Chenhall, R. H. 2003. Management control system design within its organizational context: Findings from contingency-based research and directions for the future. Accounting, Organizations and Society 28(2-3): 127-168.
Summary by Eileen Z. Taylor
Ph.D. Program in Accounting
University of South Florida, Spring 2004
Contingency Theory Main Page | Controllership/Control Main Page
This paper is a review of the empirical contingency-based literature regarding the development and structure of management control systems. It categorizes the literature by topic: meaning of MCS, outcomes of MCS, and contextual variables including external environment, technology, organizational structure, size, strategy, and national culture. The paper provides a thorough review of studies that examine these topics. Additionally, Chenhall provides recommendations for future research.
The study limits its focus to contingency-based theories developed from a functionalist perspective. In other words, the assumption is that management control systems “are adopted to assist managers achieve some desired organizational outcomes or organizational goals. The appropriate design(s) of MCS will be influenced by the context in which they operate” (p.128). The figure below has been designed to clarify the functionalist perspective.
The Meaning of MCS
There are multiple terms that have been used in describing ‘management control systems.’ The author chooses to use MCS to refer to the broad application of a system that provides both financial and non-financial information on both internal and external factors, to managers for the purposes of control and decision-making.
When studying MCS, one of two approaches can be taken. First, one can build on existing areas, trying to extend overall theories. Second, researchers may choose to evaluate new and novel ideas as they come into practice. Both types of research make important contributions. Primarily, because MCS change over time, it is important to study the implementation and success of new managerial accounting tools. Furthermore, as old MCS lose relevance, research on them does not have significant practical applications. The author also warns of the dangers of studying accounting MCS in isolation. Consideration of other organization control systems is imperative given the interactive nature of MCS with those other systems. One suggestion is to classify organizational controls as organic or mechanistic. Table 1 puts forth a taxonomy which classifies controls in this way.
Another potential issue in MCS research is the assignment of dependent and independent variables. Some studies classify performance as the dependent variable, and MCS as the independent variable. The criticism of this approach is that, given rational economic theory, organizations will always choose the MCS that leads to the highest performance; thus nullifying meaningful results. Others look at the contextual variables as independent and the MCS as dependent. Finally, some studies examine implementation of MCS and link it to performance. The author cautions that prima fascia adoption does not necessarily lead to actual use, thereby clouding the link between performance and implementation of MCS.The External Environment
The first contingent variable examined is that of the effect of the external environment on MCS. Many studies narrow this down to uncertainty and risk assessments. However, there are many other characteristics of the external environment which may be relevant.
Propositions concerning external environment
(Note: Propositions are paraphrased from the original article. For exact
wording, see original work).
Uncertainty leads to open/ externally focused MCS.
Hostility and turbulence are associated with reliance on formal controls.
Even when MCS are tightly controlled in an uncertain environment, flexibility and interpersonal interactions also exist.
Generic Concepts of Technology
The author defines technology as hardware, materials, people, software, and knowledge. The majority of studies examine complexity, task uncertainty, and interdependence as contingent factors related to MCS. When using complexity as a factor, the technology relates to the complexity of the production function. There is also a focus on the complexities of the value chain and their implications for MCS design. Overall, the more complex a process is, the more likely the MCS will be organic and less traditional.
Task uncertainty is considered as a factor associated with functional departments. For example, marketing departments have more task uncertainty than production departments, and thus need a system that includes a broad scope of information (p.140).
Interdependence, as noted earlier, can affect the study of MCS. In this case, interdependence refers to interdependence between functional areas.
Propositions concerning technology
The more technologies are standardized and automated, the more formal and traditional the MCS.
The more uncertain the task, the less formal and traditional the MCS.
The more interdependent the technologies, the less formal and traditional the MCS.
Chenhall gives space to discuss the most recent managerial accounting technological advances. These include JIT, TQM, and FM (flexible manufacturing). Overall, these new tools seem to work best when in the presence of less formal, more externally focused MCS. However, there is some indication that a hybrid MCS best serves these implementations.
Propositions concerning advanced
technologies and MCS
TQM is associated with broadly based, externally focused MCS.
Reward and compensation schemes impact the effectiveness of the combination of advanced technologies and non-financial performance measures on performance.
JIT and FMS are associated with informal controls and non-financial performance measures.
FM is associated with informal, integrative mechanisms.
Supplier partnership practices are associated with non-financial measures and informal meetings.
Basically, in this context, organizational structure refers to the “…formal specification of different roles for organizational members, or tasks for groups, to ensure that the activities of the organization are carried out”(p.144). This paper looks at some studies that classify organizational structure as mechanistic or organic. Mechanistic structures are characterized by the degree of formalization of “…rules, procedures, openness of communications, and decision processes” (p.145).
There are also studies that classify organizational structure as centralized/ decentralized or functionally differentiated/integrated.
Propositions concerning organizational
structure and MCS
Large, more decentralized organizations tend to be associated with more formal, traditional MCS.
Characteristics of functional departments, specifically task uncertainty, influence the MCS. Higher task uncertainty and higher external environment uncertainty are associated with open, informal MCS.
Leadership style can impact MCS.
Team-based structures are associated with participation and comprehensive performance measures.
Organic structures are related to future-looking MCS.
Size has been studied as a contextual, influential variable, however, many studies are limited because they only examine large firms. As firms grow larger, MCS tend to become more formal and controlling. These are primarily administrative controls and rules. In small firms, MCS focuses on interpersonal controls.
Chenhall (p.149), gives a useful analysis of possible ways to measure size. These include: number of employees, net assets, sales, and profits. Choice of a measurement depends on the dimension of MCS studied.
Propositions concerning size and MCS
Large organizations are more diversified and characterized by formalization of MCS.
Large organizations are associated with more divisionalized organizational structures.
Large organizations are associated with an emphasis on participative budgeting and sophisticated controls.
Strategy, at the business unit level, has been associated with MCS. As expected, for conservative strategies, formal controls ruled. However, as seen before, a hybrid MCS was evident in entrepreneurial strategies. Specifically, in these cases, tight controls operated together with organic communications and decision processes. Chenhall discusses ‘build and harvest” strategies, as well as prospector strategies.
There is also a discussion on possible strategy measures.
Propositions concerning strategy and MCS
Conservative strategies (defender and cost leadership) are associated with formal MCS.
Competitor-focused strategies are associated with broad scope MCS.
Entrepreneurial strategies combine formal and informal MCS.
Defender and harvest are associated with formal performance measures. Prospector strategies are associated with informal, open MCS.
Although research in this area is limited, it remains relevant. The basic premise is that culture will impact the development and effectiveness of MCS. This is especially relevant given the increased globalization of corporations. Culture classifications by Hofstede (1984) are often used in these studies. This results in some limitations. First, studies that rely on a subset of the Hofstede taxonomy may be omitting some relevant characteristics. Second, using only the Hofstede values may limit the study and exclude other relevant variables. Third, cultures change in response to globalization; therefore, the values assigned by Hofstede may not be currently applicable to the country. Last, such stereotypes may not apply equally to all individuals within a country.
Given that findings are mixed, there is only one proposition concerning culture and MCS.
National culture is associated with the design of MCS.
Although some of the previous elements will
continue to be of interest; there are new elements to be considered. For
example, there is an increased focus on environmental reporting and control.
Additionally, there is a social emphasis on employee empowerment and
fulfillment. Finally, as globalization becomes the norm, issues of size and
culture will continue to be relevant.
Various statistical and modeling variations can be used to examine MCS design. Selection studies, interactions approaches, and systems models are all considered. Statistical approaches include linear regression, structural equation modeling, and cluster analysis.Causality
Care must be taken to address causality. Is the model unidirectional (context variables influence MCS), or is it bidirectional (context variables influence MCS, which then influences context variables)?Level of analysis
The level of analysis may be individual, subunit, or organizational.Alternate Theories and contingency-based research
There is much to be gained by exploring theories from other disciplines, including psychology and economics. Psychological constructs look at how ethics and fairness can influence human behavior. Economics allows us to use agency theory to explain the development of certain MCS structures.