An Evolutionary Approach To Management Control Systems Research
An Evolutionary Approach To Management Control Systems Research
An Evolutionary Approach To Management Control Systems Research
A R TI C L E I N F O
A B S T R A C T
Article history:
Received 5 August 2020 By reviewing the theory underlying the design and interdependency of management control systems, I highlight the
Accepted 20 August 2020 important role firm dynamics plays in our understanding of these control systems. I provide both theoretical and
Available online 13 September 2020 empirical extensions encouraging authors to explore more longitudinal studies to better understand the path dependency
of control system design, how control system elements vary over time within a firm, and how the interdependent nature
of individual control elements might vary.
© 2020 Elsevier Ltd. All rights reserved.
“It is not the strongest or the most intelligent that will survive but those
cross-sectional differences, but also from discernible shifts in a single
who can best manage change.” e Leon C. Megginson
firm’s internal and external operating environments over time. Much like
cross-sectional differences in organizational context, time-varying
differences in organizational context also likely represent shifts in the
practices necessary to maintain effi- cient control. In their exploration of
1. Theoretical extensions complementarities, Milgrom et al. (1991) argue that a system begins
along a path and con- tinues along this path until an unmodeled force disturbs
Modern management control systems literature is built on the foundation it; i.e., that systems demonstrate momentum. This suggests that certain events
of contingency theory (Otley, 1980). Simply stated, this theory suggests that within the life of the firm may serve as triggers of MCS adoptions and
the design and use of management control systems (MCS) is dependent upon abandonments (Davila & Foster, 2009). What these events are, in addition to
its organizational context. In this article, I encourage management control when and how they affect control system design, is largely unexplored in
researchers to consider how the dynamic nature of the firm results in changing empirical management accounting research. The examination of MCS through
organiza- tional contexts and differing paths to control system design. While a dynamic lens will lead to an enhanced understanding of interactions among
not originally developed to provide dynamic predictions within a firm over control system el- ements. Following Hall (2016), I differentiate between
time, contingency theory is easily extended to suggest that, as the firm’s contingency based research and contingency theory itself to highlight that,
internal resources and external characteristics evolve over time, it will while contingency theory clearly speaks to the dynamic nature of the firm,
differentially utilize control systems and the control mechanisms within. the application of contingency theory in empirical research has been
However, the application of con- tingency theory in prior research largely largely static. Criticisms of the ways in which management accounting
utilizes a static reduc- tionist approach thereby limiting our understanding of researchers employ contingency theory have given way to new
the dynamic nature of control systems. conceptualizations of the combinations of management control elements
Prior research often examines how control mechanisms vary across that develop in practice (e.g., Grabner & Moers, 2013). This literature
organizational contexts by comparing MCS in different types of firms at a does not contend that organizational context does not play an important
single point in time. However, conceptualizing control elements with these role in control system design, but that organizational contingencies
static typologies “virtually ignores the origins and evolution of overlay a more complex network of control system design decisions.
organizational control” (Cardinal et al., 2004, p. 411). Variation in Un- derstanding how firms utilize control systems in practice requires an
organizational context does not only come from understanding of how organizational context influences not only the
choice of individual control mechanisms, but also its ef- fects on the
interdependencies of such systems. Examining control
E-mail address: [email protected].
https://doi.org/10.1016/j.aos.2020.101186
0361-3682/© 2020 Elsevier Ltd. All rights reserved.
2 M.A. Martin / Accounting, Organizations and Society 86 (2020) 101186
systems are capable of achieving? Conversely, can a firm alter its system through textual analysis. Using commonly employed techniques,
design focus to move between viewing their practices as a package or a researchers can scour firms’ public disclosures for words and phrases that
system, or is the system design a structural firm trait? Allowing contextual may shed light on the types of control systems employed by the firm. For
factors to influence the evolutionary impact of control system integration example, a simple textual search of public firm disclosures reveals that in
results in additional questions. Can we identify common points within a a 2020 SEC filing AmBev describes their desire to control costs by
firm’s evolution at which we see a shift in the design focus of MCS implementing a zero- based budgeting management control system (AmBev
between the package and system extremes? Following the model in S.A., March 24, 2020 20-F Filing). By collecting such data over a long
Grabner and Moers (2013), which highlights the role of interdependence, window, researchers can examine how the control systems change, how
an evolutionary approach may allow us to distinguish operational firms alter the discussion of existing controls over time, and whether this
contingencies that affect the relation between existing control mechanisms discussion can predict upcoming changes in control system design.
and/ or the payoff function to employing an additional control. By In implementing this more dynamic approach, researchers may have been
examining how the relation between controls changes over time within a deterred by the almost infinite combination of state and structural variables
firm, we can better understand whether and when control systems more (Gerdin & Greve, 2004). However, the config- uration school of thought
closely represent sets of loosely coupled elements or purposefully suggest most organizations fall within a limited number of system states
designed interdependent controls. (Miller & Friesen, 1984). The challenge to researchers is systematically
identifying both a parsi- monious set of system states as well as identifying
2. Research extensions the state of each firm at a given point in time. Life cycle theory considers a
firm’s life cycle stages as distinct phases that arise from a unique combination
While from a theoretical perspective it seems there is much to be of both the organization’s internal characteristics as well as the external
gained in our understanding of control systems by adopting an contexts in which the firm operates (Anthony & Ramesh, 1992; Dickinson,
evolutionary perspective, undertaking this research in practice is more 2011; Moores & Yuen, 2001). The life cycle liter- ature typically concludes
difficult. Starting from an analytical perspective, extending models from that a five-stage model sufficiently de- scribes the pattern of an organization’s
prior studies may provide testable empirical pre- dictions. For example, development (e.g., Gort & Klepper, 1982; Miller & Friesen, 1984).
Grabner and Moers (2013) model the choice of control mechanisms as a Identifying a firm’s life cy- cle stage provides a method to parsimoniously
function of both the relation between existing control mechanisms and a capture patterns in the set of state and structural variables, thereby creating a
predetermined vector of orga- nizational characteristics. While this model manageable framework for examining dynamic changes in control systems
consists of a single period control system decision based on the within a firm.1 Layering life cycle theory onto MCS research could be a
interdependence and expected payoff of existing controls, extensions may fruitful avenue for future research. Drake and Martin ( 2 0 2 0 ) pair publicly
allow for evolution in the system. Perhaps researchers could relax the available executive compensation and life cy- cle stage data to examine how
assumption that the vector of organizational characteristics are firms differentially reward earnings and returns in compensation contracts as a
predetermined and allow for expectations of contingencies based on firm firm evolves over time. As an alternative to using distinct changes in system
evolution. Additionally, further work could explore pre- dictions about, states to assess changes in control systems, researchers could also adopt the
for example, the interdependence of controls if the model is expanded to responsiveness and distinctness indices suggested by Demartini and Otley
maximize the payoff function over a longer window as the firm evolves. (2020), to investigate how firms’ MCS are able to evolve to changing internal
Turning to empirical research, one straightforward way to study the and external environments.
evolving nature of control systems would be to engage in a longitudinal field While no articles in this issue directly investigate the dynamic
study of a single firm. Here researchers could either utilize historical nature of MCS, it is easy to see how we might extend these studies. I first
information collected by the firm to assess changes in organizational context consider two studies that use surveys to focus on the role of MCS in
and control systems over time or embed themselves to collect data going innovation. Henri & Wouters (2020, this issue) investigate the
forward. While there is much to be gained by this approach, the costs can be interdependence of MCS for facilitating product innovation by examining the
prohibitive. As an alternative, researchers could utilize a survey that generates interactive roles of diversity in non-financial per- formance measures and cost
repeated observations within a firm over time. This approach al- lows for information. Speckbacker & Wabnegg (2020, this issue) also examine the
time-series analysis that informs changes in control practices and effect of interdependent con- trols on innovation behavior by investigating the
organizational characteristics. roles of explicit incentive design and implicit career incentives. I submit that
In addition to field research, advances in empirical archival data may the path dependent nature of firm evolution likely influences both the control
allow researchers to examine the impact of firm evolution on control system practices used by the firm as well as their integration. In adopting an
design. The length of time over which data has been collected for publicly evolutionary perspective, we could also consider how the MCS system
traded firms is reaching a critical mass to allow for time-series evolved to arrive at its current state. Have the existing mechanisms been
evolutionary style analyses. For example, BoardEx and ISS (formerly recently added to control innovation? Are the control mechanisms an artifact
RiskMetrics) provide a longitudinal sample of governance variables suitable of a prior firm state in which innovation was more/less important to value
for examining evolution in this aspect of control systems, though the data is creation? Does innovation equally create value in different states of the firm?
largely limited to US firms. Making use of this longitudinal data, Graham et Are the performance measures used to capture employees’ innovation
al. (2019) examine changing governance mechanisms over time, specifically
examining how the board and CEO interact as the firm evolves. In addition,
IncentiveLab provides approximately fifteen years of data on executive 1
Dickinson (2011) developed a methodology to classify firms’ life cycle stages from
incentive contracts, including specific contracting terms, providing another publicly available cash flow information. The ubiquitous nature of cash-flow disclosures
aspect of control systems to be examined longitudinally with publicly across time and across firms in a wide variety of countries makes this an appealing approach
available archival data. to examine dynamic properties of the firm.
As an alternative to variables routinely collected in large data- bases,
researchers could also opt for creating archival datasets
4 M.A. Martin / Accounting, Organizations and Society 86 (2020) 101186