Interfaces of control. Technocratic and socio-ideological control in a global management c

Description
This paper investigates a variety of forms of management control in a large management consultancy company. The
very high level of compliance with corporate objectives among employees is highlighted and singled out as a phenomenon
worth exploring. The company exhibits a rich variety of various formal control devices focusing on financial
issues as well as human resources. The paper shows that these do not fully work according to intention and that their
control effects do not comply that well with the bureaucratic- technocratic logic they rest upon. Technocratic control
systems are instead identified as non-obvious sources of socio-ideological control. The paper emphasizes the interface
between different forms of control and argues for a more symbolic, meaning-focusing view on bureaucratic and output
measurement control.

Interfaces of control. Technocratic and socio-ideological
control in a global management consultancy ?rm
§
Mats Alvesson*, Dan Ka¨ rreman
Department of Business Administration, Lund University, PO Box 7080, S 220 07 Lund, Sweden
Abstract
This paper investigates a variety of forms of management control in a large management consultancy company. The
very high level of compliance with corporate objectives among employees is highlighted and singled out as a phenom-
enon worth exploring. The company exhibits a rich variety of various formal control devices focusing on ?nancial
issues as well as human resources. The paper shows that these do not fully work according to intention and that their
control e?ects do not comply that well with the bureaucratic- technocratic logic they rest upon. Technocratic control
systems are instead identi?ed as non-obvious sources of socio-ideological control. The paper emphasizes the interface
between di?erent forms of control and argues for a more symbolic, meaning-focusing view on bureaucratic and output
measurement control.
#2003 Published by Elsevier Ltd.
Introduction
Forms of management and organizational con-
trol are a common theme in organization and
business studies. In fact, the whole idea of man-
agement accounting, for example, is founded on
the belief that management control is possible,
important, and, indeed, necessary. Although the
literature on organizational and management
control suggests a wide array of forms of control,
it is common to emphasize a main form of control,
either in the form of a particular organizational
structure or in the form of a speci?c mode of
control dominating. Mintzberg’s (1983a, 1983b)
?ve types are well-known, as are Edwards’
(1979) three and a variety of versions setting up
one authoritarian-hierarchical-tayloristic against
a normative-market-self-governance version (e.g.
Clegg et al, 1996; Friedman, 1977; Govindarajan,
1988; Govindarajan & Fisher, 1990; Spekle´ 2001).
One in?uential variation of the last theme is to
contrast structural control with cultural control.
Earlier enthusiastic versions of this include Peters
and Waterman (1982), but also more sceptical
researchers like Kunda (1992) have emphasized
the signi?cance of normative control—or engi-
neering culture—as a particular form of control.
There is a strong tendency to emphasis an
either-or orientation in the literature. The ?rst
author of the present paper is no exception, hav-
ing paid particular attention to cultural control
(Alvesson, 1993, 1995). To concentrate on one
type of phenomenon may often be motivated by
the need to have a focus. Theoretical interest as
well as what seems to be vital or interesting in a
particular empirical site can create a rationale for
0361-3682/$ - see front matter # 2003 Published by Elsevier Ltd.
doi:10.1016/S0361-3682(03)00034-5
Accounting, Organizations and Society 29 (2004) 423–444
www.elsevier.com/locate/aos
§
Paper presented at conference on ‘‘Information Flows in
Knowledge-Intensive Firms’’, Bocconi University, Milan, 27
November 2001.
* Corresponding author.
E-mail address: [email protected] (M. Alvesson).
investigating a particular form of control. It is
sometimes also argued that organizations typically
are structured in particular con?gurations and
tend to move towards balanced and ‘pure’ forms
(Mintzberg, 1983a). In a study of a major Amer-
ican high technology company, Van Maanen and
Kunda (1989) emphasize that for many employees
‘‘. . .‘culture’ replaces ‘structure’ as an organizing
principle and is used both to explain and guide
action’’ (p. 72). They consider that the formal
organization is not, per se, particularly important.
However, a case can be made for understanding
various forms of control as simultaneously active,
at least in reasonably complex organizations.
Indeed, Kunda (1992) maintains that certain ele-
ments of bureaucratic control remain in place,
even when normative control is predominant. In
the company studied by Kunda, control of culture
is supplemented by bureaucratic and utilitarian
control. To use Mintzberg’s (1983a) typology, a
battery of control forms seems to be used in many
contemporary organizations, typically blending
standardized output measures with standard
operating procedures (standardized work proce-
dures) and a professionalized work force (stan-
dardized knowledge). In such settings it might be
counter-productive to assume the existence of a
dominating form. Indeed, di?erent control forms
may be linked to, and supporting and sustaining
each other, rather than subdued and marginalized
by a dominant form.
On a conceptual level, it can be argued that
contrasting forms of control, while analytically
distinct, do not necessary exclude each other. On
the contrary, it can be argued that structural
forms of control also have a signi?cant cultural
dimension (Ferner, 2000; Ranson, Hinings, &
Greenwood, 1980). From an anthropological
point of view, it could be argued that structural
forms of control are cultural phenomena them-
selves, that have and take on speci?c meaning
depending on cultural contexts (taken for granted
in Western culture, signifying modernism and/or
imperialism in other cultures). Structural forms
never exist in a culture-free context, nor is this
ideational level necessarily tightly connected to a
particular social form. Structural forms of control
focusing on behaviour and/or measurable output
are supported and/or complemented, sometimes
perhaps challenged or contradicted by, cultural
ideas and values.
The present analysis focuses on the interface
between certain structural arrangements and the
ideologies and norms targeting how people are
supposed to think, feel and act. A key theme is
how a high level of compliance—including a
readiness to work very long hours and accept very
ambitious objectives in terms of keeping deadlines
and accomplishing high margins in project work—
is achieved. How di?erent forms of technocratic
control interact and merge with socio-ideological
control is investigated.
The empirical domain addressed is the manage-
ment consulting industry and profession, in parti-
cular as it is manifest in very large, international
companies whose services include advice-giving
and strategic planning as well as the implemen-
tation of suggested solutions, often involving a
large group of consultants over a long period of
time. How a fairly strict form of control is
accomplished in one such company is an interest-
ing question which is capable of giving some input
into how we can understand modes of manage-
ment control within the particular empirical
domain investigated, but also in more general
theoretical terms.
Forms of management control
Although managers do other things, the exercise
of control is a dominant part of the manager’s job
(Mintzberg, 1989; Tengblad, 2001). Management
control has been de?ned in numerous ways but
most de?nitions seem to agree that management
control includes the exercise of power (in?uence)
in order to secure su?cient resources, and
mobilize and orchestrate individual and collec-
tive action towards (more or less) given ends
(c.f. Lang?eld-Smith, 1997; Spekle´ , 2001 for
reviews). Management control typically includes
an apparatus for specifying, monitoring and eval-
uating individual and collective action. It focuses
worker behaviour, output and/or the minds of the
employees. Sometimes it attempts to focus on all
three.
424 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
Managerial activity that attempts to control
behaviour typically includes designing and super-
vising work processes. This is usually carried out
in a way that attempts to make work processes
as simple and transparent as possible, thereby
lowering knowledge thresholds (and the price
of labour). In this sense, the long dominance of
Taylorism, broadly de?ned, as the managerial
approach is unsurprising. Its approach is to con-
struct work contexts where workers discretion is
minimized, to the extent that they only can do the
prescribed thing with a minimum of e?ort and
movement.
Management control practices targeting minds,
through norms, emotions, beliefs and values, are
intended to a?ect behaviour indirectly. This type
of control has been discussed under the label of
normative (Etzioni, 1960; Kunda, 1992), cultural-
ideological (Alvesson, 1993), and concertive (Bar-
ker, 1993) control. In this article we will use the
label socio-ideological control for attempts to con-
trol worker mind-sets and technocratic control for
attempts to directly control worker behaviour (c.f.
Alvesson & Ka¨ rreman, 2001a). In the technocratic
type, management works primarily with plans,
arrangements and systems focusing behaviour
and/our measurable outputs. In the socio-ideolo-
gical version, social relations, identity formation
and ideology are basic ingredients.
Technocratic forms of control
Most studies of social control of work and
organization have traditionally focused on the
objective, behavioural aspects of control. In a
review article, Simpson (1985) for example, draw-
ing heavily on Edwards (1979), identi?es ?ve
modes of control over work: simple, technical,
bureaucratic, occupational control, and worker
self-control modes of control. Simple control
refers to the personal exercise of power of the boss
(owner, supervisor) over the worker. Technical
control is embedded in the technology of work.
Bureaucratic control is carried out through rules,
policies, formal incentives and other impersonal
devices. In occupational control, a profession
de?nes appropriate/no appropriate work beha-
viour. Finally, worker self-control labels control
contexts where the producers themselves have a
high degree of discretion.
Until some time ago, this ‘‘objectivistic’’ view on
control was dominant. In his synthesis of research
on organizational structure, Mintzberg’s (1983a)
?ve coordination mechanisms concern objective,
external control forms and hardly anything on the
signi?cance of the whole sphere of culture and
ideology. Ouchi (1979:840), for example, noted
that ‘‘present organization theory. . . concentrates
on the bureaucratic form to the exclusion of all
else’’. When ideology is considered, it is basically
as legitimating these forms of ‘‘objective’’ social
control (Burris, 1989) or as a major control device
in very special organizations, such as the ‘‘mis-
sionary’’ one (Mintzberg, 1983b).
In the context of management accounting
research, management control has largely been
interpreted in technocratic terms, as either exer-
cised through output control (through focus on
various key performance indicators, such as pro?t,
sales, and quality measurements) or through
behavioural control (such as direct supervision,
rules, standard operating procedures, and business
policies) (Govinadarajan, 1988; Govindarajan &
Fisher, 1990; Lang?eld-Smith, 1997; Lind, 2001;
Spekle´ , 2001). The insight that management
accounting seeks not only to a?ect behaviour but
also consciousness to some extent transcends a
narrow technocratic focus (Puxty, 1993). But as
Lang?eld-Smith (1997) points out, most research
on management control in management accounting
research has failed to acknowledge the distinction
between ‘‘the existence and the use of controls’’
(p. 226) as well as addressing forms of control that
extend beyond the familiar terrain of output and
formal behaviour. However, as pointed out above,
and elaborated below, management control has a
broader theoretical and practical spectrum.
Socio-ideological forms of control
Managers do not only exercise control through
prescribing behaviour or desired outputs. Man-
agers also often seek to enact a particular form of
organizational experience for others (Alvesson
& Willmott, 2002; Deetz, 1992; Mumby, 1988;
Smircich & Morgan, 1982).
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 425
They may attempt to de?ne interpretations
and meanings than can become widely
understood and shared by organization
members so that actions are guided by a
common de?nition of the situation. Those
with power are able to in?uence the course of
organizational development through control
over valued resources and through use of
symbols by which organization members
mediate their experience (Smircich, 1983:161).
In management accounting research, what we
have labelled socio-ideological forms of control
are more frequently (and misleadingly) labelled
informal controls, or in some cases, clan controls
(Lang?eld-Smith, 1997). Ouchi (1979, 1980) has
suggested that the clan can serve as an organiza-
tional control form in which the level of uncer-
tainty is too great for the classic solutions—the
market and the bureaucracy—to function. The
clan rests on ‘‘. . .social agreement on a broad
range of values and beliefs’’ and ‘‘. . .relies for its
control upon a deep level of common agreement
between members on what constitutes proper
behaviour, and it requires a high level of commit-
ment on the part of each individual to those
socially prescribed behaviours’’ (Ouchi, 1979:838).
Many writers focusing on ideology come close
to what others refer to as clans or cultures, even
though the latter concept often is understood as a
broader and more complex one. Ideology can be
de?ned as an integrated set of values, ideals and
understandings about a particular part of social
reality which justify certain commitments and
actions (c.f. Beyer, 1981; Geertz, 1973; Weiss &
Miller, 1987, etc). Sometimes ideology is seen as
control (Czarniawska-Joerges, 1988) but typically
a rather broad and general concept is used, which
can cover a wide area of possible forms of more
speci?c control. Ideology is often confronted with
a single traditional conception of management or
control. Becke´ rus, Edstro¨ m et al. (1988), for
example, argue for a view on management based
on control through ideology (ideas) rather than
through control through instructions.
Socio-ideological control can be de?ned as
e?orts to persuade people to adapt to certain
values, norms and ideas about what is good,
important, praiseworthy, etc in terms of work and
organizational life. Ideologies justify certain prin-
ciples, actions and feelings, and discourage others
(Alvesson, 1987; Czarniawska-Joerges, 1988, etc).
They are more or less elaborated and deeply held,
both on behalf of the person who tries to exercise
ideological control and the audience that has been
persuaded. When engaged in ideological control
e?orts, managers more or less consciously and
systematically, try to make the employees adhere
to the values and ideals which they believe in—or
at least the values and ideals which they believe
that the company would bene?t most from the
employees believing in. We can imagine a spec-
trum of forms of implementation of ideological
control. At the one extreme, the agent of ideolo-
gical control is reproducing ideas that he or she
takes for granted and do not re?ect upon—in this
case the controller is him- or herself also con-
trolled, and it is questionable if this is a case of
(conscious) management control. At the other
extreme, the agent is manipulating—proposing
ideals and values that he/she himself does not
believe in as means for in?uencing others. The
latter enterprise probably easily creates back-
lashes, while there is a risk that people in the long
run see through manipulations. In the normal
case, ideological control is based on a combi-
nation of the calculations and convictions of key
actors.
On the relationship between technocratic and
socio-ideological control
The literature typically portrays di?erent forms
of control as external and alternative to each
other. As indicated in the introduction, most
authors aiming to give the broader picture (Min-
tzberg, Edwards, etc) stress that di?erent organi-
zations and di?erent periods are dominated by
di?erent forms of control. Barley and Kunda (1992)
talk of cycles of control rhetorics, in which nor-
mative and rational forms dominate during di?er-
ent periods in a cyclical way. But also researchers
having conducted in depth case studies emphasize a
dominating form either focusing on structure and
behaviour or corporate culture, ideology and iden-
tity (c.f. Kunda, 1992; Martin, 1992; Alvesson,
426 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
1995). Some draw attention to the interplay
between di?erent modes of control. Barker (1993)
shows how people’s ideas and norms lead to the
construction of rules enacted in a bureaucratic
way. Ferner (2000) emphasizes that ‘formal con-
trols are underpinned by personal and cultural
control mechanisms’ (p. 524). Some critically
oriented, Foucault-inspired accounting research
also transcend the distinction between structural-
behavioural and ideological (discursive)-subjective
when emphasizing how management accounting
and control systems ‘are part of the organization’s
(and society’s) regular, routine procedures which
embed managers and employees in a disciplinary,
punishable web of discourses and practices which
go on almost unnoticed and appear as neutral
(Hopper & Macintosh, 1993:190, see also Hop-
wood, 1987; Miller & O’Leary, 1987). Also trans-
cending a simple structure-ideology divide is the
point that in certain respects ‘it is not the fore-
casting ability of a budget that is important, rather
it is the desirability of the situation that it helps to
create’ (Tinker et al., cited by Puxty, 1993:102).
An interesting question is, however, the more
precise relationship between di?erent types of
control. An alternative idea would be proceed
from the assumption that management employs a
wide variety of forms of control and it may be
di?cult to point at a single, dominant one. At
least, one could see it as good research question to
study the plurality versus domination of a single
mode(s) of organizational control. Of course, the
way one approaches this question is partly a mat-
ter of how broad are the categories one is using.
Given a broad view of ideological control, it cov-
ers a lot and easily become dominant, but one may
also favour dividing the control pie into smaller
cakes and then plurality appears.
Another interesting question concerns to what
extent di?erent forms of control are exercised in
di?erent ways, external to each other and perhaps
complementing or contradicting each other. Are
di?erent forms of control pure or fused. Does for
example bureaucratic control mainly function in
an ‘objective’ or mechanistic way, or is it heavily
dependent on a particular mode of interpretation,
based on values, convictions and norms, in order
to work? And is any ideational control grounded
mainly in the communication of ideological mes-
sages, expressed by leaders, explicitly spelling out
what is good and valuable? Or is it ?rmly
anchored and expressed in the organizational sys-
tems and practices for behavioural and output
control?
Most of the literature seems to argue that ideo-
logical control is exercised in non-bureaucratic
and non-output measurement ways. Leadership is
typically emphasized in relationship to cultural or
ideological control (e.g. Schein, 1985). Ouchi
(1980) says that the clan is the alternative when
bureaucracy and market (price mechanism) do not
work.
Based on our case study, we will argue for a
di?erent kind of understanding. The analysis pro-
ceeds by taking a close look at the various forms
of control present in a global management con-
sulting ?rm. Drawing on observational data, we
attempt to demonstrate how control forms inter-
act, occassionally merge and, sometimes, contra-
dict each other. The paper ends with a discussion
of how forms of control may interface with, rather
than substitute for, each other.
The case
Method
As pointed out by Jo¨ nsson (1998), most man-
agement accounting research lacks empirical
input, and the empirical material typically drawn
upon tends to ‘‘be limited to quick survey studies
which ?t into the publication requirements of the
main stream’’ (p. 411). Jo¨ nsson suggests the use a
more diverse set of research methods, to be rele-
vant for understanding managerial and organiza-
tional work. Consequently, this study draws upon
a broad set of practices for data generation. The
study has been conducted through interviews as
well as observations of a variety of organizational
gatherings. Observations includes following a pro-
ject group at work during two days, observations
of training sessions, competence group meetings
and the yearly meeting for everybody in managerial
positions. We have also covered events where
members of the organization have interacted
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 427
externally, as when presenting the company to
students. We have also studied internal documents
and learning materials such, as manuals for giving
feedback, project methodology, etc.
We have conducted 52 interviews with 45 per-
sons. People from all parts of the organization
have been interviewed: the CEO, partners, con-
sultants on various levels of seniority, support
sta?, newly recruited organizational members and
so on. We have focused on what in ethnographic
vocabulary is labelled as ‘studying up’: we have, in
relative terms, focused more on experienced peo-
ple than on fairly recently recruited people, which,
in numerical terms dominate the ?rm.
Fieldwork has been conducted in a way that is,
as far as possible, open and emergent. Openness is
something that one must struggle for, e.g. through
thinking critically about and broadening the
perspectives, metaphors and vocabularies the
researcher is considering before producing an
approach to interpretation. The study draws on
interpretive principles, but with a slightly critical-
sceptical edge. Rather than codifying empirical
material, this is looked upon as text, in which
one tries to go beyond the ‘surface’ and look for
something less obvious, or less easily revealed in
a (quick) coding process, and where also the text
as a totality is born carefully in mind, which
means that variation and contradiction is taken
seriously.
As our understanding of the ?eld has developed,
our lines of inquiry have followed suit. For exam-
ple, ?ndings and understanding from our ?rst 20
interviews were organized in emergent themes,
which were used as input in new interviews, both
in terms of questions asked and whom to talk to.
In this way we have been able to re?ne our
understanding of the themes that have emerged,
without providing excessive a priori closure to
?eld-work practices.
On the company
Our case company is Global Consulting, one of
the words largest management consultancy com-
panies. We have primarily studied the Nordic
subsidiary, employing roughly 800 people. The
company has grown rapidly, is very pro?table and
has generally a good reputation. As one inter-
viewee says, when asked why people consider
using the very expensive Global for a task, ‘it is
never considered wrong to ask us for an opinion’.
The company stands for quality, reliance, pro-
fessionalism and predictability.
The key organizational unit in Global Consult-
ing is always the team. Every project and customer
contact is constructed around a particular team.
Thus, management control in Global Consulting
revolves generally around team management. As a
consequence, the individual is typically rather
insigni?cant, at least as an organizational
resource. On one level the individual is highly
visible in the organization: it is the individual who
is evaluated, rewarded and punished. But on
another level, the individual is hardly visible at all.
Instead, the individual is viewed as perfectly
exchangeable and possible to replace, at least on
junior and middle levels.
Governing bodies: partner, organizational and
project control
At Global management control is exercised
from a number of distinct levels or domains. First,
control is exercised from a partner level, which in
this context covers the dominant stakeholder’s
ways and means to orchestrate the company’s
e?orts in line with their collective interests. This
includes an interest in ?nancial and operating
performance, but also strategic issues, such as
change and developments of the business model.
Other forms of control are exercised through
HRM practices and methodologies for project
work aimed to promote the correct behaviour.
There is also control being exercised at the level of
projects, which includes everyday practices to
secure that projects deliver what they are expected
to deliver, within the budget and on time.
People in the company also talk about these
arrangements in terms of culture: as a ‘control cul-
ture’—referring to the detailed monitoring of costs
and revenues—as a ‘feedback culture’ and as a
‘delivery culture’. All these ‘cultures’ can be descri-
bed as sets of understandings and values intertwined
with formal structures and monitoring practices.
428 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
The role of partners: frame control
The company is, as most other large companies
with a fairly low ratio of partner—total work-
force, strongly ‘partner-centric’. The partners
have—or are at least perceived to have—a strong
control over the business. The CEO describes his
task as monitoring market opportunities and try-
ing to match resources with these, to relocate
resources from a market segment with over time
diminishing importance to one with potential.
This means mainly to see to that we have
partners at the right place. If you get that
segment to work, the remaining part of the
forces to group themselves in the right ways.
So if we move a partner from one area to
another, a number of other resources—com-
petence—will follow this move.
The employees will not follow automatically or
instantly, but if one partner moves to another
area, businesses and sales will follow and gradu-
ally about 20 people will join him or her.
Partners control mainly through sales and
assignments and deciding upon the terms—in
particular the margin—for projects. We can talk
about sales-based control, meaning that part of
the control is exercised via the promise made to
the customer. Partners have a strong incentive to
sell. They are assessed on a number of criteria, but
the volumes they have sold during a year seem to
be the major variable. It determines how many
‘units’—i.e. shares in the company—they are
allowed to purchase next year, which in its turn is
the determinator, together with the overall pro?ts
of the company, on their future income. There are,
within the company, complaints about partners
being too one dimensionally sales-oriented and
even sometimes greedy. One interviewee talks
about ‘a horrible prioritisation’ on selling projects
even at times when the company had di?culties in
recruiting su?cient number of personnel and that
there was an agreement that recruiting and
retaining was the critical issue.
The partner group exercises control mainly
through getting assignments and contracts with
clients, making the client agreement the frame of
control for subordinates. Project managers must
agree on the terms and to do the job with a parti-
cular amount of resources allowing for a parti-
cular (high) margin. The partner group exercises
control over subordinates through the client in
combination with demand for and close monitor-
ing of project margin. This combination can be
labelled frame control, in which the project man-
ager, in particular, is ‘boxed in’ a frequently fairly
tight delivery and margin contract. This is not
exclusive for the company here studied. O’Shea &
Madison (1998) report how employees at one
consultancy company felt depressed every time
they won a new contract: ‘how in hell will we be
able to deliver all this within the budget’, they felt.
Partners are experienced as rather distanced
from the broader groups of employees. Junior
employees feel that ‘they are very small’ in relation-
ship to partners and do not know what their super-
iors do. Partners are also described as demanding
and sometimes closer to a counterpart in a business
relationship than as co-workers. They are also,
however, seen as open and receptive to good ideas.
There is some internal mysti?cation of partners,
partly a consequence of their centrality in selling and
ability to control the careers of people, but the per-
ceived distance and lack of knowledge about what
partners do and how they are seem to fuel fantasies
amongst the junior employees. Both partners and
juniors agree on there being little interaction
between top and bottom. The scope of leadership is
thus limited. Partners control through formal posi-
tion, and a combined bureaucracy-market mediated
form of control, not through being ‘leaders’.
Nevertheless, partners do have a strong grip of
business, managing a workforce partly through
controlling sales and business contracts, monitor-
ing project work and performances and control-
ling the career options of the people eager to move
up along the career ladder.
Organizational control: hierarchical levels, division
of labour, outputs and the HRM system
The organizational dimension of control at
Global is characterized by a ?ne-tuned hierarchy,
standardized work procedures, output measure-
ments and the HRM system.
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 429
Hierarchy is highly visible and pronounced at
Global. It is perceived to express genuine di?er-
ence, both in terms of say-so and power, and in
terms of experience and competence. The steps
include analysts, consultants, managers and part-
ners. People tend to take these career steps and the
titles very seriously, meaning that the level of
social di?erentiation is strong and people are
focused on their relative position. The hierarchy is
further elaborated with the practice of designating
total number of years at the position as a marker
of seniority.
Work processes are standardized in work meth-
odologies. There exists a uni?ed package of meth-
ods that consultants are expected to use in
projects. Due to the fact that most projects di?er
somewhat from each other, the methodology is
not expected to be used as a prescription, but as a
resource. Global have invested heavily in systems
for knowledge management. Knowledge manage-
ment is primarily seen as a way of taking advan-
tage of the scale and scope of the accumulated
experience generated within the ?rm. Another key
aspect is to diminish the importance of the indivi-
dual and of personal experience. The idea is that
ideas and experiences developed in one project can
be recycled and reused with minimal adaption in
other projects. The premise is that experience can
be codi?ed and rationalized in a way that suits
database storage and retrieval.
In the company there is a very strong focus on
outputs and the delivery of assignments on time.
Work procedures and management styles are
typically geared towards giving top priority to the
task at hand. This focus on the task is typically
viewed as a strength since it contributes to make
Global known as a reliable and trustworthy busi-
ness partner that delivers what is promised.
Irrespective of what kind of internal issues
that are planned and booked since a long
time—if we are needed at the client’s place
this is what guides us. And this may create a
hell internally, to put it straight, but this
priority is very strong’. (partner)
The company is well known for its capacity to
deliver on time. This is an outcome of the amount
of resources in the company making it possible to
put in extra people if needed, but perhaps even
more important is a broadly shared mentality
around the signi?cance of delivery on time. The
company has a strong ‘delivery culture’ making
holding deadlines almost a sacred ideal:
Something that penetrates the culture here
and that you always steer towards is the
deadline. A deadline is a deadline is a dead-
line, and you may almost die before you don’t
deliver on deadline.
This ‘holy’ nature of deadlines means that qual-
ity or ambition level may be adjusted accordingly.
According to one interviewee, in projects ‘quality
is the absolute most important and then ?nances
may come in secondly’. In the same interview, he
also expresses a di?erent position and says that
In a systems development project, the system
needs to work and there you can’t cheat but if
it is a matter of organizational development
or personnel education then you can always
modify how deeply and how broadly you do
things and then you can cheat a bit so that
you keep the plan.
Deadlines are an important vehicle for control
in two ways. One is related to customer control,
locating the locus of control to the client and the
company’s—the project members—obligation to
ful?l the contract. The other is related to ?nancial
margin. Through the emphasis on deadline, people
are under strong pressure to complete the project
so that costs for people’s calendar time are under
control and their labour are available for new
assignments. As people are not paid for over-
time—minor compensations such as a free travel
or a nice dinner are not that costly—the strong
emphasis on deadline/delivery facilitates the
accomplishment of high margins in projects.
Delivery culture is thus much more—and less for
that matter—than being client oriented. This is
illustrated by our observation of a project group
working until midnight to complete an internal
development project, i.e. without a client waiting
for delivery.
430 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
Financial control is tight and elaborate. Project
managers put much e?ort in budgets, as will be
developed in more detail below. Budgets are
monitored and reviewed in a systematic fashion.
Several persons work exclusively as project con-
trollers. Organizational members typically views
the systems as successful, although there are sen-
timents that voices concern that the systems may
be excessive:
‘‘We have very good ?nancial control. We
can account for every hour and every penny.
We also tightly control the business develop-
ment process. There are not many pennies
that vanishes without being accounted for.
The backside is that our ?scal focus tend to
hamper creativity and freedom. And you end
up in red tape and policies instead of thinking
for yourself. And that’s our worst enemy.’’
(Partner)
The HRM system at Global consists of systems
for recruitment, promotion, evaluation and devel-
opment.
Recruitment: selectivity and standardization.
Almost everybody working at Global has an aca-
demic degree. Consultants are mainly recruited
directly from the larger Swedish Universities. All
employees, and even partners, are expected to take
part in various recruitment e?orts, such as pre-
senting the company at Universities, interviewing
job seekers, and generally looking out for people
to hire. Recruitment is part of the control struc-
ture, as it contributes in producing a fairly homo-
genous work force, which presumably shares
certain predispositions and thus is inclined to act
in anticipated ways.
Promotion and career structure. The Firm is
generally understood to be a career company.
Initial advancement is expected to be swift for the
individual. Employees are expected to advance
within 12–18 months to analyst and within 2–4
years to consultants. They then become ‘‘man-
agers’’, i.e. they get this title and gradually are
functioning as project managers. After the man-
ager level, advancement becomes more di?cult.
Those not being promoted are expected to leave
the company in due course. The career structure
makes expectations on individuals explicit and the
pace of promotion regulates, and is seen as
directly linked to, individuals perceived degrees of
success. Thus, promotion and career structure
provide control in two ways: by prescribing the
individual’s preferred career path and grading its
relative success.
Appraisal and evaluation systems. The employees
at Global are under constant performance eval-
uation. Sometimes people in the company talk
about a ‘‘feedback culture’’, but evaluation and
feedback are fairly strictly formalized. Evaluation
is organized in two main processes. First, employ-
ees are evaluated in relation to their individual
development. This process is labelled c-mapping
and is carried out three to four times every year.
The employee’s nearest boss, usually the project
leader of the project on which the employee cur-
rently works, primarily evaluates the employee.
The general idea is that everybody should be
evaluated in similar ways and according to simi-
lar criteria. Thus, there are several tools avail-
able—policy documents, forms, and standardized
software—to ensure that everybody is treated in a
fair and unbiased way. Second, employees are
ranked by their superiors in a process labelled
banding. Banding occurs once a year. The
employee is ranked in category 1, 2 or 3. Band 1
is reserved for top performers. Category 2 is
regarded as acceptable performance. Band 3 is a
warning signal, indicating that you are perceived
to be an under-achiever. Banding in?uences com-
pensation: salary in?uences, career development
and perks.
Development. The people at Global invest a lot
of money, time and other resources in the devel-
opment of the individual, including training,
access to workshops and seminars. They are also
members of so-called competence groups—a
competence group is a number of consultants,
organized around a theme (for example e-com-
merce or performance management) that gathers
regularly and exchanges knowledge and experi-
ences about the theme. Junior members are
always paired with a senior person, who operates
as his or her counsellor—which is also the o?-
cial title for what is generally referred to as
mentors.
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 431
Project management and control
Global are typically involved in two basic types
of projects: small projects that consist of one team
(3–6 person) and Global projects that can include
as many as 50 consultants in one single project.
Obviously, project management and control dif-
fers in scale and scope between these extreme
types, but the process and the tools are basically
the same for both types of projects.
Projects emerge as a consequence of some sort
of customer demand, typically stimulated by part-
ner’s e?orts to sell more or less ready-made
e?orts, so called market o?erings. Partners are
responsible for negotiating fees with customers.
Global use three types of fees: variable fees, ?xed
fees and value billing. Variable fees puts the risk at
the client, who is billed on the basis of actual
hours worked on the project. Fixed fees put the
risk at Global who suggest a price for the whole
project and thus take any additional cost and
pro?t. Value billing is a pro?t sharing scheme,
where Global is guaranteed a minimum sum and
then have a share of the costs savings or the rev-
enue generated by the project. Fixed fee is the
most common pricing scheme currently.
The project manager is always responsible for
planning the project, in terms of manpower,
resources and cost. Project managers are nego-
tiating their budgets with the partner group who
primarily looks at the margin provided by the
project since partners have a direct ?nancial stake
in the company. If the partner group ?nds the
margin unacceptable, project managers are expec-
ted to rework the plan. As a consequence, partners
are not only part of the project, as individuals and
as resources, but they are also operating as exter-
nal stakeholders to the project, as a collective.
Planning and sta?ng. Budgets are produced by
estimating the time span and the manpower nee-
ded. Since the partner group focuses on the mar-
gin provided by the project, resources are
transformed into cost-estimates, in order to pro-
duce an estimated gross result ?gure. This is basi-
cally done by multiplying the manpower needed
with its standard cost. There is a worldwide model
for this calculation. The model uses head count
and a standard price per individual cost as the
independent variable. The standard price per
individual will vary, depending on seniority and
level of competence, thus creating an incentive to
sta? projects with more junior consultants.
The sta?ng is supposed to be made by the HR
department that will allocate people according to
the needs speci?ed by the project manager. In
practice project managers to a high degree decide
about the people he or she wants to use. For
juniors, it is viewed as important to have good
relationships with what they perceive as good pro-
ject managers so that the latter will choose/o?er
them for projects that are attractive to work in.
Monitoring. Projects are monitored by project
managers and project controllers. Monitoring is
facilitated by various information systems, which
allow project managers (and controllers) to moni-
tor the numbers down to single individuals. Pro-
ject controllers are external to the project. In
essence, they are the partner groups extended arm
into the project. Project controllers also provide
some support to project managers but mainly
operate as a controlling body.
Deviations are dealt with as they emerge. There
is, to some extent, a certain slack in budgets,
which is used to manage expected deviation.
Typically, if deviations can be accounted for in a
rational and under stable way, it is claimed that
there will be no negative e?ects for the project
manager:
You always inform your superior’s if it
becomes clear that you are going to bust the
budget signi?cantly. And when the project is
?nished you’ll make a fee adjustment. You
recalculate the project and point to the
deviations, which occurred because of that
and that, and so on And sometimes you will
have some one from way up in the hierarchy,
a global boss, that calls you and wonder why
you didn’t provide better forecasts and plans.
But I have never have any problems, nor have
I heard about anybody else having problems,
if you can explain why the project didn’t
become a success.
Evaluation. Project evaluation is mostly con-
ducted in cooperation with customers, in particular
432 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
when it comes to evaluating the relative success of
the implemented solution. The impression of success
is important for market reasons. As a consequence,
the team from Global, and the project manager in
particular, typically tries to talk down customer
expectations initially in the life span of the project in
order to exceed them with the delivered solution.
From an internal point of view, i.e. besides the
client being satis?ed, projects are exclusively eval-
uated in ?nancial terms: to achieve or exceed the
projected margin will lead to a star in the proto-
col. If the project manager has dealt with the sub-
ordinates in a way leading them to be frustrated or
leave, this is not monitored, although occasionally
it may be noted.
Projects that have been viewed as particularly
valuable form an experience point of view are also
documented and added to the company knowl-
edge management system. The knowledge man-
agement system thus mainly consists of write-ups
(usually in the power point presentation form) of
projects, which then is re-used as inputs in market
o?erings or new projects. Project outcomes thus
provide inputs to further frame control.
Excesses of control: holy margins, fake time report
cards, and the ‘‘ghosting’’ phenomenon
As is evident by the account above, the control
systems at Global are elaborate and manifold.
Since control is exercised in so many ways, from
so many directions and on so many di?erent
levels, control systems, taken together, sometimes
create situations where individuals ?nd themselves
with con?icting sets of instructions. As we will
argue here, such situations are highly instructive,
because they set members, as well as researchers,
straight on the priorities in the company.
The control system at Global will, in fact, reg-
ularly create a situation for organizational mem-
bers where they almost routinely will provide the
control systems with fake numbers. To understand
what they are doing, and why they are doing it,
will, we claim, (a) provide insight on how control
actually is performed at Global, rather than nom-
inally, and (b) provide perspective on how numer-
ical forms of control may be re-interpreted in
terms of socio-ideological control.
We stumbled on this peculiar phenomenon
when we participated in a management conference
day, where most of the upper echelons of the
hierarchy at Global’s national subsidiary were
present. An excerpt from the ?eld notes [written
by the second author, who participated in the
event] follows below:
(The excerpt starts at a demonstration of the
new balanced score-card application) Thereby
the bottom line is dealt with and the managing
partner (MP) directs the attention towards the
‘‘softer’’ performance indicators—he con-
centrates on personnel turnover and overtime
reporting.
Ludvig, a senior manager, raises his hand and
asks simultaneously: ‘‘What’s that number by
the overtime?’’
I glance at him. He looks slightly annoyed.
MP takes a step back and says: ‘‘Yes, there
could be a lot of talking about that number.
It’s not correct of course [the number says
that there is 2% reported overtime]. The
number creates some problems for us in the
relationship with the other o?ces since they
could get the impression that we are not working
[overtime]. But we know we do. So the reason
is of course that the reporting is incorrect.’’
Ludvig: ‘‘Are you going to do anything about
it?’’
‘‘Yeah, it’s not so easy to know what we
ought to do. This started to happen after the
overtime compensation was taken away. And,
as you remember, we took it away because it
could create absurd consequences. There were
cases when an analyst made more money than
more experienced managers. But the issue is
complex.’’
Partner: ‘‘It would be very unfortunate to re-
introduce the overtime compensation’’.
Voice in the audience: ‘‘The problem is that
there are no consequences for those who
report too little overtime. It encourages the
wrong behaviour since it is easier to write ‘8
hours’ no matter how much you have
worked, than to keep track of your hours.’’
A short discussion concerning the time
reporting begins, and, essentially, there are no
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 433
more concrete statements. But the partners
are throwing glances at each other all the
time, and MP, in particular, is frequently
looking in the direction where the partners
are. It is obvious that the issue already has
been raised among the partners, and it is
equally obvious that they do not want to
discuss the issue in this forum. After lunch
there is a division into di?erent groups based
on themes suggested by the participants. One
theme concerns the ‘‘overtime problem’’. I
decide to participate. It turns out to be a
small group. After some disorder we end up
being four persons in the group; two women,
one man and I. I ask them to explain the
problem. It comes out that there is a very
good reason for Global-consultants to report
incorrectly. Their managers demand that
they do not report more than eight hours a
day; irrespective of the number of hours
they actually work. The reason is very
simple. The time reported controls the most
important ratio/*measure for the evaluation
of the project, the project margin. The
higher the amount of reported hours, the
lower the margin. But this is just an internal
ratio. Most projects are sold based on a
?xed price. They thus know exactly how
much the revenue will amount to. And since
there is no overtime compensation they also
know how much the expenses will amount to.
The actual margin is thus not a?ected by the
time reporting. But a consultant appears
much more competent if she or he does not
report more than eight hours a day. This,
however, does not mean that they are not
working more than eight hours. To work
more hours than reported is so well known
that there is even a name for it: in Global-
jargon it is called ‘‘ghosting’’. After having
informed me about the situation, Helen and
Nina start, in the spirit of good problem
solving, to analyse the di?erent components
of the problem. They use a large sheet of
paper and start to write down points. The
headline is: ‘‘Why don’t people report their
time?’’ Nina and Helen bring the most
obvious points to light:
It is easier to write eight hours than
to report in detail
You want to seem good and e?ective
The margin is controlling the project
Employees are loyal to the project
Meanwhile, Nina is telling me about a col-
league who writes her time reports for four
months at a time. Helen: ‘‘I write the correct
time on the ?rst report. Certainly, there is a
manager who tells me to write eight hours
instead. I think: ‘Fine, I don’t care’.’’
She continues: ‘‘The whole thing depends
upon the fact that there are no incentives to
report the correct time. Who are you loyal to,
really?’’, she asks rhetorically. She answers
the question herself: ‘‘In the ?rst place you
are loyal to yourself. In the second place your
loyal to the project, and only in the third
place are you loyal to the ?rm.’’ Helen’s loy-
alty hierarchy is written down.
Nina: ‘‘I think people would report the correct
time if there were only one possibility to save
and allocate time. Now this is handled entirely
informally. If I need to take an afternoon o?
and I have ‘saved time’ so that I can take the
time o? without problem, then there has to be
damn much negotiating to make it work.’’
Helen is unsure if it would help: ‘‘I don’t
think there’s going to be any change until
there is something in it for the individual.’’
Nina: ‘‘But I think this kind of system would
be good for the individual too. ‘‘
Helen: ‘‘I don’t believe the individual would
think it’s enough. There must be an incentive.
I can’t see the incentive in this.’’
I mention a few words about the time
report systems that they have in larger
companies. Nina is listening with interest
but Helen seems to be sceptical about the
bene?ts of that kind of system.
Sven looks thoroughly at the points: ‘‘But
there has to be more to this. Why does the
margin have to be so high? Why do people
have to show o??’’
Sven, Nina and Helen discuss that question
for a while. They agree that it is a side e?ect
of the evaluation systems. People want to
434 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
appear to be good and e?ective because it
increases the chance of getting good evaluations.
The project managers, on the other hand,
want a high margin, because the projects are
evaluated based on their achieved margin.
The ?ipside of the coin is that people work
themselves to exhaustion and that they are
actually selling themselves with a discount.
Sven: ‘‘If we use the hours as the basis for the
price, and assume that we work 40 hours a
week when we actually work 60 hours, then
we are selling ourselves with a 25% dis-
count.’’ The others give a nod of assent.
Nina adds: ‘‘But the problem is probably a
di?erent one, actually. Do we really want to
sort this problem out?’’
In the meantime, our little group is slowly
growing. After a while we are about ten
persons who enthusiastically participate in
the discussion. One of the newcomers says: ‘‘I
think something has to be done, but I won’t
suggest anything. I got burned on the time
reporting thing. In my ?rst project there was
a crisis and we worked like crazy. Sixteen
hours a day, at a minimum. I tried to report
all my hours but the manager couldn’t accept
that I reported more than eight hours. Then I
worked eight hours. The manager got furious
and had me executed in the next evaluation.
The others look at him as if they couldn’t
believe their ears. ‘‘You worked eight hours?
And you’re still in the ?rm!?’’ ‘‘Who was it?’’
someone asks curiously. ‘‘I don’t want to
tell’’, the newcomer replies.
Helen throws a sceptical and evaluating
glance at him: ‘‘I think you would have been
more clever if you had worked.’’
The discussion starts to become technical. It
is clear that there are quite a lot of internal
systems that are built up around the time
reporting. I silently wonder how they work
since nobody seems to report the correct time.
Helen is expressing this as she once again
wants to stress the argument that the indivi-
dual has to get something out of a possible
change: ‘‘Everybody is ghosting. Everybody.
And everybody will be continuing until they
get a good reason to quit. I will be continuing
it they [the project managers] ask me to. And
they will because their evaluations depend on
the margin, which in its turn depends on the
hours. I don’t mind. I don’t give a damn.’’
It is starting to get rather crowded in our little
room. Janne, the person who is in charge of
this particular activity during the day, has
arrived. Except for him, there are few senior
persons. They don’t seem too eager to discuss
this issue o?cially. Janne, however, is senior
enough to know what it looks like on the other
side. He has been working as a project manager
and someone asks how he is dealing with this.
‘‘Well, it happens that I say we write eight
hours.’’
‘‘Irrespective of the amount of hours we put in?’’
‘‘Yes. It has to do with the fact that you don’t
want to go to the partners with a margin that is
way under the budgeted one. And it’s damn
hard to get a project if you don’t have a good
margin. So in order to keep the margin you tell
people to write eight hours.’’
‘‘But everybody knows that is can be manipu-
lated’’
‘‘It doesn’t matter. That’s the way we do it.’’
At the assembly afterwards, representatives of
the discussion group present an idea including
skipping reporting the number of work hours
but after some resistance of senior people, this
idea is dropped.
From a control point of view the excerpts pro-
vide many illuminating insights. First, it demon-
strates the hegemonic in?uence wielded by the
partner group, and thus the signi?cance of hier-
archy. Although partners may have disagreements
in private they operate as a single entity towards
the rest of the company, at least in public. This is,
in particular, demonstrated in the ?rst excerpt,
where it is quite clear that the partner group will
not discuss ‘‘the overtime problem’’ outside the
partner group, and also, perhaps more vividly, in
the last, where partner intervention e?ectively puts
an abrupt end to the discussion on ‘‘the overtime
problem’’ at the conference, at least publicly.
In fact, ‘‘the overtime problem’’ is in a sense
created by the partner’s hegemonic exercise of
control, through choosing to almost exclusively
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 435
focus on one (of many) key performance indica-
tors: the project margin. Although project man-
agers and others may have several performance
indicators at their disposal, it is quite clear that the
partner group pay most, if not all, attention
towards the project margin. This focus has ripple
e?ects. Since it is important for project managers
to look good, in order to get good evaluations,
they are disinclined to do anything that will
‘‘destroy’’ the margin. Since project participants
have ?xed salaries, and will not be compensated in
relation to actual work hours, ‘‘ghosting’’ provide
a mechanism for project managers to both pre-
serve the project margin and manage uncertainty
through making it possible to extract more work
hours, without incurring costs. For project man-
agers, ghosting makes perfect sense.
Second, the excerpt demonstrates the extent of
subordination and compliance expected by project
members. ‘‘Ghosting’’ is clearly acknowledged as
a common practice, to the extent that it, at least
for some participants, seems to be routine. But
‘‘ghosting’’ is also clearly a practice that is dubious
from a moral point of view, since it involves lying
and faking. The moral aspect is highlighted by the
person who refused to fake the numbers, and
insisted on working the reported hours. His man-
ager’s reaction, and the audience reaction in the
excerpt is instructive: The manager ‘‘executed’’
him in the evaluation, and the audience is stunned,
partly because he, as some one puts it, is still at the
?rm, partly because, as another one puts it, it was
the wrong thing to do: he should has solved the
dilemma by reporting the required hours, instead
of working the required reported hours. Obedi-
ence is the norm.
Third, the opportunistic mindset, doubtless at
least partially an e?ect of HRM practices ?ne
tuned towards career success, is literally spelled
out in the excerpt. This is, of course, highlighted
most visibly by Helen’s ‘‘hierarchy of loyalty’’ but
also in the emphasis on making it worthwhile for
individuals to report the overtime correctly. It is
asserted that nothing will ever happen without the
proper incentives for individuals. Ghosting is a
practice that ?ts rather nicely with an opportunis-
tic and instrumental mindset, in particular when
supported by the systems of evaluation at Global.
To help the project manager to look good will
undoubtedly help your project evaluation and
consequently your reputation. Your cost is negli-
gible, and mostly a matter of conscience. As a
matter of fact, it might hurt you not to cooperate,
as demonstrated by the participant who was
‘‘executed’’ in his evaluation when he refused to
fake the numbers.
From a narrow management accounting point
of view, it could perhaps be argued that fake
time-reporting does not matter that much, as it is
not a?ecting ?nancial costs, as neither revenues
nor wages are a?ected by whether an employee
does a particular job in 40 or 60 h. But although
not registered by the ?nancial measurement sys-
tem, there are huge costs involved, if a cost is
related to the use of ‘real’ resources. On average,
the overtime is probably 30–50% and not 2%, but
it varies much between di?erent projects. One can
here talk of the alternative costs, as the time spent
on ‘high-ghosting’ projects to some extent could
have been utilized on other projects or on compe-
tence development, etc. This is hidden as it may
appear as if all the overtime is for ‘free’, and is so
for the project manager. There are, of course, also
costs following from the high turnover that pres-
sure to work a lot of overtime leads to. Other
problems, with more indirect consequences for the
company’s pro?ts, are mentioned in the interview
excerpt.
Interpretations
Movements of control: interfacing technocratic
control, socio-ideological control and self-control
On the whole, the level of compliance at Global
is high. People generally seem obedient and hard-
working. They deliver on time, provide high mar-
gins on projects and avoid making clients
dissatis?ed. It can be seen as an outcome of the
fairly elaborate systems and practices for beha-
vioural and outcome control, as a sign of the vic-
tory of formal management control. However, we
do not think this exclusively or even mainly should
be interpreted as the triumph of technocratic con-
trol focusing on behaviour and output. There are
436 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
considerable loopholes in the structures and sys-
tems, as reported by the time reporting practices.
But also to the extent that the formal systems
work more accurately in relationship to their tar-
gets, it is not clear to what extent they are precise
or powerful enough to assure the prescribed
behaviour materializing.
A possible consequence of sub optimal behaviour
is poor assessment and a low ranking, which could
a?ect the wage negatively in annual revisions. But
this is not a ?ne tuned vehicle for behavioural
modi?cation. It could be argued that partners,
through two vital sources of control can create
obedience: the careful monitoring of results and
the right to decide who will become partners. But
project managers are not simply dependent on
partners in any one-dimensional way: partners
rely heavily on the project managers to deliver the
goods. And juniors may not feel that they should
not necessarily work very hard so that the project
manager should appear good in front of partners.
Deviations from the prescribed margin may
mainly fall back on the project manager, and be of
moderate concern to the other project members. A
moderate, even high sense of loyalty and the desire
to do a good job, and a moderate fear of punish-
ments (delayed rewards) from above, would not be
su?cient to account for people sometimes work-
ing 60–70 h per week. The hierarchical structure in
the company does, however, seemingly hold a
strong grip over people.
The employees are not that extremely dependent
on the employer, but could easily protest with
their feet: exit is one possibility. Given that they
are highly educated people with generally a high
self-esteem and excellent exit possibilities, one
could imagine that they could also protest through
voice. One could imagine a group of employees
complaining to senior people about project man-
agers trying to get them to fake time reporting.
Managing a project that does not meet the margin
can often be explained. Every project manager
must have access to a variety of speci?c circum-
stances that can account for less than optimal or
below calculated results: unexpected complexities
around the task, employees quitting or becoming
sick during the project, problematic customers,
etc. As one informant says, if you just have a good
explanation for additional costs, you are typically
excused. But this option is seldom used—which
makes the company di?erent from most other
companies, which are less religious about missing
a deadline or a budgeted margin.
As the technocratic control systems hardly in
themselves can fully account for the extreme
receptivity to and compliance with goals and
rules, we need to go further to ask how this is
accomplished.
The technocratic production of norms and meaning
In order to understand the high level of com-
pliance at Global, we must go further and investi-
gate the ideas, expectations, values and identities
that are created at the company and contribute to
a high level of responsiveness to demands from
seniors and management control systems. To some
extent, these can be illuminated through the con-
cept of socio-ideological control, but we also feel
that this is insu?cient and that we need to go
beyond it.
A number of distinct expressions of socio-ideo-
logical control are easy to detect at Global. Mer-
itocratic ideology puts strong imprints on the
beliefs and identities of people in the company.
This overlaps with performance anxiety, a strong
emphasis on career, instrumentalism and elite
identity constructions. Titles and promotion speed
are strongly valued. All this seems to contribute to
an extremely strong inclination to put in long
working hours. It is also connected to, indeed a key
element behind, a very high level of compliance.
In many other organizations, highly educated
people with considerable status and self-con-
?dence give some weight to autonomy and pro-
fessional ideals that at least sometimes lead to
tensions in relations to management control
arrangements and corporate demands on sub-
ordination (e.g. Covaleski et al., 1998; Empson,
2001; Robertson, 1999). At Global there is very
little of this. People subordinate themselves to
corporate demands and/or the wishes of superiors.
In Global there are strong claims about an elite
being employed. They charge clients huge
amounts of money for their services, being con-
vinced that the client gets a lot for the costs. The
large amount of corporate advertising and general
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 437
emphasis on nurturing corporate image con-
tributes to the production of a salient social iden-
tity associated with corporate belonging and even
a heightened sense of corporatively mediated self.
Title, formal position, signs of one’s relative
worth, here strongly matter. On the other hand,
the fairly elaborated control systems and high
pressure on—or internationalization of—sub-
ordination to the corporate regime is at odds with
an elite identity. There are uncertainties and ten-
sions around worth, leading to a strong inclination
to draw upon resources for securing a stable sense
of self, such as receiving promotion before one’s
peer group and look good in all assessments. Self,
socio-ideological and technocratic control thus go
together. Rather than assuming that these exist in
pure forms, the interfaces of control forms need to
be emphasized.
Values that Global employees adhere to include
money, formal status and brand name—all exter-
nal and highly visible signs of status and success.
Growth is very important and partners have
strong incentives to increase the sales. Not much
energy or skills are invested in socio-ideological
control as a special area of impact. Global exhi-
bits very little of the characteristics of Min-
tzberg’s (1983b) missionary organization and
there is not much of a corporate culture ‘outside’
regimes of bureaucracy and output control.
However, the behavioural and output oriented
forms of control are supported by ideas, beliefs
and values indicating what exists, what is good
and what is worth striving for, i.e. ideologies
(Therborn, 1980). This tends, however, to be
subtle and expressed within the use of techno-
cratic forms of control.
We suggest that these forms of control function
in a di?erent way than commonly assumed. Their
accuracy in terms of measuring and regulating
behaviour or accomplishing certain performances
through output measurement is limited. So is at
least the case with the accounting systems built
around or a?ected by time reporting. This is also
the case with the HRM system which hardly is
capable of creating a di?erentiation and ?ltering
system leading to the perfect hierarchy, in which
everybody is located according to competence.
When pushed in interviews, people’s perception of
promotion and how people are placed in di?erent
ranks do not correspond that well with this
ordering logic.
The impact of the HRM systems and practices is
not necessarily only or even mainly about locating
people on the right career level and improving
them. The impact may be as much in terms of
a?ecting the motivation to perform in ways com-
pliant with the control system. The motivation is
thus moulded after the regulative powers of the
HRM ingredients. Competence is vague and hard
to assess. The same goes for motivation, but
interviewees’ reports and also our direct observa-
tions indicate very long working weeks. So irre-
spective of how the company a?ects people’s
competence, people make strong e?orts in order
to get favourable indications of their value in
assessments.
The KM system also seems to work partly in
other ways than as a rational system for the stor-
ing and re-use of knowledge. It does not primarily
works through the re-cycling of knowledge or the
use of earlier experiences and knowledge inputs as
templates for a new project. The KM system
emphasizes that people may be contacted for dis-
cussions and aid. People are expected to be able to
phone or e-mail other people within the company
and if these can assist, they do so (Alvesson &
Ka¨ rreman, 2001a). Similar to the HRM system, its
workings are mainly through producing certain
orientations and values.
There are good reasons to emphasize the sym-
bolism of the technocratic systems and proce-
dures. They communicate certain ideals. They
draw attention to certain dimensions and frame
consciousness in a particular way. The intensity of
HRM activities does not necessarily deliver the
goods in any rational sense: a highly competent
workforce, carefully distributed according to abil-
ity. But all the systems, procedures, acts and talk
about feedback, assessments, the frequent promo-
tions, ?ne-tuned hierarchical di?erentiation sys-
tem, and the people improvement activities seen as
closely related to these—education, self-improve-
ment e?orts—signal the presence of an impressive
apparatus for the screening of highly competent
people, the improvement of them and the re-
screening of them at frequent points in their
438 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
careers. HRM symbolizes people development and
indicates competence. All this HRM means that
hierarchy is viewed as a re?ection of competence,
nothing else. People thus connect their self-esteem
and ambitions closely to the HRM systems and
the organizational hierarchy. They are more
inclined to accept hierarchy and comply with it
than in situations where it is not made credible by
the HRM apparatus and the meritocratic ideology
it simultaneously draws upon and reinforces
(Alvesson & Ka¨ rreman, 2001b).
Normative ideals as templates for self-understanding
and identity construction
To some extent associated with meritocracy is
an orientation of comparison and competition.
The company employs a homogenous group of
people in terms of age, educational background
but also stated motives and attitudes towards
work life. These people are then placed in various
categories (1, 2 or 3) and promoted with di?erent
speed. The combination of the recruitment policy
and a meritocratic ideology fuels competitiveness
and strong focus on comparisons. This forms an
important motive for performing and, in parti-
cular, getting signs of one’s relative worth from
superiors. As there is a general faith in the assess-
ment and promotion system, and self-esteem is
tightly connected to its outcomes, people tend to
want to give favourable impressions to their
superiors. The contradictory experiences of the
individual—highly visible in assessment, rewards
and punishments combined with a view of the
individual being exchangeable in a standardized
and impersonal system—also fuel a desire to
secure a positive sense of self. Any signs of success
are strongly valued in the light of contradictory
experiences.
The formal information and control systems
and processes in the company interact with this
corporately created self in a variety of ways: (1)
The systems direct attention and encourage a par-
ticular outlook and mentality, in this way the for-
mal control structures, although intended to target
behaviour and outcomes, also exercise cultural con-
trol, (2) Construction of self is partly accomplished
through subordination to control systems and e?orts
to employ these as resources in securing a self, (3) the
socialization and reproduction of a control-, feed-
back and delivery culture (or a set of cultures)
includes a ‘naturalization’ of a set of guidelines for
how to relate to corporate practices, that means an
acceptance of the control arrangements.
The possible con?ict between integrity and con-
?rmation of one’s self within a corporately de?ned
domain is, for most employees, not a very serious
one, although there is some irritation and need to
rationalize one’s willingness to fake. As Helen
says, when told to write 8 hours of working day,
‘?ne, I think, I don’t care’.
An intensive system of control encourages
instrumentalist orientations. A feedback system
pointing at a particular set of qualities/imperfec-
tions may lead the individual to change in accor-
dance with what is prescribed. This is di?erent
than if e?orts to change are grounded in intrinsic
interests or hedonistic motives. The scope and
intensity of structures and systems instructing
individuals what to do and claiming to carefully
monitor their performances and progress all draw
attention to means to be used to accomplish ?xed
goals. There is less explicit focus on values and
ideals, on the ‘pure’ ideational arena, i.e. outside
formal systems. Generally in the company, a
technocratic mentality is encouraged. Incentives
are viewed as central. The people at Global say
that time reporting is an outcome of incentives. At
Global, we ?nd that Economic Man and Woman
are at work, at least according to the self-under-
standings of people in the company. Without clear
incentives, people will fake—they will ask sub-
ordinates to fake in order to look good for their
superiors and the subordinates will respond to the
request for cheating in order to please their project
manager. Integrity appears not to matter that
much. The person that shows reluctance to com-
ply is viewed with surprise by his colleagues. What
may in other groups be seen as a moral virtue—
honesty, standing up against somebody requesting
one to fake—is interpreted almost as a sign of
stupidity.
The corporate members of Global appear as
incentive-driven. Pay and promotion are very
important, interviewees say. It is worth pointing to
the relative value of these rewards. The pay and
speed of promotion in relationship to one’s peer
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 439
group is signi?cant, indicating the role of the
symbolism and self-de?ning elements of pay and
promotion.
It is important here to emphasize that the
orientations of people in working life is not given.
It is not self-evident that pay and promotion are
central for people or that compliance is the only
route to success. Autonomy, strong social rela-
tions, interesting work tasks, integrity and self-
respect can play a role. In many companies there
are con?icts between professional and corporate
ideals and often people are not willing to sacri?ce
professional values and autonomy for increased
?nancial rewards (Covaleski, et al., 1998). Again
in other companies social climate and intrinsic
motives for working in the company are empha-
sized and hierarchy is viewed pejoratively (Alves-
son, 1995; 2000; Alvesson, Robertson & Swan,
2001).
Conducting control through constructing the
(controllable) person
Some researchers have investigated how di?er-
ent forms of knowledge and practice make possi-
ble the governable individual (Miller & O’Leary,
1987). Our approach is di?erent as we focus not
on how institutionalised discourse includes a par-
ticular construction that encourages certain plans
and arrangements, but seek to investigate how
subjectivity ‘empirically’ is being formed. (Miller
& O’Leary, like many other Foucauldians, are
interested in formal texts constructing subjects,
rather than the speci?c encounters and reactions
shaping subjectivity.)
It is important to emphasize how motives and
selves are constructed in organizations. There are
not just once and for all established needs, expec-
tations and preparedness to rank di?erent sources
of satisfaction in a clear manner, nor is there a set
of rewards to which people respond in a straight-
forward way. What motivates people is partly a
matter of how they construct themselves—gov-
erned by various agents in the organizations they
work in—and how they construct di?erent advan-
tages and frustrations in their work, and, in parti-
cular how they construct the self-work
relationship in terms of ?t, satisfactions and frus-
trations. How people, in particular relatively
young persons, relate to the work world is an open
project.
1
The desire to ‘succeed’ is not a self evi-
dently dominant idea. It is an outcome of social
processes. So is, of course, the idea of being suc-
cessful in a particular way, given a speci?c set of
criteria and constraints. In Global success is
equated with promotion speed, ranking and play-
ing the game within the rule set promulgated by
the company: including keeping the deadline and
keeping the margins.
As the theory of social information processes
(Salancik & Pfe?er, 1978) says, di?erent under-
standings of needs and indications of the elements
of signi?cance for job satisfaction are made salient
as a result of in?uences in the social environment.
Management, as well as other groups and agen-
cies, may in?uence people’s understandings of
their needs, their workplace situation (job char-
acteristics), etc. In a ‘high-intensive organization’
such as Global, the social information processes
are relatively coherent and persistent.
The motivation for carrying out tasks and con-
tinuing in an employment are not only a matter of
instrumental or even ’hedonistic’ motives (such as
pleasure from the activity in itself), but the moti-
vation may come from the meaning for the indi-
vidual, e.g. in terms of the a?rmation of his or her
identity and collective a?liations (Shamir, 1991).
Also moral obligations may serve as motives:
’moral internalization turns constraints into pre-
ferences’ (Etzioni, 1988:46). One may act in order
to feel in line with one’s cherished beliefs and self-
image and also follow the imperatives of the group
with which one identi?es. The ’motive’ is then to
maintain and enhance self-esteem and a sense of
existential security in the world, a motive that in
corporate contexts is exploited by and ‘satis?ed’
by forms of power o?ering control as a sense of
security (Alvesson & Willmott, 2002; Knights &
Willmott, 1989). The operation of this form of
power in the case of Global is salient.
1
Of course, poststructuralists suggest a radical de-centering
of subjects, an ‘emptying’ of them on any substance such as
needs or wants of a ?xed nature (e.g. Shotter & Gergen, 1989).
This paper goes some way along a de-stabilisation route, but
does not adhere to any strong anti-humanist position and does
not use poststructuralist vocabulary.
440 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
Motives are then closely related to identity and
again to forms of social regulation. Identity is a
certain set of understandings of what is appro-
priate and natural. Who am I and—by implica-
tion—how should I act? are questions answered by
the constructions of identity. In Global, as in large
parts of contemporary working and business life,
identity gets twisted into struggling with questions
such as ‘What I am worth?’ and ‘Am I performing
according to standards?’. Templates such as keep-
ing deadlines and achieving the budgeted margin
must be understood in this context.
One aspect of the motivation/identity interface
is the inclination to maintain and enhance one’s
identity and self-esteem. Another aspect means
that identity in?uences how various driving forces
are appreciated. A person who, through educa-
tion, training and on-going social relations,
experiences herself strongly as a professional may
as a result give priority to certain values—e.g.
autonomy, knowledge-development, specializa-
tion—as an outcome of the identity rather than an
e?ect of the pure motivating power of instru-
mental and hedonistic sources of grati?cation.
In ‘high-intensity’ organizations such as Global—
which make strong claims on the employees in terms
of time spent on and energy put into work—the
identities of the personnel are constructed and re-
constructed in a more powerful way than in organ-
izations that have lower pro?les and express less
distinct claims on the employees. The multitude
and variety of formal systems of control—a?ecting
behaviour as well as subjectivity—contribute to
this intensity and a heightened subjectivity associated
with the workplace and being an organizational
person. As the individual stands out so clearly as
an object of frequent assessment, feedback and
hierarchical relocation through systems and rules
focusing on behaviour, performance outcomes,
capacity and potential, feelings of personal worth
and identity become closely tied to the indications
produced by these systems.
Conclusion: interfaces of control
In Global there are strong beliefs in the capa-
cities of formal systems of KM, HRM and man-
agement accounting to deliver e?ective control
and the accomplishment of what is viewed as
desirable behaviour and outputs of various types.
However, there is considerable ambiguity
regarding what these systems really can accom-
plish in terms of behavioural and output control.
It is interesting to compare the partner cited ear-
lier claiming that ‘We have very good ?nancial
control. We can account for every hour and every
penny’ with the overwhelming evidence that time
reports are faked. Equally interesting is to note
how people can claim that the HRM system and
?ne-tuned organizational hierarchy means that
competence is mirrored in the social di?erentia-
tion system, while still expressing doubts about the
quali?cations of a number of those promoted to
managerial and partner levels.
As we have demonstrated above, almost all
control practices at Global include aspects of cul-
tural engineering more or less geared towards
organizational members’ self-de?nitions. Evalua-
tions, both formal and informal, include aspects of
conduct: for example, whether one has been a
team player or not, whether one has been ‘‘con-
structive’’ or whether one is considered to be a
‘‘showstopper’’; ‘‘constructive’’ here meaning
being cheerful rather than negative in ones con-
tribution to discussions, and ‘‘showstopper’’ here
roughly meaning working less hours than the rest
in the team. Competence groups are occasions for
intense socialization, where re-enacting Global
values and meanings is almost as dominant as a
theme, as acquiring quali?cations. Even such a
straightforward thing as the formal career struc-
ture is used as a template for self-de?nition.
Socio-ideological control is thus intimately tied
to bureaucracy and output control. It is not, as
claimed by most of the literature on control (e.g.
Ouchi, 1979, 1980), an alternative to the latter
two, useful in situations where complexity and
uncertainty make rules prescribing behaviour and
the precise measurement of results impossible. In
businesses like Global we have in a sense a lot of
complexity and uncertainty, but this has been
counteracted with e?orts to create a vast bureau-
cratic and output measuring apparatus.
This apparatus works simultaneously badly and
well. The ghosting ‘problem’ indicates that the
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 441
accounting system presents a highly distorted ver-
sion of the use of ‘real’ resources—people’s work-
ing time. The uncertainty involved in this kind of
knowledge work means that the use of time and
energy to solve problems and deliver solutions are
di?cult to predict, and the technocratic control
systems fail to deal with it. Ghosting can be inter-
preted as a pragmatic response to an obvious
problem. It is, however, also a problem: the prac-
tice wrecks havoc with the formal control and
evaluation systems, it compromises the balance
scorecard, and it makes project evaluations unre-
liable since it typically under-estimates actual
work hours, it might contribute to frustration and
cynicism and the company may sell itself more
cheaply than it thinks. Other problems regard bad
handling of junior people—often exposed to a lot
of stress and exploitation—and the company
being weak in creativity. These are well known
negative sides of (machine) bureaucracies. More
positively, for the partners and partners-wanna-
bees at least, are the e?ciency and business suc-
cesses of the company. Customers apparently
think they get value for money and the partners
become rich. It is far beyond the scope of this
paper to account for the company’s results, but we
can point at the signi?cance of the hard labour
and discipline of the employees of Global.
Socio-ideological control matters strongly here.
Technocratic forms of control work e?ciently as
hammering out vehicles for messages that ?nd a
receptive audience in the selves gradually formed
by the control, delivery and feedback cultures of
Global. The systems do not function entirely—
perhaps not even mainly—through precise book-
keeping of revenues and costs (including how
people spend their time), through the ‘objective’
selection, developing and ranking people or
through the use of earlier projects as templates for
new ones. Instead, through their systematic and
intensive communication, they are normatively
e?cient. They work in relationship to and con-
tribute to the shaping of a speci?c ‘interpretive
community’, positive to techniques and an instru-
mental logic. The magnitude of and frequency of
the control systems then contributes to the pro-
duction of a regulated social order and indivi-
duals. This does not, however, work in a
mechanical way according to an ‘objective’ logic.
The clan-like input to this gradually increasing
receptivity is important here: selective recruitment,
socialization and ritualised, ceremonial forms of
control (Ouchi, 1979:838). In particular, selective
recruitment and a highly homogenous work force
making competition and comparison important
ingredients in career aspirations and self-assess-
ments. The ceremonies of Global are couched in
technocratic rather than ‘corporate culture’ forms.
Still, the symbolism and powers of these forms in
shaping thinking, feeling and identifying are central.
We can, perhaps, talk of Global as a technocrati-
cally framed clan. As such it is probably not unique.
We can also identify bureaucracy and output
measurements as potential sources of hidden
socio-ideological control. Again, this has a poten-
tial in broad relevance. This study therefore o?ers
food for the re-thought of established manage-
ment control ideas and categorizations. In parti-
cular, the paper questions common ideas on the
existence of pure and alternative forms of control
and the assumption that technocratic and socio-
ideological controls are mutually exclusive and
function in di?erent organizations and situations.
Looking at the cultural dimension of bureaucracy
and performance measures—both in terms of what
is communicated and how the messages are being
interpreted—seems a productive way forward.
References
Alvesson, M. (1987). Organization theory and technocratic
consciousness. Berlin/New York: de Gruyter.
Alvesson, M. (1993). Cultural-ideological modes of manage-
ment control. In S. Deetz (Ed.), Communication yearbook.
Vol. 16. Newbury Park: Sage.
Alvesson, M. (1995). Management of knowledge-intensive ?rms.
Berlin/New York: de Gruyter.
Alvesson, M. (2000). Social identity and the problem of loyalty
in knowledge-intensive companies. Journal of Management
Studies, 37(8), 1101–1123.
Alvesson, M., & Ka¨ rreman, D. (2001a). Odd couple. Coming
to terms with the curious concept of knowledge manage-
ment. Journal of Management Studies, 38(7), 995–1018.
Alvesson, M., & Ka¨ rreman, D. (2001b). The perfection or
meritocracy or ritual of bureaucracy? HRM in a management
consultancy company. Paper presented at 2nd International
conference on critical management studies, Manchester July
2001.
442 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444
Alvesson, M., Robertson, M., & Swan, J. (2001). The best and
the brightest. Construction of elite identities in knowledge-
intensive companies. Paper presented at 2nd International
conference on critical management studies, Manchester July
2001.
Alvesson, M., & Willmott, H. (2002). Identity regulation as
organizational control. Journal of Management Studies,
39(5), 2002.
Barker, J. (1993). Tightening the iron cage: concertive control
in self_managing teams. Administrative Science Quarterly, 38,
408–437.
Barley, S., & Kunda, G. (1992). Design and devotion: surges of
rational and normative ideologies of control in managerial
discourse. Administrative Science Quarterly.
Becke´ rus, A
?
., & Edstro¨ m, A., et al. (1988). Doktrinskiftet. Nya
ideal i svenskt ledarskap. Stockholm: Svenska Dagbladet.
Beyer, J. M. (1981). Ideologies, values, and decision-making in
organizations. In P. C. Nystrom, & W. H. Starbuck (Eds.),
Handbook of organizational design. New York: Oxford
University Press.
Burris, B. (1989). Technocratic organization and control.
Organization Studies, 10, 1–22.
Clegg, S., et al. (1996). Management knowledge for the future:
innovation, embryos and new paradigms. In S. Clegg, &
G. Palmer (Eds.), The politics of management knowledge.
London: Sage.
Czarniawska-Joerges, B. (1988). Ideological control in non-
ideological organizations. New York: Praeger.
Covaleski, M., et al. (1998). The calculated and the avowed:
techniques of discipline and struggles over identity in Big Six
public accounting ?rms. Administrative Science Quarterly,
43, 293–327.
Deetz, S. (1992). Democracy in an age of corporate colonization.
Albany: State University of New York Press.
Edwards, R. (1979). Contested Terrain. The transformation of
the workplace in the twentieth century. London: Heinemann.
Empson, L. (2001). Fear of exploitation and fear of con-
tamination: impediments to knowledge transer in mergers
between professional service ?rms. Human Relations, 54(7),
839–862.
Etzioni, A. (1960). A comparative analysis of complex organiza-
tions. New York: The Free Press.
Etzioni, A. (1988). The moral dimension. New York: Free Press.
Ferner, A. (2000). The underpinnings of ‘bureaucratic control
systems: HRM in European multinationals. Journal of Man-
agement Studies, 37(4), 521–540.
Friedman, A. (1977). Industry and labour. London: Macmillan.
Geertz, C. (1973). The interpretation of culture. New York:
Basic Books.
Govindarajan, V. (1988). A contingency-approach to strategy
implementation at the business-unit level: integrating
administrative mechanisms with strategy. Academy of Man-
agement Journal, 828–853.
Govindarajan, V., & Fisher, J. (1990). Strategy, control sys-
tems, and resource sharing: e?ects on business-unit perfor-
mance. Academy of Management Journal, 259–285.
Hopper, T., & Macintosh, N. (1993). Management accounting
as disciplinary practice. The case of ITT under Geneen.
Management Accounting Research, 4, 181–216.
Hopwood, A. (1987). The archeology of accounting systems.
Accounting, Organizations and Society, 12, 207–234.
Jo¨ nsson, S. (1998). Relate management accounting research to
managerial work! Accounting, Organization and Society, 23,
411–434.
Knights, D., & Willmott, H. (1989). Power and subjectivity at
work. Sociology, 23(4), 535–558.
Kunda, G. (1992). Engineering culture. Control and commitment
in a high-tech corporation. Philadelphia: Temple University
Press.
Lang?eld-Smith, K. (1997). Management control systems and
strategy: a critical review. Accounting, Organizations, and
Society, 22, 207–232.
Lind, J. (2001). Control in world class manufacturing—a long-
itudinal case study. Management Accounting Research, 12,
41–74.
Martin, J. (1992). The culture of organizations. Three perspec-
tives. New York: Oxford University Press.
Miller, P., & O’Leary, T. (1987). Accounting and the construc-
tion of the governable person. Accounting, Organizations and
Society, 12, 235–265.
Mintzberg, H. (1983a). Structure in ?ves. Englewood Cli?s:
Prentice-Hall.
Mintzberg, H. (1983b). Ideology and the missionary organiza-
tion. In J. B. Quinn, H. Mintzberg, & R. M. James (Eds.),
The strategy process. Englewood Cli?s: Prentice-Hall.
Mintzberg, H. (1989). Mintzberg on management. New York:
The Free Press.
Mumby, D. (1988). Communication and power in organizations:
discourse, ideology and domination. Norwood, NJ: Ablex.
O’Shea, J., & Madison, J. (1998). Dangerous company.
Harmondsworth: Penguin.
Ouchi, W. G. (1979). A conceptual framework for the design of
organizational control mechanisms. Management Science,
25, 833–848.
Ouchi, W. G. (1980). Markets, bureaucracies and clans.
Administrative Science Quarterly, 25, 129–141.
Peters, T. J., & Waterman, R. H. (1982). In search of excellence.
New York: Harper & Row.
Puxty, A. (1993). The social and organizational context of man-
agement accounting. London: Academic Press.
Ranson, S., Hinings, C. R., & Greenwood, R. (1980). The
structuring of organizational structures. Administrative
Science Quarterly, 25, 1–17.
Robertson, M. (1999). Sustaining knowledge creation within
knowledge-intensive ?rms. Diss. Warwick Business School.
Salancik, G. R., & Pfe?er, J. (1978). A social information pro-
cessing approach to job attitudes and task design. Adminis-
trative Science Quarterly, 23, 224–253.
Schein, E. H. (1985). Organizational culture and leadership. San
Francisco: Jossey-Bass.
Shamir, B. (1991). Meaning, self and motivation in organiza-
tions. Organization Studies, 12, 405–424.
Shotter, J., & Gergen, K. (Eds.). (1989). Texts of identity.
London: Sage.
M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444 443
Simpson, R. L. (1985). Social control of occupations and work.
Annual Review of Sociology, 11, 415–436.
Smircich, L. (1983). Studying organizations as cultures. In
G. Morgan (Ed.), Beyond method, strategies for social
research. Beverly Hills: Sage.
Smircich, L., &Morgan, G. (1982). Leadership: the management of
meaning. Journal of Applied Behavioural Science, 18, 257–273.
Spekle´ , R. (2001). Explaining management control structure
variety: a transaction cost economics perspective. Account-
ing, Organizations, and Society, 26, 419–441.
Tengblad, S. (2001). Examining the stability of managerial
behavior: a replication of Henry Mintzberg’s classic study 30
years later. GRI-rapport 2001:6 Go¨ teborg, GRI.
Therborn, G. (1980). The power of ideology and the ideology of
power. London: Verso.
Van Maanen, J., & Kunda, G. (1989). Real feelings: emotional
expression and organizational culture. In B. M. Staw, &
L. L. Cummings (Eds.), Research in organizational behaviour,
vol 11. Greenwich: JAI Press.
Weiss, J., & Miller, L. (1987). The concept of ideology in
organizational analysis. Academy of Management Review, 12,
104–116.
444 M. Alvesson, D. Ka¨rreman / Accounting, Organizations and Society 29 (2004) 423–444

doc_595964044.pdf