Introduction

Since the beginning of the 21st century, we witness a Zeitgeist of deepening awareness of social and environmental challenges, the role of (big) corporations herein and subsequently a growing demand for change. Numerous cases of corporate deviance made (inter)national news headlines in the past years: the Volkswagen Dieselgate, money laundering cases at global banks, the Wirecard accountancy fraud, several cases against Shell for corruption, human rights violations and climate pollution, tax avoidance of Amazon or Starbucks or the prosecution of Purdue Pharma and the Sackler family. Attention to and disapproval of corporate deviance is partly fuelled by increasing transparency of information on corporations and their behaviour all over the globe. Cases are shared, discussed, or disapproved by people on (social) media (e.g. #StopHateForProfit boycott on Facebook), on the streets (e.g. Yellow Vests in France or actions by Extinction Rebellion targeting the (financiers of the) fossil industry), in courtrooms (e.g. lawsuits against German car manufacturers and Shell for climate impact or lawsuit against H&M for greenwashing) and in parliamentary debates.

Much literature has explored media framings of corporate scandals (Cavender & Miller, 2013; Jong & Van der Linde, 2022; Levi, 2006; Van Erp, 2013), civil society actions against corporations (Calhoun, 2013; Gunningham, 2020; Van Erp, 2021), corporation’s own responses to scandals (Benson, 1985; Gottschalk & Benson, 2020; Van Rooij & Fine, 2018), or the regulation, criminalisation and sanctioning of corporate misconduct (Braithwaite, 2020; Croall, 2003; Huisman, 2016; Levi, 2002, 2018; Parker, 2012; Schell-Busey et al., 2016; Shichor & Heeren, 2021).

There is, however, little research on responses to corporate deviance by their peers (corporations in the same industry). Just like people or groups, corporations respond to and interact with each other rather than operating in isolation. In his work on reintegrative shaming, Braithwaite (1989) translates insights from research on the role of peers in juvenile crime to the corporate setting. He emphasises that peers play a crucial role in influencing the behaviour and perceptions of deviant corporations (Braithwaite, 1989). Lang and Stulz (1992, p. 46) in their study on bankruptcies speak of a contagion effect of crisis that starts at one corporation and spreads to its competitors in the industry. Verhage (2011, p.82–83) shows that peer pressure between corporations plays a role in defining compliance benchmarks within financial institutions. Research on trust and codes of secrecy between corporations (Jaspers, 2020; Van de Bunt, 2010) suggest that condemnation by peers is unlikely to occur. Following that line, Macey (2013), in his analysis of the substitution of reputation by regulation, points out that in their own worlds (peer environment), businesspeople have a distinct reputation and can often uphold their respectability, even after big media scandals.

Existing research offers ambiguous insights on possible processes of peer responses. Also, the role of peers in responses to corporate deviance has not been their primary focus. This article fills this gap and puts peer corporations, an often-overlooked actor, to the centre stage. It aims to answer the question how corporations respond to corporate deviance of their peers.

This question is particularly relevant given the Zeitgeist of public scrutiny, in which corporations “can no longer get away with just deflecting blame” (Van Rooij & Fine, 2018, p. 3). The external ‘labelling’ of their behaviour pressures (other) corporations to respond. This is particularly visible in the banking industry that faces ongoing scrutiny since the financial crisis of 2008. Anti-money laundering (AML) violations, especially since 2018 with big cases at Danske Bank (Denmark & Estonia) and ING bank (The Netherlands) form yet another scandal attracting regulatory, political and public scrutiny. This paper focuses on AML cases in the Netherlands where the fight against money laundering has become a top priority for authorities since the ING case, prompting all banks to examine their AML frameworks. As all banks share the same responsibilities under AML regulations, a contagion effect in which violations at one corporation turn into an industry-wide problem seems likely (see Jong & Van der Linde, 2022, p. 5; Lang & Stulz, 1992, p. 46). This raises questions on how other banks respond to AML violations in the Dutch banking industry.

This article first outlines symbolic interactionism as the theoretical foundation and place it in the context of corporate deviance and interactions between peers. Subsequently, the case study of non-compliance with AML regulations in the Dutch banking industry and the qualitative methodology are outlined. In the result section, six stages of peer interactions are presented. Finally, the implications of these stages for theory and practice are discussed.

Peer responses and symbolic interactionism

Peer responses can take a variety of forms and be approached from different angles. This article takes a symbolic interactionist perspective. It pays particular attention to the perspectives, frames and interpretations in the process of social interaction (Blumer, 1969) between corporations. The article deliberately speaks of corporate deviance, instead of the more commonly used corporate crime, to underline the interactionist underpinning of deviance being socially constructed in interaction (Becker, 1963). It thereby also goes beyond violations criminalised by the law, encompassing violations of social norms and rules.

Before introducing different peer responses according to interactionist theories, a categorisation based on Goffman’s (1973) dramaturgical perspective is offered. Goffman introduces a distinction between back stage and front stage behaviour to understand social interactions. As corporations face (growing) external attention and disapproval of their behaviour, this distinction can help to interpret different peer interactions. Front stage behaviour is visible to others and guided by other’s expectations and norms. Peer interactions on the front stage are shared – either publicly or privately within the industry. They entail the dynamic interplay of action and reaction between banks, thereby focusing on responses towards each other. Back stage behaviour, in contrast, is invisible to others and closer to one’s true self (Goffman, 1973). For this research, back stage behaviour entails responses within one bank that are not shared with other banks or the public.

The labelling approach by Becker (1963) formed the departure point of this research. The persistent external labelling of AML cases by authorities and the public raises the question whether and how this affects and translates to internal labelling between banks. Labelling describes the process of creating deviance in social interaction. Outsiders are those who have been successfully labelled as such by groups in positions to create and impose their rules on others (Becker, 1963, p.9). Successful labelling results in stigma and othering. Huisman (2011) and Levi (2009) discuss the difficulties of labelling white-collar criminals who do not qualify as typical folk devils (Levi, 2009) and have the means to counteract (Huisman, 2011) or prevent efforts of labelling by authorities or the public (Levi, 2009). The same applies for corporations. Therefore, a horizontal alternative of peer labelling (Merz, 2019) is taken in which such power imbalances do not apply. The work by Becker (1963, p. 85) on labelling includes a demonstration of the process for dance musicians who labelled their audiences as squares representing “a person who is opposite of all the musician is”. This label is shared among musicians and leads to a perception of being better than other people. Likewise, corporations might perceive themselves as better than other corporations in the industry, thus engaging in the labelling of their peers. Such labels can be shared with the industry (front stage) or remain within the boundaries of the corporation (back stage).

During the research process, labelling appeared less prevalent among peers. Rather, more diverse peer responses and interactions were found. Therefore, the theoretical foundation was broadened to other symbolic interactionist perspectives. Some of them, such as shaming, are inspired by labelling. Shaming is a social sanction based on public disapproval of behaviour. Stigmatisation is a possible but not a necessary element of shaming (Van Erp, 2008). Contrary to labelling, shaming is widely used in the context of corporate deviance. Authorities use (naming and) shaming as a punishment for corporations (Van Erp, 2008). The public and civil society engages in shaming on social media and in campaigns to raise awareness of corporate deviance. Shaming by peers is possible as an informal punishment by industry (members), leading for instance to the exclusion from (conferring presidency of) partnerships. Just like labelling (and concepts such as stereotyping, scapegoating or defamation), shaming is a condemning response.

Such condemning peer responses fit the defensive orientation of face work (Goffman, 1967). Face work illustrates counteractions to incidents to save face (the image of the self). Besides the defensive orientation that is geared towards protecting oneself, counteractions can also take a protective orientation geared towards the protection of others (peer corporations or the whole industry). The protective orientation introduces a different type of peer responses such as closing ranks or cooperation between corporations. In tactic cooperation, industry members work together to enable all parties to maintain face and keep their good name (Goffman, 1967). Research by Jordanoska and Lord (2020, p. 22) on benchmark manipulations in the financial industry show efforts of closing ranks by banks acting collectively. Put differently: defense mechanisms expanding from the corporation to the industry.

Equivalent to Becker’s (1963) approach of labelling where five characteristics (actor, act, victim, situation and audience) and the interpretation of these determine the creation of deviance, the type of peer responses can be determined by several characteristics. Benson (1985, p. 599) showed that successful labelling requires the practice to be regarded as ‘out of ordinary’ (characteristic of the act). Research by Braithwaite (1989) and Levi (2002) indicate that for peer labelling to be successful, corporations need to identify with and care about these other corporations (characteristic of actor). Besides, the level of information on the violations (characteristic of the situation) and the response by the public and authorities (characteristic of the audience) can impact peer responses. For the present research, a protective orientation seems likely in cases of contagion from one AML violation to an industry-wide issue. The high external pressure on banks and their role as competitors, on the other hand, suggest the existence of condemning responses between banks.

The case study: non-compliance with AML regulations in the Dutch banking industry

Peer responses are researched using cases of non-compliance with anti-money laundering regulations in the Dutch banking industry. Under anti-money laundering and counter terrorist financing (AML/CTF) regulations banks are responsibilised to act as gatekeepers obliging them to identify, prevent and signal suspicious financial flows and clients.

The banking industry forms an excellent case study for peer responses given the ongoing pressure and scrutiny on financial institutions since the financial crisis (Jordanoska & Lord, 2020; Shichor & Heeren, 2021). First, this became visible in the movement of Occupy Wall Street as part of a broader mobilisation across the globe scrutinising the ‘1%’ financial elite (see Calhoun, 2013). Much of the campaign focused on the power of language in challenging banks, using “the imagery of the ‘bankster’ and the criminal elite for its political slogans” (Tombs & Whyte, 2020, p. 22). Later, with increasing attention for environmental challenges, the public targeting of the fossil fuel industry spread to the role of financial institutions with their major investments in these industries. And more recently with the war in Ukraine, banks were cross-examined for their allegedly slow implementation of sanctions against Russia.

A series of scandals on violations of AML/CTF across the globe (e.g., Danske Bank, Deutsche Bank, ING, Swedbank, UBS, U.S. Bancorp, Westpac) since 2018 mark another wave of disapproval of banks by authorities and the public. This paper focuses on AML non-compliance in the Netherlands. In 2018, ING bank, the largest bank of the country, settled an investigation into violations of AML guidelines for 775 million euros with the Netherlands Public Prosecution Service (NPPS). In 2021, another systemic bank of the Netherlands, ABN AMRO, settled a similar case for 480 million euros. The settlements included the publication of extensive statements of facts (NPPS, 2018, 2021), disclosing the identified shortcomings, its causes, the criminal allegations and conclusions reached by the NPPS. In December 2022, it was announced that the third systemic bank of the Netherlands, Rabobank, is under criminal investigation for suspected AML violations (Reuters, 2022). Other, smaller banks have faced administrative sanctions by the Dutch financial regulator De Nederlandsche Bank (hereafter DNB).

DNB continuously stresses that countering money laundering remains a top priority (DNB, 2020). While the regulator has acknowledged banks efforts on AML, these efforts remain ‘insufficient’, with ‘persistent’ and ‘extensive’ problems (Betlem, 2021). AML cases, especially the two settlement cases, have triggered public and political outrage in the Netherlands, leading to parliamentary debates and continuous media coverage. This tight climate of external labelling of banks on AML non-compliance increases the pressure on them to respond. This article examines how banks respond to this pressure and violations at peers in the banking industry.

Methods and operationalisation

Peers in the corporate context can consist of a broad group of corporations across different industries or networks. Yet, as little is written on peer responses in the corporate context, this research uses a narrow peer concept: peers were operationalised as corporations in the same industry. More specifically, it focuses on retail banks in the Netherlands, leaving investment or private banks aside. Retail banks, banks with financial services for individual (and corporate) consumers, were primarily targeted in the public outrage and political debates on AML cases. Likewise, initiatives of cooperation against money-laundering by private and public parties specifically focus on retail banks. The group of retail banks in the Netherlands includes three systemic banks and several medium or small sized banks (hereafter smaller banks) subdivided in online, sustainable or niche (focusing on a specific region or group of people) banks.

A qualitative research design was chosen due to the nature of the research focusing on constructionist elements of responses to corporate deviance (Wincup, 2017). Interviews often form the most accessible method in qualitative research on corporate deviance (Ho, 2009: 30) and have been widely used (in triangulation) in research on the banking industry (Beizsley, 2019; Jordanoska & Lord, 2020) and anti-money laundering (Eren, 2021; Favarel-Garrigues et al., 2011; Tsingou, 2018; Verhage, 2011). The article is based on 37 semi-structured interviews with three (former) AML analysts who acted as gatekeepers and 34 employees working in the Dutch banking industry (see Appendix for an overview of all respondents). Fieldwork was conductedbetween November 2019 and May 2022. Respondents were selected based on an affiliation with financial economic crime in their work while also representing different linesFootnote 1, departments and management levels of each bank. Most respondents were recruited via personal networks, starting with AML analysts and compliance officers with a criminological background and consequently snowballing through organisations. A few respondents were recruited via LinkedIn or an online search on people working in the industry. After listening to an interview podcast on compliance with the Head of Compliance of a smaller bank, this person was contacted, ultimately leading to four interviews with colleagues. At another bank, a contact from earlier research in the banking industry facilitated to pitch the research and recruit respondents during an online meeting on AML.

Respondents worked at six different banks, with 20 working at the three systemic banks and 17 workings at a smaller bank, of which nine worked at an online bank. Twenty-two respondents worked in the first line which is directly responsible for risks (risk owner). This foremost entails people working in the Know Your Customer (KYC) domain (including Customer Due Diligence, transaction monitoring and operational risk). Twelve respondents worked in the second line (risk monitoring) that entails compliance, conduct and integrity and risk management and three respondents in the third line, that is internal audit (independent assurance). Interviews with respondents working in the second line of defence included the richest and most elaborate data.

Most interviews (23 out of 37) were conducted online via videoconferencing, given the circumstances and restrictions under the COVID-19 pandemic. The other 14 interviews were conducted face to face, with four interviews conducted at the respondents workplace. In 2013, Deakin and Wakefield (2014: 606) stated that online interviewing is still in its infancy in research. The present research is an example of a change during and after the pandemic, making online interviewing more prominent in research. In line with Denscombe (2003) and Deakin and Wakefield (2014: 610), I encountered no or little difference in quality of responses gained through online and traditional (offline) research methods. One disadvantage of online interviewing was that most interviews ended without any room for valuable off-the-record conversations after the official interview as respondents often planned them between other meetings. An advantage was the disappearance of hierarchy or status inequalities between the interviewer and respondent that can play a role in researching up. The online setting and working from home equalised possible differences and allowed respondents to openly share their views on AML and peers. With five respondents, follow-up interviews were held to answer remaining questions or assess newly emerging developments and incidents. Another five respondents responded to follow-up questions in e-mails. The interviews took between 45 and 125 min, with an average of 83 min. All respondents provided written or verbal informed consent. Most interviews were recorded, except for four informal interviews with three gatekeepers and one compliance employee.

The interview data is complemented with content/discourse analysis of publicly available documents between September 2018 (time of the first AML case in the Netherlands) and summer 2022. The analysis focused on what was said (the content) and how it was said (the discourse), thus banks (employees) language and ways of presenting anti-money laundering (violations). Examples of analysed material are banks annual reports and job advertisements, transcripts of analyst calls, statement of facts in the settlement cases and media coverage.

The data was analysed using Atlas.ti (inductive and deductive coding). First, the data was analysed theoretically drawing on the literature and particularly the labelling approach. This first round led to several categories: the labelling of specific cases, general labelling of other banks, factors that hampered labelling, factors that enabled labelling, cooperation between banks, distancing from other banks and blaming others. In a second step, an open round of coding followed, elaborating on peer responses between banks. This second round resulted in six (inductive) stages to structure peer interactions according to respondents’ accounts of it.

Stages of peer interactions

In this article, six stages of peer interactions in the aftermath of AML cases in the Dutch banking industry were identified: (1) panic and fear, (2) comparing practices, (3) distancing, (4) investments, (5) cooperation and (6) defiance. This article refers to ‘stages’ due to parallels with Goffman’s (1973) depiction of social interaction as front stage and back stage performances. The stages are specifications of the theoretical notions of condemning responses (shaming and labelling) and face work.

The interaction stages are partly consecutive as most stages trigger others. For instance, stage 2 (comparing practices) leads to stage 3 (distancing) and stage 4 (investments). Transitions between stages can also be circular: (employees at) banks can go back and forth from one interaction stage to another (e.g. between distancing and cooperation). Stages also overlap, not having strict boundaries and some occur (almost) simultaneously like cooperation and defiance. And lastly, the six stages are indefinite as they mirror the most common reactions and interactions among respondents without presenting a complete picture or each stage applicable to each case of money laundering violations or bank.

Stage 1: From initial panic to lasting fear

As a first reaction, AML cases at other banks triggered feelings of panic, astonishment and shock (stage 1) among peers in the industry. This first stage is not a form of interaction between peers, but rather a reaction to sanctioned violations. Panic was evident among colleagues, being shared in board rooms, (team) meetings and informal conversations.

Responses of panic were in particular noticeable after the ING settlement, describing the case as a “game changer” (R32, board member, small bank), “turning point” (R26, security affairs, systemic bank), “wake up call” (R25, security affairs, systemic bank) or “big bang” (R34, compliance officer, small bank). Respondents’ narratives were coloured with expressions of waves of panic, shock and fear about the implications and consequences of this case for their bank, thus fearing a contagion effect for the industry. The ING case marks the first criminal case against a bank for shortcomings in their AML regulations, with an unprecedented settlement for 775 million euros and high level of transparency:

“All banks were turned upside down with Houston [name of criminal investigation] at ING bank. Due to the amount [of the settlement], but also as this is the first time that the Public Prosecution Service proceeds like this” (R19, compliance officer, small bank).

“The ING settlement had an enormous impact. On this bank as well. It was a bit of ‘haha ING, not us’ but also ‘gosh, that according to the interpretation of the Public Prosecution Service this is what is wrong’. And what does that mean for us and how we tackle things [on AML]?” (R26, security affairs, systemic bank).

Similar observations have been found in research by Rakké and Huisman (2020) on motivations of financial institutions to report suspicious transactions. The settlement increased the extrinsic motivation for reporting among respondents in the banking, accountancy and trust sector by raising the costs of enforcement (Rakké & Huisman, 2020, p. 18). The present research shows that other cases, such as the ABN AMRO settlement (criminal case) and administrative cases in the industry had less impact. They only led to small frights among peers.

The narratives of respondents also showed an ambivalence in feelings on (criminal) sanctions at peer banks: panic and shock were often mingled with compassion and solidarity (R5, R17, R25, R28, R29), frustration about bad publicity for the industry (R7, R8, R36), indignation (R5, R7, R8, R11, R28) but also laughter (R7, R10, R13, R26) and relief that it was not them who got sanctioned (R6, R10, R26, R32). This variety of feelings form peer interactions and come back throughout the following stages.

Besides this first response of panic and shock, a lasting fear emerged among (employees at) banks. While panic and shock are first reactions to a sanction and segue into following stages, fear stayed throughout (and influences) other stages as external pressure, public attention and strict enforcement by authorities is upheld. Fear works on two levels: respondents expressed a fear to be sanctioned (see also stage 2 and stage 4) and a fear to speak up and criticise the regulators current approach and debate on the responsibility of banks in the fight against money laundering:

“I have to say that in a conversation like this I also have some reservations. How much can you fight? How clearly do you dare to say that you think it [the current AML approach] should be different? You might get punished for doing that. That must change. Now, there is this fear. […] People don’t dare to say certain things to the regulator and that’s not okay. I also don’t dare to, even now. I will quit soon but there is a good chance that I will continue elsewhere in the financial industry […].” (R36, board member, small bank).

This fear shapes peer interaction and can ultimately lead to defiance as outlined in stage 6.

Stage 2: Comparing practices

Triggered by and immediately following upon the first stage of panic, meticulous comparisons of AML practices were carried out by other banks. Comparing practices often took the form of in-depth investigations and reconstructions. Compliance and AML experts were asked by higher management to study and analyse available information on cases (such as the statement of facts in the settlement cases) and compare the framework and shortcomings with their own procedures and systems. Going beyond requests, an AML expert (R21) elaborated that they also proactively follow (international) cases and share insights with colleagues and management. The respondent gave the example of an AML case at Swedbank, one of the largest banks in Sweden, which was fined 380 million euros in 2020 by the Swedish regulator. The case showed parallels with and therefore relevance for the procedures on the relation and responsibilities of the Head Office towards branches at the bank the respondent worked for. This links to research by Thornton et al. (2005) and Van Wingerde (2012) on deterrence of corporate crime. They show that general deterrence often works in indirect ways, serving either as a reminder for other corporations to check and possibly revise their own framework or, as a reassurance that they are still up-to-date.

Internally, comparisons have two functions. Firstly, it forms the groundwork for defence mechanisms by examining the characteristics of AML violations at peer banks. Secondly, it serves the identification of vulnerabilities in their own risk framework to mitigate the risk of being sanctioned for comparable shortcomings. Comparisons are guided by the question whether the same could have also happened at their bank. Peers at other banks recognised that (in the past) their bank also (used to) struggle with AML compliance:

“When that scandal happened at ING, everyone at our bank also thought that it could have been us too.” (R4, AML manager, systemic bank).

Most respondents spoke of AML cases (at their banks and others) while referring to the general shortcomings and challenges in the industry. This can inhibit peer condemnation. The representation of AML shortcomings as an unfortunate normality resembles what Whyte (2016, p. 175) in his study on neutralisations in the automobile industry has called denial of deviance; a practice that is normal in the industry cannot be deviant. Neutralisations can also be understood as a defence mechanism. Taking a protective strategy of face work, geared towards saving face of others, can avoid hostility or prevent a backlash (Goffman, 1967, p. 12). The condemnation of violations at another bank can result in such a backlash if the condemning bank also ends up being sanctioned for AML non-compliance. Indeed, the other two systemic banks have, since then, likewise been sanctioned (ABN AMRO) or are under investigation (Rabobank) for AML non-compliance.

Comparing practices leads to two simultaneously occurring stages: distancing (stage 3) and investing (stage 4). This again shows the ambiguity of peer responses: on the one hand, peer banks take sanctions in the industry seriously, improving and accelerating their own AML programs. On the other hand, they engage in a process of dissociation to maintain their own reputation and avoid being compared to sanctioned parties.

Stage 3: Distancing

In this third stage, peers actively seek (small) differences to distance from sanctioned parties. After having compared practices, they look for divergences in their own approach that differentiates them from sanctioned peers:

“You want to distance yourself as an organisation. You don’t want to measure yourself with a party where the ins and outs of what went wrong are publicly known. There is a fear whether the same thing could happen to us. This fear makes you compare practices but also distance yourself. […] You start looking for differences to be able to say ‘that’s them, not us’. In such a moment [publication of ING settlement] you really look for these differences” (R10, compliance officer, systemic bank).

Distancing is shared between colleagues as back stage behaviour and is usually not shared with outsiders. However, when publicly confronted with their own shortcomings, distancing can enter the front stage. In an article at the Dutch newspaper de Volkskrant, Rabobank emphasised that the AML violations for which they received a 1 million euros (regulatory) fine in 2018 on incomplete client files are incomparable to the ING case (Haegens, 2019) although incomplete client files were also part of the ING settlement. Similarly, after the Dieselgate at Volkswagen, BMW and Daimler choose to “vehemently refute similar transgressions” when being confronted with the violations in the media (Jong & Van der Linde, 2022, p. 4).

Distancing also partly correspondents with insights from a report on the support for AML compliance among civil-law notaries, brokers and appraisers in the Netherlands (Hoogenboom, 2021). Respondents in Hoogenboom’s research engaged in ‘denial’, that is recognising AML insufficiencies in general but denying own insufficiencies (‘yes there are money laundering problems, but not with us’). In the present research, banks recognised that problems with AML compliance are industry-wide but also concern their own organisation. As denial was impossible, they downplayed their own problems compared to the issues at other banks. This downplaying implies that shortcomings elsewhere were perceived as more severe and deviated from their own practices.

Employees at other banks particularly disapproved of one practice: ‘capping’ at ING bank. Capping entails that ING’s transaction monitoring system was capped to only show three alerts of potential money laundering signals per day for some risk categories (NPPS, 2018, p. 11). Several respondents in the first and second line (AML, compliance and risk) at smaller and systemic banks expressed their bewilderment and disapproval of the practice calling it “serious” (R13, compliance officer), a “no go” (R24, senior AML analyst), “fundamentally wrong” (R21, AML manager) or even “shit” (R7, conduct & ethics). The term of capping has become a common joke among employees in transaction monitoring (where capping can be done):

“Capping, that’s simply out of the question. That joke is occasionally made: “Shall we cap?“ We won’t, but it’s the one that comes back the most. […] in the Houston report [name of criminal investigation into ING], there were things that could not have been us, with ‘capping’ as an example. Do we always have our workload under control at this bank? No. But do we cap? Never. We’ll just let it happen and come up with a plan on how to fix it. But we don’t sweep it under the carpet.” (R21, AML manager, small bank).

The practice of capping is an example of a characteristic of the act that was regarded as ‘out of the ordinary’ by peers (Benson, 1985, p. 599). Respondents labelled capping to distance themselves, prevent being affiliated with (shortcomings at) ING bank, and to qualify their own shortcomings. In this example, peers choose for a defensive orientation of face work. Above all, saving the bank’s own face prevails. As no bank wants to be the poorest performer (Verhage, 2011), banks label others to uplift themselves.

Distancing and disapproval were mostly present in peer relations with lower levels of identification (characteristic of the actor). Condemnation was foremost found between systemic banks and newcomer, online banks. Then labels do not (solely) focus on the act, as it is the case in the example of capping, but on the actor. Employees at online banks use labels such as “old-fashioned”, “outdated” or “dusty” (R28 and R29) to free themselves from the traditional image of banks. This self-othering by banks can be explained by the dominant negative sentiments towards banks (culture)since the financial crisis. By explicitly paying attention to differences in culture (e.g., in hierarchy, work climate, bonding or clothing), smaller banks create their own identity as ”the other bank”, “the bank that does things differently” (online banks) or ”the bank with the human aspect” (smaller retail and online banks). Newcomer banks, in turn, are seen sceptically by the systemic and, to a lesser extent, smaller well-established banks. They are jokingly labelled as “crook banks” where all criminals go (R3, R10, R16, R35). AML practices at newcomer banks were in few cases described as a “misery” (R25, security affairs; R16, AML expert). This critical notion towards newcomer (online) banks is particularly interesting as it is the systemic banks that (as yet) have been heavily sanctioned for AML violations, as pointed out with a frown by some respondents (R32, R35) working at one of the smaller banks. In his dissertation on organisational character and integrity Fiorito (2021) found efforts of “neutralizing character work” (p.150), that is the neutralisation of negative character cues, and “accentuating character work” (p. 154), that is reshaping history and highlighting positive character cues, at corporate scandals. The findings of the present research suggest that character work also takes place between peer corporations.

The use of labels to create their identity puts the labelling approach in a different light. Traditionally, the labelling approach highlights how the self-identity is shaped by being labelled. In this case, however, we see banks actively engaging in the labelling of other banks to create their own (different) identity. They use stereotypes that others hold and further cultivate them for their own advantage.

Stage 4: Investments

Besides distancing from peer corporations and downplaying their own shortcomings, peers learn from cases in the industry. As AML non-compliance is an industry-wide problem, other banks started to invest in their own systems and procedures. Highlighting their own investments and lessons learned can be beneficial as it deflects attention to the future while emphasising their learning experience (Murphy et al., 2017).

As a reaction to the first major case at ING bank in 2018, all banks heavily invested in AML. They set up remediation trajectories, Know Your Customer related departments and hired numerous employees. The credo at that time was described by one respondent from a systemic bank as “we don’t care about costs” (R7, conduct & ethics, systemic bank).

In this stage, peers also criticise each other’s approaches. An employee who works at a systemic bank blamed the other systemic banks for being late realising the urgency of AML and acting upon it (R6, AML analyst, systemic bank). This is often fixed by enticing away employees from other banks. While high turnovers are common in the field, offering “ridiculous amounts of money” (R21, AML manager, small bank) to attract employees from other banks was disapproved by several respondents. Others (R7, R10, R22, R25, R35) criticised or joked about how other banks, or even their own bank, try to attract new employees with expensive marketing campaigns. An AML expert at a smaller bank implicitly referred to a campaignFootnote 2 by ABN AMRO to attract young professionals as ‘anti-money laundering heroes’, sharing the following:

“We are not one of the large banks [laughs]. We fix it differently. We have our own heroes to do that [laughs]. […] We have a couple of smart people in IT who also say ‘we are not really impressed by what the large banks are doing’. That’s our way of looking at it. It’s not better by definition. But we are not that easily impressed by their increase in FTE” (R22, AML expert, small bank).

AML cases mostly work as an accelerator for peers, highlighting the sense of urgency to act and thus often result in quick fixes rather than structural changes. Slowly, these quick fixes and investments driven by panic change into a determination to, in the words of several respondents, “do it [AML] right” and forestall regulation or sanctions. When asked about the main driver to comply with AML regulations, respondents stress their intrinsic motivation (front stage behaviour). Their overall narratives show that the underlying motivation, however, is rather extrinsic, with reputation management and commercial goals prevailing. This is similar to the conclusions reached by Favarel-Garrigues et al. (2011), Harvey and Lau (2009) and Verhage (2009) for reporting behaviour and risk assessments: banks foremost comply with AML to cover their back and avoid the reputational damage of sanctions. Again, defensive mechanisms prevail.

Stage 5: Cooperation

The results also show two stages of peer interactions on the defense spectrum that fit the idea of closing ranks: joining forces in cooperation (stage 5) and defiance (stage 6). This is particularly interesting in comparison to condemning responses by authorities or the public that result in stigmatisation and exclusion. While peers engage in some forms of condemnation (in stage 2, 3 and 4), this did not result in stigmatisation. Rather, with defensive mechanisms prevailing, banks chose to cooperate.

Besides fixing their own problems, banks seek out to each other under the banking association, in official partnerships, or via informal contacts. As banks face the same regulation, enforcement and challenges under AML/CTF, several private or public-private partnerships have evolved (see Favarel-Garrigues et al., 2011; Verhage, 2015). Cooperation is a move towards sharing knowledge and combining forces to be more effective in tackling money laundering but also towards saving face as an industry.

Overall, a narrative of being partners rather than competitors on AML (and other security) issues prevails in the banking industry (R5, R6, R7, R15, R25, R28, R29, R36). As for now, most partnerships are still developing and often consist of the systemic banks, public authorities, plus at times one or two smaller banks. This leads to a “large banks party” (R10, compliance officer) in cooperation. Some small banks who are eager to participate in partnerships report that they must prove themselves first as a serious player to get a place at the table. In the perception of some respondents at smaller banks (R19, R21, R32), the systemic banks are cautious in cooperation, still being entrenched in competition rather than cooperation. The systemic banks have a longer history of experience, with several remediation trajectories and outsourced consulting projects, impairing the incentive to share these with newcomers. Newcomer banks, on the other hand, can profit from their technology-based systems, not having to deal with major legacy problems. The partners narrative obscures the fact that banks remain competitors who are reluctant to share knowledge, tools and best practices in which they have heavily invested with peers. Ambivalence in peer responses as portrayed in the other stages indicates that cooperation is foremost a tactical move (see stage 6) that has its boundaries as peers still engage in distancing and condemnation. While cooperation between banks generally makes condemnation of peers’ behaviour less appealing, the two do not exclude each other. Even in cooperation, peers actively work on their own presentation towards another. Some respondents indicated that, when being among peers from other banks, they carefully select their words and framing of own procedures and problems to keep up their own image towards peers and not lose face. On other well-established partnerships, the fraud protocol for instance, non-affiliation or insufficient compliance with regulations and expectations results in newcomer banks being labelled as “the joker” (R31, AML analyst, small bank).

Stage 6: Defiance

Besides cooperation, and mainly driven by lasting fear and frustration about the effectiveness of increasing AML investments, another notion emerged from the interviews: defiance of the regulator’s approach and, to a lesser extent, societies expectations from banks. Defiance is not a singular act by one bank, but a combined approach of peers in the industry. While the other stages of peer interactions are mostly directed towards each other, defiance particularly focuses on how banks position themselves against the regulator. Such positioning is key to the concept of motivational postures (Braithwaite et al., 1994). Motivational postures shed light into the dynamic interplay between those who regulate and those who are regulated. Banks’ defiance resembles the motivational posture of resistance towards regulators, a confrontational approach of anger about how the regulator operates and uses its power (Braithwaite et al., 1994, p. 5).

A few respondents describe an active choice in the industry to stick together and follow the same line instead of scapegoating or blaming each other. Under the shared responsibility on AML/CTF, banks hold an interest in being consistent in their approach:

“What I notice in the banking industry is that this subject [AML] is something where people walk on eggshells. Everyone is a bit in panic at times: ‘We really need to do something about this.’ In one of the interbank associations that I attend, people really stick together: ‘if we as banks do this together this way, then we will at least be safe.” (R19, compliance officer, small bank).

Protecting themselves and others in the industry is secured by distancing from and confronting the regulator. Banks form a lobby or front to have a stronger standing and thereby join forces against authorities:

“There has also been a lot of contact between the banks on these [AML] files in recent years. […] Working together against the regulator too. So much is expected from banks but the party [Financial Intelligence Unit] to oversee all the reported transactions by banks is actually very small” (R15, risk manager, systemic bank).

Defiance is also visible in a lawsuit by online bank BUNQ against the Dutch regulator on BUNQs use of artificial intelligence in the onboarding of new clients to perform customer due diligence. The case follows a dispute over an earlier imposed instruction. According to a newspaper article (Betlem & Pols, 2022), other banks support BUNQ’s step and critique on the regulator’s expectations, calling it ‘brave’ and ‘necessary’. A respondent working at a systemic bank in risk management elaborates that the support of BUNQ is foremost shared on the workplace among colleagues rather than in official communication by banks. Such information hardly reaches the public. In this case, however, fear turns from back stage behaviour, being shared within banks, to front stage in the form of defiance and peer support. While many people working in the industry share doubts on the regulator interpretation of regulation, until now no bank dared to speak up and defend themselves publicly.

Defiance and blaming authorities, in contrast to peer condemnation, which was difficult to elicit in the interviews, was shared without even asking. In defiance rests a critique on the current Zeitgeist of responsibilising and criticising banks. Banks are left with a feeling that whatever they do, they will be blamed for it the one way or the other. Defiance then not only prevents the internalisation of AML at banks, but also hinders informal social control effects between corporations as peers focus on blaming authorities.

Discussion and conclusion

While much literature has focused on the condemnation of corporate behaviour by the public and the authorities, less is known on responses by corporations’ own peers. The case study of non-compliance with AML regulations in the Dutch banking industry was used to shed light on how corporations respond to corporate deviance of their peers.

To answer that question, the paper drew on theoretical foundations of (symbolic) interactionism. These foundations have been specified in six stages of peer interactions that emerged from the data: panic and fear, comparing practices, distancing, investments, cooperation and defiance. These stages are (yet) dependent on external labelling that puts corporations under pressure to respond to peer corporations behaviour.

The first big AML case in the Netherlands, the ING settlement, quickly spread to the whole industry, triggering a contagion effect. For peer responses, contagion played out in banks comparing their practices and heavily investing in AML to avoid sanctioning and reputational damage. Peers employed defence mechanisms (face work) as responses to AML cases at other banks (e.g. differentiation in stage 3 and closing ranks in stage 5 and 6). Several peer responses and interactions (stage 2, 3 and 4) fit a frame of reassurance for their own approach towards AML (see Thornton et al., 2005; Van Wingerde, 2012).

Peer labelling between corporations remained limited and, if it occurred, served the purpose of differentiating, image guarding (both related to AML cases) or creating an image as ‘the other’ (unrelated to AML cases). While formal labelling and shaming entail the stigmatisation of the violating actor, stigmatisation was not part of peer condemnation (or responses more generally). Peer stigmatisation can be tricky given the risk of backfiring and the realisation that no bank seems to have AML properly in order. After all, AML cases did not affect banks’ reputation in the industry as they retained their status among peers. The few labels in responses to AML violations that were found in this research were not attributed to the actor (the bank) but the act (AML violations). This supports earlier research on neutralisation by white collar offenders who rather focus on the act than the actor and deny a guilty mind (Benson, 1985; Stadler & Benson, 2012). The labelling of ‘capping’ at ING bank can be best understood as a strategic choice by industry members. It is questionable whether ‘capping’ would have been disapproved by peers in the industry if the case (and particularly this practice) had not been publicly scrutinised and other banks would not face similar challenges regarding AML compliance, creating a need for differentiation. In fact, most peer responses remained in the back stage (within corporations) and only entered the front stage when externally triggered.

Under the current AML framework, criminalisation increasingly overshadows responsibilisation. As a result, enhancing AML in banks is foremost a result of reacting, pleasing regulators’ demands and avoiding sanctions and reputational damage. This implies that discussions about norms on AML and an internalisation of the Zeitgeist of financial institutions contribution to societal and environmental challenges are still lacking in the Dutch banking industry. This situation leaves little room for informal social control between peers that has the potential to foster compliance. Especially in cases of corporate deviance that fall in the grey area or for corporations with extensive financial and economic power, informal social control by peers is, just like shaming, all we got (see Jacquet, 2016: 106).

While this article focused on peer interactions, during interviews respondents often engaged in blaming others. Blaming was targeted at financial regulators, the Netherlands Public Prosecution Office, (the plethora of) AML regulations, and to a lesser extent the media, the public and politics. With blaming others, banks engaged in condemning the condemners (Sykes & Matza, 1957; Whyte, 2016) in which ‘criminals’ justify deviance by pointing fingers at authorities. While this more generally fits corporate responses of deflecting responsibilities (Benson, 1985; Shichor & Heeren, 2021; Stadler & Benson, 2012; Van Rooij & Fine, 2018) and engaging in a blame-game (Huisman, 2011), it also exposes the wicked relation between authorities and financial institutions in which banks no longer dare to act. This problem has been addressed by Eren (2021) who warns of a regulatory bulimia and by Levi (2018) highlighting the pressure towards de-risking and zero tolerance. This article shows another risk of these practices: cooperation against the regulators (defiance). Defiance hampers informal social control between corporations, resulting in the negligence of the benefits of peer (dis)approval.

In other areas that are more open to self-regulation and show a strong interdependence between corporations (e.g., fraud in the banking industry), peer social control can be even more promising. Further empirical research on peer interactions, extending to cases of commission (e.g., LIBOR scandal or Wirecard fraud) in and beyond the banking industry is needed to better understand what is regarded as deviant among corporations and how this is reinforced within organisational settings. This article took one specific case, AML shortcomings in Dutch retail banking, to explore peer responses in the corporate world. The research tried to include different voices within retail banking. However, given that only six retail banks were represented it does not cover the full variety of the industry. Especially smaller banks that are not affiliated with any industry partnerships were not represented in the article. As this research selected respondents on affiliation with financial economic crime in their job, their view is overrepresented while other positions were not included. Ideally, future research goes beyond interview data and content analysis, comprising ethnographic research to explore the full range of peer responses that go beyond the spoken word.