تجزیه و تحلیل تجربی اهرمهای چارچوب کنترل

The purpose of this paper is to use the levers of control framework to explore the antecedents of control systems – various facets of strategy that drive the use of controls; to explore the relations among control systems; and to explore the costs and benefits of control systems – costs in terms of consumption of a constrained resource (i.e., management attention) and benefits (i.e., learning). Using data from a survey of 122 Chief Financial Officers, this study tests a structural equation model that relates strategic risk and uncertainty to control systems (i.e., beliefs, boundary, diagnostic, and interactive control systems), which in turn are hypothesized to affect learning and attention, and ultimately firm performance. The evidence suggests that there are multiple inter-dependent and complementary relations among the control systems. I find that strategic risk and uncertainty drive both the importance and use of performance measures in diagnostic or interactive roles. Moreover, it appears that in certain strategic conditions information processing needs are such that firms use performance measures both interactively and diagnostically. Finally, I conclude that although there is a cost of control, there is a positive effect on firm performance.

The purpose of the management control system (MCS) is to provide information useful in decision-making, planning, and evaluation (Merchant & Otley, 2006). While the management accounting literature is replete with studies that investigate control systems, many focus on only one control such as the use of performance measures (Ittner & Larcker, 1998). However, it is well-recognized that the MCS is comprised of multiple control systems that work together (Otley, 1980). Simons (2000), in his levers of control (LOC) framework, posits that four control systems – beliefs (e.g., core values), boundary (e.g., behavioral constraints), diagnostic (e.g., monitoring), and interactive (e.g., forward-looking, management involvement) – work together to benefit a firm. The LOC framework asserts that strategic uncertainty and risk drive the choice and use of control systems, which in turn, impact the organization through organizational learning and the efficient use of management attention (Simons, 2000). The theoretical framework is illustrated in Fig. 1. The purpose of this paper is to use the LOC framework to investigate the antecedents of control systems (i.e., strategic uncertainty and risk); the associations among the control systems; and the costs and benefits of control systems (management attention, learning, and firm performance). Using data from a survey of 122 Chief Financial Officers (CFOs), I perform a three-stage analysis. First, I estimate a trimmed structural equation model (SEM) based on Fig. 1, which provides evidence on the holistic LOC framework. Second, I use coefficients from the trimmed SEM to provide evidence on three sets of hypotheses: (1) the relations among the control systems, (2) the relations between both strategic risk and uncertainty and each of the control systems, and (3) the relations between each of the control systems and outcomes (i.e., attention, learning, and performance). Finally, since the existence of a well-fitting SEM does not ensure that it is the only appropriate model (Kline, 1998), I generate six alternative models for comparison against the base model. This study makes several contributions to the literature. First, it is generally well-accepted that control systems are inter-dependent (Milgrom & Roberts, 1995); however, it is unclear whether they are complements or substitutes. This study finds that when firms emphasize the beliefs system, they also emphasize each of the three other control systems. In addition, the use of performance measures in the interactive system is associated with the use of performance measures in the diagnostic system and emphasis on the boundary system. The evidence suggests that the inter-dependencies are complementary. Thus, this study provides empirical evidence on the relations among the control systems in the LOC framework and contributes to a small but growing body of work that investigates relations among control systems (e.g., Anderson and Dekker, 2005 and Kennedy and Widener, 2006). Second, this study investigates both costs and benefits of control systems. An assumption in the literature is that firms implement controls only when the benefits received outweigh the costs. However, little evidence exists to support this assumption (e.g., for a review of performance measurement (PM) literature see Ittner & Larcker, 1998). A limitation is that this research often focuses on an aggregate measure of firm performance without delineating specific costs and benefits. A recent study extends the literature and provides evidence that a strategic PM system is associated with a specific benefit (i.e., decreased role stress) (Burney & Widener, 2007); however, the cost of controls is still largely ignored. I find that control systems are associated with both a benefit (organizational learning) and a cost (consumption of management attention); but, overall, have a positive effect on firm performance. Third, not withstanding a line of research that has investigated the alignment between strategy and a firm’s MCS (Ittner and Larcker, 1997 and Langfield-Smith, 1997), Langfield-Smith (1997) concludes that knowledge is limited since studies only investigate single facets of a multifaceted construct. Moreover, Chenhall (2003) argues that accounting studies may suffer from outdated strategy constructs. This has spurred studies to incorporate additional strategic facets such as strategic resources (see e.g., Henri, 2006) and competitive advantage (Widener, Shackell, & Demers, 2006). This study investigates two elements of strategy that Simons’ (2000) argues play a central role in the LOC theory – strategic uncertainties and strategic risk, and finds that both are associated with the use of control systems. Finally, business are competing with complex, rapidly changing, and knowledge-intensive business models driving the need to better understand the role of PM systems and how they can better meet managerial needs. A control system can function in different roles (i.e., either interactively or diagnostically). This study sheds insights on the role of the PM system and finds generally that internal (external) strategic factors are associated with diagnostic (interactive) controls. However, under certain circumstances, performance measures are used in both roles. These results add to a growing body of literature that investigate how the role of control systems differs (e.g., Abernethy and Brownell, 1999, Bisbe and Otley, 2004 and Henri, 2006). This study is organized as follows. Section “Theory development and hypotheses: control systems” provides an overview of the control systems that comprise the levers of control framework and develops hypotheses for the inter-dependencies among the control systems. Section “Theory development and hypotheses: drivers and outcomes of control systems” develops the hypotheses for the drivers (strategic risk and uncertainty) and outcomes (learning, attention, and performance) of the control systems. Section “Methods” discusses the research method and measurement of the variables. The analyses and results are presented in Section “Results”. Finally, conclusions, limitations, and extensions are discussed in Section “Conclusions”.

This study provides evidence on the LOC framework. It finds that two types of strategic elements – strategic uncertainties and strategic risk – drive the importance and role of control systems. It also documents that each of the diagnostic and beliefs systems facilitate the efficient use of management attention, while the interactive system consumes management attention (i.e., a “cost” of control). Organizational learning is enhanced by emphasis on the beliefs system as well as use of the diagnostic system. Both organizational learning and attention are positively associated with performance. Finally, it finds that the interactive system influences the diagnostic and boundary systems and the beliefs system influences each of the three other systems. Similar to most studies, there are limitations. This study relies on survey data from 122 respondents. Steps were taken to ensure the reliability of the data (i.e., random sample, pre-test of instrument, construct and content validity). All diagnostic tests show that there is no reason to expect bias (common method test, non-response bias); moreover, I also use an archival measure of performance to demonstrate the robustness of the results. However, measures may be noisy and caution should be taken when generalizing the results to other populations. In addition, this study relies on cross-sectional data. Although the relations in the path model are substantiated by underlying theory, cause-and-effect relations cannot be demonstrated empirically. In spite of the limitations, this study results in five implications for theory and practice. First, this study demonstrates that many of the controls in the LOC framework are inter-dependent and complementary. Specifically, the results show that the interactive system is inter-dependent with both the boundary system and the diagnostic use of performance measures, the latter of which is consistent with Henri (2006) who argues that dynamic tension results from the use of performance measures in dual roles. An important implication for organizations is that in order to realize the full benefits of the performance measurement system they must use them both diagnostically and interactively. The findings are also consistent with Chenhall and Morris (1995) who argue that structure is necessary for interactive type controls to be effective. In this study, the diagnostic system provides the structure that enables the interactive system to be effective since the diagnostic system is a mechanism by which the employees learn of the new strategy and consequently, the new goals and objectives with which to align behavior. The boundary system also provides structure through delineating the areas off-limit to employees. I also find that use of the beliefs system positively influences all other systems. This is consistent with Pearce and David (1987) who state that the beliefs system provides the foundation for the firm’s identity and value system. The positive relations between the beliefs system and each of the three remaining control systems positively affect organizational outcomes implying that the relations are complementary. The total standardized effect of the four control systems on performance are 0.514; however, if the control systems are isolated by removing the paths among them, the total standardized effect of the four control systems on performance decreases to 0.328 (the Chi-squared difference test is significant at p < 0.01). This result suggests that managers must consider all four control systems when designing their control system. It also provides empirical evidence that the control systems are complementary. This is consistent with Simons (2000) who argues that an effective control system, comprised of the four control levels working in harmony and balance, facilitates organizational performance. Second, strategy not only drives the importance of controls, but also the role of controls. I find that two types of strategic uncertainties (competitive uncertainty and operational uncertainty) are associated with the importance of control systems. Examining the standardized coefficients for strategic uncertainties shows that operational uncertainties have the largest effect on the diagnostic and beliefs systems, while competitive uncertainties drives interactive controls. This implies that the interactive control system is used to scan the external environment while the other systems are focused more on the internal environment. I also find that the firm uses the PM system both diagnostically and interactively to manage operational risk. These results are consistent with Galbraith (1973) who says that firms implement mechanisms to process information; as uncertainty increases the information deficit increases leading to increased reliance on mechanisms that facilitate the processing of information. The results suggest that there are relatively precise measures of operational risk and uncertainty since both are used diagnostically. However, it is likely that measures of competitive uncertainty are relatively less precise since competitive uncertainty is managed with an interactive system. Both the diagnostic and interactive systems are used to manage operational risk, which implies not only that the measures of operational risk are relatively precise, but that the information deficit associated with operational risk is such that firms need both systems to effectively manage it. The results also indicate that organizations may implement other types of control systems outside the scope of this study to effectively manage asset impairment and technological risk, which future research could investigate. Third, the interactive use of the PM system is not associated with organizational learning. There is a significant bi-variate correlation between organizational learning and use of the interactive system. In addition, if paths between the other systems and organizational learning are restricted to zero in the structural equation model then the path between the interactive use of the PM system and organizational learning is positive and significant. The empirical results in this paper suggest that it is the structured, formal process of the diagnostic system that brings to life the benefits of the interactive system. This finding illustrates the importance of studying multiple control systems. Studies that focus only on interactive controls may contend that organizational learning is enhanced; however, when controlling for other control systems (i.e., beliefs and diagnostic controls), the direct link between the interactive system and learning does not contain any additional explanatory power. Rather, the interactive system affects learning through the diagnostic system. Fourth, although there is a cost of control, overall, the four control components have a positive impact on performance, with a total standardized effect of 0.514. The results demonstrate that there is a cost associated with the interactive use of performance measures since it consumes management attention. However, the net effect of the four controls on attention is positive. Moreover, if the interactive system is removed from the control environment, the total standardized effect of the remaining control systems on performance falls to 0.438. Therefore, the benefits of the control systems outweigh the costs. Finally, this study shows that emphasis on control systems influences performance through their affect on learning and management attention. Sensitivity tests demonstrate that the direct relation between control systems and performance is weak; however, the effects become apparent when LEARN and ATTEN are included in the model. This is consistent with Luft and Shields (2003) who state that if the relation between two variables is weak then including appropriate intervening or mediating variables can help researchers detect effects. Thus this paper presents a more complete model of the relation between control systems and firm performance