As part of the survey we discuss later in this book, people who told us they had been ill-treated at work were asked why they thought this had happened to them. We gave them options which included their age, gender and ethnicity, characteristics of the place they worked and about anyone whom they thought was responsible for the ill-treatment (see pp. 30–1). If none of these options fitted, they were asked to explain in their own words. A few of our respondents said it was just a ‘bad day’ or that someone else was ‘having a bad day’. The best way to explain what this book is about is to tell the stories of a random selection of respondents who had experienced a bad day at work.1 While the facts reported, and the feelings expressed, are given verbatim, we have made up some other details in order to preserve our respondents’ anonymity; for example, all the names used here, and throughout the book, are pseudonyms. We have also imagined everything happened on the same bad day.
At 5.10 p.m., Suhuur, a 25-year-old Muslim woman of Pakistani origin, is collected by her mother outside the shop where she works. As she gets into the car, her mother can see she is upset and, as she eases into the traffic, her mum asks her what’s wrong? You won’t believe what happened today, says Suhuur, and then it all comes out in a rush: ‘A customer asked me for something which didn’t make any sense. When I went to my manager to see if she could help, the customer said a completely different thing to her than what she had said to me. We sorted out what she needed and the customer blamed me for not understanding her.’ How rude, says her mother, but Suhuur says that’s not the worst of it. The customer said it was my fault because of my head scarf. ‘She said to me “you should unwrap that thing from around your ears so that you can hear better”.’
It is now 6.30 p.m. on the same day and Tanya, a 37-year-old black Caribbean woman who has a physical disability, has just arrived home from her job as a manager in local government. As she waits in the kitchen for the kettle to boil, her eldest daughter comes in: you look tired mum, she says, bad day at work? Tanya tells her that her ‘bosses harass me as a result of not meeting the unreasonable deadlines’. Her daughter is only 15 and is yet to have a job. She tells her mother it’s only work and not to take it to heart, but Tanya tells her she doesn’t know how bad it makes you feel when you are ‘unable to meet deadlines owing to unmanageable workloads. This makes you feel incompetent’.
Half an hour after Tanya gets home, Chris rings his brother from a car rental garage in Aberdeen. Chris is white, Christian, 32 and works for an estate agent. He says, did Belinda tell you ‘I flew to Scotland after work’?, and then he tells his brother that he’s ‘expected to drive back [the] same day – 11 hours’ driving’. His brother says it’s ridiculous, probably illegal, but Chris says, I know, I tried to tell them it’s against health and safety but ‘management took no notice, and said I could not stay at a hotel’. Well at least they’ve given you tomorrow off then, replies his brother. You are joking says Chris, ‘when I said “could I come in late next day?” they said no; come in at your normal time’.
Terry is white, Christian, born in South Africa and 36. He is meeting his friend Wayne at the pub at 8.30 p.m. and is running late, so he rings Wayne on his mobile and tells him to get the drinks in. Wayne says, it sounds like you need one, have you had a bad day at the golf course (where Terry works)? Terry says he ‘made a mistake by cutting the wrong piece of grass’. He tells Wayne his supervisor ‘berated me about it, shouting at me, and he assaulted me; he struck me on the jaw with his fist’.
By 11.30 p.m. Ramsey knows he is not going to be able to sleep for an hour or two yet. He is a 43-year-old Christian of Indian origin and he has a physical disability. He checks his Blackberry and sees his brother, who is working in Korea, is already at the office, and he sends him a message to say he has had a bad day at work. His brother will not be surprised by this – he knows that Ramsey has had a long battle with his private sector employer in health and social care but has not had an update on what has happened in the past few weeks. Ramsey says, ‘they said I had been sick for too much time; they then offered me a low daytime job which was not suitable for me. They then offered me redundancy.’ His brother sympathises but tells him that employers do need their employees to be available for work. This is not the point, Ramsey tells him: ‘Staff that came after me had more sickness but no action taken against them.’
It’s 1 p.m. and Nandi is working at the hospital and things are going badly as usual. He is 28, of Indian origin, a Hindu and recently qualified as a doctor. He gets a five-minute break and uses it to go to the toilet and update his Facebook status on his i-phone. He wants it to be, ‘On call – patients to be seen from A & E and then sometimes only two doctors and 20 people to be seen. Employees off, either sick or study – no proper cover’. But he hasn’t time to type all this so writes ‘pressure and stressed’ instead.
It was C. Wright Mills (1959) who taught that it was the job of sociology to explain what bigger structural causes lay behind private troubles like those of Suhuur, Tanya, Chris and the rest, including the 1,788 survey respondents whose accounts we do not have room to discuss here. Mills explained that individuals like them could not hope to understand what was really happening in their lives from their own isolated viewpoint. The virtue of sociology was that it allowed any one of us to step outside the limitations of that individual view and find out if others shared our troubles, and what the common causes of those troubles might be. The private troubles Mills had in mind entailed an element which is common in all six examples above. In each case, a person feels their values are being threatened, and it was this threat that Mills thought could form the seed of the public issue that sociology could help people to fashion from their private troubles.
Examples of the successful translation of private troubles into public issues in the world of employment are easy enough to find. There have been public debates about unemployment, job security, working hours, health and safety, wages, income differentials and discrimination, for example. Unemployment was one of Mills’s examples, but he said that it could be far from obvious how private troubles were turned into public issues, and there might be serious disagreements about the way this was done. We are now in the middle of such a period of debate and disagreement about how best to turn the troubles at work we have just described into public issues.
Mills might agree it has taken a surprisingly long time for sociologists to get involved in this process (Beale and Hoel 2011), but we must first make it clear that this late entry left the way clear for other social scientists to get to work. By far the most important contribution to this work came from psychologists and social psychologists. It is the concepts taken from these disciplines that have drawn together the private troubles of individuals into something that can be measured and investigated, and for which causes and remedies can be found. The two most influential concepts they have introduced are work-induced stress and workplace bullying.
The concept of stress conceives of workplace troubles as excessive strain on employees which impairs their ability to function normally. Ultimately, stress may harm an individual’s mental and physical health. Examples of the kinds of remedies that have been proposed when the private troubles of the workplace are translated into the public issue of work-related stress are provided by the United Kingdom’s Health and Safety Executive (HSE), for example, HSE (2007) which offers guidance for employers. The equivalent publications which follow from the translation of private troubles using the concept of bullying are provided by the Advisory, Conciliation and Arbitration Service (Acas). See, for example, Acas (2006).
This book is mainly concerned with showing what sociology – the latecomer to the debate – can add to the conceptualisation of workplace troubles, and there is no need to review the existing research on stress and bullying. We shall, however, use the remainder of this chapter to show how conceiving of workplace troubles by drawing parallels with the behaviour of school children2 shapes them into a public issue. For example, we shall demonstrate that, while there is often disagreement about who and what to include, the bullying concept omits some of the examples given at the beginning of this chapter. Those troubles, and those individuals, are not to be helped by the construction of the public issue of bullying out of trouble at work. In fact, without more information (about whether the treatment of Chris was part of a long-term pattern of behaviour, for example) all six might not be regarded as bullying. This will not matter if alternative concepts are available, but we are not convinced that any of the available alternatives (including stress – see Walker and Fincham 2011) capture the essence of the private troubles that Mills drew our attention to: the threat to people’s values.
In the next section, we demonstrate how thinking of troubles as bullying defines some troubles in the public issue and some out, though rarely does the debate about what is in and out seem to reach consensus. We shall demonstrate how bullying defines what troubles can be measured, and how to do it, and the available explanations and solutions for the public to pursue. We shall then show how sociologists have finally joined in the enterprise of turning private troubles into public issues. To begin with, they have done this on the ground picked by the psychologists, using the bullying concept to gather data and investigate explanations, but towards the end of the chapter we shall examine how sociologists have begun to find this concept limiting and, perhaps, in need of replacement with something more conducive to what C. Wright Mills famously called ‘the sociological imagination’.
The field was founded on Scandinavian psychological research beginning with the works of Leymann in Sweden (1990, 1996) and Einarsen, Raknes and Matthiesen (1994), Einarsen and Skogstad (1996) and Bjorkqvist, Osterman and Hjelt-Back (1994) in Norway. Proof of the resonance achieved by bullying as a public issue (Einarsen et al. 2011) is easy to find and continues to grow. For example, more than a quarter of the UK newspaper references to workplace bullying in the first decade of the present century appeared in 2010 (Lexis Library). It is worth saying, however, that enthusiasm about conceptualising the issue as bullying has not been universal.
Mills would not have been surprised to learn that some people have been happier to recognise private troubles as bullying at work than others. The way in which the disagreements about the application of the label were played out was demonstrated by some, more sociological, contributions to the field. Liefooghe and Mackenzie-Davey (2001) showed how people’s understandings of the bullying label were complex and derived from different experiences and perspectives, inside and outside the organisation that employed them. Lewis (2003) pointed out that bullying is a socially constructed process in which trade unionists, employees and human resource (HR) managers interpreted the causes and outcomes of bullying in remarkably different ways. Some troubles were only seen as bullying after a process of interpretation in social interaction with co-workers, family and friends. Indeed, McCarthy and Mayhew (2004) argued that some of the patchiness in the adoption of the bullying label might be due to variations in the effort put into raising awareness of it as a public issue.
The concept of bullying also had some competition within the research community, even amongst psychologists. Yet, at the time of writing in 2011, bullying had become the dominant way of conceptualising workplace troubles in many different countries (see, for example, a recent Japanese study by Tsuno et al. 2010). Even in North America – where concepts such incivility, abuse, mistreatment, social undermining and so on have had more support – bullying has gained ground. Sometimes this has happened in conjunction with the concept of harassment, but harassment is also used interchangeably with bullying. In French-speaking countries the same is true of ‘harcèlement morale’, and in some European countries the same is true of ‘mobbing’ (Einarsen et al. 2011).
Why is it so hard to decide what counts as bullying at work? The first question to consider might be whether one can be bullied by accident. Workplace bullying researchers do not agree that there has to be a bully with intent to inflict harm for there to be bullying (Hershcovis 2010). Even if they agree, there are undoubted measurement problems because intent would seem to require verification from the alleged perpetrator (Einarsen et al. 2011). Ignoring the measurement problem, if there is intent, what sort of intent does it have to be? For instance, there has been a lot of debate about whether bullying necessarily implies the intention to harm (Einarsen et al. 2011). Establishing this might create even more challenging measurement problems (Nielsen, Notelaers and Einarsen 2011). The practical solution to these problems has been to look for circumstantial evidence of intent. Einarsen (1999), who has done more than most to define the field, argued that bullying occurs regularly and carries on for a sustained period, and attracts general agreement that it is aggressive behaviour that is intended to be hostile or could be seen as such by the person on the receiving end.
This definition of bullying could rule out one or more of the private workplace troubles described at the beginning of the chapter, even though some of the people who experienced them might consider them to be bullying. Indeed, it is possible that we would not even have gathered these data at all if we had been relying on Einarsen’s definition. If researchers are not interested in irregular behaviour, or things which have happened a few times, or which are not obviously aggressive and hostile, they do not count them. Einarsen and Skogstad (1996) found the mean duration of bullying to be 18 months, while Zapf et al. (2011) showed a mean duration in their meta-analysis of between 12 and over 60 months. We do not know from these statistics whether shorter periods of ill-treatment – which the researchers made sure they did not count – were also experiences that people considered to be bullying.
Einarsen further refined the definition of bullying after 1999, and his refinements were widely adopted in the field. There is now widespread agreement that workplace is ‘harassing, offending, or socially excluding someone or negatively affecting someone’s work’. Again, it has to be sustained (for six months or more) and it has to be frequent (say once a week), but the elements of aggression and hostility are de-emphasised in favour of others:
Bullying is an escalating process in the course of which the person confronted ends up in an inferior position and becomes the target of systematic negative social acts. A conflict cannot be called bullying if the incident is an isolated event or if two parties of approximately equal strength are in conflict. (Einarsen et al. 2011: 22)
The notion of power disparity between perpetrator and victim was first introduced by Leymann (1996), and there seems to be agreement that it works best for the less formalised sources of power that exist simply because of personality factors which make one person more dominant than another (Einarsen 1999). Hoel and Cooper (2000) suggested that horizontally derived power from co-workers can be exploited through personal knowledge of a victim or through group behaviours that target some power deficit (Einarsen et al. 2011).
Applying the notion of power distance to relations between those who exercise formal authority and those who must do as they are told is, of course, more difficult. It also makes it very difficult to see how relations between employees and customers or clients could be counted as bullying. Suhuur (p. 3) had no power over her customer (but of course her troubles would have been ruled out anyway because they were not regular or sustained). But imagine a social worker who has regular and sustained contact with a client, and with every interaction the social worker remembers how unpleasant and sometimes terrifying the other person is. But when the social worker has power over the client, his or her troubles cannot count as bullying even if they have such a profound negative effect on his or her ability to do a good job (Denney 2010).
The manner in which we conceive a public issue not only rules some things in and some out, creating various measurement challenges, but the same principle applies to explanations and solutions for the public issue. One obvious place for psychologists to look for explanations for bullying is personality characteristics (Harris, Harvey and Booth 2010). For example, Baillien et al. (2009) identified differences in the capacities of bullies and their targets to cope with frustration. Coyne, Seigne and Randall (2000) showed how bullied victims were less extrovert, submissive, averse to conflict, quiet, reserved, less stable and more conscientious. There is no agreement, however, that personality profiles or psychological coping mechanisms are the right place to look for explanations (Milczarek 2010). Zapf (1999) and Zapf and Einarsen (2011) argued that bullying can have multiple causes and that personality is only one element. Indeed, a central strand of Leymann’s original argument was that personality traits of anxiety were ‘a result of and definitely not the cause of exposure to bullying’ (Glasø et al. 2007: 2). In a matched sample of victims and non-victims, Glasø et al. (2007) showed how victims displayed more neurotic and less agreeable behaviours, but two-thirds of the victim sample did not differ from non-victims in their personality profiles.
In their search for explanations, psychologists have also looked beyond the characteristics of individuals to the character of their relationships. Einarsen (1999) had seen bullying as a gradually evolving process, starting with aggressive behaviour developing into bullying, stigmatisation and severe trauma. Numerous models have been produced to understand the manner in which conflict becomes bullying (Zapf and Gross 2001). The notion of bullying as a dysfunctional interpersonal dynamic has been extended beyond the dyad to include the group that is affected as the process of action and reaction continues (Tehrani 2011). Heames, Harvey and Treadway (2006) also saw dysfunctional group dynamics as present at the start of the process, for example because people do not agree on their relative status (also see Baillien et al. 2009, and see p. 10 on role conflict).
Some psychologists have extended the enquiry beyond groups of employees to consider the workplace itself as a possible explanation for bullying. Leymann (1996) had stressed the importance of the work environment as an explanation for bullying from the start. He considered it much more fruitful to investigate the ways in which work was organised, and leadership was displayed, than looking at the personality characteristics for bullies and the bullied. Especially in Scandinavia, researchers followed this lead (Einarsen et al. 1994; Hauge, Skogstad and Einarsen 2010; Vartia 1996; Zapf, Knorz and Kulla 1996). For example, in the spirit of Leymann’s original thesis, poor leadership and management appeared as key elements of the explanations offered by Vartia (1996) and Einarsen, Aasland and Skogstad (2007). The destructive aspects of leadership outlined by Einarsen, Aasland and Skogstad (2007) occurred not simply because leaders were purposively destructive but often because of inaction and poor management of events on the ground, characterised as ‘laissez-faire’ leadership. More dictatorial forms of leadership were proposed by Ashforth (1994), who labelled this ‘petty tyranny’. The autocratic leadership which was shown to be most prominent in a British study was also thought central to the explanation of bullying (Hoel et al. 2010). Salin and Hoel (2011) argued that, whereas autocratic or laissez-faire models of leadership could lead to bullying, a participative approach was much less likely to do so.
While an interest in leadership might betray the psychological bias of most research, Vartia (1996) found poor communication, and lack of participatory structures, to be important factors. Other factors that have been found to lead to bullying include high workload (e.g. Agervold and Mikkelsen 2004; Appelberg et al. 1991; Einarsen and Raknes 1997), low job control (Einarsen et al. 1994), role ambiguity (e.g. Vartia 1996) and job conflict (e.g. Einarsen et al. 1994; Notelaers and De Witte 2003). The last two concern whether an employee believes he or she should be working in a different way, might be doing things that are not necessary or is doing things one person thinks right and another does not (Einarsen et al. 1994; Hauge, Skogstad and Einarsen 2007; Vartia 1996). It is important to clarify, however, that the psychological paradigm does not suggest that these things in themselves constitute bullying. Rather, the idea is that an employee who experiences role ambiguity or job conflict will have a lower threshold for bullying (e.g. Einarsen et al. 1994).
From the perspective of organisational psychology, all of these factors contribute to the work environment, making bullying more likely to occur (Beale and Hoel 2011), but the existence of role ambiguity and job conflict (for example) do not necessarily imply that there is anything wrong with the way work is allocated and managed. Salin and Hoel (2011) saw things from a slightly different perspective when they suggested that work design, along with organisational culture and organisational change (see below), is closely correlated with episodes of bullying. They argued that bullying thrives where there are contradictory expectations, demands and values.
Work intensification in the form of increasing job demands and pressure of insufficient resources has been associated with bullying (Baillien et al. 2011). Other aspects of organisational change have also been shown to be highly correlated with bullying. More bullying is reported when there is more change taking place (Hoel and Cooper 2000; O’Connell, Calvert and Watson 2007; Skogstad, Matthiesen and Einarsen 2007). For example, a change of manager or more widespread restructuring have been shown to be associated with bullying (O’Connell et al. 2007; Salin and Hoel 2011). Skogstad et al. (2007) found an association between bullying and changes in work tasks and workplace composition. Skogstad et al. found that change might be associated with bullying because it caused conflicts between employees and managers, but that change also had an independent influence on bullying. Baillien and De Witte (2009) found no evidence of an independent influence and that the whole effect was mediated through role conflicts and job insecurity (note De Cuyper, Baillien and De Witte (2009) thought bullying caused insecurity rather than being caused by it). In both cases, the effect was a psychological one. When individuals found change had negative outcomes for them, this elicited victimisation.
Other researchers have been more interested in the possibility that organisational change may have a more direct relationship with bullying. Hoel, Cooper and Faragher (2001), for example, found an association with bigger, even global, shifts than those taking place in a single organisation, particularly restructuring and downsizing (Hoel et al. 2001). Along with others, they wanted to raise the possibility that these pressures led to work intensification and then to bullying (e.g. Harvey, Treadway and Heames 2006; Salin 2003). Salin (2003) argued that increased pressures on resources and restructuring can lead, for example, to increased competition between managers and all manner of local political struggles which make bullying more likely. In a similar vein, researchers have argued that, with or without organisational change, some workplace cultures can be particularly conducive to bullying. Thus Harvey et al. (2009) suggested that the reaction of others in the workplace to bullying sets the parameters for what is deemed acceptable and can encourage bullying to continue within the organisation.
In order to study the effects of workplace culture, Salin and Hoel (2011) suggested focussing on socialisation processes, communication and social climate as well as interpersonal conflicts. They cited Strandmark and Hallberg (2007) on the professional and value conflicts underpinning power struggles. Much earlier, Baron and Neuman (1996) had suggested increased workplace diversity (creating difficulties in interpersonal communication), feelings of anxiety and anger brought on by work practices such as increased computer monitoring, feelings of unjustness and unfairness related to pay cuts and unpleasant working conditions could also be conducive to bullying. A decade and a half later Einarsen et al. (2011) presented a theoretical model for the management and study of bullying which comprised cultural, socioeconomic, organisational and individual elements (including the characteristics of victims). The point of studies such as these was that organisational culture could not in itself encompass bullying, but that it could provide an environment in which bullying flourished (Agervold 2007).
In a review of the existing literature, Milczarek (2010: 11) concluded that ‘in most of the cases of bullying, at least three or four of the following can be found: problems in work design (e.g. role conflicts); incompetent management and leadership; a socially exposed position of the target; negative or hostile social climate; and a culture that permits or rewards harassment in an organisation’. There is, in such arguments, also the potential to shift the focus away from the bullying that co-workers might subject each other to under stress. Moreover, as the knowledge base has grown, bullying has become less firmly located in a person or even a relationship. Indeed, the definition of bullying from Einarsen et al., which we quoted above (p. 8), included a rider about applying the label to ‘a particular activity, interaction or process’ (Einarsen et al. 2011: 22).
Researchers such as Liefooghe and Mackenzie-Davey (2001), Hoel and Beale (2006) and D’Cruz and Noronha (2009) have argued for bullying to be seen not solely as an individualised construct but also to be recognised as an organisational one. In this regard, it is not bullies but organisational practices and processes that create the private troubles. Thus D’Cruz and Noronha concluded that it was the organisational practices of Indian call centres that demeaned and abused their employees. Lopez, Hodson and Roscigno (2009: 24) even pointed to ‘routine organisational activities’ as the locus of bullying. Much of the foregoing emphasis on organisational change, which we have been presenting as environmental factors causing the appearance of bullies and bullying, might actually be interpreted as processes which bully, even as evidence of bullying organisations. Here then, we have the seeds of a more sociological approach to workplace bullying. In the first instance this approach developed as a specialised application of a long-established sociological interest in industrial relations.
For most of the time workplace bullying has been a public issue; it has been an article of faith that workplace bullying is not a standard industrial relations issue (Expert Advisory Group on Workplace Bullying 2005). Yet Fevre et al. (2009) have shown that most employees who report bullying or harassment also experience other problems with employment rights. A substantial number report unfair treatment and discrimination, including employment rights problems such as troubles with pay, health and safety grievances, hours of work, sick pay or leave, contracts and so on. Is it the job of researchers to carefully isolate bullying from these other troubles so that measurement, explanation and remedies do not become contaminated by them? Or should researchers be trying to understand if there are common causes of a wider range of workplace troubles?
Analysing a large number of ethnographic studies, mainly undertaken by sociologists, led Hodson (2001) to conclude that insufficient employee participation was a feature of many cases of ill-treatment. Mismanagement was a major cause of abuse but its effects could be tempered where managerial power was shared with workers. This would benefit organisations as a whole because managers might do anything, including ill-treating employees, in order to increase profit, whereas workers were interested in productivity and quality. In a later article, Hodson, Roscigno and Lopez (2006) argued that chaos in the workplace was a catalyst for bullying and harassment. In part, this argument recalled the environmental theories of organisational psychology since it was assumed that chaos created opportunities for bullying and harassment of those employees perceived to be weaker.
The same authors returned to the theme in Roscigno, Hodson and Lopez (2009) where they argued that organisational chaos was central to a sociology of bullying, although it did not bear the same relation to all types of ‘workplace incivilities’. Moreover, these researchers argued that chaos was immanent in all workplaces because there was always the potential for tension between the goals of managers and workers. Rationalities on one side looked like irrationalities on the other:
[I]rrationalities may easily be experienced as chaotic by those involved because the link between known causes (e.g. effort and accomplishment) and rewards (e.g. security and advancement) are disrupted. In the resulting normative vacuum, control and co-ordination can revert to a reliance on bullying rather than use of positive inducements. (Roscigno, Hodson and Lopez 2009: 761)
Roscigno, Lopez and Hodson (2009) claimed that bullying which accompanies mismanagement and chaos can be dealt with by interventions from trades unions, for example, and by appropriate adoption of policies and practices. This conclusion might be seen as over-optimistic in Scandinavia, and the United Kingdom, where considerable evidence of bullying existed even though such policies and practices were well established (Rayner and Lewis 2011; Salin 2008), and trade unions and professional bodies, such as the Chartered Institute for Personnel Development (CIPD), and government agencies such as Acas, gave free advice and guidance to organisations and individuals to help them deal with bullying. On the other hand, Norway was one of the first nations to address bullying, and the experience there suggests that a reduction in bullying is possible through judicious use of interventions, including legislation (Nielsen et al. 2009). In multivariate analysis of their representative Irish sample, O’Connell et al. (2007) showed that there was less bullying in organisations with formal policies on bullying. However, in a similar UK study, Fevre et al. (2009) showed that trade union members were more likely to report bullying.3
Like Roscigno and his colleagues, Ironside and Seifert (2003) and Hoel and Beale (2006) concluded that workplace bullying should be dealt with through the industrial relations machinery. Comparing British and Swedish employers, Beale and Hoel (2010) found British managers to be more likely to intervene to prevent bullying because bullying in Sweden was most often seen as a dispute between employees, rather than between managers and employees, and because legal regulation of bullying was more explicit in Sweden. Beale and Hoel (2011) followed Ironside and Seifert’s lead in focusing explicitly on the collective dimension of bullying. Bullying had a purpose – to reshape employee behaviour – and was therefore endemic to capitalist employment relations. So, despite the apparent evidence of the costs of bullying to employers, Beale and Hoel (2011: 14) proposed that employers benefited from bullying.
Rafferty went so far as to argue that bullying may be a tool chosen by employers to control their staff and that this is why it is so often associated with organisational change: ‘restructuring and downsizing can magnify power imbalances and job insecurities, and encourage an atmosphere of corporate bullying. Changes of management or ownership in business can also lead to the use of bullying tactics to sweep out existing staff’ (Rafferty 2001: 102). Hoel and Beale concluded from the British case, where managers were responsible for much workplace bullying, that, at the least, they would defend each other when accused of bullying, and that initial senior management sympathy towards employees who had been bullied would not lead to action when it counted. This would make employees cynical about the fashion for high-commitment human resource management (Beale and Hoel 2011).
Sociologists are not simply interested in bullying because this seems to complement their long-standing interest in industrial relations. For example, sociologists of work and occupations might naturally be interested in any evidence that suggested bullying is more common in some jobs than others. Indeed, such occupational differences have been observed, though this observation has rarely been incorporated into a convincing theory of patterns of bullying. In the review of the literature compiled by Milczarek (2010), nurses reported a higher incidence of bullying than many other occupations. Health care workers in general, and teachers, were amongst those who might be forced out of their jobs by bullying (McCormack et al. 2009; Quine 1999, 2002). University employees (Bjorkqvist et al. 1994), civil servants or those working in public administration (Rayner 1997) have all shown relatively higher levels of bullying, while Roscigno, Lopez and Hodson (2009) showed how bullying can be the product of low occupational roles and positions. Fevre et al. (2009) showed that, in multivariate analysis of a representative sample of UK employees, those with more than one job were more likely to report bullying or harassment, but in their study, as in others, there was no evidence that bullying is more common in lower paid occupations. In multivariate analysis of a representative sample of the Irish workforce, however, O’Connell et al. (2007) showed that plant operatives and casual workers were more likely to report bullying, but this research also showed that bullying was more common amongst employees with higher levels of education.
Given the prevalence of bullying amongst nurses and teachers, it is no surprise that both education (Hubert and van Veldhoven 2001; Leymann 1996; Zapf 1999) and health and social care (Piirainen, Rasanen and Kivimaki 2003) have been shown to have higher rates of bullying in studies conducted in Sweden, Germany, the Netherlands and Finland. In their multivariate analysis, O’Connell et al. (2007) showed that bullying was more common in education, public administration, personal services and transport in Ireland. Citing one of their earlier works, Zapf et al. (2011) showed how a study of 400 German workers who reported serious bullying had a sevenfold risk of being bullied if they came from health and social services sectors, with a threefold increase for public administration workers and those employed in education. Zapf et al. (2011) cited some of the earliest studies of bullying in Sweden by Heinz Leymann, who reported an ‘over-representation’ of bullying in educational, health and administrative sectors. Leymann and Gustafsson (1996) showed in their study of post-traumatic stress disorder (PTSD) and bullying that the largest groups of patients came from health, education and social services occupations and that private sector organisations were under-represented.
Other studies found that the public sector as a whole exhibited a greater propensity towards bullying. Public sector workers were more at risk of bullying behaviours in Finland, according to studies by Vartia (1996) and Salin (2001). In the United Kingdom, Hoel and Cooper (2000) demonstrated how prison services, policing, education, health and local councils were high on the list of organisations where bullying behaviours were prevalent. Lewis and Gunn (2007) showed how UK public service workers drawn from across 13 public organisations experienced a range of negative behaviours at work, ranging from being given demeaning tasks through to humiliation and excessive criticism. In both the Lewis and Gunn (2007) study and the Hoel and Cooper (2000) research, managers were the most likely source of bullying behaviours with colleagues a distant second. In the case of the United Kingdom, for example, ‘new public management’ in the public sector has been associated with bullying and harassment (Burnes and Pope 2007). Other researchers have suggested that the higher rate of bullying in the public sector may be related to the extent of public sector change (Beale and Hoel 2010; Ironside and Seifert 2003; Salin 2001) and the nature of public sector employment (Zapf et al. 2003). Salin (2001) pointed out, however, that the public sector has received more attention from bullying researchers, perhaps because of the comparative ease of access to employees in that sector. McCarthy and Mayhew (2004) also suggested that the higher rate of bullying observed in the public sector may be a product of increased awareness there of bullying through policy initiatives (which also raised people’s expectations of the standards they expect of behaviour in the workplace).
It is not just bullying that has been observed to vary by sector. Milczarek’s review (2010) found that some of the same sectors in which bullying is believed to be more prevalent also experience more violence in the workplace from third parties such as clients. Substantially higher risks of workplace violence have been observed in health care and social work, education and public administration and defence (but also in commerce, transport, and hotels and restaurants). If there are common patterns between bullying and violence then there is no good reason, perhaps, to exclude workplace violence from the field (Einarsen and Raknes 1997).
As with bullying, some studies of workplace violence have looked for explanations in the personality characteristics of perpetrators and victims (Zapf and Einarsen 2003). Other writers have emphasised factors relating to the nature of work organisation and management processes (Hodson 2001; Neuman and Baron 2003). Again in parallel to research on bullying, attention has been given to the impact of autocratic management and, of course, organisational change, including general change brought about by growing global competitive pressures, work intensification and related stresses on the social relations of work (Bowie 2010; Hoel and Salin 2003). The theories of Ironside and Seifert, and Beale and Hoel, about the predisposition of capitalist employment relations to stimulate bullying also have their counterparts in theories of violence in the workplace advanced by Bowie (2011) and others. Bowie also identified the roots of workplace violence in inconsistency in management policies, poor communication, ineffective grievance procedures, ‘perceived unjust treatment of employees, lack of mutual respect among separate work teams and departments, ethnic tensions, increased workloads with diminishing resources and rewards, and poor working conditions and security’ (Bowie 2010: 47).
Despite the obvious similarities between the theories advanced by writers such as Bowie and those elaborated to explain bullying, we should not assume that the violence that these writers are trying to explain always occurs between employees or even managers and employees. Critical criminologists such as Tombs (2007) have argued that harms to workers (and the general public) are much more widespread and deeper rooted than can be gauged by levels of interpersonal assault, and emerge from the structures and processes of employment relations and capitalist production. Catley and Jones (2002: 25–8) drew analytical distinctions between acts of physical interpersonal violence, violent speech acts, structural physical violence and structural ‘symbolic’ violence. In a similar way, Estrada et al. (2010) identified four separate categories of work-related violence. ‘Intruder violence’ includes crimes of violence against, for example, bank employees or check-out staff, and ‘client-related’ violence is taken to mean physical assaults by customers, patients or clients. ‘Relational violence’ denotes violence and harassment between workers in the same workplace, and ‘structural violence’ includes the broader systemic aspects of workplace harm emphasised by criminologists such as Tombs. The more specific aspects of organisational structures and cultures that expose workers to violent situations can also be seen as ‘structural’ forms of violence.
Bowie argued that organisations can bear responsibility for workers’ exposure to client-related violence through placing employees and clients in potentially violent situations. He expressed particular concern about many workplaces which are affected by this and other aspects of structural violence:
are part of the so called caring professions such as health, education, and social welfare. … Often organizations where you would expect, as an employee or patient, to be treated with dignity and respect are in fact the opposite and hide an economic rationalist agenda under a veneer of service. Such abusive behavior by organizations is coming to the forefront of the current debate about healthcare provision and distribution. In such situations employers and managers might argue that they are not to blame for bad supervisors and related practices, deflecting criticism back onto supervisors and their workers. There is often no recognition or denial at the higher echelons of management regarding how the organizational climate and functioning can allow or foster a violent work environment. (Bowie 2010: 52)
There has been considerable research on client-related violence in health and social work occupations, including some recent, mixed methods research on violence against doctors, probation officers and ministers of religion. Denney and O’Beirne (2003) described the way in which managers in the probation service gave little, if any, thought to preventing violence from offenders against probation officers. Indeed, there appeared to be no acknowledgement of this violence in the training and management of probation officers except through ineffectual and piecemeal responses – for example, offering counselling – which were made when violence did occur. Elston et al. (2003) found that probation officers and ministers of religion who had suffered violence believed there were ways in which their jobs could be designed and resourced to make them safer. The research team found that, without such responses, the fear of violence might have many implications, including ‘avoidance of mandatory work, ignoring possible risks, failing to take sufficient precautions for safety and eroding staff confidence and morale’ (O’Beirne, Denney and Gabe 2004: 124).
The same research team which described the ways in which managers ignored violence also described the ways in which professionals who suffered violence often minimised it. O’Beirne et al. (2003) found that probation officers and ministers of religion might not disclose violence to their colleagues, still less to the police. Clergy who suffered violence did not even have a formal reporting structure available to them. Probation officers failed to report violence because they believed they would get little help from their managers, and members of both professions believed violence was something they were expected to take in their stride. As part of the same research project, Elston et al. (2003) reported that doctors medicalised much violence by patients, which was therefore seen as a part of their job rather than a crime. They might also be given mixed messages by other professionals; for example, a psychiatrist might encourage a doctor to report violence by a patient in order to increase the chances of that patient being treated while the police would advise the opposite.
As with bullying, variations in the conception of the public issue lead to confusion over how best to measure workplace violence. We have discussed elsewhere the contrast between levels and trends in workplace violence in crime victimisation surveys, on the one hand, and broader studies of workplace relations, on the other (Jones et al. 2011). For example, criminological studies based on the British Crime Survey (BCS) have suggested a low (and falling) risk of workplace assault in recent years (Budd 1999; Upson 2004). In contrast, studies from the field of management studies suggested that workplace violence has been growing in frequency and severity (Chappell and Di Martino 2006; Flannery 1996; Serantes and Suárez 2006). Crime victimisation surveys tend to focus respondents’ minds onto formal legal categories of assault, and thus filter out various forms of violent behaviour – both serious and relatively minor – which for various reasons are less likely to be the subject of formal action (Jones et al. 2011).
It seems clear that criminological studies substantially underplay the significance of workplace violence in contemporary Britain, and as we shall see in Chapter 4, far greater numbers of workers than those suggested by the BCS are victims of violent behaviour in the workplace. That said, it may be that some estimates of spiralling workplace violence err too far in the opposite direction. Indeed, the findings of workplace surveys may reflect a lowering of the threshold of tolerance for certain forms of behaviour and a greater tendency to label problematic or otherwise harmful behaviour as ‘violence’. Estrada et al. (2010) argued that apparent increases in workplace violence in a number of countries were related both to a greater general awareness and sensitivity to such issues, and a real increase due to changing working environments that exposed more workers to risks of violence.
Research on the sociology of organisations has explored the relationship between structural power inequalities and harassment and violence in the workplace, and has often suggested that the latter bear down more heavily upon those social categories traditionally perceived as marginalised or disadvantaged (Hearn and Parkin 2001). Roscigno, Hodson and Lopex (2009: 760) found that ‘both gender and minority status are significant determinants of not only sexual harassment but of managerial bullying as well’. Lopez et al. (2009) argued that the defence of identity and of jobs was at the core of bullying and harassment and that bullying was therefore bound up with formal and informal status hierarchies and job security. Minorities were more at risk because the biggest power differentials between employers and workers existed where the workers were drawn from minorities, and because isolation and exclusion in wider society made minorities into targets (Lopez et al. 2009: 20). This was, in part, a special case of powerlessness – those who suffered general social exclusion were more vulnerable in the workplace – but it also said something about the way non-minority workers reacted to sharing a workplace with minorities:
In some kinds of situations, the association between harassment and minority workforce may be a consequence of minority workers’ concentration in bad jobs, but in other settings, it is clear that harassment of minority workers serves white identity and job-protection functions in much the same way as general and sexual harassment serve male identity and job protection. (Lopez et al. 2009: 21–2)
Particularly conducive settings for harassment were those which involve physically demanding work. In any event,
[bullying or harassment was part of a] larger process of social exclusion and closure. In this process of closure, mocking, barriers, and sometimes blatant threats are used to exclude certain groups (even potentially forcing them out of the workplace) or to keep members of these groups ‘in their place’. (Lopez et al. 2009: 23)
Hodson et al. (2006) suggested that it was not simply whether a respondent was a member of a minority or not, but what proportion of the workforce was made up of minorities, that affected the prevalence of bullying and harassment.
The theories elaborated by Hodson and his colleagues were founded on the assumption that bullying and harassment were more prevalent amongst minorities. They felt that what they learnt from analysis of a large number of workplace ethnographies (see p. 12) justified this assumption (Lopez et al. 2009: 15; also see Hodson et al. 2006). However, these studies were completed over a fairly long period during which patterns of overt prejudice and discrimination certainly changed, not least because of changes in legislation. Moreover, the ethnographies Hodson and his colleagues drew upon in their work were conducted in more than one society. It is possible that their method is therefore not sensitive to differences in workplace behaviour between different societies which result from variations in patterns of overt prejudice and discrimination beyond the workplace. As before, such differences may well result, for example, from differences in the anti-discrimination legislation in place in different societies.
Lopez et al. (2009: 24) were certainly aware of a need for additional representative studies. What other evidence do we have that employees who are members of minorities are more at risk of bullying, harassment or violence? Fox and Stallworth (2005) argued that research in this areas should be careful about distinguishing supervisor from co-worker bullying (also see Lewis and Gunn 2007), and reported their US study in which Hispanics were the only minority to report higher levels of ‘general bullying’ than whites, although all ethnic groups reported racial or ethnic bullying such as taunting and other forms of ill-treatment. Researchers in the United Kingdom claimed correlations between ethnicity and bullying, including racial or ethnic bullying (Hoel and Cooper 2000; Lewis and Gunn 2007). The representative study reported by Fevre et al. (2009) did not, however, find a correlation between bullying or harassment and ethnicity in multivariate analysis which controlled for other factors.
According to Fevre et al. (2009), women were 73 per cent more likely to report bullying or harassment. Other researchers have explored gender and bullying, but with mixed results dependent upon methodology and sample size, and the evidence presented in meta studies does not demonstrate such a clear gendered component. Thus research in Scandinavia and the United Kingdom showed that men and women had similar levels of exposure to bullying (Zapf et al. 2011). Where women are over-represented in bullying studies, this will be because occupations or sectors with a female majority have been sampled (Zapf et al. 2011). This could mean, however, that women are more exposed to bullying but because of the jobs they hold rather than because they are targeted for their gender (Hutchinson and Eveline 2010; Lee 2002). Hutchinson and Eveline (2010) also argued that power and hierarchy in organisations was often underpinned by a gender component. Rodríguez-Muñoz et al. (2010) reported just under half of 183 victims of bullying showed symptoms for the criteria for PTSD with women more likely than men to show these symptoms. Milczarek’s review of research on the correlates of workplace violence suggested that men might be at higher risk of third-party violence and that men and women would usually encounter violence in different employment situations. Women were most at risk in health care, education and retail, whereas men were most at risk in police and security work, and transport (Milczarek 2010).
According to the representative study of British employees reported by Fevre et al. (2009), lesbian, gay and bisexual (LGB) employees were 271 per cent more likely to report bullying or harassment. This confirmed an early representative study (Grainger and Fitzner 2007) and lent credence to the suggestion from Hunt and Dick (2008) that nearly one in five lesbians and gay men experienced bullying due to their sexual orientation, with one in eight of the population reporting that they have witnessed verbal bullying of gay people in the workplace, whilst nearly one in 20 witnessed physical bullying (see also Croteau 1996). Acas (2006) suggested large minorities of LGB employees have some experience of bullying/harassment. Threats of physical abuse to LGB employees also feature in these studies (Acas 2007; Hunt and Dick 2008).
Research on the relationship between age and bullying or violence has been rather less conclusive. One early study suggested that most interpersonal conflicts were found amongst younger employees (Appelberg et al. 1991), and O’Connell and Williams (2001) reported that the 26- to 45-year-old age group is more likely to report bullying. In the United Kingdom, Rayner (1997) and Hoel and Cooper (2000) reported higher levels of bullying amongst younger employees, followed by those aged 35–44. Those aged 55+ are least likely to report being bullied. Age related negatively with being a target of workplace bullying in the study by Einarsen and Raknes (1997) and De Cuyper et al. (2009), but positively in the study by Einarsen and Skogstad (1996). Einarsen and Skogstad (1996) looked at data from 14 Norwegian surveys and found that older employees have a significantly higher risk of victimisation, than their younger counterparts, with the exception of university employees aged over 50, who were significantly less likely to report having been bullied. In the representative UK study reported by Fevre et al. (2009), no relationship was found between age and bullying or harassment, but those with less than a year’s service in their current job were more likely to experience bullying. Milczarek’s (2010) review of literature on workplace violence identified an increased risk for younger workers and those with less work experience.
Finally, we turn to research on employees with a disability or long-term health condition. Fevre et al. (2009) reported that those who were not disabled were less than half as likely to report bullying and harassment. Up to this point, most bullying research had conceived impairment or ill-health as the effects of bullying (Fevre et al. forthcoming) and not factors which might be implicated in their causes (one notable exception was Hoel, Faragher and Cooper 2004). There is certainly significant evidence for the negative consequences of exposure to bullying for mental health (Vartia and Hyyti 2002), psychosomatic illness (Zapf et al. 1996) and psychological well-being (Mikkelsen and Einarsen 2002). Some studies show how bullying leads to stress and as job demands rise, support and control diminish (Tuckey et al. 2010). This evidence strongly suggests a downward spiralling process leading to further negative psychological as well as physical ill-health. De Cuyper et al. (2009) argued that bullying leads to withdrawal, absence and seeking new employment as well as emotional feelings of isolation and helplessness.
On the other hand, research in Canada (Eakin 2005) and the United Kingdom (Cunningham, James and Dibben 2004; Dibben, James and Cunningham 2001; Foster 2007; James, Cunningham and Dibben 2002, 2006; Walker and Fincham 2011) also described the way in which employees with health problems and disabilities found themselves on the receiving end of ill-treatment, particularly from managers, including bullying and harassment, from which they were supposed to be protected by anti-discrimination legislation. Schur et al. (2009) found that employees with disabilities did not feel more marginalised or disadvantaged in companies that all employees thought were more fair and responsive. Woodhams and Corby’s research showed how perceptions of disability in different workplaces were governed by the impact of an impairment or illness. Whether an impairment was disabling depended heavily on the nature of the work and workplace (Woodhams and Corby 2003). More generally, the stigmatisation of employees with disabilities may serve to reproduce the symbolic order of the workplace, reinforcing status differences and legitimating differences in power (Abberley 1987; Parker and Aggleton 2003; Walker and Fincham 2011). This, of course, recalls the theories advanced by Lopez et al. (2009) to explain the bullying of minorities as well as some of the earlier discussion of the causes of bullying.
It has been argued that the bullying concept has been widely adopted precisely because it fails to turn private workplace troubles into public issues in a convincing way. Thus McCarthy (2003) argued bullying in the workplace was a good fit with the zeitgeist of therapeutic remedies for private troubles. It was also compatible with the prevailing emphasis on codes of conduct of individuals (in this case, individual employees – see also Walker and Fincham 2011). In effect, the concept of workplace bullying did not attempt to generalise from the individual experience and, therefore, could only propose individual solutions to the issues it raised. As a result, private troubles stayed private and did not become public issues. There have been hints in the discussion that those with a more sociological interest in this field have come to a similar, pessimistic conclusion about the usefulness of the concept of workplace bullying.
We do not find modifications to the concept like ‘organisational bullying’ (or even ‘structural violence’) to be sufficient conceptual repairs and think it is more constructive to go back to the drawing board. If bullying cannot be an effective vehicle for the application of the sociological imagination, there are other contenders: counterproductive behaviour (Greenberg 1997), uncivil behaviour or incivility (Cortina and Magley 2009; Lim, Cortina and Magley 2008; Pearson, Andersson and Wegner 2001), abuse (Keashly, Hunter and Harvey 1997; Tepper 2000), negative acts (Einarsen and Raknes 1997) and mistreatment (Blase, Blase and Du 2008), for example. ‘Counterproductive behaviour’ seems unlikely to do the trick given what we have already discussed in relation to the benefits managers and employers may find in bullying. Perhaps people can be civil, and not abusive, but still make people’s work life miserable.
Mal- or mistreatment, along with negative acts or negative behaviour (Einarsen and Raknes 1997), seems to be sufficiently inclusive for our purposes, and they usefully leave open the question of intent while indicating that people’s values are under threat. In Part Two, we shall explain that we used a modified version of the negative acts questionnaire (NAQ) in our survey. The terms ‘negative act’ or ‘negative behaviour’ do, however, carry the connotation that a person is responsible. Since we do not want to rule out the possibility that practices, or even organisations, lie behind troubles at work, we prefer to standardise on an equivalent term, ‘ill-treatment’, where we can. This is not enough, however. More conceptual innovation is required if we are going to be able to give the sociological imagination free rein to help us turn the private troubles of Suhuur, Ramsey, Nandi and the others into a public issue which can be properly addressed.
We need a concept for what lies behind, or under, the ill-treatment that directs us to the sociological explanations we might examine. Concepts like mismanagement (Hodson 2001) are again too limited, in this case failing to encompass problems for which managers are not responsible, and also having some of the same drawbacks as the idea of counterproductive behaviour. The management that creates problems for employees may be good from the employer’s point of view. The alternative we propose is to turn to the vernacular idea of trouble at work. Trouble at work is not simply an easily understood way to describe the different examples of a bad day at the office we listed at the beginning of the chapter. All the same, the fact that the term (in contrast to negative workplace behaviour) features in the vernacular tells us it has the potential to take our measurements, explanations and solutions beyond the superficial behaviours to more deep-seated, long-lasting, fundamental social patterns and relationships. Trouble at work directs us towards workplace dynamics involving managers, workers and in some cases, the clients and customers of the organisation’s services that leave people feeling troubled. It also takes us to the notion of the troubled workplace in which the underlying causes of ill-treatment are built into the social relations of the workplace between employees and employees, employees and managers, managers and clients, and customers and members of the public. Best of all, both trouble at work and the troubled workplace leave us, at the onset of our research, with no need to take sides, or assign blame, in order to find evidence that shows our concepts describe real-world experiences and help us to understand and address them.
In the remainder of this book, we interrogate large data sets in order to understand how common troubles at work really are, who experiences the most troubles, why they do and what can be done about this. Although, within the troubled workplace particular kinds of workers may suffer more ill-treatment than others, the differences between troubled workplaces and other workplaces are so marked that they mask any other effects. It is, therefore, no surprise that the most effective solutions are those which address the problems of the troubled workplace.