Management Accounting Research 22 (2011) 143–159
Contents lists available at ScienceDirect
Management Accounting Research journal homepage: www.elsevier.com/locate/mar
Application of the controllability principle and managerial performance: The role of role perceptions Michael Burkert a , Franz Michael Fischer b,∗ , Utz Schäffer c a b c
University of Lausanne, Faculty of Business and Economics, Quartier UNIL-Dorigny, Bâtiment Internef, 1015 Lausanne, Switzerland European Business School, Rheingaustraße 1, 65375 Oestrich-Winkel, Germany WHU - Otto Beisheim School of Management, Burgplatz 2, 56179 Vallendar, Germany
a r t i c l e
i n f o
Keywords: Controllability principle Management control systems Role theory Role stress Role ambiguity Role conﬂict Managerial performance
a b s t r a c t The controllability principle stipulates that the evaluation of a manager should be based only on elements that are under the manager’s control. Arguments for and against its application are theoretically well understood, but empirical evidence based on the evaluation of the perceptions of managers and their implications for managerial performance is scarce. By empirically analyzing the effects on managerial performance, this paper explores managers’ responses to the application of the controllability principle. We draw on role theory and analyze how role ambiguity and role conﬂict mediate this basic relationship. Moreover, we test whether application of the controllability principle equally affects role perceptions of top-level managers and those of lower and middle-level managers. Empirical analysis of survey responses from 440 managers reveals that role perceptions completely mediate the effect of application of the controllability principle on managerial performance. This effect is insigniﬁcant in the group of top-level managers, who appear to cope with uncontrollable factors more effectively. © 2011 Elsevier Ltd. All rights reserved.
1. Introduction The controllability principle is one of the strongest tenets in management accounting, and is considered to be directly relevant to evaluations of managers’ performance (Bhimani et al., 2008; Merchant and Otley, 2006; Merchant and Van der Stede, 2007). Mostly known as a normative principle, the controllability principle stipulates that managers should be held accountable only for results that are within their control (Atkinson et al., 2007; Dalton, 1971; McNally, 1980). Extant literature suggests that non-application of the principle leads to decreased motivation and increased role stress, which result in dysfunctional behavior by individual managers (Dent, 1987; Merchant, 1987; McNally, 1980). Despite this rationale for
∗ Corresponding author. E-mail address: [email protected]
(F.M. Fischer). 1044-5005/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.mar.2011.03.002
the application of the principle, there are compelling arguments as to why organizations should not always fully conform to it. Companies can direct managers’ attention to critical performance areas and gain additional information about their hidden actions by partially disregarding the principle (Ittner and Larcker, 2002; Manzoni, 2002; Simons, 2007). Given the convincing arguments both for and against applying the controllability principle, prior studies not surprisingly found that companies differ in the extent to which they follow the principle in formal performance evaluations (Drury and El-Shishini, 2005; Huffman and Cain, 2000; Merchant, 1989; Vancil, 1979). Even though researchers have been interested in the controllability principle for some time, quantitative empirical work investigating the actual consequences of its application is sparse (Drury and El-Shishini, 2005; Simons, 2007) with respect to both the expected beneﬁts of application and the potential drawbacks of strict application. Considering the theoretical importance of the principle
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
and its practical relevance for companies, this paucity of research is startling. As Giraud et al. (2008) commented, “The position of the managers evaluated has not been much examined” (p. 34). This paper aims to foster a better understanding of responses of individual managers to the application of the controllability principle, and especially the potential negative effect of non-application on managerial performance. Building on role theory (Kahn et al., 1964, 1966; Katz and Kahn, 1966, 1978), we relate application of the controllability principle to managers’ role perceptions. We argue that application of the controllability principle affects role ambiguity and role conﬂict. We assume that role ambiguity and role conﬂict, in turn, affect managerial performance. Because role theory emphasizes the importance of the hierarchical position of individuals for their role perceptions, we also account for possible moderating effects of this variable on the relationships between application of the controllability principle and role ambiguity and role conﬂict. Investigation of these effects illuminates the dynamics between application of the controllability principle and managerial performance, which is desirable from a theoretical, as well as practical, point of view. Exploration allows the derivation of recommendations on how to effectively mitigate dysfunctional consequences of non-application when full application is not desirable or feasible. We test our theoretical model with data from 440 managers and ﬁnd support for most of the proposed hypotheses. We ﬁnd strong and signiﬁcant effects of application of the controllability principle on role ambiguity and role conﬂict—the more superiors apply the controllability principle, the less managers experience role ambiguity and role conﬂict. These effects are signiﬁcant even after controlling for managers’ trust in their superior. Moreover, we ﬁnd that it is in particular a decrease in role ambiguity that has a positive impact on managerial performance. Our analyses reveal that the initial direct effect from application of the controllability principle on managerial performance completely disappears once the indirect effect through role ambiguity and role conﬂict is taken into account. Cognitive perceptions, therefore, fully mediate the dysfunctional effect of non-application of the principle on managerial performance. Moreover, we ﬁnd that the relationship between its non-application and role ambiguity is insigniﬁcant for the subsample of top-level managers. Compared to lower and middle-level managers, top-level managers seem to be better able to deal with uncontrollable factors. They are also better at withstanding dysfunctional cognitive perceptions that result from being confronted with factors beyond their control. Consequently, companies can disregard the controllability principle for their top-level managers more easily than one might have previously assumed. Our ﬁndings also suggest that companies can mitigate the dysfunctional effects of its non-application among lower and middle-level managers. Companies should consider using more effective personnel controls, which constitute an important element of management control systems (Merchant and Van der Stede, 2007). Personnel controls comprise (a) the adequate selection of people to
achieve a high person–job ﬁt and (b) training to prepare managers for new and stressful jobs. Researchers from the ﬁeld of organizational behavior ﬁnd that companies often ignore such controls (Cooper et al., 2001). An ex ante implementation of appropriate systems would help to avoid control problems, and psychological research has in fact provided evidence of the general effectiveness of such measures (Saunders et al., 1996). Our paper is organized as follows. Section 2 reviews the previous literature on the controllability principle. Section 3 introduces role theory and relates the role episode model to application of the controllability principle. Considering the application of the principle from a role theory perspective, this section develops our hypotheses. Section 4 describes the research method, data collection, and survey instruments, and Section 5 presents the empirical ﬁndings. Finally, Section 6 discusses these ﬁndings, highlights the limitations of the study, and proposes avenues for future research. 2. Review of prior literature on the controllability principle For some time, management accounting literature has put forward strong theoretical arguments for observance of the controllability principle when evaluating the performance of managers. Application of the principle seems advisable from a motivational perspective, because it ensures adequate levels of managerial effort on the job and circumvents dysfunctional behaviors (Hirst, 1983; Huffman and Cain, 2000; McNally, 1980; Ronen and Livingstone, 1975). Agency theorists have also highlighted the motivational effects of applying the controllability principle. They identify selection and effort effects and advocate its application because it motivates more skilled managers to (a) join and remain in the company (selection effect) and (b) increase or better allocate their effort (effort effect) (Milgrom and Roberts, 1992; Shields and Waller, 1988; Waller and Chow, 1985). A related stream of literature highlights the motivational effects of managerial perceptions of fairness that arise from applying the controllability principle (Giraud et al., 2008; Huffman and Cain, 2001). Application of the controllability principle, with its alleged emphasis on fairness, seems to reduce managers’ propensity to engage in dysfunctional behavior to protect themselves from such uncontrollable factors such as budgetary slack and earnings management (Collins, 1978; Dent, 1987). Furthermore, applying the principle can move the company away from an “excuse culture,” because managers waste less time trying to convince superiors that their performance stems from uncontrollable factors rather than their effort (Merchant, 1989; Modell and Lee, 2001). Previous literature also suggests role stress as an undesirable consequence for managers when companies do not apply the controllability principle to performance evaluation (Choudhury, 1986; McNally, 1980). Role stress can occur when managers regard their responsibilities as inadequately speciﬁed, and several studies have found role stress to decrease managerial performance (Fogarty et al.,
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
2000; Burney and Widener, 2007; Rebele and Michaels, 1990; Viator, 2001). Notwithstanding these established theoretical arguments in favor of applying the controllability principle, vast empirical evidence shows that it is not always strictly implemented in practice. Previous literature has emphasized that the performance of other managers signiﬁcantly affects the outcome of managers’ actions (Dent, 1987; Eccles, 1986; Frow et al., 2005; Merchant, 1989; Moon and Fitzgerald, 1996; Vancil, 1979), as do decisions made by managers’ superiors (Drury and El-Shishini, 2005; Fremgen and Liao, 1981; Ugras, 1994), and events occurring outside the company (Budding, 2004; Merchant, 1989; Modell, 1997; Otley, 1990). All three factors impede individual controllability and lead to the conclusion that “the principle of controllability is more honoured in the breach than its observance” (Marginson and Ogden, 2005b, p. 49). In light of the seeming paradox between the intuitive logic of the controllability principle and its apparently limited application in practice, scholars have traditionally related the equivocal relevance of the principle to high levels of complexity and uncertainty (Choudhury, 1986; Gibbs et al., 2004; Hartmann, 2000; Hirst, 1983; Marginson and McAulay, 2001; Merchant and Manzoni, 1989).1 Meanwhile, evidence is mounting that—apart from complexity and uncertainty as practical reasons for the non-application of the principle—companies often do not want to apply it in their management control systems. Companies often establish accountability arrangements that purposely differ from the prescriptions of the controllability principle. Even if organizational realities allow for applying the principle, application may not always serve the interests of the company. A less than full application of the controllability principle can direct managers’ attention to critical performance areas. If managers were completely shielded from the impact of uncontrollable factors, they would be likely to devote less attention to those areas than they would if the measured performance included the impact of uncontrollable factors. According to Manzoni (2002, p. 32), “overall ﬁrm performance is unlikely to be maximized by focusing managers’ efforts on dimensions and measures that are controllable by them”. Studies conﬁrm empirically that companies include uncontrollable components in evaluations and expect managers to take these factors into consideration when performing their job (Bourguignon and Chiapello, 2005; Giraud et al., 2008; Simons, 2007). Agency-based research provides several theoretical explanations as to why companies include uncontrollable factors in the evaluation of managers’ performance. In an early work, Demski (1976) analyzes conditions under which managers may reasonably be held accountable for uncontrollable factors. He ﬁnds that, if the principal is risk-
1 Complexity implies that revenue and cost items are most often jointly earned or incurred. Very few items are under the sole inﬂuence of one manager (Atkinson et al., 2007; Bhimani et al., 2008). Furthermore, uncertainty may lead to difﬁculties not only in formalizing ex ante work using job descriptions, standards, and ﬁxed procedures (Grifﬁn et al., 2007), but also in judging ex post whether performance is the result of managerial effort or simply of circumstances (Merchant, 1989, 1998).
averse, the use of uncontrollable factors in performance evaluation leads to more balanced risk-sharing between the principal and the agent and, as a consequence, to more efﬁcient outcomes. Likewise, the actions of a risk-averse agent will better align with the interest of the principal when the agent bears some risk as well. Such an arrangement will discourage the agent from taking actions that do not reﬂect the utility function of the risk-averse principal. Later agency works show analytically that the principal may take advantage of performance measures that the agent cannot fully control as long as they yield additional information about the agent’s unobservable actions (Antle and Demski, 1988; Ittner and Larcker, 2002; Holmström, 1979; Milgrom and Roberts, 1992). Researchers have found this argument to be relevant in multi-agent settings as well (Baiman and Demski, 1980; Holmström, 1982). Holmström (1982) demonstrates analytically that the principal beneﬁts from using a relative performance evaluation based on aggregate measures such as peer group comparisons. Relative performance evaluation implies a relaxation of the controllability principle because the agent does not have the performance of peers under his/her control. Nevertheless, relative performance evaluation provides the principal with additional information about the agent’s effort (Choudhury, 1986). Empirical evidence supports the idea that companies emphasize information content in their choice of performance measures. In fact, companies often seem to favor information content over strict controllability (Bushman et al., 1995; Frederickson, 1992; Maher, 1987).2 3. Hypothesis development The review of the existing literature on the controllability principle makes obvious that, from the individual manager’s perspective, application of the principle may avoid dysfunctional effects, whereas from the organization’s point of view cogent reasons suggest disregarding it. In developing our hypotheses, we focus on cognitive perceptions of individual managers and the resulting performance implications. Our goal is to contribute to extant research by empirically testing role perceptions that are rooted in non-application of the controllability principle. The literature discussed in the previous section provides the basis for understanding the fundamental relationship
2 Baiman and Noel (1985) go one step further and show analytically that in some situations the principal may beneﬁcially include completely uncontrollable cost items in the agent’s performance evaluation. In a multi-period setting, the principal has to decide for each period whether to continue business operations. Capacity costs are relevant to the principal’s decision. Likewise, the agent has to decide on his/her actions for each period and can choose between short-term actions without an effect on future periods and long-term actions with positive, multi-period effects on outcome but requiring more effort. Baiman and Noel (1985) demonstrate the value of using the past period’s capacity costs for the agent’s compensation. These costs aid in adjusting the agent’s actions to the likely capacity decision of the principal in the next period, as present capacity costs are linked to future capacity costs. The rationale for including uncontrollable factors in performance evaluation is not to gain information about the agent’s hidden past actions but rather to align the agent’s future actions with probable future decisions of the principal.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
between the application of the principle and managerial performance. Application of the controllability principle can lead to higher levels of motivation, higher personal effort, and reduced levels of dysfunctional behavior (Bouwens and Van Lent, 2006). Overall, we expect the application of the principle to have a positive effect on managers’ individual performance.3 However, we follow role theory in assuming this relationship to be indirect in nature. We introduce managers’ role perceptions as mediators for the relationship between the application of the controllability principle and managerial performance. 3.1. Application of the controllability principle and role perceptions of managers Prior research indicates that superiors have distinct expectations of the behavior and role of managers in organizations and these expectations determine whether the controllability principle is applied. In some cases, superiors opt for a strict application of the principle because their intent is to safeguard overall motivation. They do not want to distract managers from the core of their work and place importance on reliable job performance. In other cases, they opt against following the controllability principle because they expect managers to bear some risk in performing their job, to proactively respond to uncontrollable factors, and to develop a more entrepreneurial attitude toward their role in the company (Cohen et al., 1992; Dent, 1987; Frow et al., 2005; Merchant, 1998; Simons, 2005, 2007). The notion that well-deﬁned role expectations establish the application of the controllability principle is important for this paper, as this idea represents the link between extant literature on the controllability principle and role theory. Hence, we argue that the concept of role expectations allows for integration of role theory into research on the controllability principle. Role theory centers on the interplay between a focal person—here, the subordinate manager—and his/her role senders, or in this context, mainly the manager’s superiors in business organizations. Role expectations form one of the core concepts of this theory and the starting point of the role episode. Role expectations comprise the various prescriptions and proscriptions role senders hold with regard to what they expect from the focal person. As such, expectations are evaluative standards that deﬁne behavioral requirements or limits ascribed to a particular role. Role theory assigns the term “role sending” to acts of communicating expectations that describe role pressures. This study considers these expectations as pressure on subordinate managers to behaviorally conform to their superiors’ expectations (Kahn et al., 1964; Machin, 1979). Role theory predicts that role sender’s expectations and the accompanying pressures on the focal person affect the lat-
3 The literature does not provide reasons to believe that, theoretically, organizational beneﬁts from the non-application of the principle are reﬂected in the managers’ perception of their own managerial performance.
ter’s immediate experience. The speciﬁc conﬁguration of role expectations and role pressures the superior imposes on the subordinate manager (i.e., whether the conﬁguration allows for rational problem solving) can cause the manager to experience role stress, or a feeling of being unable to accomplish all role demands. Role stress encompasses two facets: role ambiguity and role conﬂict. Role ambiguity is a “discrepancy between the amount of information a person has and the amount he requires to perform his role adequately” (Kahn, 1974, p. 59). Role conﬂict refers to “conﬂicting expectations and organizational demands in the form of incompatible policies, requests, standards” (Rizzo et al., 1970, p. 155). An integral part of role theory is the response by the focal person to role stress (Kahn et al., 1964, p. 28). Possible responses include behavioral coping efforts (such as compliance, withdrawal, or changes in performance levels) and affective symptom formation (such as satisfaction or tension). Our review of the literature on the controllability principle shows that superiors pursue speciﬁc purposes when they decide whether to apply the principle to the evaluation of managerial performance (Cohen et al., 1992; Dent, 1987; Frow et al., 2005; Merchant, 1998; Merchant and Otley, 2006; Simons, 2005, 2007). Put differently, we assume that the practice of confronting managers with uncontrollable factors represents an important communication tool and a conscious act that reﬂects a superior’s distinct expectations. However, when managers cannot inﬂuence the performance measures their superiors use to evaluate their performance, they tend to lose conﬁdence in the effectiveness of their own behaviors and decisions. As Hirst (1981) puts it, “where performance measures are not controllable, subordinates are in an ambiguous situation because they cannot be sure what actions are likely to result in favorable performance” (pp. 776–777). Hence, managers’ knowledge about behavior–reward linkages that formal management control systems customarily prescribe is likely to diminish. Without certainty about such causeand-effect relationships, managers will lack reliable and consistent information about the ways to fulﬁll role expectations imposed on them (Ilgen and Hollenbeck, 1991; Tubre and Collins, 2000). Moreover, feedback provided by management control systems is likely of poor quality when uncontrollable factors distort managers’ performance (Huffman and Cain, 2000). Many researchers state that one of the primary functions of appraisal systems is to provide feedback to employees (Ilgen et al., 1979; Luckett and Eggleton, 1991; Murphy and Cleveland, 1995). However, this function is adversely affected when superiors evaluate managers on the basis of factors beyond their control, because managers may have difﬁculty deriving strategies to improve their future performance. The intimate relationship between non-availability of information and the resulting role ambiguity leads us to argue that non-application of the controllability principle relates to the experience of role ambiguity in managerial jobs. The foregoing discussion results in the following hypothesis:
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
H1. Application of the controllability principle is negatively associated with role ambiguity. We have outlined above that role conﬂict, the second facet of role stress, is the result of incompatible policies, requests, or standards (Rizzo et al., 1970). From the perspective of the managers evaluated, we consider a superior’s decision not to apply the controllability principle to be an incompatible request. That is, if the controllability principle is not fully applied, managers are expected to meet performance targets over which they have insufﬁcient formal authority or control, possibly resulting in role conﬂict (Choudhury, 1986; McNally, 1980). The original formulation of role theory notes that a lack of formal authority is an indicator for conﬂicting role expectations (Kahn et al., 1964), potentially causing role conﬂict. With non-application of the principle, managers are torn between what is formally required of them and what they are able to directly control through the authority granted to them. Dent’s (1987) case study on the application of the controllability principle at a U.S. technology company conﬁrms that “managers ﬁnd their responsibilities quite stressful” (p. 135) when they do not have enough formal authority. We follow Kahn and Quinn (1970) in arguing that stressors such as role conﬂict may arise not only from the mere lack of authority, but especially from the discrepancies between role expectations and authority. We propose that the less the controllability principle is applied, the higher the discrepancies and the subsequent levels of role conﬂict. Conversely stated: H2. Application of the controllability principle is negatively associated with role conﬂict. 3.2. Role perceptions and managerial performance Role theorists have traditionally argued that role ambiguity is detrimental to individual performance. As Jackson and Schuler (1985) note, cognitive processes can explain the proposed negative relationship between role ambiguity and performance. High levels of role ambiguity imply conditions of uncertainty that may lead to a narrowing of the perceptual attention of individuals and a failure to recognize performance-related cues that negatively affect their job performance. In situations of role ambiguity, individuals are primarily engaged in clarifying role expectations and, thus, coping with role stressors. As this behavior diverts individuals’ endeavors away from core job functions, it reduces their ability to perform (Gilboa et al., 2008; Jamal, 1985). Likewise, individuals will “hesitate to make decisions and will have to rely on a trial and error” when suffering from role ambiguity (Rizzo et al., 1970). Therefore, it has been argued that role ambiguity may lead to poor decision making (Kahn et al., 1964). Since the early research of Kahn and his colleagues on the effects of role ambiguity, there has been a multitude of empirical studies examining the relationship between role ambiguity and individual performance or performancerelated variables. Most investigations support the negative effect of role ambiguity on performance (Gilboa et al., 2008; Tubre and Collins, 2000). Recognizing that role ambiguity mediates the relationship between budget par-
ticipation and managerial performance, Chenhall and Brownell (1988) are among the ﬁrst management accounting scholars to predict and ﬁnd the negative effect of role ambiguity on performance.4 Later work has conﬁrmed the negative relationship between role ambiguity and managerial performance in management accounting settings (Fogarty et al., 2000; Burney and Widener, 2007; Rebele and Michaels, 1990; Viator, 2001). In light of these empirical ﬁndings and concordant results reported in meta-analyses on role ambiguity (Gilboa et al., 2008; Jackson and Schuler, 1985; Tubre and Collins, 2000), we suggest: H3. Role ambiguity is negatively associated with managerial performance. The theoretical arguments presented in the foregoing discussion on role ambiguity also apply to the relationship between role conﬂict and managerial performance. Again, the literature commonly offers cognitive explanations when theorizing about the supposedly detrimental effect of role conﬂict on performance (Jackson and Schuler, 1985; Tubre and Collins, 2000). Experiencing role conﬂict hinders performance because it leads individuals to perceive “an almost impossible situation for doing everything expected” (Jackson and Schuler, 1985, p. 43). When role expectations seem to be incompatible and contradictory, the individual’s behavior is most likely to be misdirected and insufﬁcient, regardless of how much effort is being made (Jackson and Schuler, 1985; Tubre and Collins, 2000). As far as the empirical evidence for such cognitive processes is concerned, the original research by Kahn et al. (1964) on the effects of role stress indicates that role conﬂict might be detrimental to an individual’s performance because of social and psychological withdrawal. Assuming that withdrawal reduces both commitment and inﬂuence at the workplace, effectiveness of an individual is also likely to deteriorate owing to withdrawal. Early work on role theory supports this assumption (Dubinsky and Mattson, 1979; House and Rizzo, 1972; Szilagyi et al., 1976). However, more recent studies by management accounting researchers show mixed results for the relationship between role conﬂict and individual performance. In line with the aforementioned studies, Fisher (2001) ﬁnds that role conﬂict is negatively associated with the (selfreported) auditor’s performance. Marginson and Bui (2009) recently analyzed the causes of role conﬂict and its negative effect on managerial performance and found that middlelevel managers’ performance is signiﬁcantly negatively associated with role conﬂict. Related studies report a weak and non-signiﬁcant relationship between role conﬂict and self-reported measures of performance when other role constructs are included in the model (Fogarty et al., 2000; Burney and Widener, 2007; Rebele and Michaels, 1990;
4 In line with these results, Dunk (1993) ﬁnds that job-related tension (a construct incorporating role ambiguity, role conﬂict, and role overload) has a linear, negative effect on managerial performance. It is worth noting that Dunk explicitly tests whether the relationship between job-related tension and managerial performance is linear or curvilinear. Not ﬁnding signiﬁcant support for a curvilinear relationship, he concludes that the negative effect of job-related tension on managerial performance is linear.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Viator, 2001). Likewise, meta-analyses investigating the relationship between the two variables demonstrate that, even though a negative relationship exists between role conﬂict and performance, it is only moderately strong at best and weaker in magnitude and signiﬁcance than the performance effect of role ambiguity (Fisher and Gitelson, 1983; Gilboa et al., 2008; Jackson and Schuler, 1985; Tubre and Collins, 2000). However, the relationship appears to be stronger at the managerial level (Tubre and Collins, 2000; Gilboa et al., 2008), which is the position under investigation in this study. This argument in combination with the strong theoretical arguments for a negative effect of role conﬂict on performance leads us to hypothesize: H4. Role conﬂict is negatively associated with managerial performance. The relationships depicted in H1–H4 represent a mediation model. We explicitly assume that the relationship between the application of the controllability principle and managerial performance is at least partly mediated by the role stress constructs. Therefore, we assume that the direct effect of the principle’s application on managerial performance decreases signiﬁcantly in the presence of the role stress variables. 3.3. Moderating effect of managers’ hierarchical level As outlined above, we believe that the application of the controllability principle has a causal effect on role stress, which in turn may affect managerial performance. Further examination of whether the relationships between the application of the principle and role stress depend on the managers’ organizational context is important. Previous literature on the controllability principle supports this view and calls for investigating the effects on perceptions of managers in relation to the broader organizational context (Dent, 1987; Frow et al., 2005; Simons, 2005). Katz and Kahn (1966) suggest that the hierarchical level of managers may be a central contextual factor. The general importance of this variable as a moderator is well established by numerous studies in organizational behavior and applied psychology (Aquino, 2000; Schminke et al., 2002; Stahl and Caligiuri, 2005). Not surprisingly, this variable has also attracted the interest of management accounting researchers. A body of literature reports signiﬁcant moderating effects of an individual’s hierarchical level on the relationships between management accounting variables and behavioral outcomes (Davis and Kohlmeyer, 2005; Dunk, 1992; Viator, 2001). As outlined in the literature review section, a potential beneﬁt of disregarding the controllability principle is that the attention of managers may turn to important performance areas outside of their immediate control. Given the high impact of top-level managers’ decisions and actions on their organizations (Hambrick and Mason, 1984), nonapplication of the principle appears to be more suitable for this group of managers than for others. We therefore consider the manager’s hierarchical level to be a central contextual factor and potential moderator for the effect that applying the controllability principle has on role stress. Conceptually, we argue that top-level man-
agers constantly face high levels of uncertainty (Cannella et al., 2008; Carpenter and Fredrickson, 2001), which implies that top-level managers have to deal with uncontrollable factors on a daily basis. Plausibly, then, managers at higher hierarchical levels may be more prone to accept a lack of controllability than managers at lower hierarchical levels and may also experience less role stress when being confronted with uncontrollable factors. We therefore propose: H5. The effects of application of the controllability principle on role ambiguity and role conﬂict are negatively moderated by the managers’ hierarchical level. Fig. 1 depicts our hypotheses. Trust is often one of the most important determinants of employee feelings toward performance appraisals (Brion, 1989; Henderson, 1980). In the context of this study, we argue that managers’ cognitive perceptions of role ambiguity and role conﬂict are negatively associated with trust in their superiors. Role theory is not entirely clear as to whether trust might directly affect role ambiguity and role conﬂict or moderate the relationship between stress stimuli and role stress (Katz and Kahn, 1966, 1978). To analyze whether the hypothesized effects hold after controlling for trust, we use managers’ trust in their superiors as a control variable with an impact on both role ambiguity and role conﬂict. Additionally, we test for a potential moderating role of this variable as part of the robustness check in Section 5. 4. Method 4.1. Sample selection and data collection The data collection involved a survey questionnaire directed at 3500 managers of German companies. All companies had more than 200 employees, and their selection provided some degree of control over the size of the companies in our sample. In surveying business unit and functional managers, we were consistent with previous studies on the controllability principle and focused on similarly high-ranking managers (Giraud et al., 2008; Merchant, 1989; Simons, 2007). Following Giraud et al. (2008), we excluded support function managers because their performance is difﬁcult to deﬁne and measure, which makes the controllability principle less relevant. The managers were randomly chosen from the Hoppenstedt database, the market leader in Germany for providing professional contact details of business managers for direct marketing purposes. Our survey instrument was web-based. We sent an individualized e-mail to each manager with the request to participate, a unique and anonymous link to the online questionnaire, and a link to the web page of one of the sponsoring universities. In designing the survey instrument, we followed the guidelines provided by Dillman (2007), taking several measures to increase both the quality of the instrument and the response rate. First, we made the website visually appealing. Second, whenever possible, we used scales from the extant literature to measure our constructs. Third, we drafted our questionnaire in English, translated it
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
H5 Role ambiguity Application of the controllability principle
Control variable: Trust
H3 n.s. Managerial performance
H4 Role conflict
Control path Fig. 1. The proposed mediating and moderating effects in the relationship between the application of the controllability principle and managerial performance.
into German, and then translated it back into English using the translate–retranslate procedure (Brislin, 1980; Van de Vijver and Hambleton, 1996). Fourth, we tested the German questionnaire with 45 managers to assess its length, understandability, and general attractiveness, as well as to ensure adequate reliability and validity of all measurement scales. This pilot test led to some minor adaptations of the wording of some items. Finally, after sending the initial email to the respondents, we twice reminded respondents to participate in the survey. Of the 3500 managers contacted, a total of 456 participated in our survey. We excluded 16 of the responses because they were incomplete. The remaining 440 usable responses constitute a response rate of 12.6%, which is comparable to other recent management accounting surveys (Homburg and Stebel, 2009; Widener, 2006). We tested for non-response bias according to Armstrong and Overton (1977). We split our sample into two equally large groups according to the return date of the questionnaire. We compared the ﬁrst 100 respondents to the last 100 respondents and found no evidence of response bias. The demographic data of the respondents, as shown in Table 1, reveal that the managers in our sample are highly experienced and hold positions of responsibility.
4.2. Measurement instruments 4.2.1. Application of the controllability principle We developed a new measurement instrument for assessing the application of the controllability principle in management control systems. Research suggests that an individual’s behavior is not necessarily based on the actual system design but depends on how that person perceives the system and its use in an organizational reality (Robbins, 2005). In line with other recent studies examining design characteristics of management control systems (Burney et al., 2009; Hartmann and Slapniˇcar, 2009), we focus on examining managers’ perception to understand subsequent behavioral responses. We relied on the self-reported perceptions of managers of the extent to which they viewed the performance measures for their evaluation as controllable. We explicitly asked managers to evaluate only those measures that were used in the appraisal of their individual performance. Bisbe et al. (2007) argue that the application of the controllability principle is unobservable and depends on two separate dimensions that jointly determine the meaning of the construct: (1) precision of performance measures (i.e., lack of noise or variability in the measures) and (2) sensitivity of performance measures (i.e., changes in the measures triggered by the managers’ actions).
Table 1 Sample descriptives. Age
Tenure in position
<30 yrs 30–35 yrs 36–40 yrs 41–45 yrs 46–50 yrs
9 36 82 127 89
<1 yrs. 1–3 yrs >3 yrs (n) Hierarchical level
35 121 284 440 n
51–55 yrs >55 yrs (n)
55 42 440
Lower and middle mgmt. Top mgmt. (n)
285 155 440
This type of operationalization goes back to the work of Banker and Datar (1989), Demski (1994), and Feltham and Xie (1994). Managers will not perceive that the controllability principle has been applied unless the measures for performance evaluation are both sensitive and precise at the same time. We used the constructs originally developed by Moers (2006) to measure the precision and sensitivity of performance measures and adapted them to the speciﬁc context
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Application of the controllability principle
Precision of performance measures
Sensitivity of performance measures
Fig. 2. MIMIC model for the construct “Application of the controllability principle”.
of this study. A total of four items recorded the precision of performance measures (Pre 1–Pre 4, see Appendix A). The ﬁrst item relates to noise in the performance measures, the second to the possible distortion of performance measures caused by uncontrollable factors, the third directly to their precision, and the fourth to the inﬂuence of uncontrollable factors on the performance measures. These four items yield a Cronbach’s alpha of 0.89, which slightly exceeds the Cronbach’s alpha of 0.81 reported by Moers (2006) for the respective construct in his study. We drew on another four items to assess the sensitivity of the performance measures (Sen 1–Sen 4, see Appendix A). The questions capture to what extent the manager’s effort and actions inﬂuence the measures used for performance evaluation. We obtained a very satisfactory Cronbach’s alpha of 0.85 for this construct as well, which also exceeds the Cronbach’s alpha of 0.70 reported by Moers (2006) for this construct. Because serious misspeciﬁcation problems may arise from wrong epistemic relationships between constructs, dimensions, and indicators (Bollen, 1989; Bollen and Lennox, 1991; Edwards and Bagozzi, 2000; Jarvis et al., 2003; Petter et al., 2007), we followed Bisbe et al. (2007) in developing the measurement scale for the application of the controllability principle. The resulting scale depicts the construct as a higher order (formative second-order) construct. The construct is deﬁned as a combination of the two ﬁrst-order reﬂective sub-dimensions: precision of performance measures and sensitivity of performance measures. These two sub-dimensions together form the second-order formative model which represents a MIMIC (multiple indicators and multiple causes) model (see Fig. 2). MIMIC models offer a particularly suitable approach for solving error indeterminacy and identifying multi-dimensional formative measurement models (Diamantopoulos and Winklhofer, 2001; Jarvis et al., 2003). As recommended, we included three reﬂective indicators on the second-order level of the MIMIC construct that directly assess the application of the controllability principle (CONTRO 1–CONTRO 3, see Appendix A). This approach is required to make the measurement parameters more stable and less sensitive to changes in the structural relationships.
4.2.2. Role conﬂict and role ambiguity We measured both variables with the traditional items originally developed by Rizzo et al. (1970). Although recently criticized for not sufﬁciently capturing whether the focal person actually experiences psychological distress from role expectations (Siegall, 2000), Rizzo et al.’s items are widely accepted and have been used in the vast majority of role stress studies (Jackson and Schuler, 1985; Fisher, 2001). Eight items measure role conﬂict. We asked respondents about the existence of several sources of role conﬂict in their job. Rizzo et al.’s (1970) measurement instrument is the most consistently used scale in studies on role conﬂict (Gilboa et al., 2008; Jackson and Schuler, 1985; Van Sell et al., 1981) and is also widely employed in management accounting studies on role stress (Burney and Widener, 2007; Viator, 2001). We relied on the traditional six-item scale developed by Rizzo et al. (1970) to measure role ambiguity. The instrument comprises reverse-coded items that ask the respondents about the clarity of their authorities, responsibilities, and objectives. Role stress studies across many disciplines, including management accounting, have frequently used the Rizzo et al. (1970) instrument for measuring role ambiguity (Gilboa et al., 2008; Jackson and Schuler, 1985; Van Sell et al., 1981). Prominent management accounting studies adopting this scale are Chenhall and Brownell (1988) and Fogarty et al. (2000). Gregson et al. (1994) provide further evidence in an accounting setting that support the construct validity of the Rizzo et al. (1970) measures for both role conﬂict and role ambiguity. 4.2.3. Managerial performance To measure managerial performance, we used the nineitem scale developed by Mahoney et al. (1963, 1965). This instrument is frequently applied in contemporary management accounting studies (Chong and Chong, 2002; Hall, 2008; Marginson and Ogden, 2005a; Otley and Pollanen, 2000; Parker and Kyj, 2006). It consists of a single overall performance rating combined with ratings on eight subdimensions of managerial activities. Consistent with most prior research (Brownell and Dunk, 1991; Brownell and McInnes, 1986; Chong and Chong, 2002; Kren, 1992; Lau et al., 1995; Parker and Kyj, 2006), we used the single overall rating of performance as the measure of managerial performance to test our hypotheses. Since respondents were guaranteed anonymity, we could not obtain ratings of the managerial performance from their superiors. Instead, we asked the respondents themselves to indicate the extent to which their performance was below or above average. Heneman (1974) compared such self-rated performance ratings to ratings supplied by superiors and concluded that the Mahoney et al. scale “possessed less leniency, restriction of range, and halo effect than did superior ratings” (p. 638). 4.2.4. Hierarchical level of the manager We adapted Dunk’s (1992) instrument to assess the managers’ hierarchical level in their organization. We reproduced an organizational chart in the questionnaire for managers to indicate their managerial level. We dis-
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Table 2 Descriptive statistics for all latent variables. Variable
PRE SENa RA RC MP TRUST
No. of items
4 4 4 6 1 4
3.57 5.27 2.16 4.37 5.79 4.99
1.39 1.03 0.92 1.25 0.61 1.30
1.00 1.00 1.00 1.00 1.00 1.00
7.00 7.00 7.00 7.00 7.00 7.00
1.00 2.00 1.00 1.17 4.00 1.00
Cronbach ˛ Max
7.00 7.00 5.75 7.00 7.00 7.00
Good (acceptable) ﬁt > 0.7 (0.6)
0.89 0.85 0.82 0.80 – 0.84
3.40 1.37 1.99 0.99 – 0.75
0.07 0.03 0.05 0.00 – 0.00
1.00 1.00 1.00 0.99 – 0.99
1.00 1.00 1.00 1.00 – 1.00
a The application of the controllability principle is operationalized as a formative second-order construct. Evaluation tools based on internal consistency do not apply to the respective measurement model (Bisbe et al., 2007). That is why this table only depicts the descriptive statistics for the construct’s ﬁrst-order dimensions, that is, precision of performance measures (PRE) and sensitivity of performance measures (SEN). n = 440.
tinguished three levels (lower level management, middle management, and top-level management). Providing a generic organization chart in the questionnaire avoids having respondents simply report their title, which only corresponds to a meaningful level in the context of the speciﬁc companies of the respondents (Berger and Cummings, 1979).
extracted variance for each variable to the squared interpair correlation. All average extracted variances exceeded the respective squared correlations. On the basis of the results of the factor analyses and the Fornell and Larcker test, we conclude that the variables used in our study are distinct.5 5.2. Test of hypotheses
4.2.5. Trust in superior We measured the variable trust in superior as a control variable for the model. We used Read’s instrument (1962) containing items that intend to reﬂect trust or conﬁdence of respondents in the motives and intentions of their superiors. These items refer to matters relevant to the respondents’ career and status in the company. Prior studies in the ﬁeld of management accounting have used this variable (e.g., Lau and Tan, 2006).
We employed covariance-based structural equation modeling (SEM) with LISREL 8.80 (Jöreskog and Sörbom, 2001) to test our hypotheses.6 5.2.1. Base model: Hypotheses 1–4 Fig. 3 presents the test results for the direct and indirect relationships of our mediation model. We evaluated the base model using the chi-square test statistic (459.10), the chi-square divided by the degrees of
5. Analysis and results 5.1. Test of construct quality To ensure robustness in the underlying factor structure and to test the quality of the constructs, we performed exploratory and conﬁrmatory factor analyses. They included the items of all latent variables of our study. As expected, the initial factor solution from the exploratory factor analysis contains seven factors with eigenvalues greater than 1.0 (i.e., one factor for each reﬂective ﬁrst-order construct). In reﬁning the measures, we retained all factor loadings above 0.40 (Hair et al., 2006). We dropped two items because they signiﬁcantly cross-loaded on two factors (DeVellis, 1991; Verbeke and Bagozzi, 2002). To further enhance our measurement instruments, we performed conﬁrmatory factor analyses for each of the identiﬁed factors. This procedure led us to eliminate two more items owing to low standardized factor loadings. Table 2 presents a summary of the results of the conﬁrmatory factor and descriptive analyses for all latent variables in our study. The items that we eliminated as a result of the factor analyses are marked with an asterisk in Appendix A. In an attempt to ensure discriminatory validity between the latent variables, we applied the technique suggested by Fornell and Larcker (1981). We compared the average
5 Although the results of the factor analyses provide an initial indication that common method bias is not a major issue in this study, we applied another recently proposed test (Podsakoff et al., 2003). Our measurement scale for trust in superior served as a marker variable. To refute common method bias, it has to be shown that no signiﬁcant and strong correlation exists between the marker variable and one other variable of the study between which, on theoretical grounds, one would not expect any correlation (Lindell and Brandt, 2000; Lindell and Whitney, 2001). The idea draws on the fact that common method bias affects every variable in a questionnaire. If a strong correlation exists between two variables that are not expected to be correlated, the strength of such a correlation can be interpreted as the degree to which common method bias is a problem (Podsakoff et al., 2003). We analyzed the correlation coefﬁcients of our trust variable with three speciﬁc dimensions of managerial performance that we thought should not be correlated with trust (MP items 2, 3 and 7, see Appendix A). The insigniﬁcant correlation coefﬁcients ranging from 0.03 to 0.07 corroborate the ﬁndings from the factor analyses and provide additional evidence that common method bias is not a major concern for our study. 6 We chose SEM because it offers several advantages over regression and path analyses. Generally, SEM with LISREL allows for separating out measurement error at the individual item level (Smith and LangﬁeldSmith, 2004). The better measurement errors can be accounted for, the more precise the paths can be estimated. Moreover, SEM with LISREL provides the researcher with the opportunity to test more complex relationships by including several dependent and/or mediating variables (Henri, 2007). This is a particular advantage for our study, as we test a mediation model ﬁrst, before examining the moderating effect of the managers’ hierarchical level on some of the basic relationships.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Model fit statistics: χ2 df χ2 / df RMSEA GFI CFI NFI NNFI SRMR -0.22 ***
Application of the Controllability Principle
459.10 283 1.622 0.038 0.926 0.986 0.965 0.983 0.0621
Threshold for good (acceptable) fit
< 2 (3) < 0.05 (0.08) > 0.95 (0.90) > 0.95 (0.90) > 0.95 (0.90) > 0.97 (0.95) < 0.05 (0.1)
-0.49 *** Control variable: Trust
Managerial performance -0.41 *** n.s.
_______________________________ Fig. 3. SEM results for the test of the base model. Asterisks ** and *** indicate the signiﬁcance of the p-value at < 0.05 and 0.01 respectively (two-tailed); n = 440.
freedom (1.622), the root mean square error of approximation (RMSEA; 0.038), the goodness of ﬁt index (GFI; 0.926), the comparative ﬁt index (CFI; 0.986), the normed ﬁt index (NFI; 0.965), the non-normed ﬁt index (NNFI; 0.983), and the standardized root mean square residual (SRMR; 0.0621). All ﬁt indices reﬂect a good model ﬁt with the exception of the GFI and the SRMR, which are both still within the range of commonly accepted values. A valid conclusion seems to be that our model ﬁts the empirical data reasonably well (Browne and Cudeck, 1993; SchermellehEngel et al., 2003). To assess whether role stress effectively mediates the relationship between the application of the controllability principle and managerial performance, we ﬁrst tested for a direct relationship between the application of the controllability principle and managerial performance (Baron and Kenny, 1986). We ﬁnd this relationship to be positive and signiﬁcant ( = 0.10, t = 2.04, p < 0.05). Next, we analyzed how this relationship was altered by the two potential mediators. The relationships between the application of the principle and (a) role ambiguity ( = −0.22, t = −4.03, p < 0.01) and (b) role conﬂict ( = −0.41, t = −6.33, p < 0.01) are both negative and signiﬁcant, conforming to our expectations and supporting H1 and H2 (Table 3). As predicted by H3, we ﬁnd the relationship between role ambiguity and managerial performance to be negative and signiﬁcant ( = −0.26, t = −4.30, p < 0.01). Role conﬂict, however, does not have a signiﬁcant impact on managerial performance ( = −0.02, t = −0.37, not sig.). The latter ﬁnding is consistent with much previous empirical accounting research on the relationship between role stress and man-
agerial performance (Burney and Widener, 2007; Fogarty et al., 2000; Rebele and Michaels, 1990; Viator, 2001). While we ﬁnd support for H3, we have to reject H4. Having speciﬁed the full mediation model, we ﬁnd that the initial direct relationship between the application of the controllability principle and managerial performance completely disappears ( = −0.02, t = −0.38, not sig.). We infer from this result that the effect of the application of the controllability principle on managerial performance is only an indirect one and operates through role ambiguity. Furthermore, trust in the superior—the control variable in our model—has signiﬁcant negative effects on role ambiguity ( = −0.49, t = −7.71, p < 0.01) and role conﬂict ( = −0.38, t = −6.11, p < 0.01). These ﬁndings are shown in Fig. 3 and enable us to substantiate full mediation. Partial mediation would mean that the initial direct relationship between the independent variable and the dependent variable signiﬁcantly decreases in strength but still exists (Baron and Kenny, 1986). Our conclusion is further supported by comparing the restricted model (i.e., a model without the direct effect of the principle’s application on managerial performance) with the model in Fig. 3. If the direct effect was important, restricting its path in the model would signiﬁcantly impair the ﬁt indices. In particular, one would expect a signiﬁcantly higher chi-square value (2 > 3.84; p < 0.05). However, restricting the direct effect leaves the ﬁt indices essentially unchanged. The chi-square value for the restricted model, which has one additional degree of freedom, is only marginally higher than the chi-square value for the model including the direct path (459.36; 2 = 0.26,
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Table 3 Results of the moderator analysis. Path
Application of the controllability principle → role ambiguity Application of the controllability principle → role conﬂict
Hierarchical level as moderator Lower/middle
= −0.32*** t = −4.69 = −0.39*** t = −5.34
= −0.08 t = −1.10 = −0.46*** t = −5.17
Chi-square difference (df = 1) 2 = 4.89 p < 0.05 2 = 0.88 not sig.
Asterisks ** and *** indicate the signiﬁcance of the p-value at <0.05 and 0.01 respectively (two tailed); n = 440.
not sig.). The chi-square value divided by the degrees of freedom value (2 /df) represents one of the central ﬁt indices for SEM models and improves from 1.622 (unrestricted model) to 1.617 (restricted model). Almost all other ﬁt indices remain the same as reported in Fig. 3. The NNFI (0.984) and the SRMR (0.0620) are the exception, as they slightly improve for the restricted model in comparison to the unrestricted model. When two competing models provide practically the same ﬁt indices, the more parsimonious model is regarded as the better-suited solution. Therefore, the comparison of the restricted model to the unrestricted model also supports full mediation. 5.2.2. Moderated model: Hypothesis 5 We used multi-group analysis in LISREL to test our hypothesis regarding potential moderating effects on the relationships between the application of the controllability principle and the role stress variables. Prior accounting studies have also adopted this procedure (Burney and Widener, 2007; Marginson and Bui, 2009; Viator, 2001). Multi-sample analysis requires the generation of separate SEMs by dividing the available sample into different subsamples and comparing the paths under investigation across the new subsamples. We proposed that the hierarchical level of managers moderates the effects of the application of the principle on role stress. Accordingly, we formed two subsamples to test this potential moderator. The ﬁrst group consisted of 155 top-level managers; the second group comprised 285 lower level and middle managers. We created a restricted model where the parameters for the paths of interest were restricted to be equal across the two subsamples and compared it to a model that allowed the same parameters to vary freely. Lastly, we compared the chi-square values for both models. A signiﬁcant improvement in the chi-square value for the non-restricted model indicates the existence of a moderator for the effect under investigation (Jaccard and Wan, 1996; Rigdon et al., 1998). Hypothesis 5 suggests that the managers’ hierarchical level moderates the relationships between the application of the controllability principle and (a) role ambiguity and (b) role conﬂict. Our ﬁndings show a strong moderating effect of the hierarchical level on the relationship between application of the principle and role ambiguity (2 = 4.89, p < 0.05). Among managers at high organizational levels, the effect of the principle’s application on role ambiguity is not signiﬁcant ( = −0.08, t = −1.10, not sig.). By con-
trast, the same effect is strong and signiﬁcant for managers in lower and mid-level hierarchical positions ( = −0.32, t = −4.69, p < 0.01). However, we found no support for the assumption that the hierarchical level of managers also moderates the relationship between application of the principle and role conﬂict. The paths are similar in strength across the two subsamples and the chi-square statistic does not reveal a signiﬁcant result (top-level managers: = −0.46, t = −5.17, p < 0.01; lower and middle-level managers: = −0.39, t = −5.34, p < 0.01; 2 = 0.88, not sig.). Apparently, application of the controllability principle is more important for lower level and middle managers because they beneﬁt the most from its application in terms of alleviating role ambiguity. Top-level managers, on the other hand, seem to expect the principle to be applied to a lesser extent, and the effect of the principle’s application on role ambiguity is insigniﬁcant among the subsample of toplevel managers. This ﬁnding is notable because only role ambiguity and not role conﬂict has an effect on managerial performance.7 5.2.3. Robustness checks Methodologically, scholars recommend measuring latent variables with several items to effectively take the measurement error into account (Baumgartner and Homburg, 1996; Churchill, 1979; Hinkin, 1995; Peter, 1979). On the basis of this recommendation, Hall (2008) measured managerial performance using more than one overall performance item. Adopting the Mahoney et al. instrument, he employed factor analytical techniques across all nine of the instrument’s items to identify a multi-item scale for the measurement of managerial performance. This approach enabled him to eliminate all unreliable items. To check the robustness of our results, we recalculated our model using Hall’s alternative way of measuring managerial performance. The respective SEM results did not diverge from our initial ﬁndings. The relationships reported above seem to be conﬁrmed with only marginal differences in the strength of some paths.
7 As mentioned earlier in this paper, role theory is not entirely clear about whether trust directly affects role stress perceptions or moderates the relationship between stress stimuli and role stress. In an attempt to be as thorough as possible, we tested whether trust interacts in the relationship between application of the controllability principle and role ambiguity and role conﬂict. We could not ﬁnd any signiﬁcant support for potential moderating effects of the trust variable.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Moreover, although role ambiguity and role conﬂict are commonly seen as two dimensions of job-related stress, some earlier studies in management accounting relied on a composite measure of these two dimensions (e.g., Hopwood, 1972). To establish comparability of our results with these previous studies, we performed another analysis that combined the two uni-dimensional role constructs into one composite measure. The results are practically identical and consistent with prior evidence. The application of the controllability principle has a signiﬁcant negative effect on the stress variable ( = −0.39, t = −6.29, p < 0.01), which in turn signiﬁcantly and negatively inﬂuences managerial performance ( = −0.25, t = −3.75, p < 0.01). Similar to our ﬁrst analysis, we found no signiﬁcant direct effect of the principle’s application on managerial performance once the model takes the mediating variable into account ( = −0.05, t = −0.89; not sig.). Moreover, most of the ﬁt values remain at acceptable levels and enable us to judge this model as acceptable. However, the fact that the two interrelated but distinct unidimensional constructs are merged together is reﬂected in a signiﬁcant worsening of all ﬁt indices (chi-square test statistic (875.28), the chi-square divided by the degrees of freedom (3.06), RMSEA (0.0685), GFI (0.867), CFI (0.965), NFI (0.945), NNFI (0.960) and SRMR (0.0731)). Consequently, we conclude that the main model tested above is superior to the one with a composite variable for role stress. 6. Discussion and conclusion Despite the great interest of academics and practitioners in the controllability principle, few empirical studies have quantitatively examined the consequences of its application in formal performance evaluations. Most of these studies have not gone beyond descriptive analyses or actually examined the consequences of not applying the principle at the level of individual managers. Our paper addresses this oversight. Our analyses show that the effect of applying the controllability principle on managerial performance is not direct but indirect in nature. Managers’ cognitive perceptions of role stress fully mediate this relationship. More speciﬁcally, role ambiguity mediates the relationship between application of the principle and managerial performance. When managers perceive the controllability principle as not being applied or only barely applied, they ﬁnd less clarity in the role expectations imposed on them. Ultimately, this ambiguity lowers their job performance. From a theoretical perspective, our study complements previous empirical work that found the effects of control systems on individual outcomes are not necessarily direct. Instead, individual cognition, and role perceptions in particular, often partly mediate the effects (Burney and Widener, 2007; Hall, 2008; Maas and Matˇejka, 2009; Shields et al., 2000). Our moderator analysis reveals that the dysfunctional effect of non-application of the principle on role ambiguity and indirectly on managerial performance is insigniﬁcant for top-level managers. Top-level managers seem to be better capable of coping effectively with the uncontrollable factors than lower and middle-level man-
agers. The literature provides some indication as to why this is the case. Managers’ self-image, motivation, and especially attitude toward risk vary with hierarchal level (Kahn et al., 1964; Sitkin and Pablo, 1992). Moreover, numerous studies emphasize that top-level executives are constantly confronted with high uncertainty (Cannella et al., 2008; Carpenter and Fredrickson, 2001), implying that they are used to dealing and expect to be dealing with uncontrollable factors. Consequently, the negative effects associated with non-application of the controllability principle do not hold universally, but rather depend on the experience and personality type of the individual managers. On the basis of our empirical ﬁndings, we can give a more differentiated answer to the key question of whether companies should always fully apply the controllability principle. First, the question should not be answered independently from the individual manager’s position in the corporate hierarchy. A plausible assumption is that including uncontrollable performance measures in the evaluation of top-level managers is particularly appropriate. One advantage of non-application of the controllability principle is that it allows for directing top-level managers’ attention toward critical performance areas which are not fully controllable by them but which are important to the organization. This advantage is particularly notable given the pronounced impact of the decisions and actions of toplevel managers on corporate performance (Hambrick and Mason, 1984). Their actions and decisions generally affect corporate welfare much more than actions and decisions of lower level managers. Our results suggest that, for top-level managers, companies can disregard the controllability principle more easily. Apparently, the beneﬁts from such accountability arrangements come at a lower cost than generally assumed in terms of the resultant role stress among managers. Second, important practical implications arise from insight into the cognitive mechanisms that mediate how non-application of the controllability principle affects managerial performance, particularly at lower managerial levels. These mechanisms provide the basis for understanding the potential levers for mitigating dysfunctional managerial behavior (here, lower managerial performance) when companies are not able to fully apply the principle or have a rationale for not doing so. As outlined at the beginning of this paper, not only practical reasons, but also organizational interests often call for an enlargement of managers’ accountabilities and for an erosion of the controllability principle. A notable result, therefore, is that non-application of the principle does not lead to decreased managerial performance unless role stress occurs. The results of our moderator analysis support the view that managers react differently to stress stimuli. They are also consistent with the claim of stress management literature that individuals generally differ in how well they are able to deal with stress (Cooper et al., 2001). As a consequence, companies might want to (a) select and (b) prepare managers more adequately for jobs that expose them to uncontrollable factors. Such measures are referred to as personnel controls (Merchant and Van der Stede, 2007, p. 84), and as they represent important elements of effective
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
management control systems, researchers from the ﬁeld of organizational behavior are surprised to ﬁnd that many companies ignore them (Cooper et al., 2001, p. 187). A selection strategy requires a job analysis in combination with an assessment of each manager’s characteristics and skills, particularly the ability to deal with stress, as these form the basis for selecting the right manager for the right job (Spector, 2008). Speciﬁc personality traits associated with stress resistance could help guide selection and placement decisions. Training represents another suitable personnel control to mitigate resulting dysfunctional effects from nonapplication of the controllability principle. Sessions helping managers to deal with uncontrollable factors effectively could complement executive education programs, and training might be particularly useful when employees are promoted to a position that has them take over general management responsibilities. Stress researchers have labeled such ex ante attempts at modifying individuals’ responses to stress as secondary interventions (Cooper et al., 2001). They are important when primary interventions—that is, changes in the working conditions that cause stress—are not possible. Organizations might also introduce stress management programs to buffer managers against levels of stress that inevitably come with relaxation of the controllability principle. So-called tertiary interventions focus on the “treatment” of problems and consist of measures to mitigate stress symptoms after their occurrence. Counseling and employee assistance programs are examples of those interventions (Cooper et al., 2001). Like the results of other studies, our ﬁndings are not without limitations. First, the study relies on survey data, which suffer from restrictions associated with the cross-sectional survey method. Second, we used a newly developed construct for measuring the application of the controllability principle. Even though we drew on existing management accounting literature and followed state-of-the-art guidelines in developing the construct, we recommend that future studies establish the construct’s usefulness and validity in a broader context. Third, to assess the application of the controllability principle in management control systems, our study investigates managers’ subjective perceptions of the controllability of their performance measures. It does not address the speciﬁc management accounting practices such as cost allocation and transfer pricing methods that lead to forming a certain perception. Research that relates the “objective” reality of management accounting practices to perceptions of managers represents another promising avenue for future research. Fourth, although theory clearly predicts a negative relationship between role conﬂict and managerial performance, we could not ﬁnd a signiﬁcant effect. Other studies that simultaneously analyzed the effects of role ambiguity and role conﬂict on managerial performance had similar results. Future research should re-investigate this important link by using alternative measures for managerial performance (see Marginson and Bui, 2009, for use of a different proxy) and role conﬂict (see Siegall, 2000, for a critique of traditional role ambiguity and role conﬂict constructs). Fifth, despite the use of role theory to guide the development of our structural equation model, we cannot
demonstrate cause-and-effect relations empirically. Future tests of the proposed hypotheses in an experimental setting would be beneﬁcial to further validate our ﬁndings. Lastly, we focused on two aspects: the dysfunctional effects of not applying the controllability principle (in terms of a decrease in managers’ individual performance) and the inﬂuence of one speciﬁc moderator variable—hierarchical level—on this relationship. Future research should investigate other moderating variables and, more importantly, the potential organizational advantages of non-application of the controllability principle. Notwithstanding the aforementioned limitations, we believe our study contributes to management accounting theory and practice. Our empirical analysis enhances the understanding of the effects of applying or not applying the controllability principle at different hierarchical levels. The results of this large-scale empirical study should make companies aware of the importance of dealing more effectively with behavioral effects that are caused by accountability arrangements such as application of the controllability principle, particularly at lower and middle managerial levels. Overall, understanding managers’ responses to application of the controllability principle is highly relevant for any organization, as “signiﬁcant problems can arise if uncontrollables are not dealt with properly” (Merchant, 1998, p. 583). Acknowledgments We are grateful for the helpful comments we received from the editor Michael Bromwich, two anonymous reviewers, Sebastian Becker, Pascal Langevin, Matthias Mahlendorf, Martin Messner, Felicitas Morhart, Robert Scapens, Karl Schuhmacher, and from the participants at the 4th Doctoral Summer School in Management Accounting, Siena, Italy, and the 7th ENROAC Conference, Dundee, Scotland. Additionally, we thank the Hanns-Seidel-Stiftung e.V. for ﬁnancing this research. Appendix A. Measurement instruments CONTRO item 1: CONTRO item 2: CONTRO item 3: PRE item 1:
PRE item 2:
PRE item 3: PRE item 4: SEN item 1: SEN item 2: SEN item 3: SEN item 4:
My performance measures include issues which I can only inﬂuence indirectly. (R) My performance measures include aspects that are beyond my sphere of inﬂuence. (R) My performance measures include aspects that are virtually not controllable for me. (R) There is no noise in my performance measurement as uncertain or uncontrollable factors do not have an impact on my performance measures. There is no distortion in my performance measurement that might be caused by uncontrollable factors. My performance measures are precise, i.e. the inﬂuence of uncontrollable factors is minimal. My performance measures are not blurred by factors that I cannot control. With my actions I can inﬂuence my performance measures. My effort has an impact on my performance measures. My performance measures do depend on my actions. With my effort I can inﬂuence the measures according to which my performance is evaluated.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
RA item 1: RA item 2: RA item 3a : RA item 4: RA item 5: RA item 6a : RC item 1: RC item 2: RC item 3: RC item 4: RC item 5: RC item 6: RC item 7a : RC item 8a : MP item 1:
MP item 2:
MP item 3:
MP item 4:
MP item 5:
MP item 6:
MP item 7:
MP item 8:
MP item 9: TRUST item 1:
TRUST item 2:
TRUST item 3:
TRUST item 4:
I feel certain about how much authority I have. (R) I have clear, planned goals and objectives for my jobs. (R) I know that I have divided my time properly. (R) I know what my responsibilities are. (R) I know exactly what is expected of me. (R) I receive a clear explanation of what has to be done. (R) I have to do things that should be done differently. I receive assignments without the manpower to complete them. I have to buck a rule or policy in order to carry out an assignment. I receive incompatible requests from two or more people. I do things that are apt to be accepted by one person and not accepted by others. I receive assignments without adequate resources and materials to execute them. I work on unnecessary things. I work with two or more groups who operate quite differently. Planning: Determining goals, policies and courses of action (e.g. work scheduling, budgeting, setting up procedures, programming) Investigating: Collecting and preparing information for records, reports and accounts (e.g. measuring output, inventorying, job analysis) Coordinating: Exchanging information with people in the organization other than your subordinates in order to relate and adjust procedures, policies and programs; advising and liaison with other personnel in the organization Evaluating: Assessment and appraisal of proposals or of reported/observed performance (e.g. employee appraisals, judging output records, judging ﬁnancial reports, approve requests, product inspection) Supervising: Directing, leading and developing your subordinates (e.g. counseling, explaining work rules to subordinates, assigning work, handling complaints) Stafﬁng: Maintaining the work force of your responsibility area (e.g. recruiting, interviewing and selecting new employees, promoting, transferring employees) Negotiating: Purchasing, selling or contracting for products or services (e.g. contracting suppliers, selling to dealers or customers, collective bargaining) Representing: Advancing the general interests of my organization through speeches, consultations, or contacts with individuals or groups outside the company Overall, how do you rate your (individual) performance? My superior takes advantage of opportunities that come up to further my interests by his actions and decisions. I feel free to discuss with my superior the problems and difﬁculties I have in my job without jeopardizing my position or having it “held against” me later on. I feel conﬁdent that my superior keeps me fully and frankly informed about things that might concern me. If my superior makes a decision which seems to be against my interests, I still have trust that my superior’s decision is justiﬁed for other considerations.
In my company, I belong to the following management level: Lower level management, or middle management, or top management.
CONTRO = application of the controllability principle (items on secondorder level); PRE = precision of performance measures (items on ﬁrst-order level); SEN = sensitivity of performance measures (items on ﬁrst-order level); RA = role ambiguity; RC = role conﬂict; MP = managerial performance; TRUST = trust in superior; LEVEL = hierarchical level. Apart from the LEVEL-item, all questionnaire items were assessed on a 7point-Likert scale. Items marked with (R) were reverse-coded in the questionnaire. a Items eliminated due to results of factor analyses.
References Antle, R., Demski, J.S., 1988. The controllability principle in responsibility accounting. Account. Rev. 63, 700–718. Aquino, K., 2000. Structural and individual determinants of workplace victimization: the effects of hierarchical status and conﬂict management style. J. Manage. 26, 171–193. Armstrong, J.S., Overton, T.S., 1977. Estimating nonresponse bias in mail surveys. J. Market. Res. 14, 396–402. Atkinson, A.A., Kaplan, R.S., Matsumura, E.M., Young, S.M., 2007. Management Accounting, 5th ed. Prentice Hall, Upper Saddle River. Baiman, S., Demski, J.S., 1980. Economically optimal performance evaluation and control systems. J. Account. Res. 18, 184–220. Baiman, S., Noel, J., 1985. Noncontrollable costs and responsibility accounting. J. Account. Res. 23, 486–501. Banker, R.D., Datar, S.M., 1989. Sensitivity, precision, and linear aggregation of signals for performance evaluation. J. Account. Res. 27, 21–39. Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 51, 1173–1182. Baumgartner, H., Homburg, C., 1996. Applications of structural equation modeling in marketing and consumer research: a review. Int. J. Res. Market. 13, 139–161. Berger, C.J., Cummings, L.L., 1979. Organizational structure, attitudes and behaviors. In: Cummings, L.L., Staw, B.M. (Eds.), Research in Organizational Behavior. JAI Press Inc., Greenwich, pp. 169–208. Bhimani, A., Horngren, C.T., Datar, S.M., Foster, G., 2008. Management and Cost Accounting, 4th ed. Prentice Hall, Harlow. Bisbe, J., Batista-Foguet, J.-M., Chenhall, R.H., 2007. Deﬁning management accounting constructs: a methodological note on the risks of conceptual misspeciﬁcation. Account. Org. Soc. 32, 789–820. Bollen, K., 1989. Structural Equations with Latent Variables. John Wiley & Sons, Inc., New York. Bollen, K., Lennox, R., 1991. Conventional wisdom on measurement: a structural equation perspective. Psychol. Bull. 110, 305–314. Bourguignon, A., Chiapello, E., 2005. The role of criticism in the dynamics of performance evaluation systems. Crit. Perspect. Account. 16, 665–700. Bouwens, J., Van Lent, L., 2006. Performance measure properties and the effect of incentive contracts. J. Manage. Account. Res. 18, 55–75. Brion, J.M., 1989. Organizational Leadership of Human Resources: The Knowledge and the Skills. JAI Press Inc., Greenwich. Brislin, R.W., 1980. Translation and content analysis of oral and written material. In: Triandis, H.C., Berry, J.W. (Eds.), Handbook of Crosscultural Psychology. Allyn & Bacon, Boston, pp. 398–444. Browne, M.W., Cudeck, R., 1993. Alternative ways of assessing model ﬁt. In: Bollen, K., Long, J.S. (Eds.), Testing Structural Equation Models. Sage Publications, Inc., Newbury Park, pp. 136–162. Brownell, P., Dunk, A.S., 1991. Task uncertainty and its interaction with budgetary participation and budget emphasis: some methodological issues and empirical investigation. Account. Org. Soc. 16, 693–703. Brownell, P., McInnes, M., 1986. Budgetary participation, motivation, and managerial performance. Account. Rev. 61, 587–600. Budding, G.T., 2004. Accountability, environmental uncertainty and government performance: evidence from Dutch municipalities. Manage. Account. Res. 15, 285–304. Burney, L.L., Henle, C.A., Widener, S.K., 2009. A path model examining the relations among strategic performance measurement system characteristics, organizational justice, and extra- and in-role performance. Account. Org. Soc. 34, 305–321. Burney, L.L., Widener, S.K., 2007. Strategic performance measurement systems, job-relevant information, and managerial behavioral responses—role stress and performance. Behav. Res. Acc. 19, 43–69.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159 Bushman, R.M., Indjejikian, R.J., Smith, A., 1995. Aggregate performance measures in business unit manager compensation: the role of intraﬁrm interdependencies. J. Account. Res. 33, 101–128. Cannella, A.A., Park, J.-H., Lee, H.-U., 2008. Top management team functional background diversity and ﬁrm performance: examining the roles of team member colocation and environmental uncertainty. Acad. Manage. J. 51, 768–784. Carpenter, M.A., Fredrickson, J.W., 2001. Top management teams, global strategic posture, and the moderating role of uncertainty. Acad. Manage. J. 44, 533–545. Chenhall, R.H., Brownell, P., 1988. The effect of participative budgeting on job satisfaction and performance: role ambiguity as an intervening variable. Account. Org. Soc. 13, 225–233. Chong, V.K., Chong, K.M., 2002. Budget goal commitment and informational effects of budget participation on performance: a structural equation modeling approach. Behav. Res. Acc. 14, 65–84. Choudhury, N., 1986. Responsibility accounting and controllability. Account. Bus. Res. 16, 189–198. Churchill, G.A., 1979. A paradigm for developing better measures of marketing constructs. J. Market. Res. 16, 64–73. Cohen, S.I., Loeb, M.P., Stark, A.W., 1992. Separating controllable performance from non-controllable performance: the case of optimal procurement contracting. Manage. Account. Res. 3, 291–306. Collins, F., 1978. The interaction of budget characteristics and personality variables with budgetary response attitudes. Account. Rev. 53, 324–335. Cooper, C.L., Dewe, P., O’Driscoll, M.P., 2001. Organizational Stress: A Review and Critique of Theory, Research and Applications. Sage Publications, Inc., Thousand Oaks. Dalton, G.W., 1971. Motivation and control in organizations. In: Dalton, G.W., Lawrence, P.R. (Eds.), Motivation and Control in Organizations. Richard D. Irwin, Inc., Homewood, pp. 1–35. Davis, S., Kohlmeyer, J.M., 2005. The impact of employee rank on the relationship between attitudes, motivation and performance. In: Epstein, M.J., Lee, J.Y. (Eds.), Advances in Management Accounting, vol. 14. JAI Press Inc., Amsterdam, pp. 233–252. Demski, J.S., 1976. Uncertainty and evaluation based on controllable performance. J. Account. Res. 14, 230–245. Demski, J.S., 1994. Managerial Uses of Accounting Information. Kluwer Academic Publishers, Norwell. Dent, J.F., 1987. Tensions in the design of formal control systems: a ﬁeld study in a computer company. In: Bruns, W.J., Kaplan, R.S. (Eds.), Accounting and Management: Field Study Perspectives. Harvard Business School Press, Boston, pp. 119–145. DeVellis, R.F., 1991. Scale Development: Theory and Applications. Sage Publications, Inc., Newbury Park. Diamantopoulos, A., Winklhofer, H.M., 2001. Index construction with formative indicators: an alternative to scale development. J. Market. Res. 38, 269–277. Dillman, D.A., 2007. Mail and Internet Surveys: The Tailored Design Method, 2nd ed. John Wiley & Sons, Inc., Hoboken. Drury, C., El-Shishini, H., 2005. Divisional Performance Measurement: An Examination of the Potential Explanatory Factors. The Chartered Institute of Management Accountants, London. Dubinsky, A.J., Mattson, B.E., 1979. Consequences of role conﬂict and ambiguity experienced by retail salespeople. J. Retail. 55, 70–86. Dunk, A.S., 1992. The effects of managerial level on the relationship between budgetary participation and job satisfaction. Br. Account. Rev. 24, 207–218. Dunk, A.S., 1993. The effects of job-related tension on managerial performance in participative budgetary settings. Account. Org. Soc. 18, 575–585. Eccles, R.G., 1986. The Transfer Pricing Problem: A Theory for Practice. Lexington Books, Lexington. Edwards, J.R., Bagozzi, R.P., 2000. On the nature and direction of relationships between constructs and measures. Psychol. Methods 5, 155–174. Feltham, G.A., Xie, J., 1994. Performance measure congruity and diversity in multi-task principal/agent relations. Account. Rev. 69, 429–453. Fisher, C.D., Gitelson, R., 1983. A meta-analysis of the correlates of role conﬂict and ambiguity. J. Appl. Psychol. 68, 320–333. Fisher, R.T., 2001. Role stress, the type A behavior pattern, and external auditor job satisfaction and performance. Behav. Res. Acc. 13, 143–170. Fogarty, T.J., Singh, J., Rhoads, G.K., Moore, R.K., 2000. Antecedents and consequences of burnout in accounting: beyond the role stress model. Behav. Res. Acc. 12, 31–67.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with unobservable variables and measurement error. J. Market. Res. 18, 39–50. Frederickson, J.R., 1992. Relative performance information: the effects of common uncertainty and contract type on agent effort. Account. Rev. 67, 647–669. Fremgen, J.M., Liao, S.S., 1981. The Allocation of Corporate Indirect Costs. National Association of Accountants, New York. Frow, N., Marginson, D., Ogden, S., 2005. Encouraging strategic behaviour while maintaining management control: multi-functional project teams, budgets, and the negotiation of shared accountabilities in contemporary enterprises. Manage. Account. Res. 16, 269–292. Gibbs, M., Merchant, K.A., Van der Stede, W.A., Vargus, M.E., 2004. Determinants and effects of subjectivity in incentives. Account. Rev. 79, 409–436. Gilboa, S., Shirom, A., Fried, Y., Cooper, C.L., 2008. A meta-analysis of work demand stressors and job performance: examining main and moderating effects. Pers. Psychol. 61, 227–271. Giraud, F., Langevin, P., Mendoza, C., 2008. Justice as a rationale for the controllability principle: a study of managers’ opinions. Manage. Account. Res. 19, 32–44. Gregson, T., Wendell, J., Aono, J., 1994. Role ambiguity, role conﬂict, and perceived environmental uncertainty: are the scales measuring separate constructs for accountants? Behav. Res. Acc. 6, 144–159. Grifﬁn, M.A., Neal, A., Parker, S.K., 2007. A new model of work role performance: positive behavior in uncertain and interdependent contexts. Acad. Manage. J. 50, 327–347. Hair, J.F., Black, B., Babin, B., Anderson, R.E., Tatham, R.L., 2006. Multivariate Data Analysis, 6th ed. Prentice Hall, Upper Saddle River. Hall, M., 2008. The effect of comprehensive performance measurement systems on role clarity, psychological empowerment and managerial performance. Account. Org. Soc. 33, 141–163. Hambrick, D.C., Mason, P., 1984. Upper echelons: the organization as a reﬂection of its top managers. Acad. Manage. Rev. 9, 193–206. Hartmann, F.G.H., 2000. The appropriateness of RAPM: toward the further development of theory. Account. Org. Soc. 25, 451–482. Hartmann, F.G.H., Slapniˇcar, S., 2009. How formal performance evaluation affects trust between superior and subordinate managers. Account. Org. Soc. 34, 722–737. Henderson, R.I., 1980. Performance Appraisal: Theory to Practice. Reston Publishing Co., Reston. Heneman, H.C., 1974. Comparison of self and superior ratings of managerial performance. J. Appl. Psychol. 59, 638–642. Henri, J.-F., 2007. A quantitative assessment of the reporting of structural equation modeling information: the case of management accounting research. J. Account. Lit. 26, 76–115. Hinkin, T.R., 1995. A review of scale development practices in the study of organizations. J. Manage. 21, 967–988. Hirst, M.K., 1981. Accounting information and the evaluation of subordinate performance: a situational approach. Account. Rev. 56, 771–784. Hirst, M.K., 1983. The controllability of ﬁnancial outcomes. Abacus 19, 29–38. Holmström, B., 1979. Moral hazard and observability. Bell J. Econ. 10, 74–91. Holmström, B., 1982. Moral hazard in teams. Bell J. Econ. 13, 324–340. Homburg, C., Stebel, P., 2009. Determinants of contract terms for professional services. Manage. Account. Res. 20, 129–145. Hopwood, A.G., 1972. An empirical study of the role of accounting data in performance evaluation. J. Account. Res. 10, 156–182. House, R.J., Rizzo, J.R., 1972. Role conﬂict and ambiguity as critical variables in a model of organizational behavior. Organ. Behav. Hum. Perform. 7, 467–505. Huffman, C., Cain, L.B., 2000. Effects of considering uncontrollable factors in sales force performance evaluation. Psychol. Market. 17, 799–833. Huffman, C., Cain, L.B., 2001. Adjustments in performance measures: distributive and procedural justice effects on outcome satisfaction. Psychol. Market. 18, 593–615. Ilgen, D.R., Fisher, C.D., Taylor, M.S., 1979. Consequences of individual feedback on behavior in organizations. J. Appl. Psychol. 64, 349–371. Ilgen, D.R., Hollenbeck, J.R., 1991. The structure of work: job design and roles. In: Dunnette, M.D., Hough, L.M. (Eds.), Handbook of Industrial and Organizational Psychology, vol. 2. Consulting Psychologists Press, Palo Alto, pp. 165–207. Ittner, C.D., Larcker, D.F., 2002. Determinants of performance measure choices in worker incentive plans. J. Lab. Econ. 20, 58–90. Jaccard, J., Wan, C.K., 1996. LISREL Approaches to Interaction Effects in Multiple Regression. Sage Publications, Inc., Newbury Park.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159
Jackson, S.E., Schuler, R.S., 1985. A meta-analysis and conceptual critique of research on role ambiguity and role conﬂict in work. Organ. Behav. Hum. Decis. Process. 36, 16–78. Jamal, M., 1985. Relationship of job stress to job performance: a study of managers and blue-collar workers. Hum. Relat. 38, 409–424. Jarvis, C.B., MacKenzie, S.B., Podsakoff, P.M., 2003. A critical review of construct indicators and measurement model misspeciﬁcation in marketing and consumer research. J. Cons. Res. 30, 199–218. Jöreskog, K.G., Sörbom, D., 2001. LISREL 8: User’s Reference Guide, 2nd ed. Scientiﬁc Software International, Lincolnwood. Kahn, R.L., 1974. Conﬂict, ambiguity and overload: three elements in job stress. In: McLean, A.A. (Ed.), Occupational Stress. Charles C. Thomas Publisher, Springﬁeld, pp. 47–61. Kahn, R.L., Quinn, R.P., 1970. Role stress: a framework for analysis. In: McLean, A.A. (Ed.), Mental Health and Work Organizations. Rand McNally & Co., Chicago, pp. 50–115. Kahn, R.L., Wolfe, D.M., Quinn, R.P., Snoeck, J.D., 1966. Adjustment to role conﬂict and ambiguity in organizations. In: Biddle, B.J., Thomas, E.J. (Eds.), Role Theory: Concepts and Research. John Wiley & Sons, Inc., New York, pp. 277–281. Kahn, R.L., Wolfe, D.M., Quinn, R.P., Snoeck, J.D., Rosenthal, R.A., 1964. Organizational Stress: Studies in Role Conﬂict and Ambiguity. John Wiley & Sons, Inc., New York. Katz, D., Kahn, R.L., 1966. The Social Psychology of Organizations. John Wiley & Sons, Inc., New York. Katz, D., Kahn, R.L., 1978. The Social Psychology of Organizations, 2nd ed. John Wiley & Sons, Inc., New York. Kren, R.L., 1992. Budgetary participation and managerial performance: the impact of information and environmental volatility. Account. Rev. 67, 511–526. Lau, C.M., Low, L.C., Eggleton, I.R.C., 1995. The impact of reliance on accounting performance measures on job-related tension and managerial performance: additional evidence. Account. Org. Soc. 20, 359–381. Lau, C.M., Tan, S.L.C., 2006. The effects of procedural fairness and interpersonal trust on job tension in budgeting. Manage. Account. Res. 17, 171–186. Lindell, M.K., Brandt, C.J., 2000. Climate quality and climate consensus as mediators of the relationship between organizational antecedents and outcomes. J. Appl. Psychol. 85, 331–348. Lindell, M.K., Whitney, D.J., 2001. Accounting for common method variance in cross-sectional designs. J. Appl. Psychol. 86, 114–121. Luckett, P.F., Eggleton, I.R.C., 1991. Feedback and management accounting: a review of research into behavioral consequences. Account. Org. Soc. 16, 371–394. Maas, V.S., Matˇejka, M., 2009. Balancing the dual responsibilities of business unit controllers: ﬁeld and survey evidence. Account. Rev. 84, 1233–1253. Machin, J.L.J., 1979. A contingent methodology for management control. J. Manage. Stud. 16, 1–29. Maher, M.W., 1987. The use of relative performance evaluation in organizations. In: Bruns, W.J., Kaplan, R.S. (Eds.), Accounting and Management: Field Study Perspectives. Harvard Business School Press, Boston, pp. 295–315. Mahoney, T.A., Jerdee, T.H., Carroll, S.J., 1963. Development of Managerial Performance: A Research Approach. South-Western Publishing, Cincinnati. Mahoney, T.A., Jerdee, T.H., Carroll, S.J., 1965. The job(s) of management. Ind. Relat. 4, 97–110. Manzoni, J.-F., 2002. Management control: toward a new paradigm? In: Epstein, M.J., Manzoni, J.-F. (Eds.), Performance Measurement and Management Control: A Compendium of Research (Studies in Managerial and Financial Accounting). JAI Press Inc., Oxford, pp. 15–46. Marginson, D., Bui, B., 2009. Examining the human cost of multiple role expectations. Behav. Res. Acc. 21, 59–81. Marginson, D., McAulay, L., 2001. Determinants of executive observance of the controllability principle: personal and social factors. In: 5th International Management Control Systems Research Conference , London, pp. 1–36. Marginson, D., Ogden, S., 2005a. Coping with ambiguity through the budget: the positive effects of budgetary targets on managers’ budgeting behaviours. Account. Org. Soc. 30, 435–456. Marginson, D., Ogden, S., 2005b. Managers, budgets and organisational change: unbundling some of the paradoxes. J. Account. Org. Change 1, 45–61. McNally, G.M., 1980. Responsibility accounting and organisational control. J. Bus. Fin. Account. 7, 165–181. Merchant, K.A., 1987. How and why ﬁrms disregard the controllability principle. In: Bruns, W.J., Kaplan, R.S. (Eds.), Accounting and Man-
agement: Field Study Perspectives. Harvard Business School Press, Boston, pp. 316–338. Merchant, K.A., 1989. Rewarding Results: Motivating Proﬁt Center Managers. Harvard Business School Press, Boston. Merchant, K.A., 1998. Modern Management Control Systems: Text and Cases. Prentice Hall, Englewood Cliffs. Merchant, K.A., Manzoni, J.-F., 1989. The achievability of budget targets in proﬁt centers: a ﬁeld study. Account. Rev. 64, 539–558. Merchant, K.A., Otley, D.T., 2006. A review of the literature on control and accountability. In: Chapman, C.S., Hopwood, A.G., Shields, M.D. (Eds.), Handbook of Management Accounting Research, vol. 2. Elsevier Ltd., Amsterdam, pp. 785–804. Merchant, K.A., Van der Stede, W.A., 2007. Management Control Systems: Performance Measurement, Evaluation and Incentives, 2nd ed. Prentice Hall, Harlow. Milgrom, P.R., Roberts, D.J., 1992. Economics, Organization and Management. Prentice Hall, Englewood Cliffs. Modell, S., 1997. A progress-oriented contingency approach to responsibility in professional services. Int. J. Serv. Ind. Manage. 8, 306–335. Modell, S., Lee, A., 2001. Decentralization and reliance on the controllability principle in the public sector. Finan. Account. Manage. 17, 191–218. Moers, F., 2006. Performance measure properties and delegation. Account. Rev. 81, 897–924. Moon, P., Fitzgerald, L., 1996. Delivering the goods at TNT: the role of the performance measurement system. Manage. Account. Res. 7, 431–457. Murphy, K.R., Cleveland, J.N., 1995. Understanding Performance Appraisal: Social, Organizational, and Goal-based Perspectives. Sage Publications Inc., Thousand Oaks. Otley, D.T., 1990. Issues in accountability and control: some observations from a study of colliery accountability in the British Coal Association. Manage. Account. Res. 1, 101–123. Otley, D.T., Pollanen, R.M., 2000. Budgetary criteria in performance evaluation: a critical appraisal using new evidence. Account. Org. Soc. 25, 483–496. Parker, R.J., Kyj, L., 2006. Vertical information sharing in the budgeting process. Account. Org. Soc. 31, 27–45. Peter, J.P., 1979. Reliability: a review of psychometric basics and recent marketing practices. J. Market. Res. 16, 6–17. Petter, S., Straub, D., Rai, A., 2007. Specifying formative constructs in information systems research. MIS Q. 31, 623–656. Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J. Appl. Psychol. 88, 879–903. Read, W.H., 1962. Upward communication in industrial hierarchies. Hum. Relat. 15, 3–15. Rebele, J.E., Michaels, R.E., 1990. Independent auditors’ role stress: antecedent, outcome, and moderating variables. Behav. Res. Acc. 2, 124–153. Rigdon, E.E., Schumacker, R.E., Wothke, W., 1998. A comparative review of interaction and nonlinear modeling. In: Schumacker, R.E., Marcoulides, G.A. (Eds.), Interaction and Nonlinear Effects in Structural Equation Modeling. Lawrence Erlbaum Associates, Inc., Mahwah, pp. 1–16. Rizzo, J.R., House, R.J., Lirtzman, S.I., 1970. Role conﬂict and ambiguity in complex organizations. Adm. Sci. Q. 15, 150–163. Robbins, S.P., 2005. Organizational Behavior, 11th ed. Prentice Hall, Upper Saddle River. Ronen, J., Livingstone, J.L., 1975. An expectancy theory approach to the motivational impacts of budgets. Account. Rev. 50, 671–685. Saunders, T., Driskell, J.E., Johnston, J.H., Salas, E., 1996. The effect of stress inoculation training on anxiety and performance. J. Occup. Health Psychol. 1, 170–186. Schermelleh-Engel, K., Moosbrugger, H., Müller, H., 2003. Evaluating the ﬁt of structural equation models: tests of signiﬁcance and descriptive goodness-of-ﬁt measures. Methods Psychol. Res. Online 8, 23–74. Schminke, M., Cropanzano, R., Rupp, D.E., 2002. Organization structure and fairness perceptions: the moderating effects of organizational level. Organ. Behav. Hum. Decis. Process. 89, 881–905. Shields, M.D., Deng, F.J., Kato, Y., 2000. The design and effects of control systems: tests of direct- and indirect-effects models. Account. Org. Soc. 25, 185–202. Shields, M.D., Waller, W.S., 1988. A behavioral study of accounting variables in performance-incentive contracts. Account. Org. Soc. 13, 581–594. Siegall, M., 2000. Putting the stress back into role stress: improving the measurement of role conﬂict and role ambiguity. J. Manage. Psychol. 15, 427–435.
M. Burkert et al. / Management Accounting Research 22 (2011) 143–159 Simons, R., 2005. Levers of Organization Design: How Managers Use Accountability Systems for Greater Performance and Commitment. Harvard Business School Press, Boston. Simons, R., 2007. Revisiting the controllability principle in the twentyﬁrst century. In: American Accounting Association Annual Meeting , Chicago, pp. 1–34. Sitkin, S.B., Pablo, A.L., 1992. Reconceptualizing the determinants of risk behavior. Acad. Manage. Rev. 17, 9–38. Smith, D., Langﬁeld-Smith, K., 2004. Structural equation modeling in management accounting research: critical analysis and opportunities. J. Account. Lit. 23, 49–86. Spector, P.E., 2008. Industrial and Organizational Psychology: Research and Practice, 5th ed. John Wiley & Sons, Inc., Hoboken. Stahl, G.K., Caligiuri, P., 2005. The effectiveness of expatriate coping strategies: the moderating role of cultural distance, position level, and time on the international assignment. J. Appl. Psychol. 90, 603–615. Szilagyi, A.D., Sims, H.P., Keller, R.T., 1976. Role dynamics, locus of control, and employee attitudes and behavior. Acad. Manage. J. 19, 259–276. Tubre, T.C., Collins, J.M., 2000. Jackson and Schuler (1985) revisited: a meta-analysis of the relationships between role ambiguity, role conﬂict, and job performance. J. Manage. 26, 155–169.
Ugras, Y.J., 1994. Factors affecting allocation of noncontrollable costs for performance evaluation use: a survey. In: Epstein, M.J., Lee, J.Y. (Eds.), Advances in Management Accounting, vol. 3. JAI Press Inc., Amsterdam, pp. 255–278. Van de Vijver, F., Hambleton, R.K., 1996. Translating tests: some practical guidelines. Eur. Psychol. 1, 89–99. Van Sell, M., Brief, A.P., Schuler, R.S., 1981. Role conﬂict and role ambiguity: integration of the literature and directions for future research. Hum. Relat. 34, 43–71. Vancil, R.F., 1979. Decentralization: Managerial Ambiguity by Design. Dow Jones-Irwin, Homewood. Verbeke, W., Bagozzi, R.P., 2002. A situational analysis on how salespeople experience and cope with shame and embarrassment. Psychol. Market. 19, 713–741. Viator, R.E., 2001. The association of formal and informal public accounting mentoring with role stress and related job outcomes. Account. Org. Soc. 26, 73–93. Waller, W.S., Chow, C.W., 1985. The self-selection and effort effects of standard-based employment contracts: a framework and some empirical evidence. Account. Rev. 60, 458–476. Widener, S.K., 2006. Associations between strategic resource importance and performance measure use: the impact on ﬁrm performance. Manage. Account. Res. 17, 433–457.