/quality of pieces of evidence had been obtained for each period of time, I could write the finding section with the certainty that I have surpassed the “comfortable fit.” This means maximizing one of the strong suits of this methodology, which is providing high levels of internal validity and a rich quantity of data of great quality (Gerring, 2007).

The authority delegated to the FATF from 1989 to 2001

The FATF was the first IO ever established to deal solely with the problem of money laundering internationally (FATF, 1989), and its founding states put a lot of effort to highlight that they did not want to delegate much authority to the IO. As expressed by the founding members themselves:

The essential objective should be to maintain the informality which the FATF has adopted and to avoid a rigid bureaucratic approach (FATF, 1991: 17, emphasis added)

An interviewee confirmed that members indeed did not want to delegate authority to the IO itself, but instead control tightly how IO activities were performed by using their own national civil servants. ⁷ Another interviewee explained that this decision about the design derived from:

The clear dissatisfaction on the part of great powers with the perceived sluggishness of the United Nations’ handling of the issue of global money laundering. ⁸ (emphasis added)

Therefore, the founding members of the FATF and its allies, which together comprised the developed countries of the early 1990s, ⁹ had a strong preference toward governing the issue area of money laundering through an informal IO which did not have the same characteristics that the UN, including the delegation of authority to international bureaucrats ¹⁰ (FATF, 1991). This in line with the theoretical reasoning that states members of informal IOs will prefer to avoid delegating authority to the IO if possible because it provides benefits that they cannot acquire from their formal counterparts (Vabulas and Snidal, 2013). It also reinforces that informal designs can be intentional. In contemporary global governance when a new IO is created, it may be in spite of already having formal organizations dealing with the problem. In the case of the FATF, the informal design was not a “second choice” or a “stepping stone” for a more formal organization, but a choice of a design that was regarded as better suited to fulfill the interests of the IO members at that moment. ¹¹

The FATF was able to navigate its two first years without any sort of IO bureaucracy under the Chairmanship of France until 1991 when they decided that a small FATF Secretariat should be established in Paris within the OECD headquarters (FATF, 1991). This, however, did not translate into a direct increase in authority delegated to the FATF. Those who were at the FATF decision-making table during the 1990s confirm that the secretariat was tasked with an “assistance” role and only a couple of low-paid staff. ¹² As explained by one interviewee:

At the end of the day, practically all the work was carried out by state members, as the secretariat did not have the means or the capacity to actually get involved (in meaningful work). ¹³ (emphasis added)

This explanation is reinforced by statements from the FATF’s members reiterating their desire to avoid a big secretariat:

Size and cost of this secretariat should be extremely limited, probably in the range of 2 to 4 million francs each year. ¹⁴ (FATF, 1991: 19)

This position with regard to the FATF’s design was maintained during the 1990s (FATF, 1991–1999), where state members were able to handle practically all the operations of the institutions with their national staff. Although money laundering was an important issue for the G7 and the interests of some of their founding members can be understood in the context of the “War on Drugs” led by the United States, in some ways it was just another issue on a long list at the international level without necessarily being the first priority on many national agendas. ¹⁵ Furthermore, outside the developed financial world at the time, little attention was paid to this problem ¹⁶ and members were able to manage the function of the organization with their national experts. ¹⁷ Also, during the 1990s, terrorism financing was not a problem under the FATF mandate and was still only addressed at the UN as at that time there was insufficient will to incorporate it into the FATF mandate. ¹⁸

The authority delegated to the FATF in the aftermath of 9/11

The day after 9/11, everything changed in the regime. Going back to that period, it is impossible to overstate the political support that the US received after 9/11. It just burst everybody. ¹⁹

The topic of terrorism financing had been on the UN and G7 agendas for years. Until 9/11, there had simply not been enough interest to raise this issue as a priority on the international agenda. ²⁰

These are two quotes from interviewees in this study, recalling the importance of 9/11 when asked about its impact on the changes to the design of the FATF in the aftermath of the events, including the increase in delegation that I am analyzing.

When 9/11 occurred, the topic of terrorism financing was already being handled at the UN. In fact, the UN had already a sanctions regime to counter the financing of the Taliban and Al Qaeda, and a couple of years earlier had passed the UN Convention on the Suppression of Terrorism Financing (1999). After 9/11, the UN Security Council quickly issued Resolution 1373, an important part of which sets out measures to reduce terrorism financing.

However, “by 2001, the FATF’s success as an anti-money laundering standard-setting body led to a significant change to its mandate” (FATF, 2018: 12).

The-then US Treasury Secretary, Paul O’Neill, expressed the US interest in positioning the FATF as the institution leading the collective action against terrorism financing soon after 9/11 during an extraordinary FATF Plenary (O’Neill, 2001). He further proposed an action plan for the FATF, which included the adoption of special Recommendations on terrorism financing, and urged the IO to ensure that all countries, not just FATF members, were complying with the standards, by extending the self-assessment requirement to every country in the world (O’Neill, 2001).

This change brought an increase in the complexity of the FATF’s mandate and daily activities, as suddenly the organization was tasked with simultaneously dealing with two issues rather than focusing on just one. As recalled by one of the interviewees:

not even the FATF’s members were fully trained or knowledgeable about the technical details of the differences between AML and CTF at that stage. ²¹

Consequently, immediately after 9/11 and its aftermath the assistance from the secretariat sharply increased and members started to delegate more functions to this body. As theorized in expectation two, this unforeseen demand for technical expertise led states to delegate in its secretariat as explained by the FATF itself: “The secretariat prepares and produces policy papers discussed in working groups and/or in the plenary, organises mutual evaluation missions, and produces the related assessment reports” (FATF, 2008: 17).

The secretariat went from preparing the monitoring reports “with the supervision of the FATF (members)” (FATF, 1991: 18) to lead teams of evaluators provided by states members after 9/11 (FATF, 2006: 6). By 2006, the Executive Secretary thanked the FATF’s members for “substantially increasing their contributions” and added “with the arrival of these new staff members at the FATF Secretariat, we believe that we will now have the appropriate level of human resources to better support the work of this organisation” (FATF, 2006: 7). In October 2007, the FATF expanded its secretariat once again highlighting the increased authority it had experimented in recent years: “The work of the secretariat has expanded considerably over the years. In October 2007, the plenary decided to expand the Secretariat to ensure sufficient personnel are in place to meet the demands” (FATF, 2008: 17). During these years, the secretariat’s staff also intensified their visits to and contacts with the FATF’s regional branches and increased the FATF’s representative roles (FATF, 2007, 2008). Furthermore, when the budget of the FATF Secretariat was first made available in 2004, it was already 1,636,000 Euros, three times the maximum established when the secretariat was initially designed (FATF, 2004).

Thus, after the shock of 9/11, the FATF Secretariat went from solely operating in an assistance role to actively participating with its discretion in important functions of the organization such as the monitoring of its members and associated countries. In this particular case, both scenarios introduced in the proposed theory were present, although they did not appear separately as theorized but instead in sequential order.

First, the US (and allies) domestic audiences demanded greater effectiveness dealing with terrorist financing and reducing the number of countries that could serve as safe havens. ²² Different participants recalled how 9/11 and the global commotion that followed those events significantly influenced the changes during those years in how the regime was managed and how IOs were reformed. The fact that the hegemonic global power at that time had been attacked on its own soil made citizens around the world think that “if it was happening in the US, it could happen anywhere.” ²³ This led the FATF members to expand the mandate to cover this topic and to actively seek compliance with its regulations from a wider number of jurisdictions (including expanding its membership and convincing non-members to comply, too).

Second, all these changes overwhelmed the initial human capacities of the IO. To effectively deal with the expansion of its membership (coordination functions) and with the particularities of the new topic added to its mandate (substantive functions). The data analysis also indicates that it was crucial for the United States and its allies to establish the FATF as the focal point institution for addressing these issues as further corroborated by Zarate (2013) in his book Treasury’s War: The Unleashing of a New Era of Financial Warfare. This decision rendered other techniques of acquitting technical support, such as borrowing resources from other IOs, inefficient on their own.

The authority delegated to the FATF from 2008 to 2018

The number of staff hired did increase during these years slowly, but the duties delegated to them and how the secretariat was organized did not undergo major changes. The FATF Secretariat was finally consolidating in the years after 9/11 (FATF, 2008), when the global international financial crisis occurred in 2008. During those years, the FATF highlighted the relevance of the institution to boost financial stability and integrity (FATF, 2010). For instance, the FATF increased its focus on corruption (FATF, 2012). However, the FATF was not the chosen IO to take the lead internationally in the post-crisis financial restructuration.

Instead, the FATF engaged with support activities to increase financial stability while continuing to deal with other topics, including money laundering and terrorism financing. Furthermore, it also continued to expand its mandate to topics unrelated to the financial crisis. While corruption was important for those using the FATF at that stage as an international IO to represent economic preferences, proliferation financing was important for those using the organization as a broader geopolitical tool. ²⁴ Nonetheless, in any of those cases, neither domestic audiences nor heads of state perceived the FATF as their focal institution to obtain their main goals in those issue areas, as has happened with terrorism financing. Instead, states were resorting to the FATF to deal with new topics such as proliferation financing due to the ability shown by the organization in the previous years to get things done quickly (Zarate, 2013). In fact, the IO continued to highlight the preference of its members for an informal institutional design as the following quote shows:

(. . .) the task force structure of the FATF has served it well. The FATF should therefore retain this flexibility as it continues to consolidate and build on what it has already achieved. (FATF, 2012: 43)

At the same time that the FATF was adding new topics to its mandate, those that it was already addressing were increasing in complexity. For instance, the increased complexity of the financial system went hand-in-hand with the new technologies and new sources being used by terrorists to fund themselves, such as crowdfunding. Furthermore, apart from expanding the scope of issues covered in the mandate and the number of participants, the FATF also underwent a revision of its methodology, which was revised and updated to increase how the IO measured the effectiveness with which its standards were being implemented around the world (FATF, 2013).

All this required again an amount of technical expertise not available within the FATF. However, we cannot observe major changes to the delegation of authority to the secretariat during this period. For instance, when the methodology was reviewed and a risk-based approach adopted, expertise from the private sector was actively sought by FATF members. ²⁵ Also, when the FATF was discussing the drafting of the Recommendations on proliferation financing, its members resorted in its current secretariat and sought out complementary expertise from the UN to assist them. ²⁶

In fact, an official present during discussions on this topic shared that “(. . .) the feeling in the room was that this decision was more political than technical,” and that it was “detached from the rest of the regime.” ²⁷ She also explained that European countries were “pressuring very hard to have the IO bureaucracy from the UNSC for the drafting of the recommendations on proliferation financing.” ²⁸

In this case, a strong demand for greater effectiveness in dealing with the core topics of the FATF mandate was not coming from domestic audiences. And, contrary to what was observed in the previous period, the theorized expectation that a higher demand for technical expertise from the IO’s bureaucracy will lead alone to greater delegation was not met. In fact, in 2013 the FATF secretariat stated that

the resources available to the FATF Secretariat to carry out its tasks have become increasingly stretched. (FATF, 2013: 9)

The number of staff hired did slowly increase during these years to face the increasing workload within the remits that were delegated after 9/11, but neither the duties delegated to them nor how the secretariat was organized did undergo major changes during that period, as further confirmed in interviews for this study. ²⁹

The amount of authority delegated did not increased either during the rise of ISIS and its attacks in Europe, which affected domestic audiences in the founding members of the FATF and brought a number of issues requiring new technical expertise, from particular regional knowledge to understanding of new forms of funding, and how to combat issues such as foreign fighters (FATF, 2016). Contrary to what happened with 9/11, in 2016, the FATF was better prepared with a consolidated secretariat and network of actors to deal with the problem without increasing delegation and without damaging the reputation of the IO when doing so. ³⁰

In fact, between 2016 and 2018 long conversations and negotiations took place on whether the design of the FATF should change toward that of a FIGO (FATF, 2018). This path was rejected, as the members highlighted the benefits of the current design in comparison with a more formal one. ³¹ Notably, the most powerful members pushed hardest against the FATF becoming a formal IO. ³² This decision highlights how the FATF’s leading members continued to prefer a design closer to that of its foundation than to a more formal organization, which could further increase the authority delegated to the IO’s bureaucracy.

Furthermore, as a state delegate to the FATF explained, until 2019 the FATF continued to have two types of staff: those hired independently and those seconded by their countries (usually high-ranking officials). The number of staff from the secretariat varies from one area to another; ³³ for instance, those parts of the FATF with greater distributional implications such as the working group in charge of blacklisting countries only have a very limited number of people from the secretariat, and usually, some of those are seconded from governments. Other groups such as those handling the relationship with the regional branches have more staff from the secretariat. ³⁴

This shows that while in the aftermath of 9/11 there was a redefinition of preferences about the design of the FATF by its members, this was not sustained over the years despite the increase need for technical expertise. The FATF response to the 9/11 shock was regarded as effective and successful for its most powerful members (Zarate, 2013). With the design that resulted from the changes experimented in that period, the organization showed that was able to effectively influence the law and financial sectors of members and non-members (Morse, 2021).

Furthermore, although the effectiveness with which the international community deals with the topics that fall under the FATF mandate can be questioned, the IO itself been able to maintain over the years the image of an effective IO due to its “technical” profile and has not been displaced by other IOs. ³⁵ This has allowed FATF members to maintain an equilibrium in its design. The FATF continues to be an informal organization that carries lower cost than its formal counterparts (the UN, IMF, and World Bank (WB) primarily) while leading the collective action against money laundering (ML) and terrorism financing (TF) globally, with the support of a well-developed secretariat compromising independent staff.

Discussion of the findings

It was only under the exceptional circumstances following 9/11 that FATF members decided to increase delegation to the FATF. This occurred through the following sequence: (1) terrorism financing became a priority for domestic audiences of FATF members, (2) FATF members decided to turn the organization into a focal point institution within the global regime and expanded its mandate, (3) these new demands required more specialized human capital than was available within the organization, and (4) the urgency created by the concerns of domestic audiences and the need to address the issue quickly to position the FATF as a focal point institution, led states to soften their positions regarding the amount of delegated authority granted to the organization by design.

Proposition 1 anticipated that if the demand for effectiveness increases from domestic audiences to deal with an international cooperation problem, states will increase the authority delegated to the IO. While this pattern was evident after 9/11, it notably did not repeat following subsequent events, such as the 2008 financial crisis or ISIS’s attacks in Europe. First, these findings uncover how proposition 1 could be redefined to include the need for the IO to be portrayed as the focal point or main actor to deal with the problem at hand. We observed this after the 2008 financial crisis, when the institution was not used as a commitment device by its members to address the problem as it happened after 9/11.

Second, there is another key difference between the periods analyzed: the technical expertise already integrated within the FATF. Proposition 2 anticipated that if the IO needs an unforeseen amount of technical expertise to address problem effectively over a long period, states will increase the authority delegated to the informal IO. While this did occur in the case of 9/11—when the FATF was initially set up to deal with money laundering—it did not occur in later cases, as the IO had already consolidated a technical body with dedicated functions on the issue. For example, by the time ISIS launched its attacks, FATF members had already delegated authority for a number of technical functions related to CTF. Once the members delegate certain authority to deal with a topic, it may be enough to face sequent shocks on the same area, highlighting that is not enough that the IO needs technical expertise to deal with a given issue but that the amount needed it has to clearly surpass the one available.

Another potential scenario for states to increase delegation to an informal IO other than the one on effectiveness argued in this article could be when the reputation of the IO is at stake due to calls for more legitimacy or accountability. In the case of the FATF, calls for legitimacy and accountability have not focused on the size and autonomy of the secretariat. To start with, the FATF has been quite successful in presenting its image as a “technical IO” (de Oliveira, 2018), despite the politics behind the scenes, ³⁶ and its legitimacy has not been roughly questioned. ³⁷ While the professionalization of the FATF’s secretariat may have helped the IO preserve its “technical” image externally, the amount of delegation that the IO holds has never been widely discussed as a major design flaw in criticisms of the organization’s legitimacy, which often focus on its enforcement mechanism (the blacklist) and restrictive membership. ³⁸

Furthermore, the argument that an informal IO may be a stepping stone into a more formal organization that allows states to earn trust does not apply to the FATF either. In fact, the initial group of IOs were like-minded and could have agreed on any other design. The argument could be considered in reverse, though. Over the years, the FATF’s expansion has led to increased heterogeneity among its members’ preferences, which could have led states to seek independent control of the flow of information within the IO. Once again, this is not often linked to the amount of authority delegated to the FATF, mainly because powerful countries have agreed to control the amount of authority delegated to the IO, prioritizing their own national interests (regardless of whether those are shared or not with other FATF members). ³⁹

Moreover, while states may access assistance from independent bureaucracies by establishing links with more formal organizations, as proposed by Roger (2022), the FATF case also uncovers how informal IOs can develop their own bureaucracy, which may experience an increase in authority over time. This study reveals how states use the uneven distribution of staff across the functions of the FATF as a mechanism to exercise control and influence the IO, something that has broader implications for research on informal governance.

Finally, this research has also shown the importance of contextual events. However, it also highlights the risk we may incur by directly linking exogenous changes to the increase in authority delegated to an IO. As the FATF case shows, this may lead us to overlook important intermediate steps in the process.