The Value of Autoethnography in Leadership Studies, and its Pitfalls

The field of leadership studies frequently focuses on defining leadership traits in abstraction from the context in which leadership operates. The first aim of this article is to provide a brief overview of reasons why this might be the case. Reasons include: leadership studies being dominated by the perspectives of leaders; the lack of definition and visibility of followership studies; the status and limitations of much qualitative research; and a predominant focus on good leadership. Consequently, many people who experience the effects of leadership, and particularly those of bad leadership, do not recognise their experiences in the literature. However, recognising that followers who experience bad leadership personally increasingly engage in autoethnographic studies, my second aim is to draw out some examples of bad leadership from the autoethnographic literature, as well as their effects. In spite of its negative effects, bad leadership frequently remains unchallenged. My third aim is to explain why this might be the case, where I argue that this stems partly from the pitfalls of autoethnographic studies. I also sketch how these pitfalls might be overcome and how doing so and adopting the principle of ‘accountability for reasonableness’ might help to tackle bad leadership.


Introduction
Whilst some claim that management can be separated from leadership in that only the latter would require vision (Bohl 2019), I doubt whether it is a good idea to manage without vision. Vision is unlikely to be the prerogative of leaders, and some leaders may lead without much vision at all. Whilst vision is important for both management and leadership, I understand both to be not about vision, but about power (Laurie and Cherry 2001). Whilst good management is unlikely to flow from bad leadership, I will concern myself here exclusively with the notion of leadership.
What, then, is leadership? Much has been written about the distinction between leaders and followers. Robert Merton (1969Merton ( , 2614Merton ( -2615, for example, speaks of leaders as having 'directive influence' on those who follow them, with followers allowing themselves to be directed as long as leaders' decisions are perceived to lie within the 'zone of acceptability'. Following Merton, the concepts of leader and follower are understood here as polar opposites: the former is defined as a person whose decisions are expected to be followed, and the latter as a person who is expected to follow the leader's decisions. Whilst leaders can behave badly, for example by stealing from their companies, bad leadership behaviour is not necessarily bad leadership. For their behaviour to be classed as bad leadership, Schyns and Schilling (2013, 139) remark correctly that leaders must aim at 'follower-targeted influence'. The fact that I define these concepts as polar opposites does not imply that leaders and followers can always be separated neatly. Additionally, it must also be borne in mind that someone who leads on some matters may also be a follower in relation to some other affairs.
The vast majority of the management and leadership literature simply takes for granted that it is a good thing to have leaders. However, Laurie and Cherry (2001, 5) point out rightly that an important task for the philosophy of management is to question 'the legitimacy of managerial power'. Exploring the notion of leadership is central in this endeavour as it is invariably connected with those of power and influence. In many societies, people perceive themselves and are perceived by others to be situated on different rungs of a social ladder, with people having more decision-making power the higher they are on that ladder. Wilkinson and Pickett (2010, 40) have argued that, all else being equal, health outcomes are worse for societies that have long ladders and that they are particularly bad for those on the lower rungs, writing: 'To do well for yourself or to be successful is almost synonymous with moving up the social ladder'. This raises the questions whether life might be better if we did not have any social ladders at all, as well as whether one is necessarily worse off if one is lower on the ladder. One might wonder whether poor health outcomes should be attributed to the sheer fact that some people have greater decision-making power, or to the fact that some people who have greater decision-making power exert that power over others in inappropriate ways. On this issue, I agree with a silent assumption in the leadership literature, that it is not necessarily wrong to have societies with leaders and followers, even if it may not always be easy to distinguish the former from the latter in light of the 'continuous social flow' (Crevani et al. 2010). Indeed, I believe that it is quite appropriate for people to want to choose and to follow leaders, and that following a leader does not, per se, produce negative health outcomes. What is key is for leaders to be good leaders.
What defines as good leadership has frequently been associated with the possession of particular moral traits, such as honesty (Kirkpatrick and Locke 1991), integrity (Palanski and Yammarino 2007), and charisma (Howell and Avolio 1992). A recent review of effective leadership, specifically in the context of higher education, identified another three traits that would mark out good leadership: ambidexterity (balancing duties appropriately), being knowledgeable, and openness (Weißmüller 2019). Whilst I do not wish to deny that there may be good leadership traits, regardless of the context in which they are displayed, the problem with some of the literature on leadership traits is that it abstracts from particular behaviours and, in doing so, either ignores or neglects the importance of context (Gini 1997;Ciulla 2019). Whereas all good literature on leadership will start from concrete leadership experiences, it is frequently hard to see how they touch base again and connect with these experiences. The abstractions mask concrete reality (see e.g. Betta 2017), thus ignoring what Whitehead (1929, 7) considered pivotal for proper investigation: 'The true method of discovery is like the flight of an aeroplane. It starts from the ground of particular observation; it makes a flight in the thin air of imaginative generalization; and it again lands for renewed observation rendered acute by rational interpretation.' Whilst Whitehead may not have been quite true to this principle himself, I believe that it is a valid point. To highlight the context of leadership, it is therefore important that particular behaviours are scrutinised (Liu 2017). There are many reasons, however, why this has occurred insufficiently.
In this article, I delve into these reasons, highlighting that this is a particular problem for the study of bad leadership, the focus of this paper. I proceed by pointing out that, in spite of this, scholars are increasingly publishing autoethnographic studies to scrutinise bad leadership, where I provide some examples of such leadership, as well as of its negative effects. I continue by exploring why bad leadership frequently remains unchallenged, which I attribute to the pitfalls associated with autoethnographic studies themselves. In the last two sections before my conclusion I sketch how these pitfalls might be overcome and how adopting the principle of 'accountability for reasonableness' might help to tackle bad leadership (Daniels 2000).
Starting with the exploration why leadership behaviours are underexplored, I identify several reasons. Firstly, much of the literature on leadership has been written by leaders, who may have a vested interest in the avoidance of scrutinising leadership behaviour. Whilst Alvesson and Kärreman (2016) may overstate their case in claiming that much of the field is driven not by the desire to understand the moral aspects of leadership, but by the desire to legitimise it, it would be wrong to ignore the possibility that some authors who write on leadership may be driven excessively by the desire to defend or promote their leadership practices. Leaders who scrutinise leadership behaviour do so, by definition, from a privileged position. Whereas the claim that privilege is invisible to those who possess it may not always be correct (Coston and Kimmel 2012, 97), the sheer fact that people embody particular perspectives from particular hierarchical positions may hamper their ability to understand the perspectives of others. Consequently, the experiences of followers, or those who are led by leaders, may not be represented adequately in the policies that leaders make, as well as in the leadership literature.
Secondly, whereas leadership studies have recently been complemented by followership studies (Meindl 1995;Kellerman 2008;Chaleff 2009), there are at least two reasons why these studies shed little light on the perspectives of followers on leadership. The first is that much of this literature is not about leadership, but about good followership (Malakyan 2014, 7). The second is that the field is not well-established, which is partly related to the fact that the word 'followership' is not widely used. This hampers visibility, impairing the ability to research the experiences of followers who have written about leadership, for example by means of research tools such as Google Scholar.
Thirdly, the status and nature of much qualitative research impair the ability to understand leadership. Much social research that gets into the nitty-gritty of understanding behaviours is, by definition, qualitative. Whilst this has been questioned, a long academic tradition favours quantitative over qualitative research, which has also affected leadership studies (Sutherland 2018). Consequently, relatively little research on leadership analyses the detailed, complex perspectives of followers on leadership. In addition, much qualitative research occurs by external researchers taking the initiative to interview followers who belong to particular organisations. As these researchers are distinct from the people whose experiences they report, the fact that they are separated from these experiences entails a significant risk of misrepresentation. Also, as they may not necessarily (be able to) target interviewees with the most interesting perspectives, the experiences from followers who possess the most reflexive experiences about leadership are rarely published. Many of their experiences, therefore, remain invisible to the research community and the wider public (Ford and Harding 2017).
A fourth reason is that, whilst some ethnographic research adopts a different method, prospective participant observation, which may be more likely to report actual behaviour, in most studies researchers remain external to the organisation under scrutiny, participating in the life of the organisation only for the duration of the research project (Sutherland 2018). The fact that there is a dearth of such studies may be related to the fact that unethical leaders are unlikely to grant their permission to being observed by an external party, fearing public disclosure. In addition, a well-known problem with overt observation is the Hawthorne effect, or the possibility that leaders may alter their behaviour in the knowledge that they are being observed (Wickström and Bendix 2000;Bryman 2004). An alternative is covert observation, which is not free from problems either. It may be unethical not to disclose one's true motives, and the fact that the real goal remains hidden may impede access to targeted study populations as well (Lipson 1994).
Finally, as the field of leadership is dominated by the perspectives of leaders, it is also dominated by the search for what defines as good leadership (Harrison 2017;Bryman 2007). As people are aspirational beings, much of this literature is about what leaders ought to do, rather than about what they actually do. Whilst it is by no means irrelevant to explore how leaders might be good leaders, an exclusive focus on good leadership runs the risk of ignoring the fact that how leadership operates in the real world may be far from ideal. Whilst some work has also considered the dark side of leadership (Padilla et al. 2007;Tourish 2013), relatively few studies have focused on it, and there is a particular need for more empirical studies (Harris and Jones 2018;Schyns and Schilling 2013).

The Rise of Autoethnographic Studies of Bad Leadership
In a seminal paper on destructive leadership, the traits that are associated with such leadership are listed as: charisma, personalised need for power, narcissism, negative life themes, and an ideology of hate (Padilla et al. 2007, 180). These would be 'necessary for destructive leadership' (Padilla et al. 2007, 182), in spite of the fact that 'charisma' has also been associated with good leadership (Howell and Avolio 1992). Whilst I have seen some examples of bad leadership, I find it hard to recognise some of these traits, as well as to connect them with leadership practice. The same applies to etiological explanations of bad leadership. Padilla et al. (2007, 182) relate the concept of 'negative life themes', for example, to those of 'parental discord, low socioeconomic status, paternal criminality, maternal psychiatric disorder, and child abuse'. Whilst the view that these might inform bad leadership sounds plausible, it is hard to identify these in, and to connect these to leadership practice. Whilst the authors provide limited evidence for causal links between their themes and destructive leadership, my concern is that, by focusing on them, we fail to understand how these themes play out in leadership practice. The same problem plagues our understanding of the notion of despotic leadership, which has been associated with domination, vengefulness, and selfaggrandisement (Egorov et al. 2019).
To really understand how these concepts inform bad leadership, they must be understood in the context of bad leadership practice. Padilla et al. (2007, 190) write correctly that 'destructive leadership should be studied in its natural ecology, in terms of the interactions among leaders, followers, and contexts'. However, this also raises two questions that have so far been largely ignored. The first is how it should be studied. Whilst traditional qualitative methods rely on interviews or observation, research has found that the themes that are thus generated differ from those created by participants in personal writings, which are rarely studied (Einola and Alvesson 2019). The second question is: who should be doing the studying? The perspective of an outsider looking in is likely to differ significantly from that of an insider, and the perspectives of followers may be quite different from those of leaders too. Whilst the perspectives of insiders and followers have largely been ignored in the literature, this is changing. Both groups increasingly make use of a relatively new qualitative research method, that of autoethnography.
The concept of autoethnographical research has been held to encompass a wide range of meanings, including a research method that values personal experience, reflexivity, and even social justice (Chang 2008;Adams et al. 2015;Jones et al. 2012). Arguably, these feature in many kinds of research. If we focus on the unique nature of autoethnography, however, I understand this method to be about the exploration ('graphy') of an aspect of social reality ('ethno') that draws particular attention to the feelings that reality evokes in researchers themselves ('auto'), from the conviction that their 'own experiences can and should serve as rich sources of data' (Murphy 2008, 169). This concurs largely with the definition of Lapadat (2017, 589), who also highlights its narrative element, referring to it as 'an approach to qualitative inquiry in which a researcher recounts a story of his or her own personal experience, coupled with an ethnographic analysis of the cultural context and implications of that experience'. Whilst no research can take place without being informed by the researcher's own experiences, in autoethnography these experiences take centre stage, an approach that can be immensely rewarding for researchers as well as for their audiences as narratives might foster mutual understanding in ways that more traditionally academic, discursive texts may not be able to do (Harder et al. 2020).
Whilst the traits that authors such as Padilla et al. (2007) and Egorov et al. (2019) have described might be relevant, the autoethnographic literature suggests that bad leaders use a range of strategies that are effective for their bad leadership to emerge and to continue. Whilst these strategies come in many shapes and forms, I would like to pick out a few to illustrate their effectiveness in a bad leadership context. Whilst lying, non-inclusive decision-making, belittling, and unresponsiveness might seem to be unlikely to be good strategies for bad leaders, given that followers are unlikely to be attracted to them, the benefit of (analysing) auto-ethnographic narratives is that they allow us to understand how, in particular contexts, these might nevertheless be good strategies for bad leadership to emerge, to be sustained, and to be reinforced.
A leader with a great need for power can use lying, for example, as an effective strategy to maintain power (Vickers 2007). It can help a leader to maintain power if they do not wish to acknowledge responsibility for something that they have (not) done and that might make them look bad if they own up to it. However, they are unlikely to rise to and to maintain leadership positions if they lie all the time. Lying will only be effective if it is selective, for example when the leader can be reasonably sure that the follower does not have the means to challenge it or is too meek to do so, that they can defend themselves against any allegations, or that the institutional framework is too weak to challenge it.
Another example is exclusive (or non-inclusive) decision-making (Zabrodska et al. 2011). This may be perceived to be innocuous as leaders, by definition, have decision-making powers that are not possessed by followers. However, whilst they may have the power to make some decisions unilaterally, bad leaders also use their positions of authority to make decisions unilaterally where a good justification for doing so is lacking. In doing so, they reaffirm and legitimise questionable leadership and followership positions in the organisational hierarchy. Belittling followers, for example through asking them questions to test whether they can work out why the leader has called for a meeting, through isolating them from those who might empower them, or through the creation of micro-management structures and a constant culture of surveillance, serves the same function (Vickers 2007;Wylie 2020;Zabrodska et al. 2011). Whilst (analyses of) auto-ethnographic studies show, once again, that leaders must be careful about when to use these strategies, Sims (2019, 5) has remarked rightly -even if it might be questioned to what extent this idea is exclusive to neoliberalism -that the neoliberal idea that 'inequity (is) desirable' may facilitate exclusive decision-making as it insulates leaders, to some extent, from critique. It also aids belittlement.
As shown by Wylie (2020), unresponsiveness, the practice of not responding, can also be a useful strategy. In spite of this, a search in Scopus (a large scientific database), conducted on 6 May 2019, yielded no articles for a search using 'leadership' in the title and 'unresponsiveness' in 'any field'. The same search repeated one year later revealed only one study that had been published since 2019 (Einola and Alvesson 2019), which applied the notion of 'unresponsiveness' to followers becoming more passive after selecting a hard-working leader. The reason I think that unresponsiveness can be a particularly useful strategy for a leader, however, is related to the perception that there is decreasing culpability for the following: intentional unresponsiveness; unintentional hierarchy-affected unresponsiveness (forgetful unresponsiveness, caused by the unconscious perception that not responding to those who are lower in the organisation's hierarchy deserves lower priority); and finally, unintentional non-hierarchyaffected unresponsiveness (forgetful unresponsiveness, not caused by the unconscious perception that not responding to those who are lower in the organisation's hierarchy deserves lower priority). The temptation to present intentional unresponsiveness as unintentional nonhierarchy-affected unresponsiveness may be quite great, given that the last is much less likely to be perceived as a significant moral problem.
In an age when much communication happens by email and much email is spam, leaders with significant responsibilities may struggle to separate the wheat from the chaff. In such a context, the attraction of not responding may be particularly great. By not responding, leaders ensure that they can only be held accountable for not responding, rather than for writing inappropriate things. It may also be excused quite easily, for example by leaders saying simply that they are too busy. The fact that much communication occurs by email, however, also presents followers with a good opportunity to challenge leaders. When a leader says something that is inappropriate, it may be hard for the follower to prove that it was said. When there is a written record, the leader exposes themselves to a potential challenge that may be much harder to counter. A related strategy is to respond in a selective fashion, either with the aim to create confusion or with the effect of doing so, for example through the use of 'management speak' (Laurie and Cherry, 2001, 9;Sims, 2019, 5;Loughlin, 2004;Loughlin, 2002).
As many people may identify lying, exclusive decision-making, and unresponsiveness as problematic in many situations, it may perhaps be surprising why some leaders adopt these behaviours. What they have in common is that they leave followers with a sense that their thoughts, values, and feelings do not matter. Leaders are unlikely to adopt these practices when they engage with their superiors in the organisational hierarchy, as they are likely to backfire. Leaders who selectively adopt such behaviours are therefore likely to excel in cognitive empathy, a trait that has been associated with psychopathy. Following Book et al. (2007, 541) one might speak of 'callous empathy', or a 'lack of feeling for others while exhibiting definite understanding of their mental states', used to further 'their own ends'. I am aware of only one contemporary case study of a leader who has been claimed to possess strong psychopathic traits (Boddy 2017). Whilst his behaviour appears to differ markedly from those of others that have been reported in the autoethnographic literature on bad leadership, a common element may be a lack of humility, a quality that has been associated with good leadership (Sadler-Smith et al. 2016).
Why Bad Leadership May Not Be Challenged (More), in Spite of its Bad Consequences The autoethnographic literature on bad leadership has contributed not only to our understanding of what bad leadership is, but also to our understanding of its negative effects. The negative health impacts of lying, non-inclusive decision-making, belittling, and unresponsiveness are significant. Many will perceive that these behaviours are, in many situations, wrong. Bad leaders must therefore live with the thought that some behaviours that they adopt to legitimise their positions are likely to be perceived as problematic by others, which may cause them significant stress. In a competitive environment where owning up to a mistake may be perceived to be a weakness, rather than an opportunity for development, people may learn to avoid challenges from others. Bad leaders may find it difficult to admit doubts and weaknesses, and learn to stand their ground, at all costs. This is how they alienate followers and isolate themselves, where the adoption of defensive attitudes is likely to be accompanied by elevated stress (Lipman-Blumen 2004;Sutton 2007).
Meanwhile, bad leadership also continues to undermine the health of followers, and particularly of those with less institutionalised job security compared to their leaders (Erdemir 2020). Negative effects that have been reported include 'emotional exhaustion (…) reduced family well-being (…) and intention to quit' (Schyns and Schilling 2013, 139). These are likely to result from a feeling of powerlessness which, at least in an academic setting, has been associated with an increased risk of being bullied (Björklund et al. 2020). Unresponsiveness, for example, can instil in followers the feeling that their questions are unimportant, thus potentially undermining their self-confidence as well as their desire to pose awkward questions in the future. Whilst Schyns and Schilling (2013, 139) also report 'resistance behaviour' and 'deviant work behaviour' amongst the effects on followers, submissive behaviour is likely to be much more common. In spite of these negative effects, bad leadership frequently fails to be tackled. Whilst I do not claim to know why this is the case, several reasons may help to explain this.
Firstly, social experiments have shown that people have a strong desire to follow leaders, even if they are prompted to do unethical things by them. In experiments carried out by Milgram (1963), for example, people were, when they were requested to do so by a man with a white coat, more prepared to administer electric shocks to learners whenever they gave the wrong answers to a number of questions that they were asked. The likelihood of follower compliance appears to be correlated positively with leaders' abilities to cultivate followers' capacities to identify with their goals (Haslam et al. 2015), as well as with low levels of job security.
A second reason is that conflicts between leaders and followers are likely to be resolved in favour of the former. Leaders wield some power by virtue of the fact that they have been trusted with the task of managing others by their superiors in the organisational hierarchy. A failure of a leader, therefore, is likely to look bad also on those who entrust them with this position. Consequently, any clash between leaders and followers is likely to be resolved in favour of the leader. This is compounded by the fact that bad leaders who manage to maintain their positions are trained in the art of standing their ground. As followers are likely to be less productive when they are managed badly, leaders' defensive power is enhanced by this lack of productivity. Any allegations of bad leadership made by followers who are less productive because of it may backfire. If leaders are perceived to be able to talk themselves out of any conflict situations, followers are provided with a strong disincentive to challenge them.
A third reason lies in the absence or non-usage of good performance indicators in some organisations (Westhues 2004;Erdemir et al. 2020). Some organisations also include staff who work in many different specialties, where work that is being done in one specialty may be hard to assess by those who operate in different specialties. This is a particular problem in situations where a specialty contains a dyad, comprising one leader and one follower. As few people in the organisation may be able to evaluate the specialty's productivity, the semblance of productivity may be maintained with relatively great ease, particularly by a leader who is trained in standing their ground. Where no hard data on productivity exist or where it is difficult to produce such data, senior management may be unlikely to instigate change on the basis of a perceived lack of productivity. In the absence of accurate data about productivity, any queries about a lack of productivity that might be made by senior management to those who work in any specialty that is occupied by a dyad are likely to be attributed by the leader to the follower, and vice versa. The leader might have a strong case here, due to the effects of their actions being particularly negative for followers (Wilkinson and Pickett 2010).
Fourthly, those who challenge bad leadership may expect even worse leadership. This concern is summarised aptly by Schyns and Schilling (2013, 149), who write that 'followers might shy away from direct resistance to avoid further destructive leadership behaviour and a spiral of abuse'. In spite of these reasons, followers are increasingly challenging bad leadership by means of autoethnographic studies. This takes me to a final reason, the problems associated with autoethnographic studies themselves (Sambrook and Herrmann 2018).

The Pitfalls of Autoethnographic Studies
Whilst some followers have initiated their own studies of leadership by reflecting on the social structures of which they are parts, autoethnographic studies require great courage as they involve exposing one's 'vulnerabilities' (Murphy 2008, 166). This is not a problem that is intrinsic to such studies, but it implies that many autoethnographic stories are never published. Many are deterred from doing so or from exposing their worst experiences of leadership out of fear of negative consequences, for example personal retributions, including losing one's job, a particular concern for those with relatively little job security (Lee 2019). Leaders who engage in unethical behaviour are unlikely to cherish either public exposure or any consequent financial or reputational loss that their organisations may suffer. Consequently, they are likely to disincentivise public disclosure, for example through firing employees who speak out, deterring such disclosures. Another disincentive is the widely held view that public exposure demonstrates a lack of loyalty towards one's organisation (Duska 2007). Accordingly, people may think that public exposure should only be considered in extreme cases, after attempts to remedy the matter internally have failed consistently and the consequences of not reporting are very grave. The public disclosure of questionable leadership by insiders is predominantly associated with the concept of whistleblowing, a term that refers to the public reporting of wrongdoing. As the term's historical usage refers to police officers blowing the whistle to call upon assistance from others when witnessing a crime (Bolsin et al. 2011), many may associate the term primarily with the reporting of criminal acts, even if a wider understanding may be preferable (Vandekerckhove 2006). Consequently, non-criminal, more mundane types of questionable behaviour remain underreported, underexplored, and therefore unchallenged.
In spite of these barriers, some followers have published their autoethnographic accounts of leadership (behaviour), and the number of such studies has grown significantly in recent years. Whilst deciding to publish one's story may be hard, significant difficulties remain after the decision has been made (Bantjes and Swartz 2018;Winkler 2018;Doyle and Buckley 2017). The pitfalls of autoethnographic studies are that those who engage in them land in an ethical quagmire, which creates significant insecurity, at least in part because there has been little discussion about how to resolve these ethical issues. Whilst I tease out some of these problems and provide some clues as to how they might be overcome, these will remain significant pitfalls until autoethnographers address these issues systematically and discuss them in greater detail.
One issue is whether autoethnographers should name those who are implicated by their experiences. I have not come across any autoethnographic studies that name bad leaders. The reason is rarely articulated, but stems presumably from a concern that doing so might produce negative impacts, for example potential reprisals, either for followers or for leaders. Even if followers may decide to take the risk of potential retribution, the decision not to name leaders may be based either on the view that their right to privacy should outweigh the potential benefits that might flow from public disclosure, or on the view that the important message that followers broadcast when they disclose bad leadership does not lie in knowledge about the agent, but in the behaviour that they display. Whilst I sympathise with these views, I do not believe that they justify the view that public disclosure of the leader's identity must be avoided in all situations.
Let me illustrate this with a personal example. Nearly a decade ago I sent a paper off for peer review to an academic journal. Five months after submission, I enquired about it, and I found out that the paper had not yet been sent out for review. Rather, I was asked to resubmit the paper using a different submission system. Eleven and a half months after my original submission, I posted a message on a large academic mailing list, inviting people for a lunchtime meeting two weeks later to discuss editorial ethics, to mark the first anniversary of my paper being under review by the journal in question. I named the journal in my email. Within minutes after posting the message, I received an email from a person on the editorial team, who wanted to phone me to discuss the situation. We conversed over the phone on the same day. I was told that the handling editor experienced personal difficulties and that I would hear about the paper soon. The handling editor mailed shortly afterwards that the paper had been reviewed. Even if I did not mention the name of the handling editor in my mail to the mailing list, it would not have been difficult for any reader to work out who might have been responsible for the problem. I did not ask for her consent before going public. I have no regret about this, and wrote to the editors that I had not wanted to shame anyone, but merely wanted to bring this example of poor practice to the attention of peers so that it might inform their decisions as to where to place their papers. In the same email, I also offered to send another mail to the same mailing list explaining the peer review procedures used by the journal, asking what they wished me to include in that email. Whilst this offer was welcomed, it was not acted upon.
After further reflection, I realise that I probably did want to shame the handling editor. However, the meaning of the verb 'to shame' is ambiguous, given that one can try to shame, but not succeed. Whilst one can name someone who is perceived to do something wrong, it does not imply that they will be shamed. In spite of the fact that I know, being an editor myself, that waiting for a long time for reviewers to provide review reports can be very frustrating, my email exchanges about the paper had left me with the impression that the handling editor did not care much about me. By going public whilst naming the journal, I hoped to achieve that the handling editor would question her apparent lack of care. I underlined that this was important by sharing the experience with a large community of scholars, hoping that she might either recognise that she should have shown more care towards me or be at least concerned about the bad publicity for the journal, and therefore about what others might think about her (Braithwaite and Drahos, 2002). In this case, I think that I achieved at least one of these objectives, and probably both.
It is debatable whether this experience pertains to a leadership context as the editorial team may not have wished to influence my behaviour, even if there is no question that they had power over me. The lesson which can be drawn from this experience, which shows that authors can challenge editors who exercise their editorial power in inappropriate ways by naming them or by sharing information about them that allows them to be identified relatively easily, is relevant in a leadership context. Followers may also have very significant interests that cannot be satisfied unless they name leaders or allow them to be identified easily, even if they might be less likely to do so out of fear, either of worsening relationships, or of personal repercussions. Think for example of a follower who has exploited all internal mechanisms, but who has remained unable to tackle bad leadership. Should we really expect them to avoid naming the bad leader? I do not think so. Doing so runs the risk, at least in some situations, for the bad leadership to continue, and for this to affect many more apart from the follower. Disclosing the leader's name might be motivated by the wrong reasons, for example by the follower's desire to become a leader themselves. However, it might also be motivated by appropriate reasons, and fulfil several useful functions. It underlines the perceived severity of the negative leadership impacts, increasing the likelihood of leaders and others with high positions in the organisational hierarchy taking the allegation more seriously than they did before. This could either lead to leaders being replaced by others or for them to be provided with opportunities for reflection and training. Another benefit is that it provides the public with potentially valuable information about how an organisation is being run, as well as about which individuals it might be best to avoid giving new leadership positions to, for example where leaders intend to move from one organisation to another.
In situations where some of these benefits can be achieved without naming leaders or where these benefits are outweighed by the value of privacy, it is preferable to avoid naming leaders. In autoethnographic studies, however, reported experiences pertain to an identifiable other, which may allow others to be identified by association (Lapadat 2017). Even if others are not named, they may be identifiable. Autoethnographers may therefore decide to publish their work under a nom de plume, and anonymise the organisations that they describe their experiences of as well. Doing so is not free from problems, however, as some publishers may not allow it, and authors who publish under a pseudonym may forgo opportunities to receive academic recognition and critique.
Regardless of whether or not one's experiences are anonymised, a significant problem that besets autoethnographic research is the widely held presumption that consent from those with capacity whose behaviours are reported in research publications ought to be obtained in all, or at least in nearly all situations (Tolich 2010;Gelinas et al. 2016). This raises a problem for studies of bad leadership, as bad leaders are unlikely to consent. In addition, regardless of whether or not they do, some autoethnographers may not be comfortable in asking for consent in the first place. Whilst it is generally assumed that competent people's behaviours should only be reported for the purpose of research if they consent, some autoethnographers do not mention whether or not the people whose behaviours are reported consented, which suggests that they might not have done so (Hill et al. 2019;Wylie 2020). Even renowned autoethnographers are unclear on this (Ellis 2009;Ellis et al. 2011). Others claim that there may be situations where it may not be appropriate to ask for consent. In relation to the latter, Lee (2019) writes, drawing on Muncey (2005), that there may be situations where doing so would imply that 'power remains with the perpetrator and the victim continues to be silenced'. This is highly relevant in the context of bad leadership studies, but I do not think that it is quite correct. It is quite possible to ask the perpetrator for consent, but to publish one's account anyway, even if no consent is given. The fact that several autoethnographers do not appear to think that consent is a necessary prerequisite (Sims 2019;Wylie 2020), however, suggests that there may be situations where it may not even be appropriate to ask for consent in the first place. Others disagree, claiming that consent should necessarily be both asked for and obtained (Tolich 2010).
Where do I stand on this debate? A consideration that is relevant in this context is that much autoethnographic research takes place after the data has been collected. Researchers frequently work from memory. Some people whose behaviours are recorded in one's memory may no longer be alive or be hard to trace. Whilst asking those whose behaviours are scrutinised in one's memories for consent provides them with an opportunity to correct narratives (Ellis et al. 2011), it also provides them with opportunities to distort them. Whilst cross-checking one's data may improve reliability, there are likely to be situations where researchers feel that asking for consent from those whose behaviours they question may elevate tension within one's organisation, as well as personal risk. In light of this, I believe that there are cases where it is entirely appropriate to publish, even without asking for consent. If Schoepflin (2009), for example, had known the name of the person who beat him up, it would seem to be inappropriate to insist that he should ask for his permission to publish his experience. Ellis et al. (2011, 281) suggest that it may 'on many occasions' be one's duty to show one's work to others who are implicated by one's work, 'allowing these others to respond'. I doubt whether Schoepflin would be convinced, and neither am I.
In such cases, as well as in many others, I think that it may be generally advisable to seek approval from institutional review boards or ethics committees. I disagree with Sims's (2019, 3) claim that 'auto-ethnography is not subject to traditional ethic committee approvals', even if there may be situations where seeking approval may neither be necessary nor appropriate, for example where doing so runs a significant risk of researchers experiencing negative health impacts (Carr 2015), or where there is a good case for researchers publishing their research against their advice. Such may be the case, for example, where a committee is overly concerned with 'reputation management', to the detriment of academic freedom (Hedgecoe 2016). In many other cases, however, seeking the approval from a committee may provide researchers with valuable feedback and, if the study is approved, with some reassurance that their research might be appropriate. Regardless of whether or not one seeks or acts on the advice of a committee, I believe that Tolich (2010) is right to suggest that researchers should be prepared for everyone to read their work. I also think that they should be prepared to accept challenges, which in this context includes autoethnographers being prepared to defend their allegations in court if those whose leadership is being scrutinised decide to take legal action against them.
To sum up. Many autoethnographic studies are never published because it would be unethical to do so, and many are never published because it is wrongly assumed that it would be unethical to do so. This stems in part from widely accepted methods of qualitative research, which demand that study participants are anonymised and that they consent, as well as that approval is obtained from a research ethics committee. These demands are questionable. However, until researchers start to dismantle the orthodoxy about how qualitative research should be done, these obstacles will remain significant pitfalls for ethnographic studies. These pitfalls are accentuated in a field like leadership studies where many autoethnographies are unlikely to be published due to concerns that leaders ought to consent, ought not to be named, or that studies ought only to be published if permission from a research ethics committee has been obtained.

The Usefulness of Daniels's Concept of 'accountability for reasonableness' in Discussions of Leadership
In spite of these limitations, autoethnographic accounts of bad leadership are important as they show in very concrete ways how such leadership can emerge and persist in spite of its bad effects. They also provide opportunities for correction. Whilst Tourish (2013, 5) has remarked correctly that 'passivity … provides the unscrupulous with opportunities to manipulate others against their own real best interests', exposure and confrontation provide opportunities for change. In the remainder of this article I develop some guidance on how bad leadership might be prevented and cured, drawing on the work of Daniels (2000), who introduced the principle of 'accountability for reasonableness' as a decision-making tool that is often used to determine what and who should be prioritised in the delivery of health care. Autoethnographic studies are also useful to explore to what extent organisations abide by this principle. Many have applied the concept of 'accountability' to the field of leadership studies (Ghanem and Castelli 2019). However, whilst it has been used in many other contexts as well, I am not aware of 'accountability for reasonableness' having been proposed as a method to prevent bad leadership, in spite of its simplicity and obvious benefits to prevent and tackle such leadership. Daniels (2000Daniels ( , 1301 writes: 'Key elements of fair process will involve transparency about the grounds for decisions; appeals to rationales that all can accept as relevant (…); and procedures for revising decisions in light of challenges to them.' Transparency is important as it makes the invisible visible. If no reasons for decisions are provided or if they are hidden, followers are clutching at straws as to why particular decisions are made. This requires them to expend more energy in challenging a range of potential, speculative reasons, which provides leaders with the opportunity to attribute any allegations that are made to followers not seeing the full picture. Transparency is also important as it allows those who are invisible to become visible. Conversely, a lack of transparency signals to followers that they do not count. Whereas any working relationships can lack transparency, dyadic relationships may be particularly prone to it. The saying 'numquam duo, semper tres' ('never two, always three') has been used in Catholic monasteries to discourage particular sexual relationships, but the specific nature of dyadic relationships has also been studied and compared to other kinds of relationships in philosophy and psychology (Goss 2004, 58). The fact that not many people know about the modus operandi of dyads provides fertile ground for the development and the persistence of a kind of oppressive relationship that also features in many degenerations of intimate relationships between monogamous partners. Whereas the oppressed partner in such relationships may be unwilling to go public for a number of reasons, for example the fear of undermining their economic dependency, their joint responsibility for children, or their willingness to take the rough with the smooth, economic interests may drive followers in dyads to avoid significant disclosure to others within or outside their organisation.
Mere transparency, however, is insufficient. Senior management must also have the desire to scrutinise outputs and working relationships, particularly of those who work together in small groups, and particularly of those who work in dyadic relationships. This takes us to the second of Daniels's criteria, 'appeals to rationales that all can accept as relevant'. I believe that this demand for universality pushes an acceptable position too far. Whilst many may see particular reasons as relevant, they may not be accepted by all. In addition, the majority may not always be right. Nevertheless, I am at one with Daniels that not everything goes, and that it is generally good to demand that people provide reasons for the decisions that they make, and that these reasons are debated. These reasons must be open to scrutiny by those within, as well as by those outside particular organisations. Pace Friedman (1962), no organisation should focus on the narrow financial interests of its members and its shareholders, excluding all other interests, but allow public scrutiny to ensure that private interests promote the common good (Downie and Macnaughton 2001). Neither autocratic nor laissez-faire leadership styles are sufficiently open to constructive critique (Erdemir 2020). This is why organisations must ensure that they adopt and continuously revise processes to evaluate outputs as well as strategies that are recognised as relevant, perhaps not by all, but at least by many, including those outside the organisation. A lack of interest in such processes may be more likely to exist where the financial stakes are insufficiently high (Huisman 2018).
With regard to Daniels's third criterion, which is also called the 'appeals condition' as it refers to the availability of mechanisms to challenge and revise decisions, it is paramount not only for such procedures to be in place, but also for there to be sufficient desire to follow them and to question decisions in light of these. The need for this desire to be present is also recognised by Daniels and Sabin (2002) where they speak of a fourth criterion, sometimes called a regulative, and at other times an enforcement condition (Daniels and Sabin 1998;Daniels and Sabin 2008). It refers to the requirement to adopt rules to offer assurance that not only the third, but also the other conditions are met. Accordingly, dynamic organisations should also review hierarchical positions as a matter of routine to enhance decision-making in accordance with the principle of equality of opportunity and to bolster productivity, paying particular attention to the possibility that those who lack leadership experience may not choose to adopt followership roles voluntarily.

Conclusion
The field of leadership studies has been dominated by the search for character traits that would mark out good leaders. Insufficient attention has been paid to the study of particular behaviours that are associated with bad leadership, the object of this paper. Several factors were identified as contributing to this, including the fact that much of the leadership literature has been written by leaders, a lack of clarity and transparency about followership studies, an academic bias against qualitative research, the fact that much of this research is framed heavily by researchers who are external to the leadership context, the limitations of prospective participant observation, and a predominant focus in the literature on good leadership.
In spite of these obstacles, some qualitative studies of bad leadership exist, and a recent trend is a rise in autoethnographic studies. These studies of bad leadership highlight that, whilst bad leaders may need to possess some traits that have traditionally been associated with such leadership, good strategies are needed for bad leaders to maintain their positions. Examples include lying and exclusive decision-making, as well as unresponsiveness, in spite of the dearth of studies that mention it. Autoethnographic studies also document that bad leadership frequently fails to be challenged, in spite of its bad effects. I provided several reasons why this might be the case, and argued that the pitfalls of autoethnographic studies must be explored more systematically and discussed in greater detail to challenge bad leadership. In spite of these limitations, I argued in the final part that autoethnographic studies may be useful to explore to what extent organisations adopt the principle of 'accountability for reasonableness', which may be a good tool to counter bad leadership. I hope that this article may help to counter bad leadership, as what Souba (2010Souba ( , 1609 once wrote about leaders might also apply to followers like myself: leaders create 'linguistic distinctions that prompt cognitive shifts in others, jarring them loose from their entrenched worldviews'.