Dysfunctional behaviour in university accounting schools: a tale of management control Free

Dysfunctional behaviour within the workplace may be broadly thought of as “any behavior that brings harm, or is intended to bring harm, to an organization, its employees, or stakeholders” (Van Fleet and Griffin, 2006, p. 698). Such behaviour adversely affects work quality and work relations and has high financial and social costs for the organisation (Peterson, 2002). Perpetrators of such behaviour have been variously and evocatively described as “snakes in suits” (Babiak and Hare, 2006), “workplace monsters” (Clarke, 2002), “dark triad personalities” (Babiak and Hare, 2006) and corporate psychopaths (Lingnau et al., 2017). The causes, effects and implications of dysfunctional behaviour have attracted the increasing attention of researchers in a diverse range of disciplines including organisational behaviour (Spector, 2011); management (Aasland et al., 2010); human resource (HR) management (Cohen, 2016); psychology (Volmer et al., 2016); business ethics (Mackey et al., 2018); tourism and hospitality (Lugosi, 2019); public administration (Boddy et al., 2010); neurology and psychiatry (DeSilva, 2014); and law (Babiak et al., 2010). This range of disciplines not only highlights the disparate contexts within which dysfunctional behaviour in the workplace has been observed but also reflects the multidisciplinary nature of this field of research. Unsurprisingly, this inventory of research clearly recognises that dysfunctional behaviour can and does occur at the organisational as well as the individual level (Cohen, 2016) and is not confined to particular industries or occupations. Rather, it rather appears to be a pervasive aspect of organisational life particularly in environments characterised by high autonomy, limited monitoring and minimal control (Young et al., 2016). Although based on evidence gathered in university accounting schools, this study aims to enhance and enrich our understanding of the role of management control in dealing with dysfunctional behaviour in public sector organisations and beyond.

Despite the apparent ubiquity of dysfunctional behaviour in the workplace and its potentially deleterious effects, what is implicit in any discussion of dysfunctionality is that as a violation of formal and informal organisational norms (Jaworski and Young, 1992, p. 18), it can be fundamentally cast as a problem of management control (Hartmann, 2000). Thus, central to any investigation of dysfunctional behaviour – be it exhibited at the organisational or at the individual level – is an understanding of the processual relationship between prevailing management control practices and the behaviour they seek to regulate.

This relationship is an important one. Dysfunctional behaviour at the organisational level has been shown to have serious ramifications (MacKenzie et al., 2011), including reduced productivity (Boddy et al., 2010); increased absenteeism (Bennett and Robinson, 2000); erosion in organisational commitment (Willness et al., 2007); higher employee turnover (Hauge et al., 2010); legal action (Ege, 2015); and role conflict (Jaworski and Young, 1992). At the individual level, research has established that certain behaviours – for example, overt and covert aggression, bullying, verbal abuse, undermining, harassment and mistreatment, (Hershcovis, 2011; Tepper and Henle, 2011) have a destructive impact on victims, including lowered physical health (Lim et al., 2008); psychological ill-health (Niedl, 1996; Zapf et al., 1996); depression and emotional exhaustion (Hershcovis and Barling, 2010); lowered mental health (Willness et al., 2007); a sense of helplessness (Harvey et al., 2017); besmirched reputations (Bergman et al., 2011); physical and mental health (Giacalone and Promislo, 2010); incivility (Estes and Wang, 2008); fear and frustration; as well as stress-related symptoms (Michalak and Ashkanasy, 2018).

Universities, a major economic sector in most developed countries, may be particularly prone to such dysfunctional behaviour in view of the considerable autonomy traditionally enjoyed by academics, the relative freedom they have been afforded in the course of discharging their teaching and research responsibilities (Martin-Sardesai et al., 2016) and a climate characterised by subjective performance evaluation and conflicting goals (Westhues, 2004). It may be that while too much accountability in the form of increased management control might reduce academic autonomy and freedom (and might itself have unintended dysfunctional consequences) (Martin-Sardesai et al., 2016), too much academic autonomy and freedom might lead to a lack of accountability (for particular behaviours) due to insufficient management control generally. Indeed, a number of recent studies have highlighted the prevalence of dysfunctional behaviour exhibited within higher education contexts (Björklund et al., 2020; Keashly and Jagatic, 2011; Shelley et al., 2017). This has increasingly been set in a university environment of major university strategic and identity change as its managements have increasingly pursued commercial imperatives, behaving as corporate organisations in pursuit of brand revenues and financial bottom lines (Parker, 2020). Universities pursuit of these commercial imperatives are intensified in an economic environment of increasing external pressures which includes significant reductions, in real terms, of government funding for tertiary education in Australia. Anecdotally we contend that it is not difficult to find senior, mid-career or early-career academics who have directly or indirectly been privy to numerous instances of dysfunctional behaviours over the course of their careers. The nature, prevalence and consequences of such behaviour at an organisational level, however, is not as easily discernible. This oftentimes opaque dysfunctionality stands in marked contrast to the performance focus of today’s commercially oriented universities where performance is managed and reported from internal management systems through to university-wide public performance reporting (Rana et al., 2022).

Aside from studies that have examined dysfunctional behaviours as responses to government-mandated research assessment exercises in universities (Martin-Sardesai et al., 2016, 2017a, 2017b, 2018), research that has investigated the incidence of dysfunctional behaviour more generally as exhibited within the higher education sector has been limited. Universities have over the past decade experienced considerable organisational change as a consequence of pressures associated with “accountingisation” [¹]. There is evidence that these pressures have increased the incidence of dysfunctional behaviour at the organisational and individual levels in different ways, (Martin-Sardesai et al., 2020b). As processes “[…] used by managers to direct employee attention and influence employee behavior in ways that increase the probability of achieving organisational goals” (Grabner and Speckbacher, 2016, p. 31), management controls (presumably) play or potentially play an instrumental role in mitigating the possible incidence and consequences of dysfunctional behaviour (Martin-Sardesai et al., 2020b), yet the role of management control in dealing with dysfunctional behaviour is largely unknown. Towards resolving this tension is what this paper seeks to achieve. Accordingly, an investigation of such behaviour within a higher education context will bring additional insights relating to the range, prevalence and control responses of dysfunctionality to this conversation.

This study was conducted via semi-structured interviews with 28 current or previous heads of accounting schools, research deans, deans of teaching and learning, school managers and HR managers from 16 Australian universities. Our decision to constrain our investigation to university accounting schools is quite calculated. As noted above, we do not by any means suggest that the incidence of dysfunctional behaviour is unique to the accounting discipline. However, in addition to making the scope of the study more manageable, restricting the study to a single discipline provides a point of departure and comparison for future research in this area. Moreover, focusing on the unique characteristics and requirements of a specified disciplinary context that is familiar to the research team provides the opportunity to mitigate the influence of other disciplinary idiosyncrasies that could account for the incidence of dysfunctional behaviours. Further, accounting as a discipline remains one of the most popular undergraduate areas of study at Australian universities and it is by far the most popular among business students; thus, as a significant contributor to organisational outcomes the implications of dysfunction can be expected to be similarly significant.

In setting out to better understand the role of MCS in dealing with dysfunctional behaviour, we find a discernible preference by managers for informal control processes rather than formal MCS in managing these behaviours. This paper presents a new paradigm for understanding this response in suggesting that managers conceive of dysfunctional behaviours as fitting into three broad forms of dysfunction: legal, hierarchical and interpersonal. This categorisation is in turn shown to impact on the preference for informal versus formal control processes as a mechanism to resolve dysfunctional behaviours. Specifically, informal control processes are preferred to resolve hierarchical and interpersonal dysfunction while formal control processes are preferred for resolving legal dysfunction.

The remainder of this paper proceeds as follows. Firstly, we provide an overview of the relevant prior literature that has considered the occurrence of dysfunctional behaviour in the workplace. Next, we outline the research context in which the study is situated and describe our approach to the design of this inquiry. We then present and interpret our findings and discuss their theoretical significance and practical importance. Finally, we offer our concluding reflections, identify the limitations of this study and suggest directions for future research.

Defined broadly as “behavior that the organization, given control, would prefer not to have exhibited by its employees” (Griffin and Lopez, 2005, p. 988), what constitutes “dysfunctional behaviour” appears at face value to be intuitively obvious – simply, “employees behaving badly”. Nevertheless, it is a term that embodies more complex meanings, encompassing any number of workplace practices and negative outcomes for both individuals and organisations. This study adopts an interpretive approach to the investigation, and in so doing accepts the premise that different people experience and understand the same “objective reality” in different ways (Bryman, 2016). It is therefore unsurprising that the diversity in behaviour characterised as dysfunctional has been reflected in academic literature examining a plethora of related issues. Behaviour within organisations that has been cited as dysfunctional has included bullying (Mackey et al., 2018); harassment (Lim and Cortina, 2005); theft (Cohen, 2016); rigid bureaucratic behaviour (Hirst, 1981); verbal abuse and disrespectful language (Cortina et al., 2001); intimidation and threatening behaviour (Thoroughgood et al., 2011); violation of rules (Fisk et al., 2010); blaming others (Lugosi, 2019); accountability avoidance (Michalak et al., 2018); sabotaging the efforts of co-workers (Aasland et al., 2010); cynicism (Sabagh et al., 2018); resistance to change (Warren, 2005); and passivity and apathy (Väyrynen and Laari-Salmela, 2018).

The literature amply demonstrates that the prevalence of such behaviour does not appear to respect organisational boundaries, having been studied in a diverse range of settings including school principals (Bergman et al., 2011); health care and social services industry (Volmer et al., 2016); undergraduate students (DeShong et al., 2015); graduate business students (Harlos, 2010); MBA graduates (Dean et al., 2010); marketing managers (Jaworski and Young, 1992); manufacturing, professional/technical, education, emergency services and administrative groups (Michalak et al., 2018); financial institutions and public service organisations (Mathieu et al., 2014); banking, information technology and fashion retailers (Cugueró-Escofet and Rosanas, 2017); information and communication technologies (Väyrynen and Laari-Salmela, 2018); and marketing managers (Saini and Krush, 2008). These studies, although by no means exhaustive, provide clear examples of the detrimental outcomes that dysfunctional behaviour typically invokes and underscores the importance and pervasiveness of this problem across sectors, industries and organisations generally.

Despite the various forms and guises through which it is manifested, it could be argued that the incidence and consequences of all dysfunctional behaviour is a management control issue. Control has been portrayed as “any process used by managers to direct employee attention and influence employee behaviour in ways that increase the probability of achieving organizational goals” (Grabner and Speckbacher, 2016, p. 31). Surprisingly, however, control as a frame of reference from which to examine dysfunctionality is one that has been implicit rather than explicit in much of the empirical literature. Thus, conceptualising behaviour in terms of its functionality (i.e. “functional” or “dysfunctional”) confers upon this construct a normative status that implies a desirable versus an undesirable state or process (Bergman et al., 2011). However, different behaviour(s) can have different meanings to different observers, and as such, what constitutes functional or dysfunctional behaviour is not necessarily an absolute judgement but often a relative one, driven by the subjective interpretation of the actors and observers involved. Nor is dysfunctionality an either/or proposition. As suggested by studies investigating the various forms of dysfunctionality (Griffin and Lopez, 2005; Martinko et al., 2002; Forsyth et al., 2012, among others), what is clear is that dysfunctional behaviour should be regarded as a matter of degree, with some behaviours (such as theft) being clearly dysfunctional, while others (such as apathy and passivity) can be open to various construals. Formal university management control systems (MCS) [²] governed by HR management, operational, financial and other policies are typically prescriptive as to how such dysfunctional behaviours should be addressed. The extent to which such formal MCS are successful in identifying, regulating and resolving such behaviours is largely unknown, and the role of informal control processes [³] is similarly equivocal. Following this line of argument, distinguishing between the relative use of formal MCS and informal control processes in responding to dysfunctional behaviour is important from the perspective of management control research. One of the more influential views in management control research is that an organisation’s control efforts comprise the use of both formal and informal controls collectively operating as a “package”. The notion of management control operating as a package comprising both formal MCS as well as informal control processes dates back to Otley’s seminal work (Otley, 1980). Despite this relatively long tradition, management control researchers have repeatedly called for more empirical evidence on the interplay between formal MCS and informal control processes. There remains little empirical analysis of how control mechanisms combine (Bedford and Malmi, 2015) and indeed, as recently as 2016, Otley (2016, 45) observes, “The ‘package’ concept has not yet been taken seriously in the design of most empirical studies”. This has been argued to be partly attributable to difficulties associated in more fully understanding the nature of informal control (Tucker, 2019), mixed findings of studies undertaken in this area (Demartini and Otley, 2020) and because insights into the isolated or stand-alone effects of controls could be fundamentally different from the use of controls in combination (Bedford, Malmi, and Sandelin, 2016). This study adds to the emerging body of research investigating the interdependencies between formal MCs and informal control practices as it relates to the control of dysfunctional behaviour.

Thus, the attention of this study is directed first, to the question of how dysfunctionality is defined by internal university academics and professional administrators responsible for its management, and second to the role of prevailing management controls in dealing with the range of dysfunctional behaviours exhibited in the workplace.

Notwithstanding the diverse range of settings within which various forms of dysfunctional behaviour has been investigated, one context that has received limited attention is that of the higher education sector. This sector is for many countries a major economic driver, which engages large numbers of people, is subject to increasing government regulation and exhibits a range of complex characteristics. As large-scale employers and economic drivers, universities constitute an appreciable component of national and local economies. For example, the significance of the Australian higher education sector, which comprises 40 public universities, two international universities, a private specialist university and approximately 130 other higher education providers, is reflected in the enrolment of over a million students annually (Norton and Cakitaki, 2016); international education exports reported to be in the order of AU$21.8bn in 2016 [Australian Bureau of Statistics (ABS), 2017]; and employment of approximately 100,000 staff (Department of Education and Training, 2015).

As an example of the higher education sector’s regulatory environment, recent research assessment exercises such as the Australian Government’s ERA (Excellence in Research for Australia), the UK Government’s REF (Research Excellence Framework) and the New Zealand Government’s PBRF (Performance Based Research Fund) are becoming an increasingly common fixture in the global environment within which universities function. The trend has been followed in countries including Austria, Denmark, Sweden, Canada, Germany and the Netherlands (Parker, 2011). The rating of universities through such exercises forms part of an MCS designed to allocate public funding to universities by measuring, monitoring and evaluating their research “impact”, “relevance” or “usefulness”. Such national research assessment systems are potent examples of national and university level metrics that drive right down to measuring the research performance of individual academics (Beattie and Goodacre, 2012). As indicators of performance, they become used as management control tools for driving the behaviours of university academics and ultimately shape the branding and market attractiveness of individual academics, business schools and universities (Broadbent, 2010; Martin-Sardesai et al., 2018). All of these carry pronounced financial incentives for the participants. Despite protestations to the contrary, the academic accounting community and its business school deans consider these national research performance evaluation systems to be primarily driven by formal or (reviewers’) informally held journal rankings. These produce direct pressures on academics to engage in a sector wide gaming of the system that includes academics, journal editors, journal reviewers, business deans and university bureaucrats (Moizer, 2009). The potential for dysfunctional academic behaviours directly prompted by university MCSs is clearly evident (Broadbent, 2010). As O’Regan and Gray (2018, p. 545) put it:

What these global trends towards university teaching and research performance measurement, and rankings reflect is the accountingisation of the higher education sector (Martin-Sardesai et al., 2020a, 2020b). Accountingisation represents the growing influence of accounting and audit in this sector, taking the forms of performance management systems, management controls, metrics and rankings – applied from the organisation-wide level right down to the individual manager and academic levels. These are focussed on delivering financial outcomes, cost efficiency, economic performance. Martin-Sardesai et al. (2020a, 2020b) produce evidence of accountingisation imposing increasing workloads, calculation and control of individual behaviour and performance and associated staff stress. Pressure is applied and experienced for ever-rising teaching evaluations, student success rates, research “scores”, research grant winning and more. These outcomes are almost universally translated in financial terms. Increasingly the targets and procedures are viewed by academic staff as unrealistic and unattainable, thereby inducing staff frustration, bureaucracy avoidance tactics, self-protecting individual research and teaching strategies and a prioritisation of self-interest and self-protection over organisational aims. Such accountingisation intersects with university strategizing, a process occurring at school, faculty and university levels that has been shown susceptible to both formal and informal power structures and individual influence by Hutaibat (2019).

The dysfunctional consequences of such accountingised control regimes has been well documented in the management accounting literature, with recent studies noting among other things the incidence of gaming (Agyemang and Broadbent, 2015); opportunism (Broadbent, 2007); citation counts and H-scores (Tourish and Willmott, 2015); “reactive conformance”, in which performance measures and targets become an end in themselves rather than a means to an end (Pollock et al., 2018); and misrepresentation, misinterpretation, reclassification or fabrication of information to meet performance requirements (Cardinaels and Yin, 2015). Moreover, consequences attributed to such controls have included increasing levels of stress and decreasing job satisfaction (Martin-Sardesai et al., 2017a); feeling overworked, pressured, demoralised and frustrated (Pop-Vasileva et al., 2011); impaired innovation and creativity in both research and teaching (ter Bogt and Scapens, 2012); and aggressive bullying behaviour by academic management (Du and Lapsley, 2019).

From this literature, two important insights emerge. Firstly, empirical research on dysfunctional behaviour has to date been primarily oriented towards the macro issues of governance and resource allocation, driven largely by responses of academics to government-imposed research assessment exercises in various guises (Abramo et al., 2011; Broadbent, 2016; Evans, 2014; Guthrie et al., 2019; Malina and Tucker, 2020). Research into dysfunctional behaviour in the higher education extending beyond that invoked by national research assessments is somewhat sparse. The second insight is that much of the literature noted above reflects an accounting orientation, namely, dysfunctionality in response to accounting controls and performance measures within universities or their effects upon accounting academics. Thus, what may appear at first sight to be a general management challenge of addressing dysfunctional academic behaviour and its causes, is arguably rooted in today’s corporatized and commercialised university culture with its primary focus on financial outcomes, aided and abetted by the accountingisation of university management and control systems. Hence, this study’s investigation of control systems and dysfunctional university staff behaviour provides an individual level insight into the operation and relationships between management control design, use and reactions thereto in a higher education sector that has become distinctly accountingised and commercialised.

The theoretical development and framing conducted within this study proceeded in two iterative stages (Hesse-Biber and Leavy, 2011). Firstly, within the qualitative tradition of inductive theorising from field data, the study set out to elicit interviewees’ implicit concepts of dysfunctional behaviour. From their interview transcripts it then thematically induced the major emergent categories of dysfunctional behaviour as observed by actors in the field. This allowed us to draw from that social and organisational phenomenon, theoretical concepts of dysfunctional behaviour, supported by field-derived evidence and examples (Silverman, 2000; Payne and Williams, 2005). In then exploring the three major emergent conceptual categories and their characteristics, we developed a deeper understanding and representation of behaviours and beliefs associated with each category (Williams, 2000). Hence, our conceptual theorising has initially been data driven, triggered from interviewees’ reflections on the dysfunctional behaviour phenomena they have witnessed (Hennik et al., 2011; Harding, 2013). This reflects a “bottom up” approach, developing local interpretations of particular occurrences before moving to a more generalised theoretical position (Eriksson and Kovalainen, 2008; Gibson and Brown, 2009; Harding, 2013). Thus, the emergence of the three categories of dysfunctional behaviour identified in this study are the product of both emic and etic level analysis and interpretation: i.e. analytical induction (Madden, 2010; Hesse-Biber and Leavy, 2011; Hennink et al., 2011). These constitute this paper’s three major themes which are then interpreted from the frame of reference of Rational Choice Theory (RCT) with a view to providing further explanations (Simons, 2009). This approach has thus included the employment of RCT to assist in making sense of the attitudes and behaviours being explored in the field data and assisting in proffering explanations of those (Goddard, 2017; Lukka and Modell, 2017).

The use of RCT to explain human behaviour has a long tradition in accounting research, dating back over three decades (Ansari and Euske, 1987; Ansari and Bell, 1991; Carruthers and Espeland, 1991; Covaleski et al., 1993; Hoque and Hopper, 1994; Jones and Dugdale, 1994; Vamosi, 2000; Rahaman and Lawrence, 2001; and Tucker and Lawson, 2020). As a long-standing paradigm in neoclassical economics (Kari, 2014), RCT describes, explains and predicts decision-making in terms of a cost–benefit analysis of a set of alternatives, in which the actors are seen as rational utility maximisers who select the decision option that promises the highest benefit with the lowest cost (Scott, 2000). This perspective considers the challenge of deciding between alternative courses of action in any given situation to be essentially an optimisation problem (Moll and Hoque, 2018), in which individual action is motivated by the maximisation of future utility (Tucker and Lawson, 2020). It should be noted that in RCT, neither benefits nor costs are limited to economics, nor are they limited to expression in monetary or financial terms. As Quackenbush (2004, p. 94) observes, “[…] there is in no way a requirement that rational actors are only (or at all) motivated by economic benefit. Demonstrating that people do not choose according to economic preferences does not demonstrate that they do not choose rationally; it only means that they are not solely motivated by economic gain”. Thus, benefits and costs can include items that are not readily quantifiable such as perceived advantages/disadvantages, “strategic” decisions/preferences, “politically acceptable” choices, preferences based on social, organisational, institutional, political factors or personal opportunity costs for example. Adopting a wider view of benefits and costs to include social as well as economic ones and quantifiable as well as non-quantifiable items reflects a more balanced approach to evaluation in a higher education context, recognising the statement attributed to Albert Einstein that, “Not everything that counts can be counted and not everything that can be counted counts” (Malina and Tucker, 2020, p. 118).

Nevertheless, as with most theories, RCT is not without its criticisms and limitations, many of which relate to its underlying assumptions that individuals possess perfect information enabling a definition of a problem and knowledge of all alternatives; are capable of identifying relevant criteria and their accurate weightings in evaluating alternative courses of action; and purposively seek to maximise their utility (or gain) solely in terms of the consequences for themselves (Burns and Roszkowska, 2016). However, it is well established that individuals often behave in ways that may be inconsistent with predictions based on the assumptions of rational decision-making (Chenhall, 2003), a point reiterated by Williams (2009, p. 277) who noted, “humans are irrational or at least do not function as the economically rational hypothetical beings that populate the world of neoclassical economics”. Further, it is unlikely that individuals possess access to the same information networks or use the same criteria upon which they evaluate this information, nor do they have an unlimited ability to absorb, process and interpret the information at their disposal (Simon, 1991). In addition, individuals typically evaluate and distinguish between the outcomes associated with the perceived costs and benefits of alternative actions, primarily with concern about the consequences for themselves (Burns and Roszkowska, 2016). Thus, as costs and benefits are not only defined by the individual on the basis of his or her perception, but also according to their unique individual and often different personal criteria and preferences, making the notion of “costs” and “benefits” relative rather than absolute.

With respect to the three major dysfunctional categories that will be identified inductively in this paper, these observations are particularly apposite to the hierarchical and interpersonal forms of dysfunctional behaviour. While a legalistic form may predispose managers towards a formal rational choice because of the rules of law and of the organisation, observations concerning less rational forms of behaviour do, for example, relate to the hierarchical form, as senior managers in organisations often have better access to information and resources than junior managers. Similarly, those observations of less rationality also relate to the interpersonal form of dysfunctional behaviours because managers and staff may not know what the criteria are, in that they may not be readily overt, explicit or easily identified.

In response, the notion of bounded rationality (Simon, 1959, 1991) serves to refine RCT (Cyert and March, 1963), relaxing the assumptions underpinning RCT by acknowledging that an individual’s evaluation of the options occurs through a simplification of the complexity of spatial, temporal or causal relations among choices and outcomes (Amit and Schoemaker, 1993). Adopting bounded rationality to replace the “absolute” rationality of RCT recognises that people often do not make the “best” choices; rather, they base their choices on reasonable options that meet expectations (van der Meer-Kooistra and Vosselman, 2000). Bounded rationality adopts an explicitly behavioural stance and takes into account the cognitive limitations of decision makers, their limited access to information processing capacity, potential for selective perception and inclination to satisfice rather than optimise in selecting courses of action (Adler, 2020). Thus, to abandon RCT as a lens through which to view dysfunctionality simply because complete rationality cannot be assumed may be premature.

The current study involving both inductive field data theorising as well as RCT theoretical explanations, capitalises upon the opportunity to engage more broadly with dysfunctional staff behaviour and the ways in which such behaviour is typically addressed in a university setting, in this case, schools of accounting. This exploration will focus on the individual and small group level, at the coal face of university operations, with a particular concern for the processes involved. The study design in pursuit of this research agenda is detailed in the following section.

This study explores the role of management control in dealing with of dysfunctional behaviour within accounting schools in Australian universities. We use a qualitative design to understand and interpret the nature of dysfunctionality as it is perceived as a precursor (Stake, 2010) to considering the ways in which management control might address and reduce such behaviours. However, concepts such as dysfunctional behaviour are “difficult beasts to theorize” (Malmi and Granlund, 2009, p. 633), not least because they often carry specific meanings for actors using and responding to them. Perceptions of the range of dysfunctional behaviours will invariably reflect the differing realities and experiences of individuals. Questions of not only “what” but also “how” and “why” underpin the understanding of the range of dysfunctional behaviours and their ramifications. Thus, in capturing the direct and vicarious observations, narratives and impressions of witnesses to dysfunctional behaviour, a qualitative investigation is well suited to the design of this study.

This study draws on interview-based evidence from accounting academics and university managers. We seek to gain a nuanced understanding and portrayal of what, how and why actions take place the way they do, the processes and mechanisms involved and an appreciation of the ways in which they occur in particular settings through accessing the lived experiences of the actors (Parker and Northcott, 2016). Participants were therefore selected using a non-random, purposive sampling technique. This sampling strategy is characterised by small sample sizes that aim to obtain richness and deeper understanding of the phenomena under investigation (Fossey et al., 2002). To achieve this requires the identification and inclusion of those individuals “that can best inform the focus of their inquiries and provide the in-depth information relevant to the study’s research question” (Parker and Northcott, 2016, p. 1116).

Because the object of our study was focused on an examination of dysfunctional behaviour and the MCS and processes used in response to such behaviours, we strived for maximum variation within those two investigative criteria (Miles and Huberman, 1994; Patton, 2002). Consequently, our sample comprised two broad groups: senior accounting academics who may have been directly or indirectly privy to incidences of behaviour they considered to be dysfunctional; and university managers (deans of research, deans of teaching, heads of school and school managers [⁴]) who were well placed to comment on the nature and adequacy of management controls used within universities to manage such behaviours.

The views of these two groups inform our research aim, which is to enhance and enrich our understanding of the role of management control in dealing with dysfunctional behaviour in university accounting schools. The voices of these informants are therefore pivotal in offering experiential insights into the ways in which formal MCS and informal control processes engage in this process (Beyers et al., 2014).

Semi-structured interviews formed the primary evidence base informing this study (Glesne, 1999). Our interviews aimed to identify interviewee perceptions relating to the nature, range, prevalence and consequences of dysfunctional behaviour, as well as the MCS and processes typically used in response to such behaviour. Given that the focus and scope of this study is on the role of management control in dealing with dysfunctional behaviour, the line of questioning was necessarily directly related to this phenomenon. The questions guiding the interviews were deliberately designed to address the study focus by allowing interviewees to speak free of any influences or preconceptions, interviewers expressing no opinions and taking care not to influence the interviewees. To this end, interviewees had ample opportunity to declare they had not witnessed or heard of any dysfunctional behaviour, to decide against answering any particular questions and to withdraw from interview. Given the seniority and experience of interviewees, the researchers did not observe any exhibited tendency to respond with a view to obtaining interviewer approval. Accordingly, interviewers made sure not to suggest or influence interviewees with any suggestion of a predetermined agenda or view. It is also important to note that ethics committee approval for this study was subject to the requirement that interviewees could identify the absence of dysfunctionality if that was their experience. In addition, our research design explicitly empowered interviewees to adopt their own definition and construals of dysfunctionality rather than us imposing our view of dysfunctionality upon them. It should be further noted that our review of the range of universities, hierarchical levels of interviewees and the academic/administrative staff division of interviewees was sufficiently broad that that there were no discernible contextual factors that appeared to predominate. Given the semi-structured interview method, the related guided conversations and probe questions, no particular influence from any attitudes that interviewees might have held towards their university was apparent.

The semi-structured interviews were designed to provide the opportunity to explore the perceptions of interviewees in an open and flexible manner (Shank, 2006), enabling the identification of both common and divergent understandings of interviewees’ (Fontana and Frey, 2000). Interview questions (presented in  Appendix) were constructed so as to enable interviewees to elaborate upon observations they believed to be relevant and significant and to present their understandings and interpretations (Kvale and Brinkmann, 2009).

The two interviewee groups, senior accounting academics and university managers, were invited via email to participate in this study. We conducted a total of 28 interviews with individuals drawn from 12 (of 34) Australian universities. Of the 28 interviews, 11 were senior accounting academics and 17 were university managers (nine pro vice-chancellors/deans of research/deans of teaching and learning, five school managers and three HR managers).

As interviewees were geographically dispersed across Australia, face-to-face interviews were impractical for all cases. For this reason, in addition to the nine interviews conducted in the traditional face-to-face mode, 15 interviews were conducted via telephone and four via Skype.

The length of interviews was determined by the available time of the interviewees and the extent to which they responded to the questions raised. The duration of interviews ranged from 50 to 95 min. Interviews were recorded and subsequently transcribed. Detailed notes were taken during and following the interviews. Names of interviewees and their institutional affiliation(s) are not published to ensure anonymity.

Our analysis of the qualitative data broadly followed Eisenhardt’s (1989) approach and involved data summarisation, categorisation and interpretation. This entailed identifying common themes, unique insights and discrepancies apparent in the data.

We undertook a thematic analysis of the interview transcripts. This enabled the development of emergent codes which formed the basis for cross-case analysis, which in turn assisted in the identification of emergent patterns and themes in the data (Huberman and Miles, 1998). We used the software package (NVivo) to facilitate the efficient recovery of quotations made by interviewees, as well as to aid in detecting commonalities evident in interviewees transcripts.

Interview transcripts were coded using multiple categories, because more than one code might be applicable to a particular discussion. Categories with inadequate supporting data were discarded or integrated with similar categories. Original transcripts were then rationalised around predominant themes (such as how dysfunctional behaviour is defined) and issues (such as how effective prevailing management controls were perceived to be in addressing dysfunctionality) and compared. This enabled the scope of the themes raised and their associated contexts and meanings to reach saturation point when no new ideas were forthcoming and (consequently) no new categories were developed (Silverman, 2006).

A number of measures were used to ensure the field data was captured and delivered authentically and genuinely (Parker and Northcott, 2016) and to establish “credibility” and “dependability” of the data (Lincoln and Guba, 1985). Procedures undertaken to promote credibility were principally by means of peer debriefing, in which the researchers debated the emergent findings to identify and confirm the meaning of the themes and patterns that appeared to be apparent. Dependability was fostered by the maintenance of contact records with interviewees, the dates, times and venues of interviews and the transcription of interview notes (Gelman and Basbøll, 2014).

Our findings are organised around two principal themes that emerged from our discussions:

1. defining dysfunctional behaviour and the associated forms of dysfunctional behaviour identified as prevailing within accounting schools; and

2. the ways in which both formal MCS and informal control processes are perceived to address such behaviours.

In reference to the first theme, three broad forms of dysfunctional behaviour were identified: legal, hierarchical and interpersonal. In reference to the second theme, the perceptions of interviewees regarding the relative effectiveness of formal MCS and informal control processes used in accounting schools in response to incidences of dysfunctional behaviour were identified.

The quotations reported in this section are used to illustrate emerging themes and points representative of the discussions and the participants are identified by pseudonyms to preserve their anonymity. The convention used in citing direct quotes is to preface comments made by academic staff with the letter A, while comments made by university managers are prefaced with the letter M.

The findings echo the “voices” of interviewees, underscoring their uppermost observations, principal experiences and primary concerns. Close examination of the transcripts revealed an absence of any discernible difference between the perceptions of participants based on:

• whether they were academics or university management;

• years of experience; or

• gender.

The similarities and convergence in views expressed by participants were pronounced and suggests a degree of transferability of the ways in which dysfunctionality is perceived across the broader university and organisational contexts within which this study was undertaken.

Rather than impose a preconceived definition or rigid idea on what dysfunctional behaviour is or might be, our opening question to all interviewees invites the participants to expand upon what they perceive to be “dysfunctional behaviour”. As shown in the following list, although this term was variously defined by interviewees, common to most construals is the perceived incommensurability of personal and organisational interests and the consequent disruption to the work of others.

Dysfunctional behaviour as defined by the participants:

• People not contributing at the level they should be contributing at. (A4)

• People not acting in the best interest of the organisation. (M2)

• People having their own agendas, which are different to the organisation and just doing things for their own personal interest rather than the greater good of the whole. (A6)

• When people put their own self-interest first. (A8)

• Individuals disrupt the operation of the institution in terms of limiting the productivity and performance of others. (M7)

• Where the values and expectations of the organisation differ significantly to those of the employee. (M8)

• People not contributing to a community. (A9)

• Behaviour that disrupts others or does not contribute to, or acts against, the overarching goals and mission of the unit. (M11)

• Those that disrupt or cause harm to others or who are antagonistic to what we’re seeking to achieve as an educational grouping. (M13)

• Working against the common good. (M15)

• Purposeful misbehaviour that affects others. (A10)

• Behaviour that gets in the way of people doing their jobs. (M16)

• Being mean to other people. (M17)

The definitions of dysfunctional behaviour as presented in Dysfunctional behaviour as defined by the participants identify some common views as to what constitutes dysfunctional behaviour. However, despite this confluence of views, what was apparent from our interviews was the diverse range of interviewee-portrayed behaviours that were reflected in these definitions. Consistent with our research method, Table 1 sets out the behavioural examples cited by our interviewees, which can be broadly grouped in terms of the potential implication of the behaviour into three dysfunctional forms:

1. Legal dysfunctionality is behaviour perceived to have legal ramifications for the university.

2. Hierarchical dysfunctionality encompasses those behaviours seen to affect relations within hierarchies, both academic and administrative.

3. Interpersonal dysfunctionality includes those behaviours pertaining to relations between staff irrespective of academic and administrative hierarchies.

The three forms of dysfunctionality as portrayed in Table 1 represent a distillation of interviewees’ observations. However, these three forms are by no means mutually exclusive and should be regarded as situational, rather than absolute, rigid and immutable. For example, interpersonal dysfunctional behaviours such as gossiping may carry with them legal implications in the form of potential defamation charges. Hierarchical dysfunctionality in the form of subversive behaviour may stem from problematic interpersonal relationships, and the potential legal implications arising from threatening and intimidatory behaviour may be attributable to hierarchical conflict. Moreover, legal dysfunctionality may be ultimately resolved at an interpersonal or hierarchical level, while hierarchical or interpersonal dysfunctionality may conceivably escalate to become litigious.

Nevertheless, what differentiates these forms of dysfunctional behaviour is that they not only recognise and categorise a considerable assortment of dysfunctionalities but also potentially explain the different responses to these behaviours. While care has been taken in the coding of participants’ observations, we recognise the need for caution in treating the categories as clearly distinguishable or reducing the complexity of dysfunctional behaviours into simplistic discreet stereotypes. As observed above, the relationship between the three forms of dysfunctional behaviour will almost certainly be more fluid and complementary than might appear at first glance. This characterisation of dysfunctional behaviour as identified by the participants provides a useful way of contextualising the reported observations and provides a structure for the findings that follow.

Legal behaviour dysfunction relates to those behaviours and actions that are “governed by a system of rules that are created and enforced by recognized authorities” (Pearce, 2013, p. 498). In the context of workplace dysfunctionality, such forms of behaviour have been defined and are assessable and punishable through the prevailing legal system – both statutory and via common law civil proceedings – and accordingly, they may have potential legal ramifications for the institution and the individual if not controlled. It appears to be a widely held view that if uncontrolled or unmanaged, these types of actions could result in court cases, legal costs and staff time costs and university reputational damage could be considerable. Accordingly, interviewees repeatedly emphasised that legal action and the associated damage it might bring was an outcome regarded as one to be avoided if at all possible:

The incidence and observation of this form of dysfunctional behaviour, including a variety of behaviour pattern types ranging from interpersonal to financial, was mentioned by almost all of our interviewees, who noted that such cases could result in court cases:

Nonetheless, it was also acknowledged that by no means did all potential legally contravening behaviours produce legal action. This was argued by a number of interviewees:

A number of interviewees called attention to the ramifications of action being pursued in relation to unlawful behaviours. What became apparent was that for many, as soon as potential legal implications were recognised, the recording process that might build a foundation for formal MCS implementation appeared to be a common response by those with managerial responsibility and accountability. Interviewees described them opting for recording careful and detailed notes (by themselves or other formally appointed personnel) relating to events, information, meetings etc. A common action was to document all communications by email so that a formal documented trail was built. This was linked to careful observation of university procedures and rules to avoid staff complaints of unfair treatment or resorting to stress leave and to build an evidential base that could be used should court cases ensue:

A particular theme emerging from our interviews pointed to dysfunctionality associated with tensions according to relative status, power or authority. Such forms of dysfunctionality have been labelled hierarchical. Observations of this form of dysfunctionality were repeatedly conveyed to us, in terms of dysfunctional behaviour displayed by academics to other academics; by academics towards administrative staff; by administrative staff towards other administrative staff; and between administrative staff and academics.

The two predominant forms of hierarchical dysfunctionality referred to by interviewees were bullying and micro-management. Bullying was primarily observed in the treatment of more junior academics by senior academics, who appeared to feel their rank permitted them to issue instructions to junior colleagues. This tendency was observed among senior academics who have no line management responsibility over any colleagues. This behaviour was particularly evident in senior academics’ treatment of newly appointed and less experienced academics, who were often required to take on duties and work the senior academic was normally expected to undertake:

A less toxic but nonetheless prevalent form of dysfunctional hierarchical behaviour took the form of micro-managing, whereby senior academics with managerial responsibilities might delegate work but then closely supervise and intervene in detailed implementation by subordinates:

Reasons underlying hierarchical dysfunctionality exhibited by academics, whether it be towards other academics or administrative staff (often unreported), was attributed to a number of influences. Firstly, there appears to be a significant perceived power imbalance between senior and junior academics, whereby the latter perceive the senior academics to be in a position to influence their chances of achieving tenure or promotion by virtue of positions as referees, panel members, networks, etc. In an accountingised university management control environment, staff careers are governed by individual level metrics and associated targets applied to them. Hence, the junior academics are more likely to defer to the senior academics, and bullying or micro-management by senior academics may, therefore, go unreported. Secondly, a further influence, also arguably reflecting university accountingisation, relates to the increasing employment of metrics as formal control mechanisms over staff performance measurement, rewards and sanctions. These appear to have promoted an instrumental culture among academics whereby senior academics were seen by interviewees to be prepared to use dysfunctional strategies and behaviours towards academic colleagues and administrators to enhance their metrics, if they felt it necessary:

A second influence appearing to trigger hierarchical dysfunctionality by academics was related by interviewees with respect to academics’ treatment of administrative staff. A recurring influence identified by interviewees was the notion of entitlement. This appears to involve academics’ perceptions of their role and position within universities, whereby they may see themselves not as employees but as independent experts working within universities and exempt from what they perceive to be constraints imposed by universities’ accountingised rules, expectations and control systems. This can include a perception that administrative staff are subordinate to them and expected to support whatever endeavours or expectations the academics choose:

In contrast to dysfunctional behaviour between academics, the incidence of dysfunctional behaviour exhibited between administrative staff appeared less common:

The most common type of dysfunctionality exhibited between administrative staff related directly to work priorities, where sometimes junior staff politely refused senior staff requests to undertake particular roles or duties. It was also manifested in the avoidance of responsibility or blame for undesired outcomes. Furthermore, there appeared to be a reluctance to engage in any form of whistleblowing, which appeared to be perceived as counterproductive to one’s career prospects:

While seemingly a rare occurrence from interviewee reflections, administrative staff were occasionally the perpetrators of bullying academics, by giving them orders. This was seen to be sourced in some administrators exercising what they perceived to be their formal organisational power and justifying the authority of their formal positional role:

The third broad form of dysfunctional behaviour relates to interpersonal conflicts, disagreements or tensions that give rise to, perpetuate or result in the incidence of dysfunctional behaviour irrespective of to whom it is directed and from whom it originates and as distinct from those arising from hierarchical differences. A succinct interpretation of this form of dysfunctionality, labelled interpersonal, was offered by one academic interviewee:

This form of behaviour was seen as particularly destructive and insidious because much of it is not readily explicit, overt or provable. It was cited as occurring “behind closed doors” in informal conversations and interactions away from formal meetings and events. Indeed, cases of people taking positions and actions outside formal meetings and occasions contrary to those they profess during formal meetings were provided as examples of this dysfunctional behaviour. It appeared to be seen as part of the behaviours exhibited by those who actively played “organizational politics” in pursuit of their own self-interests, particularly if they related to the accountingised derived performance metrics discussed above. Yet, while such activities were often hidden from public view, individuals engaging in such dysfunctional behaviours were seen by interviewees to often be well known by their colleagues for such proclivities and their behaviours are seen as duplicitous and particularly damaging for the targets of this behaviour:

Some of our interviewees offered reasons for why individuals would engage in this type of dysfunctionality. Virtually all of their diagnoses related to the individual characteristics of the perpetrators. The participants made some strong claims about their colleagues’ tendencies and insecurities:

A pronounced observation emerging from the accounts offered by interviewees is that dysfunctionality – irrespective of the form it takes – can be and is addressed either by formal or informal means. In this context, formal means involve recourse to prevailing university policies, procedures and other official MCS to control dysfunctional behaviour. In contrast, informal means relate to efforts to resolve instances of dysfunctional behaviour without recourse to formal MCS, typically through informal, “off-the-record” interactions and communications. In this section, the interviewees’ perceptions of the role and effectiveness of these formal and informal control processes with respect to the ways in which dysfunctionality is addressed are discussed.

The majority of interviewees outlined an extensive number and diverse range of university policies, procedures and other elements of the formal MCS available to deal with dysfunctional behaviour. As one senior academic commented:

However, invoking formal procedures was described as requiring a long and involved process that included meetings, ongoing performance assessment over time, extensions of time for improvement, involvement of HR personnel, recording and reporting. The length of time taken was repeatedly emphasised by interviewees as a significant weakness in MCS processes, as was a lack of confidence that formal MCS procedures could produce tangible outcomes:

Further limitations of the procedures available for managing dysfunctionality were also raised. Firstly, the difficulty in academic managers understanding and working through the MCS procedures:

Secondly, the decision to commence a process of performance management or disciplinary proceedings was seen to carry significant consequences for all involved, in that once started, it was difficult to stop the process. Furthermore, interviewees felt that formal proceedings can aggravate dysfunctional behaviours, impacting on the persons involved and causing reactions from staff in ways that do not help solve the problem. These perceived consequences can lead to managers avoiding formal MCS processes:

Thirdly, the formal MCS process for managing dysfunctional behaviour was seen to risk subjectivity and inherent bias. Interviewees considered dysfunctional behaviour, relationships and treatment of others difficult to identify and measure. Taking action in relation to these was seen to leave managers open to claims of bias, especially in the absence of agreed behavioural objectives and assessment criteria:

Interviewees also identified the failure of the formal MCS to produce constructive outcomes, remediations or penalties with respect to identified dysfunctional behaviour. Even where penalties might be specified, they were often seen not to be enforced. Severe penalties such as dismissal were seen to be extremely rare, while admonition was sometimes the most likely outcome, which in itself might not produce behavioural change. In these formal MCS processes, interviewees cited managers as not feeling their role and actions were supported by the university senior management:

With regard to the effectiveness of formal MCS in regulating dysfunctionality, two beneficial aspects of the policies, procedures and processes in place to address dysfunctionality were repeatedly expressed. Firstly, to the advantage of the university’s self-interest, the existence of and adherence to formal written policies and procedures was seen to be important for reducing risk, especially exposure to legal liability:

Secondly, to the advantage of the managers, interviewees recognised the value provided by the involvement of HR staff in supporting managers with professional advice through the process:

Virtually all interviewees indicated their preference for addressing dysfunctional behaviour via informal means. This informal approach usually involved personal, confidential conversations with the person exhibiting the dysfunctional behaviour or the person to whom the dysfunctional behaviour was directed and was designed to encourage resolution of the problematic behaviour. It also appeared to trigger confidential consultation with trusted senior colleagues, an important step in taking an informal route towards controlling dysfunctional behaviour:

Reasons for the predilection to manage dysfunctional behaviour through informal as opposed to formal means consistently related to the time, risk and difficulties such as paperwork, time expended and stress associated with instigating and pursuing a formal course of action. It also appeared to reflect a preference for dealing with dysfunctional behaviour at the local level within the organisation, while also assessing whether such behaviours are individually based or a more systemic pattern reflecting some wider problem within the organisational culture. The latter might suggest the need for a more formal approach:

Some interviewees also stressed that informal control approaches might move on to the implementation of formal MCS processes when it becomes evident that the informal approach is not producing the desired outcome:

In summary, this exploratory investigation of the role of management control in dealing with dysfunctional behaviour in university accounting schools indicates that definitions of dysfunctionality converge broadly to the incommensurability of personal and organisational interests and the disruption to the work of others. This definition was found to encompass a diverse range of behaviours converging to three broad groupings, which reflect the legal, hierarchical and interpersonal sources of dysfunction. From this overall framing of dysfunctionality, it is apparent that both formal MCS and informal control processes are typically used to various extents in addressing these behaviours. However, it is informal control processes that are the preferred means by which interviewees believe dysfunctionality is most effectively and commonly addressed. As a corollary, the coexistence of formal MCS and informal control processes point to a distinction between the predominantly formal design of management control as opposed to the actual formal and informal use of management control. The implications of this finding and a theoretical position that may explain these practices in the use of control in response to dysfunctional behaviour form the basis of the ensuing discussion.

In this section, we direct our attention to considering the implications of the findings presented above. In the first instance, we consider the predominance of informal control processes in the management control package as a preferred means of dealing with dysfunctional behaviour. We then invoke RCT as one means by which this preference may be explained.

This study’s finding of the reliance of interviewees on formal MCS for legal dysfunctionality but on informal control processes for hierarchical and interpersonal dysfunctionality resonates with the notion of management control collectively operating as a package (Demartini and Otley, 2020; Falkenberg and Herremans, 1995; Grabner and Moers, 2013; Otley, 1980). As has been previously observed in control research, the coexistence of formal and informal means of control highlights the potential for formal MCS and informal control processes to complement each other, operate as substitutes or act in opposition (see e.g. Abernethy and Chua, 1996; Bisbe and Otley, 2004; Ferreira and Otley, 2009; Fisher, 1995; Malmi and Brown, 2008; Sandelin, 2008; Tucker, 2019). That is to say, researchers’ identification and recognition of the control package comprising both formal and informal control, points to the limitations of any control strategy that attempts to exclusively rely on formal controls. This highlights formal and informal control as essential components of the overall controls used by management.

As outlined earlier, formal MCS are based on explicit and quantifiable standards and refer to that which is written or official (but not necessarily practiced) and are defined in terms of explicit and codified rules, procedures and policies (Cardinal et al., 2017). Informal control in contrast, is based on personal values, and in turn, beliefs, attitudes and behaviours (Norris and O’Dwyer, 2004) and social interactions (Dekker, 2004). Informal means of control influence and are influenced by group norms, common values and beliefs, culture and traditions (Ouchi, 1980) that direct the behaviour of group members (Collier, 2005). Formal MCS and informal control processes do not exist in isolation. Rather, control outcomes can be achieved through formal MCS or by more informal means through vertical as well as lateral networks comprising social interactions, relationships and ties, informally existing between employees (Tucker, 2019). By its very nature then, informal control is propagated through both formal as well as informal structures. However, management controls are relatively more informal than formal when they are less obtrusive and indirectly focussed (Merchant and Van der Stede, 2012). As Pan Fagerlin and Lövstål, 2020, p. 501) observe, “The dividing line between formal and informal controls can thus be drawn on the extent to which they are explicitly designed or planned, and their reliance on human interactions”.

As applied to the current study, this two-tiered approach to the control of dysfunctionality suggests a certain complementarity in the means of control, with formal MCS used in specific (narrow) circumstances (legal dysfunction) and informal control processes seen as more appropriate for other (broader) circumstances (hierarchical and interpersonal dysfunction). The predominance of informal control processes over formal MCS in certain circumstances is not without precedent in the management control literature, and as such our findings, although extended to a different context, are nevertheless largely unsurprising. As with the current study, informal control processes have been found to be more influential than formal MCS in a range of differing contexts within, for example, a major British-based communications organisation (Marginson, 1999, 2002); an Australian not-for-profit religious organisation (Parker, 2001, 2002); an international retailer (Norris and O’Dwyer, 2004); and an owner-controlled multinational packaging equipment supplier (Collier, 2005). Our finding that interviewees maintained a predilection for the use of informal control processes over formal MCS is consistent with this corpus of research investigating the relative importance of control based on its formality. What has yet to be satisfactorily resolved, however, is why an informal rather than a formal pathway to management control is typically preferred and adopted. In the current study, the inclination towards adopting an informal approach appears to prevail despite the extensive formal policies and procedures guiding how dysfunctional behaviour should be dealt with in all the universities from which our sample was drawn. This may be attributed to the fundamental difference between the design of formal MCS and its use.

This seemingly obvious point is well established in the control literature (Langfield-Smith, 1997; Simons, 1995; Tucker et al., 2013). As Bisbe et al. (2019) noted, although controls (particularly formal controls) may have been designed with an intended purpose, it is by no means inevitable that the ways in which such controls are used results in a manner that is compatible with, or corresponds to, the purpose for which they were originally designed. Formal MCS may be used as intended, used selectively on an exception basis or not used at all. Although this observation is consistent with the findings of the current study, it nevertheless falls short of explaining why a discrepancy appears to exist between the design of an extensive formal MCS and, at the same time, an apparent proclivity to use informal control processes in addressing dysfunctionality. However, our study findings go further with respect to how dysfunctional behaviour is addressed. Viewed through the lens of RCT, they provide a more specific understanding of the evident predisposition towards the use of informal control processes in preference to implementing formal MCS.

From an RCT perspective, the notion of bounded rational choice provides valuable insights into the apparent predilection of individuals to pursue informal control processes rather than formal MCS in dealing with dysfunctionality. These insights can be appreciated by considering the costs and benefits of using formal MCS, as perceived by interviewees. A distillation of overall findings as costs and benefits presented by interviewees is depicted in Table 2, which groups the overarching evaluations of formal MCS prevailing within universities as perceived by interviewees in to three categories: outcome, processual and personal.

These groupings can be thought of as “drivers” influencing the decision to use formal MCS in addressing dysfunctional behaviour (Shaver, 2005). The outcome driver reflects cost–benefit evaluations of the perceived consequences of using formal MCS. The processual driver refers to the cost–benefit evaluation associated with the procedures necessitated by the formal MCS. The personal driver represents the cost–benefit evaluation relating to interpersonal or individual ramifications perceived in using the formal MCS.

The classification of drivers as depicted in Table 2 was inductively derived based on the evidence presented in this study and presents a nuanced picture of the cost–benefit evaluations of using formal MCS. Specifically, this classification offers insights into the underlying use and usefulness of formal MCS in addressing dysfunctionality within the accounting schools investigated in the current study.

On inspection of Table 2, at least three observations emerge as significant. Firstly, the potential costs associated with using formal MCS are perceived to be considerably greater than the range of potential benefits these MCS provide in addressing dysfunctional behaviour. These costs relate to the outcomes of the formal MCS (the effectiveness of formal mechanisms to resolve dysfunctionality), the processes necessitated by the formal MCS (the overly bureaucratic character of the procedures) and personal considerations in using the formal MCS (the stressful nature of having to undertake the process). Secondly, of the three drivers, outcomes are perceived to carry the most potential benefits from using the formal MCS, arguably representing a fall-back position (or “avenue of last resort”) or a means of protection against legal liability. Unsurprisingly, these benefits were largely cited in relation to legal forms of dysfunctionality. Thirdly, formal MCS is seen to offer a limited range of potential benefits in relation to the processual and personal drivers. The types and numbers of potential costs were perceived to be far more prevalent and were most associated with hierarchical and interpersonal forms of dysfunctionality.

Consistent with RCT, in any given decision scenario where the perceived costs exceed the perceived benefits of a particular option for addressing an observed dysfunctional behaviour, that particular option will be eschewed by decision makers in favour of other alternatives in which benefits exceed costs. Accordingly, it may well be that the perceived excessive costs of formal MCS appears to provide one explanation for the preference expressed by interviewees in this study for pursuing informal means by which dysfunctionality may be addressed. This is not to say informal processes provide a panacea or “silver bullet” in dealing with dysfunctionality; clearly, they do not. Nevertheless, our finding that university managers typically turn to informal control processes rather than formal MCS to address instances of dysfunctional behaviour adds to the empirical inventory of evidence that management control operates as a package. That is to say, decision makers retain the option to use either or both formal MCS and informal control processes in addressing a dysfunctional behaviour incident or pattern. In addition, this study provides evidence of the primacy of informal control processes over formal MCS in certain contexts. This evidence assists in explaining why such choices are made by revealing that the choice relates to the decision maker’s perception of a predominance of formal MCS costs over benefits. It is these perceived processual and personal costs that appear to be influential drivers of decision makers opting to use informal means of control in addressing dysfunctional behaviour.

Our study makes three principal contributions to the management control literature. Firstly, consistent with prior research in this area, our findings indicate that the umbrella term “dysfunctionality” comprises an assortment of behaviours. Hitherto no academic studies have offered classifications of dysfunctionality. Our study thus makes an important further contribution through its identification and articulation of the characteristics of three primary, albeit broad, forms of dysfunctionality: legal, hierarchical and interpersonal. Such classifications are important because they provide a framework for understanding the diversity of dysfunctional behaviour and thus provide a foundation for a comparison of the effectiveness of control outcomes available for use, contextualised in an organisational change environment. Moreover, classification is important, because by attributing different types of dysfunctionality to different root causes, we can gain insights into the potential effects of dysfunctionality. This classification which has emerged from the qualitative evidence presented in this study recognises that dysfunctionality extends beyond the violation of rules, regulations, codes of conduct or codes of practice. Certain conduct, while not strictly illegal, unlawful or in breach of policies, procedures or organisationally mandated guidelines, may nevertheless be described as “unethical”, “immoral”, “anti-social” or merely “objectionable”.

Beyond the identification of nuances in understandings of dysfunctionality, a second contribution of this study lies in the insights into the distinctions in the way control mechanisms are used in response to address these forms of dysfunctionality; principally, the predilection and inclination of individuals to use, in certain circumstances, informal control processes in preference to formal MCS in dealing with instances of dysfunctional behaviour. This contribution confirms and substantiates the findings of Parker (2001, 2002), Norris and O’Dwyer (2004), Marginson, (1999, 2002) and Collier (2005) in the relatively higher prominence that informal control has in comparison with formal control within the control package. That is, the inclination to use informal rather than formal controls persists, even when dealing with dysfunctional behaviour. Thus, in finding that dysfunctional behaviour is a significant threat to the use of formal MCS, this study contributes to management control knowledge by identifying distinct relative effects of formal MCS and informal control processes on organisational level outcomes. In particular, our findings point to workplaces in which dysfunctional behaviour occurs as another potential context in which informal control appears to predominate over prevailing formal controls in contemporary organisations. This has clearly been evident in the changing university environment within which this study has been set.

Our third contribution, discernible through the interpretive lens of RCT, points to and reinforces the distinction made in previous management control research between the design of formal MCS and its use. This distinction provides a diagnostic tool for, as well as insights into how control can be more effectively designed to deal with dysfunctionality depending on the predominant form of dysfunctionality and, consistent with RCT, the qualitative and quantitative costs and benefits perceived to be associated with the use of these control designs. Thus, in the parlance of this paper’s title, our study is a tale of management control within the context of dysfunctionality, and it signals a clear message as to the limits of formal MCS design and its use, and the important role played by informal control processes in supplanting formal mechanisms when the costs of using formal procedures are seen to be greater than the benefits of so doing.

For this study, a qualitative research design has been used for its relevance to the research aim of understanding how MCS processes interact with, reflect and create specific organisational events, activities and changes (van der Meer-Kooistra and Vosselman, 2006; Parker, 2012). Our purposive approach to constructing the sample in which our informants were primarily senior academics and managers necessarily precluded more junior staff. Consequently, we do not purport to provide a complete picture across the university hierarchy of the nature as well as the extent to which dysfunctionality is present in any of the universities forming part of this investigation. Nonetheless, our intent has been to deliver a foundational understanding of “what is perceived” from the standpoint of (some of) the involved senior and middle level organisational actors (Lukka, 2014; Parker and Northcott, 2016; Yin, 2014).

In so doing, we seek to kindle further dialogue on this important topic. Indeed, the three broad forms of dysfunctionality we have advanced, represent generalisable themes offering insights into the way management control is implicated in dysfunctional behaviour. This conception is consistent with the “skeletal” or “middle range” approach advocated by Laughlin (1995) that provides a conceptual language by which to explore the empirical situation without necessarily defining it (Agyemang and Broadbent, 2015), but is not “unnecessarily restrictive in forcing the empirical detail into theoretical categories” (Broadbent and Laughlin, 2005, p. 9). Our findings then, may in a sense be seen as offering “a theoretical skeleton in need of empirical flesh, providing a language by which to develop empirical understandings” in future investigations [⁶].

Similarly, this investigation is situated within an Australian academic accounting schools’ setting in an environment of significant national university sector competitive landscape and accountingised change. While it reflects an Australian university context and the particular sample of academics and professional administrators interviewed, the applicability of this study’s findings both across the Australian university sector, but also to other professional, public sector and industrial settings, has significant potential. This argument is based on Parker and Northcott’s (2016) exposition of qualitative generalisation to which this study adheres. The potential translation of this study’s generated empirical themes to other settings is facilitated by its theoretical generalisation delivered through its RCT based contextualised theorising that includes interpreting the induced meanings of the phenomenon investigated. This study’s findings’ translation to other settings is also facilitated potentially through naturalistic generalisation whereby the context and evidence outlined in this study draws on the accumulated analyses of the actors and researchers while offering a collaborative possibility for readers to draw on their own experiential and tacit knowledge to judge the findings’ applicability to other wider settings (Parker and Northcott, 2016).

In terms of investigating the wider possible pervasiveness of this study’s findings and conclusions, investigating universities in different geographical and different faculty, settings and involving lower levels of university hierarchies would also add to our MCS knowledge. Further research could also address any differences in perspectives and behaviours between teaching focussed versus research and teaching focussed academics and between tenured and contract or casual staff. From a broader perspective, as was highlighted at the onset of this paper, dysfunctionality appears to be a pervasive aspect of organisational life particularly in environments characterised by high autonomy, limited monitoring and minimal control. Thus, extending this study to such non-academic contexts and environments would also be a worthwhile avenue of further research.

Finally, as with many phenomena, dysfunctionality is amenable to investigation from alternative vantage points. In this study the theoretical lens of RCT has been used to guide our understanding of dysfunctional behaviour. However, RCT is by no means the only theoretical standpoint from which to view dysfunctionality. Almost certainly, psychological, sociological or cultural theories may provide alternate insights into our findings and may be used as theoretical platform for further investigation. Thus, an avenue for further research is the replication of this study using alternative theoretical perspectives upon which dysfunctionality may be problematised and findings may be interpreted.

This study offers insights into the lessons that might be learned about the types of dysfunctionality exhibited within accounting schools, the ways in which this dysfunctionality is addressed (and not addressed) and the relative interplay between the use of formal and informal means of control. It provides a point of departure for university managers, policymakers and academics alike for further reflection, consideration and determination. In particular, because the universities forming the evidence base comprising this study are all publicly funded, our findings are likely to be particularly relevant to public sector organisations as they relate ultimately to whether or not public universities because:

• they relate ultimately to whether or not public universities are producing desired outcomes for society;

• they have a strong interest in internal organisational processes and their contribution to achieving desired outcomes (the avoidance or mitigation of dysfunctional behaviour); and

• they may often prioritise institutional and political agendas/outcomes over traditional private sector concerns with efficiency and profit.

Arguably, every research study has its own particular “feel”, “tone” or “sense” for its researchers. As the study progresses from inception to completion, it can leave distinct impressions upon them. This study was no exception. Indeed, for the research team, it could be said that this was a salutary experience, because in many ways it reflected their “lived” experience as accounting academics. The accounts of how individuals within accounting schools sometimes treat each other was enlightening, yet on occasions disturbing. Indeed, we considered whether to include verbatim observations conveyed to us by some interviewees alleging that some colleagues (both academic and administrative) suffered from a host of medical/psychiatric/personality disorders/illnesses (including Asperger syndrome, sociopathy, psychopathy, narcissism and paranoia). We chose to not report all these comments in the paper as we recognise our limitations in being able to attest to the veracity of these allegations.

We also considered expanding upon the implications for practice suggested by our findings. Again, we decided against this given the likelihood of this section resembling kindergarten classroom posters, which frequently display exhortations such as Respect others, Respect yourself, Be safe, Be kind, Be honest, Voices are not for yelling, Words are not for hurting. In the adult world of the university, one might expect similar values to be conveyed through university values statements. Yet the accounts of dysfunctional behaviour conveyed to us by interviewees, both on and off the record, ranged from disconcerting to distressing. As experienced academics, however, many of them were not entirely surprising. Nonetheless we persist in our view that our work will in some measure contribute to the revision and employment of controls to mitigate dysfunctional behaviour and the ramifications for the casualties of dysfunctionality as well as those who have observed it.

The authors gratefully acknowledge the support provided by the Accounting and Finance Association of Australia and New Zealand (AFAANZ) in providing partial financial support for this research. This paper has also benefited from constructive comments from those who attended research seminars at Curtin University, RMIT and UniSA Business, University of South Australia as well as CPA (Australia). In addition, the authors thank all academics and university managers participating in this study. The authors are also indebted to Sue Wright, John Dumay, James Guthrie and Zahirul Hoque for their valuable comments and suggestions on previous versions of the paper.

1. How long have you been in your current position?

2. In a university environment, what do you define as “dysfunctional behaviour”?

3. What makes these behaviours dysfunctional?

4. Do you draw any distinction between dysfunctional behaviour exhibited by professional as distinct from academic staff?

5. What management controls are used to address these behaviours?

6. In what ways are such management controls implemented?

7. What outcomes are sought by these management controls?

8. How do you evaluate the effectiveness of these management controls?

9. In what ways might these controls be enhanced in managing dysfunctionality?