From tax havens to cryptocurrencies: secrecy-seeking capital in the global economy

Abstract

The global backlash against tax havens has pushed secrecy-seeking capital to explore alternative opportunities in non-tax-haven countries and new financial technologies (FinTech). We identify two major corporate practices—organizational ring-fencing and swarming—that have enabled secrecy-seeking capital to adapt to new regulatory realities and illustrate these practices empirically with the extreme case of Estonia. In the 2010s, several Nordic banks turned their Estonian offices into hotbeds of high-risk transactions, ring-fencing their Baltic affiliates from their group-level systems and generating several money laundering scandals with global repercussions. More recently, secrecy-seeking capital ‘swarmed’ into Estonia’s large cryptocurrency sector and thereby thwarted effective supervision of the activities of the firms involved. Neither swarming nor organizational ring-fencing have been sufficiently explained by existing approaches in International Political Economy (IPE) as new core practices of secrecy-seeking capital. We study both practices in a mixed-methods research design and provide novel empirical insights to illuminate this phenomenon. In filling this gap, our study paves the way for a second generation of global tax governance scholarship amidst the cryptocurrency and FinTech boom, and calls for a research agenda that addresses these new practices that take advantage of the lack of administrative capabilities in non-tax-haven jurisdictions.

Keywords:

• Estonia
• cryptocurrencies
• FinTech
• tax governance
• tax evasion
• money laundering

Introduction

‘We hope that no one thinks that the money that flew through the Baltic countries few years ago is gone. No, it’s flowing somewhere else geographically or has taken some other form. Maybe it’s not money anymore, maybe it is cryptoassets.’ – Kilvar Kessler, Chairman of the Board of The Estonian Financial Supervision Authority, May 2021 (Kessler, 2021)

Following the 2007–2009 global financial crisis, data leaks from tax havens and civil society campaigns have paved the way for a gradual crackdown on tax havens.¹ Anti-money laundering (AML) regulations, the exchange of tax information between countries, and compliance work have been significantly strengthened in recent years (Ahrens et al., 2022; Eggenberger & Emmenegger, 2015; Rossel et al., 2021). Research on global tax avoidance, evasion and the regulatory responses against tax havens has evolved into a mature research field within IPE and beyond (Temouri et al., 2022; Ylönen & Finér, 2023). These developments notwithstanding, the demand for secrecy-seeking capital has persisted and has perhaps even grown amid sanctions on Russian elites. This poses challenges to the achievements in international regulatory guidance and coordination of tax information exchange, AML, and cross-border banking supervision (Chang et al., 2023). Secrecy-seeking capital—here defined as cross-border assets that are transacted or invested to conceal the identity of owner(s)—has historically received attention in IPE research through studies on authority shifting from the nation-state toward cross-border actors like accountancy firms or even organized crime (Kindleberger, 1970; Strange, 1996). In today’s global economy, the drivers of secrecy-seeking capital include tax evasion and avoidance, money laundering, and circumvention of sanctions. Fears of expropriation have also encouraged the usage of secrecy-seeking capital, especially in the context of sanctions against Russia (Cockfield, 2018).² Alongside increased demand for secrecy, financial and technological advances have generated new opportunities for circumventing rules and exploiting weak administrative capacities (Wójcik, 2021).

Consequently, the commonly used binary onshore–offshore distinction has become increasingly ‘futile’ (Stausholm, 2022, p. 15; see also Binder, 2019; Clark et al., 2015; Finér & Ylönen, 2017). Individuals and firms have adapted to regulatory and technical advances, which challenges IPE to move beyond the first generation³ research on regulatory responses to ‘classical’ tax avoidance and evasion, and towards new strategies of secrecy-seeking capital. In this paper, we extend the boundaries of scholarship on secrecy-seeking capital to include ‘onshore’ countries and bring it into dialogue with the nascent FinTech scholarship. We do so by mapping alternative practices that have emerged amid regulatory and technological advances, which induced new secrecy-seeking arbitrage strategies from private agents (Palan et al., 2023).

We identify two core practices—organizational ring-fencing and swarming—which explain how secrecy-seeking capital has adjusted to the global pushback against tax havens. Both practices exploit either the lack of administrative capacity to enforce existing rules, the non-existent cross-border coordination between administrations, or both. The first practice, organizational ring-fencing, occurs in the traditional banking sector and relates to the operations between the banks’ branches and subsidiaries and their parent companies. It captures how banks or other financial institutions insulate certain activities (e.g. IT systems and financial reporting) or corporate-level assets from regulatory oversight. The second practice we analyze is swarming, which describes the rapid influx of numerous firms into a single jurisdiction. As a result, the volume of incoming companies overwhelms regulatory authorities and hinders their effective monitoring.

We use the extreme case of Estonia to empirically illustrate the logic of these two practices, which have crucially undermined new stricter know-your-customer (KYC) requirements and risk assessment. Swarming also demonstrates how contemporary secrecy-seeking capital exploits relevant FinTech instruments like cryptocurrencies in a qualitative shift away from ‘offshore’ and toward virtual assets that rely on novel geographies (see, e.g. Campbell-Verduyn, 2018). We support our argument empirically by triangulating 30 semi-structured interviews; observational data from Estonian industry seminars; media sources and reports; and corporate-level data. This rich empirical basis allows us to comprehensively illustrate these emerging secrecy-seeking practices, laying the groundwork for further research that incorporates more cases and data sources.

We argue that organizational ring-fencing became an important practice when the financial secrecy of tax havens had been weakened through international policy measures such as the introduction of the first Organisation for Economic Co-operation and Development (OECD) tax haven blacklists in the early 2000s (Sharman, 2009). While these initiatives had loopholes, they nevertheless created incentives for developing arbitrage strategies that channel secrecy-seeking capital from tax havens to less controversial locations. As ‘onshore’ banks became enmeshed in such capital relocations, challenges in AML supervision emerged for the supervisors in banks’ home and host countries (Demetriades & Vassileva, 2020). Despite advances in the coordination of tasks between host and home supervisors in the 2000s, cross-border AML supervision remains difficult (Beck et al., 2013; D’Hulster, 2012; Isolauri & Ameer, 2023), which benefits secrecy-seeking capital. We illustrate how organizational ring-fencing practices allowed the Estonian offices of Nordic banks to serve dubious international clients while maintaining an impeccable public image. This generated one of the largest global money laundering scandals (Ewing, 2019; Milne & Winter, 2018). Organizational ring-fencing helps to explain how some of the largest money-laundering operations in the world could occur in esteemed Nordic banks—and in Estonia, with its poster-boy image of digital transformation and low corruption.

In turn, swarming has helped secrecy-seeking capital to thrive amid the most recent wave of anti-tax-haven and AML initiatives. We illustrate swarming by describing how the rapid influx of hundreds of cryptocurrency service providers into Estonia in the late 2010s led to supervisory overload and regulatory failure. More than half of the global cryptocurrency service providers were registered in Estonia in the early 2020s (RAB, 2022), making it a highly interesting case study for swarming practices. Advances in global AML and tax rules imply that regulators should be better equipped to monitor these firms. Yet, a sudden concentration of cryptocurrency firms in Estonia (i.e. swarming) obstructed regulatory oversight. Our analysis of swarming shows how small firms can successfully overcome regulatory state power in the age of FinTech and the virtual economy.

Existing analyses of secrecy-seeking capital or FinTech—in IPE and beyond—explain neither organizational ring-fencing nor swarming. This is a major research gap, also because of the financial and geopolitical significance of these practices. Transactions emanating from Russia were central to the money laundering scandals of the 2010s, and the cryptocurrency boom has been driven by entrepreneurs with connections to both Russia and Ukraine (RAB, 2022). Both practices are hence at the forefront of attempts to carve out spaces and assets for secrecy-seeking capital amid geopolitical and technological change and disruptions of ‘classical’ tax havens (Campbell-Verduyn, 2018). Our study lays important building blocks for a new wave of scholarship in IPE that is likely to emerge as an answer to the rise of novel secrecy-seeking practices and the associated new regulatory challenges and battles ahead.

We proceed as follows. Section Research design and data outlines our analytical strategy and methodological approach. Section The evolution of research on secrecy-seeking capital and its governance introduces the first generation of tax governance studies and makes a case for its closer integration with the AML and financial supervision agendas. Section New avenues for secrecy-seeking capital establishes our analytical framework. Section Organizational ring-fencing and the money laundering scandals of the 2010s tackles the Estonian money laundering scandals of the 2010s and Section Swarming: the Estonian cryptocurrency sector and its offshore linkages analyzes recent, swarming-related money laundering concerns. We conclude by highlighting our contributions to the broader research agendas of tax governance and IPE.

Research design and data

We follow an abductive analytical strategy (Peirce, 1878; Tavory & Timmermans, 2019). Abductive approaches enable the development of plausible explanations of empirical phenomena that are novel or inadequately understood. Given the novelty of the phenomena addressed here, an abductive approach is fruitful in allowing us to probe the emerging practices of secrecy-seeking capital. This enables us to develop plausible explanations about the conditions under which these practices arise and thrive, and which can be tested and refined in future research.

We began with the observation that secrecy-seeking capital has become increasingly established outside the confines of traditional tax-havens and with the help of newly emerging financial technologies. To illuminate this phenomenon, we introduce and analyze the practices of organizational ring-fencing and swarming using the extreme case of Estonia. As Yin (2002) explains, an extreme case can help illuminate the studied phenomena in a particularly ‘pure’ way. First, given that the offices of the Nordic banks in Estonia became the epicenter of globally significant money-laundering scandals in the mid-2010s, the Estonian experience offers valuable insights about how organizational ring-fencing emerged as a viable practice for secrecy-seeking capital. Second, since Estonia became a host to more than half of the cryptocurrency service providers in the world by the end of the 2010s, it offers revelatory insights about swarming.

We support our argument with several data sources. First, we rely on 30 semi-structured interviews conducted with civil servants, entrepreneurs, and service providers in 2020–2021 (Supplementary Material Appendix A). The interviews focused on the Estonian e-residency program, which has fostered the growth of Estonia’s cryptocurrency sector by allowing straightforward incorporation online. The interviews covered three groups: Civil servants, financial service providers and e-residents. We began interviewing people who had publicly commented on e-residency-related issues. Further interviewees were sought by contacting agencies tasked to regulate e-residents (e.g. relevant ministries and government agencies), from media sources, and through snowball sampling (Noy, 2008). We also contacted people via social media groups. Following our abductive methodology, the interviews continued until saturation points were reached to create a comprehensive picture of this emerging phenomenon.

Second, we draw on information obtained from three industry seminars that we observed in 2020–2021. They discussed Estonia’s advantages for entrepreneurs and regulatory challenges. Third, our mixed-method setting draws on media sources, assessments, and reports, particularly regarding money-laundering scandals that made headlines globally but have received little attention in IPE. We chose the dual strategy of relying on interviews for studying e-residency and crypto firms, and written materials for the analysis of the money laundering scandals of the Nordic banks to cover a longer period for our study and to counter data availability issues.

To illuminate the operational logic of swarming, we use data on the board composition of more than 1,300 cryptocurrency service providers obtained from the Estonian Register of Economic Activities, and Bureau van Dijk’s ORBIS database. This enables us to map the transnational connections of Estonian cryptocurrency companies by combining board membership data from ORBIS and the Estonian cryptocurrency license database on current and past license-holders. To understand the relationship between the nascent cryptocurrency industry and global corporate networks, we analyze the board interlocks of the Estonian-based cryptocurrency firms. Board interlocks occur ‘when a person affiliated with one organization sits on the board of directors of another organization’ (Mizruchi, 1996, p. 271). Interlocks depict direct linkages between board members, amounting to a corporate community. Interlocking directorates ‘simultaneously create direct links between the top decision-making bodies of corporate governance and tie together the corporate elite in a social network’ (Heemskerk & Takes, 2016, p. 93). The cryptocurrency industry is a ‘footloose’ community that traverses jurisdictions (Sigler et al., 2020, p. 622). Practices like swarming require at least some coordination and exchange between firms. Interlock analysis is hence a useful proxy for the information exchange networks that facilitate such practices (Brown & Drake, 2014). Our analysis provides insights on the linkages between this novel technological field and tax havens, which could not have been excavated from the interviews or existing assessments.

The evolution of research on secrecy-seeking capital and its governance

The first generation of modern global tax governance scholarship in IPE emerged as a response to the race to the bottom dynamics in international taxation and the ensuing governance failures, such as non-taxation, caused by the opaqueness of private wealth (Christensen & Hearson, 2019; Palan et al., 2010). An early milestone was the OECD’s (1998) report on Harmful Tax Practices, which outlined the regulatory challenges that IPE research dealt with in the following years. Further significant steps involved the blacklisting of tax havens and the gradual expansion of an automatic exchange of information on bank deposits and other assets (Ahrens & Bothner, 2020; Picciotto, 2022). Early American and European initiatives inspired the OECD’s Common Reporting Standard, which established an automatic exchange of information on financial instruments between approximately 100 participating countries (Hakelberg, 2016; Lesage et al., 2020). The policy agenda against corporate profit shifting was expanded in the 2010s in the OECD and beyond, as reflected in the global tax governance and IPE scholarship (Christensen & Hearson, 2019; Lips, 2019). Early evidence suggests that these initiatives have helped to reduce the use of tax havens (Ahrens & Bothner, 2020; Hakelberg & Rixen, 2021). Many tax havens have increased their sharing of tax information with foreign authorities, while data leaks have made these jurisdictions less safe for secrecy-seeking capital.

For the respective scholarship in IPE and beyond, these advances beg the question of how private actors respond to regulatory initiatives and what new arbitrage strategies are emerging (see Baker & Murphy, 2019; Seabrooke & Wigan, 2022). This question arises against the backdrop of two recent important developments that shape the newly emerging strategies of secrecy-seeking capital: The increased relevance of ‘onshore’ states and the rise of FinTech and cryptocurrencies as novel opportunities for regulatory arbitrage. For both developments, we argue that it is the lack of administrative capacity more than the lack of rules that facilitates the development of new practices of secrecy-seeking capital in the contemporary global political economy.

First, so-called ‘onshore’ states receive greater scrutiny today regarding their role in global tax avoidance and evasion (Binder, 2019; Finér & Ylönen, 2017; Stausholm, 2022). Recent studies on the integration of various jurisdictions in global capital circuits showed the limits of the commonly used on/offshore distinction in global tax governance (Garcia-Bernardo et al., 2017; Sharman, 2017). Newly emerging practices of secrecy-seeking capital are therefore unlikely to follow ‘old’ patterns of on/offshore jurisdictions, but exploit existing regulations in novel ways. Hence, the opportunities for secrecy-seeking capital today are made possible by low administrative capabilities for effective cross-border coordination between supervisory authorities (in ‘onshore’ and ‘offshore’ states) rather than obvious regulatory loopholes. In the first part of our empirical study, such capability gaps relate especially to cross-border banking supervision and the division of labor between the bank’s headquarter country regulators (home country) and the regulators in the country of the subsidiary or the branch (host country) (e.g. Beck et al., 2013; D’Hulster, 2012; Singer, 2007; Walter & Sen, 2009).

Allocating responsibilities between banks’ host/home supervisors have historically focused on micro-level issues, such as capital adequacy requirements, and less on cross-border AML supervision, which is an important entry point for secrecy-seeking capital (Demetriades & Vassileva, 2020; Isolauri & Ameer, 2023; Jones & Knaack, 2019). One central coordination issue is information access. According to the core principles of effective banking supervision, home country supervisors should have access to all information required for effective consolidated supervision. Cross-border operations should also be effectively overseen in both home and host countries (BCBS, 1996, 1997; Fiechter et al., 2011). Host countries are tasked to oversee AML supervision of local subsidiaries, ‘while the home country is responsible for prudential supervision at the consolidated level’ (Huizinga, 2018, p. 7; Fiechter et al., 2011). This requires the home supervisor to incorporate information on AML compliance in the host country (Huizinga, 2018).

Despite the clear regulatory and supervisory burdens and tasks, the actual enforcement ability of both home and host states determines whether and how secrecy-seeking capital can exploit the global financial system. In the words of Kirschenbaum and Véron (2018, p. 86), the EU’s AML system is ‘extremely complex, with an awkward coexistence of national and supranational features, [with] many new and untested aspects[.]’ The ensuing complexities in the division of responsibilities and access to information created a fertile ground for the money laundering scandals discussed in Section Organizational ring-fencing and the money laundering scandals of the 2010s.

A second key development after the crackdown on tax havens is the rise of FinTech and especially the associated cryptocurrency industry, which has introduced additional problems for administrative supervision and enforcement of rules to regulate secrecy-seeking capital (Lai & Samers, 2021; Wójcik, 2021). Cryptocurrencies have been described as potential ‘super tax havens’ due to their ability to undermine existing monetary and fiscal regulations (Marian, 2013). Hence, they enable secrecy-seeking capital to circumvent classical routes that fall under stricter regulatory supervision. To prevent such illicit activity, AML and KYC policies target the infiltration of criminal money into global financial systems and networks (Rossel et al., 2021). Global AML standards were first formulated in 1989, following the establishment of the Financial Action Task Force (FATF) (Kirschenbaum & Véron, 2018). The FATF stipulates 40 recommendations that encompass AML rules for financial institutions to be enforced by public authorities (ibid.). At the European level, the EU amended its AML Directive twice (2015, 2018) and now classifies tax evasion as a predicate offence to money laundering (ibid.). Yet, IPE research on tax havens has largely focused on tax information exchange and corporate taxation, rather than AML-related concerns (but see Sharman, 2017; Tsingou, 2022). However, the increasingly tighter navigation space of secrecy-seeking capital heightens the relevance of AML for a second generation of tax governance scholarship.

Reflecting the recent evolution of FinTech, organizations such as FATF have also addressed the issue of virtual assets and cryptocurrencies since 2014 and issued best-practice recommendations and updates for regulators (FATF, 2014, 2022). The EU also finalized its comprehensive cryptocurrency regulation regime, ‘markets in crypto-assets’ (MiCA), in June 2023. Likewise, various other jurisdictions have implemented AML guidance on virtual assets and cryptocurrencies. However, this can facilitate jurisdiction-shopping of secrecy-seeking capital, with the associated risks of money-laundering (Pavlidis, 2020). These developments demonstrate that cryptocurrencies and their underlying (FinTech) technologies present a major AML challenge for their global governance (Campbell-Verduyn, 2018). As our analysis of swarming shows, these challenges often arise from the lack of administrative capacity to deal with crypto-related practices of secrecy-seeking capital.

New avenues for secrecy-seeking capital

We identify and operationalize for the first time the practices of organizational ring-fencing and swarming to understand how secrecy-seeking capital has responded to the changes in the regulatory environment. These practices help us to understand why secrecy-seeking capital persists in a situation where progress in the exchange of tax information has led to conclusions that the ‘prerequisites for tax evasion have largely been eliminated’ (Hakelberg & Rixen, 2021, p. 1143). We argue that, stronger formal rules for information exchange notwithstanding, the key to effective supervision is whether states can actually utilize those rules. The organizational fragmentation of multinationals into complex organizational networks and the rise of FinTech have introduced major challenges to their effective utilization (Christensen et al., 2022; Reurink & Garcia-Bernardo, 2021). Both swarming and organizational ring-fencing have been enabled, in different ways, by these new developments.

Organizational ring-fencing builds on the well-established idea of juridical ring-fencing. The OECD (2022) defines juridical ring-fencing as a ‘theoretical enclosure established by tax legislation around certain profits, losses, transactions or groups of transactions in order to isolate them for tax purposes’. It has allowed tax havens to tax their citizens while attracting disproportionate amounts of untaxed wealth from agile investors and corporations. In contrast, organizational ring-fencing entails the practices of banks or other financial institutions that insulate certain corporate-level activities or assets from the oversight of relevant authorities and act as intermediaries for secrecy-seeking capital.

Organizational ring-fencing is related to the concept of corporate arbitrage, i.e. ‘the use of corporate legal entities located in diverse jurisdictions to arbitrate a third country’s rules and regulations’ (Palan et al., 2023, p. 1). Legal affordances convey a similar idea by referring to absences, ambiguities and arbitrage realized via strategies that allow firms to exploit legal differences between countries (Grasten et al., 2023). However, rather than exploiting national differences in ‘laws of incorporation and corporate governance’ (Palan et al., 2023, p. 15), organizational ring-fencing exploits differences in governments’ effective capabilities to monitor firms across countries (c.f., Baker & Murphy, 2019; Ylönen et al., 2023). Hence, organizational ring-fencing adds an important dimension to analyses of corporate arbitrage and legal affordances. Understanding these capability differences matters, because failures in AML work often arise from national difficulties in financial supervision. Such difficulties are regularly highlighted in the country reviews conducted by international organizations that monitor national AML progress (Tsingou, 2022). It is hardly a surprise, then, that banks may help secrecy-seeking capital to exploit national administrative deficiencies and weak cross-border coordination of AML supervision between host and home country supervisors.

Organizational ring-fencing can be used to separate the control and reporting practices in country offices from the rest of the group. Just as tax havens aim to collect ‘rents’ from mobile capital by providing access points to global finance (Palan et al., 2010), organizational ring-fencing enables banks to collect rents by granting secrecy-seeking capital entry points to the global financial system.

Our Estonian case study resembles experiences from other money laundering scandals. Consider two recent decisions of the British Financial Conduct Authority (FCA). It imposed major fines in the money laundering cases of Deutsche Bank (in 2017, £167 million) and Commerzbank (in 2020, £38 million), citing reasons that resonate with organizational ring-fencing (FCA, 2017, 2020). Both schemes enabled secrecy-seeking capital from Russia to enter global financial markets through German banks. Banks used their subsidiaries in London to exploit gaps in monitoring and supervision. Deutsche Bank’s arrangement also relied on its subsidiary in Moscow (FCA, 2017). As a third example, the so-called Russian Laundromat money laundering scheme centered around Moldova and its deficient administrative practices (Tofilat & Negruta, 2019).

Our second practice, swarming, refers to the influx of numerous firms into a single jurisdiction. In such a situation, the large-scale influx of companies may hinder their effective monitoring. Swarming relies on the decentralization of secrecy-seeking capital, in contrast to the centralized practice of organizational ring-fencing. Analyzing swarming helps to address a major gap in FinTech research, given its dominant focus on transactive instead of its systemic dimensions (Omarova, 2020). The relation between the governance of FinTech and global tax governance constitutes one such systemic dimension. The second part of our case study shows how the decentralized nature of FinTech helps secrecy-seeking capital to uphold its secrecy amid tax haven crackdowns. Again, the regulatory challenges arising from this decentralization relate to global imbalances in effective supervisory capacity more than to national regulators having insufficient formal powers. A country may fully commit to international tax, AML and financial governance but fail to control swarming. This happened when the influx of cryptocurrency firms overwhelmed supervisory capacity in Estonia.

Beyond our case, the statements of two important regulatory agencies underline the seriousness of this global risk. First, the Financial Stability Board (2022, p. 19) warns that ‘[m]any regulatory authorities do not have adequate resources’ to verify the accuracy and reliability of on-chain transactions. The second statement relates to the European Parliament’s recent proposal to force automatic reporting of large transactions with un-hosted crypto wallets to the authorities. Data Specialist Joana Neto from the European Banking Authority notes how this could overwhelm authorities across Europe: ‘It’s very resource intensive […] Who’s going to handle this? […] If it’s going to be the competent authority, what are they going to do with that information?’ (quoted in Schickler, 2022).

Such phenomena thrive in the fast-growing cryptocurrency sector, with its networks of nimble service providers. In late 2022, an industry source listed 528 cryptocurrency exchanges globally (Gravity Team, 2022). Furthermore, this industry involves digital wallet companies and similar actors. The small size of most service providers enables them to relocate quickly to jurisdictions with enticing regulatory environments. While Estonia stood out as a crypto hub, others are catching up. For example, Lithuania (with a population of 2.8 million) has become a European center for cryptocurrency exchanges, reflecting similar trends in Singapore and the United Arab Emirates (BCCS, 2023; Chadha, 2022). This signifies a trend of mid-range financial centers aiming to profile or rebrand themselves as cryptocurrency hubs, which can generate swarming-related problems (Hendrikse et al., 2020). IPE research has advanced in analyzing the role of these technological systems within broader financial governance frameworks (Bernards & Campbell-Verduyn, 2019; Langley & Leyshon, 2021). However, the use of the cryptocurrency ecosystem to exploit regulatory gaps through practices like swarming remains understudied within the field (see also Nesvetailova et al., 2018).

In sum, the practices of organizational ring-fencing and swarming underline the need to broaden the research of global tax and AML governance to the actual capabilities of monitoring secrecy-seeking capital. Both are hence crucial for understanding contemporary obstacles for effective global tax governance. We illustrate these practices empirically below.

Organizational ring-fencing and the money laundering scandals of the 2010s

In the mid-2010s, Estonia became the epicenter of one of the largest money laundering scandals globally (Ewing, 2019; Milne & Winter, 2018). In March 2017, a Danish newspaper published a series of articles about money laundering at Danske Bank’s Estonian branch (Bjerregaard & Kirchmaier, 2019; Milne & Winter, 2018). Danske Bank Press Releases (2017) acknowledged that ‘major deficiencies’ had allowed its Estonian branch to be used ‘for criminal activities such as money laundering’. An internal investigation and several official inquiries followed. Danske’s investigation concluded that the non-resident portfolio of its Estonian branch—with many UK shell companies—contained over €200 billion in suspicious transactions originating in Russia, Azerbaijan, and other former Soviet countries from 2007 to 2015 (Bruun & Hjejle, 2018; Coppola, 2018). Under pressure from the Estonian Financial Supervisory Authority (FSA), Danske closed this portfolio in 2015. The Danish FSA reprimanded Danske Bank in 2018. A year later, the Estonian FSA ordered Danske to end its Estonian business.

The scandal led other Nordic banks to distance themselves from Danske, but they were soon found to have used similar practices (Milne, 2019a, Milne, 2019b). Journalists uncovered how Swedbank channeled approximately €135 billion of ‘high-risk non-resident’ flows through its Estonian subsidiary between 2008 and 2018 (Milne, 2019a). Eventually, public pressure forced Swedbank to commission an internal investigation, but only for the years 2014 to 2019. The report found Swedbank’s AML systems to be inadequate and estimated the scale of transactions in Swedbank’s Baltic subsidiaries with AML risks to be €37 billion, half of which passed through Estonia (Clifford Chance, 2020).

The New York Times was shocked to see Scandinavian banks ‘accused of helping Russian oligarchs, corrupt politicians and organized crime lords send hundreds of billions of ill-gotten dollars to offshore tax havens’ (Ewing, 2019). The Financial Times saw Nordic banks as ‘part of a system that allowed oligarchs and criminals from Russia and elsewhere to move money’ through western banks (Milne, 2019a). The emerging picture was one of criminals exploring ‘the weakest links in the chain surrounding the EU’ (Bjerregaard & Kirchmaier, 2019, p. 4). Finally, in 2022, news broke that Luminor bank’s investigation and its predecessors had discovered suspicious transactions worth over €3.9 billion in 2008–2018 in their investigation (Hänninen, 2022). Luminor was created to merge the Baltic operations of the Finnish-Swedish Nordea and the Norwegian bank DNB in 2017. Consequently, five out of the eight biggest banks in the Nordics had been found culpable to facilitating money laundering in the Baltics.

Based on our data collection and analysis, we devised a taxonomy of the factors that enabled this money laundering to continue without public knowledge for years. Figure 1 builds on our framework (Section New avenues for secrecy-seeking capital) and the empirical material from our case study to illustrate how organizational ring-fencing works. It shows how large banks can separate their IT systems, reporting and management functions and AML practices of their subsidiaries from their group-level equivalents. This separation allows banks to comply with the regulatory standards in their headquarter countries, while facilitating the international movement of secrecy-seeking capital.

Figure 1. How ring-fencing works.

Our taxonomy includes six aspects. First, local Estonian banks had a heritage of non-resident deposit portfolios that predated their acquisitions. Second, the existing AML procedures of the banks were insufficient. Third, in the case of Danske, money laundering went undetected because the bank had separated the IT systems of its Estonian branch from its group-level systems. Fourth, high profits motivated headquarters to ignore warnings. Fifth, cross-border financial supervision was ineffective. Finally, the existing legislative framework was too weak to detect and prevent ring-fencing. We will examine each aspect through Danske and Swedbank.

First, after the collapse of the Soviet Union, banking governance in the Baltic countries was laissez-faire. Banks amassed non-resident portfolios from the former Soviet countries (Ewing, 2019). After joining the EU, the Baltics became an attractive market for Nordic banks (Liuhto et al., 2007). Estonian authorities hoped foreign investments would consolidate and ‘clean up the Baltic financial sector’ (Milne, 2018). By 2010, foreign banks owned 99% of the Estonian banks (Kattel & Raudla, 2013), with the largest market shares held by Swedbank, SEB and Danske, which inherited non-resident portfolios through these acquisitions. For example, in 2007, Danske acquired the Finnish Sampo Bank. Its Estonian subsidiary (Sampo Pank) became Danske’s branch. The predecessor of this branch had attracted considerable Russian clientele in the 1990s (Bruun & Hjejle, 2018). At the time of the acquisition, Sampo Pank had around 1,550 customers who were suspected of illegal activity (Bjerregaard & Kirchmaier, 2019, p. 12). In the early 2010s, non-resident customers accounted for approximately 20% of all Estonian savings account balances (Kessler, 2019b).

Second, Estonian branches and subsidiaries had deficient AML practices (Bruun & Hjejle, 2018; Clifford Chance, 2020; Finanstylnet, 2018). The management of the Estonian branch of Danske seemed ‘to have been negligent to the point of criminality’ (Coppola, 2018). An internal investigation found AML procedures at the Estonian branch of Danske ‘manifestly insufficient and in breach of international standards’ and Estonian law (Bruun & Hjejle, 2018, p. 27). The bank had neglected the supervision of customers, beneficial owners, and sources of funds. It had hesitated to act on suspicious customers and transactions (Bruun & Hjejle, 2018; Finanstylnet, 2018). Transactions were conducted by known criminals who used notorious foreign banks. Many customers shared identical addresses and phone numbers. They regularly established Estonian accounts as conduits for offshore transactions, obfuscating the origins of their wealth (Ewing, 2019). The Estonian subsidiary of Swedbank approved risky customers without knowing the ultimate beneficial owners or sources of funds (Clifford Chance, 2020). Companies had opaque ownership structures with tax haven entities, masking ultimate beneficiaries with trusts or similar vehicles. Swedbank Estonia allowed listing intermediaries as beneficial owners and employees disregarded signs of suspicious transactions (ibid.).

Third, Danske left the IT systems of the Estonian branch unintegrated with the rest of the corporate group. Consequently, the Estonian branch never established transaction and risk monitoring or other group-level AML procedures. The branch operated on an outdated and inadequate Estonian-language monitoring system with little automation (Bjerregaard & Kirchmaier, 2019). Most documents were only available in Estonian or Russian (Bruun & Hjejle, 2018). The headquarters attributed blame to ‘poor communication’ between the executive board and a ‘corrupt’ Estonian management (Logan, 2019). Swedbank operated as a subsidiary in Estonia, with more independent management. However, it failed to match its AML resources with the strategy of pursuing high risk-customers (Clifford Chance, 2020).

Fourth, the headquarters repeatedly ignored warning signs of potential criminal activity. Dysfunctional IT systems notwithstanding, concerns of suspicious activities emerged in Danske between 2007–2015. In 2007, both Estonian and Danish FSAs flagged money laundering risks in Sampo Pank, but the non-resident portfolio continued to balloon (Milne & Winter, 2018). Subsequent years saw critical reports from the Estonian FSA, following a 2013 whistleblower report and internal audits (Bjerregaard & Kirchmaier, 2019; Bruun & Hjejle, 2018; Finanstylnet, 2018). AML concerns led J.P. Morgan to terminate its corresponding banking relationship with Danske’s Estonian branch in 2013 (Finanstylnet, 2018; Milne & Winter, 2018). Profit considerations and a weak AML culture overruled money laundering concerns (Coppola, 2018; Ewing, 2019). The financial crisis and the Eurozone crisis generated losses for Danske, and the Estonian branch became a useful buffer for the company (Coppola, 2018; Logan, 2019). In 2011, the return on equity for Danske’s Estonian branch was extremely high at 47% (Coppola, 2018), and the branch generated approximately 11% of Danske’s profits (Milne & Winter, 2018). In 2013, the return on allocated capital in the non-resident portfolio of the Estonian branch was 402% (Finanstylnet, 2018). Similar dynamics occurred in Swedbank, where executives ignored repeated warnings about money laundering (Clifford Chance, 2020; Finantsinspektsioon, 2020).

Fifth, supervisory authorities in Estonia, Denmark and Sweden failed to tackle ongoing, pervasive money laundering in their financial systems. Despite advances in the international regulatory landscape (see Section The evolution of research on secrecy-seeking capital and its governance), cross-border supervision of multinational banking groups remains fragmented, with coordination problems ranging from overlapping responsibilities to an unclear division of labor, especially regarding AML supervision (Beck et al., 2013; Demetriades & Vassileva, 2020). This enabled blame shifting between countries and regulators (Heinkelmann-Wild & Zangl, 2020). The Danish FSA handled the overall supervision of the Estonian branch of Danske, while the Estonian FSA handled AML compliance. This created legal uncertainty for the Estonian FSA about their authority (Milne, 2018). According to the current director of the Estonian FSA, Kilvar Kessler (2019b), AML work was a secondary goal for them. Supervisors prioritized processes over transactions, which were considered to be the responsibility of the Financial Intelligence Unit (FIU) (Kessler, 2019a, 2019b). AML tasks lacked human and financial resources, and Baltic authorities deprioritized them (Ewing, 2019). Kessler notes that the Estonian FSA mostly focused on supervisory issues like solvency, and few people worked on AML (Kessler, 2019a). The concerns voiced by the Estonian FSA notwithstanding, its 2009 report on Danske’s Estonian branch found ‘no significant breaches of internal procedures or legal requirements’, warranting no ‘immediate regulatory action’ (Bjerregaard & Kirchmaier, 2019, p. 17). This attitude changed only in 2014 after leadership changes and the Russian annexation of Crimea, which heightened interest in AML controls (Kessler, 2019a, 2019b).

In 2019, the Estonian FSA demanded Danske to cease activities in Estonia (Finantsinspektsioon, 2019). Kessler (2019a) argues that the power of Nordic banks had postponed such decisions: ‘These were relatively large, efficient and well-equipped organizations by Estonian standards, with influence in the public space and in legal disputes, and as lobbyists and employers’. Journalists also highlighted the ‘clubby nature’ between the regulators and the banks in Sweden and Denmark (Milne, 2019a). The head of the Swedish FSA was close friends with one of Swedbank’s directors, and the former finance director of Danske became the chair of the Danish FSA (ibid.). Regulators may have also hesitated to ‘police big, politically connected lenders’ (Ewing, 2019).

Finally, the existing legislative framework was deficient (Tammer, 2022). The maximum fine that the Estonian FSA could issue for money laundering was a mere €32,000 (Kessler, 2019a). Cumbersome misdemeanor proceedings favored banks over regulatory concerns (ibid.). Effective AML procedures were extremely difficult under this framework and further exacerbated the problems listed under the first five points. Taken together, these aspects enabled the emergence of organizational ring-fencing practices throughout the Baltic branches of Nordic banks.

The Swedbank and Danske scandals provoked Baltic regulatory authorities to demand a more coordinated AML approach, including the creation of an EU-level AML authority (Milne, 2019b). In 2020, the European Commission introduced such a proposal (Demetriades & Vassileva, 2020). Subsequently, the Estonian FSA has prioritized AML work (Interviews 5; 7; 10; 19), as the case of the Estonian Versobank demonstrates. It was described as ‘a rising star in the money laundering market’, before the European Central Bank withdrew the bank’s license on the advice of the Estonian FSA (Kessler, 2019b). The FSA has also demanded banks align their risk controls with their risk appetite (ibid.; Interviews 17; 23). Consequently, the share of non-residents in the Estonian banking system has plummeted from approximately 20% to below 7% (ibid.).

Swarming: the Estonian cryptocurrency sector and its offshore linkages

As Estonia was reeling from the shockwaves of money laundering scandals, novel concerns emerged in its nascent cryptocurrency sector. Since the early 2000s, Estonia has branded itself internationally as ‘e-Estonia’, spearheaded by the digital ID card system (establised in 2002) and the blockchain-based X-Road (established in 2001), which facilitates information exchange between Estonian authorities (Drechsler, 2018; Tammpuu & Masso, 2019). The e-residency program that Estonia introduced in 2014 further supported the emergence of the cryptocurrency industry (Interview 19). The program grants foreigners an electronic ID number for accessing digital services, including online incorporation and digital signatures (Interviews 6; 7). The e-residency advocates have highlighted its competitive benefits for Estonia, whereas the tax authority and the FIU were worried about risks of money laundering and financial crime (Interview 5).

The Estonian cryptocurrency boom commenced in 2016, when Estonia adopted a licensing system for cryptocurrency providers. In some respects, this system presaged the registration requirements of the EU’s 5^th AML directive. Yet, whereas the directive issued various new obligations for customer identification, the license system was laissez-faire. Cryptocurrency-related trades below €15,000 did not trigger KYC checks. Any trades below this threshold were anonymous. The licensing scheme gained international attention when the head of the e-residency program suggested that Estonia should issue its own cryptocurrency (Korjus, 2017). The Bank of Estonia rebuked the idea, but it bolstered the global image of Estonia as a crypto-friendly nation (Interview 7). The licensing scheme was rubber-stamping: Almost anyone could get a license, and an official recognition from an EU country was a marketing asset. The licenses followed the requirements for local pawn shops and currency exchanges, which were low-risk, unlike most cryptocurrency providers (Interview 19).

Skyrocketing applications overwhelmed supervisory capabilities. At the peak, some 1,300 licensed cryptocurrency service providers resided in Estonia (RAB, 2020). While the number of active firms has since dropped to around 180 (see below), more than half of the global cryptocurrency service providers remained registered in Estonia in mid-2021 (RAB, 2022). The combined turnover of companies with cryptocurrency licenses was €590 million in 2018, climbing to €1.2 billion in 2019. This income came from 500,000 unique customers in 2019 (RAB, 2020). Between the summers of 2020 and 2021, the combined turnover of these firms climbed to €20.3 billion (RAB, 2022). We can contextualize this turnover with a comparison to the nationally important tourism sector, which stood at €2 billion in 2018 (OECD, 2020). Authorities saw money laundering risks growing under the ‘layer of respectability’ of the licensing scheme (Interviews 5; 19). Two-thirds of Estonia-based crypto firms resided in only four Estonian addresses, highlighting the centrality of service providers for this boom (RAB, 2022).

Following tighter regulations, most firms left as quickly as they arrived. Given the ‘virtual’ nature of these firms and their practices, their relation to traditional tax havens is a key question to determine the linkages between swarming and secrecy-seeking capital. We address this question by analyzing the board interlocks of Estonian cryptocurrency license holders in firms that had been registered in the database by March 2020, when the tougher regulations were established. Figure 2 highlights the countries where board members of Estonia-based cryptocurrency firms also have corporate board memberships. We arranged the countries in the figure with Gephi’s Geo Layout for a map-like projection. We aggregated the firms which share board members at the national level and kept the self-loops of the nodes in the network—i.e. circumstances when a board member sits on two firms from the same country. To carve out the main pattern of the network, we excluded repeated connections between countries—i.e. if a board member sits on several boards of Estonian and British or any other firm, we treated this as a singular connection. Hence, the analysis does not consider the weight of each connection.

Figure 2. Visualization of board member interlocks of crypto firms registered in Estonia, 2016–2020.

As Table 1 below demonstrates, the top 10 of these countries includes five jurisdictions that feature in the top-20 of the 2020 Financial Secrecy Index of the Tax Justice Network (2020). These countries are the UK (#1), Malta (#4), Singapore (#8), Switzerland (#7), and the USA (#6). The top ten jurisdictions of our board interlock analysis also include Cyprus (#5). While Cyprus has improved its transparency and cooperation in tax matters, it boasts a decades-long long history of being a major tax haven (Haberly & Wójcik, 2015).

Table 1. The top 10 jurisdictions of the board interlock analysis.

Country	Unweighted degree in the interlock network
United Kingdom	48
Spain	42
Lithuania	41
Malta	39
Cyprus	37
USA	36
Switzerland	35
Singapore	35
Latvia	34
Italy	33

We also compared this list to the snapshot of interlocks formed between cryptocurrency firms in Estonia and other jurisdictions in late 2022 (Figure 3). The number of firms fell by almost 90% to 180 and the network shrank (the average degree fell from 15.474 to 10). However, the target jurisdictions remained relatively stable: Within the top 10, the USA was replaced by Portugal. Hence, despite the significantly lower number of cryptocurrency firms registered in Estonia today, linkages with secrecy jurisdictions seem to persist. We also compared our results to recent research by the Estonian FIU. In addition to highlighting the role of the tax havens Belize and Luxembourg as significant client bases for Estonian-based crypto companies, the FIU states that ‘a significant share of Estonian-based cryptocurrency firms have linkages with Ukraine or Russia’ (RAB, 2022, p. 11). The absence of these two countries from the results of our board interlock analyses may point to the use of intermediaries or ownership/management structures that are chained through multiple countries or countries with opaque company registers, as the FIU can access information beyond the scope of academic researchers.

Figure 3. Visualization of board member interlocks of crypto firms registered in Estonia, October 2022.

The prevalence of secrecy jurisdictions adds weight to the challenges highlighted by our interviewees and in Estonian assessments. Tax authorities and the understaffed Estonian FIU found monitoring cryptocurrency companies with little or no economic activity in Estonia strenuous (Interview 5). Authorities were better equipped than before to monitor individual cryptocurrency firms, given their small size and the advances in international information exchange and investigative methods. However, monitoring hundreds of cryptocurrency companies swarming to Estonia with little real economic presence created major difficulties. As summarized in Figure 4, we highlight four kinds of supervisory challenges that swarming creates for the authorities.

Figure 4. How swarming works.

First, the e-residency system and the network of business and company service providers facilitated easy online incorporation. Authorities were strained to monitor hundreds of new foreign-operated companies (Interview 5). This problem was in many ways by design: At the beginning of the program in 2014, Estonian authorities aimed to reach 10 million e-residencies until 2025—almost eight times the population of Estonia (Sullivan & Burger, 2017). Reducing bureaucratic hurdles of economic participation was thereby a core goal of the e-residency program (ibid., p. 471), which laid the basis for later regulatory overburdening through practices like swarming.

Second, whether national authorities can gain information on firms depends on the extent of their business activities in the country. When firms do not declare employee or value-added taxes, authorities know little about them. In the Estonian case, even financial reports are often deficient or submitted late, with no penalties for such behavior. Insufficient administrative capacity required prioritizing and targeting supervision. This became problematic with numerous small firms that posed a systemic risk to financial integrity (Interview 8). The cryptocurrency boom spiraled supervisory efforts ‘out of control’ because authorities did not know ‘what this company is doing and how much its activities’ related to Estonia (Interview 10).

Third, performing effective background checks of individuals in foreign countries is time-consuming and difficult (Interviews 5; 8). Entrepreneurs may have many namesakes in their home countries (Interview 19), and the national commitment to international information exchange varies. As a further worrying sign, in 2021, almost three-fourths of cryptocurrency service providers submitted zero reports on suspicious transactions or customers to the Estonian FIU (RAB, 2022), even though close to half of bitcoin transactions are suspected to involve illegal money (Foley et al., 2019). As one interviewee notes, ‘the lawmakers are always behind the real life’, as are the administrative and IT systems for exchanging information within and between countries (Interview 22).

Fourth, while the 2010s saw substantial advances in the exchange of information on financial investments and beneficial owners of companies, national authorities sometimes cannot share this information between them. The police and other authorities would also benefit from sharing information on bans of business operations and convictions related to economic crimes, which are often excluded from international criminal record databases (Interviews 5; 20). Together, these four aspects hindered the ability to monitor swarming in Estonia, despite new tools for the automatic exchange of tax information and AML control. The e-residency program and the ease with which entrepreneurs can register companies in Estonia created administrative bottlenecks that foreign cryptocurrency firms were able to exploit.

In March 2020, the government expanded KYC requirements to all cryptocurrency trades. To renew a license, service providers had to provide the FIU with information on, e.g. bank accounts, board members, and employees (Alev, 2020). The rising costs of licenses pushed many cryptocurrency firms to relocate (Ministry of Finance, 2021), as our board interlock comparison also suggested. Yet, a 2021 government evaluation reported that the industry was again growing, and regulatory challenges persisted (Ministry of Finance, 2021). As a regulatory reaction, Estonia banned crypto companies from having anonymous accounts in June 2022 and issued a capital requirement of €350,000—a tenfold increase (Euronews, 2022). The effect of these rules remains to be seen.

Discussion and conclusion

Advancements in global tax governance and AML work notwithstanding, the demand-side of secrecy-seeking capital has persisted. Secrecy-seeking capital has found new avenues, developing practices such as those described in this paper. High profits incentivize financial intermediaries to seek new channels for bypassing the regulatory advances of the 2010s. The looming regulatory battles point to the need to progress beyond the first wave of global tax governance scholarship. In this paper, we have argued that organizational ring-fencing and swarming have emerged as two new key practices that have enabled secrecy-seeking capital to thrive despite the stronger international regulatory frameworks. Our rich empirical illustration of the Estonian case traced such practices for the first time in the real world and excavated their intimate connections to offshore finance after the recent crackdowns on tax havens.

These shifts have major implications for IPE scholarship in global tax governance, AML, FinTech, and beyond. While drawing a clear line between tax havens and non-tax-haven countries has always been difficult, various tax haven indices have enabled such a distinction. This distinction has also influenced the policy agenda of the first wave of global tax governance. Indices have typically relied on analyses of legal characteristics of countries and facilitated quantitative studies. In contrast, our paper illuminates two practices where secrecy-seeking capital has exploited differences in administrative capabilities to monitor the financial sector and shortcomings in cross-border coordination of financial supervision. Mapping such difficulties requires qualitative studies of countries and/or financial actors (Baker & Murphy, 2019; Ylönen et al., 2023). We have shown how secrecy-seeking capital has used organizational ring-fencing and swarming to exploit these difficulties. Both phenomena underline the national-bureaucratic foundations of global tax, financial and AML governance. Organizational ring-fencing and swarming also highlight the importance of advancing the coordination of cross-border financial supervision via regulatory standards and supervisory cooperation.

Organizational ring-fencing is driven by massive profits generated by granting high-risk capital flows entry points to the global financial system. The trickier the use of traditional tax havens, the higher the likely profit margins. While regulatory advances may make similar events in EU countries unlikely, other non-tax haven jurisdictions may offer banks similar entry points. The combination of such avenues, the profit motive, and the large number of banks globally increase the likelihood that some bank somewhere will replicate organizational ring-fencing—or is doing so already. The AML enforcement is highly heterogeneous in the EU, let alone globally (Kirschenbaum & Véron, 2018). In some respects, the global financial governance system is as strong as its weakest links.

Swarming—the second practice we describe—was facilitated by a deeply inadequate supervision practice of cryptocurrency firms in Estonia and beyond. When the crypto firms saw that their transactions were not really monitored, they were further incentivized to swarm to Estonia. As Section Swarming: the Estonian cryptocurrency sector and its offshore linkages illustrated, the influx of cryptocurrency firms in Estonia relied on several regulatory, reputational, and business-related factors. Our board interlock analysis and the FIU analyses demonstrated that a large share of the board members of Estonian crypto firms also sit on the board of firms in well-known offshore financial centers (Section Swarming: the Estonian cryptocurrency sector and its offshore linkages), pointing to linkages between this novel technological field and traditional tax havens.

What do our results mean for broader debates in current IPE scholarship? We can seek answers to this question from long-standing debates on corporate power and global tax governance. Following the global financial crisis, the question of corporate power and its relation to state authority has resurfaced in IPE (Babic et al., 2022). However, corporations exercise power in more complex ways than before, exploiting and amplifying the decentered nature of firms (Reurink & Garcia-Bernardo, 2021). Both organizational ring-fencing and swarming allow corporate and private capital to circumvent regulatory initiatives that should constrain corporate power and impunity. Both emerged as direct answers to attempts of reinserting state authority in a globalized political economy through regulatory advances against tax havens (Ahrens & Bothner, 2020).

We demonstrated that the exercise of corporate power is not restricted to big multinational corporations and their lobbying and instrumental power efforts to extract concessions from their home or host states. Such a view corresponds to the classical IPE emphasis on corporations as challengers to nation-states when it comes to exercising instrumental and structural power (Gilpin, 1976; Vernon, 1974). Swarming in particular illustrates that the determinants of corporate power in the global political economy go beyond size. It is a higher-order phenomenon that emerges from the agency of numerous smaller and medium firms aiming to exploit administrative incapacities and supervisory loopholes. To paraphrase Peter Katzenstein (1985), IPE should also tackle ‘small firms in global markets’, which will continue to influence global (regulatory) politics in an era of increased geoeconomic competition and weaponized interdependence.

Next to this core theme, our case study also speaks to the research agenda of mapping the corporate strategies in Fintech (Bernards & Campbell-Verduyn, 2019), as major centers for international capital from Luxembourg to Singapore strive to become hubs for cryptocurrency-related business (Haberly et al., 2019; Hendrikse et al., 2020; Lai & Samers, 2021). Organizational ring-fencing and swarming can help to analyze such dynamics in an era characterized by geoeconomic sanction regimes (Olsen, 2022). In addition to cryptocurrencies, virtual assets and infrastructures also transform the global economy. This happens, for example, by centralizing market power ‘through democratization’, where seemingly democratic forms of virtual ‘sharing’ evolve into winner-take-all situations (Haberly et al., 2019). The newly emerging virtual asset economy might be even more unequal and unjust than before, and IPE needs to better understand the varying corporate and political strategies of the actors that exploit it. A mere focus on regulatory issues would fall short of this, as the phenomena described in this paper involve practices of financial actors that are designed to circumvent these very regulatory efforts.

Our findings hence provide a steppingstone to expand the repertoire and knowledge of corporate financial practices and strategies in an evolving global political economy. Taking such corporate practices seriously implies moving beyond the first generation of modern global tax governance scholarship. This shift is already happening, as we have documented with references to existing studies. It occurs in a situation where the challenges of tackling the (non-)taxation of multinationals are more global than ever, with issues such as global minimum taxes and digital taxation dominating the agenda (Gelepithis & Hearson, 2022). Yet, simultaneously, secrecy-seeking capital turns increasingly toward ‘onshore’ practices discussed in this paper, and even efforts to tackle corporate profit shifting are hampered by the lack of administrative capacity—especially in the Global South (Oguttu, 2016). By following the money in small and large firms alike, we can identify the range of governance challenges that delineate global tax governance in countries that may or may not align with traditional definitions of ‘offshore’. Understanding the full impact of this shift will also require further expanding the research agenda on tax intermediaries (Boussebaa & Faulconbridge, 2019; Christensen et al., 2022; Raitasuo & Ylönen, 2022) in onshore states. Such lines of inquiry are particularly important in a situation where the evolution of FinTech—and its nimble service providers—may upend global financial patterns and associated regulatory challenges at a faster pace than before.

Finally, our findings are also pertinent in a world where rising complexity and fragmentation upend ‘classical’ institutions of global economic governance in favour of (networked) regimes with diffuse hierarchies and rules (Alter, 2022; Eskelinen & Ylönen, 2017; Kim, 2020). Such developments have already ‘gridlocked’ several global economic governance mechanisms and fuelled the search for a post-Bretton Woods mode of global economic governance (Fioretos & Heldt, 2019; Hale et al., 2013; Rauh & Zürn, 2020). As our analysis shows, overcoming such gridlocks will not only require a stronger focus on new global rules, but, importantly, also on national administrative capacities in enforcing these rules—in and beyond the fields of tax and AML governance. With this, we hope to contribute to the ongoing debates on how to salvage and transform global economic governance in times when it’s (again) most needed (see Hale et al., 2013).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data on board interlocks are available from Bureau van Dijk’s ORBIS database (https://www.bvdinfo.com/). We used the data under license and restrictions apply to its availability. Data are available from the authors with the permission of Bureau van Dijk. The nationally aggregated and processed data, more supplementary material, and the Python script are freely available at the Open Science Framework (https://osf.io/5ju2k/).

Additional information

Funding

This project has received funding from the European Union’s Horizon 2020 research and innovation programme (grant 867454) and from the Estonian Research Council (grant PRG1125).

Notes on contributors

Matti Ylönen

Matti Ylönen is a University Lecturer in World Politics at the University of Helsinki, with a focus on International Political Economy. He has published extensively on the politics of multinational enterprises, tax havens, consultancy firms, and global development, and is currently working on a research project on the power of big tech companies in the European Union.

Ringa Raudla

Ringa Raudla is Professor of Fiscal Governance, Ragnar Nurkse Department of Innovation and Governance, Tallinn University of Technology, Estonia. Her main research interests are fiscal policy, public budgeting, public administration reform, financial regulation, and institutional economics. Her recent publications include articles in Public Administration Review, Governance, Policy Studies Journal, Journal of Public Policy, and Journal of Comparative Policy Analysis.

Milan Babic

Milan Babic is an Assistant Professor in Global Political Economy at the Department of Social Sciences and Business at Roskilde University and PI of the DECARB project. His work deals with the transformations of the global political economy in the transition from a neoliberal towards a post-neoliberal global order. His latest book is The Rise of State Capital (Agenda 2023).

Notes

1 Tax havens are sometimes called offshore financial centers or secrecy jurisdictions. The term corporate haven is also sometimes used for centers that rely more on exploiting loopholes in international corporate taxation.

2 While secrecy-seeking capital is typically associated with tax evasion, money laundering or other illegal activities, avoiding expropriation or sanctions can be legal. Furthermore, multinationals may have incentives to hide assets or transactions that have not been deemed illegal. Such behavior can occur if the legality transactions is uncertain, or if companies want to hide asset-specific financial or reputational risks from stakeholders.

3 We thank an anonymous reviewer for suggesting the division between two generations of global tax governance research.