Accountability is of ‘strategic importance’ to organisations because, without it, it would be difficult to coordinate activities and individuals would act without regard for consequences (Romzek, 2015, p. 27), especially where mine safety is concerned. Accountability is a ‘foundational social force’ that regulates individual behaviour (Brees & Martinko, 2015, p. 63) and requires individuals to provide reason and justification for their behaviour or conduct in some form of social exhange as suggested by Messner (2009). It is also considered a complex, nebulous and ever-expanding concept, with multiple definitions (Goodman et al., 2021; Mansouri & Rowney, 2014; McKernan, 2012; Romzek, 2015; Willems & Van Dooren, 2011). Therefore, it is important to define it more tightly to improve its efficacy.

The characterisations of accountability are wide-ranging. Some people regard accountability as a ‘virtue’ encompassing values such as transparency, integrity and responsibility, while others consider it to be a social mechanism (Bovens, 2010; Hall & Ferris, 2011; Romzek, 2015). Furthermore, accountability does not always yield ‘universally positive’ results, with some perceiving it to be a stressor (Hall & Ferris, 2011, p. 132; Laird et al., 2015; Wikhamn & Hall, 2014). The multi-faceted nature of the concept often makes understanding, implementation and measurement challenging. Indeed, Cordery et al. (2010, p. 795) suggest that accountability is ‘more readily identified when it is absent than when it is present’ because it is often when individuals do not take accountability for poor outcomes that it is noticed or recognised.

As there are multiple interpretations of accountability (Hall et al., 2017; Mansouri & Rowney, 2014), it is necessary to seek a common definition of the term within a mining organisation and to understand how accountability influences individual behaviour, as illustrated in the next section:

Beyond the definitional issues, accountability is a multi-disciplinary construct and, accordingly, is developed differently across a range of disciplines. From an economics perspective, accountability theory is largely based on agency theory which postulates that accountability is derived from a contractual relationship between the principal and an agent (Bosse & Phillips, 2016; Eisenhardt, 1989; Goodman et al., 2021; Mansouri & Rowney, 2014; Mero et al., 2014). However, from a public administration perspective (Royle, 2017, p. 20), accountability is a contextually bound concept consisting of a ‘wide array of phenomena’, including individuals’ attitudes, behaviours and subjective interpretations.

The public administration view is echoed in organisational studies (Busuioc & Lodge, 2017). Therefore, accountability cannot be limited to the regnant principal‒agent model underpinning agency theory which prioritises contractual relationships, because consideration needs to be given to the ‘individual, subjective and internal nature of accountability’ (Hall & Ferris, 2011, p. 132). Using a more process-based, as opposed to a contract-based, approach to accountability, Hall et al. (2017) reviewed the extant theoretical and empirical literature on accountability and identified seven key concepts from a number of theoretical frameworks, as presented in Table 1.

The source of accountability, or the ‘to whom’ aspect of the relationship, is an important feature of the accountability environment (Hall et al., 2017). Yet the interdependent relationship between an account holder and an account giver is often neglected in the extant literature on accountability, which focuses primarily on the influence of a single moderator or antecedent (Busuioc & Lodge, 2017; Hall & Ferris, 2011; Hall et al., 2017; Joannides, 2012; McKernan, 2012; Mero et al., 2014; Messner, 2009). Understanding the ‘to whom’ aspect is crucial because individuals may alter their behaviour according to the source of accountability and the investment made in the relationship (Busuioc & Lodge, 2017; Goodman et al., 2021; Hall et al., 2017). The literature refers to four main sources of accountability: to oneself, to peers, to managers, and to systems.

Accountability research shows the popularity of the hierarchical approach, where the focus is on a single factor that drives accountability (Hall & Ferris, 2011; Mero et al., 2014). Pearson and Sutherland (2017) developed the ‘Systems of Accountability Model’ (p. 434) to show how a range of factors relate to and interact with one another. The factors that feature in the Pearson and Sutherland (2017) model differ from those identified in previous studies. For example, peers and managers are not considered to be key drivers of accountability. Instead, the five primary drivers are as follows: systems, the culture of the organisation, clarity of role and tasks, strategic leadership, and the individual (Pearson & Sutherland, 2017). In the model, each factor has a critical threshold that limits its influence on accountability, after which additional factors may come into play.

Having considered the various sources of accountability, it is necessary to identify and understand ‘to whom’ employees view themselves to be accountable, a critical feature of the accountability environment and an area of accountability that warrants research (Hall et al., 2017), particularly in the mining sector. Furthermore, it is important to understand the reasons why and the extent to which an individual, in a mining context, prioritises different sources of accountability. This is a further area of accountability that warrants research:

Although accountability is often associated with virtuous behaviour, personal ethics may not be sufficient to ensure accountability within organisations. External constraints and controls, or accountability mechanisms, are often required (Dubnick, 2003; Goodman et al., 2021). Organisational responses to the need for accountability include the introduction of formalised mechanisms or systems, such as formal reporting processes and procedures (Frink & Klimoski, 2004). Alternatively, informal mechanisms such as values, cultural norms and interpersonal relationships are used to drive accountability within organisations (Frink & Klimoski, 2004; Romzek et al., 2012). The use of both informal and formal accountability mechanisms establishes a ‘web of accountabilities’ that occurs at varying levels within an organisation, further illustrating the complexity of the accountability construct (Frink & Klimoski, 2004, p. 3).

Individuals are required to make sense of the accountability web and prioritise sources of accountability within it (Pearson & Sutherland, 2017). While there are various types of accountability mechanisms, little is known about the specific mechanisms used for identified sources in the context of the mining sector (Prinsloo & Hofmeyr, 2022). It is necessary, therefore, to identify which mechanisms are implemented by the organisation in response to the need for accountability (Frink & Klimoski, 2004). More importantly, understanding which mechanisms can be implemented for each identified source of accountability will enable the individual to experience the source:

The population consisted of mine managers, front-line mine supervisors and blue-collar mine workers employed in a mining organisation, with knowledge of and expertise in safety management and accountability. Managers are individuals who ‘manage managers’ and supervisors are first-level management or persons who ‘manage others’ (Drotter, 2010, p. 12). Supervisors are responsible for assigning tasks to their direct reports, the workers. Finally, blue-collar workers are individuals who produce results through their own efforts by ‘managing self’ (Drotter, 2010, p. 12); they are skilled, working-class people performing manual work. The above-mentioned categories constituted individual units of analysis.

The individual workers, supervisors and managers from a single mining organisation, employing 1200 employees, were the units of analysis in this study. The mining organisation was selected to mitigate the effects of extraneous variables on participants’ perceptions, such as varying mining company cultures, and different organisational structures or labour relations matters. Participants were selected using non-probability quota sampling techniques involving judgmental and quota sampling (Saunders & Lewis, 2012; Zikmund et al., 2013). The use of a judgment sampling technique required participants in the sample to represent certain characteristics of the population (Saunders & Lewis, 2012). The characteristics were defined as individuals working in high-risk areas within the mining organisation, with knowledge and expertise in the fields of safety management and accountability.

Data were collected from interviews conducted with participants from the three different levels in the mining organisation. A total of 21 face-to-face, semi-structured, in-depth interviews were conducted. Of these, 10 interviews were conducted with blue-collar workers, 5 with supervisors, and 6 with managers. In discovery-orientated qualitative research, it is considered generally acceptable for the source of the qualitative data to originate from a relatively small sample size or a ‘handful of people’ (Zikmund et al., 2013, p. 135). Therefore, a sample size of 21 participants was deemed appropriate. The sample size was selected with the aim of concluding the interviews once no new insights were being revealed (Burden & Roodt, 2007). Saturation was achieved at each of the levels for the given sample size (Burden & Roodt, 2007). Furthermore, the sample size was benchmarked against a related qualitative study and found to be suitable (Pearson & Sutherland, 2017). The interviewer identified potential candidates to be interviewed based on the defined sample characteristics.

Candidates were invited to interviews via telephonic communications and were formally invited through electronic mail. The interviews were conducted face-to-face with the participants. Prior to each interview, the participant in question received an interview guide and a consent form and was assured that their responses would be presented without identifiers. The information gathered from the literature review was used to draft standardised interview questions that served as a guideline for the researcher. The interviewer used an interview guide to facilitate a discussion by asking nine open-ended, non-leading questions (Yin, 2009). The participants were encouraged to answer at will, sharing their own insights and opinions based on their personal experiences. The open-ended question approach was used to support the exploratory nature of the research and encourage open dialogue. The semi-structured interview guide is presented in Table 2 and comprised nine questions. To ensure data credibility, the same open-ended questions were used at each level of participant.

Two pilot interviews were conducted to evaluate the interview guide and the technique used by the interviewer. Adjustments were made to the interviewer’s technique as the analyses of the pilot interviews revealed that the interviewer followed up dialogue points made by the interviewees with questions that were deemed leading (based on the theoretical knowledge gained). The interviewer adjusted techniques by allowing the dialogue to flow freely and by asking clarifying, non-leading questions. The pilot investigation prompted another change being made to the interviewer’s technique, as the feedback revealed that some explanation was required for the term ‘accountability source’. Furthermore, the blue-collar worker pilot interview emphasised the need for the individual to be familiar with the English spoken language to ensure that they arrived at the correct understanding and interpretations. During the data-collection process, notes were made in an attempt to conduct a partial analysis.

The interviews were recorded with the permission of the interviewee and transcribed to form part of the data analysis. Interviews ranged from an average of 28 min to as long as 1 h, which was generally dependent on the organisational level of the individual being interviewed.

The data were analysed using thematic analysis, a common method applied in qualitative studies (Guest et al., 2012) to identify and analyse emergent themes (Gibson & Brown, 2009). The different stages of thematic analysis, as described by Braun and Clarke (2006), were followed. These were:

This section outlines what is understood by the term accountability by individuals working in the mining sector and how it influences an individual’s behaviour. The first interview question dealt with participants’ understanding of the concept of accountability. Table 3 shows the five most frequently mentioned aspects of accountability for each subsample.

The subsamples’ understanding of accountability can be summarised as follows:

Blue-collar workers’ accountability was tied to task completion, for supervisors it was tied to the maintenance of social order in a hierarchical structure, and for managers it was tied to authority and obligation in order to limit liability. Therefore, accountability in mining appeared to have different meanings at different levels of the organisation, thereby suggesting differences in people’s understanding of accountability at different hierarchical levels within the organisation. A closer review of the data, however, indicates that individuals across organisational levels viewed accountability as having a common feature, which is that accountability is an outcome of tasks, order and obligation; it is not a process. Accordingly, the views provided by blue-collar workers, supervisors and managers, respectively, were consolidated to form an overarching definition for the purpose of this study:

The second interview question was designed to determine how a person’s perceived sense of accountability influenced their behaviour. Behaviours were ranked for each subsample, based on the frequency with which each was mentioned. The results are shown in Table 4 and thereafter discussed:

The second research question required participants to identify ‘to whom’ they viewed themselves accountable, that is, the source of their accountability, within the different levels in a mining organisation. The identified sources of accountability per subsample were ranked and are shown in Table 5, along with the main motivation or reason provided for the accountability relationship. Each of the sources of accountability is then discussed in turn:

Interestingly, only one supervisor listed a manager’s leadership skills as a motivating factor in the relationship. Another supervisor argued that leadership skills were not necessary; the fact that someone is a boss automatically gives them authority. Some managers regarded the accountability relationship as ‘more of a personal’, mutually beneficial relationship:

However, what should not be overlooked is the fact that the reputation of a mining operation is heavily dependent on adherence to safety standards. Unsafe practices or careless monitoring (even on the part of just a few people) can – in the event of mining accidents – tarnish the organisation’s reputation in the eyes of the public (Svobodova et al., 2020):

One supervisor questioned the merits of horizontal accountability in the mining sector due to its informal nature, stating: ‘It is not monitored or managed officially on a regular basis’. It was regarded as ineffective because neither party had the authority to pass judgment and prescribe sanctions. Surprisingly, managers did not view their peers as highly influential, thus supporting the findings of Pearson and Sutherland (2017), although one manager acknowledged that he still valued his peers:

The authors of this article used the identified sources of accountability from each organisational level in the mining organisation to compile a conceptual accountability model (Figure 2). The model shows that accountability occurs within a system of interrelationships, both with internal and external sources (Busuioc & Lodge, 2017; Mero et al., 2014). The direction of the arrows represents the ‘to whom’ aspect or source of accountability. For example, managers said they felt accountable to their line managers, their supervisors, their mentors and regulatory bodies. The findings expand on the main accountability sources identified in the literature.

The third research question sought to establish how participants prioritised the identified sources of accountability and the reasons for the prioritisation. Each participant was given 100 points and asked to distribute the points across the sources, with the expectation that sources of higher priority would be allocated a larger share of points. The sources were then ranked according to the percentage of points allocated per subsample. The results are shown in Table 6 and the top three sources per subsample are then discussed:

Participants referred to the limited effectiveness of other sources given the central role played by the individual within a mining context. Interestingly, managers rated themselves significantly higher than any other source. This may indicate that people in powerful positions often disregard the advice of others (De Wit et al., 2017):

Pearson and Sutherland (2017) found that managers were not regarded as highly influential. This study’s findings dispute this assertion at the blue-collar and supervisory levels but support it at the managerial level. Managers rated line managers as only the fourth most effective source of accountability, after regulatory bodies and systems:

Some managers viewed systems (in particular, disciplinary procedures) as a ‘last line of defence’, with one stating: ‘Policing or, according to the company framework, your disciplinary code is there to help keep the people in line who don’t associate with the culture’:

A key finding was that peer accountability appeared to become less effective the higher one rose through the organisational hierarchy of the mine. One manager explained that peer accountability was only possible once an organisation had developed a culture of self-regulation. This may reinforce the argument by Frink and Klimoski (2004) that individuals do not always accept informal accountability systems:

The fourth research question focused on the mechanisms used to drive accountability in a mining organisation (Frink & Klimoski, 2004). Table 7 summarises the main mechanisms for the top three identifiable sources at each organisational level. Each mechanism is then discussed in turn:

Blue-collar workers and supervisors stressed the importance of an individual’s personal values being considered during the recruitment process. One worker said that background checks should be done before people are appointed. Managers listed other factors, such as career aspirations and self-satisfaction, as effective mechanisms to drive accountability.

In addition, much emphasis was placed on informal mechanisms, such as effective communication, informal feedback sessions and the supervisor ‘knowing’ their subordinates, thus enabling them to delegate tasks according to people’s strengths and weaknesses. This was also regarded as an essential platform ‘to gain trust’ – a recurring theme among blue-collar workers and supervisors. The concept of delegating responsibility to show trust in a subordinate’s abilities was also raised. Some considered it a positive reflection of their abilities and felt the need to live up to those expectations. Intrinsic motivating factors such as trust and transparency could therefore be leveraged to reduce the need for external monitoring and control (Mansouri & Rowney, 2014):

The results of research questions 3 and 4 provided the basis for the development of an accountability prioritisation heat map (see Figure 1). The top three sources for the identified organisational levels in the mining organisation and their affiliated mechanisms represented the impact of the combinations. The perceived priority of the sources was depicted on a colour scale, while the relative importance of the mechanism was depicted by the size of the icon. The map can be used by managers, specifically mine managers, as a point of reference to determine the importance and priority of the combinations at each level in an organisation.