Varieties of Capitalism and the Institutionalization of Whistleblowing: Centralization and Civil Society Engagement

The scholarly definition of whistleblowing in a work context is ‘the disclosure by organization members (former or current) of illegal, immoral or illegitimate practices under the control of their employers to persons or organizations that may be able to effect action’ (Near & Miceli, 1985, p. 4). Efforts to institutionalize whistleblowing have increased since the beginning of the 21st century, at least in a regulatory sense. Calland and Dehn (2004) count only six countries with whistleblowing legislation prior to the year 2000. In 2004, Vandekerckhove (2006) counted seven countries with legislation and a further six where legislative bills were being discussed. A decade later, Thüsing and Forst (2016) included 23 countries in their review of whistleblowing legislation. And as recent as 2025, the ILO published an analysis of national laws containing whistleblowing provisions in the public sector across 67 ILO Member States (ILO, 2025).

Comparative studies of whistleblowing have focused either on individual intentions to blow the whistle and values around whistleblowing (for an overview see Vandekerckhove et al., 2014) or on what legal protection for whistleblowers is in force, i.e. ‘who is protected, what kind of behaviour is protected, and what kind of behaviour whistleblowers are protected against’ (IBA, 2021; Thüsing & Forst, 2016). Such legislation increasingly converges on a three-tiered model, as articulated in Vandekerckhove (2006, 2010) and taken up in the Council of Europe Recommendation (CM/Rec(2014)7), and the EU Directive on whistleblowing (EU2019/1937), which implies that protection measures are in place for whistleblowers in relation to their reporting of wrongdoing (1) inside their organization, (2) to regulatory agencies and (3) to the wider public.

However, comparative law studies do not exhaust institutional analysis. In the context of whistleblowing, Vandekerckhove (2010, p. 31) notes:

In this chapter, we are interested in the ‘operational forms’ taken to make whistleblower protection real. We deem that gaining insight into different ‘operational forms’ of institutionalization is relevant to current debates on whistleblowing, in particular within the European Union. The EU’s principle of subsidiarity stipulates that horizontal legislation (i.e. an EU Directive) needs to be justified by evidencing that leaving a particular intervention to the member states would harm the interests of the European Union. DG GROW (2017) and DG JUST (ICF, 2017) provided these justification studies, and the EU Directive ‘on the protection of persons who report breaches of Union law’ (EU2019/1937) entered into force on 16 December 2019. The European Commission published in 2024 a first report on the transposition of the Directive into Member States’ national legislation (EC, 2024), finding many differences remain and we are far removed from horizontal legislation. The Directive nevertheless includes stipulations for institutional arrangements, which is the focus of our analysis in this chapter. The EU Directive stipulates that member states need to mandate new or existing organizations and agencies to handle whistleblower reports (Art 12, Recital 64) and provide support measures (Art 20, Recital 89). Given that countries (also within the EU) have different institutional histories and make-ups, the research question this chapter answers is: what differences between countries can we expect in institutionalization of whistleblowing?

Loyens and Vandekerckhove (2018a, 2018b) carried out a comparative study of institutional landscapes. They observe two main points of difference: (1) in what government agencies do and (2) in civil society activism. However, Loyens and Vandekerckhove (2018a) provide description only for a limited sample (n = 11), without any attempt to provide a theoretical explanation. This chapter uses a ‘varieties of capitalism’ (VOC) approach to provide a theory driven explanation of these observations.

VOC denotes a new body of work (Hall & Soskice, 2001) within comparative political economy. The theory stipulates that within a political economy, two or more institutions will be complementary to each other. The theoretical contribution of our study is first to provide a further confirmation of VOC, extending empirical work into yet another sphere of political economy, namely whistleblowing. A second theoretical contribution of our chapter is to use our findings to provide a discussion on institutional emulation, which is something Hall and Gingerich (2009) speculate on. However, our main aim of the analysis in this chapter is practical rather than theoretical. We seek to use VOC as a theoretical lens to generate a critical discussion of institutional reforms currently debated in Europe, in the context of whistleblowing.

To that end, the chapter is structured as follows. The next section puts forward the main tenets of VOC. The section after that operationalizes the VOC model to explore the institutionalization of whistleblowing. We formulate propositions and explain our sample and measurements. The section after that, presenting and discussing our findings, is the longest one of the chapters and contains the contributions: corroboration of the propositions, discussing emulation and drawing a positional matrix with regard to institutionalization of whistleblowing in Europe. We conclude by summarizing our study and pointing out its limitations.

Historically, comparative political economy has concerned itself with how national variations in institutions affect economic performance. An important line of work consists of looking for interaction effects among institutions across different spheres of the political economy. For example, interaction effects have been found between monetary institutions and institutions that govern wage coordination. Soskice (1990, 1991) started to develop a theory based on neocorporatism and regulation scholarship, which became known as ‘varieties of capitalism’ (Hall & Soskice, 2001).

VCO is a theory about the nature of institutional complementarity that can be found in the political economies of the developed world. Paradigmatically, VOC is based on the notion that political economies are characterized by how firms and other actors coordinate their endeavours. It starts from the observation that countries have very different institutions but are nevertheless capitalist economies. Theoretically, its two propositions (Hall & Gingerich, 2009) are that (1) countries cluster into identifiable groups based on the extent to which firms rely on market or strategic modes of coordination and (2) complementarities exist between institutions in different spheres of the political economy. VOC distinguishes two idealtypes although other have been suggested (e.g. Nölke & Vliegenthart, 2009; Schneider, 2009): in liberal market economies (LME), relations between firms and other actors are primarily coordinated by competitive markets and in coordinated market economies (CME), more strategic interaction with trade unions, financial institutions and other actors can be found. Institutional landscapes rather than one particular institution is the object of VOC research. The main assumption of VOC seems to be that either institutions aim at market perfection (LME) or institutions do what markets don’t (CME) and hence why institutional complementarities can be found.

Most research using VOC theory was conducted as comparative qualitative case studies (e.g. Campbell & Pedersen, 2007; Feldmann, 2006). Hall and Gingerich (2009), however, have provided some quantitative testing of VOC. They test VOC’s two propositions for what they argue are the two most important institutional spheres of the political economy: labour relations and corporate governance. Hall and Gingerich (2009) note a difficulty in obtaining comparable measurements and resort to calculating factor scores based on OECD data, which limits the number of countries that can be included in the sample. They find measures for labour relations and corporate governance correlating strongly and significantly and find that a plot of those measures shows countries in LME and CME clusters.

Hall and Gingerich (2009) then go on to hypothesize systematic variation across other spheres. For example, they argue that organizations that coordinate wages can also be used to coordinate training systems and provide confirmatory analysis for this. They are furthermore able to confirm that institutionalized practices in the spheres of social protection, and product market regulation vary across countries in correlation with measures for labour relations and corporate governance and hence in correspondence with the characterization of the political economy across the LME–CME spectrum. Hall and Gingerich’s (2004) quantitative confirmation of VOC was an important milestone in the development of the VOC research stream. However, the authors note that whilst it is important to quantitatively confirm that uniformity exists for countries, how such uniformities come about remains an open question:

This chapter will explore how differences in institutionalization of whistleblowing can be understood through VOC. In the discussion of our findings, we will have a closer look at institutional emulation.

Loyens and Vandekerckhove (2018a, 2018b) made the (to our knowledge) first comparative study of institutionalization of whistleblowing. As legislation protecting the employment status of whistleblowers is enacted, Loyens and Vandekerckhove (2018b) note that an emerging policy question is what institutions will implement the legislation. Their research question does not relate to what court will be hearing claimants but rather which organizations and actors will take up roles in handling the whistleblowing throughout a whistleblower’s journey, from innocuously raising a concern with their employer to then approaching an external agency that might effect action, to seeking protection from retaliation, to perhaps approaching journalists and eventually seek vindication in a court (Kenny, 2019; Rothschild & Miethe, 1999; Vandekerckhove & Phillips, 2019). For Loyens and Vandekerckhove (2018a), the question of institutionalizing whistleblowing is: who does what? They distinguish 8 roles, summarized in Table 1.

Yet, Loyens and Vandekerckhove (2018a) remain descriptive; they do not attempt to theoretically explain their findings or derive further implications. We assert in this chapter that what Loyens and Vandekerckhove (2018a) observe is not a coincidence. Rather, the institutionalization of whistleblowing – i.e. the uptake of roles in handling whistleblowing – happens in an already existing institutional context. As can be seen from Table 1, the handling of whistleblowing requires several professional skills and mandates (cf. advisory, investigative and adjudicative functions), which require coordination. VOC posits that coordination within a political economy can be characterized as either competitive (LME) or strategic (CME) and that there is congruency across institutional spheres (Hall & Gingerich, 2004). Thus, if VOC is a valid theory, institutionalization of whistleblowing will be similar to other institutional spheres in the political economy. We therefor believe it is possible to explore to what extent VOC corroborates the observations of Loyens and Vandekerckhove (2018a).

In order to formulate propositions, let us first consider the constructs and sample Loyens and Vandekerckhove (2018a) used. They selected countries that had whistleblowing legislation in place and where there is at least one government agency that has at least one role in handling whistleblowing. They arrived at a sample of 11 countries: Australia, Belgium, France, Ireland, Israel, Netherlands, Norway, Republic of Korea, Serbia, the United Kingdom and the United States. Loyens and Vandekerckhove (2018a) used a sample of countries that had institutionalized whistleblowing – although in some, this was very recent (e.g. the Irish and Serbian legislation was from 2014, the French and Dutch was from 2016).

In nearly all of the countries included in the sample, recent legislative changes are still being implemented into institutional mandates as we write this chapter. Hence, an important limitation of our research presented here is that we arguably use outdated data. We acknowledge this but remain confident about our analysis on these two points. First, we are still seeing a lot of institutional mandates changing as we are writing this chapter (end of 2024). If we would have done a new data collection, it would also have been outdated by the time the book goes to print. Second, because the data we use are pre-transposition of the EU Directive, we position our research here as the baseline measurement for future studies.

How do Loyens and Vandekerckhove’s (2018a) observations corroborate with VOC characterizations of political economies as LME or CME? They observe that although every country had at least one government agency with an explicit mandate to take up a handling role, in some countries government agencies would take up and combine more roles than in other. They also observe that in some countries, there would be one or a few government agencies with a vast market scope whereas in other countries, many government agencies would take up the same role but each for a distinct part of the economy. For example, ASIC (Australian Securities and Investment Commission) in Australia can receive whistleblower reports and investigate wrongdoing on any breach of law by any private sector corporation. In France, the Defender of Rights (Défenseur des Droits, the national ombudsman) is mandated to do so for both public and private sector. In contrast, the SEC (Securities and Exchange Commission) in the United States, well known for its whistleblower-reward scheme mandated by the Dodd-Frank Act, is one of 23 Federal regulatory agencies and only occupies itself with fraud by companies noted on the stock exchange. Hence, what Loyens and Vandekerckhove (2018a) observe here are different degrees of centralization of whistleblowing handling roles. A higher degree of centralization implies a more strategic coordination (CME) of different handling roles.

Loyens and Vandekerckhove (2018a) also observe that in some countries, civil society organizations (CSOs) take up handling roles whereas in other countries, CSOs hardly play any role in the handling of whistleblowing. Hence, what they observe here are different degrees of civil society engagement. They understand CSOs as non-governmental organizations that receive whistleblower disclosures and – often in a more activist way – try to enhance whistleblower protection; journalists and media organizations are not included. Bair and Palpacuer (2012) found that activist CSOs in the United States and Europe had different structures and campaigned differently, corresponding to VOC characterizations. They observe ‘NGOs substituting for the state’ and ‘raw struggle’ in CME contexts (Bair & Palpacuer, 2012, p. 538). In line with their reasoning, we deem that a high degree of civil society engagement stems from weak formal strategic coordination power and thus that more civil society engagement implies more competitive coordination (LME) of whistleblowing handling roles.

We further note that Hall and Gingerich (2009) argue that labour relations and corporate governance are the two most important institutions in the political economy. They also confirmed that measures for these two institutions correspond with LME and CME characterizations.

We thus formulate the following propositions for our VOC model of whistleblowing institutionalization.

• P1. Strong coordination of labour relations and corporate governance enhances the centralization of whistleblowing handling roles.

• P2. Weak coordination of labour relations and corporate governance coincides with high civil society engagement in whistleblowing handling roles.

• P3. High civil society engagement in whistleblowing handling roles coincides with low centralization of whistleblowing handling roles.

Hall and Gingerich (2004) point out that obtaining measures for institutions in different spheres of the political economy is difficult and error prone. They obtain measures from OECD data. We will use Hall and Gingerich’s (2004) measures for the coordination of labour relations and corporate governance to test our hypotheses. Hall and Gingerich (2004) further assert that obtaining measures requires ‘intense observation’ of a particular sphere. We will therefore use Loyens and Vandekerckhove’s (2018b) observations to calculate measures for centralization of handling roles and civil society engagement.

One implication of our approach to measurements is that our sample is limited to countries used in both Hall and Gingerich (2004) and Loyens and Vandekerckhove (2018a, 2018b). We further do not include Belgium because the whistleblowing legislation only covers the public sector (in 2018) whereas in the other countries, both public and private sectors are covered by the whistleblowing legislation. This leaves us with a sample of seven countries: Australia, France, Ireland, Netherlands, Norway, the United Kingdom and the United States. We acknowledge that this small sample size poses a limitation to our research, especially with regard to the testing of our hypotheses.

As a measure for centralization of whistleblowing handling roles, we use:

• CTR = (highest #roles by one organization/total #roles performed) x private sector coverage

  This is the ratio of the highest number of whistleblowing handling roles taken up by a single governmental actor, per the total number of roles performed in that nation, taking into account the scope of that the organization’s mandate.

  As measure for civil society engagement, we use:

• ENG = #roles by civil society organizations/total #possible roles

  This is the ratio of the number of whistleblowing handling roles taken up by CSOs in a nation, by the total number of possible roles, which is 8 (cf. Loyens & Vandekerckhove, 2018a; 2018b). Table 2 gives an overview of the resulting measures.

  Table 2.

  Measures Used in This Chapter.

  Nation	Highest #Roles by One Gov Actor	Total #Roles in Nation	#Roles by Civil Society	Total #Roles Possible	Ctr	Eng	LRel	Gov
  Australia	2	6	1	8	0.33	0.12	0.29	0.47
  France	5	7	3	8	0.71	0.37	0.6	0.82
  Ireland	2	6	3	8	0.33	0.37	0.28	0.35
  Netherlands	8	8	1	8	1	0.12	0.53	0.74
  Norway	2	4	0	8	0.5	0	0.81	0.74
  UK	2	8	4	8	0.25	0.5	0.04	0.14
  USA	4	7	4	8	0.03	0.5	0	0

  Nation	Highest #Roles by One Gov Actor	Total #Roles in Nation	#Roles by Civil Society	Total #Roles Possible	Ctr	Eng	LRel	Gov
  Australia	2	6	1	8	0.33	0.12	0.29	0.47
  France	5	7	3	8	0.71	0.37	0.6	0.82
  Ireland	2	6	3	8	0.33	0.37	0.28	0.35
  Netherlands	8	8	1	8	1	0.12	0.53	0.74
  Norway	2	4	0	8	0.5	0	0.81	0.74
  UK	2	8	4	8	0.25	0.5	0.04	0.14
  USA	4	7	4	8	0.03	0.5	0	0

  LRel = Labour relations (taken from Hall & Gingerich, 2004).

  Gov = Corporate Governance (taken from Hall & Gingerich, 2004).

  #roles variables taken from country descriptions in Loyens and Vandekerckhove (2018b).

  Ctr (Centralization) and Eng (Engagement) calculated based on formulas provided in the text the other variables.

  View large

We mapped our variables along our three propositions. The results are shown in Figs. 1–3. For labour relations and governance, we used measures from Hall and Gingerich (2009) as discussed in the previous section. The other two variables – centralization and engagement – are our own measures for institutionalization of whistleblowing, based on observation of Loyens and Vandekerckhove (2018b).

Looking at the plot in Fig. 1 for centralization and corporate governance, we can discern three groups. The LME group consisting of the United States and the United Kingdom; the CME group consisting of France, Netherlands and Norway; and a third ‘middle’ group of Ireland and Australia. This grouping corroborates with how these countries are often characterized in the VOC literature (Schneider & Paunescu, 2012). What we can thus say is that in countries that tend towards the CME idealtype, institutionalization of whistleblowing tends to occur through governmental agencies and actors whose mandate combines many of the whistleblowing handling roles and with civil society actors performing very few of those handling roles. In contrast, in countries that tend towards the LME idealtype, institutionalization of whistleblowing tends to occur through civil society actors who take up a number of whistleblowing handling roles and governmental actors with a limited mandate for handling roles.

Looking at the plot for engagement and labour coordination (Fig. 2), we see that most countries in our sample are where we expect them to be, i.e., high labour relations coordination (CME idealtype) coincides with low engagement from civil society, whereas low labour relations coordination (LME type) coincides with high engagement from civil society, taking up handling roles. France and Australia are the exceptions here. Our findings suggest that whistleblowing in Australia may be under-institutionalized whereas in France, we see both strong labour relations coordination (CME) as well as a high level of civil society engagement with whistleblowing.

An important observation from Fig. 3, where we plot our measures for engagement and coordination, is we can discern three groups. The first (the United States, the United Kingdom, Ireland) is placed within the expectations of the VOC model. The second group (Australia and Norway) suggests a lack of institutionalization of whistleblowing, i.e., given the low level of centralization of handling roles, we would expect more CSOs picking up those handling roles. A third group (France and Netherlands) is most exceptional because these countries score high on both centralization and civil society engagement in handling roles. Given that VOC is an idealtype approach – i.e. CME and LME are two idealtypes defining a range (Hall & Gingerich, 2004, 2009; Hall & Soskice, 2001) – the truly interesting cases are those that turn out not to be as expected. Empirical phenomena are never fully in line with an idealtype, rather, reality more or less approximates an idealtype. It is precisely where a particular case deviates from the ‘rational’ idealtype that we can reach understanding (Vandekerckhove, 2006) and explanation of social reality (Ringer, 1997). What does that mean for our findings and what insights can we derive with regard to institutional emulation?

In the United States, institutionalization of whistleblowing happened in a piecemeal way (Vaughn, 2012). There are more than 40 laws that have whistleblowing provisions, each with a specific scope and jurisdiction. At the federal level, there are 23 regulatory agencies that have a mandate to fulfil a handling role. The most well-known one in the context of whistleblowing is the SEC, mandated through the Dodd-Frank Act to carry out investigation of wrongdoing, retaliation, offer protection, and de facto through settlements, issue corrective actions. In some cases, the SEC also issues financial rewards for whistleblowers. The SEC has an internal team called the ‘Office of the Whistleblower’. Yet, the SEC only handles whistleblower reports that concern possible violations of the federal securities law. There is no strategic coordination with other regulators that handle whistleblowing reports for other laws. In the United States, a number of non-profit CSO focus on whistleblowing. Perhaps, the best known one is Government Accountability Project (GAP), which is also the oldest whistleblowing CSO in the world, established in 1977 (Vaughn, 2012). GAP takes up a number of handling roles, mainly the ones that regulators do not (Loyens & Vandekerckhove, 2018b).

The United Kingdom did not take a piecemeal approach, at least not in its approach to legislation. It enacted in 1998 one whistleblowing legislation (Public Interest Disclosure Act 1998) that covered all sectors. It was one of the first ‘stand-alone’ whistleblowing laws. However, it was fully implemented through existing organizations and amending existing legislation (Employment Act), without any strategic coordination. There are more than 50 regulatory agencies mandated to receive whistleblowing reports. Adjudication on the retaliation is separate from that on the alleged wrongdoing. A dedicated CSO became active in 1993 and has more recently been joined by others, readjusting the handling roles taken up between them.

If we look at the other side of the spectrum, we find Norway as the most pronounced CME nation in our sample. Its whistleblowing legislation was also implemented through existing organizations, mandates the labour inspection to take up handling roles rather than spread these across regulatory agencies. However, very few handling roles are actually performed in Norway. Thus, Norway is characterized by a strategic coordination of a limited number of roles, without CSOs taking up other handling roles.

Hence, in the countries in our sample that most resemble the LME and CME idealtypes, institutionalization of whistleblowing seems to have happened through emulation, i.e. ‘adoption […] of a programme already in effect in another jurisdiction’ (Rose, 1991, p. 22). Importantly however, we see auto-emulation, i.e. not making any institutional change by copying what other countries do but rather expanding one’s own existing procedures and mandates. In other words, ‘business as usual’, which is surprising for a phenomenon that is claimed to seek effecting change (cf. the definition of whistleblowing, Near & Miceli, 1985) – it does so by maintaining the status quo.

Ireland and Australia are considered to be countries in the middle of the LME–CME range. How is whistleblowing institutionalized in these countries? Ireland has one of Europe’s more progressive whistleblowing legislation, yet its implementation very clearly is an emulation of that in the United Kingdom: mandating existing regulatory agencies with limited handling roles and without any strategic coordination. In terms of CSO engagement, we also see similar handling roles taken up as in the United Kingdom but more concentrated across fewer CSOs.

We also see international emulation in Australia where the Australian Securities and Investments Commission (ASIC), mandated through the Corporations Act to take up handling roles. Loyens and Vandekerckhove (2018b, p. 72) note that ASIC was ‘built on SEC model’ but takes up slightly different handling roles than the SEC. Further, international emulation in Australia has not finished although Australia now seems to take its cue from Europe, more precisely Netherlands and France. We will now discuss these countries and their emulation.

Our exploration of the institutionalization of whistleblowing perceived through a VOC lens suggests the Netherlands and France are interesting precisely because we are not seeing what we would expect. In the Netherlands, a new specialist agency was created. In 2012, an ‘advice centre for whistleblowers’ was created as an independent government agency. In 2016, new whistleblowing legislation created the House for Whistleblowers (Huis voor klokkenluiders, hereafter House). The agency represents the highest possible centralization of whistleblowing handling roles (cf. Fig. 1), namely all eight (Loyens & Vandekerckhove, 2018b). This centralization involved the usurpation of roles that were previously performed in a slightly more decentralized way. For example, the Bureau of Public Sector Integrity (BIOS) saw its budget and personnel go to the House. Another example is psycho-social support, which was initially taken up by a CSO (the Expertgroep). However, in 2016 that CSO stopped its activities and took a seat on the governing body of the House.

Loyens and Vandekerckhove (2018b) note that during their research on whistleblowing institutions in 10 other countries than the Netherlands, many of their interviewees were awe-stricken about the Dutch House. No one seemed to know how exactly (or even whether) the House worked, yet many of their interviewees regarded the House as the new template for institutionalizing whistleblowing: one specific governmental organization with a more or less holistic provision of whistleblowing handling roles. There are indications of international emulation of this institutional template in our sample, which we will briefly comment on.

The French whistleblowing legislation – known as Sapin II – also dates from 2016 but did not create a new agency. Rather, the parliamentary ombudsman (Défendeur des Droits) was mandated to take up varied whistleblowing handling tasks. However, a parliamentary ombudsman usually mediates in interactions between citizens and government agencies. Interventions in private sector organizations is not straightforward. In the Netherlands, this was sufficient reason to create a new organization. France decided to give the mandate to the parliamentary ombudsman, yet it took a while to figure out how these new mandates could be incorporated. The executive decree shaping the operation of new mandates was not signed until late 2018. The emphasis of the French parliamentary ombudsman seems to be less on carrying out its own investigations – it does not have a budget for that – but rather ensuring strategic coordination (Loyens & Vandekerckhove, 2018b). Another difference between Netherlands and France is that in France the legislation did not result in the usurpation of CSO engagement but instead gave it a boost. Before the 2016 legislation, a small number of CSOs provided support and advice to whistleblowers on an ad-hoc basis without any of these perceiving themselves as a dedicated whistleblowing CSO (Vandekerckhove & de Graaf, 2017). It was TI-France who around 2014 started to campaign for whistleblowing legislation and the Dutch model of the House (which was then at the stage of law proposal). After the French executive decree in 2018, a new CSO was created – called the House of the Whistleblowers (Maison des lanceurs d’alerte) – and functions as a coordination of almost 20 other CSOs who can provide whistleblowing handling roles. Thus, the French emulation of the Dutch model amounts to a strategic coordination rather than usurpation of actors’ roles and a stronger CSO engagement.

The emulation of the Dutch House model – or at least attempts to emulate that model – can also be seen in Australia and the United Kingdom. Brown (2019, pp. 40–41), an influential voice in Australian whistleblowing policy making, advocates a stronger centralization of handling roles in the Australian ASIC, based on the Dutch model. In the United Kingdom, an All Party Parliamentary Group (APPG) was established in 2018. One of its central points of discussion is an ‘Office for the Whistleblower’, modelled on the Dutch House. There is also a new Bill, which would entail an overhaul of the 1998 legislation and which includes more strategic coordination of handling roles through a central agency – again a key tenet of the Dutch House model.

However, we are also seeing a potential reverse-emulation. The Dutch House has recently published a document that entails its new strategic vision (House, 2020), formulated on the basis of an independent evaluation (Ruys, 2017) and a letter of the national ombudsman (Van Zutphen, 2019). In its renewed vision text, the House announces that it will move away from a full centralization of all handling tasks towards a French-like emphasis on strategic coordination. The upshot of this change in direction was that Expertgroep, the CSO that had ceased its activities when the House was launched, has now reactivated itself. If this reverse-emulation materializes – i.e. if the House indeed puts down some handling roles, then updated measures for the Netherlands would make our correlations stronger, thus making the institutionalization of whistleblowing more in line with what we would expect from a VOC perspective or indeed, ‘business as usual’. Is that what we should hope for?

Before we go into that, let’s first summarize what we discussed so far. We can explain institutionalization of whistleblowing in our sample through emulation, but the countries in our sample suggest there are different kinds of emulation at play. In our sample, countries that were closest to one of the idealtypes (LME or CME) institutionalized whistleblowing through auto-emulation, i.e. emulation of existing institutional arrangements within the nation. This is done through piecemeal legislation (USA), expanding mandates of investigative and adjudicative agencies (regulators in the United Kingdom, labour inspection in Norway). In contrast, countries that are in the middle range, in-between LME and CME, institutionalize whistleblowing by emulating institutional arrangements of other countries (Ireland and Australia), i.e. international emulation. We also saw some institutional innovation (the House in the Netherlands), which was emulated in France, and will not potentially get reverse-emulated back into the Netherlands. Hence, what our study suggests is that (1) it might be useful to distinguish types of emulation and (2) type of emulation can corroborate with the proximity of a nation to a particular VOC idealtype. However, is this what we should hope for if we want whistleblowing to lead to systemic change?

Whistleblowers disclose wrongdoing to others ‘who may be able to effect action’ (cf. definition of Near & Miceli, 1985). Thus, whistleblowers call upon others to bring about a change which they themselves cannot make happen (Kenny et al., 2020; Moberly, 2014). It follows that the way in which recipients handle the whistleblowing determines whether change will occur (Rothschild & Miethe, 1999; Vandekerckhove, Brown, & Tsahuridu, 2014). Our findings suggest that the institutionalization of whistleblowing, i.e. the configuration of handling roles across different actors, happens in ways that are no different from how institutionalization usually happens. That is to say, our VOC approach suggests that CME type countries institutionalize whistleblowing through centralization of handling roles and low civil society engagement whereas LME type countries have high civil society engagement in handling roles that interact in a decentralized way.

But how can institutional change – which whistleblowers strife for – come about through institutions that need to change? Near and Miceli (1995, p. 681) defined effective whistleblowing as ‘the extent to which the questionable or wrongful practice (or omission) is terminated at least partly because of whistleblowing and within a reasonable time frame.’ Note that according to this definition, effective whistleblowing means bringing about change rather than winning a court case. Near and Miceli (1995) however only considered internal whistleblowing for developing a model. They regard external whistleblowing channels as one of the variables but worked with the assumption that external whistleblowing would be effective and thus drive effective internal whistleblowing. However, research suggests that whistleblowing escalates to public disclosures in the media because external whistleblowing to regulatory agencies is not effective. Thus very often, systemic or institutional change is precisely what is required to make whistleblowing effective (Andrade, 2015; Kenny, 2019; Weiskopf & Tobias-Miersch, 2016). Here then is the problematic nature of institutionalizing whistleblowing: the configuration and coordination of handling roles that is supposed to make whistleblowing effective in bringing about change tends to occur in ways that stabilizes the existing institutional spheres within a political economy. Theoretically worded, VOC predicts an institutional status quo. That might be desirable for a political economy in general or for its economic productivity specifically, but it is perhaps less desirable in the context of whistleblowing. Thus, we can posit that institutionalization tends to neutralize whistleblowing. The example of the Dutch House suggests that international emulation does offer opportunities to institutionalize a-typically and non-neutralizing, but this is not self-evident, and there remains always a tendency to, through reverse-emulation, revert to the status quo. If we want whistleblowing to be effective in bringing about change, then the way we institutionalize it matters, and we need to somehow ‘go against the grain’, as presented in Fig. 4.

Fig. 4 is a contingency matrix of institutionalization of effective whistleblowing along two dimensions, centralization of handling roles in a dedicated government agency and civil society engagement in handling roles. The LME and CME idealtypes can be found in the top left (LME) and the bottom right (CME) quadrant, as hypothesized from VOC and confirmed in our findings. The bottom left quadrant – low centralization and low civil society engagement – points at a lack of institutionalization of whistleblowing. If there are any actors performing whistleblowing handling roles, these are not strategically coordinated nor are there any civil society groups who ‘calls out’ regulatory agencies that neglect whistleblowers or fail to prevent retaliation, campaign for better whistleblower protection, provide support, etc.

The institutionalization of whistleblowing to hope for, within the contingency matrix, is the top right quadrant. It is the institutionalization characterized by a high level of centralization and a high level of civil society engagement. Handling tasks are extremely well coordinated strategically, and there is a body of CSOs that holds government agencies to account for how handling roles are performed. It follows that because in LME type countries, it is likely that institutionalization happens through a high CSO engagement and a low centralization, whistleblowing will be more effective if effort is made to increase centralization of handling roles. It also follows that because in CME type countries it is likely that institutionalization happens by centralizing handling roles, whistleblowing will be more effective if also CSOs engage in handling roles.

In this chapter, we used a ‘varieties of capitalism’ approach to provide insight into how institutionalization of whistleblowing is taking place. We used existing observations of institutional particularities and institutional measures to formulate propositions around the congruency of the institutionalization of whistleblowing with institutionalization in other spheres of the political economy. For whistleblowing, we distinguished eight different handling roles. We calculated measures for how many handling roles for private sector whistleblowing are taken up by government agencies and for to what extent CSOs engage with these handling roles.

Our findings suggests that overall, CME type countries institutionalize whistleblowing through centralization of handling roles and low civil society engagement whereas LME type countries have high civil society engagement in handling roles that interact in a decentralized way. However, from our discussion of how institutionalization takes place in countries that closely resemble VOC idealtypes, compared to countries that are more in-between, we were able to discern different types of emulation. We suggest that: (1) it might be useful to distinguish types of emulation (auto-emulation, international emulation, reverse-emulation) and (2) the type of emulation can corroborate with the proximity of a nation to a particular VOC idealtype. We believe further research would be useful into how distinct phenomena or issues involving private sector employers, government agencies and civil society actors have institutionalized across countries.

Our discussion also led to a questioning of the institutional status quo when it comes to whistleblowing. Our findings, in line with our VOC hypotheses, suggest that institutionalization tends to neutralize whistleblowing. On that basis, we developed a positional matrix for campaigners, indicating where the status quo will lead in terms of centralization of handling roles and civil society engagement and which of the two dimensions needs to be strengthened to obtain whistleblowing institutions that are capable to bring about change.

Given that the new EU Directive on whistleblowing triggered institutionalization of whistleblowing in EU Member States and that these countries diverge on their VOC characterization, we believe this chapter offers both a framework for research as well as for practice.

Our findings are however subject to limitations of the research. First, we used a small sample (n = 7) because we used extant measures and observations. This small sample size poses a limitation to the validity of our exploration of the model. Further research could collect primary data on whistleblowing institutions in more countries for which VOC characterizations exist. A second limitation is that we used two VOC idealtypes (LME and CME) whereas there might be other idealtypes. Where these have measures for certain institutional spheres of the political economy, further research could expand our hypotheses. Third, institutionalization of whistleblowing is quite recent. The observations for a number of countries in our sample had been made during, shortly after, or even with announced institutional reform. The implication is that what currently looks like reverse-emulation could end up looking very differently.