تفکر مدیریتی

In this article we discuss management accounting theory. We discuss the motivation for and role of
theory in management accounting, arguing that theories in an applied field such as management
accounting should provide explanations that are useful for those we study - managers, organizations
and society. We evaluate the nature of theories currently used and developed. Those theories that
are considered theories by the research community are largely imported from other social sciences,
but have hardly anything that makes them unique to management accounting. Those theories that do
not currently deserve the status of theory attempt to explain how to apply management accounting
to achieve superior performance. We argue that both forms of theories, at present, largely fail to
provide valid support for practitioners. We contend that management accounting theory should help
us to answer questions of what methods we should apply, how, in what circumstances, and how to
change management accounting. We provide suggestions on how management accounting research
could proceed to produce better theories.
4th draft, September 2005
Acknowledgements: We would like to thank Thomas Ahrens, Chris Chapman, Trevor Hopper, Jari
Huikku, Marko Järvenpää, Taru Lehtonen, Sari Lounasmeri, Kari Lukka, Kai Luotonen, Jan
Mouritsen, Robert Scapens, Jani Taipaleenmäki, Tuija Virtanen and the participants of research
seminars at the University of New South Wales, University of Technology Sydney and University
of Queensland, as well as the participants of the 2nd Global Management Accounting Research
Symposium (Sydney) for helpful comments.
2
1. The purpose of management accounting theory
Why do we research management accounting?1 A recent debate provoked by Zimmerman (2001) on
the state of management accounting (MA) research and theory (Ittner & Larcker, 2001;
Zimmerman, 2001; Hopwood, 2002; Ittner & Larcker, 2002; Luft & Shields, 2002; Lukka &
Mouritsen, 2002) provides some answers, but also raises a number of further questions. The
following is particularly puzzling: what is the purpose and role of theory in management accounting
research? Given that we can answer to this, the next obvious question is do current theories fulfill
this purpose and how we should, as an academic community, proceed in our theory development?
These are the issues we address in this study.
In their response to Zimmerman, Luft & Shields (2002) refer to their extensive literature review on
theory consistent research in management accounting (Luft & Shields, 2003). The review focuses
on articles that explain the causes and effects of management accounting. This is to say that the
theory of, or theorizing in, management accounting is about explaining its causes and effects.
Hopwood (2002, p. 783) refers to the emergence, functioning and impacts of management
accounting practices as an object of inquiry, thus extending the scope from causes and effects to
include how management accounting is practiced2. Although not all accounting researchers
subscribe these views (see Baxter & Chua, 2003, for a review of some of this literature; see also
Chua 1986), this seems to be the mainstream conception prevailing in today’s management
accounting research community. But why do we, as an academic community, try to understand the
causes, effects and functioning of management accounting? We believe that the ultimate reason for
this is to be able to use this understanding, or theory, in creating “better” MA practices, both in
terms of content/form and use (cf. Ittner & Larcker, 2001, p.399; Chenhall, 2003, p. 159). The
definition of “better” is not crucial here. The argument is that we undertake research and develop
theories in management accounting to be used by someone to accomplish something.3
It is several years ago that two studies took as their task to explicitly analyze the issue of how to
conduct practice relevant MA research and produce theoretical contribution (theorize) accordingly.
Kasanen et al. (1993) argued for constructive research approach in management accounting (see
also Labro & Tuomela, 2003) and Kaplan (1998) for innovation action research, both sharing a
normative agenda and an aim to contribute directly to practice trough scientific MA research. As
Kasanen et al. (1993, p.262) summarize it, “..management accounting is, in the end, a practical field
where theory without pragmatic implications is empty”.4 These two pieces seem to have had a very
limited impact on research practice ever after. The question is: why? Of course there are problems
related to normative research agendas – as there are to any research approach – which could explain
the lack of studies called for by these authors. One can probably find convenient arguments to
object normative research approaches, which we suggest to be only one possible way to pursue in
future MA research. But, it is much more difficult to come up with good arguments to object
production of practice relevant research results/theories in an applied science like management
accounting: “…accounting is fundamentally an applied research area that should ultimately provide
new insights for practice” (Ittner & Larcker, 2002, p.788). Indeed, as we will show below, nobody
truly seems to object that we should do practice relevant research. Therefore, one conclusion seems
1 By management accounting we refer to both accounting for decision-making and various forms of managerial control.
2 Functioning may be captured in studying causes and effects by the variance in MCS as a result of causes, or as a
reason for variance in effects. This interpretation would suggest these authors use different words for similar objects of
inquiry.
3 Mattessich (1984, p.2) argues that applied science must be concerned with goals, norms and prescriptive conclusions
(see also Reiter & Williams, 2002, p. 581). We will return to this issue below.
4 Pragmatism is according to standard definitions a distinctly American philosophical movement founded by Charles S.
Peirce and expounded upon by William James and John Dewey. Essentially, pragmatism asserts that if a certain
proposition has practical meaning or produces practical results, then the proposition is determined to be true.
3
to be that there is some disagreement of what is deemed as practical or practice relevant among
researchers (cf. Miller & O’Leary, 1990), and especially of how we could produce practice relevant
management accounting theories (cf. Mattessich, 1995a, b). We present in this study arguments and
critique regarding this issue by leaning on the ideology of pragmatism, according to which practical
consequences are the criteria of knowledge, meaning and value (e.g. James, 1955; see also Kasanen
et al., 1993).5 We also subscribe to the fundamental ideas of Mattessich (1995a, b) regarding the
problems of current theorizing in accounting research, although we focus merely on management
accounting.
Watts and Zimmerman, in their Positive Accounting theory (1986), justify the importance of
accounting theory by referring to a number of interest groups that have to make decisions about
external accounting reports (p.2). Moreover, they argue that one important criterion for a theory’s
success “is the value of the theory to users” (see also Demski, Dopuch, Lev, Ronen & Searfoss,
1991). Hence, advocates of positive accounting theory subscribe to the practical purpose for a
theory in accounting, although they beware of being normative.
Baxter and Chua (2003) review alternative management accounting research. They summarize the
radical alternative (p.99-100) as “…mobilizing research to provide a platform for critique, change
and improvement within organizations, in particular, and society, in general”. Therefore, in a wide
sense the radical (critical) research seems to have a similar purpose of change and improvement as
positive and constructive approaches. Although it is difficult to judge the various purposes and
motivations of research undertaken in various streams of alternative management accounting
research, Baxter and Chua summarize it as follows: “…this legacy enables us to appreciate the
limitedness of management accounting inscriptions and the technologies that generate them”. It
appears that such appreciation of limitedness is required to be able to act upon it. Similar thoughts
may be drawn based on Chua’s earlier assessment on the interpretative perspective (Chua, 1986).
Although explicitly denying the technical application of interpretative science, she argues that
“…the aim of the interpretative scientist is to enrich people’s understanding of the meanings of their
actions, thus increasing the possibility of mutual communication and influence” (p. 615). If our
interpretation is right, then more or less all strands of current MA research seem to share a common
purpose for theory and research: the aim at the end is to come up with some guidelines or
suggestions as to how to apply, or not to apply, management accounting.6
Despite its practical purpose, explicit or implicit, management accounting research is often
criticized of not having an impact on practice, let alone leading it. Swieringa (1998, p.43) points to
the fact that most academic researchers are invisible to broader audiences than just their colleagues
(cf. Lee, 2003):
“They do research, publish the results, review the work of other researchers, and build reputations. Academic
researchers usually write for other researchers and experts in their field. They are expected to be objective, to
meet high technical standards and to subject their work to rigorous peer reviews, and they are concerned about
the quality and cleverness of their work and the reactions of colleagues”.
5 We also lean on our wide experience of theoretical and empirical management accounting research, and
comprehensive experience of working with practitioners.
6 Organization theorists seem to share this view for theory to be useful for users (other than researchers). Hinings &
Greenwood (2002; see also Clegg, 2002; Bazerman, 2005) argued that originally the research agenda in organization
theory (or organizational sociology) was: what are the consequences of the existence of organizations? After the field
became business school driven, the question has been: how to understand and thus design efficient and effective
organizations. Although the shift has been from policy making to management, there has always been an assumed user
for the analysis and developed theory. Hambrick (1994) called for relevance in the field of management studies, arguing
that “our responsibility is not to ourselves but to the institutions around the world that are in dire need for improved
management, as well as those individuals who seek to be the most effective managers they possibly can be.”
4
We believe this orientation towards colleagues, not external users of our theories has partly
contributed to the situation where academic management accounting research has had a very limited
effect on practice during the last few decades7. As considerable amounts of state and other funding
are granted to perform research and education in the area, it seems that Swieringa’s description of
the role of accounting researchers is not feasible in the log run (cf. Hambrick, 1994). It should not
be in anybody’s interest to generate research articles that, at worst, nobody cites (this is not tied to
the position of the publication outlet in journal rankings) and the results, if any, of which could not
be of less interest to practitioners. We do recognize, however, the underlying tenure and élite
systems as well as the myopic university and education program rankings that perpetuate this
process (prospering in the US in particular; see Lee, 1997, 2003; Lee & Williams, 1999). Therefore,
we understand that such change cannot take place overnight. But we do suggest that changing partly
what we perceive as the purpose and role of theory in management accounting may aid in making
management accounting research relevant to a broad audience. As Bazerman (2005, p.29-30) puts
it: “…there are barriers to the perceived efficacy of non-economic social scientists. […] These are
real tradeoffs and barriers. However, I believe the primary barrier to having influence is the failure
to even try”.
Before we move on to discuss what is theory, we should revert to the very beginning of the study,
because in the end the question of why we research management accounting is an epistemological
one. The different strands of research referred to above derive the conclusion that the aim of
research at the end is to come up with some guidelines as to how (or not) to apply management
accounting from different directions. This is because underlying their knowledge creation processes
we can find different interests of knowledge and thus different epistemological assumptions
(Habermas, 1971; see also Lukka & Granlund, 2002; cf. Latour, 1987).8 The technical interest of
knowledge is known to dominate in the natural sciences. According to Habermas (1971), natural
sciences are seen to be meaningful and their results valuable because they make it possible to
forecast and control the events of reality. Many fields of social sciences, including economics,
largely share a technical interest of knowledge. This is an appropriate objective also regarding the
arguments of the present paper, since guidelines or suggestions as an ultimate outcome of research
refer directly to the ability to control events.
The practical interest of knowledge aims for a profound understanding of cultural phenomena and
an enhancement of our ability for self-reflection through communication. While not denying the
merits of research committed to this, in light of the arguments presented here this may not be
enough alone: understanding a phenomenon – how ever deep it be – does not produce directly
usable results for various user groups. However, such understanding could be developed to
guidelines or suggestions; it could be combined into the agenda argued for in this paper, as, for
example, understanding the effects of organizational power distribution may well be relevant for
certain types of management accounting theories. One could probably also claim that at best this
interest of knowledge may provide practitioners with new insights to be applied in their own
problem solving, though the problem still being how to make practitioners aware and to care about
such research results. So while not saying that all earlier research would be impractical and not
trying to contribute to practice, we think the success in pursuing this has been very limited or
underwhelming.9
7 A more complete account of various reasons why research has had limited impact on practice is beyond the scope of
this paper. Lee (2003) provides one recent and persuasive account on this (see also Inanga & Scheider, 2005).
8 The fact that epistemology is concerned with what is true (adequate) and what false (inadequate) knowledge, i.e. how
knowledge accumulates, practically translates into the question: how can one develop theories that are better than
competing theories?
9 Habermas (1971) identifies also a third interest of knowledge, the emancipatory one, which can also be found in the
field of accounting research (though maybe today to a lessening degree). This interest of knowledge is driven by the
aim to gain a consciousness about the illegitimate modes of repression as well as by the attempt to dissolve them,
5
Our position in this paper is in favor of the technical interests of knowledge, although we do
acknowledge that research based on other interests of knowledge may at best provide potentially
useful insights in developing theories that are in the interest of, and useful for the examined
people/organizations.10 Ideas presented in this paper are still likely to be most relevant for those
sharing the technical interest of knowledge. However, adopting this interest of knowledge does not
mean that all the (ontological) assumptions prevailing in natural sciences will apply in social
sciences, even though the interest of knowledge may well be the same. Predictability and
controllability can in our view apply at several levels and also vary with regard to intensity.
Therefore it is important to also acknowledge the many local, contextual meanings and influences
that may be at work as we analyze and build theories of the uses and effects of MA systems.
As a starting point for the remaining analysis, we regard management accounting as an applied
science, where the aim of research and theory in the long run is to facilitate attempts to make
organizations and societies better off. There seems to be widespread concurrence to this, but much
less accord on how we should proceed towards that aim. In the next section we will define what we
mean by theory. We then assess briefly what type of theories we have at the moment to help us in
building better management accounting practices and what the main shortcomings of those are. This
is followed by what we see should be the core of management accounting theory, or a set of
theories. We then provide some avenues to develop such theories. Finally, we summarize the main
arguments.
2. What is theory?
Social scientists have had troubles in defining what theory is (Sutton & Staw, 1995; Llewelyn,
2003). There is also great debate about what kind of theories we should be looking for in social
sciences in general (about the “Science Wars”, see e.g. Flyvbjerg, 2001). Although some authors,
like Flyvbjerg, argue that social science has set itself an impossible task when it attempts to emulate
natural science and produce explanatory and predictive theory, we are inclined to think that while
this may hold for some strands of social science, it requires reconsideration with regard to applied
streams of science, such as management accounting.
Weick (1989) quotes Sutherland in defining what we mean by theory - “an ordered set of assertions
about a generic behavior or structure assumed to hold throughout a significantly broad range of
specific instances”. This corresponds fairly well to a standard dictionary definition: “a coherent
group of general propositions used as principles of explanation for a class of phenomena”
(Webster’s; our italics). We define theory in similar lines. Theory is understood here as a coherent
set of propositions used as principles of explanation for a class of phenomena assumed to hold
throughout a broad range of specific instances. However, we refer neither to a significantly broad
range of instances nor general propositions, as these words tend to support the idea of “grandtheory”
or theory as covering laws (DiMaggio, 1995). We believe that our field of inquiry,
management accounting, is relatively heterogeneous across industries, cultures, firm life cycles,
ultimately leading to criticism towards predominant ideologies. Modern management control techniques are regarded in
this context as one of the instruments used to support the prevailing status quo in society (cf. Hopper & Powell, 1985).
Practitioners easily neglect such an approach, typical of critical social analysis: provision of mere criticism without
suggesting concrete remedies is undoubtedly unappealing for people aiming to solve an imminent problem (Lukka &
Granlund, 2002)
10 We are neither saying that we should abandon other types of research agendas as those suggested in this study. We
are rather suggesting that there should be something else too: something that we see practically unavoidable in the long
run. It may be an extension or modification regarding current research agendas, or it could be a new agenda to be
pursued alone or side by side with the existing ones.
6
organization size, and so on, and is likely to benefit from a broader set of theories instead of one
grand theory.
Whetten (1989) argues that a complete theory must contain four elements. The first is which factors
(variables, constructs, concepts) logically should be considered as part of the explanation of the
phenomena of interest. The second element is concerned with how they are related. Together what
and how constitute the domain or subject of the theory. In research papers providing formal models,
constructs are boxes and relations are arrows.11 The third element relates to the question why. What
are the underlying psychological, economic, or social dynamics that justify the selection of factors
and the proposed causal relationships? The fourth element focuses on the conditions that place
limitations on the propositions generated from a theoretical model. Questions such as who, where
and when address the temporal and contextual factors that set the boundaries of generalizability.
This view is a more detailed account of what theory is than the definition we provided above. We
do not regard these accounts to be contradictory in any way. For us it appears that this view
provided by Whetten (1989) is more or less the view held by most positive accounting researchers.
However, sometimes accounting scholars seem to refer to the theory being simply the
psychological, economic, or social dynamics providing answers to why questions. Our view on
theory is akin to that of Whetten and we could apply equally well Whetten’s account to define what
we mean by theory.
Watts and Zimmerman (1986, pp. 7-9) make a clear distinction between positive and normative
theory. Theory for them yields no prescription for accounting practice. They argue that prescription
requires the specification of an objective and an objective function. In principle, we agree that
prescription requires the specification of an objective and an objective function. However, we argue
that most of the economics based management accounting research assumes implicitly the objective
of economic efficiency or shareholder value maximization. Therefore, theories capable of
explaining and predicting how certain forms and uses of management accounting 1) lead to
decisions that are in line with these implicit objectives, 2) increase goal congruence and 3)
ultimately financial gains are likely to provide prescriptions relevant for business managers
pursuing such goals. Moreover, we argue that researchers should develop management accounting
theories devoted to value maximizing, theories devoted to social equality, theories devoted to
environmental sustainability, etc. In other words, we do not believe that there can be only one base
theory in management accounting we can build on just by varying our objective function. Research
on management accounting would be better off if researchers would clearly assume and explicate an
objective and objective function, and build theories to support that objective. To avoid any possible
misunderstanding it is perhaps important to note that all this does by no means imply that we as
researchers should operate purely on the conditions of the research objects. The above neither
means that we could simply assume objective functions and choose between them according to our
own values. Our view is here in line with Flyvbjerg (2001), who suggests that researchers can carry
their own values and mobilize them with consideration in research: even to question the
assumptions and intentions of the examined organizations (Jönsson & Lukka, forthcoming). The
conclusion would be that researchers should assume an objective function so that their own values
would not be offended while developing theories. Values and ethic codes, for instance, are reflected
already in what kind of problems the researcher wants to examine in the first place, and how they
are to be solved.
To sum, even if we call for theories to be useful in practice, we do not mean that theories need to be
normative as such. Theories capable to explain and predict are likely to provide practical insights
given that they deal with the issues that are of practical interest, i.e. explain and predict phenomena
11 These elements of theory are illustrated in the summary maps by Luft & Shields (2003).
7
related to practitioners’ objectives. On the other hand, the definition of theory as a coherent set of
propositions used as principles of explanation for a class of phenomena assumed to hold throughout
a broad range of specific instances does not specify the nature of the set of propositions. These
propositions may as well be normative, including objectives.
3. The types of theories in management accounting
Research and practice in management accounting relies largely on two types of theories. Those
theories that are considered theories by the research community are imported mainly from other
social sciences, but have hardly anything that makes them unique to management accounting. Those
theories that do not currently deserve the status of theory attempt to explain how to apply
management accounting to achieve superior performance. Let us address them both in turn.
3.1 Theories that have theory status
Based on Luft & Shields (2003), in management accounting research a number of MA related
issues are explained by a number of theories. Typically MA is either an independent or a dependent
construct, but the whole causal chain is not addressed. Theories applied to explain causes, effects
and various interrelationships of MA are mainly from the fields of economics, organization theory
(contingency theory), sociology and psychology. There is nothing wrong with this; we need these
theories and many MA issues of interest can be explained by these theories. It is interesting though
to think about this practice in terms of Whetten’s definition of theory. These independent borrowed
theories have their own initial constructs, relationships, explanations and assumptions, depending
what is the phenomenon they are about to explain. Some of the explanations and assumptions (but
not constructs and relationships) are then applied to explain MA constructs and relationships. In
other words, we borrow only parts of these theories, mainly to provide answers to why questions.
Again, while this in itself is not necessarily problematic, we have four main concerns with current
theorizing in MA.
First, none of these theories is something we could build on if we wish to distinguish MA theory
from some other theory. For us, these seem to be theories about management accounting, not
theories of management accounting (Humphrey & Scapens, 1996). Management accounting theory,
as judged based on the publications in “top tier” academic research journals, is determined by the
object of inquiry. Researchers use principal/agent theory, information economics theory,
structuration theory, actor-network theory, goal setting theory and a countless number of other
theories to explain issues of interest in management accounting. Sometimes it seems that what is
interesting is determined by the pre-adopted theory (and/or method tool-box), not the accounting
phenomena of interest.12 These same theories can be used to explain a number of issues other than
MA. We argue that there is a need for theories of management accounting. This is not to say that we
do not need the explanations provided by and insights derived from these other theories. We
certainly do. But we argue there is room and need for both.
Why do we need unique theories of management accounting? We believe that these borrowed
theories, even when applied to explain management accounting related issues, provide little insight
on practical issues as such. A set of unique theories of management accounting would be closer to
practitioners’ concerns than the current theories used in management accounting research. Thereby
12 See Reiter & Williams (2002) and Mattessich (1984) for positive economic theories, and Scapens (1990) in the
context of case studies informed by sociology. In addition, in the mainstream research tradition the quality criteria for
research are borrowed from natural sciences. This has led to situations where especially statistical generalizability has
been used as “an amulet” to demonstrate contribution, while at the same time the economic significance of such
statistically significant results may have remained practically non-existent (see Lukka & Kasanen, 1995).
8
theories would be easier to communicate (directly) to various user groups. Furthermore, this would
mean more influential research with (direct) practical effects. The consequence would be a stronger
identity to management accounting. This should significantly enhance the acknowledgement and
respect of management accounting research outside the academia. It is sometimes depressing to
encounter managers dealing with management accounting issues that do not know that management
accounting research even exists. Furthermore, perhaps this would also attract new good quality
students to our field.
A limited insight for practitioners is the second concern we have with current theorizing. Let us use
a recent summary on contingency research as an example of how current “science” based theories
fall short in providing practice relevant guidance. We use the contingency framework as a frame of
reference as we believe it is fairly explicit in its aim of providing insights to practice and as we see
theories of management accounting necessarily contingent (as opposed to covering laws). Chenhall
(p.138) provides the following summary concerning research findings relating MCS to the external
environment:
1. The more uncertain the external environment the more open and externally focused the MSC
2. The more hostile and turbulent the external environment the greater the reliance on formal
controls and an emphasis on traditional budgets
3. Where MCS focused on tight financial controls is used in uncertain external environments they
will be used together with an emphasis on flexible, interpersonal interactions.
Although we understand this is only a summary, it persuades us to ask what it tells to managers.
The first two propositions claim that both externally and internally focused management control
systems are observed to exist (to be useful?) in uncertain and/or turbulent environments. So any
system that has external focus, or internal focus, would do? It does not matter how systems are
used? This is true in all industries? These claims are also contradictory, unless we consider
uncertainty and turbulence as clearly distinct phenomena. If we assume that firms are optimizing
their MCS and that the observations reflect good practice to follow, the first point suggests that as
uncertainty increases, more emphasis should be placed on externally focused systems (more
emphasis with respect to what? More than internally focused?). The second one proposes that more
emphasis should be placed on formal controls and traditional budgets. Of course, they may both
hold at the same time, indicating that as uncertainty increases, both should receive more emphasis
(though, more with regard to what? With regard to a stable environment? Is this not too obvious?).
The third proposition is a bit more useful for practitioners suggesting that in uncertain environments
tight financial control needs to be supported by informal interactions. Our argument is that these
propositions or findings are so general, and partly self evident, that they are of little use or
incremental value in practice. There is a need to advance from this stage to be able to argue that as
uncertainty increases, certain forms of MCS used in a certain way would provide better decisionmaking
support, or more likely achievement of goal congruence.
The third major concern with current theorizing is the “meta theoretical” nature of some of the used
theories. The problem with meta theories is that you cannot falsify them, or you cannot even
generate propositions based on them which could be turned into a falsifiable form. For example,
Giddens’ (1984) theory of structuration, a meta theory, points us to look for certain issues, such as
the relationship between norms, resources and power, in organizational and societal practices. There
is nothing wrong with this per se. However, as Giddens himself says, it is not necessarily the task of
“meta theorists” to examine practice: he leaves this to others. But how that should be done in a
management accounting context? Giddens, for instance, seems to have no preference for research
methods to be applied in different circumstances. At worst, the application of such meta theories
leads to accounting research where empirical practices are described and explained using
complicated concepts (totally strange to practitioners) and constructs borrowed from these theories.
9
Often such use of meta theory does not make the complicated world more easy to comprehend, let
alone manage; rather, it may add confusion. Although the merits of research applying meta theories
are to be acknowledged (though they are debatable) and such theories may be useful at the early
stages of research projects, we argue that there is also need for an other research agenda. In this
agenda meta theories should be left in the background at subsequent phases, and emphasis should
be placed on how to capture the praxis into easily understood concepts and models.13 We do not
mean here that praxis as such would be simple or easy to model, but rather point to the
communicative aspect; how to present issues so that they are as comprehensible as possible to
practitioners.
Fourth, we do not have a clear picture of which one(s) of these theories provide(s) the strongest
explanations for various MA forms and outcomes in a given time frame and circumstances. For
example, economics based theories explain the existence of MA practices as means to enhance
efficiency. Political theories explain these practices by referring to various interest groups and their
struggle for power. Institutional theories provide signaling as an explanation as firms seek for
legitimacy from their stakeholders. Empirical research has demonstrated that all these have had an
impact on management accounting practices. But do we know when each of these explanations is
likely to be valid? Are these mutually exclusive? And even if we can answer these questions, so
what? In other words, even if we understand that accounting is practiced for various reasons, what
are the implications of this knowledge? Do these various reasons or motives to adopt practices have
an impact on how these practices are implemented and what the expected benefits are? In the case
where we do not understand the relative power of various theories explaining the same phenomena,
we do not actually understand the issues surrounding MA in sufficient depth (and breadth) to be
able to provide clues about how to apply MA. One may argue that this is only a question of
insufficient accumulation of research knowledge, and will be defeated as the field matures. We
doubt this as we seldom see attempts to address the strengths and weaknesses of various theories in
a single study (see e.g. Ittner et al., 2003, for an exception). Accounting scholars seem to believe
one theory at a time, and make few attempts to modify or further develop, let alone reject those
borrowed theories they use (see also Reiter & Williams, 2002). If we are not willing to develop
theories we borrow from elsewhere to better explain issues we are interested in, and if we are not
prepared to assess the relative strengths and weaknesses of various theories, there is little hope for
coming up with practice relevant management accounting theories14. We reject Zimmerman’s
(2001) solution to rely on only economics based theories (see also Lukka & Mouritsen, 2002); for
us it is once again an example of putting the carriage before the horse. Rather, the explanatory and
predictive power of various theories needs to be assessed, and theories, whatever the origin, need to
be developed, modified or combined to increase their explanatory power in a management
accounting context.15
3.2 Theories that do not have theory status
In management accounting we also have a number of “normative” theories or constructs. These
include activity-based costing for overhead allocation, Balanced Scorecard for control system
design, Quality Costing framework to manage and reduce quality costs and Value Based
Management framework to guide decision-making and control to ensure shareholder returns. These
13 DiMaggio (1995) discusses three views on theory. The use of some of these meta theories may conform to the view
of theory as enlightenment, where theory is complex, defamiliarizing and rich in paradox.
14 In their review, Luft and Shields (2003) considered only studies that found support for the proposed hypothesis.
Although probably a reasonable choice for the review purposes, it may be seen to illustrate the interplay between theory
and empirical data in current management accounting research. If the data does not support the hypotheses, then it is not
the fault of improper (borrowed) theory, but inappropriate data or methods.
15 Many of the remarks we make here regarding the first and the fourth concern are tightly connected to the earlier
mentioned academic systems (e.g. Lee, 2003).
10
are not regarded as theory by academic researchers, although these theories claim how we should be
doing something and why. Zimmerman (2001) argues this as follows: “while it (VBM) resembles
various theories, it is not a positive theory in the sense that it neither explains nor predicts firmrelated
phenomena”.16 Even if we accept Zimmerman’s quest for positive theory, we disagree with
his view that VBM neither explains nor predicts firm-related phenomena. It could be argued that the
VBM framework aims to explain firm performance, measured by shareholder returns. Adherence to
six managerial steps forming the core of VBM (Ittner & Larcker, 2001; Zimmerman, 2001; Martin
& Petty, 2000; Morin & Jarrell, 2001) explain why one of the otherwise similar firms perform
better, or why a performance of a certain firm improves over time. This same logic can be extended
to prediction, and these predictions can be refuted by evidence.
Recall our definition of theory: a coherent set of propositions used as principles of explanation for a
class of phenomena assumed to hold throughout a broad range of specific instances. But is this not
exactly what VBM proponents claim? There is a coherent set of propositions regarding how
organizations should be managed in order to maximize shareholder wealth, which are assumed to
hold in a broad range of instances. These propositions of VBM are not tested, but it can still be seen
as a theory. Watts and Zimmerman (1986) contend that the objective of accounting theory is to
explain and predict accounting practice (p.2). Sure, VBM is not a theory of accounting practice but
a theory of organizational performance, including accounting related issues as a mechanism of
explaining outcomes. We can even argue, that as VBM literature explains the mechanism by which
better performance is about to emerge, it deserves a status of theory more than some other so called
theories, that we use to predict outcomes (usually with low R squares), but which treat the actual
mechanism (i.e. firm) as a black box (e.g. S-curve models in innovation diffusion literature).
How does VBM comply with the four elements of theory as discussed by Whetten (1989)? We may
think that the six managerial steps provide a frame for a managerial construct, organizational
performance being the other major construct in this theory. This managerial construct consists of
managerial choices and actions, including management control related choices. We may call the
main choices and actions illustrated in VBM literature as sub-constructs. Six managerial steps now
suggest which values those sub-constructs may get in firms practicing VBM. For example, step 4
(see Ittner & Larcker, 2001) in VBM theory argues that action plans, performance measures and
target setting should be based on identified value drivers. In practice, performance measures may be
defined relying on e.g. value drivers, strategy, TQM framework or they may be ad hoc.
Performance measures can now be considered a sub-construct and part of the managerial construct.
Values for the performance measure sub-construct may vary, including e.g. “measures based on
value drivers” or “measures defined ad hoc”.
The relationship between the managerial construct and organizational performance is simple: If all
these managerial steps are taken as suggested, or more precisely, if all values of sub-constructs are
aligned with VBM theory, firm performance will improve. Selecting some alternative methods to
accomplish these managerial tasks will produce less shareholder value.
Why does this relationship between these constructs then exist? We do not suggest that this theory
is unambiguous about why these steps are to produce better performance; though we consider the
explanation fairly simple and straightforward. Considering the management control part of it, for
example, it seems that expected benefits are based on an assumption that you get what you measure.
Moreover, making sure that measures reflect value creation (either outcome or drivers) should
ensure that the system creates right incentives for managers to behave in the best interest of
16 Ittner & Larcker (2002) echo this in their reply by saying “we did not intend this framework to be a theory of
managerial accounting practice, or a complete depiction of the many economic and non-economic factors affecting