Dealing with uncertainty in knowledge-intensive firms: the role of management control syst

Description
Little research on knowledge-intensive firms has focused specifically on management control issues. This paper aims
to consider such issues. Starting from the limitations of the definition of uncertainty,especially when applied to contexts
characterised by knowledge intensity,this study investigates the relationship between knowledge complexity and
management control systems. This relationship is analysed in the realm of knowledge-intensive firms’ teams where it is
particularly critical due to the double coordination and knowledge integration role played by management control
systems. A field research conducted in three project teams of a software firm supports the relevance of knowledge
complexity in explaining the variation of management control systems. The paper concludes with some avenues for
future research.

Dealing with uncertainty in knowledge-intensive ?rms:
the role of management control systems as knowledge
integration mechanisms
§
Angelo Ditillo*
L. Bocconi University and SDA Bocconi, Viale Isonzo, 23, 20135 Milan, Italy
Abstract
Little research on knowledge-intensive ?rms has focused speci?cally on management control issues. This paper aims
to consider such issues. Starting from the limitations of the de?nition of uncertainty, especially when applied to con-
texts characterised by knowledge intensity, this study investigates the relationship between knowledge complexity and
management control systems. This relationship is analysed in the realm of knowledge-intensive ?rms’ teams where it is
particularly critical due to the double coordination and knowledge integration role played by management control
systems. A ?eld research conducted in three project teams of a software ?rm supports the relevance of knowledge
complexity in explaining the variation of management control systems. The paper concludes with some avenues for
future research.
#2004 Elsevier Ltd. All rights reserved.
Introduction
The understanding of how a ?rm can manage
knowledge is an issue that has received increasing
attention in both theory and practice over the past
ten years: on the one hand, we have seen the
emergence of the knowledge-based theory of the
?rm, on the basis of which, knowledge and the
capability to create and utilise such knowledge
are the most important sources of competitive
advantage (Prahalad & Hamel, 1990; Nelson,
1991; Henderson & Cockburn, 1994; Nonaka &
Takeuchi, 1995; Boland & Tenkasi, 1995; Grant,
1996; Kogut & Zander, 1996; Nonaka et al., 2000);
on the other hand, there has been an attempt to
de?ne knowledge-intensive ?rms and explain their
organizational and management features (Ber-
nardi & Warglien, 1989; Greenwood, Hinings, &
Brown, 1990; Hinings, Brown, & Greenwood,
1991; Starbuck, 1992; Winch & Schneider, 1993;
Alvesson, 1993, 1995, 2000; Nurmi, 1998). In gen-
eral terms, knowledge-intensive ?rms refer to those
?rms that provide intangible solutions to customer
problems by using mainly the knowledge of their
individuals. Typical examples of these companies
are law and accounting ?rms, management, engi-
neering and computer consultancy organizations,
and research centres. The category overlaps with
0361-3682/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2003.12.001
Accounting, Organizations and Society 29 (2004) 401–421
www.elsevier.com/locate/aos
§
The research has been ?nanced with Government/L. Boc-
coni University funds allocated to the ‘Young Researchers’
Project 1999’. The paper has bene?ted also from ?nancial
resources by the Research Division of SDA Bocconi.
* Tel.: +39-02-5836-2576; fax: +39-02-5836-2561.
E-mail address: [email protected] (A. Ditillo).
the concept of professional service ?rms, but is
broader, and does not focus on the features ascribed
to a typical profession, such as a code of ethics, a
strong professional association, monopolization
of a particular market through the regulation of
entry and so on (Raelin, 1985). In this paper I will
discuss some central aspects of the management
control systems
1
of this kind of organizations. To
date contributions on both knowledge-intensive
?rms and management control systems have
almost completely neglected this issue.
The literature on knowledge-intensive ?rms has
focused mainly on the reasons for, and con-
sequences of, the distinctiveness of this type of
?rms from other kinds of organizations and has
mostly devoted its attention to the more obtrusive
mechanisms of management such as the profes-
sional control of tasks, culturally based forms of
co-ordination, and ideological modes of control
(e.g. Smigel, 1963; Hall, 1968; Alvesson, 1993,
1995; Abernethy and Stoelwinder, 1995; Cooper et
al., 1996; Dirsmith et al., 1997; Montagna, 1968;
Morris & Empson, 1998). Management control
mechanisms have not been explicitly addressed in
the relevant contributions on the topic (Alvesson,
1995) or even have been considered, under certain
conditions, counterproductive (Raelin, 1985;
Nelson, 1988; Van Maanen & Kunda, 1989;
Winch & Schneider, 1993; Alvesson, 1993).
Much of the management control thinking has
concentrated its attention on the speci?cs of con-
trol systems design in manufacturing settings,
where the activities are considered to be well sui-
ted to the use of such mechanisms. Only recently,
has some attention been devoted to understanding
their role in other contexts where the tasks di?er
substantially from the physical production of
goods and are likely to include some tasks which
are relatively extreme in terms of task uncertainty,
as in knowledge-intensive ?rms. Some instances
of the contributions included in this stream of
research refer to the design of control systems in
research and development organizational units
(Abernethy & Brownell, 1997; Birnberg, 1988;
Brownell, 1985; Hayes, 1977; Kamm, 1980; Rock-
ness & Shields, 1984, 1988)
2
and the use of these
mechanisms in product innovation projects (Koga
& Davila, 1998; Nixon, 1998; Davila, 2000).
3
Yet,
1
The de?nition of management control systems has evolved
over the years from one focusing on the provision of more for-
mal, ?nancially quanti?able information to assist managerial
decision making to one that embraces a much broader scope of
information (Chenhall, 2003). Here, the term management
control systems is used to name the design as well as the use
of coordination mechanisms based on the standardization of
either input, action or results (Thompson, 1967; Mintzberg,
1979, 1983). In this way we follow Merchant (1985), according
to whom, the array of controls available for coordinating
and controlling tasks spans from the use of result, to action and
personnel/cultural controls.
2
The studies on research and development mainly suggest
that management control systems constrain or are irrelevant in
R&D settings. Some contributions focus on how R&D depart-
ments use accounting controls (Hayes, 1977; Brownell, 1985;
Rockness & Shields, 1988) and show that ?nancial indicators
do not assume speci?c relevance in these departments other
than signalling the commitment of the organization to its R&D
e?orts. Other contributions, by adopting a wider de?nition of
control systems, ?nd only limited relationships between them
and project characteristics. For example, Abernethy and
Brownell (1997) demonstrate that ‘‘reliance on accounting
controls has signi?cant positive e?ects on performance only
where task uncertainty is lowest’’ while ‘‘behavior controls
appear to contribute to performance in no situation’’ (p. 245).
This evidence seems to suggest that management control sys-
tems have a minor role to play in contexts characterised by a
high level of uncertainty (Hirst, 1983; Brownell & Hirst, 1986;
Brownell & Dunk, 1991).
3
The literature on product development suggests that when
management control systems provide information directed to
coordination and learning, they a?ect performance in a positive
way (Koga & Davila, 1998; Nixon, 1998). But alternative
arguments and evidence (Eisenhardt & Tabrizi, 1995) propose
that such a relationship does not exist or is negative. Manage-
ment control systems, by imposing rules and constraints on
behaviour, reduce the level of creativity necessary to develop
new products and, thus, negatively a?ect performance (Ama-
bile, 1998; Davila, 2000). These arguments are in line with the
traditional view of product development, according to which,
successful new products derive from avoiding control proce-
dures that could restrict the level of freedom available to
researchers (Lothian, 1984; McNair & Leibrfried, 1992). The
e?ect of the use of management control systems on product
development performance is, therefore, unclear. So far, only
Davila (2000) has tried to explain this lack of clarity by sug-
gesting that these contradictory results might be the result of a
di?erent interpretation of the role of management control sys-
tems that should be considered as information tools to face
uncertainty rather than control mechanisms to reduce goal
divergence (Hirst, 1983; Brownell & Hirst, 1986; Brownell &
Dunk, 1991; Hartman, 2000).
402 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
the evidence contained in these contributions is
still sparse, and that which does exist is mixed.
One of the possible explanations of the contra-
dictory results reported in these earlier contribu-
tions may be related to both the variables that
have been used to describe uncertainty and the
role assigned to management control systems. The
concept of uncertainty adopted in the manage-
ment accounting literature mainly refers to either
Perrow’s model of technology and structure (Per-
row, 1970)-which considers task analysability and
the number of exceptions as the relevant dimen-
sions of analysis-or ‘‘the di?erence between the
amount of information required to perform a task
and the amount of information already possessed
by the organization’’ (Galbraith, 1973, p. 5) (e.g.
Hirst, 1981; Rockness & Shields, 1984; Abernethy
& Brownell, 1997; Davila, 2000). Yet, more recent
contributions, mainly in the organization ?eld of
research, have enlarged the meaning and extension
of this concept by including new dimensions in its
de?nition, like for example the nature of the
knowledge used to carry out a speci?c task (Min-
tzberg, 1979; Williamson, 1985, 1993; Grandori,
1997). Therefore, in this study, these additional
dimensions are incorporated into the analysis
of knowledge-intensive ?rms and the nature of
uncertainty is assumed to be much more complex
and ?ne-grained than the traditional models allow.
Uncertainty is seen as being related to the speci?c
characteristics of the knowledge applied to work
activities and determining the way in which knowl-
edge is transferred and controlled. As a consequence,
the understanding of its impact on management
control systems needs to be further explored.
In addition, this paper, in contrast to previous
contributions in the literature, assumes that one of
the relevant roles of management control systems
is to coordinate activities by integrating di?erent
sources of knowledge expertise instead of simply
to either supply information to deal with uncer-
tainty (e.g. Khandwalla, 1972; Gordon & Nar-
ayanan, 1984; Simons, 1987; Davila, 2000) or to
reduce goal divergence (Ouchi, 1979; Vancil,
1979). This alternative perspective may help re-
interpret the existing contradictory empirical
results related to the control of activities char-
acterized by di?erent levels of uncertainty.
Given these premises, the objective of this paper
is to develop a conceptual model of the role of
management control systems in knowledge-inten-
sive ?rms by considering a new variable that
expresses the level of uncertainty, here called
knowledge complexity. More speci?cally, the aim
is to analyse the way in which knowledge com-
plexity a?ects coordination and knowledge inte-
gration and, in turn, management control systems.
The relationships of these variables are examined
in the realm of knowledge-intensive ?rms’ teams
(considered as the level of analysis), where
the greatest problems of consistency between the
coordination and knowledge integration modes
occur. In this way this study addresses the issues
associated with the manner in which uncertainty is
de?ned and provides a more complete under-
standing of how knowledge activities are con-
trolled and integrated. Based on a review of the
literature on management control systems,
the likely in?uences of the type of uncertainty on the
use of di?erent control mechanisms are explored.
For this purpose we consider a wide range of tools
available for co-ordinating knowledge-intensive
activities by examining the use of result, action
and personnel/cultural controls (Merchant, 1985).
The present contribution has a number of dif-
ferent purposes. The general aim is to advance the
understanding of knowledge-intensive ?rms by
analysing the contingency features of the manage-
ment control systems of their teams. In addition,
by introducing a new variable (knowledge com-
plexity), new insights concerning the relationship
between uncertainty and management control
systems are provided. Finally, because the knowl-
edge-intensive setting is neither particularly well
understood, nor extensively researched in the ?eld
of management accounting, exploratory case
studies are used to illustrate the arguments.
The remainder of the paper is organized into
four sections. First, drawing together existing
strands of research on knowledge-based organiza-
tions and on knowledge in organizations, the rele-
vance of knowledge integration and the role of
teams in knowledge-intensive ?rms are presented.
Second, by means of a brief review of the relevant
prior research, the revised concept of uncertainty
and its implications for the design of management
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 403
control systems are illustrated. Third, the metho-
dology and data analysis are described and the
results shown. The last section is dedicated to
presenting conclusions and an agenda for future
research.
Knowledge-intensive ?rms and knowledge
integration: the role of teams
The analysis of the literature on knowledge-
intensive ?rms reveals a remarkable heterogeneity
between the various contributions that have
appeared in books, reviews and journals. These
contributions di?er in terms of theoretical
approaches and empirical investigation in a way
that prevents a homogeneous framework and the
generalisation of empirical results. For example,
Winch and Schneider (1993) and Starbuck (1993)
examine respectively the strategic management
issues facing the architectural practices and the
elements of exceptional success in a law ?rm. Ber-
nardi and Warglien (1989), Ekstedt (1989) and
Starbuck (1992) analyse the processes of learning
and knowledge renewal. But, while the former
study a research centre and a computer ?rm, the
latter analyses mainly professional service organi-
zations. Furthermore, other contributions exam-
ine speci?c management issues concerning
knowledge-intensive ?rms. In particular, Alvesson
devotes his attention to both the cultural-ideolo-
gical modes of control (Alvesson 1993a), and the
relevance of social identity and loyalty in prevent-
ing the (unwanted) exit of personnel (Alvesson,
2000).
4
Knowledge-intensive ?rms have been de?ned in
di?erent ways by the various contributions as:
?rms that use, more than the average, employees
in ?elds that require a sophisticated knowledge
and whose expertise is the source of a competitive
advantage (Bernardi & Warglien, 1989; Ekstedt,
1992; Winch & Schneider, 1993); ?rms ‘‘in which
. . . experts are at least one-third of the personnel’’
and experts are ‘‘those with formal education and
experience equivalent to a doctoral degree’’ (Star-
buck, 1992). Thus, in general, according to these
de?nitions, knowledge-intensive ?rms’ capital
consists predominantly of human capital, their
critical elements are in the minds of individuals
and heavy demands are made on the knowledge of
those who work in them (Ekstedt, 1989, p. 7).
Alternatively, such a type of ?rms also has been
characterized as those that deploy their ‘‘assets in
a distinctive way, for they sell a capacity to pro-
duce, rather than a product’’ (Winch & Schneider,
1993, p. 923) and ?nally those that process what
they know into unique knowledge products and
services for their customers, or possibly goods in
combination with services. They are less capital-
intensive than companies in the manufacturing
industries and more learning-intensive than those
operating in other service industries (Nurmi,
1998).
The diversity of perspectives and de?nitions
provided in the literature makes the concept of
knowledge-intensive ?rm—and the related notion
4
There is also an alternative institutional-constructivist
perspective to study knowledge-intensive ?rms. According to
this approach, ‘Knowledge-intensive ?rms can be viewed as
providers of institutionalised myths’ (Alvesson, 1993b) because
they incorporate ambiguities and uncertainties involved in their
work and results. As a consequence knowledge-intensive ?rms
have to put a great e?ort, both internally and externally, to
emphasise, for employees as well as for customers and other
actors, that their experts should be relied upon. In addition,
according to this perspective, an aspect that di?erentiates
knowledge-intensive ?rms from non-knowledge-intensive ?rms
is ‘the degree of elaboration of the language code through
which one describes oneself, one’s organization, regulates cli-
ent-orientations as well as identity’ (Alvesson, 1993b, pp. 1007).
Furthermore, knowledge plays roles such as (a) a means for
creating community and social identity through o?ering orga-
nizational members a shared language and promoting their
self-esteem; (b) a resource for persuasion in, for example, public
relation work and interactions with customers; (c) providing
the company with a pro?le (an intended image targeted at the
market); (d) creating legitimacy and good faith regarding
actions and outcomes; and (e) obscuring uncertainty and
counteracting re?ection. With these knowledge roles in mind,
management is much more a matter of in?uencing employees
on a broader scale, including securing and developing work and
organizational identities (Alvesson, 1993b). However, this
institutional-constructivist perspective is not considered here.
This is because in this paper we use a functionalist approach
that considers the utility of management control systems in
achieving purposeful outcomes and ‘caution must be directed at
any approach providing some uni?cation between functionalist
and ‘‘alternate’’ approaches’ (Chenhall, 2003, p. 160).
404 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
of knowledge
5
—problematic.
6
In brief, it is di?-
cult to characterise knowledge-intensive ?rms as a
distinct, uniform category. The di?erence between
knowledge-intensive ?rms and other companies is
not self-evident because all organizations involve
knowledge. In addition, it becomes even more
opaque if these ?rms are substantiated with refer-
ence not only to formal, science-based knowledge
but also to other more embedded and encultured
versions of it (Blackler, 1995; Alvesson, 2000).
Nevertheless, there are many crucial di?erences
between many professional service and con-
sultancy companies on the one hand, and more
routinized service and manufacturing ?rms on the
other. It is therefore useful to refer to knowledge-
intensive ?rms as a vague but meaningful cate-
gory. The category per se does not lead to a pre-
cise de?nition or delimitation, as in any case often
happens in social science; it includes organizations
which are neither unitary nor unique. Yet, it draws
attention towards phenomena that are beyond the
single case without aspiring to describe organiza-
tions in general, loosely focusing on an organiza-
tional category about which new insights may be
developed (Alvesson, 2000).
More speci?cally, for the purposes of this paper,
knowledge-intensive ?rms are viewed as organiza-
tions that use mainly the knowledge of their indi-
viduals to develop and trade immaterial responses
7
to customer requirements. The one feature such
?rms possess is that their expertise is used to solve
varied problems by o?ering a di?erentiated range
of innovative responses to customers (Ekstedt,
1989; Starbuck, 1992). In addition, their knowl-
edge is mainly embedded in human capital, even if
this knowledge may be partially institutionalised
and localised at the organisational level in the
form of collective frames of reference, systema-
tised methods of work, sophisticated routines and
processes (Starbuck, 1992; Alvesson, 1995; Morris
& Empson, 1998).
Knowledge-intensive ?rms have become more
prevalent and more important as the business ser-
vices sector has grown equally over the last twenty
years (Winch & Schneider, 1993) and the world
has been moving toward the so-called ‘‘post-
industrial’’ economy (Drucker, 1993; Nonaka,
1994). Yet, research has only just started to
scratch the surface in this area of business and
most of the existing writings have suggested sim-
plistically that managing these organizations is
mainly based on both attracting and keeping the
key professional workforce—the most signi?cant
‘resource’ of knowledge-intensive companies—and
developing organization-speci?c knowledge of an
informal nature, inscribed in organizational cul-
ture and a certain style of working (Maister, 1982;
Alvesson, 2000).
However, the management of knowledge-inten-
sive ?rms is de?nitely more di?cult than sug-
gested. This is because knowledge-intensive ?rms
need not only to attract the right individuals with
the right expertise, but also to integrate the
knowledge of those recruited in order to carry out
activities mostly characterized by uncertainty,
knowledge asymmetries and resulting observa-
bility problems (Winch & Schneider, 1993; Austin
& Larkey, 2002). In fact, uncertainty comes from
the fact that work in this type of ?rm tends to be
oriented toward innovation and problem solving,
and may require e?orts on dimensions that are
unanticipated and whose criticalness often evolves
dynamically (Ekstedt, 1989). In addition, activities
generate pronounced asymmetries—based not only
on information, but also on knowledge disparities—
between a manager and those (s)he manages. A
manager who has the same information as a
5
An analysis of the concept of knowledge is outside the
scope of this study. Yet, in the discussion that follows we are
consistent with the de?nition provided by Morris and Empson
(1998) when they say that ‘knowledge is viewed as information
which professionals acquire through experience and training,
together with the judgement which they develop over time
which enables them to deploy that information e?ectively in
order to deliver client service. Thus, knowledge is not limited to
technical or product based expertise (professional know-how as
Svieby & Lloyd, 1987, call it) but may also be knowledge of
clients or industries and how they operate (managerial know-
how). In turn, knowledge takes particular forms as it accumu-
lates over time depending on the historical development of the
?rm (Dodgson, 1993)’.
6
For a review of the di?culties related to the de?nition of
this concept see, for example, Starbuck (1992) and Alvesson
(1992, 1995).
7
These solutions are immaterial in that they refer to value-
creating transformations occurring in the realm of ideas or
symbols, or, alternatively, in which a substantial amount of
productive activity is intellectual rather than physical (Austin &
Larkey, 2002).
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 405
worker can still lack the expertise needed to
understand, attribute, evaluate, and act on what
(s)he observes. This leads to observability pro-
blems that are particularly severe and persistent in
knowledge-intensive ?rms due to the immaterial
nature of their activities (Austin & Larkey, 2002).
Moreover, the same problems of uncertainty,
asymmetry and observability occur not only in the
manager-subordinate relationship but also with
reference to individuals possessing di?erent kinds
of expertise and operating together at the same
level. To reduce uncertainty and overcome both
asymmetry and observability problems, they need
to integrate the many types of knowledge they
have (Grant, 1996; Boland & Tenkasi, 1995; Aus-
tin & Larkey, 2002). Integration of knowledge
helps overcome uncertainty and reduce knowledge
disparities making visible what the other indivi-
duals think and do (Boland & Tenkasi, 1995).
The literature on the topic has mainly analysed
only one way of integrating knowledge, by means
of formal knowledge transfer (e.g. Kay, 1979;
Levitt & March, 1988; Boisot, 1995), and has
made only limited progress in dealing with this
issue when most of the relevant knowledge is tacit
(Polanyi, 1966). Some notable exceptions to this
void are represented by Nonaka (1994) who
describes the conversion of tacit into explicit
knowledge (and vice versa) and other authors who
emphasize the role of codi?cation and di?usion
processes in making tacit knowledge visible and
available (Boisot, 1998).
However, the knowledge-based theory of the
?rm has suggested that knowledge transfer is not
always an e?cient approach to integrating knowl-
edge. According to the same theory, when work
requires the combination of many individuals’
specialist knowledge, the key to e?ciency is to
achieve e?ective integration while minimizing
knowledge transfer through cross-learning by
organizational members. Such processes of cross-
learning takes place within teams that directly
involve individuals (Lave & Wenger, 1991, Brown &
Duguid, 1991; Cohen, 1993; Mohrman, 1993; Wen-
ger, 1998; Scott & Tiessen, 1999; Easterby-Smith,
Crossan, & Nicolini, 2000; Ancori et al., 2000).
Teams create synergy to increase the integrated
application of specialized knowledge, so that the
performance of the whole is greater that the sum
of its parts (Cohen, 1993). In fact, complex tasks
require the search for and evaluation of various
alternatives, and this is likely to require more
information than is possessed by a given indivi-
dual (Bamber & Bylinski, 1982; Ismail & Trot-
man, 1995; Scott & Tiessen, 1999). This additional
information may be provided by groups which
generally should have greater collective knowledge
than individuals (Burton, 1987; Hare, 1976; Stein,
1975; Taylor, Berry, & Block, 1958; Yetton &
Bottger, 1982; Scott & Tiessen, 1999). Prior
research suggests that groups bene?t from the
combination of di?erent backgrounds, compe-
tencies and perspectives of their members (Ismail
& Trotman, 1995; Shaw, 1976). Even if one indi-
vidual has more knowledge and experience, the
speci?c knowledge of the less informed individuals
can integrate the knowledge of the group as a
whole (Burton, 1987; Maier, 1970). The group can
combine the knowledge of di?erent group mem-
bers and reach judgements of higher quality as a
result of exchanging and sharing information and
perspectives (Casey, Gettys, Pliske, & Mehle, 1984;
Stocks & Harrell, 1995). Accordingly, teams which
have this knowledge integrative function tend to
be widely used in complex environments requiring
the use of di?erentiated knowledge and the provi-
sion of novel and multidimensional solutions
(Grant, 1996, 1997; Scott & Tiessen, 1999).
8
These conclusions are especially appropriate for
knowledge-intensive ?rms. In fact, as argued in the
literature (Boland & Tenkasi, 1995), integration
8
Nevertheless, even within team-based processes, hierarchy
is still necessary in order to link di?erent sub-systems (e.g. the
various project teams) together. The principles of hierarchical
design are fundamental to ‘modular’ design in organizational
structures. Critical to the design of modular structures is the
separation of the total system into a number of modular sub-
systems and then to design and standardize the interfaces
between these sub-systems. A key distinction here is between
the ‘component knowledge’ required by the sub-systems and
the ‘architectural knowledge’ required for the linking of the
various sub-systems (Grant, 1997, p. 453). Knowledge-intensive
?rms include both kinds of knowledge. In this paper, however,
we mainly devote attention to the ‘component knowledge’
related to project teams considered as the main locations where
knowledge activities take place.
406 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
through teams is particularly suitable to deal with
a dynamic environment that requires fast and
innovative responses (Cohen, 1993; Scott & Ties-
sen, 1999),
9
as is the case of knowledge-intensive
?rms (Boland & Tenkasi, 1995). In this type of
environment, teams allow a process of distributed
cognition in which multiple communities of spe-
cialised knowledge workers, each dealing with a
part of an overall organizational problem, interact
to create the patterns of sense-making and beha-
viour displayed by the organization as a whole.
This distributed cognition is necessary because the
critically important processes and the diversity of
environments and technologies to face are too
varied and complex for an individual to under-
stand in its entirety (Boland & Tenkasi, 1995).
Yet, as suggested in the introduction, the issue
of control has received only fragmented and
unsystematic attention in the knowledge-intensive
?rms’ literature. Few contributions have been
focused on the questions related to what controls
are used in the teams of knowledge-intensive ?rms
and on how these controls guarantee the necessary
integration of knowledge.
10
This study has the
objective to ?ll this void.
The main research questions are: How are
knowledge-intensive ?rms’ teams controlled? And
how does uncertainty a?ect management control
systems? Additional questions are: How does
uncertainty in?uence knowledge integration? How
do management control systems act as knowledge
integration mechanisms?
To be informed as much as possible by the exist-
ing writings, we start our explorations of the
research questions by reviewing the published
management control literature. We then analyse
these questions with a ?eld study of three software
development project teams in order to provide
some tentative explanations on the existing man-
agement control systems di?erences. We conclude
with an agenda for future research.
Uncertainty and management control systems in
knowledge-intensive ?rms
Literature review: uncertainty and management
control systems
Contingency-based research has a long tradition
in the study of management control systems
(Hayes, 1977, 1978; Tiessen & Waterhouse, 1978;
Otley, 1980; Hopwood, 1989; Chapman, 1997;
Chenhall, 2003). Researchers have attempted to
explain the e?ectiveness of management control
systems by examining designs that best suit the
nature of the context in which they operate.
Perhaps the most widely researched aspect of
the context is uncertainty (Chenhall, 2003) and its
relevance has been rea?rmed recently by many
commentators who have stressed its fundamental
role in the management control systems design
(Chapman, 1997; Hartmann, 2000; Chenhall,
2003).
In the literature, the concept of uncertainty has
been mainly associated with environment and
technology (e.g. Burns & Stalker, 1961; Wood-
ward, 1965; Lawrence & Lorsch, 1967; Thompson,
1967; Perrow, 1967). More speci?cally, on the one
hand, environmental uncertainty has been cap-
tured in terms of dynamism, heterogeneity (Gor-
don & Miller, 1976; Amigoni, 1978), predictability
(Waterhouse & Tiessen, 1978), controllability
(Ewusi-Mensah, 1981) and equivocality (Daft &
Macintosh, 1981). On the other hand, technology
has been expressed with reference to complexity
(Woodward, 1965), task uncertainty (Perrow, 1967;
Ouchi, 1979) and interdependence (Thompson,
9
It is important to notice here than ?rms can use various
mechanisms to deal with dynamic and complex environments
(Galbraith & Kazanjian, 1986). One possibility is to use infor-
mal mechanisms such as co-location, the establishment of
interpersonal networks, or rotation of individuals (Mohrman,
1993). Another possibility is through more formal responses
like teams, which serves to integrate di?erent elements and
foster innovation within organisations (Scott & Tiessen, 1999).
The objective of this paper is to analyse how the success of
teams is related to the use of various types of controls and not
to evaluate the relative e?cacy of alternative mechanisms.
10
In this paper we investigate only teams that are formal,
secondary to an individual’s department in the organisation’s
structure, and that can exist entirely within a department (intra-
departmental) but especially across departments (inter-depart-
mental). They are often established to solve a particular prob-
lem or meet a speci?c objective.
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 407
1967).
11
These three latter concepts are normally
considered as separate constructs but they are
common because they include some overlapping
themes concerning uncertainty. It seems likely that
conversion of inputs into outputs within less com-
plex, mass production technologies is more pro-
grammable and predictable than in job or batch
styled technologies servicing customized products
(Chenhall, 2003).
Typically, the arguments supporting uncer-
tainty-related expectations have not always dis-
criminated clearly between these various types and
sources of uncertainty leading to some indistinct-
ness of the di?erent dimensions of the concept
(Hartmann, 2000; Chenhall, 2003). As a result,
while the relevance of this construct is still widely
recognized in extant management control systems
research, its ambiguity has been advocated as one
of the main reasons for the vagueness in capturing
the impact of uncertainty over management con-
trol systems.
12
This ambiguity may produce even
more severe problems in understanding if the same
variables are applied to completely di?erent con-
texts, without any speci?c caution.
In this respect, it is worth noting that many of
the dimensions related to uncertainty have been
developed with reference to the physical produc-
tion setting [see for example Woodward (1965)],
and that many of the contingency-based manage-
ment control systems contributions have con-
sidered manufacturing organizations (Chenhall,
2003). Yet, the extent to which the same dimen-
sions of uncertainty maintain their relevance in
other more knowledge intensive contexts is ques-
tionable. It is doubtful, for example, that two of the
most common variables used to describe uncer-
tainty—task analysability/programmability and the
number of exceptions—represent discriminatory
dimensions to study management control systems
in knowledge-intensive ?rms. In fact, in such type
of ?rms, on the one hand, analysability is prevented,
in part, by the lack of observability of action and,
on the other hand, the number of exceptions is by
de?nition high due to the high level of innovation
involved in executing tasks. In addition, knowl-
edge-intensive ?rms are characterized by a more
complex and multidimensional environment. As
compared with the physical activities of manu-
facturing organizations, knowledge-intensive pro-
cesses involve an extra dimension of uncertainty
because they require the application of a wide
range of di?erentiated knowledge that needs to
be integrated e?ectively to produce appropriate
responses to customers’ needs. As already suggested,
this integration mainly takes place within teams.
Therefore, to study the impact of uncertainty over
management control systems in knowledge-inten-
sive ?rms, the nuances of the project teams and
the dimensions of knowledge need to be better
expressed. More speci?cally, advances could be
made by giving a speci?c theoretical meaning to
the concept of uncertainty when applied to tasks
characterised by knowledge intensity.
The purpose of the section that follows is to
provide a concept of uncertainty that takes into
consideration the speci?cs of knowledge-intensive
?rms’ teams and to suggest a framework that
represents the impact of this variable on both the
coordination and knowledge integration modes,
and in turn on management control systems.
Dealing with knowledge complexity in knowledge-
intensive ?rms’ teams: the role of management
control systems as knowledge integration
mechanisms
There is little understanding of how knowledge-
intensive ?rms control their teams. This problem is
11
The review is selective and illustrative of issues pertinent to
the development of a contingency-based framework for the
design of management control systems in knowledge-intensive
?rms, and does not attempt a comprehensive coverage of rele-
vant research. In addition, the management control systems
contingency-based literature has been already successfully
reviewed by many authors, recently (e.g. Chapman, 1997;
Lang?eld-Smith, 1997; Hartmann, 2000; Chenhall, 2003).
12
Other reasons for the unfruitfulness of results are related,
as many commentators have suggested (Briers & Hirst, 1990;
Chapman, 1997; Hartmann, 2000), to the theoretical ?aws in
the management control systems contingency-based paradigm.
The particular claim is that this ?eld of research has come to
overemphasize statistical sophistication to the detriment of
construct and theory development. Critique has also been
directed at both research models and methods for data collec-
tion and analysis (Otley, 1980; Schoonhoven, 1981; Chapman,
1997; Hartmann, 2000). Yet, the notion of management control
systems contingency-based theory has never been really chal-
lenged (Young, 1996).
408 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
particularly multifaceted in such a type of organi-
zation because the selection of control mechan-
isms should take into consideration at the same
time both the coordination and the knowledge
integration aspects. A framework is needed to
explore how the various forms of control act as
complex knowledge integrative devices and how
the contextual factors a?ect their selection and
e?ectiveness.
13
In knowledge-intensive ?rms, one of the dis-
tinctive factors potentially a?ecting control system
choices is the nature of knowledge and its com-
plexity. Wood (1986) discusses three aspects of
task complexity that are strictly interlinked to
knowledge complexity: component complexity,
coordinative complexity and dynamic complexity.
The component complexity of a task is determined
by the number of distinct information cues that
must be processed and the number of distinct acts
that must be executed in performing the task. The
coordinative complexity is the result of the form
and strength of the relationships between infor-
mation cues and acts, including the content, timing,
frequency, and location requirements for perfor-
mances of demanded acts. Finally, the dynamic
complexity relates to the need to adapt to changes
occurring in the cause-e?ect relationships or
means-ends chain during the execution of the task.
The task is a?icted by component complexity
when knowledge is characterised by computa-
tional complexity, arising from the high number of
agents and activities, and their interconnections
(Grandori, 1997; Simon, 1962). In this case, orga-
nizational economic analyses and classic organi-
zation theory suggest the extension of the
codi?cation and formalization of information by
using formal processing supports to manage oper-
ations (Galbraith, 1977; Mintzberg, 1979; Grandori,
1997). The management literature recommends
the di?usion of codi?cation as a mechanism to
integrate the di?erent pieces of knowledge (Cowan
et al., 2000). Obviously, this is possible only when
knowledge is characterized by codi?ability, that is
to say the possibility to formally articulate it in
documents and software. This knowledge may be
substantive, e.g. in blueprints, or it may be proce-
dural, e.g. in a recipe for carrying out a task
(Simon, 1979; Winter, 1987; Zander & Kogut,
1995; Cohendet & Steinmuler, 2000).
On this basis, we predict that when a task
involves computationally complex knowledge it
will be regulated by codifying procedures, actions,
rules and instructions in order to both coordinate
e?ciently the complementary activities of many
agents by adopting an action oriented approach to
control and, at the same time, guarantee the
e?ective integration of knowledge.
In contrast, a task may be subject to coordina-
tive complexity, which is the result of using
knowledge involving technical complexity, de?ned
as the number of distinctive skills, or compenten-
cies belonging to many di?erent (groups of)
experienced people (Cohen & Levinthal, 1990;
Iansiti & Clark, 1994; Zander & Kogut, 1995).
The more knowledge is di?erentiated among
agents, the more it fosters interactions and triggers
mechanisms for knowledge integration due to the
multiplicity of ways the problems are perceived
and dealt with. In this case, integration is not
achieved by the transmission of tacit knowledge
(and by its formalisation) but through its coordin-
ation aimed at pursuing a common objective. In
other words, in a context of diversi?ed knowledge,
the ‘constraint’ of tacit knowledge can be solved
by means of coordination mechanisms more than
codi?cation processes (Grant, 1996; Cowan et al.,
2000). In addition, coordination of actions is
e?ciently achieved by means of joint problem
13
Knowledge-intensive ?rms can use di?erent mechanisms
to control their teams. One e?ective and complete way to clas-
sify these mechanisms, according to the object of control, is by
distinguishing between result, action and personnel/cultural
control (Merchant, 1998). In the result oriented approach,
control is mainly exercised via setting targets, reporting
achievements, accountability, and reward structure that serve
to foster output-directed behaviour, whereas control of action
is based primarily on procedure guides, operating manuals,
codi?cation of actions and supervising observance of rules and
instructions. On the contrary, personnel controls consist of
helping individuals perform their tasks by building on the nat-
ural tendency to control themselves and cultural control repre-
sents a set of values, social norms and beliefs that are shared by
members of the organization and that in?uence their actions. It
is based on the belief that by fostering a sense of solidarity and
commitment towards organizational goals individuals can
become immersed in the interests of the organization (Mer-
chant, 1998; Ouchi, 1979). The di?erent controls tend to be
usable in di?erent contexts (Merchant, 1998; Ouchi, 1979).
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 409
solving, di?erentiation of decoupled specialized
sub-systems and output exchange (Grandori,
1997, 1999), and needs to be further reinforced
when the di?erent specialized sub-systems are
characterised by a diversity of interests, leading to
potential competitive and hostile uses of knowl-
edge (Ouchi & Bolton, 1988; Grandori, 1999).
Therefore, we predict that a task involving
technically complex knowledge requires the use of
result oriented control mechanisms hinging pri-
marily on setting objectives and monitoring
achievements with reference to content, timing,
frequency, and location of required outputs. This
allows at the same time an e?cient control of
action and an e?ective integration of knowledge.
Finally, in some other situations, the main fea-
ture of the task is dynamic complexity and
knowledge is characterised by cognitional com-
plexity. The processes are either new for the agents
involved or entail innovative problem solving and
are subject to many possible ‘serendipities’ and
unexpected outcomes. This implies the discovery
of cause-e?ect relations and relevant goals, and
the transformation in explicit knowledge is expec-
ted to fail (Grandori, 1997; Perrow, 1967; Burns &
Stalker, 1961). Therefore, such form of knowledge
is communicated between individuals by means of
common history, shared experiences and collective
social and organizational frames—if these indivi-
duals have the time, the occasions for socializa-
tion, and the broader institutional incentives to
perceive the game as basically integrative (Grand-
ori, 1997; Ouchi, 1980; Ouchi & Bolton, 1988). In
addition, in this setting coordination is achieved
mainly by means of ‘professional or collegial’
structural arrangements (Perrow, 1970).
Thus, we predict that tasks that use cognition-
ally complex knowledge require regulation
through self and group controls achieved with the
use of selection and training policies which ensure
that individuals have been exposed to appropriate
training and socialization processes. In this way,
by fostering a sense of community both control
and knowledge integration are in place.
In brief, what is hypothesized here is that dif-
ferent types of knowledge complexity give rise to
di?ering relations between reliance on each of the
three kinds of controls, and management perfor-
mance (see Fig. 1). Yet, in many instances, there
will be no radical choice between the various
forms of control because the task involves many
pieces of knowledge, each characterised by a spe-
ci?c complexity, and more complex integrative
devices need to coexist to assist control.
Tentative theoretical propositions
The preceding discussion can be reported in
brief in the form of the following tentative theo-
retical propositions, which are, of course, subject
to re?nement:
P1: A context involving knowledge character-
ized by computational complexity will tend to
be regulated through action oriented controls
aimed to coordinate and guarantee the integra-
tion of knowledge.
P2: A context involving knowledge subject to
technical complexity will tend to be regulated
through result oriented control mechanisms
allowing coordination and integration of
knowledge.
Fig. 1. Knowledge complexity, knowledge integration and management control systems.
410 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
P3: A context involving knowledge character-
ized by cognitional complexity will tend to be
regulated by means of personnel/cultural forms
of control leading at the same time to coordi-
nation and knowledge integration.
Comparative case studies: management control
systems in three software development project
teams
To explore the tentative research propositions
discussed above and, more generally, the whole
area of control system choice in knowledge-inten-
sive ?rms’ teams, we solicited the cooperation of a
multinational software ?rm in the UK.
14
Research design
15
Because the existing contributions on manage-
ment control systems in knowledge-intensive ?rms
are still quite limited, I decided to do a case study
as the preferred methodology to build knowledge
about the phenomenon (Yin, 1988). In addition,
as management control systems vary across ?rms
of di?erent industries and the organization’s hier-
archical levels—and this variation in the research
setting can be detrimental to the power of the
research design—two speci?c decisions have been
made: the ?rst was to limit the study to a ?rm
belonging to the software industry; the second was
to select the project team as the unit of analysis,
because it is in the project teams that ideas are
transformed in executable software applications
through the use of di?erentiated knowledge and
where the greatest problems of knowledge inte-
gration and control occur.
Initial contact with the ?rm was through its
Chief Financial O?cer, with whom the purpose of
the study was discussed and agreement to partici-
pate secured.
Data were collected through a combination of
document analysis, direct observation and sys-
tematic interviewing. The analysis of documents
spanned from the annual reports, to budget and
reporting statements, to the speci?c project doc-
umentation. Observation was carried out within
three software development projects in which I
spent 20 days full time over a period of two
months, analysing meetings, practices and docu-
ments. Interviews were administered to managers
at di?erent levels (15 h at the top management
level and more than 20 h at the project team level).
Interviews were structured around a set of ques-
tions about the management control systems and
the software development process itself. The
questions were open-ended in order to adapt the
interview to the expertise of each interviewee
without losing the overall direction. An average of
six to seven people (managers, leaders and devel-
opers) were interviewed in each project team. The
use of multiple informants and sources of evidence
allowed for the triangulation of data.
The evidence collected was repeatedly analysed.
At the very beginning it seemed that transcripts
and notes were incoherent. But di?erent ways of
organising the material were considered and a
pattern began to emerge. At each phase, the
interpretations of the data collected were checked
against the next round of data, in an attempt to
verify the level of understanding gained. This pro-
cess continued until succeeding data started to
become predictable. All the data collected and
their interpretation became part of a report dis-
cussed with the Chief Financial O?cer. This dis-
cussion helped verify the correct understanding of
the practices adopted in the ?rm and challenge the
corresponding interpretations.
Next, I provide a description of three illustrative
software development project teams dealing with
knowledge complexity and the design and use of
management control systems. The primary criter-
ion for project teams selection was to ensure that
the variety of knowledge complexity and control
patterns would emerge. The paper concludes with
14
The name of the ?rm has been obscured to preserve con-
?dentiality.
15
The case study is based on data gathered as part of a much
broader research study into the management control systems in
knowledge-intensive ?rms carried out during my Ph.D thesis.
In the course of the study, a broad cross-section of managers
and professionals were interviewed in the ?rm, from the chief
?nancial o?cer to the most junior, least experienced sta?.
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 411
an analysis of the similarities and di?erences in the
software development control systems and some
tentative theoretical conclusions.
The software development process
‘Computer software consists of any set of instruc-
tions and data which is read, interpreted and exe-
cuted by the control units of a computer system’
(Grindley, 1988). It can be distinguished in three
broad classes: operating systems, the basic software
controls of the operations of a computer, including
network controllers and compilers; application
tools, which are software tools that support speci?c
applications in software engineering or database
management, among other areas; and applications
solutions, which are software instructions that
enable a computer to perform speci?c tasks of
interest to the end user, such as accounting and
word processing. All these three types of software
can be provided in either ‘standard’ (packaged
software) or ‘custom’ form (Mowery, 1996). It is,
however, only the control of the development of
this latter kind of software that is analysed here.
The process of software development is struc-
tured around a well-de?ned sequence of phases.
The ?rst phase is typically represented by the
requirement speci?cation phase to indicate the
complete, validated set of required functions,
interfaces and performance outcomes for the soft-
ware product. The outcome of the initial phase is a
project plan with a description of the software
characteristics and project speci?cations: techno-
logical performance, customer interfaces, target
costs, and organizational resources. The second
phase- design- goes into more detail to specify the
overall hardware-software architecture, control
structure and data structure for the software pro-
duct, along with such other necessary components
as data user’s manuals and test plans. The third
phase—software development—is the actual
development of the set of instructions in terms of
lines of code. It is in this phase when trade-o?s get
resolved and information is transformed into an
executable programme. The last two phases—test-
ing and implementation—con?rm that the software
meets the requirements and the customer’s needs.
Yet, even if described as apparently linear, the
process is iterative in its nature: software speci?-
cations or even the design can be re-evaluated in
light of new information generated throughout the
process (Rook, 1986; Bohem, 1988; Berso? &
Davis, 1991).
Project team A
The ?rst project analysed was a ?xed-price pro-
ject to develop software for the exploitation of
meteorological satellites in space division to pro-
vide a system which enables the generation and
enhancements of meteorological products from
satellite data.
Because the project was a re-run of a previous
similar project with the same client, few doubts exis-
ted regarding technology, client and management.
The software is developed on the basis of the
previous experience and uses a lot of the ideas
developed in the preceding project (Team leader
of product quality monitoring).
In terms of technology there is nothing new
(Team leader of the monitoring control and
communication on-line subsystem).
Most of the ideas are the same, even if we have
to write everything from scratch. There is a lot of
use of experience. The advantage is also that the
client is known (Team leader of algorithms).
The source of uncertainty came mainly from the
project scope. Evidence of this emerges with
reference to the duration of the project (24
months), the number of people involved (30), the
number of project requirements (5000 classi?ed in
three di?erent categories: the facilities, the algo-
rithms, and the general ground requirements), and
?nally the wideness of project documentation
(5000 pages only for the architectural design
phase). The project is therefore characterized by
what we have called computational complexity,
due to the number of people, activities and inter-
connections that need to be coordinated.
Due to its low level of newness, the project team
possessed su?cient knowledge to decide in
advance the way in which activities were to be
412 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
executed to achieve the objectives. The action and
knowledge of many agents through a long period
of time needed to be coordinated and integrated
by means of e?cient information carriers. To
achieve these objectives, the management control
system was authoritative and prescriptive in nat-
ure, centred around action, featuring rules of
behaviour and focusing on compliance with pre-
speci?ed norms and plans.
In fact, the authoritative nature of the project
was witnessed by a clear de?nition of lines of
responsibility between the project manager, sup-
port sta? (quality engineering, hardware procure-
ment, computer system, system design), team
leaders (algorithms, on-line systems, o?-line sys-
tems, product quality monitoring, on-site support
and assembly-integration-test) and developers.
The structure of the project is de?nitely hier-
archical (A developer of the product quality
monitoring team).
The structure is hierarchical. There is a clear
de?nition of objectives, responsibilities and
delegation (The team leader of monitoring
control and communication on-line subsystem).
Interviewees reported that control was exerted,
on a regular basis, around activities which were
weekly assigned to developers and monitored by
speci?c metrics, such as the number of modules
completed, the time spent on the modules, the time
to go, the daily number of written lines of code
and so on. These various pieces of information were
communicated and integrated constantly in formal
written documents and weekly formal meetings.
The detail of action control emerges by con-
sidering the words of some project members, who
stressed the role of detailed plans (specifying time,
modules and lines of codes) and the continuous
monitoring of productivity by counting the num-
ber of lines of code generated:
At the end of the architectural design phase
there was a plan with the number of modules
and information about how long it would take
to produce one. Then the modules were assigned
to the di?erent members of the team. Every
week the project manager monitored which
module was ?nished, how long it would still take
and so on. At the end of the detailed design
phase there was a rede?nition of the plan (The
team leader of product quality monitoring).
In a project like this we thought we would be
able to achieve twenty lines of code per day and
during this phase we have achieved forty. So
the productivity has been higher (The project
manager).
We started thinking that it was 260,000 lines of
code, then gradually things changed over the
project. For example, the algorithm team had
overestimated the number of lines of code and
the on-line and o?-line teams had under-
estimated them. We ended up with more or less
260,000 fortuitously (The project manager).
As for the control of the work, the team leader
reviews everything we do on a regular basis not
only at the end, but incrementally. There is also
a weekly monitoring concerning the time spent
on a speci?c model and the time to go (A
developer of the team of monitoring control and
communication on-line sub-system).
In addition, the observation of the project showed
that the same action control mechanisms (plans,
schedules, forecasts, policies and procedures, and
standard information and communication systems)
were also the main channels for knowledge inte-
gration. Knowledge was integrated mainly through
the codi?cation of action in documents embodying
both the content and the processes inherent in the
application of technical knowledge. It was cap-
tured in project documentation and was made
accessible to all the members of the project team.
In brief, the project was controlled in detail, on
a continuous basis, by action control mechanisms
which were at the same time means for integrating
knowledge.
Project team B
The second project observed was a tele-
communication division’s mobile billing and
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 413
customer care project for an Israeli mobile opera-
tor. The project was about the customisation of a
speci?c product that was produced by another
company. The product aimed to manage all the
problems linked to the calls of a mobile phone, the
invoices, the payments, and so on. Therefore,
the kernel of the product needed some interfaces
with other systems to manage all these activities.
For these reasons, the ?rm had to integrate the
activities of di?erent work groups (sub-con-
tractors), each providing software that formed
part of the total solution.
The most important phase of the project will be
the integration phase (The delivery manager).
Each work group was characterised by speci?c
knowledge and competencies: one work group
knew the product very well and was in charge of
making all the kernel changes; another work
group was dedicated to making changes in the
payment product; the third work group was aimed
to the provisions; and ?nally, the ?rm under ana-
lysis had the knowledge to introduce all the non-
kernel changes. In other words, the distinctive
knowledge and compentencies belonging to the
di?erent work groups represented what has been
here called technical complexity.
Due to the participation of di?erent autono-
mous work groups, the project was characterised,
also, by a very demanding timeframe.
The project is time critical because the various
phases are executed in parallel and there are a
lot of dependencies (The delivery manager).
The strategy adopted for the project implemen-
tation was based on the ?rm’s previous experience
of business support and control systems imple-
mentations.
Interviewees reported that the main task of
control mechanisms was to coordinate the e?ort
of work groups to meet the tight schedule. Each
work group was working autonomously and what
needed to be secured was that the groups’ outputs
were presented in time and according to speci?ca-
tions. For this reason control was centred around
results by using reporting and performance
measurement mechanisms in which the outputs
were presented and potential problems and excep-
tions illustrated.
The control mechanisms help coordinate the
activities and manage the interdependencies of
di?erent groups. The project managers and the
team leaders wanted also to know whether the
outputs described in the plan were achieved,
because a variance could have had an impact on
other activities (The integration test teamleader).
Apart from this, within groups, the work was
left completely free and activities were carried out
by the di?erent individuals without a rigid hier-
archy, with a lot of autonomy and self coordin-
ation guaranteed by direct and informal
communication.
The structure of the project is not rigid. I believe
that developers can go to the project manager if
they want to, and normally they would report
to the line manager. So the developers would
report to the delivery manager and the mem-
bers of my team would report to me. But if
they want they can go straight away for
whatever speci?c matter they have (The
acceptance test team leader).
As far as the team goes I talk to them everyday,
I ?nd what they have been doing and if they
have any problems, questions, technical issues, I
arrange for either ?nd themselves the answer or
I ?nd the answer for them (The acceptance test
team leader).
The observation of the project showed that the
work groups were working autonomously accord-
ing to the detailed objectives that were assigned to
everyone, with a very limited formal and informal
communication in between. In this case knowledge
integration was achieved via exchanging perfor-
mance measures and outputs, embodying the
di?erent group’s expertise, while minimizing
knowledge transfer.
In summary, control was exerted via setting
targets and reporting achievements and perfor-
mance, and was not only a way to coordinate the
414 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
activities of di?erent work groups, but also a
means to guarantee the integration of di?erent
groups’ specialist knowledge.
Project team C
The last project analysed was a ?nance divi-
sion’s initiative aimed to develop a solution for
Visa cash card loading using mobile phones and
short message service system. It was a completely
new project with a lot of innovation. Neither
the technology was known by the company nor the
market in which the product would be introduced.
We look for what are the latest technologies and
how they can be used via us to help our clients
do things better. Technology innovation is the
application of new things to do business better
and the mobile commerce project comes as a
part of this (The technical manager).
The project is innovative because it is in a market
sector in which we have not built a product
before (The technical manager).
With this project, we are in a situation of cog-
nitional complexity because the processes are new
for the agents involved and entail innovative
problem solving leading to unknown outcomes
and potential unexpected exceptions.
At the very beginning the objectives of the pro-
ject were very general and this prevented the pos-
sibility to de?ne speci?c targets to achieve.
The only requirement for the project was to load
cash over the mobile phone. . .that was the total
scope in the requirement at the start of the build
in the project (The technical manager).
The time was not considered as a critical issue.
Therefore, deadlines were self de?ned by the
members of the team and were quite far in terms
of time.
We didn’t have details about when to achieve
results; we were saying to each other ‘Ok I will
do that’ but we did not de?ne formal cut-o?
dates to sort out when we would deliver the
results, when we would complete the design
(The technical manager).
Interviewees reported that the project manager
was not part of the project from the very begin-
ning and when he was involved, he played mainly
a supportive role (reinforcing intentions, giving
advice, questioning decisions, asking for explana-
tions, contributing to solving problems, providing
resources and so on). He was a ‘‘lightweight’’
project manager with no one reporting directly to
him, but only coordinating the development e?ort
of very competent and talented people.
In the ?rst meetings there was no project man-
ager . . . it was a group of peers (The technical
manager).
They (the programmers) produced a very
exceptional program. . . and this was a function
of having a good team . . . The ?rm has a lot of
talented people and part of my job was to turn
talented individuals into a talented team . . .
create the motivation, create the working
atmosphere (The project manager).
My style was to let the technicians get on with
the project . . . therefore I didn’t attend any of
the meetings because having the project man-
ager there I think it would sti?e creativity ‘oooh
my boss is in here I don’t want to appear fool-
ish’ . . . I have given them the job of designing . . .
let them go into the design meetings, be creative
(The project manager).
People were de?ning their own deadlines and I
was just reviewing what they were doing and
support them one hundred percent providing
them with the program, the desks, equipment . . .
I had the role of facilitator . . . I didn’t have that
much in?uence on the technical solution (The
project manager).
Observation showed that control was mainly
informal in nature, with di?use responsibilities, a
lack of explicit guidance and a predilection for
frequent but ad hoc communication. Preplanning
was not possible and so a greater need existed for
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 415
information acquisition on an ongoing basis. In
addition, as the members of the project were reli-
ant upon one another for the accomplishment of
their tasks they tended to share information,
cooperate, and continuously adapt to new insights
as they were emerging.
We had technical meetings to discuss the tech-
nical architecture and everybody was saying ‘Ok
this is what we need to do’, but the control of
specifying those requirements and documenting
the requirements was really loose . . . we started
with the general let’s talk about this and what
sort of things do we need to do . . . We didn’t
de?ne each of our responsibility . . . there was a
lot of informal communication (The technical
manager).
This form of control was also the main knowl-
edge integration channel. The project relied heavily
upon interpersonal processes and communication-
intensive forms of information exchange as the
primary means for integrating knowledge, and
placed little emphasis on codi?cation. Individuals
were combining their specialist knowledge through
‘personal’ and ‘group’ communication, the last
taking the form of meetings. As the range of pro-
blems and decisions to be covered tended to be
wide, hence a great volume of information was
communicated among individuals, often at the
peer level.
To sum up brie?y, informal and intensive inter-
action and communication were the ways to exert
cultural control and, at the same time, achieve
e?ective knowledge integration between the indi-
vidual specialists belonging to the project team.
Discussion of projects
The previous projects provide a diverse set of
software development experiences and di?erent
roles and characteristics of management control
systems. Each project team required di?erent
management control mechanisms depending on
the knowledge complexity of the project.
Knowledge complexity spanned from the com-
putational complexity of project A due to the size
of the project, its duration and the number of
requirements to consider, to the technical com-
plexity of project B, resulting from the participa-
tion of di?erent groups of specialists and, ?nally,
to the cognitional complexity of project C, deriv-
ing from its level of innovativeness.
This variation a?ected the projects’ manage-
ment control practices in terms of the detail of the
de?nition of objectives, the level of detail of
information reporting, the frequency of informa-
tion updating and the usage of information. Pro-
ject A was characterised by an extreme precision
in terms of the de?nition of objectives, a frequent
reporting of detailed and codi?ed information,
and the use of this information to keep track of
the activities carried out (action control). By con-
trast, in project B, while there was still a detailed
de?nition of targets, the information exchanged
during the project was kept at a minimum and was
mainly incorporated in both the outputs delivered
and the performance reported at each of the main
milestones (result control). Finally, project C had
a completely di?erent pattern. The scope of the
project had a very broad de?nition, and information
was continuously exchanged informally between
the members of the team. The purpose was to sort
out potential problems and exceptions as they were
emerging over time (personnel/cultural control).
The same variation was also impacting the
modes of knowledge integration: the codi?cation
of action in project A, output exchange and per-
formance reporting in project B and informal
information sharing in project C.
The comparison of the projects should highlight
the impact of knowledge complexity on the coor-
dination and knowledge integration modes con-
sistently incorporated in the management control
systems. Therefore, management control systems
play a double role in knowledge-intensive ?rms in
that, on the one hand, they help coordinate activ-
ities and, on the other hand, foster a speci?c mode
of knowledge integration. It is therefore their cap-
ability to perform this double role that makes
them e?ective mechanisms in knowledge-intensive
?rms.
To conclude, the theoretical discussion and the
case descriptions suggest that knowledge com-
plexity is a driving force in the design and use of
management control systems.
416 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
Epilogue and agenda for future research
Very little is known about the control practices
of knowledge-intensive ?rms. Past studies have
mainly focused on their organization and man-
agement features, and have only partially addres-
sed control issues. The present paper tries to
overcome this limitation and examines the vari-
ables that a?ect the management control systems
of such type of ?rms.
Starting from the limits of the de?nition of
uncertainty, especially when applied to the context
of knowledge-intensive ?rms, this study focuses on
the impact of a new variant of uncertainty—
knowledge complexity—on management control
systems. This e?ect is evaluated through the inter-
vening role of coordination and knowledge inte-
gration modes. In other words, it is investigated
by examining how knowledge complexity a?ects
coordination and knowledge integration, and in
turn management control systems. These con-
tingent relationships are analysed in the realm of
knowledge-intensive ?rms’ teams, where the pro-
blems of consistency between coordinating indivi-
duals and integrating knowledge mostly occur.
More speci?cally, the analysis suggests that,
according to the type of knowledge complexity
faced by teams—whether computational, technical
or cognitional—knowledge integration tends to be
centred around documentary sources, output
exchange or informal communication, and coor-
dination around action, results or values/beliefs.
Consistent with this framework, management
control mechanisms are to be designed to coordi-
nate individuals and support knowledge integra-
tion at the same time. It is, therefore, the
capability to play both the coordination and
knowledge integration roles that makes manage-
ment control mechanisms e?ective in knowledge-
intensive ?rms. If they were useful mechanisms
only to coordinate individuals’ action but pre-
vented the appropriate integration of knowledge
they would be detrimental to the health and func-
tioning of knowledge-intensive ?rms’ teams.
The analysis started with a review of the litera-
ture on knowledge-intensive ?rms and manage-
ment control systems and proceeded on that prior
knowledge to develop some tentative propositions.
These propositions were further explored by
means of a ?eld study research method. In fact, as
the study of management control systems in
knowledge-intensive ?rms is at an early stage, we
thought that the development and re?nement of
theory was de?nitely more desirable in this phase
than testing pre-de?ned hypothesis. The evidence
collected con?rmed the expectations concerning
the impact of knowledge complexity on manage-
ment control systems.
This study contributes to the literature in several
ways. First, it provides new insights concerning
the control practices of knowledge-intensive ?rms.
Second, by addressing some de?nition issues
associated with the manner in which uncertainty
has been normally described in the management
control contributions, it contributes to a re-inter-
pretation of the contradictory results reported in
the literature on management control systems in
research and development units. Finally, by ana-
lysing the relationship between knowledge com-
plexity and management control systems, it
suggests a framework that could be also fruitfully
applied to other potentially knowledge-intensive
environments like for example marketing depart-
ments, human resources units, and cultural orga-
nizations.
This study has only started to scratch the sur-
face of the complex knowledge-intensive ?rms’
control practices. As research advances, certainly
better classi?cations, descriptions and roles of
management control mechanisms can be devel-
oped as well as implications of combinations of
controls and their interactive e?ects examined.
Further research can also enlarge the set of vari-
ables that a?ect both management control systems’
choices and their e?ectiveness in knowledge-
intensive ?rms. It could also contribute to asses-
sing the relative weight of the causal variables and
their interactions. In addition, new insights could
emerge by considering other industries and levels
of analysis.
Notwithstanding these limitations, this study
does provide some insights into the nature of
controls in knowledge-intensive ?rms and illus-
trates how two complementary needs—coordina-
tion and knowledge integration—can be e?ectively
satis?ed by management control systems.
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 417
Acknowledgements
The author would like to thank M. Abernethy,
M. Agliati, F. Amigoni, S. Beretta, A. Caglio, D.
Cooper, A. Davila, A. Dossi, A. G. Hopwood, P.
Miller, M. Power and M. D. Shields for their
helpful comments and suggestions on earlier drafts
of this paper. The paper bene?ted also from the
discussion with the participants of the Conference
‘Information Flows in Knowledge-Intensive Firms’,
L. Bocconi University, Milan, 2001.
References
Abernethy, M. A., & Brownell, P. (1997). Management control
systems in research and development organizations: the role
of accounting, behavior and personnel controls. Accounting,
Organizations and Society, 22, 233–248.
Abernethy, M. A., & Stoelwinder, J. (1995). The role of pro-
fessional control in the management of complex organiza-
tions. Accounting, Organisations and Society, 20(1), 1–17.
Alvesson, M. (1993a). Cultural-Ideological modes of manage-
ment control: a theory and a case study of a professional ser-
vice company, Communication Yearbook, SAGE, Newbury
Park, vol. 16.
Alvesson M. (1993b). Organisations as rhetoric: knowledge-
intensive ?rms and the struggle with ambiguity, Journal of
Management Studies, 30, (6 November).
Alvesson, M. (1995). Management of knowledge-intensive com-
panies. Berlin: Walter de Gruyter.
Alvesson, M. (2000). Social identity and the problem of loyalty
in knowledge-intensive companies. Journal of Management
Studies, 37(8, December).
Amabile, T. M. (1998). How to kill creativity. Harvard Business
Review, 76, 77–87.
Amigoni F. (1978). I sistemi di controllo direzionale, Giu?re´ ,
Varese.
Ancori, B., Bureth, A., & Cohendet, P. (2000). The economics
of knowledge: the debate about codi?cation and tacit
knowledge. Industrial and Corporate Change, 9(2).
Austin, R., & Larkey, P. (2002). The future of performance
measurement: measuring knowledge work. In A. Neely (Ed.),
Business performance measurement—theory and practice.
Cambridge: Cambridge University Press.
Bamber, E. M., & Bylinski, J. H. (1982). The audit team and the
audit review process: an organisational approach. Journal of
Accounting Literature, 1, 35–58.
Bernardi B., & Warglien, M. (1989). I dilemmi dell’apprendi-
mento in una impresa a intensita` di conoscenza, Sviluppo e
Organizzazione, n. 116 Novembre–Dicembre.
Berso?, E. H., & Davis, A. M. (1991). Impacts of life cycle
models on software con?guration management, communi-
cation of the ACM, August, pp. 104–118.
Birnberg, J. G. (1988). Discussion of an empirical analysis of
the expenditure budget in research and development. Con-
temporary Accounting Research, 4, 582–587.
Blackler, F. (1995). Knowledge, knowledge work and organi-
zations: an overview and interpretation. Organization Stud-
ies, 16(6), 1021–1046.
Bohem B. W. (1988). A spiral model of software development
and enhancements, Computer, May, pp. 61–72.
Boisot, M. H. (1995). Is your ?rm a creative destroyer? com-
petitive learning and knowledge ?ows in the technological
strategies of ?rms. Research Policy, 24, 489–506.
Boisot, M. H. (1998). Knowledge assets-securing competitive
advantage in the information economy. New York: Oxford
University Press.
Boland Jr. R. J., & Tenkasi R. V. (1995). Perspective making
and perspective taking in communities of knowing, Organi-
zation Science, 6(4) July–August, 350–372.
Briers, M., & Hirst, M. (1990). The role of budgetary informa-
tion in performance evaluation. Accounting, Organizations
and Society, 15(4), 373–398.
Brown, J. S., & Duguid, P. (1991). Organizational learning and
communities-of-practice: toward a uni?ed view of working,
learning, and innovation. Organization Science, 2(1, Feb-
ruary), 40–57.
Brownell, P. (1985). Budgetary systems and the control of
functionally di?erentiated organizational activities. Journal
of Accounting Research, 23, 502–512.
Brownell, P., & Dunk, A. S. (1991). Task uncertainty and its
interaction with budgetary participation and budget empha-
sis: some methodological issues and empirical investigation.
Accounting, Organizations and Society, 16, 693–703.
Brownell, P., & Hirst, M. (1986). Reliance on accounting
information, budgetary participation and task uncertainty:
tests of a three-way interaction. Journal of Accounting
Research, Autumn, 241–249.
Burns, T., & Stalker, G. M. (1961). The Management of Inno-
vation. London: Tavistock.
Burton, E. E. (1987). The clustering e?ect: an idea-generation
phenomenon during nominal group. Small Group Behaviour,
18, 224–238.
Casey, J. T., Gettys, C. F., Pliske, R. M., & Mehle, T. (1984). A
partition of small group performance into information and
social components. Organisational Behavior and Human Per-
formance, 34, 112–139.
Chapman, C. S. (1997). Re?ections of a contingent view of
accounting. Accounting, Organizations and Society, 22(2),
189–205.
Chenhall, R. H. (2003). Management control systems design
within its organizational context: ?ndings from contingency-
based research and directions for the future. Accounting,
Organizations and Society, 28(2-3), 127–168.
Cohen, S. G. (1993). New approaches to teams and teamwork.
In: Galbraith J. R. & Lawler E. E., (Eds.) Organizing for the
Future (pp. 194–226), San Francisco: Josey Bass.
Cohendet, P., & Steinmuller, E. (2000). The codi?cation of
knowledge: a conceptual and empirical exploration. Indus-
trial and Corporate Change, 9(1) March.
418 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
Cooper, D., Hinings, C. R., Greenwood, R., & Brown, J. L.
(1996). Sedimentation and transformation in organisational
change: the case of canadian law ?rms. Organization Studies,
17(4), 623–647.
Cowan, R., David, P. A., & Foray, D. (2000). The explicit
economics of knowledge codi?cation and tacitness. Industrial
and Corporate Change, 9(2).
Daft, R. L., & Macintosh, H. J. (1981). A tentative exploration
into the amount and equivocality of information processing
in organizational work units. Administrative Science Quar-
terly, pp. 207–244.
Davila, T. (2000). An empirical study on the drivers of man-
agement control systems’ design in new product develop-
ment. Accounting, Organizations and Society, 25, 383–409.
Dirsmith, M., Heian, J., & Covaleski, M. (1997). Structure and
agency in an institutionalised setting: the application and
social transformation of control in the big six. Accounting,
Organisations and Society, 22(1), 1–27.
Dodgson, M. (1993). Organisational learning: a review of some
of the literature. Organisation Studies, 14(3), 375–394.
Drucker, P. (1993). Post-capitalist society. London: Butter-
worth Heinemann.
Easterby-Smith, M., Corssan, M., & Nicolini, D. (2000).
Organizational leaning: debates past, present and future.
Journal of Management Studies, 783–796.
Eisenhardt, K. M., & Tabrizi, B. N. (1995). Accelerating
Adaptive Processes: Product Innovation in the Global Com-
puter Industry. Administrative Science Quarterly, 40, 84–110.
Ekstedt, E. (1989). Knowledge renewal and knowledge compa-
nies, Research Report no. 22, Uppsala Papers in Economic
History: pp. 1–17.
Ewusi-Mensah, K. (1981). The external organizational envir-
onment and its impact on managerial information systems.
Accounting, Organizations and Society, 6(4), 310–316.
Galbraith J. R. (1973). Designing complex organizations. Addi-
son-Wesley.
Galbraith, J. R. (1977). Organization design. Reading: Addison
Wesley.
Galbraith, J. R., & Kaanjian, R. (1986). Strategy implemen-
tation: the role of structure and process. St. Paul: West
Publishing.
Gordon, L. A., & Narayanan, V. K. (1984). Management
accounting systems, perceived environmental uncertainty
and organization structure: an empirical investigation.
Accounting, Organizations and Society, 1, 33–47.
Grandori, A. (1997). An organizational assessment of inter-
?rm coordination modes. Organization Studies, 897–925.
Grandori A. (1999). ‘Reti Organizzative e Governo delle Con-
oscenze. Relazione al Convegno Aidea, Le Relazioni Inter-
aziendali nella dinamica competitive, Parma, 29 ottobre.
Grant, R. M. (1996a). Prospering in dynamically-competitive
environments: organizational capability as knowledge inte-
gration. Organization Science, 7(4, July-August), 375–387.
Grant, R. M. (1996b). Toward a knowledge-based theory of
the ?rm. Strategic Management Journal, 17(Winter Special
Issue), 109–122.
Grant, R. M. (1997). The knowledge-based view of the ?rm:
implications for management practice. Long Range Planning,
30(3), 450–454.
Greenwood, R., Hinings, C. R., & Brown, J. (1990). P2-Form
strategic management: corporate practices in professional
partnerships. Academy of Management Journal, 33(4), 725–
755.
Grindley, P. C. (1988). The UK software industry: a survey of
the industry and evaluation of policy. Centre for Business
Strategy, London Business School.
Hare, A. P. (1976). Handbook of small group research (2nd ed.).
Collier: Massachusetts Free Press.
Hartmann, F. G. H. (2000). The appropriateness of RAPM:
toward the further development of theory. Accounting,
Organizations and Society, 25(4–5).
Hayes, D. D. (1977). The contingency theory of managerial
accounting. The Accounting Review, 52(1), 22–39.
Hayes D. C. (1978). The contingency theory of managerial
accounting: a reply, The Accounting Review, Vol. LIII,
2(April), pp. 530–533.
Henderson, R., & Cockburn, I. (1994). Measuring Compe-
tence: Exploring Firm-e?ects in Pharmaceutical Research.
Strategic Management Journal, 15(Winter Special Issue), 63–84.
Hinings, C. R., Brown, J. L., & Greenwood, R. (1991). Change
in autonomous professional organization, Journal of Man-
agement Studies, 28:4, July.
Hirst, M. K. (1981). Reliance on Accounting Performance
Measures, Task Uncertainty and Dysfunctional Behavior.
Journal of Accounting Research, 21(2), 596–605.
Hirst, M. K. (1983). Reliance on accounting performance
measures, task uncertainty and dysfunctional behavior: some
extensions. Journal of Accounting Research, 596–605.
Hopwood, A. G. (1989). Organizational contingencies and
accounting con?gurations. In B. Friedman, & L. Ostman
(Eds.), Accounting development—some perspectives: a book
in honour of Sven-Erik Johansson (pp. 23–44). Stockolm:
Economic Research Institute.
Iansiti, M., & Clark, K. B. (1994). Integration and dynamic
capability: evidence from product development in auto-
mobiles and mainframe computers. Industrial and Corporate
Change, 3(3).
Ismail, Z., & Trotman, K. T. (1995). The impact of the review
process in hypothesis generation tasks. Accounting, Organi-
zations and Society, 20, 345–567.
Kamm, J. (1980). The balance of innovative behavior and con-
trol in new product development, DBA Dissertation, Gradu-
ate School of Business Administration, Harvard University.
Kay, N. M. (1979). The innovating ?rm. New York: St Martin’s
Press.
Khandwalla, P. (1972). The e?ects of di?erent types of compe-
tition on the use of management controls. Journal of
Accounting Research, 275–285.
Koga, K., & Davila, A. (1998). What is the Role of Performance
Goals in Product Development? A study of Japanese Camera
Manufacturers, Working Paper, Harvard Business School.
Kogut B., & Zander U. (1996). What ?rms do? Coordination,
identity, and learning. Organization Science, 7(5), Septem-
ber–October, 502–518.
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 419
Lang?eld-Smith, K. (1997). Management control systems and
strategy: a critical review. Accounting, Organizations and
Society, 22, 207–232.
Lave, J., & Wenger (1991). Situated learning: legitimate periph-
eral participation. Cambridge, MA: Harvard University
Press.
Lawrence, P. R., & Lorsch, J. W. (1967). Organization and
environment. Boston, MA: Division of Research, Harvard
Business School.
Levitt, B., & March, J. G. (1988). Organizational learning.
Annual Review of Sociology, 14, 319–340.
Lothian, N. (1984). How companies manage R&D: a survey of
major UK companies. London: Chartered Institute of Man-
agement Accountants (CIMA).
Maier, N. R. F. (1970). Problem solving and creativity in indivi-
duals and groups. Monterey: Brooks/Cole.
Maister, D. H. (1982). Balancing the professional service ?rm,
Sloan Management Review, Fall.
Merchant, K. A. (1985). Control in business organizations.
Boston: Pitman.
Merchant, K. A. (1998). Modern management control systems.
Upper Saddle River: Prentice Hall.
Mintzberg, H. (1979). The structuring of organizations. Englewood:
Prentice Hall, Inc.
Mintzberg, H. (1983). Structure in ?ves. Englewood Cli?s:
Prentice-Hall, Inc.
Mohrman, S. A. (1993). Integrating roles and structure in the
lateral organization. In J. R. Galbraith, & E. E. Lawler
(Eds.), Organizing for the future (pp. 109–141). San Fran-
cisco: Josey Bass.
Montagna, P. (1968). Professionalization and bureau-
cratization in large professional organizations. American
Journal of Sociology, 74(September), 138–145.
Morris, T., & Empson, L. (1998). Organisation and expertise:
an exploration of knowledge bases and the management of
accounting and consulting ?rms. Accounting, Organizations
and Society, 23(5–6), 609–624.
Mowery, D. C. (1996). Introduction to Mowery D. C. (Ed.),
The international computer software industry, New York:
Oxford University Press.
Nelson, R. L. (1988). Partners with power: the social transfor-
mation of the large law ?rms. Berkeley: University of Cali-
fornia Press.
Nelson, R. R. (1991). Why do ?rms di?er, and how does it
matter? Strategic Management Journal, 12(Winter Special
Issue), 61–74.
Nixon, B. (1998). Research and development performance
measurement: a case study. Management Accounting
Research, 9, 329–355.
Nonaka, I. (1994). A dynamic theory of organizational knowl-
edge creation, Organization Science, Vol. 5, No. 1, February:
pp. 15–37.
Nonaka, I., & Takeuchi, H. (1995). The knowledge-creating
company—how japanese companies create the dynamics of
innovation, Oxford University Press.
Nonaka, I., Toyama, R., & Nagata, A. (2000). A ?rm as a
knowledge-creating entity: a new perspective on the theory of
the ?rm. Industrial and Corporate Change, Vol. 9, No. 1,
March.
Nurmi, R. (1998). Knowledge-intensive ?rms. Business Hor-
izons, May–June.
Otley, D. T. (1980). The contingency theory of management
accounting: achievement and prognosis. Accounting, Organi-
zations and Society, 5(4), 413–428.
Ouchi, W. (1979). A conceptual framework for the design of
organizational control mechanisms. Management Science,
25(9), 833–848.
Ouchi, W. G., & Bolton, M. K. (1988). The logic of joint
research and development. California Management Review,
30(3), 9–33.
Perrow, C. (1967). A framework for the comparative analysis
of organizations. American Sociological Review, 32(April),
194–208.
Perrow, C. (1970). Organizational analysis: a sociological view.
California: Wadsworth Publishing Company.
Polanyi, M. (1966). The tacit dimension. London: Routledge &
Kegan Paul.
Prahalad, C. K., & Hamel, G. (1990). The core competence of
the corporation. Harvard Business Review, May–June, 79–91.
Raelin, J. A. (1985). The basis for the professional’s resistance
to managerial control. Human Resource Management, Sum-
mer, 24(2), 147–175.
Rockness, H. O., & Shields, M. D. (1984). Organizational
control systems in research and development. Accounting,
Organizations and Society, 9, 165–177.
Rockness, H. O., & Shields, M. D. (1988). An empirical analy-
sis of the expenditure budget I research and development.
Contemporary Accounting Research, 4, 568–581.
Rook, P. (1986). Controlling software projects. Software
Engineering J., January, 7–16.
Scott, T. W., & Tiessen, P. (1999). Performance measurement
and managerial teams. Accounting, Organizations and
Society, 24, 263–285.
Schoonhoven, C. B. (1981). Problems with contingency theory:
testing assumptions hidden within the language of contingency
‘theory’. Administrative Science Quarterly, 26, 349–377.
Shaw, M. E. (1976). Group dynamics: the psychology small
group behavior. New York: McGraw-Hill.
Simon, H. A. (1962). The architecture of complexity. Proceed-
ings of the American Philosophical Society, 106, 467–482.
Simon, H. (1979). The architecture of complexity, the sciences of
the arti?cial. Cambridge: MIT Press.
Simons, R. (1987). Accounting control systems and business
strategy: an empirical analysis. Accounting, Organizations
and Society, 12, 357–374.
Smigel, E. O. (1963). The Wall Street lawyer: professional
organisation man?. New York: Free Press.
Starbuck, W. H. (1992). Learning by knowledge-intensive
?rms. Journal of Management Studies, 29(6 November).
Starbuck, W. H. (1993). Keeping a butter?y and an elephant in
a house of cards: the elements of exceptional success. Journal
of Management Studies, 30(6 November).
Stein, M. K. (1975). Stimulating creativity. New York: Academic
Press.
420 A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421
Stocks, M. H., & Harrel, A. (1995). The impact of an increase
in accounting information level on the judgement quality of
individuals and groups. Accounting, Organizations and
Society, 20(7–8), 685–700.
Svieby, K., & Lloyd, T. (1987). Managing know-how. London:
Bloomsbury.
Taylor, D. W., Berry, P. C., & Block, C. H. (1958). Does group
participation when using brainstorming facilitate or inhibit
creative thinking? Administrative Science Quarterly, 2, 23–
47.
Thompson, J. D. (1967). Organizations in action. New York:
McGraw-Hill.
Tiessen, P., & Waterhouse, J. H. (1978). The contingency the-
ory of management accounting: a comment. The Accounting
Review, April, 523–529.
Vancil, R. F. (1979). Decentralization: managerial ambiguity by
design. Homewood: Dow Jones-Irwin.
Van Maanen, J., & Kunda, G. (1989). Real feelings: emotional
expression and organization culture. Research in Organiza-
tional Behavior, 11, 43–103.
Waterhouse, J., & Tiessen, P. (1978). A contingency framework
for management accounting systems research. Accounting,
Organizations and Society, 3(1), 65–76.
Wenger, E. (1998). Communities of practice-learning as a social
system. Systems Thinker, June.
Williamson, O. E. (1985). The economic institutions of capital-
ism. New York: Free Press.
Williamson, O. E. (1993). Transaction cost economics and
organization theory. Industrial and Corporate Change, 2(2),
107–156.
Winch, G., & Schneider, E. (1993). Managing the knowledge-
based organization: the case of architectural practice. Journal
of Management Studies.
Winter, S. (1987). Knowledge and competence as strategic assets.
In D. Teece (Ed.), The competitive Challenge. Cambridge:
Ballinger.
Wood, R. E. (1986). Task complexity: de?nition of the
construct. Organizational Behavior and Human Decision
Processes, 60–82.
Woodward, J. (1965). Industrial organization: theory and
practice. London: Oxford University Press.
Yetton, P. W., & Bottger, P. C. (1982). Individual versus group
problem solving: an empirical test for a best member strategy.
Organisational Behavior and Human Performance, 30, 307–321.
Yin, R. K. (1988). Case study research: design and methods.
Newbury Park: Sage Publications.
Young, S. M. (1996). Survey research in management account-
ing: a critical assessment. In A. J. Richardson (Ed.), Research
Methods in Accounting. Vancouver: CGA Research Founda-
tion.
Zander, U., & Kogut, B. (1995). Knowledge and the speed of
the transfer and imitation of organizational capabilities: an
empirical test. Organization Science, 6(1, January–February),
76–92.
A. Ditillo / Accounting, Organizations and Society 29 (2004) 401–421 421

doc_232977983.pdf