JOURNAL ANALYSIS

Vol.:(0123456789)1 3

J Bus Ethics (2019) 159:201–219 https://doi.org/10.1007/s10551-017-3727-8

ORIGINAL PAPER

Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

Wim Vandekerckhove1 · Arron Phillips1 

Received: 23 May 2016 / Accepted: 27 October 2017 / Published online: 4 November 2017 © Springer Science+Business Media B.V., part of Springer Nature 2017

are perceived as viable and also what the initial responses are in terms of retaliation and effectiveness.

Keywords Effective whistleblowing · External whistleblowing · Internal whistleblowing · Power · Retaliation

Introduction

This paper uses secondary data from an independent whistleblower advice line in the UK—Public Concern at Work (PCAW) to explore whistleblowing as a protracted process. Previous whistleblowing research has examined whistleblowing mainly as a one-off rational decision. The suggestion from Rothschild and Miethe (1999) that whistle- blowing is a sequence of attempts someone makes to raise a concern, and that people find themselves to be inadvertently turned into whistleblowers by management responses, has not been followed up by a systematic field study tracking the path of how a concern is repeatedly raised by whistleblow- ers, and identifying patterns of responses and effectiveness of each recipient along that path. We deem that to do so is crucial if research wants to inform the various actors that occupy themselves with making whistleblowing more suc- cessful—from the lawyers that represent whistleblowers or organisations, whistleblower support groups, unions, conflict mediators, HR departments, compliance, and ethics officers, to regulators and legislators—about patterns in whistleblow- ing processes. All of these actors would benefit from insights into through what sequences or ‘routes’ whistleblowers tend to escalate their concern, which of these routes (rather than any single recipient) make whistleblowing more effective in stopping wrongdoing, and which of these routes are less harmful for whistleblowers. Thus, our exploration in this

Abstract This paper provides an exploration of whistle- blowing as a protracted process, using secondary data from 868 cases from a whistleblower advice line in the UK. Pre- vious research on whistleblowing has mainly studied this phenomenon as a one-off decision by someone perceiving wrongdoing within an organisation to raise a concern or to remain silent. Earlier suggestions that whistleblowing is a process and that people find themselves inadvertently turned into whistleblowers by management responses, have not been followed up by a systematic study tracking the path of how a concern is repeatedly raised by whistleblowers. This paper provides a quantitative exploration of whistleblow- ing as a protracted process, rather than a one-off decision. Our research finds that the whistleblowing process generally entails two or even three internal attempts to raise a concern before an external attempt is made, if it is made at all. We also find that it is necessary to distinguish further between different internal (e.g. line manager, higher management, specialist channels) as well as external whistleblowing recip- ients (e.g. regulators, professional bodies, journalists). Our findings suggest that whistleblowing is a protracted process and that this process is internally more protracted than previ- ously documented. The overall pattern is that whistleblowers tend to search for a more independent recipient at each suc- cessive attempt to raise their concern. Formal whistleblower power seems to determine which of the available recipients

* Wim Vandekerckhove [email protected]

Arron Phillips [email protected]

1 Business School, University of Greenwich, Park Row, London SE10 9LS, UK

202 W. Vandekerckhove, A. Phillips

1 3

paper of whistleblowing as a protracted process proceeds via the following research questions:

RQ1: If whistleblowing is a protracted process, what is the typical sequence of recipients? RQ2: At what point in the protracted process does whistleblowing become an effective way to stop wrong- doing? RQ3: How does retaliation change throughout the pro- cess?

We answer these research questions using a data set of 1000 whistleblower cases from the Public Concern at Work (PCAW) database. PCAW is an independent agency in the UK providing free advice to people who want to raise a con- cern about wrongdoing in their workplace. PCAW is not a regulator nor does it operate corporate hotlines. PCAW does not investigate concerns. It advises individuals on how to best raise their concern inside or outside of their workplace whilst remaining within the protection boundaries of the UK Public Interest Disclosure Act.

The contributions of this paper are twofold. The first is that our exploration of whistleblowing as a protracted process offers a more differentiated insight into the roles of those who receive whistleblower concerns. Rather than merely distinguishing internal from external whistleblow- ing, our research design provides knowledge of how whistle- blowing unfolds through a combination of different inter- nal recipients, and takes into account the internal process preceding external whistleblowing. We also distinguish regulators—mandated as prescribed persons under the UK whistleblowing legislation—from other external recipients (professional bodies, NGOs). We find some indications that rather than merely reverting to external after internal whistleblowing, whistleblowers might very well be search- ing for a more independent recipient at each successive attempt to raise their concern. A second contribution of this paper is that our exploration offers support for theorising power at the institutional level rather than the individual or organisational level to explain whistleblowing outcomes. More precisely, our findings make it plausible to theorise that it might not be formal whistleblower power that deter- mines one’s legitimacy and credibility, but rather that the formal power of the whistleblower influences which of the available recipients are perceived as viable. Our findings further suggest that both effective and safe whistleblowing might depend on the whistleblower’s ability to break the organisation’s control over the whistleblowing process.

The paper is structured as follows. We first review the whistleblowing literature in terms of how whistleblowing is conceptualised: a one-off decision versus a process. We then proceed in the section after with developing hypoth- eses around our research questions, which structure our

exploration. The section after that explains how we collected and analysed our data and what the uniqueness as well as the limitations of our data are. After that, we present our find- ings for each hypothesis. We then discuss these in a separate section, before concluding with a summary of our findings.

One‑Off Decision Versus Process

In the early 1970s, the term whistleblowing was coined as behaviour resulting from a moral choice, more precisely as

an act of a man or woman who, believing that the public interest overrides the interest of the organiza- tion he serves, blow the whistle that the organization is involved in corrupt, illegal, fraudulent or harmful activity (Nader et al. 1972: vii).

This definition presents whistleblowing as a cogni- tive, one-off ethical decision. Research into whistleblow- ing from an organisational behaviour perspective from the 1980s onwards not only maintained whistleblowing as a prosocial phenomenon, but also sustained the assump- tions of whistleblowing as a cognitive and one-off decision (Jones et al. 2014; Watts and Buckley 2015). The standard definition of whistleblowing today (Brown et al. 2014) is the prosocial definition of Near and Miceli (1985: 4) ‘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’. Dozier and Miceli (1985) constructed a prosocial organisational behaviour (POB) model that depicts whistleblowing as involving both egoistic and altruistic motives. Moreover, the POB model postulates that whistle- blowing behaviour is determined by interactions of stable personality characteristics and individuals’ perceptions of organisational situations. Dozier and Miceli (1985) see the whistleblowing decision process as a cognitive process con- sisting of the following steps (Jones et al. 2014): (1) aware- ness that an act is questionable, (2) decision to react, and (3) deciding what action to take. Miceli and Near (1992) added a fourth step: organisation members react to the whistle- blowing (often against the whistleblower). This begs the question to what extent the anticipated consequences peo- ple take into account when making the assumed cost–ben- efit decision whether or not to blow the whistle, match the actual responses from organisational members. However, this research question has had very little resonance amongst whistleblowing scholars.

Since 2000, whistleblowing research suggests the whistleblowing decision-making process might not be all that rational or cognitive as previously assumed. Gundlach et al. (2003) and Blenkinsopp and Edwards (2008) argued the decision is to a great extent influenced by emotional

203Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

responses. Once made, the decision nevertheless becomes cognitive through information processing and rationalisa- tion (Blenkinsopp and Edwards 2008; Gundlach et al. 2003; MacGregor and Stuebs 2014). MacGregor and Stuebs (2014) built a decision-making model for what they term ‘fallacious silence’, a situation in which individuals refrain from raising a concern about illegal or immoral issues that violate per- sonal, moral, or legal standards. Their model consists of a triangle of determinants including incentives, opportunities, and rationalisation possibilities. Whilst this model does not necessarily regard the decision to blow the whistle or remain silent as a cognitive process—rationalisation allows for post hoc cognitive justification of a predetermined decision—it still assumes that to blow the whistle or to remain silent is a one-off decision.

Whistleblowing research has also advanced by focus- ing on who whistleblowers raise their concern with. Near and Miceli’s (1985: 4) definition leaves this quite open: ‘to persons or organisations that may be able to effect action’. Research into whistleblowing recipients has predominantly used an internal/external dichotomy with little or no differ- entiation within these categories. Internal whistleblowing entails raising a concern about alleged wrongdoing to an authority within the organisation. External whistleblowing involves raising a concern to an outside agency, such as a professional organisation, a regulator, or the media. Sims and Keenan (1998) found external whistleblowing to be predicted by perceived supervisor support, perceived infor- mal policies, and ideal values about whistleblowing. Here again, whistleblowing is casted as a decision to do it either internally or externally, without considering the possibil- ity of doing one after the other. Kaptein (2011) measured the impact of organisational culture on how people raise a concern about wrongdoing, differentiating between inac- tion, confronting the wrongdoer, reporting to management, calling an internal hotline, and external whistleblowing. This study does not include a measurement of how people continue to raise the concern—e.g. does organisational cul- ture influence how one would raise a concern after confront- ing the wrongdoer, or after reporting to management? No doubt culture is important to the whistleblowing process, but the data used for the analysis in this paper do not allow to address the role of culture. Hence, we do not elaborate on organisational culture in this paper.

Vandekerckhove et al. (2014) define successful whistle- blowing as whistleblowing that is both effective in stopping the wrongdoing and safe for the whistleblower. Effectiveness of whistleblowing is 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. (Near and Miceli 1995: 681)

Indeed, if the primary motivation of raising a concern is to effect a change in practice, then the impact of rais- ing the concern on the perceived wrongdoing needs to be separated from responses to the whistleblower. Nevertheless, Near and Miceli’s (1995) proposition is that whistleblow- ing will be more effective if the concern is raised with a powerful recipient, but only if that recipient is supportive of the whistleblower. They do not formulate the proposition in terms of internal versus external recipients. Mesmer-Magnus and Viswesvaran (2005) noted a small negative correlation between whistleblowing and effectiveness. However, all cases in the sample were external whistleblowing. In con- trast, Dworkin and Baucus (1998) found external whistle- blowers to be more effective in triggering investigations and remedial actions. They analysed 33 legal cases of employees fired for whistleblowing, looking for patterns of internal or external whistleblowing that might determine effectiveness of the whistleblowing and retaliation against the whistle- blower. They found that internal whistleblowers were usually ineffective in stopping the wrongdoing, partly because they ‘seem to poorly select their complaint recipient’ (Dworkin and Baucus 1998: 1295) with most of them raising the con- cern directly with managers involved in the wrongdoing. External whistleblowers tended to be more effective in trig- gering investigations and remedial actions. In terms of retali- ation, internal whistleblowers tended to be dismissed imme- diately, while external whistleblowers were dismissed after a while. Interestingly, the sample used by Dworkin and Bau- cus (1998) also included a number of cases where whistle- blowers had gone external only after first raising their con- cern internally. However, while Dworkin and Baucus (1998) insist that ‘[we] limit our understanding of whistleblowing if we continue to treat internal and external whistleblowers as one group’ (p. 1296), their analysis does not acknowledge that the complexity of the whistleblowing process might be even greater when it is protracted, i.e. whistleblowers go external after raising their concern internally. Nevertheless, Dworkin and Baucus (1998) called for future research to identify patterns of retaliation and effectiveness in relation to specific internal and external recipients of whistleblowing. This paper provides such an analysis.

Smith (2014) argues that ‘we do not know which types of retaliation and suffering will prove more or less impor- tant for whistleblowers in different circumstances’ (p. 236). This was already acknowledged in early whistleblowing research (Miceli and Near 1989; Near and Miceli 1986), and the standard solution has been to give all types of retali- ation the same weight and add these up along with factors capturing the breadth of retaliation such as number of types of retaliation and number of types of organisation members involved in the retaliation (Miceli et al. 2008). Smith (2014) comments that the effect of this solution to the objective/ subjective suffering ‘has been to help direct attention away

204 W. Vandekerckhove, A. Phillips

1 3

from analysis of specific types of whistleblower suffering’ (p. 236). Hence, if the aim is to gain insight into specific types of retaliation and patterns of protracted retaliation, we should avoid aggregating different types of retaliation into a single index. Dworkin and Baucus (1998) found that internal and external whistleblowers differed in the type of retaliation they experienced. Internal whistleblowers tended to be dismissed more quickly that external whistleblowers.

To our knowledge, only five studies of protracted whistle- blowing exist. Rothschild and Miethe (1999) interviewed more than 200 whistleblowers in the USA. They find that ‘[many] came to their whistle-blowing almost by accident’ (Rothschild and Miethe 1999: 119). Employees raised their concern to their line manager and then to senior manage- ment, each time believing the recipient would step into cor- rect the situation. Only after voicing their concern inter- nally twice did they consider going to authorities outside the organisation. Whistleblowers almost never accurately anticipated the consequences of voicing their concern. Nevertheless, one of the most important findings of their research is that

[it] is management’s response that shapes the poten- tial whistle-blower’s subsequent actions. Specifically, our interviews revealed a common pattern in which management’s efforts to discredit or retaliate against the claimant become the major catalyst for the politi- cal transformation of the concerned employee into a “persistent resister” (Rothschild and Miethe 1999: 119–120).

The implication is that if whistleblowers make a cost–benefit decision prior to voicing a concern, they are bound to make that decision on mistaken terms. Moreover, they appear to be unaware that the decision they make is a whistleblowing decision. Rather, they voice their concern inadvertently, not knowing that what they do is whistleblow- ing. With regard to the ethicality of whistleblowing deci- sions or the post hoc rationalisations suggested in the litera- ture, Rothschild and Miethe (1999: 120) write that

[it] is by virtue of the organization’s retaliatory response that whistle-blowers become convinced of the moral correctness of their actions and strengthened in their conviction to fight—both to exonerate them- selves and to correct the organization’s wrong.

Donkin et al. (2008) used data from two surveys of Aus- tralian public sector workers in an attempt to track reporting paths. Unfortunately, the two surveys differed in the way they captured the sequence of reporting. One survey asked respondents to indicate with whom they raised their con- cern the first time, second time, etc. The other survey asked respondents to indicate with whom they raised first and list all subsequent recipients without indicating the sequence.

Findings show that 97% of whistleblowing starts as voicing a concern internally, and 90% remains internal. It is also clear that line managers and senior management are generally the first two ports of call, with the number of whistleblowing attempts averaging at 1.9 and 4.3 in the respective surveys (Donkin et al. 2008). It is possible that the research design did not anticipate finding whistleblowing to be so heav- ily protracted, and the research report is not able to show clear figures for each recipient linked with specific response patterns or their effectiveness as whistleblowing channel. Mansbach and Bachner (2010) use an experimental research design with vignettes to measure Israeli nurses’ willing- ness to raise a concern about possible harm to patients as a protracted process. They differentiated internal recipients into ‘wrongdoer’ and ‘a central ethics committee’. External recipients were ‘a professional body’ and the press. Par- ticipants expressed a higher likelihood to raise a concern internally than externally, and their willingness decreased at each step of escalation. Skivenes and Trygstad (2010) researched whistleblowing in the Norwegian public sector and distinguish between weak, strong internal, and strong external whistleblowing. The notion of strong whistleblow- ing puts an emphasis on process, as Skivenes and Trygstad define it as

where there is no improvement in, explanation for, or clarification of the reported misconduct from those who can do something about it. In such cases, an employee must report the misconduct again and is thus engaging in strong whistleblowing. (p. 1077)

The study found that 36% engaged in strong whistleblow- ing, with 29% raising the concern four times. The study uses a power model to hypothesise retaliation and effectiveness of weak and strong whistleblowing, but finds little support for the power model, and does not provide findings for specific internal and external recipients. Finally, Lewis and Vande- kerckhove (2015) use data from a survey of NHS Trusts in the UK, showing that internally more than half of the Trust employees raise their concern first with their line manager either informally (52.3%) or in writing (7.3%). The first external recipient of concerns is their trade union (38%), followed closely by professional body (35%). The study does not give further specification about internal or external paths. However, it does show that those who followed their organisation’s whistleblowing procedure were more likely to proceed to raise their concern externally when internal recipients showed to be ineffective or resulted in retaliation.

Hence, whilst some research suggests whistleblowing is a process of both real and anticipated responses to the rais- ing of a concern, the whistleblowing literature has so far mainly developed on the assumption that whistleblowing is a one-off decision. This paper aims to provide insights into whistleblowing as a protracted process, based on data

205Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

tracking the path of how concerns are raised, identifying patterns of responses, and effectiveness along that path. We aim to provide insights that are more detailed than merely distinguishing between internal and external whistleblow- ing. To facilitate our exploration of the data, we formulate hypotheses in the next section.

Hypotheses

This section develops hypotheses around the three research questions driving our exploration of whistleblowing as a pro- tracted process: what is the typical sequence of recipients, at what point in the process does whistleblowing become effective in stopping wrongdoing, and how does retaliation change throughout the process?

RQ1: If whistleblowing is a protracted process, what is the typical sequence of recipients?

In an experimental study, Park et al. (2008) found that people prefer to raise a concern inside their organisation. Mansbach and Bachner (2010), also in an experimental study, found that people’s willingness to raise a concern decreased with each further attempt, regardless of whether they were asked about their willingness to report internally or externally. Response options with internal recipients were preferred over those with an external recipient.

Hence, if whistleblowing is a protracted process, we expect to find that even though the number of people tails off as the process protracts, people tend to raise a concern more than once. Also, because people prefer to raise a concern inside the organisation, in a protracted process we expect people to raise a concern with an internal recipient before doing so with an external recipient.

H1: People raise a concern more than once.

H2: People raise a concern internally before doing so externally.

In an Australian field study, Donkin et al. (2008) found that line managers and higher management were by far the most used ‘ports of call’ for whistleblowers. This is in line with what Rothschild and Miethe (1999) found, namely that people inadvertently find themselves in a whistleblowing situation, the making of which is codetermined by how line managers and higher management react to workers who raise a concern. Donkin et al’s (2008) study was carried out with a sample of public sector workers. The sample in our research includes whistleblowers from the public, private, and not- for-profit sectors. However, there is no reason to anticipate any differences across different sectors. Hence in terms of

which internal recipients are most used by whistleblowers to raise a concern with, we expect to find the same in our UK sample as in Donkin et al’s (2008) Australian sample.

H3: Line managers and higher management are the most used recipients for internal whistleblowing.

Skivenes and Trygstad (2010) used Near and Miceli’s (1995) power model to examine whistleblowing in Norwe- gian organisations. This approach assumed that membership of a decision-making group—or being in communication with those who make decisions—is a source of power rele- vant to one’s ability to raise a concern, because such contacts entail opportunities not only to report problematic situations to important people in the organisation, but also to explain the concern and formulate arguments. The variables exam- ined by Skivenes and Trygstad (2010) include education, tenure, type of employment, hierarchical position, and union contact. Based on this, they expected that whistleblowers with more power would be more likely to continue to raise their concern. They found that only union contact made it more likely that people would raise their concern outside of line management but internal to the organisation. Moreover, a higher level of education correlated with a lesser likeli- hood to raise the concern to an external recipient. Thus we expect, following Near and Miceli’s (1995) power model and Skivenes and Trygstad (2010) findings of internal and external differences, that when whistleblowers escalate their concern, those with more power tend to remain inside, using channels outside of line management.

H4: Whistleblowers with more power are more likely to use different internal recipients before going outside.

RQ2: At what point in the protracted process does whistleblowing become an effective way to stop wrong- doing?

In the previous section, we mentioned how Near and Miceli’s (1995) model propositions that whistleblowing will be more effective if the concern is raised with a powerful recipient. They argue that internal recipients can be just as credible as external recipients. Regulators and independent bodies (police or professional bodies) are external agencies bestowed with the authority to safeguard ethical standards. The same holds for specialist channels such as internal audit or compliance as the internal safeguards of professional standards. The extant literature remains ambivalent on this. Miceli and Near (2002) test differences in effectiveness of internal and external whistleblowing in three field studies of US federal employees and North American auditors, but find no significant difference. Apaza and Chang (2011) ana- lysed two cases from Peru and South Korea and found that

206 W. Vandekerckhove, A. Phillips

1 3

external recipients and media support were crucial in mak- ing whistleblowing effective. Dworkin and Baucus (1998) find in their multi-case analysis that in contrast to internal whistleblowers, external whistleblowers ‘often triggered investigations, remedial actions or other changes by the organization’ (p. 1295). Dworkin and Baucus (1998) theo- rise that blowing the whistle externally puts the organisation under greater pressure to respond to them than with internal whistleblowing because an external recipient gives the con- cern legitimacy and organisations cannot simply ignore the concern. This then leads to more effective whistleblowing.

Hence, in a protracted process, where external whistle- blowing tends to happen after ineffective internal whistle- blowing, we would expect that only powerful external recipi- ents can make the escalated whistleblowing effective. We use a sample from the UK where regulators are mandated as ‘prescribed persons’ under the Public Interest Disclosure Act. Thus, we expect regulators to be more effective recipi- ents than other external recipients.

H5: External whistleblowing to a regulator leads to more effective whistleblowing.

RQ3: How does retaliation change throughout the pro- cess?

Smith (2014) cites O’Day (1974) who researched patterns of intimidation towards employees who suggest reforms and found a pattern of successive rituals that lead to the expul- sion of the whistleblower: nullification (denial that there is a problem) is followed by isolation (this would both be infor- mal retaliation and blocking resources in our coding); after that there is defamation of the whistleblower (this would be blocking resources and formal retaliation). These three ‘ritu- als’ aim to cause the whistleblower to quit, either by leaving the organisation or going on sick leave. O’Day (1974) writes that the final ritual is to force an ‘involuntary withdrawal’ by firing the whistleblower. Smith (2014) comments that the cases in Dworkin and Baucus (1998) were almost never subjected to O’Day’s phases, and that they presented cases where management fired whistleblowers immediately.

Mesmer-Magnus and Viswesvaran (2005) found that external whistleblowing correlates with more retaliation. This, however, does not take into regard whistleblowing as a protracted process nor does it distinguish between types of retaliation. Smith (2014) also finds more retaliation fol- lowing external whistleblowing but explains this differently depending on at what stage the whistleblower goes external. When the first recipient is an external one, the whistleblower suffers retaliation because ‘initial reporting through external avenues breaks with the “proper channels” set out in the law’ (p. 242). However, when the whistleblower goes external at a later stage, retaliation can ‘reflect the fact that by the time

whistleblowers report externally, internal processes for han- dling their reports have already gone badly wrong, so that further action from any source is unlikely to improve the situation’ (p. 243, emphasis in original). This leads Smith (2014) to suggest that more retaliation for external whistle- blowing at a later stage is merely a continuation of retali- ation following previous internal whistleblowing. Rather than seeing the internal/external position of the recipient as influencing the likelihood of dismissal, the argument is that it is the protractedness of the whistleblowing that makes dismissal more likely.

H6: Dismissal becomes more likely as the whistleblowing process protracts.

Method

Data Collection

Every time someone comes to PCAW for advice, a PCAW advisor takes notes on the nature of the concern and the unique situation of the whistleblower. This helps PCAW to give advice when the whistleblower calls back. All notes of a particular whistleblower together form a whistleblower case in the PCAW database. Hence, for each whistleblower PCAW advises, the PCAW database includes a narrative of their whistleblowing journey. For this paper, we use data from a wider research project that included a content analy- sis of 1000 of these journey narratives (PCAW 2013). The ‘narrator’ is not the whistleblower; it is the PCAW advisor who makes these notes. Some information is registered in a structured way (formal position, industry, type of wrong- doing). Most of these advisor notes are made in an attempt to get a chronological account of what has happened to the whistleblower. In this sense, the notes can be regarded as resulting from a semi-structured interview between the advice giver and the whistleblower.

This research has been approved by the University of Greenwich Research Ethics Committee (Ref 11/12.3.5.21). To avoid using case files from ongoing cases, we started with entries in the PCAW database from December 2010 and worked backwards in time until 1000 journey narratives had been coded. Thus, the sample consists of 1000 cases from the PCAW advice line between 30 December 2010 and 20 August 2009. We only included entries where the contact with the whistleblower was through the phone. We excluded entries where the call for advice came from those other than the whistleblower. We further excluded entries where there was no information on the type of wrongdoing or industry sector of the organisation the whistleblower was working for.

At the start of the project, two researchers independently coded the same 90 narratives. These 90 were first cleared of

207Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

any content that would allow the identification of the organi- sation or individuals, by PCAW to ensure confidentiality in relation to the users of the advice line. The researchers first coded 20 and then 30 narratives to develop the code book. A further 10 narratives were double-coded at three subse- quent instances to get a shared understanding of the coding categories and to ensure consistent coding. At each instance, differences in coding would be discussed and clarified. A shared understanding was reached after the third session. A final double coding of ten random narratives did not reveal any differences.

All variables were nominal-level variables, except the timing of the contact with PCAW (after how many attempts to raise a concern did the whistleblower contact the PCAW advice line). We coded variables relating to industry, type of wrongdoing, frequency and duration of the wrongdo- ing, formal position of the whistleblower and whether the wrongdoer’s hierarchical position was higher, equal to, or lower than that of the whistleblower. We also coded vari- ables relating to the first four times someone had tried to raise a concern: with whom they had raised their concern, anticipated and actual responses from management and co- workers, whether the wrongdoing had been investigated or stopped, and what happened to the wrongdoer.

Sample

It is important to point out this research is based on second- ary data. The narratives were written by PCAW staff for the purpose of giving advice, not for research purposes. The implication is that for most variables not every case included data. Another implication is that we do not have informa- tion relating to gender or age of the whistleblowers in our sample, as PCAW does not record this data.

The top six industries advice seekers to PCAW worked in were: health (14.8%), care (14.3%), education (11.1%), char- ities (8.9%), local government (6.9%), and financial services (6.7%). According to Census 2011 data, 13% of employed people in England and Wales work within the health and social work sector, 10% work in education, 6% work in pub- lic administration, and 4% in financial services. In our sam- ple, callers working in the health and care sectors add up to 29%, compared with the 13% in the Census data. This means that those working in caring environments, where patients, vulnerable people, or children are involved, tend to contact the advice line more than workers in other sectors. Perhaps this is due to the nature of their work, media attention for NHS whistleblowers, policy agendas, scandals or awareness of PCAW. Other industries included in our sample were: retail (4.4%), manufacturing (3.3%), food/beverage (2.9%), transport (2.7%), construction (2.1%), leisure/hospitality (2.1%), utilities (1.5%), science/technology (1.2%), and a number of industries had a very marginal (< 1%) presence in

our sample (including insurance, central government, police, housing, legal services, pharmaceutical, quango, agriculture, forestry, armed services, mining, oil, media).

The top five concerns in the sample (n = 1000) relate to the following types of wrongdoing: ethical (19%), financial malpractice (19%), work safety (16%), public safety (11%), and patient safety (8%). Other concerns are about environ- ment, discrimination/harassment, or consumer and competi- tion regulation. Ethical concerns include abuse of position, cronyism and breach of policy, confidentiality, manipulation of scientific research, etc. Financial malpractice includes accounting fraud, corruption, bribery, and attempts to cover these up. Work safety concerns usually relate to health and safety in the workplace principally affecting employees. This includes unsafe machinery, no appropriate safety equip- ment, etc. Public safety is where the safety of the public is at stake and can also include the safety of the employees. This includes faulty wiring on a train track, unsafe meat in a supermarket, etc. Patient safety includes malpractices where patients are or might be harmed, e.g. lack of qualified personnel, breach of procedures to distribute medication in a hospital, physical or psychological harm to patients, etc.

Wrongdoing might be a one-off incident or re-occurring malpractice, e.g. one incident of abuse in care or ongoing neglect. In 86% of the cases, someone raised a concern about a re-occurring wrongdoing. In 43% of the cases, the duration of the wrongdoing was less than 6 months. In 33%, this was between 6 months and 2 years. The duration of the wrong- doing did not necessarily imply that the whistleblower had been raising the concern for that duration.

In terms of the whistleblower’s formal power in the organisation, our sample ranges from unskilled workers at the bottom (13%, e.g. carers, support workers, bartenders) followed by administrative/clerical positions (8%, e.g. office administrators, secretaries, advisers), skilled workers (27%, e.g. brokers, chefs, engineers), professionals (26%, e.g. nurses, doctors, teachers, accountants), managers (15%, e.g. line managers, general managers) and executives at the top of the organisational structure (2%). Others (9%) are scien- tists and researchers. Given the wide variety of organisations and sectors people in our sample worked for, and the often inflated use of job titles, we coded the whistleblowers in one of the above categories based on a combination of their posi- tion and authority. It is important to note that even an execu- tive can be faced with a wrongdoer who is equally powerful than they are—imagine a COO who becomes aware of a CFO’s wrongdoing. The notion executive means part of the top management team. Perhaps the most famous example of an ‘executive whistleblowers’ is Michael Woodford who as the CEO of Olympus blew the whistle about payments to obscure companies in the Cayman Islands (Woodford 2012). Of course, he had raised a concern about these payments vis-a-vis his executive team, and after that with the board.

208 W. Vandekerckhove, A. Phillips

1 3

He was dismissed from his role of CEO, but kept his seat on the board. He nevertheless raised his concern with the shareholders, some of whom demanded an investigation.

In the majority of instances in our sample, the wrongdoer has a hierarchically higher position than the whistleblower in the organisation (72%). In 24%, the wrongdoer has an equal position, and in 4% a lower position.

We also coded, for the first four attempts a whistleblower makes to raise a concern, with whom they raised their con- cern, what the response from management was, and what happened to the wrongdoing. We clarify and discuss our nominal codings for these variables when we present find- ings from our exploration in the paper.

Limitations

A further reflection with regard to our data and methods of analysis is required here. Our data set is unique, but at the same time suffers from limitations. As mentioned in the previous section, we used notes advisors had made during interactions with whistleblowers. For each case, the notes unravelled the chronology of events: a sequence of who whistleblowers raised their concern with, how management had reacted both towards them as well as towards the wrong- doing at each instance, etc. This is excellent data to docu- ment whistleblowing processes. Most of the case narratives are the result of more than one interaction, as 44% of our sample had called the advice line twice, 10% three times, another 10% four times, and 4% more than four times. We know of no other whistleblowing research using this type of data.

Despite its uniqueness and suitability to study the whistleblowing process, our data have serious limitations. First of all, one implication of using secondary data was that we could not code every variable for every case. In many cases, the advisor notes did not include informa- tion needed to obtain codes for certain variables. Obvi- ously, missing values restrict the possibilities of analys- ing the data. For example, if a whistleblower did not call the helpline back, we did not have data on the effect of their whistleblowing in terms of effectiveness or retalia- tion. Thus, some of the analyses are carried out on small subsamples. We report the n cases where we had data for each crosstab and also how we arrived at a measure for significance. Our decision to use nominal variables no doubt contributes to the limitation in techniques we could use to analyse the data, e.g. our data and the way we code it does not allow multivariate analysis. However, we believe maintaining nominal values rather than ordinal ones or collapsing too many variances into few categories, provides a view of what precisely is neglected in previous quantitative whistleblowing research. This has of course

implications for the generalisability of our findings. On the one hand, we believe that what we find questions the generalisability of other whistleblowing research assum- ing whistleblowing as a one-off event, or theorising the effectiveness based on the legitimacy of the whistleblower within the organisation. On the other hand, we use a data set and analysis that make it difficult to generalise our own findings. Therefore, our claim can only be a modest one. The claim we make with our research is that our explora- tion of whistleblowing as a process provides findings that call for more distinction between whistleblowing variables in relation to timing, i.e. it is a call for more attention in whistleblowing research for the protracted process.

A second limitation of our research is that our sam- ple might not be representative of all whistleblowing in the UK. Rather, our sample is very likely to be biased towards unsuccessful whistleblowing, as there would be no reason to call an advice line if raising a concern had been effective in stopping the alleged wrongdoing without resulting in retaliation against the whistleblower. How- ever, we believe this bias is less severe than relying on legal case documents (e.g. Dworkin and Baucus 1998) as those taking their case to court is a subset of unsuccess- ful whistleblowing. We also believe that the bias towards unsuccessful whistleblowing in our sample is less severe than when a sample is constructed based on media reports, as this tends to only use high-profile cases (e.g. Weiskopf and Tobias-Miersch 2016). We suspect the bias in our sample to be similar to that in the sample of Rothschild and Miethe (1999) who used a list of 1000 people asking a whistleblower support group for advice. A difference between their and our sample is that ours is not self-selec- tive. Rothschild and Miethe (1999) obtained a final sample of 293 whistleblowers who had self-selected to participate by either filling in a survey or agreeing to be interviewed over the phone. Our final sample consisted of 868 people who had contacted a whistleblower advice line.

A third limitation of our data set relates to possibly important variables we did not code. For example, we were unable to code gender and age of the whistleblowers, as this information is not withheld by PCAW. Also, we did not code severity of wrongdoing. The whistleblowing lit- erature suggests type and severity (or intensity) of wrong- doing as an independent variable. However, we believed that calling an advice line implies that the whistleblower perceives the severeness of the wrongdoing beyond a cer- tain threshold. Finally, we were unable to code mistaken whistleblowing. People might have been raising a con- cern holding a reasonable yet mistaken belief that a certain practice constituted wrongdoing. It is of course impossible for researchers to investigate the correctness of whistle- blower perceptions.

209Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

Findings

RQ1: If whistleblowing is a protracted process, what is the typical sequence of recipients?

From our total sample of 1000 advice line cases, in 132 cases people expressed only an intention to raise a concern. Whilst we also coded intentions to raise with a specific recipient further on in the whistleblowing process, for the analyses presented in this paper we only use instances where the concern was actually raised. This resulted in a sample of n = 868. Because we rely on secondary data from notes made for advising whistleblowers when they called PCAW, in reality the whistleblower process might have been further protracted than what we had data for, i.e. a whistleblower might continue to raise their concern without calling the PCAW advice line. Hence, in 868 cases a concern had been raised at least once, in 484 at least twice, in 142 at least three times, and in 22 cases the concern had been raised four times. This means that in 55.8% a concern is raised more than once. This confirms H1 (People raise a concern more than once).

Compared to Donkin et al. (2008) who found the average length of the protracted whistleblowing process to be 1.9 and 4.3, the average number of times a concern was raised in our sample was 1.74. However, we only coded the first four instances in which a concern was raised.

When developing the codebook, it was clear that whistle- blowers can raise concerns either within the organisation (internal), outside of the organisation (external) or with their union if they are a member. We further distinguished within the internal category: wrongdoer, line manager, higher man- ager, grievance (where the individual raises their concerns together usually with personal issues via the formal griev- ance procedure), and specialist channel (audit, compliance, hotline). In the external category, we distinguished between regulator, other independent bodies (police, MP, NGO), and media. We coded ‘union’ as a separate recipient, following Lewis and Vandekerckhove (2016) who assert that unions are an exceptional recipient with both internal as well as external presence. We did not code whether the whistle- blower was a union member, although we did come across cases where the whistleblower mentioned joining a union to seek help with raising a concern. We coded the recipient as ‘other’ where the whistleblower raised a concern to someone who falls outside of the above categories, for example to customers or funders. This also covers the situation where they have spoken to a colleague about their concern (not typically raising a concern). Table 1 gives an overview of internal, external, and union recipients in the whistleblowing process of our cases, excluding ‘other’.

Whilst we found a substantial decrease in the number of whistleblowers who go on to raise a third time, it must be

noted that of those who raise a concern a third time about 60% still do so with an internal recipient. Even at the fourth instance, an equal number of recipients do so internally or to a union (45.5 and 4.5%, respectively) as with an external recipient (50%). Hence, our findings confirm H2 (People raise a concern internally before doing so externally).

Our findings give some depth to the previously existing (and mainly experimental) knowledge about internal versus external whistleblowing. Not only do whistleblowers tend to raise their concern internally before they do so externally, but they tend to raise internally more than once before going external, if they go external at all. In our sample of 868 cases, a concern was raised 1516 times, 80.7% of which was internal, 16.6% external, and 2.7% to a union representa- tive. Whilst the number of those making external disclo- sures increases as the whistleblowing process becomes more protracted, it never surpasses the amount of those making internal disclosures.

Figure 1 shows a series of frequencies of who whistle- blowers raised their concern with at each attempt. This offers a more detailed overview of the different recipients used by whistleblowers at each instance of raising their concern. It shows that 52% of whistleblowers first raised their concern with their line manager and 22% with higher management. When whistleblowers raise their concern a second time, 14% do this with their line manager and 33% with higher management. Taken together, line manage- ment and higher management are the initial recipients of 74% of whistleblower concerns, and second-attempt recip- ients of 47% of whistleblower concerns. This confirms H3 (line managers and higher management are the most used recipients for internal whistleblowing) and is in line with Donkin et al. (2008), although our UK data allow to give a more detailed picture of which recipients are used by whistleblowers the first four times they raise their con- cern. Only 7% raised their concern initially directly with the wrongdoer, whilst more than half of the whistleblow- ers in our sample (52%) first raised their concern with their line manager. More than one in five (22%) do so with higher management in the first instance, but a higher proportion (33%) goes to higher management when rais- ing their concern a second time. Other recipients showing

Table 1 Internal and external whistleblowing

Internal External Union Total

First attempt 778 (89.6%) 75 (8.6%) 15 (1.7%) 868 (100.0%) Second

attempt 350 (72.3%) 115 (23.8%) 19 (3.9%) 484 (100.0%

Third attempt 85 (59.9%) 51 (35.9%) 6 (4.2%) 142 (100.0%) Fourth

attempt 10 (45.5%) 11 (50.0%) 1 (4.5%) 22 (100.0%)

210 W. Vandekerckhove, A. Phillips

1 3

an increase at the second instance are specialist channels, regulators, external bodies, and grievances. Their usage increases even more at the third instance. Figure 1 also shows that by the third and fourth attempts whistleblow- ers are most likely to pursue the matter via a grievance procedure. It must be noted that a grievance procedure is

unsuited to a whistleblowing concern. A grievance proce- dure is a complaint procedure and hence places the onus on the individual to prove the complaint. A whistleblower, however, is a witness, passing on information to those with a responsibility to address the problem. A witness should not have to prove their concern.

7

52

22

4 41

14

33

10 14

1 4

16 14

25

0 0 9 9

27

0

15

30

45

60

WD LM HM SP GR

1st a�empt (n=868) 2nd a�empt (n= 484) 3rd a�empt (n= 142) 4th a�empt (n= 22)

2 4 2 1 24

11 11

1 14

20

13

2 15

18 18

9 5

0

15

30

45

60

U REG EXT M OT

Fig. 1 Recipients of concerns throughout the whistleblowing process (rounded %). WD wrongdoer, LM line manager, HM higher management, U union, SP specialist channel, REG regulator, EXT external recipient (professional body, NGO), GR grievance procedure, M media, OT other

211Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

In order to identify the most used routes amongst the vari- ous recipients, we sorted the sequences according to with whom the concern was last raised and then by length of the sequence. Concerns raised with specialist channels, regula- tors, external agencies, and grievances tend to have longer histories of raising with other recipients. We then looked for patterns within the sequences of recipients used by whistle- blowers to raise a concern. We disregarded sequences where only one attempt to raise a concern was made. A sequence of two or more attempts that occurred at least three times for each final recipient was regarded as a salient pattern. There were no patterns in the sequences of four attempts. Table 2 gives an overview of the salient patterns. These pat- terns further confirm the role of management as recipients of whistleblower concerns. In more than half of the cases, the whistleblowing process ended with either the line manager (27.8%) or higher management (25.2%). The patterns also

indicate that other recipients tend to receive concerns after the line manager or higher management had not been able to prevent escalation of the concern.

For example, salient patterns where a specialist channel was the final recipient tend to involve line manager, higher management or both as previous recipients. This finding has an important implication for organisational actors operating such a specialist channel—e.g. audit, compliance, hotline. These channels are often named in organisational whistle- blowing or speak-up policies and nearly always guarantee to the whistleblower that their identity will be kept confidential. Our data suggests that the majority of those using a special- ist channel have already raised their concern with someone else before using the specialist channel. The implication is that in most cases, either the whistleblower’s line manager, higher management or both know who has raised the con- cern that audit or compliance might investigate. Specialist

Table 2 Salient patterns in whistleblowing sequences

WD wrongdoer, LM line manager, HM higher management, U union, SP specialist channel, REG regulator, EXT external recipient (professional body, NGO), GR grievance procedure, M media, OT other

Salient sequences (# occurence) Rounded  % of cases with that final recipient

LM (188) 78 Cases with LM as final recipient n = 241 (27.8% of all cases)LM-LM (18) 7

WD-LM (12) 5 HM-LM (9) 4 HM (93) 42 Cases with HM as final recipient

n = 219 (25.2% of all cases)LM-HM (64) 29 WD-HM (13) 6 SP-HM (10) 5 LM-HM-HM (5) 2 LM-LM-HM (3) 1 HM-SP-HM (3) 1 HM-GR (24) 23 Cases with GR as final recipient

n = 106 (12.2% of all cases)LM-GR (19) 18 GR (19) 18 LM-HM-GR (8) 8 LM-EXT-GR (5) 5 SP-GR (4) 4 LM-REG (27) 29 Cases with REG as final recipient

n = 93 (10.7% of all cases)REG (18) 19 HM-REG (12) 13 LM-HM-REG (5) 5 LM-SP (19) 29 Cases with SP as final recipient

n = 66 (7.6% of all cases)SP (12) 18 LM-HM-SP (10) 15 HM-SP (8) 12 LM-EXT (16) 25 Cases with EXT as final recipient

n = 65 (7.5% of all cases)HM-EXT (11) 17 EXT (11) 17 LM-HM-EXT (11) 6 REG-EXT (3) 5

212 W. Vandekerckhove, A. Phillips

1 3

channels as recipients of whistleblower concerns might act wisely to make whistleblowers aware of this risk, and in any case carry out investigations assuming others will know who the whistleblower is.

Other patterns also underline the huge role of manage- ment’s responsiveness (Vandekerckhove et al 2014). Where a regulator was the final recipient, the emerging patterns also involved line managers and higher management. Similar patterns emerge where the final recipient was an external agency other than the regulator. A substantial amount of our whistleblower cases ended in whistleblowers opting for a grievance route, meaning they raise their concern through a formal grievance procedure in the context of retaliation they suffer. These also tend to be protracted and salient sequences involved higher management, line managers, or both.

We also analysed our data to see whether the whistle- blower’s formal power in the organisation made a difference with regard to whom they raised their concern with (H4). We first cross-tabulated for each of the four attempts, the whistleblower’s formal power within the organisation with the recipient they raised their concern with. In order to get some idea of the statistical significance, we created dummy variables for each of the formal power categories. For example, in the dummy variable for ‘unskilled’ we coded unskilled worker as 1 and the others as 0, in the dummy vari- able ‘skilled’ we coded skilled worker as 1 and the others as 0, etc. We ran crosstabs for these with recipients for the first three attempts. Table 3 gives an overview of the cross- tabs that showed to be significant at the first whistleblowing

attempt. Table 4 gives an overview of the significant cross- tabs at the second whistleblowing attempt.

Table 3 shows unskilled workers, skilled workers, man- agers, and executives are significantly different from each other in who they initially raise their concern with. At the second attempt, unskilled and professional are significantly different from the others. Hence, it seems that formal power of the whistleblower influences how the process begins, but not so much how it protracts.

In order to test H4, we also collapsed the recipient vari- able into internal recipients (line manager, higher manage- ment, wrongdoer, specialist channel, grievance), external recipients (regulator, external, media, other), and union. We ran crosstabs for the formal whistleblower power dummy variables with the collapsed recipients for the first three attempts. The collapsed recipients internal/external/union did not show any significant differences across the formal whistleblower power categories. However, when recipients are not collapsed into the categories internal/external/union, we do see significant differences for formal power (Tables 3, 4). This leads us to reject H4 (Whistleblowers with more power are more likely to use different internal recipients before going outside). Formal whistleblower power is related to how the whistleblowing process starts, but not in terms of internal versus external.

Our findings show that lower down the ladder the worker was less likely to approach higher management; this is in direct contrast with those in more senior positions: unskilled 14.2%, skilled 12.8%, management 37.6%. Unskilled

Table 3 Crosstab formal whistleblower power per first attempt recipient

Note recipients: 1 = wrongdoer; 2 = line manager; 3 = higher management; 4 = union; 5 = specialist channel (audit, compliance, HR); 6 = reg- ulator; 7 = external body (NGO, professional body); 8 = through grievance procedure; 9 = media; 10 = other

Formal whistleblower power

n Recipient Crosstab sig (per dummy)

1 (%) 2 (%) 3 (%) 4 (%) 5 (%) 6 (%) 7 (%) 8 (%) 9 (%) 10 (%) % X2 cells with count < 5

Unskilled 113 11.5 56.6 14.2 1.8 0.0 7.1 2.7 3.5 0.9 1.8 100.0 0.025 8 (40%) Skilled 235 6.4 64.3 12.8 1.3 2.1 3.8 2.6 4.3 0.4 2.1 100.0 0.001 4 (20%) Manager 125 7.2 36.8 37.6 2.4 6.4 3.2 1.6 2.4 0.0 2.4 100.0 0.002 6 (30%) Executives 22 9.1 9.1 45.5 0.0 13.6 4.5 9.1 0.0 9.1 0.0 100.0 0.000 10 (50%) All 868 7.0 51.6 22.5 1.7 4.3 3.6 2.5 4.3 0.5 2.1 100.0

Table 4 Crosstab formal whistleblower power per second attempt recipient

Note recipients: 1 = wrongdoer; 2 = line manager; 3 = higher management; 4 = union; 5 = specialist channel (audit, compliance, HR); 6 = reg- ulator; 7 = external body (NGO, professional body); 8 = through grievance procedure; 9 = media; 10 = other

Formal whistleblower power

n Recipient Crosstab sig (per dummy)

1 (%) 2 (%) 3 (%) 4 (%) 5 (%) 6 (%) 7 (%) 8 (%) 9 (%) 10 (%) % X2 cells with count < 5

Unskilled 63 1.6 25.4 25.4 3.2 1.6 14.3 17.5 9.5 0.0 1.6 100.0 0.036 5 (25%) Professional 144 1.4 11.8 36.8 4.9 15.3 9.0 8.3 9.7 0.0 2.8 100.0 0.039 6 (30%) All 484 1.4 14.0 32.9 3.9 9.9 11.2 11.0 14.0 0.4 1.2 100.0

213Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

workers were more likely than other workers to approach the wrongdoer in the first instance (11.5%) and also to make a disclosure to external recipients other than regulators as a second step (17.5%). Specialist channels are not on blue collar workers’ radar, and it seems as though this key group of workers did not explore internal options aside from their line manager. Those in more senior positions such as profes- sionals, management, and executives, were more likely to raise concerns through a specialist channel at any step. This group is less likely to raise concerns with a regulator but more likely to go to the media at an early stage.

RQ2: At what point in the protracted process does whistleblowing become an effective way to stop wrong- doing?

In our data, we coded how the whistleblower perceived the impact of raising the concern on what the organisation was doing about the wrongdoing. Coding categories were: ‘nothing is done’ (where the whistleblower expresses that no action has been taken, for example where the whistleblower has been ignored or the wrongdoing is denied), ‘investigat- ing but low expectations’ (where the recipient of the infor- mation is conducting an investigation but the whistleblower lacks confidence in the investigation, for example where the investigation is being headed by someone the whistle- blower does not trust or fear that there will be a whitewash), ‘investigating with high expectations’ (as previous but where the whistleblower has confidence in the investigation), and ‘stopped’ (where the wrongdoing is stopped).

We initially did a cross-tabulation of how the whistle- blower perceived the impact of raising a concern, per recipi- ent and attempt. In order to get an idea about significance, we created dummy variables for the first three attempts with collapsed nominal categories for change to wrongdoing: ‘nothing is done’ and ‘investigating (low expectations)’ were coded 0, ‘investigating (high expectations)’ and ‘stopped’ were coded as 1. We then reran the crosstabs with the col- lapsed variables. Table 5 shows the results of crosstabs for the first and second attempts. These indicate that the recipi- ent does make a difference in terms of how effective whistle- blowers perceive their whistleblowing to be, at least at the first and second attempts.

Overall, the most common response is that nothing is done about the concern. The data also suggests that per- sistent whistleblowers tend to perceive themselves as more effective. The percentage of whistleblowers who believed their whistleblowing triggered a serious investigation increases from only 16.1% at the first attempt, to 24.7% at the second.

Three recipients appear to be more effective than others: specialist channels, regulators, and external independent bodies. However, although whistleblowing to a specialist channel appears to trigger more investigations, in at least half of these instances the whistleblower did not have trust in the seriousness of the investigation.

Table  6 shows a crosstab detail for regulators and for other external recipients. We see two salient points. The first is that regulators seem less likely to neglect a whistle- blower concern if it has been raised previously with another

Table 5 Crosstab recipient per perceived change to wrongdoing after first and second attempt

Note perceived change to wrongdoing: negative  =  nothing is done, or low expectations of investigation; positive = high expectations of investigation, or wrongdoing stopped Note recipients: 1  =  wrongdoer; 2  =  line manager; 3  =  higher management; 4  =  union; 5  =  specialist channel (audit, compliance, HR); 6 = regulator; 7 = external body (NGO, professional body); 8 = through grievance procedure; 9 = media; 10 = other

Recipient Perceived change to wrongdoing after first attempt

Perceived change to wrongdoing after second attempt

Negative (%) Positive (%) n Total (%) Negative (%) Positive (%) n Total (%)

1 91.7 8.3 36 100.0 100.0 0.0 4 100.0 2 89.1 10.9 349 100.0 86.0 14.0 50 100.0 3 79.7 20.3 153 100.0 86.7 13.3 113 100.0 4 100.0 0.0 6 100.0 66.7 33.3 3 100.0 5 61.9 38.1 21 100.0 69.2 30.8 26 100.0 6 47.4 52.6 19 100.0 40.6 59.4 32 100.0 7 35.7 64.3 14 100.0 46.9 53.1 32 100.0 8 92.3 7.7 26 100.0 87.0 13.0 23 100.0 9 66.7 33.3 3 100.0 100.0 0.0 1 100.0 10 91.7 8.3 12 100.0 75.0 25.0 4 100.0 Total 83.9 16.1 639 100.0 75.3 24.7 288 100.0 Crosstab sig X2 = 0.000

8 cells (40%) with count < 5 X2 = 0.000 8 cells (40%) with count < 5

214 W. Vandekerckhove, A. Phillips

1 3

recipient: cf. in Table 6 ‘nothing is done’ shows a downward trend (attempt 1, 2, and 3) for ‘regulator’ and an upward trend for ‘external’. The second salient aspect is that whistle- blowers seem to perceive regulators as more effective later in the whistleblowing process, and other external bodies less effective: cf. in Table 6 ‘investigating (high expect)’ and ‘stopped’ taken together shows an upward trend (attempt 1, 2, and 3) for ‘regulator’ and a downward trend for ‘exter- nal’. Hence, external recipients seem to lead to more effec- tive whistleblowing, and regulators even more so than other external recipients. This confirms H5 (External whistleblow- ing to a regulator leads to more effective whistleblowing).

RQ3: How does retaliation change throughout the pro- cess?

In only 38.1% of our case narratives (331 out of 868 cases) did whistleblowers mention a response from management, suggesting that in 61.9% of cases the

management response was not of significance to the whistleblower when they contacted PCAW for advice. Whilst we must assume that our sample is biased towards unsuccessful whistleblowing—a whistleblower would not contact PCAW when their whistleblowing was success- ful—our finding that in only 38.1% of our sample were the responses from managers towards whistleblowers salient for the latter suggests that in the other cases whistleblow- ers contacted PCAW mainly because their whistleblowing had so far been ineffective.

We noted in our sample that there were a number of common responses from management that whistleblowers may experience. These were coded as follows: ‘informal’ (closer monitoring, ostracised, verbal harassment), ‘blocking resources’ (blocking access to emails, information, training, hours), ‘formal’ (relocation, demotion, job reassigned, sus- pended, disciplined), ‘dismissal’ (being fired), and ‘support’ (the recipient is taking the issue seriously or has expressed support). Figure 2 shows a bar chart of relative frequency of

Table 6 Crosstab perceived changes to wrongdoing per attempt (regulator and external bodies)

Recipient Attempt n Perceived change to wrongdoing (%) Total (%)

Nothing is done Investigating (low expect)

Investigating (high expect)

Stopped

Regulator 1 19 42.1 5.3 52.6 – 100.0 2 32 31.3 9.4 53.1 6.3 100.0 3 18 27.8 5.6 66.7 – 100.0

External 1 14 28.6 7.1 50.0 14.3 100.0 2 32 34.4 12.5 46.9 6.3 100.0 3 11 45.5 9.1 36.4 9.1 100.0

20

13

31 29

7 10

13

37 37

37 5

39 44

50

12

44 44

0

informal blocking res formal dismissal support

first a�empt (n=331) second a�empt (n=213) third a�empt (n=57) fourth a�empt (n=9)

Fig. 2 Retaliation throughout the whistleblowing process (rounded %)

215Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

different responses whistleblowers experienced from their managers throughout the whistleblowing process.

Our finding is that both dismissal and formal reprisals such as demotion, relocation, or reassigning job responsibili- ties are the most common type of retaliation, at each attempt. When we look at the bar chart in Fig. 2, we do see a pattern towards more formal and more dismissal as the whistleblow- ing process protracts. Hence, we can only partially confirm H6 (Dismissal becomes more likely as the whistleblowing process protracts); our data are inconclusive as to the differ- ences between formal retaliation and dismissal.

Although the overall pattern in our findings is in accord- ance with O’Day’s phases, there is a substantial proportion of whistleblowers in our cases who suffered formal retalia- tion or dismissal immediately. In order to get more insight into what the contingencies might be with regard to retalia- tion patterns, we present findings on recipients and formal whistleblower power separately.

Our findings suggest that in contrast to effectiveness, there is less reason to distinguish between different external recipients when it comes to retaliation. We did not arrive at convincing findings for maintaining the question whether blowing the whistle to a regulator might lead to different whistleblower experiences than blowing the whistle to a professional body or NGO. We created dummy variables to perform a Chi-square test, one comparing responses after whistleblowing to a regulator with other recipients (regulator coded as 1, others as 0), and one comparing responses after whistleblowing to an external body with other recipients (external body coded as 1, others coded as 0). We found only responses after blowing the whistle to an external body (not a regulator) at the third attempt to be significantly different from other recipients. People who raised their concern to an external body at the third attempt experienced less dismissal or formal retaliation. However, absolute numbers are very low (n = 58 for all cases with data for management response at third attempt), and we cannot see how to explain why the difference would only occur at the third attempt and not sooner.

The formal power of the whistleblower does seem to make a difference for how retaliation develops. We recoded responses from management as follows: ‘support’ as 1, ‘informal reprisals’, ‘blocking resources’, and ‘formal repris- als’ as 2, and ‘dismissal’ as 3. We ran crosstabs for formal whistleblower power with the collapsed variable. Chi-square tests showed only the crosstab for the first attempt to blow the whistle was significant. These are presented in Table 7 and indicate that initial responses relate with formal power of the whistleblower. Executives received more support, and unskilled workers experienced most retaliation short of dis- missal. Managers were least likely to receive support, and managers and administrative staff were most likely to get fired at their first attempt to raise their concern.

Discussion

Protracted Whistleblowing

Our findings, especially the support for H1, H2, and H3 sup- port whistleblowing as a protracted process. The increased use of higher management in the second instance, and spe- cialist channel, regulator, and external bodies in the second and third instances suggests whistleblowers seek to raise their concern with increasingly independent recipients, rather than merely decide to blow the whistle externally if an internal channel is unsuccessful. Our exploration of the data with regard to H4 indicated that formal power of the whistleblower did not determine whether the process pro- tracted internally or externally. Rather, formal whistleblower power did determine with which recipient the process starts. Near and Miceli (1985: 4) stipulated that whistleblowers raise their concern with ‘persons or organisations that may be able to effect action’. The ‘may be able’ refers of course to the whistleblowers’ perception of that actor’s ability. Our findings suggest this perception is dynamic. Following Roth- schild and Miethe (1999), we can argue that the perception of someone’s ability to ‘effect action’ is determined by the response of the previous recipient. However, this might not be the only factor at play. The whistleblower’s perception of who is able to effect action to stop the wrongdoing might also be influenced by the awareness of oneself as a whistle- blower. Weiskopf and Tobias-Miersch (2016) show in a case analysis of the whistleblowing process of Guido Strack—a European Commission whistleblower—how the self-iden- tification as a whistleblower came to articulation only after a number of internal attempts to question a practice and raise a concern about supplier contract breaches. Andrade (2015) also suggests that the moment of self-identification as a whistleblower happens after internal questioning and is a constitutive element of the turn from internal to external

Table 7 Crosstab formal whistleblower power per retaliation

X2 = 0.045 6 cells (33.3%) with count < 5 Note retaliation: 1  =  support; 2  =  informal reprisals, blocking resources, or formal reprisal; 3 = dismissed

Formal whistle- blower power

n Retaliation Total (%)

1 (%) 2 (%) 3 (%)

Unskilled 39 5.1 71.8 23.1 100.0 Skilled 93 8.6 63.4 28.0 100.0 Admin 23 8.7 60.9 30.4 100.0 Professional 104 5.8 66.3 27.9 100.0 Manager 65 3.1 61.5 35.4 100.0 Executive 7 42.9 28.6 28.6 100.0 Total 331 6.9 64.0 29.0 100.0

216 W. Vandekerckhove, A. Phillips

1 3

whistleblowing. Our data do not allow us to determine when exactly people calling the PCAW advice line started to self-identify as a whistleblower. However, the moment they call the PCAW advice line suggests this is somewhere during the whistleblowing process. In our sample, only 5% called the PCAW advice line before raising their concern with anyone. The majority called the advice line after rais- ing their concern once (38%) or twice (39%). A further 15% called after raising their concern three times, and 3% did so after raising their concern four times. Public Concern at Work brands itself as ‘The Whistleblowing Charity’; hence, we can assume that callers recognise what they have done as whistleblowing when they call the PCAW advice line. The literature (Andrade 2015; Rothschild and Miethe 1999; Weiskopf and Tobias-Miersch 2016) suggests that reactions to earlier attempts to raise a concern triggers such a recogni- tion and thus the self-identification as a whistleblower.

Given the apparent importance of line managers and higher management in protracted whistleblowing, we must acknowledge that often these recipients are where the buck seems to stop. A closer look at sequences where line manag- ers are the final (but not first) recipient shows that in 23% of these cases a concern was previously raised directly with the wrongdoer. This sequence suggests that the individual believed they could rectify the perceived wrongdoing them- selves through open and direct communication with the alleged wrongdoer. They subsequently call upon their line manager to step in when their direct attempt fails. Often two subsequent attempts were made with the line manager. This suggests that the individual perceives their line manager has ignored their first attempt to raise the concern, but still trusts the line manager to be willing and able to take action. In some cases, an attempt to raise the concern was made with higher management before turning to the line manager. The individual might be following instructions from higher man- agement who believes the concern needs to be dealt with at lower level. The question remains whether there was any follow-up with regard to how the individual’s line manager has dealt with the concern.

The emerging patterns where higher management was the final recipient point at a more hierarchical rationale of whistleblower strategies. For sequences of two attempts, these were LM-HM, W-HM, and SP-HM. Sequences of three attempts are LM-LM-HM, LM-HM-HM, and HM-SP- HM. Most of these sequences resonate with a hierarchical rationale: if the line manager fails to take action or where the line manager is the wrongdoer, the individual raises the concern with higher management. More surprising are the sequences where the individual makes two attempts to raise their concern with higher management or returns to higher management after raising their concern with a specialist channel (such as internal audit or compliance). These sug- gest that whistleblowers are quite persistent in assuming

higher management acts in good faith. In other words, it often takes a protracted process of raising a concern and interpreting the response before a whistleblower realises those in control of the organisation are not able or not will- ing to deal adequately with concerns about wrongdoing.

Overall, our findings suggest two things. One is that if the formal power of the whistleblower is a determining factor in who they raise their concern with, it is more a question of when rather than whether they will approach a particular recipient. The other suggestion is that the relation between choice of recipient and whistleblowers’ formal power might be more usefully theorised not in terms of ‘how far is the whistleblower willing to go’ but rather which recipient is most in line with the whistleblower’s mode of thought. For example, Skivenes and Trygstad (2010) rely to a great extent on union density in the Norwegian workplace to explain the high incidence of internal whistleblowing in Norway. It is the contact with union representatives that makes the union viable as a whistleblowing recipient. We can explain in a similar vein why unskilled and skilled workers tend not to use specialist channels. Organisations often advertise these as whistleblowing hotlines (either to internal audit or to externally operated anonymous channels). It is possible that blue collar workers do not differentiate between differ- ent levels of management—management is simply the other side. Thus, after having raised their concern with their line manager, relatively few of them see any recourse in an inter- nal route such as a hotline to management.

We can now answer our first research question: ‘If whistleblowing is a protracted process, what is the typical sequence of recipients?’ Our findings suggest that whistle- blowing is a protracted process and that this process is inter- nally more protracted than previously documented. Line management as well as higher management are by far the most used recipients for whistleblowers. The sequence of recipients seems to be further influenced by the previous recipient: most sequences follow a hierarchical rationale indicating the whistleblower searches a more independent recipient rather than merely an external one. However, the formal power of the whistleblower seems to influence which recipients are in scope as a further whistleblowing option.

Effective Whistleblowing

Our findings corroborate with those of Dworkin and Baucus (1998), suggesting legitimacy and credibility of the recipi- ent leverages the effectiveness of whistleblowing only for external recipients. Yet our findings also suggest that regu- lators—who in the UK are mandated as prescribed persons in the whistleblowing legislation—are more effective than other external recipients (H5). Hence, the answer to our second research question ‘At what point in the protracted process does whistleblowing become an effective way to

217Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

stop wrongdoing?’ is when the organisation is no longer in control over whether or not the concern is legitimate.

This provides ground to search for an alternative theoreti- cal framing of power than the one Near and Miceli (1995) have put forward. In their framework, it is the power of the recipient which can enhance the power of the whistleblower. Skivenes and Trygstad (2010) offer an alternative framing. Their focus on union mechanisms gives much greater weight to institutional arrangements beyond the organisational level than Near and Miceli (1995) seem to do. Skivenes and Trygstad (2010) use a power model to theorise in a simi- lar vein that employees with higher education will be have more credibility and hence be more effective. However, their findings contradict their expectations. They note that ‘Surprisingly, highly educated employees were less likely to achieve changes when they blew the whistle than others’ (Skivenes and Trygstad 2010: 1086). Our findings suggest this is not surprising at all. In our sample cases, external whistleblowing was more successful than internal whistle- blowing. Hence, if education level correlates with less exter- nal whistleblowing (Skivenes and Trygstad 2010), then we would expect education level to have a negative relationship with effectiveness. The conclusion must be that for effective whistleblowing the power of the whistleblower matters far less than the choice of the recipient.

Retaliation

Smith and Brown (2008) found that in the Australian pub- lic sector the most common type of whistleblower retalia- tion was informal reprisal. The overall pattern in our UK findings is that formal retaliation and dismissal increase at each attempt. Smith (2014) reviews the literature on whistle- blower retaliation and notes that position and power in the organisation is the most consistent predictor of retaliation, with those in more senior occupational levels less likely to experience retaliation for blowing the whistle. Along with the partial confirmation of H6, further exploration of our UK data suggests that workers with less formal power (unskilled, skilled, admin) seem to be tolerated longer. For them, O’Day’s four rituals of expulsion seemed to describe more accurately the protracted retaliation they experienced. Those with more formal power (professionals, managers, and exec- utives) experience dismissal earlier in the whistleblowing process. Mesmer-Magnus and Viswesvaran (2005) found that external whistleblowing correlates with more retalia- tion. This, however, does not take into regard whistleblowing as a protracted process nor does it distinguish between types of retaliation. Smith (2014) also finds more retaliation fol- lowing external whistleblowing but explains this differently depending on at what stage the whistleblower goes external. When the first recipient is an external one, the whistleblower suffers retaliation because ‘initial reporting through external

avenues breaks with the “proper channels” set out in the law’ (p. 242). However, when the whistleblower goes external at a later stage, retaliation can ‘reflect the fact that by the time whistleblowers report externally, internal processes for handling their reports have already gone badly wrong, so that further action from any source is unlikely to improve the situation’ (p. 243, emphasis in original).

Our findings do not corroborate those of the Australian research (Smith 2014). Just as with our first research ques- tion—how the whistleblowing process protracts—we find that with regard to retaliation formal whistleblower power can influence how the process starts, but this is of much less or no importance for how the process protracts or how retaliation develops further.

Conclusion

In this paper, we used secondary data in the form of whistle- blower narratives from an advice line in the UK to explore whistleblowing as a protracted process. Our findings suggest that formal whistleblower power influences how the whistle- blowing process starts in terms of to whom the whistle is blown and what happens to the whistleblower, but that this influence dies out as the process protracts. Our findings also suggest that a recipient’s institutional power rather than the formal whistleblower power influences the effectiveness of the whistleblowing.

We believe our findings lead to a questioning of the gen- eralisability of research designs that conceptualise whistle- blowing as a one-off event. In particular, we found that the whistleblowing process generally entails two or even three internal attempts to raise a concern before an external attempt is made, if it is made at all. We have also provided ground for further research to reconsider internal/external differentiation by seeking to measure the independence of the recipient. The overall pattern found in our research makes it plausible that whistleblowers search for a more independent recipient at each successive attempt to raise their concern.

Most of the extant whistleblowing literature theorises successful whistleblowing in terms of the legitimacy and credibility of the whistleblower within their organisation. Our findings do not contradict this line of reasoning, but they support Skivenes and Trygstad’s (2010) emphasis on institu- tional power beyond the organisational level. Where whistle- blowing is immediately successful—effective and safe—this might be because the whistleblower has legitimacy and credibility. Yet when whistleblowing is unsuccessful at the first attempt, and hence becomes a protracted process, the whistleblower seems to have little avail in seeking to estab- lish legitimacy within the organisation. Current whistle- blowing theory has little to offer for those whistleblowers.

218 W. Vandekerckhove, A. Phillips

1 3

We believe our research in this paper supports the search for an alternative, more institutional-level analysis of whistle- blowing. This resonates with Skivenes and Trygstad (2010) who explain the high rate of successful whistleblowing in Norway through high levels of unionisation and Norwegian industrial relations. These have established a general attitude that employers do not have full control over how work is organised and how human relations within workplaces are shaped. Union density in the UK is a fraction of that in Nor- way, and whistleblowing to a union represented a very small proportion in our sample. However, our findings suggest that whistleblowing might be more successful if its protracted process involved a third, external player earlier on.

Funding This research was funded by the University of Greenwich (Work and Employment Relations Unit), and Public Concern at Work.

Compliance with Ethical Standards

Conflict of interest Wim Vandekerckhove and Arron Phillips de- clare that they have no conflict of interest.

Ethical Approval All procedures performed in this study involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual partici- pants included in the study. The research was approved by the Uni- versity of Greenwich Research Ethics Committee (Ref 11/12.3.5.21).

References

Andrade, J. A. (2015). Reconceptualising whistleblowing in a complex world. Journal of Business Ethics, 128(2), 321–335.

Apaza, C. R., & Chang, Y. (2011). What makes whistleblowing effec- tive: Whistleblowing in Peru and South Korea. Public Integrity, 13(2), 113–130.

Blenkinsopp, J., & Edwards, M. S. (2008). On not blowing the whis- tle: Quiescent silence as an emotion episode. Emotions, Ethics and Decision-Making Research on Emotion in Organizations, 4, 181–206.

Brown, A. J., Lewis, D., Moberly, R., & Vandekerckhove, W. (Eds.). (2014). International handbook on whistleblowing research. Chel- tenham: Edward Elgar Publishing.

Donkin, M., Smith, R., & Brown, A. J. (2008). How do officials report? Internal and external whistleblowing. In A. J. Brown (Ed.), Whistleblowing in the Australian public sector. Enhancing the theory and practice of internal witness management in public sector organisations (pp. 83–108). Canberra: ANU Press.

Dozier, J. B., & Miceli, M. P. (1985). Potential predictors of whistle- blowing: A prosocial behavior perspective. Academy of Manage- ment Review, 10(4), 823–836.

Dworkin, T., & Baucus, M. S. (1998). Internal vs. external whistle- blowers: A comparison of whistleblowering processes. Journal of Business Ethics, 17(12), 1281–1298.

Gundlach, M. J., Douglas, S. C., & Martinko, M. J. (2003). The deci- sion to blow the whistle: A social information processing frame- work. Academy of Management Review, 28(1), 107–123.

Jones, J. C., Spraakman, G., & Sanchez-Rodriguez, C. (2014). What’s in it for me? An examination of accounting students’ likelihood to report faculty misconduct. Journal of Business Ethics, 123(4), 645–667.

Kaptein, M. (2011). From inaction to external whistleblowing: The influence of the ethical culture of organizations on employee responses to observed wrongdoing. Journal of Business Ethics, 98(3), 513–530.

Lewis, D., & Vandekerckhove, W. (2015). Does following a whistle- blowing procedure make a difference? The evidence from the research conducted for the Francis Inquiry. In D. Lewis & W. Vandekerckhove (Eds.), Developments in whistleblowing research (pp. 85–105). London: International Whistleblowing Research Network.

Lewis, D., & Vandekerckhove, W. (2016). Trade unions and the whistleblowing process in the UK: An opportunity for strategic expansion? Journal of Business Ethics. https://doi.org/10.1007/ s10551-016-3015-z.

MacGregor, J., & Stuebs, M. (2014). The silent Samaritan syndrome: Why the whistle remains unblown. Journal of Business Ethics, 120(2), 149–164.

Mansbach, A., & Bachner, Y. G. (2010). Internal or external whistle- blowing: Nurses’ willingness to report wrongdoing. Nursing Eth- ics, 17(4), 483–490.

Mesmer-Magnus, J., & Viswesvaran, C. (2005). Whistleblowing in organizations: An examination of correlates of whistleblowing intentions, actions, and retaliation. Journal of Business Ethics, 62, 277–297.

Miceli, M. P., & Near, J. P. (1989). The incidence of wrongdoing, whistle-blowing, and retaliation: Results of a naturally occurring field experiment. Employee Responsibilities and Rights Journal, 2(2), 91–108.

Miceli, M. P., & Near, J. P. (1992). Blowing the whistle. New York: Lexington Books.

Miceli, M. P., & Near, J. P. (2002). What makes whistle-blowers effec- tive? Three field studies. Human Relations, 55, 455–479.

Miceli, M. P., Near, J. P., & Dworkin, T. M. (2008). Whistleblowing in organizations. New York: Routledge.

Nader, R., Petkas, P. J., & Blackwell, K. (Eds.). (1972). Whistle blow- ing: The report of the conference on professional responsibility. New York: Grossman.

Near, J. P., & Miceli, M. P. (1985). Organizational dissidence: The case of whistle-blowing. Journal of Business Ethics, 4, 1–16.

Near, J. P., & Miceli, M. P. (1986). Retaliation against whistle blowers: Predictors and effects. Journal of Applied Psychology, 71(1), 137.

Near, J. P., & Miceli, M. P. (1995). Effective-whistle blowing. Academy of Management Review, 20(3), 679–708.

O’Day, R. (1974). Intimidation rituals: Reactions to reform. The Jour- nal of Applied Behavioral Science, 10(3), 373–386.

Park, H., Blenkinsopp, J., Oktem, M. K., & Omurgonulsen, U. (2008). Cultural orientation and attitudes toward different forms of whistleblowing: A comparison of South Korea, Turkey, and the UK. Journal of Business Ethics, 82(4), 929–939.

PCAW. (2013). Whistleblowing—The inside story. London: Public Concern at Work/University of Greenwich.

Rothschild, J., & Miethe, T. D. (1999). Whistle-blower disclosures and management retaliation: The battle to control information about organization corruption. Work and Occupations, 26(1), 107–128.

Sims, R. L., & Keenan, J. P. (1998). Predictors of external whistle- blowing: Organizational and interpersonal variables. Journal of Business Ethics, 17, 411–421.

219Whistleblowing as a Protracted Process: A Study of UK Whistleblower Journeys

1 3

Skivenes, M., & Trygstad, S. C. (2010). When whistle-blowing works: The Norwegian case. Human Relations, 63, 1071–1097.

Smith, R. (2014). Whistleblowing and suffering. In A. J. Brown, D. Lewis, R. Moberly, & W. Vandekerckhove (Eds.), International handbook on whistleblowing research (pp. 230–249). Chelten- ham: Edward Elgar.

Smith, R., & Brown, A. J. (2008). The good, the bad and the ugly: Whistleblowing outcomes. In A. J. Brown (Ed.), Whistleblowing in the Australian public sector (pp. 109–136). Canberra: ANU E Press.

Vandekerckhove, W., Brown, A. J., & Tsahuridu, E. E. (2014). Mana- gerial responsiveness to whistleblowing: Expanding the research

horizon. In A. J. Brown, D. Lewis, R. Moberly, & W. Vandekerck- hove (Eds.), International handbook on whistleblowing research (pp. 298–327). Cheltenham: Edward Elgar.

Watts, L. L., & Buckley, M. R. (2015). A dual-processing model of moral whistleblowing in organizations. Journal of Business Eth- ics. https://doi.org/10.1007/s10551-015-2913-9.

Weiskopf, R., & Tobias-Miersch, Y. (2016). Whistleblowing, parrhesia and the contestation of truth in the workplace. Organization Stud- ies, 37(11), 1621–1640.

Woodford, M. (2012). Exposure: Inside the olympus scandal. London: Penguin.

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.