Editorial Free

More than 20 years ago, Adler and Borys (A&B) (1996) wrote a seminal article aiming to reconcile the conflicting views on bureaucracy in organizational research that formalizing work can either abrogate personal freedom and demotivate employees or that it can clarify responsibilities and reduce role stress. In particular, since its importation to accounting research by Ahrens and Chapman (2004), it has become a major theoretical framework to analyze accounting and control aspects in organizations (Chapman and Kihn, 2009; Englund and Gerdin, 2015; Free, 2007; Goretzki et al., 2018; Jordan and Messner, 2012; Jørgensen and Messner, 2009; Wouters, 2009; Wouters and Roijmans, 2011; Wouters and Wilderom, 2008).

Although accounting and control researchers generated manifold insights into the organizing of formal bureaucracies, not all aspects of the original framework have been exploited yet, and there is still room for potential contributions to the domain of accounting and control research. For example, A&B argued that future research should focus on the individual differences and variability in how enabling characteristics are perceived over time. In addition, the societal and organizational conditions, which surrounded not only the conceptualization of enabling and coercive control but also management accounting and control, have changed tremendously in the past 20 years. While in the former times of industrialization and mass production, the value creation of firms was based on transforming raw materials into standardized products, individualization and customization of increasingly digital products start to dominate the economic value creation. Furthermore, the societal attitude toward work has shifted from a “disutility interpretation” (also the basic assumption of economics) which assumes that working has to be compensated by material values (such as salaries) to a “utility of personal fulfilment interpretation” which sees work as a way to satisfy the personal need of self-actualization in many Western countries (Delhey, 2010).

Against this background, the purpose of our introduction to this special section is to develop future avenues for revisiting A&B’s enabling and coercive framework in light of contemporary issues, as well as expand our conceptual understanding of these concepts. Therefore, the remainder of this introduction is structured as follows. First, we give a short overview of the original framework. We then review the literature to summarize how the framework has helped to solve some of our domain specific theoretical puzzles and how accounting studies in general and in this special section in particular have contributed to the framework’s development. Finally, we develop some more general ideas on which avenues for future research might be promising.

Adler and Borys (1996) tried to understand how workflow formalization could be designed to either alienate employees or enable them to perform their tasks better. At that time, two opposing views dominated the understanding of bureaucracy in organizational research. The first view treats bureaucracy as a negative aspect of organizing work because bureaucracy abrogates individual freedom to minimize deviations from pre-set targets or behavior. Accordingly, the main function of formalizing work is to coerce employees to behave as prescribed because work is a disutility which fosters dissatisfaction and demotivates employees (Adler and Borys, 1996; Arches, 1991) that will therefore be avoided by employees (Adler and Borys, 1996). The second view assesses bureaucracy as positive because formalizing work to organize it can be experienced as cooperative and, thereby, transforming work itself as fulfilling instead of a disutility. The reason for this argument is that formalizing work reduces role stress and supports individuals to be and feel more effective; i.e. it enables employees to perform their tasks better (Adler and Borys, 1996, p. 61).

Stimulated by these two conflicting views of bureaucracy in organizational research, A&B tried to develop a new conceptualization of bureaucracy that might reconcile the two different views. A&B conceptualized formalization as an organizational technology that is assessed by actors based on whether it helps them to perform their tasks better or if it functions as a management tool that coerces them to comply. The former would lead to a positive assessment, the latter would lead to a negative one. The main assessment criteria for actors are the features of formalization (the so-called design characteristics) and its development process. The design characteristics are repair, internal and global transparency and flexibility. Repair describes to which extent actors are able to repair the formalization themselves rather than asking for help in the case of a breakdown. Coercive formalization suspects that any breakdown, i.e. individual deviations from pre-set norms, is resulting from actors’ opportunistic behavior. Consequently, coercive formalization does not allow employees to (simply) repair controls. In contrast, enabling formalization interprets breakdowns as learning opportunities as to why formalization might not work because of specific (local) contingencies and, consequently, allows (local) actors to improve it. The second design characteristic, internal transparency, refers to the extent to which formalization’s logic about its internal functioning and its status not only is transparent to actors but also is designed in such a way that actors understand both aspects (e.g. using a design which presents the status of the formalization as understandable for the users). Coercive formalization is characterized by presenting the status of the formalization to actors only in case of a malfunction and in an inaccessible language; in contrast, enabling formalization provides actors with its underlying logic to clarify the raison d’être of the rules and make them understand the current formalization status. Global transparency, the third design characteristic, “refers to the intelligibility of employees of the broader system within which they are working” (Adler and Borys, 1996, p. 73). Whereas coercive formalization does not provide actors insights into their connections to the overall formalization, enabling formalizations provide employees a broad range of information on “how their own tasks fit into the whole” (Adler and Borys, 1996, p. 73). Finally, flexibility describes to what extent actors can “modify the interface and add functionality to suit their specific work demands” (Adler and Borys, 1996, p. 74). Accordingly, in a coercive formalization, actors’ possibilities to adapt the formalization are minimized; in an enabling formalization, however, employees would be offered some room for individual maneuver. If the four features are designed in an enabling (coercive) way, it will be more likely that employees’ attitudes to the formalization will be positive (negative) which means it is seen as “needed guidance and clarifies responsibilities, thereby easing role stress and helping individuals be and feel more effective” (or “stifles creativity, fosters dissatisfaction and demotivates employees”) (Adler and Borys, 1996, p. 61).

In addition to the features of formalization per se, Adler and Borys also argue that the design and implementation process of the formalization can influence employees’ attitudes toward the formalization. Involving actors in the formulation process of formalization and, thereby, giving them the chance to elaborate their (local) needs and to formalize them, has positive effects on attitudinal and technological outcomes (Adler and Borys, 1996, p. 75). Finally, the implementation of the formalization also influences actors’ assessment of it. If the implementation does not allow any adaptation to local contingencies, actors’ attitudes will be negative and so even formalization with enabling features can be perceived as coercive.

The enabling and coercive framework was first published in the field of organization studies. As such, in the context of accounting research, the framework can be considered as a method theory, i.e. “a metalevel conceptual system, or theoretical lens, which originates from another field” (Lukka and Vinnari, 2014, p. 1312). Lukka and Vinnari (2014) argue that the first few papers using a new method theory usually focus on demonstrating that it does “fit” the domain theory, i.e. a particular set of knowledge on a substantive topic such as budgeting or performance management. However, these authors argue that to contribute to knowledge, subsequent papers need to move beyond the fit test by using the method theory to contribute to a domain theory phenomenon, by contributing to the enhancement of the method theory, or both.

Since Ahrens and Chapmans’ (2004) seminal paper, we have seen both types of contribution in the accounting literature. In this section, we will give a brief overview of the literature regarding the contribution of A&B’s framework to solve accounting domain specific theoretical puzzles and the contributions of accounting studies to the development of the original framework.

The enabling and coercive framework has helped us further develop our understanding of hierarchical relationships between corporate level/top management and local level/middle management. Indeed, “[i]t provides a framework for understanding how organizations seek to elicit flexible and local attempts to streamline and refine work processes with no necessary implications for hierarchical relationships” (Ahrens and Chapman, 2004, p. 277). For example, in the context of performance evaluation, enabling development processes offer possibility for exchange and communication between middle managers and top managers and therefore foster vertical relationships. As such, developing enabling performance measurement systems (PMS) that involve lower levels, improves knowledge integration (Englund and Gerdin, 2015; Wouters and Roijmans, 2011), which in turns increases transparency (Englund and Gerdin, 2015).

The papers in this special section offer additional insights in regard to vertical relationships. While the framework is “not meant to reflect radical change in the nature of organizations and organizational control” (Ahrens and Chapman, 2004, p. 277) and is “not mainly about decentralization” (Ahrens and Chapman, 2004, p. 296), one of the special section papers uses the framework in a radically decentralized organization (O’Grady, 2019). As O’Grady (2019) explains, this setting is interesting since this type of organizations relies less on bureaucratic forms of controls in the sense that hierarchical relationships are reduced (or even eliminated) and procedures are not imposed by senior management. Such a setting would naturally support enabling controls and leave little room for coercive controls. Yet, her case study shows that some controls have coercive characteristics. Analyzing coercive features in the light of action and results controls, O’Grady (2019) shows that the lack of repair and flexibility affects predominantly results controls. Coherent with the organizational structure of her case study, managers’ actions remained flexible. Moreover, the lack of repair of the results controls was only minimally constraining and managers’ actions. As such, O’Grady (2019) suggests that these carefully selected coercive controls (imposed by senior management) are actually supporting the highly decentralized organization. Indeed, performance measurement systems, if standardized across the organization, can provide managers from different hierarchical levels “with a common language informed by accounting and non-accounting measures and information that are spread across the different parts of the organisation” (Busco et al., 2008, p. 104).

The A&B framework has also been used to understand lateral relationships, both intra-organizational relationships, such as those fostered by transfer pricing (Cools et al., 2008) and product development management (Jørgensen and Messner, 2009), and inter-organizational relationships, such as supply-chain relationships (Free, 2007). One main finding is that lateral controls often imply the suppression of the repair and flexibility characteristics to preserve standardization (Cools et al., 2008). This is supported to some extent by the study of Jørgensen and Messner (2009) which highlights the tension between efficiency and flexibility in the context of product development, as well as potential problems with local repair. However, users seem to accept that lateral controls imply less flexibility (Chapman and Kihn, 2009; Cools et al., 2008), mainly because the loss of flexibility and repair is compensated by an increase in both forms of transparency (Cools et al., 2008). In the context of product development, transparency can be created through formal control mechanisms and informal information sharing (Jørgensen and Messner, 2009). For example, Jørgensen and Messner show a case in which formal controls such as budgets and profitability calculation tools provide internal transparency and formal gate meetings provide global transparency. To prevent a loss of global transparency for those actors who are not present at the gate meetings, communication across teams, a form of informal controls, is encouraged (Jørgensen and Messner, 2009). In other words, informal controls are used to foster lateral relationships, which in turn facilitate global transparency.

Moreover, by using the A&B framework in inter-organizational contexts, researchers have been able to reveal that coercive forms of control affecting lateral relationships can have a profound impact on the nature of these relationships. Enabling controls can enhance cooperation between buyer-supplier (winwin), however, lack of global transparency and flexibility can increase coercive negotiation oriented toward the extraction of concessions (win-lose) (Free, 2007). According to Free (2007), the level of global transparency regarding accounting systems, such as financial information, can actually reinforce the enabling/coercive approach to lateral relationships. For example, he explains that:

The framework has also contributed to our understanding of accounting information systems, specifically cost systems (Mahama and Cheng, 2012), local accounting systems (Goretzki et al., 2018; Wouters and Roijmans, 2011), performance management systems (Jordan and Messner, 2012; Wouters and Roijmans, 2011; Wouters and Wilderom, 2008), new budget practices such as rolling forecast (Henttu-Aho, 2016) and system integration (Chapman and Kihn, 2009; Goretzki et al., 2018). Accounting systems, especially PMS, are often characterized as providing incomplete information (Jordan and Messner, 2012). However, if designed in an enabling way, accounting systems can reduce the negative impact of incompleteness. For example, developing PMS in an enabling way, by engaging those whose performance will be measured “may compensate for the inherent incompleteness of performance measures” (Wouters and Wilderom, 2008, p. 511). Furthermore, in addition to an enabling development process, flexibility is also required to compensate for PMS incompleteness to allow employees to treat them “as means rather than ends when carrying out their work” (Jordan and Messner, 2012, p. 544). Flexibility can be realized through loose PMS (Jordan and Messner, 2012) and PMS that build on existing practices and concerns (Englund and Gerdin, 2015; Jordan and Messner, 2012). Tight PMS, however, reduce flexibility, which “translate into managers’ perception of the control system as being a coercive rather than enabling one” (Jordan and Messner, 2012, p. 544).

Transparency also plays a key role in the perception of enabling accounting systems (Wouters and Wilderom, 2008) and incompleteness (Henttu-Aho, 2016; Jordan and Messner, 2012). However, this role is not completely clear. On the one hand, Jordan and Messner (2012) found that internal transparency provided the necessary information to identify incompleteness of PMS. On the other hand, Henttu-Aho (2016) found the opposite i.e. that it was a lack of transparency that led to the recognition of incompleteness of the budget information. Moreover, perceiving incompleteness does not mean that the system is considered as problematic. Indeed, research shows that it is the combination of perceived incompleteness and reduced flexibility that makes it a concern for managers (Jordan and Messner, 2012). In some cases, incompleteness of accounting information can actually be repaired through new accounting practices such as rolling forecasts (Henttu-Aho, 2016). Rolling forecasts increase internal and global transparency for actors, which in turn improves their ability to provide more accurate forecasts. Moreover, the inherent flexibility of rolling forecasts improves budget work (Henttu-Aho, 2016).

In regards to local accounting systems and system integration, Goretzki et al. (2018) found that selfdeveloped systems, i.e. vernacular accounting systems (VAS) (Kilfoyle et al., 2013), can have an impact on the perception of the development process. Indeed, local actors evaluate a new global system as enabling or coercive depending on both their ability to manipulate the knowledge codified within as well as the consequences that the codified knowledge has for their authority in the local knowledge creation process. Goretzki et al. (2018) show that VAS play a vital role in both local actors' evaluation of a new system (as points of reference) and their influence on knowledge integration (as knowledge transformation devices). Furthermore, they demonstrate that local actors proceed to rely on their VAS if they realize that the proposed global system does not fit their needs. As a consequence, local actors can use their VAS as negotiation devices to reinforce their position in the development process by “threatening” the system designers of continuing to use their VAS (Goretzki et al., 2018).

One of the papers in this special section (Doornich et al., 2019) builds on this as well as on vertical relationships, by investigating how tensions between corporate managers’ intentions for the control system and local managers’ attention to it can foster change toward an enabling approach. Indeed, corporate managers’ intention of implementing a coercive and global control system led to a lack of local manager’s attention. By describing several versions of the system, the authors are able to show that the more coercive the system was, the less local managers paid attention to it. The resulting tension eventually led to a complete redesign of how the system could be used, allowing more flexibility to account for local contingences and providing more transparency regarding the system. This in turn allowed for corporate managers’ intention (compliance to rules) and local managers attention (following rules that applied locally) to become aligned.

Overall, research shows that users will use accounting systems, such as costs systems or budgets, more if they perceive them as enabling, which will then make users feel empowered, and improve task performance (Mahama and Cheng, 2012). As Chapman and Kihn (2009, p. 166) explain, “it is the enabling approach to budgeting that has direct effects on both system success and performance.” Enabling forms of accounting systems can even relate to the identities of those actors who are using the system. For example, Henttu-Aho (2016) shows a case in which controllers’ own role identity is challenged by enabling characteristics of a new budgeting practice thus providing insights into the evolution of the controller’s role toward a more business-oriented one.

Most of the papers in the accounting literature have focused on the four design characteristics of enabling and coercive controls (repair, flexibility internal and global transparency) (Ahrens and Chapman, 2004; Chapman and Kihn, 2009; Cools et al., 2008; Englund and Gerdin, 2015; Free, 2007; Henttu-Aho, 2016; Jordan and Messner, 2012; Jørgensen and Messner, 2009). As such, these studies have contributed to our understanding of these characteristics per se. For example, some studies have highlighted ambiguity in the definition of transparency. Indeed, Englund and Gerdin (2015) identified two different definitions, transparency of the design of the control itself and whether the control provides transparency into the work employees are performing. While Chapman and Kihn (2009) defined transparency as the latter (i.e. whether the budget process helps clarify activities, increase understanding of operations, etc.), Englund and Gerdin (2015) argue that these two “versions” of transparency are highly interrelated and both should be considered. Henttu-Aho (2016) makes a separate, but related point, and argues that from a global level, the distinction between internal and global transparency becomes unclear.

Some studies have argued that enabling aspects can actually appear coercive (Ahrens and Chapman, 2004). For example, Henttu-Aho (2016) found that users have a negative attitude toward enabling formalization that challenges professional competence such as modern budgeting practices (e.g.: rolling forecast), while Jørgensen and Messner (2009) suggest that incremental repair efforts could hide the need for more radical redesign. Moreover, Free (2007) found that transparency within the supply-chain relationships was associated with coercive form of control because it highlighted poor performance.

Prior research has also provided new understandings on the relationships between the characteristics. According to Chapman and Kihn (2009), repair is positively associated with the other three characteristics, internal and global transparency are positively and significantly correlated with each other, but flexibility does not seem to be correlated with the other variables. In addition, several studies have found ambivalent results regarding transparency. For example, while Englund and Gerdin (2015) found that low transparency sparked flexible use and repair of control, which in turn increase transparency, others have found that an increase in transparency resulted in a loss of flexibility and repair opportunities (Cools et al., 2008). Therefore, Englund and Gerdin (2015) explicitly challenged the “standalone” nature of the four characteristics and instead suggest a complex and evolving relationship between the four characteristics.

O’Grady (2019) further develops our understanding of the different design features by investigating coercive controls in the context of a radically decentralized organization. To understand why the observed managers do not perceive controls as coercive in such an environment, the author builds on the idea that enabling and coercive controls are on a continuum (Burney et al., 2016) and suggest that when controls are “minimally coercive,” managers will not perceive them as coercive. This concept, borrowed from the systems literature which labels it “minimal constraint,” is refined by the author’s use of action and results controls, as explained in the previous section. Indeed, according to O’Grady’s (2019) findings, control users perceive missing features, such as for example flexibility or repair, as coercive depending on what type of controls – for example, results or action controls – lacks the enabling features. This implies that to better understand why users perceive controls as they do, researchers should look at both the design features and the types of controls.

Some papers have investigated the design and implementation dimensions of enabling control (Cools et al., 2008; Englund and Gerdin, 2015; Jordan and Messner, 2012; Wouters and Roijmans, 2011; Wouters and Wilderom, 2008), providing further development of the enabling and coercive framework. For example, because design and implementation are hard to distinguish in practice, the accounting literature suggest these two dimensions should be combined under the label “development process” (Wouters and Wilderom, 2008). Furthermore, three characteristics of an enabling development process have been identified: it is experience-based (existing local knowledge), it allows for experimentation (test and review), and it builds on employees’ professionalism (orientation toward learning) (Wouters and Wilderom, 2008), with specific characteristics of experimentation in the context of enabling development processes further developed by Wouters and Roijmans (2011). Moreover, research has shown that even controls that are implemented top-down, with no room for flexibility and repair, can lead to an enabling development process if combined with other enabling information systems that do provide flexibility and options for repairing them (Englund and Gerdin, 2015). These authors also suggest that a lack of transparency plays a key role in fostering an enabling development process and vice versa. For example, target pressure combined with low transparency (i.e. how to achieve these targets is not clear), can foster an enabling development process which attempts to improve the PMS. Inversely, perceived “increased internal and global transparency worked to cease the efforts to further develop an enabling PMS) (Englund and Gerdin, 2015, p. 279). Finally, research has shown that an enabling development process can be influenced by self-developed systems (VAS) (Goretzki et al., 2018) as explained in the previous section.

Several papers that focused on more than one dimension of the framework (e.g. development process and design characteristics) suggest that these dimensions are not always aligned with each other. Indeed, while the control is being developed, there is room for experimentation, sharing of knowledge, flexibility, repair, which in turn increases transparency (Cools et al., 2008; Wouters and Wilderom, 2008). However, once the control is set, flexibility and repair are suppressed to protect transparency (Cools et al., 2008). Thus, an enabling development process does not necessarily lead to an enabling control in the end. On the other hand, an initially coercive development system can be followed by a switch to a more enabling approach. Indeed, one of the papers in this special section (Doornich et al., 2019) follows the long-term development of a control system through three iterations and shows that a coercive system can revert back to an enabling one when tensions between corporate management and local management are resolved through a dialectic motor of change (Tessier and Otley, 2012b; Van de Ven and Poole, 1995). In their case, it was the mindfulness of the rule followers that forced the rule setters to readjust the system.

Very few papers (Ahrens and Chapman, 2004; Free, 2007; Jordan and Messner, 2012) have explicitly included the goal, i.e. the manager’s view of employees, as a dimension of enabling control. In fact, it would seem that the goal is usually used as a general definition of what an enabling approach to control is, where controls are meant to enable employees to do their job better (Englund and Gerdin, 2015; Henttu-Aho, 2016; Jordan and Messner, 2012; Wouters and Roijmans, 2011) and support rather than constrain (Ahrens and Chapman, 2004; Free, 2007; Jordan and Messner, 2012; Mahama and Cheng, 2012; Wouters and Roijmans, 2011). Additionally, only a few papers explicitly characterized enabling as a matter of perception (Chapman and Kihn, 2009; Englund and Gerdin, 2015; Henttu-Aho, 2016; Jordan and Messner, 2012; Mahama and Cheng, 2012), using the description of the goal as a definition for perception and the characteristics as design features that could promote an enabling perception (Mahama and Cheng, 2012). These differences seem to result from the level of analysis of the studies, meaning that when the focus is on managers implementing controls and the design of the control itself, the goal is used as a definition of enabling control, but when the focus is on employees being controlled and the perception of the control, the goal is used as a definition of enabling perception. However, because most papers do not explicitly state which level of analysis is investigated, with a few exceptions such as Henttu-Aho (2016) (controllers’ point of view) and Mahama and Cheng (2012) (users’ point of view), the distinction between design and perception is unclear. Under such conditions, knowledge building from the accumulation of results becomes hazardous. This reinforces the point made by Tessier and Otley (2012a) that studies should clearly state which level of analysis they are focusing on, i.e. intention of managers or perception of users or both. In fact, investigating both level of analysis has highlighted interesting tensions between enabling design and enabling perception. Indeed, when controls are flexible and therefore enabling for users (middle managers), they may not be useful for the designers (top managers) (Jordan and Messner, 2012).

Most of the papers cited above have used the enabling and coercive framework in combination with another theoretical lens. Interestingly, early papers in the accounting literature combined the framework with mainstream theory such as contingency theory (Ahrens and Chapman, 2004; Cools et al., 2008) and domain theory such as typologies of control (Cools et al., 2008). This emphasized the novelty of the framework. Since then, the framework has been used with other method theories such as process view (Jordan and Messner, 2012), sensemaking (Englund and Gerdin, 2015; Jordan and Messner, 2012), empowerment (Mahama and Cheng, 2012), knowledge integration (Wouters and Roijmans, 2011), mental model of operations (Englund and Gerdin, 2015), motors of change and attention-based view (Doornich et al., 2019) and systems theory (O’Grady, 2019). Furthermore, over the years, efforts have been made to provide more precise definitions of the four design characteristics to better suit accounting needs. While flexibility, repair and transparency were originally developed for bureaucratic forms of formalization (Adler and Borys, 1996), authors such as Jordan and Messner (2012) have redefined them to better suit another form of formalization, i.e. performance indicators. This suggests that, 15 years after being introduced as a method theory, the enabling and coercive framework is now a well-established theory in the accounting domain (Lukka and Vinnari, 2014).

Albeit the A&B framework provided a useful theoretical lens for addressing theoretical accounting questions, management and accounting research is a social science. While many researchers would agree with this characterization, the consequences of it do not seem to always be reflected in the development of our research domain (Bodrožić and Adler, 2017). Especially, the influence of social, organizational and technological structures of the time when theories were developed on the results of management and accounting studies that build on those theories can be critically discussed (Cardinal et al., 2017). Of course, it could be argued that theories in social science – e.g. in management and accounting – are developed in such a way that their key aspects are independent from time and specific contemporary situations. However, this rather strict assumption can particularly be contested in social science where theory development often relies on empirical observations of contemporary social, organizational and technological structures. For example, whereas management and accounting studies in the 1950s, 1960s and 1970s made the assumption that significant institutional changes can be predicted based on cyclical patterns or known periodicity in the changes of technology or consumer tastes, studies of the twenty-first century suggest that firms face complex task environments characterized by tremendous modifications from one period to the other and, as a consequence, in which significant changes are difficult to anticipate in both timing and type (Cardinal et al., 2017). Specifically, the underlying assumptions of theories, theoretical frameworks or concepts that have gained a certain popularity – which is mainly driven by the prestige of the journals in which they are published (Judge et al., 2007) – might be less challenged than others. Depending on the specific research interest of the study, applying a theory from another “institutional decade” can be problematic mainly for two reasons. First, when interpreting the original framework, researchers might overlook or marginalize certain aspects of the phenomenon under investigation (e.g. adaptability of software) because of their contemporary scientific framing, even if those aspects were central when the framework was developed or could be important again in the future. Second, the building blocks of the original theory – such as, for example, the dominating type of work or form of organizing it – have changed over the course of time.

Against this background, we develop four main avenues for future research that can contribute to revisiting the original framework of A&B to adapt it to the twenty-first century. The first avenue is focusing on aspects of the original framework that have been marginalized or overlooked so far. The second avenue is dedicated to ideas on how to revise the framework in the light of current developments in our society. The third avenue considers new forms of organizing.

One aspect that was integrated into the original framework of A&B but has mostly been overlooked in management accounting research is the role of perception. A&B argued that actor’s assessment of bureaucracy as either enabling or coercive depends on the perception of the four design characteristics. More precisely, they stated that their:

Although A&B clearly distinguish between objective characteristics and perceptions of them, many studies don’t consider this difference or at least not explicitly (Mahama and Cheng, 2012 being an exception). Rather, as mentioned in the previous section, authors implicitly define enabling controls in a way that suggests it is a matter of perception. Yet, prior literature does provide evidence that is explicitly acknowledging perception and situating it terms of time and level of analysis is important. For example, Jordan and Messner (2012) found that although the objective design characteristics of an accounting system did not change over time, their perception by actors did. This means that depending on when actors’ perceptions are investigated, the results of the study may change.

Moreover, it seems to be important not only to distinguish between design characteristics and perception but also to investigate the reasons for diverging perceptions to understand how to “build” enabling perception as argued by Mahama and Cheng (2012). Therefore, we submit that future research should put the role of perception center stage when drawing on A&B’s framework or at least consider it as complementary to the objective design characteristics.

More precisely, we argue for three potential aspects that might be worth studying. First, future studies could focus on how individual attributes of actors might influence their perception of bureaucracy. For example, the degree of acceptance of formal hierarchy of an actor might also influence how he/she is perceiving bureaucratic structures. If he/she is more reluctant to accept formal authority than, for example, to technical authority, flexibility, repair and local transparency might play more important roles than global transparency. Second, in addition to the more general attributes of an individual, situational attributes, such as the current mood of an actor or the relationship to a specific system designer at a particular point in time, might also influence the assessment of bureaucracy. For instance, if actors have experienced a couple of breakdowns of formal bureaucracy in a very short period, repairing the system might be the most relevant design characteristic from their perspective in this specific situation. Furthermore, it could be argued that if the system designer is someone who actors dislike, the perception of the bureaucracy might be more likely coercive than enabling because the assessment of the bureaucracy is biased by the personal disfavor against the system designer.

Third, if situational attributes affect the perception, it could be interesting to unravel whether and how managers (or those responsible for implementing the formal bureaucracy) can intentionally influence the situational attributes to affect the perception of the bureaucracy in specific situations. For example, it could be interesting to reveal how managers can increase the likelihood that actors perceive new formal bureaucracy like accounting systems as enabling shortly after an acquisition to make sure that all actors of the newly formed organization use the same (accounting) systems.

Studies drawing on A&B’s framework often used it to explain why specific forms of formalization such as, for example, accounting systems have been adopted or bypassed by respective users (Doornich et al., 2019; Goretzki et al., 2018; Mahama and Cheng, 2012). One key argument is that if actors perceive formalization as enabling, they are more likely to adopt it (Mahama and Cheng, 2012). However, the assessment of the four design characteristics is influenced not only by general and situational attributes of actors but also by alternative organizational and technological structures available to actors. For example, Goretzki et al. (2018) found that actors mobilize their VAS as points of reference, knowledge transformation and negotiation devices when a new global accounting system should be developed and implemented. This insight means that the assessment of the design characteristics of a system as enabling or coercive is influenced by alternative systems and their characteristics (such as VAS) that are available to local actors in their local context. If the systems – as points of reference – available at the local context change, it can be submitted that local actors’ assessment of bureaucracy as enabling or coercive also changes. As a consequence, we argue that new technological developments, that are main building blocks of formalization in contemporary organization (Bodrožić and Adler, 2017), will influence the relevance of the four design characteristics. For example, at the beginning of the industrial digitalization in the 1990s, when first ERP systems were introduced, the customization options were kept to a minimum. Consequently, the introduction of ERP systems led to a homogenization not only of the IT landscape in organizations but also of organizational processes which were adopted to fit the IT systems’ structures (Heinzelmann, 2017; Kallunki et al., 2011). More recently, further IT development and its focus on usability has led to an increasing adjustability of IT systems to customers’ demands. For example, technological developments like high speed internet connections or cloud technology tremendously changed the cost structure of IT investment (capital expenditures were transformed to operative expenditures) and, thereby, also increased the number of options to individualize software toward original organizational processes (Seethamraju, 2015). By providing more flexibility than older versions of ERP systems, new systems will be perceived as more enabling. However, over time, this new degree of customizability might increase the relative level of expected individualization when systems are assessed. Accordingly, the importance of some design characteristics such as local transparency and flexibility might be higher today than 20 years ago when the original framework was developed. We therefore argue that perception of control is based on not only features of the control itself but also the whole ecosystem of (accounting) systems of an organization and that this is why it can change over time as suggested by Jordan and Messner (2012).

Furthermore, while prior technological influence was restricted to the physical boundaries of the organization, IT developments – such as LTE (Long Term Evolution) or cloud services – extend the boundaries of prior organizational bureaucracy to outside the organization. Remote working places become increasingly popular and employees use organizational IT systems (as a form of bureaucracy) in very different situations and settings. Accordingly, their demands on an enabling system might also change from a purely organizational perspective to a more individual perspective.

Against this background, we submit that future research should consider current technological developments and how these developments might influence the relative relevance of the four design characteristics and the overall understanding of bureaucracy in our society. Moreover, and building on O’Grady’s (2019) paper, because technological developments are reshaping hierarchical relationships and bureaucracies (e.g. remote working facilitated by cloud technology), we suggests that using the A&B framework in a wider (and maybe more “extreme”) range of organizational contexts and structures will provide learning opportunities that will nourish the development of the framework.

The abovementioned technological developments also led to new forms of organizing and management models. For example, platform organization such as Uber or Airbnb, are characterized by “distributed and often switch-role producers (sellers) and consumers (buyers) interacting with each other, digitally mediated by a third party, the platform owner. Platforms organize distributed production (Benkler, 2002) and collaborative consumption (Botsman and Rogers, 2010) without direct control over the value creation process” (Kornberger et al., 2017, pp. 79-80). This new form of organizing is granted a license to operate based on its ability to establish trust between the distributed actors through the evaluation system of ongoing transactions (ibid.). The characteristics of platform organizations also have an effect on the relevance of bureaucratic structures such as information systems because they not only are vital for the interaction of internal actors but also are the primary value creation tool of the organizations for their (external) customers. Therefore, we argue that, while the assessment of (organizational) bureaucracy is limited to members of that organization in classical (manufacturing based) organizations, in the case of platform organizations the assessment of bureaucratic structures as enabling or coercive crosses organizational boundaries. For example, in their interaction with the system, users – i.e. sellers and buyers – have to follow the bureaucratic structure that is created by Airbnb. This means that Airbnb determines what kind of information is relevant as well as how and when it is presented to users. If users such as, for example, sellers do not follow the rules inscribed into Airbnb’s bureaucratic structures – for example, showing an identification number in France – they are punished for it by exclusion from the platform. From the perspective of buyers, the platform’s structure also determines how they perform their task of renting a room – for example, by the way they can compare different rooms based on the information provided by the platform and how they can select their preferred room. As shown by Free (2007), extra-organizational relationships can be damaged by coercive controls where one dominant organization tries to extract profits from another one or supported by enabling controls that foster learning and adaptation. Investigating the relationships between Airbnb and different types of users (i.e. sellers and buyers) and the evaluation system at the basis of these relationships through the lens of the A&B framework could provide some valuable insights.

Against this background, we submit that the assessment of bureaucratic structures crosses the organizational border and might include extra-organizational members in the case of new forms of organizing such as platform organizations. This extension of the circle of potential actors assessing the bureaucratic structures of the organization might influence the way bureaucratic structures are assessed. For example, it can be argued that some kind of shared common goal of all organizational members exists which is also the basis for assessing global transparency because actors assess how their specific contingencies are reflected in the global system which supports the entire organization in achieving its goals. If individual actors who do not share a common (organizational) goal start to assess the bureaucratic structures, the relevance of global transparency might be questioned. As a consequence, it could be the case that certain design characteristics are not relevant and not considered in the assessment anymore.

Furthermore, another consequence of technological developments and a shift toward more “digital value creation” is shortened product life cycles which are enhancing the time pressure imposed on firms to successfully develop more products or portfolios of products (Cooper, 2005; Dougherty, 2001; Edmondson and Nembhard, 2009; Moll, 2015). This trend was also a main driver for the development of new management models that take these new contingencies into consideration. Management models of the 1950s to 1970s, though, were based on the assumption that significant changes in the organizational environment are based on cyclical patterns or periodicity and, as a consequence, their influence on customer preferences and demands can be predicted (Cardinal et al., 2017). In this vein, tasks and goals can be well described for employees and shared between them, because they are relatively stable, or their change can be anticipated. Consequently, formal organizational structures such as bureaucracy have been an efficient way to organize work. However, today, where technological inventions or new digital business models can emerge everywhere in the world because development costs are low, and a higher volatility of entire markets can be observed, significant changes cannot be reliably predicted anymore. In turn, this development has reduced the efficiency of classical bureaucratic structures and new forms of management interactions between members of an organization have been developed. For example, self-management (Bernstein et al., 2016) has been introduced to offer a management concept that draws on people’s own intrinsic motivation and their strive for “post-material values” such as autonomy instead of materials values such as financial incentives (Delhey, 2010). Rigid classical organizational structures such as bureaucracy that were designed in a way to reduce deviations from pre-set standards and, thereby, limited personal freedom and autonomy, are getting less accepted and, finally, become ineffective (Cennamo and Gardner, 2008; Simons, 1995), particularly under the new agile and volatile conditions.

As mentioned above, the A&B framework was developed in the field of organization studies. Using it as a method theory had been helpful to understand how management control systems can be designed and implemented in an enabling way. However, as pointed out by Radtke and Widener (2016) and Burney et al. (2016), in the field of accounting, which includes fraud prevention amongst other things, coercive forms of control could actually be beneficial. These authors call for more research on the beneficial role of coercive controls (Burney et al., 2016; Radtke and Widener, 2016). Indeed, organizations need to balance tensions between performance objectives and compliance objectives (Tessier and Otley, 2012a). Building on Simons’ Levers of Control framework, and specifically on boundaries control system, Tessier and Otley (2012a) argue that compliance objectives, supported by strategic boundary controls and operational boundary controls, are essentials to any organization, yet remain under-investigated. The concept of coercive controls could be useful in understanding the tensions between performance and compliance. Yet, to the best of our knowledge, most of the studies using the A&B framework investigate the trade-offs between efficiency/predictability and flexibility in regards to accounting systems such as PMS (Jordan and Messner, 2012), cost systems (Mahama and Cheng, 2012), budgets (Henttu-Aho, 2016), etc.

This also seems important considering recent societal developments, such as the need to protect the environment and the pressure on corporations to be good citizens, which require compliance to laws and social norms. Therefore, we argue that more research could be done using the A&B framework in circumstances which calls for the implementation of coercive controls.

It has been more than 20 years since A&B introduced the concepts of enabling and coercive control, and 15 years since it was used for the first time in the accounting and control literature by Ahrens and Chapman (2004). One of our goals in writing this introduction was to critically revisit prior accounting literature using the framework to assess the contribution of the framework to our field and the contribution of accounting scholars to the framework itself. Another goal was to propose research avenues that incorporate contemporary issues to continue the development of the A&B framework and ensure it remains relevant in the future.

A first step on those avenues was made by the two papers in this special section. O’Grady’s (2019) paper extends our understanding of enabling and coercive controls beyond typical hierarchical settings. It draws on insights from a case study of a highly decentralized organization and analyses a control package and its individual controls. One of her key insights is that control users perceive missing features, such as for example flexibility or repair, as coercive depending on what type of controls – for example, results or action controls – lack the enabling features. Accordingly, this result adds another factor (to design characteristics per se and the available local controls) that is influencing the assessment of controls. Additionally, Doornich et al. (2019) contribute to the development of the enabling and coercive framework by exploring how changes in the coercive and enabling orientations of the organizational rule system influence the attention managers pay to rules and how these dialectic changes lead to a rebalancing of the enabling and coercive orientations of the rule system. Based on insights from a case study covering a multinational energy company, they explain how rules can direct attention of local managers without limiting their perceived autonomy.