Introduction

Jones and Millar (2010) argued it is not only desirable but also a matter of organisational survival to ensure organisations behave ethically. Yet examples of large-scale unethical behaviour in organisations continue to be plentiful including for example, at the BBC (Dobbs, 2016), Rotherham Child Sexual Exploitation (Jay, 2013), the Volkswagen emissions scandal (Rhodes, 2016), NHS Mid-Staffs (Francis, 2013), Enron (Tourish & Vatcha, 2005) and Theranos (Debapratim et al., 2020). Voice is one of the most important ways for organisations to learn about unethical behaviour as it allows employees to raise concerns and issues so they can be addressed by managers (Hirschman, 1970; Sherf et al., 2020). The voice and silence literature has identified many important variables which influence voice but gaps in our understanding remain, with the role of time in voice about unethical behaviour highlighted as a key area for exploration (Morrison, 2023). Even in the most egregious scandals it is typically found that a) unethical behaviour started small and grew out of control, b) despite misgivings many people did not speak up at this stage but c) some people did and their concerns went unheeded by management (Pinder & Harlos, 2001) and d) if these voices had been heard and their concerns acted upon the problem could have been addressed while still relatively manageable. What we can see very clearly from these statements is that the timing of voice and subsequent management attention appears crucial to the stopping of the unethical behaviour.

An examination of the role of time could thus help us better understand the voice about unethical behaviour, in several ways. Firstly, to understand the process by which people decide if and how to voice. Secondly, to understand how and why managers react at the time when voice takes place. And finally, to understand how managers can build voice mechanisms which encourages the voicing of unethical behaviour at an earlier time. The aim of this paper is to provide a frame of reference which helps managers and researchers to apply a temporal lens to examine voice about unethical behaviour in organisations. Our paper is organised around the question of what insights can be derived from using a temporal lens to analyse voice about unethical behaviour. We draw three important conclusions. First, voice must continue to take place over time to generate evidence for unethical behaviour. Second, perceptions of what represents unethical behaviour change over time. Third, that it is most difficult to voice about unethical behaviour at the time it is most needed.

The structure of the paper follows a theory adaptation process (Jaakkola, 2020) which shows how a given field can be advanced through the application of a new theoretical lens. Firstly, we carried out a review of the voice, silence and whistleblowing literatures which demonstrates how all are important for understanding voice about unethical behaviour despite the differences in the way they define and measure it. The review identified areas in which there is considerable evidence, for example that leadership is important for creating voice climates and voice mechanisms which encourage voice about unethical behaviour, but also areas in which there are gaps in our understanding. In particular we found that an absence of processual perspective has inhibited our ability to gain a deeper understanding of the way in which individuals make decisions about voice over time. We therefore outline two temporal frameworks which allow us to analyse the way in which organisational processes draw attention towards or away from voice. Based on this analysis, we are able to generate three novel insights—that the passage of time is required for the emergence of evidence about unethical behaviour, that perceptions of unethical behaviour change over time, and that voice attempts are likely to be most challenging and least effective at the very time when they are most needed. Finally, we suggest new opportunities for further research based on these insights.

Voice and Unethical Behaviour

There are several reasons why individuals in organisations might not speak up about unethical behaviour, or take remedial action to stop the behaviour once it is first-known. These reasons relate to fear of being ostracised or vilified (Milliken et al., 2003), obtaining or maintaining power (Currie et al., 2019), aggressive pursuit of business targets (Rhodes, 2016), or not knowing what to do with the information they hold. With regard to aggressive pursuit of targets, the Volkswagen (VW) emissions scandal of 2016 acts as a good example. VW had equipped 11 million diesel engine vehicles with software controls which falsely enabled vehicle emission outputs to meet USA emission standards, despite the vehicles emitting more than 40 times that amount (Rhodes, 2016). This unethical business practice went unhindered from 2009 until when VW was served a notice of violation by the USA environmental protection agency. Prior to this notice, VW had been describing these test discrepancies as a ‘technical glitch’. Even though a “few people” at VW had intentionally installed the software as a way to cheat the standards, many senior executive and board members declared ignorance of the root cause of the issue (Kresge & Weiss, 2015). As can be seen from this example, voice took place, managers did not act on it, and it was only much later that the scale of the problem became evident. We turn to the voice and silence research now to see how voice has been defined and measured in that literature, and to consider the role of time in how past ways of thinking about this phenomena can be enhanced.

Defining and Measuring Voice About Unethical Behaviour

Morrison (2011) called for the voice field to explore different types of voice, recognising that the variables which affect voicing are likely to be different depending on the content of the voice. One of the problems when studying voice about unethical behaviour is that various terms have been used to describe it, making it difficult to identify studies which have captured insights. For example, studies have used terms such as ethical voice (Zheng et al., 2021), prohibitive and promotive voice (Bai et al., 2019; Mo & Shi, 2018; Wang et al., 2020), challenging voice and defensive voice (Bharanitharan et al., 2019) or simply employee voice (Avey et al., 2012; Peng & Wei, 2020).

As the definitions of the various terms are broadly similar it is tempting to assume they are simply different labels for the same behaviour. However, as the measurement scales used in the studies vary we cannot be confident the same thing is being measured. For example, Lee et al., (2017) use the term employee moral voice to describe “the act of speaking out against unethical issues” (p. 52) and measure it using the three items adopted from the moral courage scale developed by Hannah and Avolio (2010). On the other hand, Zheng et al. (2021) use the term ethical voice which they define as “employees discussing and speaking out opinions against unethical issues in the workplace” (p133) but measured it using four items from a safety voice measure (Tucker et al., 2008). The most common scale for measuring employee voice by Van Dyne and LePine (1998) does not include questions about workplace issues which contravene legal and moral principles (Morrison, 2023). Therefore, studies that have used this as a scale (e.g. Peng & Wei, 2020) provide little insight into voice about unethical behaviour specifically and this makes it difficult to look at past research to draw conclusions. However, one place that might provide some insight into voicing about legal and moral issues is the whistleblowing literature.

Whistleblowing can be defined as the disclosure by organisational members of perceived organisational wrongdoing to authorities who can take action (Near & Miceli, 1985). Whistleblowing and voice have tended to be researched separately, though they are clearly related phenomena. Someone raising a concern with their line manager about ethical issues can equally well be described as enacting voice about unethical behaviour or as an internal whistleblower, while Vandekerckhove and Phillips (2019) found almost all external whistleblowing comes only after the individual has raised their concerns about wrongdoing within the organisation. Observing that initial voice about unethical behaviour is not generally framed as whistleblowing by either the employee or the manager, Blenkinsopp et al. (2019) suggest it can be useful to think in terms of a continuum from raising concerns to speaking up to whistleblowing. Given the focus of our article is on the role of time in voice about unethical behaviour we will ground our analysis within the voice literature, but there is clearly scope for future research seeking to develop greater connection between the two fields. Whistleblowing is presently one of the commonest routes by which large-scale unethical behaviour is exposed. This suggests voice systems currently found within many organisations may not be optimally designed to deal with voice about unethical behaviour.

Leadership and Voice About Unethical Behaviour

There has been a strong focus on the way that leadership styles such as ethical leadership (Avey et al., 2012; Bai et al., 2019; Bharanitharan et al., 2019; Chen, 2010; Lee et al., 2017; Mo & Shi, 2018) and authoritarian leadership (Zheng et al., 2021) shape voice about unethical behaviour. Ethical leadership demonstrates to employees how they are supposed to act and gives clear signals to employees considering voicing about unethical behaviour that it is the correct thing to do (Lee et al., 2017). Studies looking at variables which shape voice about unethical behaviour identified that a combination of ethical leadership and organisational climate encourages voice most successfully (Bai et al., 2019; Kim & Vandenberghe, 2020; Peng & Wei, 2020).

Voice Climate

Studies looking at voice climate have prioritised the contextual influences which shape voice (Frazier & Fainshmidt, 2012) and identified that climate appears to be related to time and voice in differing ways. For example, in a climate where leaders and employees share perceptions that unethical behaviour should be dealt with, voice is more likely to happen and more quickly (Peng & Wei, 2020). Leaders with high levels of integrity encouraged employees to voice about deeply held personal and moral convictions by providing consistency of response, thereby reducing the perceived risk that results from uncertain leader response. In another study, where staff in a healthcare setting felt safe to speak up about issues, the patients reported better timeliness of care (Nembhard et al., 2015). Nembhard et al. (2015) identified that employees at the lower end of the hierarchy had better potential for influencing timeliness of care because they had a broad and deep knowledge about organisational processes and procedures whereas more senior members of the team had narrower access to less information directly relating to patient care. Hence, we concern ourselves next with the types of voice mechanisms that employees might utilise to enact voice, whether they are formally known as part of existing procedures where the climate supports it, or those which develop emergently on an informal basis.

Voice Mechanisms

Formal voice mechanisms are recognised as a key indicator of voice climate as they symbolise managerial interest in hearing what employees have to say (Kwon & Farndale, 2020; Mowbray et al., 2014; Vakola & Bouradas, 2005). Similarly, an absence of formal voice mechanisms is likely to signal that voice is not welcome (Donaghey et al., 2011; Morrison & Milliken, 2000). Formal voice mechanisms, defined as “codified, pre-arranged, and regular/concrete structures that foster consistent implementation and that reduce the discretionary powers of voice managers” (Mowbray et al., 2014 p.8) include open-door policies, grievance procedures, appraisal schemes, employee suggestion boxes, scheduled meetings with managers and email (Kwon & Farndale, 2020; Mowbray et al., 2014). Informal voice mechanisms are generally considered to be ones which are unplanned and therefore fall outside of the formal organisational structure, such as bumping into a manager in a corridor or calling them without prior arrangement. With informal voice mechanisms there are unlikely to be procedures about how to use them and how managers should respond (Mowbray et al., 2014).

When considering the sensitive nature of voice about unethical behaviour, the choice of voice mechanisms is likely to be especially important. Whereas voice in general is considered to be a risky act owing to its challenging nature (Milliken et al., 2003; Peng & Wei, 2020), voice using formal mechanisms can be considered more risky because mechanisms are often recorded meaning information can be shared with those other than the target of the voice (Brooks, 2018). Furthermore, formal voice mechanisms are often accompanied by documentation and need to be scheduled in advance to accommodate the multiple parties who need to be present, all of this taking time to set up. If the voicer has an opportunity to use a less risky informal voice mechanism, that might appear to be a much simpler and less time-consuming option. However, although there is an advantage provided by the lack of formality surrounding informal voice mechanisms, the informality can also make voice less effective as managers are then able to use discretion about how they respond to the voice (Harlos, 2001). Inaction by managers is more likely to occur where the information being voiced is considered insignificant to the manager or goes against performance objectives they are striving towards. Harlos (2001) refers to this phenomena as deaf ear syndrome, explaining how employees’ experiences of workplace injustice result from voice mechanisms which failed to adequately circle back to their concerns and increased dissatisfaction among employees. The effect was found in both formal and informal mechanisms, with the latter found to be particularly susceptible to failure (Harlos, 2001).

While these concepts of leadership, voice climate, and voice mechanisms are useful in defining foundational conditions for whether and how voice is enacted, we argue that a temporal lens will enhance a processual understanding of how and why voice events play out over time to enable greater insight into how voice about unethical behaviour can be efficacious. To elaborate this point we turn now to providing an outline of the temporal frameworks upon which we have drawn.

Adopting a Temporal Lens

Temporality can be defined as “a non-spatial continuum in which events occur in apparently irreversible succession from the past through the present to the future” (Ancona et al., 2001b, p.513). In an organisational context it refers to the ways organisations sequence and manage activities. The management of voice, being one such activity, would benefit from analysis using a temporal lens given growing recognition that cross-sectional research is limited in its ability to help us answer questions such as how people decide to voice and how managers respond to it (Morrison, 2023). We therefore present a framework by Ancona et al., (2001b) which highlights three ways in which time can be analysed in organisations. Given our focus upon voice mechanisms we also incorporate a second analytical framework suggesting that knowledge of attentional scales could highlight how organisations could be missing opportunities to hear voice about unethical behaviour (Bansal et al., 2018). Subsequently, we will use both of these frameworks in relation to voice about unethical behaviour and present our insights and future avenues of research.

Temporal Framework 1

The temporal framework developed by Ancona et al., (2001a, b) comprises three categories—conceptualisation, mapping activities, and actors.

Category 1: Conceptions of Time

This first category encourages analysis of the types of time—clock-time and event-based time—used to manage organisational activities. Clock-time can be characterised as activities which are scheduled to take place at a particular time, such as meetings, budgeting cycles and performance appraisals. Event-based time takes into account the unplanned and planned happening of events but unlike clock-time, these activities take place outside of any specific plan or schedule that dictates how the organisation works. For example, employees might refer to activities needing to be done before or after a temporal referent such as Christmas holidays but these do not form part of the temporal rhythm that sustains successful organisations in the same way that yearly planning meetings and daily operations meetings do.

Category 2: Mapping Activities to Time

This category focuses on how activities within organisations are explicitly and deliberately mapped within the time available. This category is not about individuals directly but the way in which work activities are planned, either by them, or by managers within the organisational setting. Although individuals in roles with high levels of autonomy may have more control over scheduling and planning of their own work this work nevertheless forms part of a larger process which needs to be taken into account when scheduling activities. For example, the operations manager could change the time of the daily operations meeting to suit themselves but this might not suit the shift manager who requires performance data by a particular time to meet daily production schedules.

Category 3: Actors Relating to Time

This category is concerned with the way in which individuals perceive time and subsequently act in relation to those perceptions of time. Each individual has a temporal personality, that is, their own unique way of interacting with time. The temporal personality describes the way an individual thinks about the past, the present and the future (Fabbri et al., 2020), how they wish to spend their time and the prioritisation they apply to organisational activities (Blount & Janicik, 2001). In other words, the way in which time is considered by individuals shapes the way that they make decisions.

Having highlighted three ways in which time can be analysed in organisations, we turn now to a consideration of how time is perceived within organisations, showing how different organisational archetypes vary in their attitudes towards time and exploring the implications for voice about unethical behaviour.

Recognising Different Attitudes to Time in Different Organisations

To make sense of the different ways in which the types of time could be understood in organisations Butler (1995) characterised organisations along a continuum from bureaucratic to ‘garbage can’. A bureaucratic organisation has close links between its past, present and the future which show up as formalised features which are designed to guarantee success (Butler, 1995). For example, procedures which explain how things should be done are built on past learning about what works and what to do if things go wrong. As a result, the dominant feature of bureaucratic organisations is that work is highly ordered and predictable. They are characterised by a strong top-down management style and are driven by KPIs and targets which are used to ensure targets are met on time and in sync with other temporal cycles relating to budgeting and strategy. There is a strong desire for the organisation to continue to operate in these ways, because the formalised features have been created over time from organisational learning and are therefore considered important for the success of the organisation. Managers in bureaucratic organisations are likely to use clock-time (Ancona et al., 2001b) to decide whether attention should be given to an activity. Generally speaking, with only limited time available, under-performance will receive more attention than over-performance, and serious under-performance will receive more attention than moderate under-performance (Butler, 1995). Where there are multiple problems of under-performance, the least risky option will be dealt with.

In ‘garbage can’ organisations (Butler, 1995) there is very little relationship between the past, the present, and the future because they are new, dynamic organisations (e.g. business startups) who have yet to find out what works for them. As a result, the most common type of temporal referents in these organisations are ‘events’ and so time is said to be event-based. Given the nascence of organisations that use event-based time, there are unlikely to be processes and procedures telling managers exactly what to do so situations are likely to appear unpredictable and irregular (Ancona et al., 2001a). In ‘garbage can’ organisations things that need attention are likely to be dealt with as and when they occur because there are no pre-determined templates about what does or doesn’t work.

Using the framework by Ancona et al., (2001b) to explore voice about unethical behaviour enables us to see that timing of voice is likely to be important given the prevalence of silence in clock-based bureaucratic organisations (Brinsfield, 2013; Morrison & Milliken, 2000). However, the framework does not offer a way for us to explore what happens when voice takes place in a clock-based organisation. The framework indicates activities are planned into the organisation in order to meet performance targets. However, voice can be planned and unplanned, denoted by the use of formal and informal voice mechanisms (Kwon & Farndale, 2020). As such, we were faced with questions such as what do managers do if voice is raised outside of the schedule i.e. informally? What determines if managers subsequently build activities such as voice into their schedule? What difference does it make if voice is planned and was expected by managers as part of the schedule? To explore these questions we draw upon a second temporal framework proposed by Bansal et al. (2018) to explain how management attention is directed to activities.

Temporal Framework 2: Attention

Where issues appear in an organisation which are not addressed, it can be the result of attentional failure. That is, managers did not give attention to a situation. Bansal et al. (2018) suggest attentional failure occurs when issues appear in a way which is not expected by managers. Attentional scales highlight that both temporal and spatial aspects of an organisation are important for ensuring attention is paid to organisational phenomena. A spatial scale can be defined as “the geographical area in which the dominant process(es) of interest manifest” whereas the temporal scale refers to the “patterned variations in processes over time” (Bansal et al., 2018 p221). There is likely to be a relationship between the temporal scale and the spatial scale in that more senior managers tend to focus more on longer-term trends and larger-scale patterns compared to those lower down the organisation. It is therefore possible there will be a lack of understanding and interest when lower level issues are raised to senior managers. This lack of interest and understanding means voice often goes unheard because attention is not paid to issues which do not make sense to those focusing on different temporal and spatial scales (Harlos, 2001; Nembhard et al., 2015). To explain some of these attention failures, Bansal et al. (2018) suggest using the concepts of attentional grain and attentional extent.

Attentional Grain

Attentional grain represents the unit of measurement from fine to coarse that an organisation uses to observe a process in time and space. The attentional grain indicates an organisation’s ability to narrow in on specific issues by looking for precise examples. The more an organisation’s management uses a fine attentional grain, the more chance they have of becoming aware of problems (or potential) problems and addressing them. While it may be appropriate for senior management to operate with a different attentional grain to shop floor workers, they also need to be open to attending to voice which raises issues at a finer attentional grain, just like a cancer specialist needs to be willing to respond to a patient’s request that they take a look at a freckle or listen to a cough. For example, looking for small amounts of unaccounted for money in a department might identify small-scale theft, and attending to this might reveal weaknesses in the firm’s finance systems which have been allowing small-scale thefts to become common right across the firm.

Attentional Extent

Attentional extent relates to the range of measurement from narrow to broad used to measure a problem. If a problem is large, such as accounting fraud within an organisation over many years, a coarse attentional extent would pick up trends and patterns by looking for large sums of money across different years and different departments. A narrow attentional extent would only look for specific things so it could be more useful for identifying accounting fraud at a local level by looking for small sums of money in that department alone. Looking for small sums of money among many millions is likely to be time consuming and inefficient, whereas looking for large amounts of money in one department is unlikely to identify small-scale theft. Therefore, identifying organisational issues requires an appropriate attentional extent and grain. While different levels in the hierarchy are likely to operate for much of the time with particular attentional extent and gain, an inability to be flexible when required can lead to failures to attend to serious problems.

How a Temporal Lens Helps Us Reframe Voice about Unethical Behaviour

Our article highlights three ways in which a temporal lens shifts our understanding of voice about unethical behaviour. First a temporal lens highlights how evidence of wrongdoing can build over time such that with hindsight management can see how they might have addressed it earlier. Second it draws attention to how behaviour can come to be seen as unethical as social and cultural norms shift over time. Finally it highlights why timing of voice behaviour matters so much which is important for understanding why voice behaviour is so often unsuccessful in effecting change. We have developed a simple, heuristic model (see Fig. 1) to allow us to make sense of our three insights and how they relate to time.

Fig. 1
figure 1

The continuum of time and voice about unethical behaviour

Full size image

This simplified model contains three time points, T0, T1, and T2. T0 represents the moment where individuals within an organisation become aware of an unethical issue. T1 is the moment where individuals have come forward and enacted voice through either formal or informal mechanisms, and T2 is the time where action, in response to the voice at T1, is initiated. The lines between the time points are the lapse in time between each event. Next, our three insights are presented and refer to these time points.

It Takes Time for Voices to Generate Evidence for Unethical Behaviour

It takes time to gather sufficient evidence about unethical activity in an organisation to realise that concerning events have occurred. There are two stages to this. The time between T0 and T1 is when employees have become aware of a potential issue and are deciding whether there really is an issue and if so whether they are prepared to voice their concerns. At T1 individuals are actively seeking ways to voice about unethical behaviour because it has reached a point where they are no longer able to ignore it. However, at this point some misconduct might be overlooked where organisational leaders suffer from deaf ear syndrome (Harlos, 2001) and information is considered insignificant, or goes against organisational performance goals. Also at this point there may be insufficient attentional scales (Bansal et al., 2018) at a fine enough grain to manage voice about certain concerns because there is not enough evidence that unethical behaviour needs to be dealt with at an organisational level. As a result major unethical behaviour can go undetected at management level over a period of many months or years, until the problem has reached a critical point. Failure to identify and react to early signs of unethical behaviour at T1 can allow failures in organisations to stay hidden (Balch & Armstrong, 2010; Weick & Sutcliffe, 2003). This can lead to collective denial, which encompasses the misconduct perpetuated by many employees simultaneously (Hendy & Tucker, 2021), where they unify around the non-disclosure of deteriorating standards (Morrison & Milliken, 2000).

To illustrate this insight, consider the Bristol Royal Infirmary (BRI) in England where over 30 children undergoing heart surgery died between 1981 and 1995 (Crown Copyright, 2002). In the first year, performance began in line with other similar heart surgery centres, yet across the next seven years all others improved while BRI did not. Surgeons at BRI held a particular mindset towards this evidence, they believed differences in mortality rates were due to patient cases being unusually complex, saw the unit as being on a learning curve, and took an overly optimistic view of evidence of gradual improvements (Weick & Sutcliffe, 2003). It did not take long for employees within the organisation to raise concerns about the unacceptable practices of the surgeons. Concerns about the quality of care were raised as early as 1986 and over the next 9 years, more than 100 formal concerns were raised about the service (Crown Copyright, 2002), yet it was not until 1995 that they were taken on-board by management. An anaesthetist employed at BRI filed a complaint with the CEO. The CEO dismissed his concerns, saying the issue was a clinical matter clinicians should sort out on their own (Weick & Sutcliffe, 2003). A report published in 1989 by a pathologist at BRI found that out of 76 post-mortem examinations on children treated at the centre, 29 listed anomalies and surgical flaws as contributing to death. By 1991 the Royal College of Physicians refused to accredit the BRI as a place to train for paediatric cardiology. These concerns and reports had a cumulative effect, and eventually the evidence of poor quality at BRI could not be explained away by surgeons or management. A climax was reached in 1995 when a child died during a procedure which had been resisted by everyone involved, except the two surgeons who performed it. Services were halted and a public inquiry was launched in 1996 and ran until 2001 (Crown Copyright, 2002).

Reflecting upon this example, we see the CEO demonstrating a form of deaf ear syndrome (Harlos, 2001) where his organisation failed to listen and respond to the repeated voice of employees. One reason this persisted was because some organisational leaders had a say in designating BRI as paediatric heart centre in the early 1980s and did not want to listen to dissenting views and data, which went against their own judgements about the suitability of the organisation (Weick & Sutcliffe, 2003). As a result, no priority was placed on implementing fine enough attentional scales to record and act upon these concerns. The other reason is the surgeons engendered a ‘culture of entrapment’. Described as “the process by which people get locked into lines of action, subsequently justify those lines of action, and search for confirmation that they are doing what they should be doing” (Weick & Sutcliffe, 2003 p. 73) the culture of entrapment enabled a collective denial about quality issues over time. Thus, managers at BRI did not have a fine enough attentional grain towards the quality of paediatric heart services specifically to interrogate further the defensive rhetoric of surgeons over learning curves and complex patient cases. It took many years, unfortunate events, and reports by employees, for this organisation to recalibrate their attentional scales towards voice about these types of unethical activities.

Perceptions of Unethical Behaviour Change Over Time

Our first contribution shows the time between T1 and T2 represents the period where evidence is accumulated to demonstrate that unethical behaviour has occurred, with initial scarcity of evidence being a potential reason for lack of voice or for management inaction in response to voice. We might call this the “had we known all this at the time” defence. In this second contribution what changes is not the volume of evidence (though this may also be significant) but the view of that evidence taken both by workers and management. The change of view on the evidence may take three forms. Sometimes the change is really just a willingness finally to admit there is a problem (cf. Mayor Quimby, “In light of these new facts, of which I now realize I was largely aware, I must take action”). Other times the individual may come to a genuine reappraisal of the same evidence, in some cases triggered by additional evidence.Footnote 1 Finally, sometimes what changes is not so much the individual’s view of the facts of the situation but their understanding of how others view it. The impact of the #metoo movement on Hollywood offers a useful example. The revelations about Harvey Weinstein shocked even insiders to the business, and their reactions reflect the three forms described above. For some it made them acknowledge a problem of which they already had some awareness, for others it changed their understanding of the nature of what was happening. Others were appalled by Weinstein but unconvinced by the idea that he represented the tip of the iceberg and were concerned about the risk of witch-hunt. The backlash to such criticism led some to retract their criticisms. Some of the retractions seemed more a product of PR concerns than a genuine change of view.

The wider #metoo movement also led to a re-evaluation of behaviour which had previously been considered unremarkable. Debating its own stance on workplace relationships several decades ago the European Commission drew upon statistics suggesting a third of long-term relationships had started in the workplace and concluded it was unfeasible to treat the initiation of such relationships as, by default, a form of sexual harassment. Yet in the last decade more and more employers are treating such relationships as simply unacceptable and creating policies which forbid their employees from engaging in them. This can be viewed as simply an excess of caution, HR taking a view that an outright ban is the easiest way forward. However, discussion around high profile cases of executives such as Brian Krzanich (Intel CEO) and Steve Easterbrook (McDonalds CEO), both fired for breaking their firms’ no-fraternisation policies, suggests the behaviour itself (not merely the breaking of a company policy) is now viewed more negatively, especially when hierarchical differences create the potential for power imbalances (cf. Mainiero, 2020). This shows how change over time in attitudes towards a set of behaviours can lead to them being viewed as unethical. But the example also illustrates how the expectations placed on management to respond can also change over time. These shifts have important implications for voice.

Most Difficult to Voice About Unethical Behaviour at the Time it is Most Needed

The third contribution is to highlight that voice is likely to be most difficult yet most needed at a time when managers aren’t seeking it. At T1, individuals are starting to voice because they are concerned by things happening in the organisation but at that point managers are not yet fully aware (or do not accept) that the situation requires their intervention. T1 represents a time when issues are just starting to surface but before large-scale damage has yet to happen, a time when action to address the issues could prevent them escalating. Many such examples indicate that unethical behaviour often starts with incidents which increase over time (see earlier examples of VW, BRI). One of the many tragedies of reports about large-scale and serious breaches of ethical behaviour is that it becomes clear, in retrospect, that people had attempted to voice their concerns but their concerns had been dismissed or fallen on deaf ears.

Drawing upon the framework presented by Ancona et al., (2001b) we can see that within organisations that use clock-time to plan their activities (most Western bureaucratic organisations) the absence of formal voice mechanisms is a sign that hearing about unethical behaviour has not been built into the temporal rhythm of the organisation. At the other end of the scale, where event-based time prevails, the absence of formal voice mechanisms is likely to be less important because in such organisations, there are no established processes and procedures for dealing with anything. Instead events occur and they are dealt with as and when they appear to be important. This could lead such organisations to deal with unethical behaviour sooner and more effectively, though we acknowledge that in startups survival and growth may be of greater concern than ethics so they too may ignore voice about unethical behaviour (cf. Theranos).

The ‘garbage can’ model of decision-making is often used in organisations which use event-based time, with decisions made based on events rather than KPI targets and strategic outcomes as might be the case in bureaucratic organisations (Cohen et al., 2012). ‘Garbage can’ models are more likely to view voice as an opportunity to learn more about an organisation, rather than as a problem, because it is identifying something which does not fit into the neat organisation of a bureaucratic organisation (Cohen et al., 2012).

Organisational norms indicate voice is not of interest to managers where formal voice mechanisms do not exist (Pinder & Harlos, 2001). With regard to time and voice about unethical behaviour, the presence of formal voice mechanisms indicates that voice could be safe and welcomed by managers (Kwon & Farndale, 2020), and which direct management attention to particular issues, giving them the time and space to deal with them. Without such recognition, management can more easily sweep voice under the carpet and prioritise other issues (Harlos, 2001). Therefore, when using the attentional scales framework presented by Bansal et al (2018), we can see formal voice mechanisms are very effective at signalling that an organisation is aware of particular issues and interested in hearing about them at that time.

Implicit in our argument is an assumption that voice will be easier where formal mechanisms exist, yielding higher levels of voice overall. However, the voice and silence literature has researched extensively the different variables which create conditions favourable to voice. What the literature shows overwhelmingly is that voice is seldom easy. At best, it is a carefully considered decision by employees who, over time, weigh up the risks and benefits of voicing and frequently can result in silence (Blenkinsopp & Edwards, 2008; Brinsfield, 2013; Milliken et al., 2003). Therefore, how can we use our temporal framework insights to suggest a way forward?

Using the temporal frameworks presented, we can see that where managers do not have time allocated within their role to listen to and deal with voice, they are less likely to be receptive to it. This is supported by studies which find that where voice is perceived to be an in-role behaviour, managers are more likely to welcome it (Morrison, 1994). Where managers consider voice to be part of their role, they are more likely to be open, honest and transparent and are likely to be considered more approachable and voice is likely to be higher, even for sensitive issues (Detert & Burris, 2007; Tangirala & Ramanujam, 2008). Building upon our three insights into the temporal aspects of voice about unethical behaviour detailed above, we suggest new avenues for future research.

Avenues for Future Research

The idea that time should be central to research on voice about unethical behaviour is supported by policy and practice. In the aftermath of scandals involving unethical behaviour in organisations, questions relating to time are often of primary interest to the authorities. Determining who did what and when, who knew what and when, and what efforts (if any) were made to try and address the problem, are all crucial whether the desire is to apportion blame or learn lessons. Although a rich and complex picture typically emerges from such investigations, the timeline of the scandal remains central to understanding how it arose and unfolded. Although inquiries have been surprisingly incurious about the role of voice during scandals (Powell et al., 2022) their focus on the importance of time nevertheless offers voice researchers a potential steer. Drawing up our proposed model we identify potential implications for future research.

Perhaps the most important implication is the need for different research designs and methodologies. Treating time as a key feature of voice behaviour is potentially difficult when designing research into voice about unethical behaviour, as we rarely know when or where voice will emerge. Recalling Fig. 1, it is difficult to imagine how researchers could be involved at T0, unless the unethical behaviour started at a time when the researchers already happened to be in the organisation studying something else. Involvement between T1 and T2 is possible in principle but would require an organisation willing to take the unusual step of inviting researchers to get involved in examining voice about unethical behaviour at a point where management themselves have not yet decided whether they agree with the concerns being raised and/or whether they will take action.Footnote 2 Researchers are therefore likely to have to accept the limitations of gathering data retrospectively, which is an obvious problem for research using a temporal lens. The design requires a robust methodological approach which takes into account the potential risks inherent with retrospective work. Since voice research is predominantly quantitative (e.g.Chamberlin et al., 2017; Sherf et al., 2020), it has historically identified correlation between different antecedents and voice behaviour; however, it does not describe the underlying processes which produce the observed relationships. Furthermore, as identified earlier, the establishment of an agreed definition and measurement of voice about unethical behaviour are required to allow both future and retrospective research. To make significant advances in the field of voice research, Engemann and Scott (2020) emphasise the need for methodological diversification. We concur and suggest there is a need for future qualitative research which facilitates a processual understanding of how and why events play out over time (e.g.Langley, 1999; Pettigrew, 1992). The use of a timeline (e.g. Mazzetti & Blenkinsopp, 2012) to identify critical incidents (Chell, 1998) in the span of voice episode(s) would highlight important milestones. Other fields in business research, such as strategy and careers, have successfully employed retrospective methods to understand influences and their impact over time.

The use of retrospective methods to develop a better understanding of how both the unethical behaviour and the voice in response to it emerges over time opens up the possibility for researchers to establish the clarity on sequencing and causation which public inquiries have historically achieved. While we cannot know exactly what the data might look like, paying attention to temporal reference points seems important. Future research could begin with the point (T0) where the voicer is aware of an unethical issue, to the enactment of voice (T1), and the corresponding action by managers in response to the voice (T2). The same could be plotted for attentional scales, for example, to determine at which point the attentional grain was fine enough to permit successful reporting of unethical behaviour. Overall the ambition would be to gain clarity about the start and end point of the decision-making process about voice, both the voicer’s decision to speak up and management’s decision to act (or not).

Despite scholars identifying that formality is an important influence over voice and silence (Klaas et al., 2012; Kwon & Farndale, 2020; Mowbray et al., 2014), there is still a lack of knowledge over the ways that formal and informal voice mechanisms shape propensity for voice. We have demonstrated that formal voice mechanisms are more likely to signal that voice is welcome and there is more likely to be a process and procedure supporting managers to deal with the voice. Yet voice may also be sometimes perceived as more risky using formal voice mechanisms because by their nature they are recorded and therefore do not allow for confidentiality. Attempts have been made to distinguish them (Brooks, 2018; Mowbray et al, 2014) but the features academics use to distinguish between formal and informal voice do not appear to be applied in the same way by practitioners (Brooks, 2018). Therefore, in order to better understand the role of time, an exploration of the relationship between the formality of voice mechanisms and the speed (or latency) of voice, and the variables which influence these effects are required.

Practical Implications

The insights generated using a temporal lens provide a call to action for organisations to ensure they are searching specifically for unethical behaviour appropriately in their organisation and signalling to employees that these types of voice are important to the organisation. Since choosing the exact moment to enact voice could be frustrated by concepts of clock and event-based time enforced by the organisation’s temporal rhythm and subject to perception using each individual’s temporal personality, voice might not be easily planned by employees and managers. The use of a temporal lens thus highlights that influence over voice about unethical behaviour does not lie solely with the manager or the employee.

The concept of attentional scales, grain and extent (Bansal et al., 2018) help us understand that looking for small-scale unethical behaviour requires different techniques than large-scale unethical behaviour. Organisations often use voice mechanisms as a way of encouraging employees to voice their concerns. However, voice mechanisms are typically generically designed across the organisation without providing very specific examples of how to use the voice mechanisms for different types of issues. As a result, voicers are not given clear signals which are meaningful to lower levels in the same way as they are meaningful to senior levels.

Conclusion

Drawing upon frameworks developed by Ancona et al., (2001b) and Bansal et al., (2018), we have outlined a temporal lens with which we can better understand the role of time in voice about unethical behaviour in organisations. We have drawn upon two temporal frameworks by way of introduction to a temporal lens which enables a reframe of voice about unethical behaviour in organisations. The application of a temporal lens to the study of voice yields new opportunities which benefit both researchers and practitioners alike. Taking a time-based perspective enables us to problematise existing dilemmas regarding voice about unethical behaviour. For example, how long does it take someone to make a decision to voice? How does someone decide that it is the right time to voice? How long does it take for a manager to decide that voice is worth taking action about? A consideration of temporal rhythms within organisations is likely to increase our understanding of voice about unethical behaviour by showing why timing is an important variable in encouraging voice and ensuring it is attended to. As it pertains to practical applications for managers in organisations, we have highlighted the importance of both clock and event-based time. We have shown how timely voice about unethical behaviour might help improve organisational reputation or financial circumstance, indeed in some cases might prevent scandals like Enron or VW’s ‘Dieselgate’ which had economy-wide implications.