Combinations of Strategy and Control: A Set-Theoretic Approach
Abstract
A significant amount of attention has been devoted to understanding the relationship between strategy and control. While much progress has been made a number of important questions have so far received little attention: (1) how do management controls combine to achieve effective control outcomes for different firm strategies? and (2) are there multiple, equally effective control combinations in a given strategic context? This study addresses these questions by drawing on the Miles and Snow (1978) typology to hypothesize the control combinations expected to be effective for different strategic contexts. Expectations are tested through a novel method termed fuzzy set qualitative comparative analysis (fsQCA) using data obtained from a cross sectional survey. The study contributes to the literature by providing evidence on how control attributes combine within and across strategic contexts and demonstrating that effective control can be achieved through multiple configurations. In doing so the study provides insight into the interdependence and relative importance of control attributes for achieving effective control outcomes in different strategic settings.
Keywords: Management control, strategy, configurations, qualitative comparative analysis
Introduction
The interface between strategy and management control (MC) is one of the most enduring concerns in the management accounting literature. Much of the research in this space builds upon strategic typologies, such as prospectors/defenders (Miles & Snow, 1978), entrepreneurial/conservative (Miller & Friesen, 1982), build/harvest (Gupta & Govindarajan, 1984), and differentiation/cost leadership (Porter, 1980). These typologies represent important developments as they allow researchers to empirically capture the complex patterns of action and distinctive competencies that constitute the strategy of an organization (Chapman, 1997; Dent, 1990). Although there are noted inconsistencies, the resulting body of research has been insightful, identifying not only how strategy influences the choice of certain MC, but also revealing how controls such as accounting might play a more active role in shaping the strategic direction of the firm (Langfield-Smith, 2008).
Despite this progress there remain fundamental questions that have received little attention. First, while the literature has been relatively successful in identifying the individual MC attributes associated with higher performance for different strategic positions, there is little understanding of how these attributes actually combine. This is somewhat surprising as many of the foundational theories of this research, such as Miles and Snow (1978), argue that effectiveness is maximized only through specific configurations of organizational attributes. The implicit assumption of much of the strategy-MC literature is that the optimal combination is an aggregation of the individual MC attributes observed to have incremental effects on performance. Yet, without empirical evidence it remains less than clear whether the MC attributes studied in prior literature are in fact all necessary, or what the relative importance of each attribute is, for achieving desired firm outcomes.
Second, the literature generally assumes that there is a single MC combination that leads to high performance for a given strategy, ignoring the possibility that there may be multiple and equally effective control arrangements available to an organization. The omission of choice in contingency models is identified by Dent (1990, p. 10) as a possible explanation for the unrewarding results of early MC research:
If organizations are only loosely coupled to their environments, there may be a range of viable responses to specific contingencies […] and even in exogenously constrained situations there may be degrees of freedom within which organizations operate.
This idea is captured by the term equifinality, which states that social systems are capable of achieving an outcome through different structures and processes even when facing similar contextual conditions (Gresov & Drazin, 1997; Katz & Kahn, 1978). Equifinality is a fundamental assumption of many structural-functionalist and systems theories of social organization, from which contingency theory has its origins (Ashmos & Huber, 1987). This is recognized by early theorists in the field such as Galbraith (1973), Miles and Snow (1978) and Thompson (1967). But as contingency researchers began to move away from firsthand accounts of organizations towards arms-length statistical analyses, the circumstances originally conceived as the conditions that limit or constrain organizational choices became the factors that essentially determine them (Galunic & Eisenhardt, 1994). Although prior research cites equifinality as a possible reason for insignificant or unexpected results (e.g. Gerdin, 2005), there has been little explicit attempt to investigate the potential for equally viable MC alternatives.
To explore these questions the study draws upon the strategic types of Miles and Snow (1978) and the strategy-MC literature to hypothesize and test the MC combinations associated with effective control outcomes. The initial expectation is that the most effective MC combinations will be those comprised of the control attributes that theory and prior research indicates are individually associated with firm outcomes conditional on strategic context. The theoretically derived combinations are then tested empirically using a novel method termed fuzzy set qualitative comparative analysis (fsQCA). This set-theoretic approach is uniquely suited to the research questions of this study as it explicitly examines how attributes combine to produce an outcome of interest (Ragin, 2008). Furthermore, the method allows for examination of the possible equifinality of multiple combinations, and unlike comparable techniques such as cluster analysis and profile deviation analysis, it provides insight into the interdependence and relative importance of attributes within a configuration for achieving an outcome (Fiss, 2007).
This study contributes to our limited knowledge on how MC attributes combine to achieve effective control in different strategic contexts in a number of ways. First, the study extends the contingency MC-strategy literature by demonstrating that although many of the hypothesized associations are supported, not all MC attributes need to be simultaneously present in an organization for effective control to be realized. It is shown that there are multiple alternate and functionally equivalent ways to combine different MC attributes in the same strategic context, providing empirical support for the notion of equifinality in control system design (Gerdin, 2005; Sandelin, 2008). Second, it is widely assumed that accounting and other control mechanisms are interdependent (Chenhall, 2003; Otley, 1980; Milgrom & Roberts, 1995) but there is little understanding of how relationships between MC attributes might be affected by firm context (Grabner & Moers, 2013). This study adds to the literature by showing how the interdependence of MC attributes for achieving effective control is conditional on the strategic context of the firm. Third, recent literature argues that firms pursuing mixed or joint strategies require control systems that encourage managers to balance, rather than tradeoff, competing priorities (Gibson & Birkinshaw, 2004), yet so far research in this area is limited to the choice and use of performance measures (Lillis & van Veen-Dirks, 2008; Dekker et al., 2013). By examining a wider range of MC attributes this study provides insight into how effective MC is achieved in this complex setting. Finally, the study adds to recent literature that demonstrates the usefulness of set-theoretic methods for MC research (Erkens & Van der Stede, 2013). In allowing examination of more complex combinations of control the method holds particular promise for gaining insights into unresolved puzzles and gaps in the literature. Specifically, this study speaks to concerns surrounding the effectiveness of tight accounting controls in the context of innovation and the relationship between formal and social controls (Dent, 1990; Langfield-Smith, 2008).
This study is organized as follows. The next section describes the configurational approach that underpins the analysis of MC combinations and the concept of equifinality. The study then proceeds to develop a set of hypotheses regarding effective MC combinations for different strategic contexts. Next the survey design, set-theoretic method and measurement of constructs are described, followed by the results of the analysis. The final section discusses the findings of the study and the limitations of the research.
Literature review and hypothesis development
Overview of management control combinations
The conceptual basis for understanding MC combinations stems from the configurational approach to organizational analysis. This perspective suggests that organizations are best understood as complex combinations of interconnected structures and processes (Fiss, 2007; Meyer et al., 1993). Organizational outcomes are assumed to be dependent not only on the incremental contributions of individual components, but on their overall arrangement. This logic underpins many of the foundational theories used to investigate strategy-MC linkages. For instance, although the term is not explicitly used, the Miles and Snow (1978) typology represents a configurational theory of organizational effectiveness (Doty & Glick, 1994; Gerdin & Greve, 2004). The central contention is that there are a limited number of arrangements that maximize fit and effectiveness; variations from these configurations result in lower performance.[1] Miles and Snow identify three ideal types, termed the defender, analyzer and prospector, each representing a consistent and equally effective combination of organizational attributes.
Despite the centrality of typological theories to strategy-MC research, there have been few attempts to study how MCs actually combine, or the effectiveness of different combinations, for particular strategic types. In a seminal investigation, Chenhall and Langfield-Smith (1998) use cluster analysis to investigate the performance effects of different configurations of strategy and managerial practices, including accounting. The study provides insight into the combinations of practices and techniques that are frequently used when pursuing different strategic priorities. One limitation of their approach is the inability to determine the relative importance of each mechanism, or whether all observed components are actually necessary to achieve high performance (Ittner & Larcker, 2001).
Conceptually this concerns the distinction between the core and periphery of a configuration. Configurational theorists distinguish core attributes in terms of the extent of interaction or connectedness. These tightly integrated components are surrounded by peripheral attributes that reinforce but are loosely coupled to the central core (Grandori & Furnari, 2008; Siggelkow, 2002). Recently Fiss (2011) has built on these insights to define core attributes as those “for which the evidence indicates a strong causal relationship with the outcome” and peripheral attributes as “those for which the evidence for a causal relationship with outcome is weaker” (p. 398). This emphasizes the relative importance of components within a configuration for achieving an outcome. While core attributes are necessary parts of a configuration to lead to an outcome, they may not be sufficient by themselves unless combined and reinforced with certain peripheral attributes. But as peripheral attributes are weakly connected, organizations are able to substitute or interchange these, resulting in multiple permutations that are potentially equally effective. This possibility is captured by the notion of equifinality.
Equifinality and management control
Equifinality, or functional equivalence, refers to the idea that a system can achieve the same final state or goal “from different initial conditions and by a variety of paths” (Katz & Kahn, 1978, p. 30; von Bertalanffy, 1968). In organizational analysis this has come to mean that the required performance of an organization can be achieved through multiple strategic and structural alternatives despite facing the same environmental imperatives (Gresov & Drazin, 1997). This is a central assumption of configuration theories, such as Miles and Snow (1978, p. 260), who recognize that the pattern of adaptive response selected by an organization is not wholly determined by exogenous forces:
It has been frequently noted that organizations adopt a variety of forms in response to apparently similar environmental demands [...] Thus, there is evidence to refute a “functional imperative” of organizational structure and behaviour.
While managers are noted to have strategic choice in deciding how their organization should be aligned to the external environment, Miles and Snow (1978) contend that for each strategic profile there is a single, optimal (ideal) structural solution, and variations from this arrangement will result in lower performance. The theory of Miles and Snow (1978) therefore posits equifinality at a population or strategic level – that organizations pursuing different strategies can be equally effective – but not at a structural level – only one structural combination will be optimally effective for each strategy. However, as most research has focused upon between-group differences in structural attributes and performance, there is little evidence to reveal the within-group effects of employing different combinations (Zahra & Pearce, 1990; for exceptions see Doty et al., 1993; Fiss, 2011).[2]
The understanding that there may be equally effective structures for a given organizational context is largely derived from Merton’s (1968) critique of the central postulates of comparative sociology. In particular Merton takes issue with “the postulate of indispensability” (1968, p. 87). Merton argues that researchers frequently treat structures as having a single function that are necessary for the survival and effective performance of the overall system.[3] This assumption underpins conventional contingency theory whereby “the independent variable [function] is a necessary and sufficient condition for the dependent variable [structure]” (Schreyögg, 1982, p. 75). In effect, structures are assumed to be equivalent to their functions as organizational contingencies dictate a single optimizing structure (Donaldson, 2001).[4] The central claim of Merton is that while a structure may fulfill a particular function, it is not synonymous with that function. Rather, social systems are permissive of a “range of possible variation” in structural components to effectively meet functional requirements (Merton, 1968, p. 106).
A challenge to empirically investigating equifinality is identifying a suitable basis for assessing the equivalence of structural alternatives. Effectiveness or fit in MC-strategy literature is typically evaluated in terms of financial performance (Langfield-Smith, 2008). While an obvious concern for most firms, the connection with MC is largely implied, with Chenhall (2003, p. 132) noting that “there is no compelling evidence to suggest that such links exist”. Instead it may be better to assess more proximate outcomes by focusing on the functions of MC, that is, the control problems they are intended to solve (Grabner & Moers, 2013). In this study the control problem relates to the alignment of organizational behavior to various strategic objectives. The next section develops hypotheses for the MC combinations expected to achieve effective control in different strategic contexts.
Hypotheses among strategy and MC attributes
The mechanisms available to top management to achieve effective control are potentially extensive (Malmi & Brown, 2008; Merchant & Van der Stede, 2012). The concern of this study is, however, with those MC attributes expected to vary in effectiveness within and between strategic orientations. In this respect certain control elements have received greater attention in strategy-MC literature, presumably because of their perceived importance for effective control for different strategic types (Chenhall, 2003; Dent, 1990; Langfield-Smith, 2008; Simons, 1987). Most of this research adopts a contingency approach to examine the association of individual MCs with firm outcomes. The general assumption is that MC attributes have an additive effect, such that an organization will maximize performance by combining individual attributes. The remainder of this section reviews the strategy-MC literature and identifies the structural, accounting, incentive and social control attributes found or theorized to be associated with improved firm outcomes for defender, prospector and analyzer strategic types.[5] The MC combinations hypothesized to lead to high MC effectiveness for each strategic type are then formally stated.