Description
Much consideration has been given over the years to what may be described as the ‘negative’ aspect of budgeting;
that budgets may constrain innovation and learning, and that budgetary pressure may lead to unintended behavioural
side effects. In contrast to this, the present study examines the extent to which budgets have a more positive, ‘comforting’
role to play in the individual’s work experiences
Coping with ambiguity through the budget: the positive
e?ects of budgetary targets on managers’ budgeting behaviours
David Marginson
*
, Stuart Ogden
1
Manchester School of Management, UMIST, P.O. Box 88, Manchester M60 1QD, UK
Abstract
Much consideration has been given over the years to what may be described as the ‘negative’ aspect of budgeting;
that budgets may constrain innovation and learning, and that budgetary pressure may lead to unintended behavioural
side e?ects. In contrast to this, the present study examines the extent to which budgets have a more positive, ‘com-
forting’ role to play in the individual’s work experiences. We argue that managers confronted with uncertainties
associated with role ambiguity may respond by becoming positively committed to achieving budgetary targets as
budgets o?er a source of structure and certainty. We ?nd that the use of budgets as an antidote to role ambiguity is a
powerful in?uence on the manager’s budgeting behaviour. We test the strength of this e?ect and we ?nd that budgetary
commitment brought on by the experience of role ambiguity may over-ride the potential for recognised explanatory
variables such as leadership style, the expectations of the superior, and occupational socialisation, to inform managers’
budgeting behaviours in these circumstances. Budgets, it seems, may be as useful to the individual as they are prob-
lematic.
Ó 2004 Elsevier Ltd. All rights reserved.
Introduction
Budgets often receive a ‘bad press’. Not only
are they accused of sti?ing innovation and learning
(e.g. Ansari, 1979; Argyris, 1977; Bartlett &
Ghoshal, 1993; Emmanuel, Otley, & Merchant,
1990; Hedberg & Jonsson, 1978; Hope & Fraser,
2003; Johnson & Gill, 1993), but a high emphasis
on budget attainment is deemed liable to create
various behavioural side e?ects that are likely to
prove dysfunctional to the ?rm. Issues examined
include data manipulation (Hopwood, 1972),
interdepartmental strife (Argyris, 1960), job-re-
lated tension (Brownell, 1981; Hopwood, 1972),
group based ‘anti-management’ behaviour (Argy-
ris, 1952), ‘gaming’ (Hofstede, 1968) and short-
termism (Merchant, 1990). The rather negative
view of budgets is also evident in the number of
studies that have, following the pioneering work of
Hopwood (1972) and Otley (1978), sought to
understand the extent to which factors such as
participation (e.g. Brownell & Dunk, 1991;
Brownell & Hirst, 1986), locus of control (Gov-
indarajan, 1988; Harrison, 1993), task uncertainty
(Brownell, 1983; Hirst, 1981), and degree of stra-
tegic change (Abernathy & Brownell, 1999), may
moderate the expected dysfunctional consequences
of a high reliance on accounting measures of
*
Corresponding author. Tel.: +44-161-200-3459.
E-mail addresses: [email protected] (D.
Marginson), [email protected] (S. Ogden).
1
Tel.: +44-161-200-3455.
0361-3682/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2004.05.004
Accounting, Organizations and Society 30 (2005) 435–456
www.elsevier.com/locate/aos
performance (RAPM) in the context of perfor-
mance evaluation (see reviews by Brownell, 1982;
Hartman, 2000; Kren & Liao, 1988; Otley &
Fakiolas, 2000; Shields & Shields, 1998). Such re-
search continues (e.g. Otley & Pollanen, 2000; Van
der Stede, 2000). However, despite considerable
evolution in the conceptualisation and speci?ca-
tion of variables deemed relevant in this context,
researchers appear to be still some way from
reaching de?nitive conclusions as to what condi-
tions are most likely to allow RAPM to operate
e?ectively without incurring the ‘problems’ that
Hopwood had predicted would accompany budget
constrained styles of supervision. Part of the
explanation for this lack of convergence is attrib-
uted to confusion and ambiguity in the meaning
and measurement of the RAPM construct itself;
problems with the use of methods for data col-
lection and analysis; and the uncritical use of
contingency theory (Briers & Hirst, 1990; Brownell
& Dunk, 1991; Chapman, 1997; Hartman, 2000;
Lindsay, 1995; Otley & Fakiolas, 2000; Vagneur &
Peiperl, 2000). Nevertheless, whatever reasons we
may invoke to explain the inconsistency of the
research ?ndings, one consequence of the persis-
tence with the research agenda raised by these
debates has been to perpetuate analysis of the
dysfunctional e?ects of budgets to the exclusion of
any consideration of the more positive role(s) that
budgets may play in people’s work-related expe-
riences. Since the work of Argyris (1952) and
others (e.g. Caplan, 1966; Stedry, 1960), the per-
spective adopted in the budgeting literature has
been one of ‘‘Human problems with budgets’’
(Argyris, 1953).
The lack of a counter-balancing ‘positive’ per-
spective on budgeting is surprising, not least be-
cause the potential for budgets to play a more
functional role in managers’ work-related experi-
ences has been suggested by the human relations
movement in accounting. Path-goal theory of
?nancial controls, for example, suggests that
where managers do not have self-evident paths and
clear-cut goals, they will welcome accounting
based controls such as budgets for the structure
and certainty they provide. This is particularly true
of senior managers, whose roles are frequently
ambiguously de?ned. Macintosh (1995, p. 215)
argues that budgets ‘‘serve to reduce sharply role
ambiguity’’. At the same time, interesting and
persuasive as these arguments are, they remain
under-developed and certainly under-researched.
In particular, little is currently understood about
how or why managers may use the budget to cope
with uncertainty and ambiguity, what this means
in terms of budgeting behaviours, and more
importantly, perhaps, what the implications may
be for individual and organizational performance.
The question of whether or not budgets may play a
positive role in people’s work-based experiences
remains a neglected topic.
The importance of understanding the potential
functionality of budgets at the level of the indi-
vidual is evident in recent empirical research on
contemporary attitudes of senior managers to
performance evaluation based strictly on account-
ing measures of performance. Storey, Edwards,
and Sisson (1997), for example, found that man-
agers, faced with increasingly tight budgetary
targets, not only accepted but positively endorsed
the rigorous degree of accountability they experi-
enced. The main reason for this was the extent to
which managers welcomed the certainty created by
the combination of clear goals and well-speci?ed
objectives, and a performance evaluation system
that was tightly focused on how successfully those
objectives have been achieved. Contrary to the
expectations associated with the dysfunctional as-
pects of RAPM, the degree of close monitoring
and evaluation of performance did not produce
discontent. Rather, Storey et al. (1997, pp. 201–
202) found that ‘‘tight evaluation was linked to a
favourable view of the system of evaluation’’ and
that managers were ‘‘most likely to be strongly
motivated when they felt that their activities were
closely evaluated by their companies’’. Storey et al.
(1997, p. 203) explain the sense of satisfaction ex-
pressed by managers as the result of their sense
that companies with demanding targets appear to
know where they are going and what they want
from their managers. Managers generally appre-
ciated the sense of strong direction from the top, as
expressed in the detailed speci?cation and tight-
ness of their performance targets, and welcomed
the sense of certainty and mission this communi-
cated.
436 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
In the example quoted above two distinct, but
clearly related, processes are discernible in man-
agers’ responses to RAPM. On the one hand, there
is positive endorsement by managers of the goal
speci?city and goal clarity that operates in the
context of performance focused companies. On the
other hand, there is also appreciation of the use
made of budget constrained styles of supervision
which focus on evaluating how successfully speci-
?ed goals have been achieved. Viewed from this
perspective, the absence of such clarity and speci-
?city and the reliance on a more relaxed supervi-
sory style with regard to APM, as envisaged, for
example, by Hopwood’s (1972) notion of a ‘pro?t
conscious’ style of supervision, may paradoxically
cause more problems than it solves. It may be
suggested that where this certainty provided by a
clear emphasis on budget goals is lacking manag-
ers may seek ways to recreate it for themselves. In
this context, identifying with the budget may be
seen as functional by managers precisely because it
is seen as providing a clearly de?ned performance
target which, if achieved, will provide a degree of
security amidst the inherent ambiguity of the
managerial role (Macintosh, 1995).
The presence of role ambiguity provides the
focus for our examination of the potential func-
tionality of budgets at the level of the individual.
Speci?cally, we examine the extent to which man-
agers may actually commit to meeting budgetary
targets, not because of the threat of accountability
or the promise of reward, but because doing so
o?ers a sense of clarity and security, particularly in
circumstances where their role is fraught with
uncertainties. In e?ect, we seek to understand how
far managers who experience anxiety brought on
by ambiguity and uncertainty within their role may
use budgets as a means of coping with this anxiety.
We address this aim through the analysis of evi-
dence garnered from a recently conducted com-
pany case study. The sources of role ambiguity are
explained, and managers’ reactions to this and
their attempts to secure for themselves a more
gratifying role experience through commitment to
budgetary targets are traced. Using quantitative
evidence we are also able to test the strength of the
e?ect that role ambiguity may have on managers’
behaviour towards budgetary targets against the
more usual explanations for budgetary commit-
ment by managers suggested in previous research.
We argue that budgetary commitment brought on
by the experience of role ambiguity may over-ride
the potential for recognised explanatory variables
such as leadership style (Argyris, 1952), hierarchi-
cal contagion (Hopwood, 1974) and occupational
socialisation (Brownell, 1985; Hirst, 1981) to in-
form managers’ budgeting behaviours in these
circumstances. Moreover, as the case study
company exhibited a distinct lack of emphasis
on RAPM and a marked absence of formal
accountability and reward from the budgetary
system, we are able to examine the role played by
budgets, of themselves, as an antidote to role
ambiguity in circumstances unencumbered by the
presence of factors which would undoubtedly
confuse the analysis and limit interpretation of the
results.
The section that follows introduces the case study
and presents the qualitative evidence that identi?es
managers’ experience of role ambiguity, its causes,
and their reactions to it. Section 3 develops several
arguments relating to the use of budgets as coping
mechanisms. Section 4 describes the quantitative
methodology that was used to test the resulting
hypotheses, while Section 5 presents the results of
this analysis. Section 6 outlines the study’s limita-
tions and concludes the discussion. Our ?ndings
suggest that the experience of role ambiguity may
exercise considerable in?uence on managers’ bud-
geting behaviours; it may ensure their budgetary
commitment in the search for a more gratifying role
experience. The implications of these ?ndings are
discussed in the concluding section.
The case study: Infotain plc
The case study company, which we have called
Infotain plc, is a major UK (FTSE100) based
organization which operates in what it describes as
the ‘global communications business’. Infotain plc
is a prime example of a modern ‘professional adh-
ocracy’ (Euske, Lebas, & McNair, 1993) or infor-
mation-age company (Bartlett & Ghoshal, 1993;
Simons, 1995, 1999). De?ning features include
fast-moving circumstances, jurisdictional and
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 437
decisional ambiguity, ?uid role responsibilities,
frequent job transformations, a constant need to
prioritise among tasks, and front-line empower-
ment. The company supplies and maintains a range
of communication and infotainment products and
services to both major private sector companies and
large public corporations. It employs around
12,000 people, and operates from several sites
located primarily in the UK. Turnover exceeds £4
billion per annum.
The aim of the case study was to examine the
design and operation of the company’s control
processes and procedures, with particular attention
being paid to investigating the ‘human dimension’
of management control. Of the areas investigated,
research into budgeting’s role in the process of
management control was intentional, not least be-
cause budgets have long been viewed as a, if not the,
pivotal organizational control mechanism (Ansari,
1979). Several other issues, including the company’s
empowerment initiative, managers’ experience of
role ambiguity, and the notion of budgetary com-
mitment, were also explored in some detail, given
initial insights which showed these to be interlinked
and potentially signi?cant to organizational per-
formance. All issues were addressed through inter-
views with senior and line managers; meetings with
personnel from both the accounting and human
resource functions; and scrutiny of company
memoranda, reports and other documentary evi-
dence. Interviews were used to elicit managers’
views, beliefs, and actions about the control systems
at their disposal, and were conducted over the two-
year period of this case study. During this time,
initial ideas and arguments were developed and
re?ned in the light of additional evidence through,
for example, further interviews with managers. The
following seeks to elaborate on the key issues to
emerge from the qualitative ?eld work.
For reasons relating to the company’s highly
organic context and the need to ensure a high rate
of strategic adaptation and change at ?rm level,
given a fast-moving business environment (Dut-
ton, Ashford, O’Neil, Hayes, & Wierba, 1997),
Infotain had largely abandoned budgeting in its
traditional form. It adopted a broader control
environment which sought to encourage strategic
adaptation and change through, for example, the
use of beliefs systems (Simons, 1995). Although
‘prudent’ budgetary goals and other ?nancial tar-
gets were set by the accounting function to instil
cost consciousness throughout the ?rm, there was
little formal accountability or reward attached to
budgetary performance. Only the most senior
managers had ?nancially based accountabilities.
Cost centre managers, although allocated re-
sources, were not explicitly assessed or rewarded
on their ability to remain within budget. Indeed,
tolerance limits were attached to overall budget
allocations, which meant that budget targets could
be exceeded (or under-achieved) by upwards of
10% for ‘strategic reasons’. ‘Out of spends’ were
also possible if the manager had a good reason to
exceed the 10% tolerance limit. Managers could
also renegotiate their budgets periodically in the
light of unfolding events. The reasoning behind
this ?exibility, according to top management, was
to ensure, or at least not to deter, strategic adap-
tation in response to a rapidly changing, highly
uncertain environment. Managers were therefore
given considerable licence to sacri?ce initial bud-
getary targets in favour of unexpected, but
potentially pro?table, opportunities if and when
they arose during the budgetary period. They were
speci?cally encouraged to balance the need for
variance correction with (1) the need to meet
launch deadlines for new products and services
and (2) the possibility of making improvements to
on-going projects as new information became
available, notwithstanding the extent to which
meeting agreed time-frames and modifying pro-
jects often required additional resources.
The degree of discretion that managers were
a?orded in terms of budgetary performance mir-
rored the ethos of empowerment which pervaded
the organization. For reasons similar to those
which appeared to drive the company’s seemingly
radical approach to budgeting, there was a general
reluctance to de?ne and limit individual responsi-
bilities. This was epitomised in the company ’motto’
‘‘tell me what you’ve done, don’t ask me what to
do’’. Middle-ranking managers were particularly
empowered to indulge in ’autonomous strategic
behaviour’ (Burgelman, 1983a, 1983b), ‘issue-sell-
ing’ (Dutton et al., 1997), and other grass-roots
activity that typi?es an emergent strategy process
438 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
(Mintzberg & McHugh, 1985). Most were members
of at least one autonomous multifunctional project
team, and several remarked on how they were
‘‘free’’ to commit to accountabilities that would
take them ‘‘completely away from any form of
hierarchical relationship’’. Managers saw them-
selves in terms of having ‘‘multiple stakeholders’’ as
a result of ‘‘mutual accountabilities’’ and omni-
present interdependencies (Spekle, 2001).
The company’s empowerment strategy ap-
peared to be supported by managers. The majority
of those interviewed suggested that the dissipation
of responsibility for strategy formation through-
out the managerial ranks was necessary as the
rapid technological advances, short product life
cycles and a ?ercely competitive environment
made the notion of detailed strategic planning
problematic (Mintzberg, 1979). At the same time,
some interviewees did appear uneasy with the high
levels of ambiguity and uncertainty that a lack of a
clear ‘‘authority map’’ and a highly organic con-
text created. This lack of role clarity extended to
performance evaluation. For example, some
managers commented how they received little
formal feedback from their superiors on their role
performance, while their superiors commented
how it was di?cult to do so, given a person’s
activities did not necessarily support the line rela-
tionship. These conditions were prevalent
throughout the ?rm and prompted some managers
to express concern about the extent of their ?uid
role responsibilities, their mutual accountabilities,
and the continual need to prioritise amongst tasks.
For example, one manager remarked:
Things are in a constant state of ?ux around
here, and this can be a bit of an issue for me
personally at times. I like to be clear about
what I am supposed to do, but I appreciate
this can’t really be the case around here.
Another manager commented on how ‘‘the
constant need to prioritise’’ often led to people to
‘‘doubt their choice of priority’’. Other managers
suggested, somewhat more negatively, that the
blurring of roles and responsibilities was ‘‘a nec-
essary evil’’ or just had to be accepted and was
therefore ‘‘something to live with’’.
In examining how managers sought to ‘‘live
with’’ the ambiguity and uncertainty prompted by
the company’s empowerment initiatives, evidence
emerged from the ?eld study to suggest a degree of
‘coping through the budget’. A number of inter-
viewees remarked how committing to meeting
budgetary targets ‘‘provided an anchor’’ or ‘‘focal
point’’ in the midst of jurisdictional and decisional
ambiguities, and the constant push for innovation.
They made it clear that they wanted to achieve pre-
determined budgetary targets, and expected their
subordinates to do likewise. One manager re-
marked: ‘‘I don’t like all this change. If there’s a
budget and a target, then it should be met. End of
story’’. Several others commented on the impor-
tance of this role of budgets through references to
the company’s ethos: when asked to ‘‘tell me what
you’ve done, not ask me what to do’’, they ruefully
said they would ‘‘tell them they had met the bud-
get!’’ One manager, in more serious vein, remarked:
I am no longer sure where my job ends and
someone else’s begins. There’s an incredible
amount of uncertainty, and people are always
being asked to do many di?erent things at
once. The whole thing is a mess. This does
bother me; a hell-of-a-lot at times; I just don’t
know what I should be doing half the time.
But, at least I know where I stand with the
budget. If I keep this in order and hit the tar-
gets, and make sure my guys are doing the
same, then no one can argue with that, not
even the chief executive!
The above evidence suggests both causal direc-
tion and a rationale for the surprising degree of
budgetary commitment that was evident at Info-
tain. Managers’ propensity to pursue budgetary
targets was, in part at least, driven by their desire
to create for themselves a degree of structure and
certainty in the face of the ambiguities and
uncertainties that had become characteristic of
their situation as a result of their experience of
empowerment. Moreover, the evidence indicates
that the pursuit of structure and certainty through
the budget in these circumstances may well over-
ride other factors such as organizational position
and leadership style in determining attitudes
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 439
towards the budget in these circumstances. One
manager, for example, remarked how his man-
ager’s ‘‘harsh’’ attitude towards budgetary targets
was inconsistent with his supervisory style, which
was ‘‘generally supportive and understanding’’.
Further, the use of standard practices for quali-
tative data analysis, including interview coding
and cross-tabulations (Miles & Huberman, 1984),
revealed that those who expressed varying degrees
of concern about their role also tended to show
commitment towards the budget, irrespective of
their organizational position or what could be
interpreted at this stage as their more general ap-
proach to the job. In contrast to this, factors such
as business unit a?liation did emerge as possible
predictors of budgeting behaviours for the group
of interviewees who, through an absence of rele-
vant remarks, showed little or no concern for the
uncertainty they faced as part of their role. These
are potentially important observations, not only
because they challenge longstanding assumptions
about the conditions required to ensure budgetary
commitment, but also because they suggest that
the behavioural literature in accounting may have
omitted a signi?cant determinant of budgeting
behaviours: the experience of role ambiguity.
These insights drawn from the qualitative data
may be further explored through the analysis of
quantitative data. While this complements the
qualitative data in examining how and why bud-
gets may be used as a coping mechanism against
the ambiguity experience brought on by empow-
erment initiatives, it is distinctive in that it pro-
vides a basis for testing the strength of the
relationship between role ambiguity and budget-
ary commitment.
2
Role ambiguity and budgetary commitment
In this section, we develop a number of argu-
ments in order to test the strength of the relation-
ship between role ambiguity and budgetary
commitment. We argue that empowerment may
constitute a source of role ambiguity. We then
consider the concept of role ambiguity before pre-
senting the arguments as to why budgets may be
useful mechanisms to alleviate stress and comfort
managers a?ected by role ambiguity. We conclude
by arguing, ?rstly, that the use of budgets as a
coping mechanism leads managers to commit to
meeting budgetary targets, irrespective of the pres-
ence of some factors (i.e. recognised explanatory
variables such as leadership style and the superior’s
expectations), and the absence of others (e.g. formal
accountability and reward), and secondly that
coping is related to a gratifying role experience
(Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964).
The various arguments and proposed relationships
are depicted in Fig. 1, which presents the frame-
work for the forthcoming discussions.
Empowerment and role ambiguity
A fundamental aim of empowerment initiatives
is to embed notions of strategy formation
throughout the ?rm (Otley, 1994). As middle-level
managers are often closer to the customer and
other stakeholders than top management, and
therefore have more detailed knowledge of what
strategic issues require attention, empowerment
initiatives frequently envisage middle-level man-
agers playing pivotal roles in detecting new ideas
and in mobilising resources around these new ideas
(Dutton et al., 1997; Kanter, 1982; Simons, 1995).
In pursuing these ‘strategic forcing’ roles (Burgel-
man, 1983a, 1983b), managers are encouraged to
undertake ‘boundary-spanning’ activities, apply
individual initiative, and work as ‘team players’
(Euske et al., 1993). However, such activities tend
to result in the blurring of role responsibilities and
traditional line relationships, and, at times, poor
communications between superior and subordi-
nate. In this respect, empowerment initiatives
undoubtedly disrupt the chain-of-command prin-
ciples characteristic of formal hierarchical man-
agement structures (Bartlett & Ghoshal, 1993;
House & Rizzo, 1970) and, in so doing, may create
or increase role ambiguity (King & King, 1990).
Role ambiguity occurs when an individual is
unsure about others’ expectations of him-or her-
2
In view of the ?eld evidence, and in anticipation of the
analysis to follow, causal direction is implied in the arguments
that are developed.
440 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
self. It has been conceived as the ‘‘discrepancy
between the amount of information a person has
and the amount s/he requires to perform his or
her role adequately’’ (Kahn, 1974, p. 426). The
increased job transformations and ?uid role
responsibilities precipitated by empowerment ini-
tiatives may leave the manager unsure about what
his or her duties and responsibilities are. The
manager may also be unclear about the activities
which should be performed in order to ful?l these
responsibilities, while the consequences of action
(or inaction) may be unknown. Such ‘objective’
role ambiguity is, in many respects, a fact of
managerial life (Macintosh, 1995), but people
have an inherent, if varying, need for role clarity
(Elovainio & Kivimaki, 2001; Eysenck, 1954;
Kahn et al., 1964).
3
As a result, a close corre-
spondence is expected between objective and
subjective, or experienced, role ambiguity (Kahn
et al., 1964; Netemeyer, Johnston, & Burton,
1990). In general, we prefer to know what our
duties and responsibilities of o?ce are, and how
we are supposed to discharge them, while from a
socioemotional perspective, it is important that we
know what the consequences of performance or
non-performance will be, particularly for our-
selves, but also, perhaps, for our colleagues and
the organization in general (King & King, 1990).
However, this sort of means-ends knowledge is
unlikely to be forthcoming from the ?uid role
responsibilities and the high levels of jurisdictional
and decisional ambiguity created by empower-
ment initiatives. Rather, the ?exible approach to
task requirements and the boundary spanning
activities prompted by empowerment are likely to
lead to imprecise lines of authority and lack of
clarity about what is required. Consequently, we
may argue that managers’ experience of empow-
erment is likely to be related to their experience of
role ambiguity. We therefore hypothesise that:
H1 (A source of role ambiguity): Empower-
ment is associated with role ambiguity.
Coping with role ambiguity through the budget
Experienced role ambiguity is an unwanted
psychological state, which occurs when the person
is concerned or ‘stressed’ by a lack of clarity and
Role ambiguity
(e) H5 -ve
Empowerment
(a) H1+ve
Performance
Budget commitment
(b) H2 +ve (d) H4 +ve
(a) A source of role ambiguity
(b) For high levels of role ambiguity
(c) Sig. only in the absence of role ambiguity
(d) Regaining a gratifying role experience
(e) A gratifying role experience regained
Superior’s
expectations
Leadership
style
(c) H3 sig (c) H3 sig (c) H3 sig
Occupational
socialization
Fig. 1. Diagrammatic representation of the arguments.
3
The need for role clarity can vary according to a person’s
tolerance of objective role ambiguity (Eysenck, 1954).
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 441
structure to the role (King & King, 1990; Siegall,
2000). Doubts about how others evaluate us,
about how satis?ed they are with our performance,
are frequent and deep-seated sources of anxiety
(House & Rizzo, 1970; Kahn et al., 1964; King &
King, 1990). As a result, those who experience role
ambiguity must confront it in some way (Kahn,
1974; Siegall, 2000), usually by invoking ‘coping
strategies’. These are attempts to regain clear, or-
derly and meaningful cognitive experiences. Such
attempts may involve the use of defence mecha-
nisms, which are able to distort the reality of the
ambiguous situation in order to relieve the anxiety
of the undistorted experience (Dougherty & Prit-
chard, 1985). However, within the role theory lit-
erature there is little elaboration of what these
defence mechanisms may consist of. Consequently,
little is actually known about the nature of these
defence mechanisms, except for evidence to sug-
gest that it may involve the person leaving the role
(King & King, 1990). Notwithstanding this, our
evidence indicates that it is plausible to argue that
identifying with the budget by emphasising the
achievement of short-term budgetary targets con-
stitutes an important defensive strategy against the
experience of role ambiguity; it is certainly one
which most managers within the ?rm can invoke.
There are several reasons why budgets o?er a
useful coping strategy. First and foremost, bud-
getary control is organizationally ‘good citizen-
ship’. Those who are controlling their costs are
bene?ting the ?rm (Merchant, 1998). Moreover,
because of the visibility of accounting information
(Roberts, 1991), the manager can be seen to be an
organizationally ‘good citizen’. This is important,
as it means that the manager is able to predict,
with a high degree of con?dence, the consequences
of performance. More importantly still, the
acquisition of such knowledge does not depend,
for example, upon clear communication with the
hierarchical superior about what should be done,
how it should be done, or what kinds of behav-
iours are likely to result in reward. Rather, the
manager autonomously ‘knows’ what needs to be
done to acquire a positive organizational identity,
knowledge which is aided by the ‘mechanistic’
credentials of budgetary systems (e.g. assumptions
of certainty and independence of responsibility––
Ezzamel & Hart, 1987). These make it clear what is
required (achieve the budget), who is responsible
(the manager), and, given the arithmetical nature
of accounting reports (Hopwood, 1976), what
regulatory action must be taken if the manager is
to meet his or her own budgetary target: instruct
the subordinate to do likewise (Hopwood, 1976).
Those who experience role ambiguity may
therefore value the budget for the comfort and
security it o?ers. But, such psychological return is
conditional upon the achievement of targets; there
is little comfort to be gained from exceeding pre-
established targets or missing the budget. Thus, if
budgetary control is to be used as a coping strategy,
the manager must commit to the budget, and must
exercise ‘tight budgetary control’ to achieve bud-
getary targets. Anything less decreases the proba-
bility of meeting the budget and, thereby,
jeopardises the possibility of gaining the sense of
socioemotional security that is desired. There may
be valid ‘strategic reasons’ for exceeding the budget,
but these reasons will be less visible to others and
longer-term in their fruition. Role ambiguity is not
avoided by these actions; rather ambiguity increases
as objects and events recede in time and space, a fact
which most people appear to accept (King & King,
1990). Those who experience role ambiguity are
thus unlikely to adopt a ?exible approach to the
budget or even to forgo budgetary targets in favour
of longer-term payo?s; such behaviour serves to
increase perceptions of uncertainty (e.g. how much
?exibility) and ambiguity (e.g. is it for the ‘right’
strategic reason that the subordinate is being in-
structed to forgo the budget) at a time when cer-
tainty is desired. The coping strategy is committing
to the budget. The reward is a more gratifying role
experience (House & Rizzo, 1970).
An over-riding in?uence?
The corollary of the above argument is that
those who experience little or no role ambiguity
may adopt a more ?exible attitude towards the
issue of budgetary control, because they have little
need to ‘cope’. The absence of role ambiguity re-
moves the need for the manager to commit to
meeting the budget in order to counter or distort
the ambiguity experience. However, in view of the
442 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
?ndings of prior research, we may argue that
budget ?exibility in these circumstances is likely to
be shown only to the extent that variables such as
leadership style (Argyris, 1952; Hopwood, 1974),
the supervisor’s behaviour towards the budget
(Hopwood, 1974), or even the manager’s previous
work experiences and current functional location
(Hirst, 1981) do not themselves result in budgetary
commitment. People for whom the experience of
role ambiguity is minimal or absent may still
exercise tight budgetary control because of their
task-orientated approach to the job (Hopwood,
1974) or because of the e?ects of hierarchical
contagion (Bonini, 1963; Hopwood, 1974). Equal-
ly, of course, these managers may not exercise tight
budgetary control because their style of leadership
is more considerate or employee-oriented, or be-
cause this is what the superior expects, or indeed
because they are merely re?ecting established views
within the particular business unit or function in
regard to the importance of budgets (Hirst, 1981).
In e?ect, therefore, it cannot be suggested cate-
gorically that budget ?exibility is associated with an
absence of role ambiguity. But, what we can argue
with some conviction is that, because role ambigu-
ity triggers a seemingly powerful desire for structure
and certainty (Kahn et al., 1964; Siegall, 2000), and
because budgets o?er a potentially e?ective coping
mechanism, those experiencing role ambiguity are
likely to commit to meeting the budget in these
circumstances. The ambiguity experience crystal-
lises attitudes towards the budget in a way that
non-experience does not. The strength of the rela-
tionship between role ambiguity and budgetary
commitment may be assessed by examining the
extent to which the experience of role ambiguity
overrides or suppresses other explanatory variables
in determining the manager’s behaviour towards
the budget. The most appropriate variables to in-
clude in the analysis are those, such as leadership
style and hierarchical contagion, which have been
shown previously to shape managers’ budgeting
behaviours. We will therefore examine the argu-
ment that the experience of role ambiguity leads the
manager to commit to meeting the budget irre-
spective of (1) his or her natural inclination to be-
have otherwise, as is the case with a considerate or
employee-orientated style of leadership (Hopwood,
1974), (2) what the superior may expect of the
manager (DeCoster & Fertakis, 1968; Hopwood,
1974), and (3) the e?ects of occupational socialisa-
tion, as demonstrated through business unit a?li-
ation.
The formal hypotheses for the above arguments
are as follows:
4
H2 (A crystallisation of attitudes): Those who
experience high levels of role ambiguity are
more likely to commit to meeting the budget
than those whose experience of role ambiguity
is minimal or absent.
H3 (An over-riding in?uence): The experience
of role ambiguity over-rides recognised
explanatory variables (leadership style, hierar-
chical contagion) and other predicted factors
(e.g. occupational socialisation) in shaping
attitudes to the budget in these circumstances.
H4 (A coping strategy for gaining or regain-
ing a gratifying role experience): The higher
the commitment to the budget, the higher
the level of self-reported performance.
5
4
The ?ve hypotheses seek to re?ect the various components
of the ‘theory’ (i.e. source of role ambiguity, asymmetry,
suppression of recognised explanatory variables, and ‘coping’).
Their separate examination is necessary, given the intricacies of
the arguments. In fact, a more complete picture is possible
through this cumulative approach to the empirical analysis.
5
The role theory literature stresses how people invoke
coping mechanisms in a bid to obtain a more gratifying role
experience, but o?ers little guidance on what may constitute a
‘gratifying role experience’ or how this may be measured. The
present study sought to construct this phenomenon in terms of
performance. The reasons are twofold. First, it is reasonable to
argue that one measure of ‘grati?cation’ is performance; those
who are satis?ed with their role are likely to feel that they are
performing well (and vice versa). Second, and more importantly
perhaps, given a lack of formal accountability and reward at the
research site, and given a general emphasis on innovation and
learning, the manager has much less need to evaluate his or her
worth to the organization against accounting numbers (Rob-
erts, 1991). Self-evaluation of performance in this context will
thus represent a feeling of general well-being within the role
rather than a speci?c measure of task accomplishment.
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 443
H5 (A gratifying experience regained): Per-
formance is inversely related to role ambigu-
ity.
Method
The arguments were tested on data collected
from a questionnaire survey involving ‘level three’
managers drawn from ?ve of Infotain’s eight
strategic business units. The business units were
chosen to represent a spectrum of functions,
including engineering, product development and
sales and marketing. All level three managers from
these business units were asked to participate in
the questionnaire survey. Completed question-
naires amounted to 221 out of a possible 270, of
which all were usable. The response rate was 82
percent.
The choice of managerial level was dictated by
several factors. First, level 3 managers represent,
for the ?rm in question, junior executives with
budget responsibilities and regulatory duties (level
4 managers were also allocated resources). Sec-
ond, meetings with accounting and human re-
source personnel and the perusal of documents
relating to formal accountabilities established
several important facts about the way budgetary
control was exercised at this level. One is that
these junior executives were neither assessed nor
rewarded for staying within budget. Another is
that most managerial ranks, including the one in
question, were not, by and large, able to partici-
pate in the setting of their budgetary targets.
Instead, ‘prudent’ budgets tended to be imposed
by the accounting function in a bid to establish
cost consciousness throughout the managerial
hierarchy. When questioned about whether or not
they accepted their budgetary targets, managers
responded by saying they had ‘‘no choice’’! This
is important in the context of the present study as
the lack of participation in setting budgetary
targets is normally expected to make managers
less committed to achieving them. Commitment
to budgetary targets, even for intrinsic reward, is
considered to be inexorably bound with notions
of perceived legitimacy and ‘ego involvement’,
which come from managers’ ability to participate
in the setting of budgetary targets (Collins, 1982).
As regards the achievability of their budgetary
targets, managers when asked in interviews
commented that they were moderately di?cult or
di?cult. None suggested that the budget was
easily achievable.
A ?nal reason for focusing on level 3 managers
is that managers at this level and below were
empowered to behave ‘strategically’; they could, if
necessary, exceed their original budgetary targets
for ‘strategic reasons’. The same applied at the
next level down, but it was felt that lower-ranking
managers may generally feel more insecure, and
therefore might be more predisposed to experi-
encing role ambiguity. The intent of the study was
to focus attention on the highest managerial level
possible where any feelings of anxiety and insecu-
rity may be minimised, thereby establishing a
context which would make signi?cant results less
rather than more likely.
Measurement of the variables
Commitment to the budget: tight budgetary control
Ten questions were initially devised to measure
managers’ commitment to the budget as manifest
in the exercise of ‘tight budgetary control’ (Van
der Stede, 2001). Tightness may be exhibited in
two main ways. One is that which is imposed by
the budgetee on him or herself. The other is the
tightness that is imposed by the budgetee on the
subordinate. Both are necessary if the budgetee is
to obtain security from meeting the budget
(Hopwood, 1976). Hence, both these approaches
were applied in devising the questions. The ten
questions are shown in the Appendix (questions 1–
10).
Con?rmatory factor analysis was employed to
determine validity. This resulted in the original
ten-item instrument being reduced to eight, as
survey items that load above 0.40 were retained
(Dutton et al., 1997; Van der Stede, 2001). A
Cronbach Alpha score of 0.76 was obtained for
this eight-item measure, which is comfortably
above lower limits of acceptability, generally con-
sidered to be around 0.50–0.60 (Nunnally, 1967),
and may be regarded as reasonable for an instru-
ment which is being used for the ?rst time. Cron-
444 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
bach Alpha for all ten original items was 0.66
(Cronbach, 1951). Thus, omitting the two items
with weak factor loadings ‘‘not only makes the
scale more parsimonious, it also enhances its
overall reliability’’ (Van der Stede, 2001, p. 129).
Factor loadings are reported in Table 1. The two
omitted items are statements 3 and 7 (see the
Appendix).
The underlying principle that was applied in
devising the research instrument was that of
‘trade-o?’; budgetary commitment through ‘tight
budgetary control’ (Van der Stede, 2001) was
measured, not by examining, for example, ‘the
amount of emphasis placed on attaining budget
targets’ per se, but by examining the extent to
which such emphasis may be exhibited in the
context of the need for strategic adaptation and
change. Tensions between these important aspects
of organizational endeavour are inherent to ?rms
(Simons, 1995); they were a particular feature at
Infotain. It is likely that most ?rms would like
managers to strike a balance between budgetary
control on the one hand, and innovation and
learning on the other. To that end, getting
respondents to demonstrate their preferences and
priorities in trade-o? situations is a potentially
useful way of establishing strength of budgetary
commitment. Respondents were therefore asked to
what extent they expected budgetary targets to be
met, given the need to consider (1) long-term
?nancial performance, (2) problems of controlla-
bility (Choudhury, 1986), (3) unfolding events and
opportunities, and (4) strategic developments.
Items 6 and 10 are used as overall anchors (Ma-
honey, Jerdee, & Carroll, 1963, 1965). Each item
was measured on a ?ve-point scale, with 1 repre-
senting strong agreement and 5 strong disagree-
ment with the statement presented. Items 1, 7, 8
and 9 were reverse scored.
An expectations approach is used to measure
the degree of tightness as applied to the subordi-
nate. This is in keeping with the ?ndings of the
?eld study and suggestions in the literature (e.g.
Machin, 1979). The communication of expecta-
tions represents an in?uence attempt, which is
aimed at ensuring compliance with what is ex-
pected (Kahn et al., 1964; Machin, 1973, 1979).
Thus, expecting pre-determined budget targets to
be achieved in the context of trade-o? represents
budgetary commitment. It is ‘tight budgetary
control’.
Experienced empowerment
The notion of empowerment has been exten-
sively discussed, but there are surprisingly few re-
search instruments that have been devised to
detect and measure the extent of a manager’s
empowerment in a business context. In this study,
we use a ?ve-item instrument (reduced from six
items on the basis of con?rmatory factor analysis)
to measure respondents’ perceptions as to the de-
gree of strategic empowerment they were a?orded
in the discharge of their role. Consistent with
?ndings from the qualitative study, respondents
were asked to indicate the extent to which they
were expected to indulge in various ‘boundary-
spanning’ activities which took them away from
the line relationship with the superior. Each item
was measured on a ?ve-point scale with 1 repre-
senting strong agreement with the statement pre-
sented and 5 strong disagreement. The Cronbach
Alpha score for the ?ve-item instrument is a highly
respectable 0.72 (Nunnally, 1967). The research
instrument is shown in the Appendix (items 11–15).
Role ambiguity
Role ambiguity is assessed using a slightly
adapted version of the well-tested measure ?rst
developed by Kahn et al. (1964) (see questions 16–
21 in the Appendix). A reliability measure of 0.80
Table 1
Factor loadings for budgetary commitment measures
Items in order shown
in Appendix A
Rotated
factor loading
Squared multiple
correlations
1 0.716 0.525
2 0.651 0.566
3 0.340 0.511
4 0.533 0.483
5 0.814 0.879
6 0.583 0.482
7 0.243 0.548
8 0.708 0.624
9 0.488 0.475
10 0.510 0.897
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 445
was obtained for the resulting six-item instru-
ment.
6
The superior’s budgetary expectations
Hopwood’s (1974) research suggests that tight
budgetary control (in the form of a budget-con-
strained evaluative style) is passed on from one
level to the next through the notion of hierarchical
contagion (Bonini, 1963). However, what is of
greater interest here is the extent to which the
budgetee may be committed to meeting the budget
even though the superior may expect otherwise.
Use of an expectations approach (Machin, 1979)
works both ways, in that the superior may expect
the manager to trade budgetary targets in favour
of strategic adaptation. This was evident at Info-
tain: the few level 2 managers that were inter-
viewed suggested that they would be prepared to
forgo budgetary targets for strategic reasons. The
extent to which level 2 managers expected their
subordinate(s) to trade budgetary targets for
strategic reasons was assessed by asking level 3
managers what they perceived to be expected of
them in relation to the budget. A ?ve-item
instrument was devised and reduced to four items
following factor analysis. The four items are
shown in the Appendix (questions 22–25). A
Cronbach Alpha score of 0.79 was obtained for
the four-item instrument.
Occupational socialisation/business unit a?liation
Occupational socialisation (DiMaggio & Pow-
ell, 1983) at the business unit level was assessed
according to membership of a particular business
unit. This was represented as three separate dum-
my coded (0 or 1) current position variables. For
example, in the ?rst variable members of business
unit one (engineering) were coded 1 and the
remainder 0 (Donabedian, McKinnon, & Bruns,
1998; Sim & Killough, 1998). The process was
repeated for each of the other four business units.
Leadership style
A ?ve-item instrument was devised in order to
detect and measure leadership style in action at the
research site.
7
Broadly speaking, each question
sought to measure leadership style along a high
task/low consideration––low task/high consider-
ation continuum. For example, Question 26 asks
respondents to indicate the extent to which they
expect subordinates to revise their responsibilities/
commitments as circumstances change; the more
revisions are sanctioned or expected, the more
employee-orientated the manager, given the cir-
cumstances encountered. (Expecting original tar-
gets to be met in the context of rapidly changing
conditions demonstrates a higher regard for the
task than the individual, who must confront these
‘di?cult’ circumstances.) Each question in the
present instrument was measured along a ?ve-
6
While the paper focuses on anxiety brought on by the
experience of role ambiguity, all 15 items that Kahn et al. (1964)
?rst composed to measure job-related stress and tension
manifest as role ambiguity, role con?ict and role overload were
included in the questionnaire in order to check the validity of
the role ambiguity measure through factor analysis. The six
items for role ambiguity loaded onto one (of three factors) at
0.50 or above. The analysis yielded three Eigen values >1 which
accounted for 68% of the variance. Only the six items relating to
role ambiguity are shown in Appendix A for reasons of brevity.
7
Due to top management placing restrictions on the size of
the questionnaire, the general circumstances encountered, and
the focus of the study (i.e. the manager rather than the
subordinate), a decision was taken not to use the Stodgill (1963)
two-dimensional instrument of initiating structure and consid-
eration. This measure contains a range of questions which, in
the eyes of Infotain’s senior managers, rendered the question-
naire ‘to big’ in terms of resource implications for the ?rm. In
the event, while this instrument may be well tested, it has several
features which limit its usefulness in the present analysis. First,
it is mainly used to measure subordinates’ perceptions of their
manager’s leadership style; the measure is generally regarded as
being unsuitable for direct assessment because it is vulnerable to
‘favourable self-biasing’ (Lowin & Craig, 1968; Otley & Pierce,
1995). Second, the Stodgill measure has been criticized for not
providing an indication as to how leadership style may manifest
‘on the job’ (Yukl, 1989). This is limiting in the context of the
present analysis, given the focus on examining practical
expressions of both leadership and budgetary style. Finally, it
is possible for the Stodgill measure to identify managers who
score highly on both initiating structure and consideration
dimensions (suggesting that they are both highly task-orien-
tated and highly employee-orientated). The present analysis
seeks a more straightforward instrument which will measure the
relationship between general leadership style (along a broad
continuum from task––employee-oriented) and commitment to
the budget.
446 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
point scale. A Cronbach Alpha reading of 0.71 was
obtained. Factor analysis revealed no sub-com-
ponents, while all items loaded above 0.40 (Van
der Stede, 2001).
Managerial performance
In an adaptation of the Mahoney et al. (1963,
1965) overall self-rating measure, a three-item
instrument was used to measure managers’ views
of their performance. This instrument includes
both an absolute (the Mahoney et al measure) and
two relative measures of performance (see ques-
tions 31–33 in the Appendix). These relative
measures were crucial, given the focus on under-
standing how well budgetary performance enables
the manager to achieve a sense of security in the
context of his or her place of employment. A
Cronbach Alpha score of 0.75 was obtained for
this three-item instrument.
While concern has been expressed in the liter-
ature regarding the use of self-report measures to
assess performance, a study by Venkatraman and
Ramamujam (1987) suggests that this concern is
somewhat misplaced. Managers’ self-rating of
performance tends to be less biased than what one
might expect. Moreover, Dunk (1993) points out
that self-rating of performance tends to be less
lenient than supervisory ratings (Heneman, 1974),
while earlier studies by Parker, Taylor, Barret, and
Martens (1959) and Kirchner (1965) provide fur-
ther support ‘‘for the utility of self-report mea-
sures’’ (Dunk, 1993, p. 580). Dunk (1993, p. 580)
goes on to explain that, ‘‘even though the rating of
subordinate has traditionally been the responsi-
bility of supervisors, superior managers may not
necessarily be well placed to rate the performance
of their subordinates.’’ This was precisely the case
at Infotain.
Pre-testing and piloting
Several instruments have been devised for the
?rst time in this study, and as such, particular
attention was paid to the pre-testing and piloting
of the questionnaire. The aim was to ensure that
the questions used to measure each variable were
unambiguous and captured the constructs of
interest. Standard practices were employed,
including the use of a panel of experts to assess the
questionnaire for understanding, style of question
(e.g. trade-o? approach), length of question, lay-
out and so on. The questionnaire was revised on
the basis of comments received. A second phase of
pre-testing and piloting was then undertaken, in
which the 26 managers who formed the interview
set during the earlier ?eld study were given a copy
of the questionnaire and asked to assess the sub-
stance, relevance and clarity of the proposed
questions. Final adjustments to the questionnaire
(mostly to the wording of questions) were made on
the basis of feedback received.
Analysis and results
Descriptive statistics
Table 2 reports means and standard deviations
and, where appropriate, both theoretical and ac-
tual ranges for the variables measured in the study.
Test of hypotheses
Empowerment and role ambiguity
Hypothesis 1 predicts that empowerment will be
related to the experience of role ambiguity. The
Pearson correlation coe?cient is 0.29 ðp < 0:01Þ,
suggesting a signi?cant positive relationship
between perceived empowerment and role ambi-
guity.
8
A crystallisation of attitudes
Hypothesis 2 seeks to re?ect the argument that
attitudes towards the budget are crystallised
as levels of subjective or experienced ambiguity
8
Of course, given the relationships depicted in Fig. 1, it is
important to test for a possible direct link between empower-
ment and budgetary commitment. While the Pearson correla-
tion coe?cient suggests a marginally signi?cant association
between these variables ðp < 0:05Þ, path analysis (Duncan,
1966; Kerlinger, 1986) demonstrates that role ambiguity com-
pletely moderates the relationship between empowerment and
budgetary commitment. From this, we conclude that empow-
erment is linked to budgetary commitment, not directly, but
through the experience role ambiguity.
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 447
increase. To test this, scores for role ambiguity are
divided at roughly the mid-point on the theoretical
range and compared against budgetary commit-
ment (also divided at the mid-point) by way of
Chi-squared analysis.
9
Results are signi?cant at
p < 0:001, indicating support for H2 (see Table 3);
those who experience high levels of role ambiguity
are, it seems, more likely to commit to meeting the
budget than those whose experiences of role
ambiguity are minimal.
10
An over-riding in?uence
Hypothesis 3 predicts that the crystallising of
commitment in situations of high role ambiguity
will suppress the potential for other factors to
in?uence budgeting behaviours in these circum-
stances. This argument is tested using a stepped
process of analysis. In the ?rst instance, it was
necessary to establish the extent to which each of
the variables––role ambiguity, leadership style, the
superior’s expectations of budgetary performance
and occupational socialisation––is signi?cantly
associated with budgetary commitment in its own
right. Table 4 presents Pearson correlation coe?-
cients. All are signi?cant to varying degrees, sug-
gesting that each variable is associated with
budgeting behaviour. Moreover, results for both
leadership style and the expectations of the supe-
rior may be claimed to be in the expected direc-
tion, given the results obtained by Hopwood
(1974).
11
The second stage of the analysis involved
examining the extent to which the experience of
role ambiguity over-rides the ability of each of the
other variables to in?uence budgetary commit-
ments. This was done, in the ?rst instance, through
two multiple regression analyses, which are dif-
ferentiated only by the role ambiguity data used.
For one, the measure of role ambiguity comprises
those scores equal to and in excess of the theo-
Table 2
Descriptive statistics
Variable Items Theoretical range Actual range Mean SD
Min. Max. Min. Max.
Budgetary
commit
8 8 40 11 38 25.87 7.03
Role
ambiguity
6 6 30 11 29 20.35 5.91
Empower-
ment
5 5 25 7 24 14.23 3.76
Leadership
style
5 5 25 8 21 15.67 4.02
Superior’s
expectations
4 4 20 8 18 11.09 3.58
Performance 3 3 15 5 15 10.04 2.56
9
For the purposes of this paper, scores below the theoretical
mean (i.e.
Much consideration has been given over the years to what may be described as the ‘negative’ aspect of budgeting;
that budgets may constrain innovation and learning, and that budgetary pressure may lead to unintended behavioural
side effects. In contrast to this, the present study examines the extent to which budgets have a more positive, ‘comforting’
role to play in the individual’s work experiences
Coping with ambiguity through the budget: the positive
e?ects of budgetary targets on managers’ budgeting behaviours
David Marginson
*
, Stuart Ogden
1
Manchester School of Management, UMIST, P.O. Box 88, Manchester M60 1QD, UK
Abstract
Much consideration has been given over the years to what may be described as the ‘negative’ aspect of budgeting;
that budgets may constrain innovation and learning, and that budgetary pressure may lead to unintended behavioural
side e?ects. In contrast to this, the present study examines the extent to which budgets have a more positive, ‘com-
forting’ role to play in the individual’s work experiences. We argue that managers confronted with uncertainties
associated with role ambiguity may respond by becoming positively committed to achieving budgetary targets as
budgets o?er a source of structure and certainty. We ?nd that the use of budgets as an antidote to role ambiguity is a
powerful in?uence on the manager’s budgeting behaviour. We test the strength of this e?ect and we ?nd that budgetary
commitment brought on by the experience of role ambiguity may over-ride the potential for recognised explanatory
variables such as leadership style, the expectations of the superior, and occupational socialisation, to inform managers’
budgeting behaviours in these circumstances. Budgets, it seems, may be as useful to the individual as they are prob-
lematic.
Ó 2004 Elsevier Ltd. All rights reserved.
Introduction
Budgets often receive a ‘bad press’. Not only
are they accused of sti?ing innovation and learning
(e.g. Ansari, 1979; Argyris, 1977; Bartlett &
Ghoshal, 1993; Emmanuel, Otley, & Merchant,
1990; Hedberg & Jonsson, 1978; Hope & Fraser,
2003; Johnson & Gill, 1993), but a high emphasis
on budget attainment is deemed liable to create
various behavioural side e?ects that are likely to
prove dysfunctional to the ?rm. Issues examined
include data manipulation (Hopwood, 1972),
interdepartmental strife (Argyris, 1960), job-re-
lated tension (Brownell, 1981; Hopwood, 1972),
group based ‘anti-management’ behaviour (Argy-
ris, 1952), ‘gaming’ (Hofstede, 1968) and short-
termism (Merchant, 1990). The rather negative
view of budgets is also evident in the number of
studies that have, following the pioneering work of
Hopwood (1972) and Otley (1978), sought to
understand the extent to which factors such as
participation (e.g. Brownell & Dunk, 1991;
Brownell & Hirst, 1986), locus of control (Gov-
indarajan, 1988; Harrison, 1993), task uncertainty
(Brownell, 1983; Hirst, 1981), and degree of stra-
tegic change (Abernathy & Brownell, 1999), may
moderate the expected dysfunctional consequences
of a high reliance on accounting measures of
*
Corresponding author. Tel.: +44-161-200-3459.
E-mail addresses: [email protected] (D.
Marginson), [email protected] (S. Ogden).
1
Tel.: +44-161-200-3455.
0361-3682/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aos.2004.05.004
Accounting, Organizations and Society 30 (2005) 435–456
www.elsevier.com/locate/aos
performance (RAPM) in the context of perfor-
mance evaluation (see reviews by Brownell, 1982;
Hartman, 2000; Kren & Liao, 1988; Otley &
Fakiolas, 2000; Shields & Shields, 1998). Such re-
search continues (e.g. Otley & Pollanen, 2000; Van
der Stede, 2000). However, despite considerable
evolution in the conceptualisation and speci?ca-
tion of variables deemed relevant in this context,
researchers appear to be still some way from
reaching de?nitive conclusions as to what condi-
tions are most likely to allow RAPM to operate
e?ectively without incurring the ‘problems’ that
Hopwood had predicted would accompany budget
constrained styles of supervision. Part of the
explanation for this lack of convergence is attrib-
uted to confusion and ambiguity in the meaning
and measurement of the RAPM construct itself;
problems with the use of methods for data col-
lection and analysis; and the uncritical use of
contingency theory (Briers & Hirst, 1990; Brownell
& Dunk, 1991; Chapman, 1997; Hartman, 2000;
Lindsay, 1995; Otley & Fakiolas, 2000; Vagneur &
Peiperl, 2000). Nevertheless, whatever reasons we
may invoke to explain the inconsistency of the
research ?ndings, one consequence of the persis-
tence with the research agenda raised by these
debates has been to perpetuate analysis of the
dysfunctional e?ects of budgets to the exclusion of
any consideration of the more positive role(s) that
budgets may play in people’s work-related expe-
riences. Since the work of Argyris (1952) and
others (e.g. Caplan, 1966; Stedry, 1960), the per-
spective adopted in the budgeting literature has
been one of ‘‘Human problems with budgets’’
(Argyris, 1953).
The lack of a counter-balancing ‘positive’ per-
spective on budgeting is surprising, not least be-
cause the potential for budgets to play a more
functional role in managers’ work-related experi-
ences has been suggested by the human relations
movement in accounting. Path-goal theory of
?nancial controls, for example, suggests that
where managers do not have self-evident paths and
clear-cut goals, they will welcome accounting
based controls such as budgets for the structure
and certainty they provide. This is particularly true
of senior managers, whose roles are frequently
ambiguously de?ned. Macintosh (1995, p. 215)
argues that budgets ‘‘serve to reduce sharply role
ambiguity’’. At the same time, interesting and
persuasive as these arguments are, they remain
under-developed and certainly under-researched.
In particular, little is currently understood about
how or why managers may use the budget to cope
with uncertainty and ambiguity, what this means
in terms of budgeting behaviours, and more
importantly, perhaps, what the implications may
be for individual and organizational performance.
The question of whether or not budgets may play a
positive role in people’s work-based experiences
remains a neglected topic.
The importance of understanding the potential
functionality of budgets at the level of the indi-
vidual is evident in recent empirical research on
contemporary attitudes of senior managers to
performance evaluation based strictly on account-
ing measures of performance. Storey, Edwards,
and Sisson (1997), for example, found that man-
agers, faced with increasingly tight budgetary
targets, not only accepted but positively endorsed
the rigorous degree of accountability they experi-
enced. The main reason for this was the extent to
which managers welcomed the certainty created by
the combination of clear goals and well-speci?ed
objectives, and a performance evaluation system
that was tightly focused on how successfully those
objectives have been achieved. Contrary to the
expectations associated with the dysfunctional as-
pects of RAPM, the degree of close monitoring
and evaluation of performance did not produce
discontent. Rather, Storey et al. (1997, pp. 201–
202) found that ‘‘tight evaluation was linked to a
favourable view of the system of evaluation’’ and
that managers were ‘‘most likely to be strongly
motivated when they felt that their activities were
closely evaluated by their companies’’. Storey et al.
(1997, p. 203) explain the sense of satisfaction ex-
pressed by managers as the result of their sense
that companies with demanding targets appear to
know where they are going and what they want
from their managers. Managers generally appre-
ciated the sense of strong direction from the top, as
expressed in the detailed speci?cation and tight-
ness of their performance targets, and welcomed
the sense of certainty and mission this communi-
cated.
436 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
In the example quoted above two distinct, but
clearly related, processes are discernible in man-
agers’ responses to RAPM. On the one hand, there
is positive endorsement by managers of the goal
speci?city and goal clarity that operates in the
context of performance focused companies. On the
other hand, there is also appreciation of the use
made of budget constrained styles of supervision
which focus on evaluating how successfully speci-
?ed goals have been achieved. Viewed from this
perspective, the absence of such clarity and speci-
?city and the reliance on a more relaxed supervi-
sory style with regard to APM, as envisaged, for
example, by Hopwood’s (1972) notion of a ‘pro?t
conscious’ style of supervision, may paradoxically
cause more problems than it solves. It may be
suggested that where this certainty provided by a
clear emphasis on budget goals is lacking manag-
ers may seek ways to recreate it for themselves. In
this context, identifying with the budget may be
seen as functional by managers precisely because it
is seen as providing a clearly de?ned performance
target which, if achieved, will provide a degree of
security amidst the inherent ambiguity of the
managerial role (Macintosh, 1995).
The presence of role ambiguity provides the
focus for our examination of the potential func-
tionality of budgets at the level of the individual.
Speci?cally, we examine the extent to which man-
agers may actually commit to meeting budgetary
targets, not because of the threat of accountability
or the promise of reward, but because doing so
o?ers a sense of clarity and security, particularly in
circumstances where their role is fraught with
uncertainties. In e?ect, we seek to understand how
far managers who experience anxiety brought on
by ambiguity and uncertainty within their role may
use budgets as a means of coping with this anxiety.
We address this aim through the analysis of evi-
dence garnered from a recently conducted com-
pany case study. The sources of role ambiguity are
explained, and managers’ reactions to this and
their attempts to secure for themselves a more
gratifying role experience through commitment to
budgetary targets are traced. Using quantitative
evidence we are also able to test the strength of the
e?ect that role ambiguity may have on managers’
behaviour towards budgetary targets against the
more usual explanations for budgetary commit-
ment by managers suggested in previous research.
We argue that budgetary commitment brought on
by the experience of role ambiguity may over-ride
the potential for recognised explanatory variables
such as leadership style (Argyris, 1952), hierarchi-
cal contagion (Hopwood, 1974) and occupational
socialisation (Brownell, 1985; Hirst, 1981) to in-
form managers’ budgeting behaviours in these
circumstances. Moreover, as the case study
company exhibited a distinct lack of emphasis
on RAPM and a marked absence of formal
accountability and reward from the budgetary
system, we are able to examine the role played by
budgets, of themselves, as an antidote to role
ambiguity in circumstances unencumbered by the
presence of factors which would undoubtedly
confuse the analysis and limit interpretation of the
results.
The section that follows introduces the case study
and presents the qualitative evidence that identi?es
managers’ experience of role ambiguity, its causes,
and their reactions to it. Section 3 develops several
arguments relating to the use of budgets as coping
mechanisms. Section 4 describes the quantitative
methodology that was used to test the resulting
hypotheses, while Section 5 presents the results of
this analysis. Section 6 outlines the study’s limita-
tions and concludes the discussion. Our ?ndings
suggest that the experience of role ambiguity may
exercise considerable in?uence on managers’ bud-
geting behaviours; it may ensure their budgetary
commitment in the search for a more gratifying role
experience. The implications of these ?ndings are
discussed in the concluding section.
The case study: Infotain plc
The case study company, which we have called
Infotain plc, is a major UK (FTSE100) based
organization which operates in what it describes as
the ‘global communications business’. Infotain plc
is a prime example of a modern ‘professional adh-
ocracy’ (Euske, Lebas, & McNair, 1993) or infor-
mation-age company (Bartlett & Ghoshal, 1993;
Simons, 1995, 1999). De?ning features include
fast-moving circumstances, jurisdictional and
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 437
decisional ambiguity, ?uid role responsibilities,
frequent job transformations, a constant need to
prioritise among tasks, and front-line empower-
ment. The company supplies and maintains a range
of communication and infotainment products and
services to both major private sector companies and
large public corporations. It employs around
12,000 people, and operates from several sites
located primarily in the UK. Turnover exceeds £4
billion per annum.
The aim of the case study was to examine the
design and operation of the company’s control
processes and procedures, with particular attention
being paid to investigating the ‘human dimension’
of management control. Of the areas investigated,
research into budgeting’s role in the process of
management control was intentional, not least be-
cause budgets have long been viewed as a, if not the,
pivotal organizational control mechanism (Ansari,
1979). Several other issues, including the company’s
empowerment initiative, managers’ experience of
role ambiguity, and the notion of budgetary com-
mitment, were also explored in some detail, given
initial insights which showed these to be interlinked
and potentially signi?cant to organizational per-
formance. All issues were addressed through inter-
views with senior and line managers; meetings with
personnel from both the accounting and human
resource functions; and scrutiny of company
memoranda, reports and other documentary evi-
dence. Interviews were used to elicit managers’
views, beliefs, and actions about the control systems
at their disposal, and were conducted over the two-
year period of this case study. During this time,
initial ideas and arguments were developed and
re?ned in the light of additional evidence through,
for example, further interviews with managers. The
following seeks to elaborate on the key issues to
emerge from the qualitative ?eld work.
For reasons relating to the company’s highly
organic context and the need to ensure a high rate
of strategic adaptation and change at ?rm level,
given a fast-moving business environment (Dut-
ton, Ashford, O’Neil, Hayes, & Wierba, 1997),
Infotain had largely abandoned budgeting in its
traditional form. It adopted a broader control
environment which sought to encourage strategic
adaptation and change through, for example, the
use of beliefs systems (Simons, 1995). Although
‘prudent’ budgetary goals and other ?nancial tar-
gets were set by the accounting function to instil
cost consciousness throughout the ?rm, there was
little formal accountability or reward attached to
budgetary performance. Only the most senior
managers had ?nancially based accountabilities.
Cost centre managers, although allocated re-
sources, were not explicitly assessed or rewarded
on their ability to remain within budget. Indeed,
tolerance limits were attached to overall budget
allocations, which meant that budget targets could
be exceeded (or under-achieved) by upwards of
10% for ‘strategic reasons’. ‘Out of spends’ were
also possible if the manager had a good reason to
exceed the 10% tolerance limit. Managers could
also renegotiate their budgets periodically in the
light of unfolding events. The reasoning behind
this ?exibility, according to top management, was
to ensure, or at least not to deter, strategic adap-
tation in response to a rapidly changing, highly
uncertain environment. Managers were therefore
given considerable licence to sacri?ce initial bud-
getary targets in favour of unexpected, but
potentially pro?table, opportunities if and when
they arose during the budgetary period. They were
speci?cally encouraged to balance the need for
variance correction with (1) the need to meet
launch deadlines for new products and services
and (2) the possibility of making improvements to
on-going projects as new information became
available, notwithstanding the extent to which
meeting agreed time-frames and modifying pro-
jects often required additional resources.
The degree of discretion that managers were
a?orded in terms of budgetary performance mir-
rored the ethos of empowerment which pervaded
the organization. For reasons similar to those
which appeared to drive the company’s seemingly
radical approach to budgeting, there was a general
reluctance to de?ne and limit individual responsi-
bilities. This was epitomised in the company ’motto’
‘‘tell me what you’ve done, don’t ask me what to
do’’. Middle-ranking managers were particularly
empowered to indulge in ’autonomous strategic
behaviour’ (Burgelman, 1983a, 1983b), ‘issue-sell-
ing’ (Dutton et al., 1997), and other grass-roots
activity that typi?es an emergent strategy process
438 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
(Mintzberg & McHugh, 1985). Most were members
of at least one autonomous multifunctional project
team, and several remarked on how they were
‘‘free’’ to commit to accountabilities that would
take them ‘‘completely away from any form of
hierarchical relationship’’. Managers saw them-
selves in terms of having ‘‘multiple stakeholders’’ as
a result of ‘‘mutual accountabilities’’ and omni-
present interdependencies (Spekle, 2001).
The company’s empowerment strategy ap-
peared to be supported by managers. The majority
of those interviewed suggested that the dissipation
of responsibility for strategy formation through-
out the managerial ranks was necessary as the
rapid technological advances, short product life
cycles and a ?ercely competitive environment
made the notion of detailed strategic planning
problematic (Mintzberg, 1979). At the same time,
some interviewees did appear uneasy with the high
levels of ambiguity and uncertainty that a lack of a
clear ‘‘authority map’’ and a highly organic con-
text created. This lack of role clarity extended to
performance evaluation. For example, some
managers commented how they received little
formal feedback from their superiors on their role
performance, while their superiors commented
how it was di?cult to do so, given a person’s
activities did not necessarily support the line rela-
tionship. These conditions were prevalent
throughout the ?rm and prompted some managers
to express concern about the extent of their ?uid
role responsibilities, their mutual accountabilities,
and the continual need to prioritise amongst tasks.
For example, one manager remarked:
Things are in a constant state of ?ux around
here, and this can be a bit of an issue for me
personally at times. I like to be clear about
what I am supposed to do, but I appreciate
this can’t really be the case around here.
Another manager commented on how ‘‘the
constant need to prioritise’’ often led to people to
‘‘doubt their choice of priority’’. Other managers
suggested, somewhat more negatively, that the
blurring of roles and responsibilities was ‘‘a nec-
essary evil’’ or just had to be accepted and was
therefore ‘‘something to live with’’.
In examining how managers sought to ‘‘live
with’’ the ambiguity and uncertainty prompted by
the company’s empowerment initiatives, evidence
emerged from the ?eld study to suggest a degree of
‘coping through the budget’. A number of inter-
viewees remarked how committing to meeting
budgetary targets ‘‘provided an anchor’’ or ‘‘focal
point’’ in the midst of jurisdictional and decisional
ambiguities, and the constant push for innovation.
They made it clear that they wanted to achieve pre-
determined budgetary targets, and expected their
subordinates to do likewise. One manager re-
marked: ‘‘I don’t like all this change. If there’s a
budget and a target, then it should be met. End of
story’’. Several others commented on the impor-
tance of this role of budgets through references to
the company’s ethos: when asked to ‘‘tell me what
you’ve done, not ask me what to do’’, they ruefully
said they would ‘‘tell them they had met the bud-
get!’’ One manager, in more serious vein, remarked:
I am no longer sure where my job ends and
someone else’s begins. There’s an incredible
amount of uncertainty, and people are always
being asked to do many di?erent things at
once. The whole thing is a mess. This does
bother me; a hell-of-a-lot at times; I just don’t
know what I should be doing half the time.
But, at least I know where I stand with the
budget. If I keep this in order and hit the tar-
gets, and make sure my guys are doing the
same, then no one can argue with that, not
even the chief executive!
The above evidence suggests both causal direc-
tion and a rationale for the surprising degree of
budgetary commitment that was evident at Info-
tain. Managers’ propensity to pursue budgetary
targets was, in part at least, driven by their desire
to create for themselves a degree of structure and
certainty in the face of the ambiguities and
uncertainties that had become characteristic of
their situation as a result of their experience of
empowerment. Moreover, the evidence indicates
that the pursuit of structure and certainty through
the budget in these circumstances may well over-
ride other factors such as organizational position
and leadership style in determining attitudes
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 439
towards the budget in these circumstances. One
manager, for example, remarked how his man-
ager’s ‘‘harsh’’ attitude towards budgetary targets
was inconsistent with his supervisory style, which
was ‘‘generally supportive and understanding’’.
Further, the use of standard practices for quali-
tative data analysis, including interview coding
and cross-tabulations (Miles & Huberman, 1984),
revealed that those who expressed varying degrees
of concern about their role also tended to show
commitment towards the budget, irrespective of
their organizational position or what could be
interpreted at this stage as their more general ap-
proach to the job. In contrast to this, factors such
as business unit a?liation did emerge as possible
predictors of budgeting behaviours for the group
of interviewees who, through an absence of rele-
vant remarks, showed little or no concern for the
uncertainty they faced as part of their role. These
are potentially important observations, not only
because they challenge longstanding assumptions
about the conditions required to ensure budgetary
commitment, but also because they suggest that
the behavioural literature in accounting may have
omitted a signi?cant determinant of budgeting
behaviours: the experience of role ambiguity.
These insights drawn from the qualitative data
may be further explored through the analysis of
quantitative data. While this complements the
qualitative data in examining how and why bud-
gets may be used as a coping mechanism against
the ambiguity experience brought on by empow-
erment initiatives, it is distinctive in that it pro-
vides a basis for testing the strength of the
relationship between role ambiguity and budget-
ary commitment.
2
Role ambiguity and budgetary commitment
In this section, we develop a number of argu-
ments in order to test the strength of the relation-
ship between role ambiguity and budgetary
commitment. We argue that empowerment may
constitute a source of role ambiguity. We then
consider the concept of role ambiguity before pre-
senting the arguments as to why budgets may be
useful mechanisms to alleviate stress and comfort
managers a?ected by role ambiguity. We conclude
by arguing, ?rstly, that the use of budgets as a
coping mechanism leads managers to commit to
meeting budgetary targets, irrespective of the pres-
ence of some factors (i.e. recognised explanatory
variables such as leadership style and the superior’s
expectations), and the absence of others (e.g. formal
accountability and reward), and secondly that
coping is related to a gratifying role experience
(Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964).
The various arguments and proposed relationships
are depicted in Fig. 1, which presents the frame-
work for the forthcoming discussions.
Empowerment and role ambiguity
A fundamental aim of empowerment initiatives
is to embed notions of strategy formation
throughout the ?rm (Otley, 1994). As middle-level
managers are often closer to the customer and
other stakeholders than top management, and
therefore have more detailed knowledge of what
strategic issues require attention, empowerment
initiatives frequently envisage middle-level man-
agers playing pivotal roles in detecting new ideas
and in mobilising resources around these new ideas
(Dutton et al., 1997; Kanter, 1982; Simons, 1995).
In pursuing these ‘strategic forcing’ roles (Burgel-
man, 1983a, 1983b), managers are encouraged to
undertake ‘boundary-spanning’ activities, apply
individual initiative, and work as ‘team players’
(Euske et al., 1993). However, such activities tend
to result in the blurring of role responsibilities and
traditional line relationships, and, at times, poor
communications between superior and subordi-
nate. In this respect, empowerment initiatives
undoubtedly disrupt the chain-of-command prin-
ciples characteristic of formal hierarchical man-
agement structures (Bartlett & Ghoshal, 1993;
House & Rizzo, 1970) and, in so doing, may create
or increase role ambiguity (King & King, 1990).
Role ambiguity occurs when an individual is
unsure about others’ expectations of him-or her-
2
In view of the ?eld evidence, and in anticipation of the
analysis to follow, causal direction is implied in the arguments
that are developed.
440 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
self. It has been conceived as the ‘‘discrepancy
between the amount of information a person has
and the amount s/he requires to perform his or
her role adequately’’ (Kahn, 1974, p. 426). The
increased job transformations and ?uid role
responsibilities precipitated by empowerment ini-
tiatives may leave the manager unsure about what
his or her duties and responsibilities are. The
manager may also be unclear about the activities
which should be performed in order to ful?l these
responsibilities, while the consequences of action
(or inaction) may be unknown. Such ‘objective’
role ambiguity is, in many respects, a fact of
managerial life (Macintosh, 1995), but people
have an inherent, if varying, need for role clarity
(Elovainio & Kivimaki, 2001; Eysenck, 1954;
Kahn et al., 1964).
3
As a result, a close corre-
spondence is expected between objective and
subjective, or experienced, role ambiguity (Kahn
et al., 1964; Netemeyer, Johnston, & Burton,
1990). In general, we prefer to know what our
duties and responsibilities of o?ce are, and how
we are supposed to discharge them, while from a
socioemotional perspective, it is important that we
know what the consequences of performance or
non-performance will be, particularly for our-
selves, but also, perhaps, for our colleagues and
the organization in general (King & King, 1990).
However, this sort of means-ends knowledge is
unlikely to be forthcoming from the ?uid role
responsibilities and the high levels of jurisdictional
and decisional ambiguity created by empower-
ment initiatives. Rather, the ?exible approach to
task requirements and the boundary spanning
activities prompted by empowerment are likely to
lead to imprecise lines of authority and lack of
clarity about what is required. Consequently, we
may argue that managers’ experience of empow-
erment is likely to be related to their experience of
role ambiguity. We therefore hypothesise that:
H1 (A source of role ambiguity): Empower-
ment is associated with role ambiguity.
Coping with role ambiguity through the budget
Experienced role ambiguity is an unwanted
psychological state, which occurs when the person
is concerned or ‘stressed’ by a lack of clarity and
Role ambiguity
(e) H5 -ve
Empowerment
(a) H1+ve
Performance
Budget commitment
(b) H2 +ve (d) H4 +ve
(a) A source of role ambiguity
(b) For high levels of role ambiguity
(c) Sig. only in the absence of role ambiguity
(d) Regaining a gratifying role experience
(e) A gratifying role experience regained
Superior’s
expectations
Leadership
style
(c) H3 sig (c) H3 sig (c) H3 sig
Occupational
socialization
Fig. 1. Diagrammatic representation of the arguments.
3
The need for role clarity can vary according to a person’s
tolerance of objective role ambiguity (Eysenck, 1954).
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 441
structure to the role (King & King, 1990; Siegall,
2000). Doubts about how others evaluate us,
about how satis?ed they are with our performance,
are frequent and deep-seated sources of anxiety
(House & Rizzo, 1970; Kahn et al., 1964; King &
King, 1990). As a result, those who experience role
ambiguity must confront it in some way (Kahn,
1974; Siegall, 2000), usually by invoking ‘coping
strategies’. These are attempts to regain clear, or-
derly and meaningful cognitive experiences. Such
attempts may involve the use of defence mecha-
nisms, which are able to distort the reality of the
ambiguous situation in order to relieve the anxiety
of the undistorted experience (Dougherty & Prit-
chard, 1985). However, within the role theory lit-
erature there is little elaboration of what these
defence mechanisms may consist of. Consequently,
little is actually known about the nature of these
defence mechanisms, except for evidence to sug-
gest that it may involve the person leaving the role
(King & King, 1990). Notwithstanding this, our
evidence indicates that it is plausible to argue that
identifying with the budget by emphasising the
achievement of short-term budgetary targets con-
stitutes an important defensive strategy against the
experience of role ambiguity; it is certainly one
which most managers within the ?rm can invoke.
There are several reasons why budgets o?er a
useful coping strategy. First and foremost, bud-
getary control is organizationally ‘good citizen-
ship’. Those who are controlling their costs are
bene?ting the ?rm (Merchant, 1998). Moreover,
because of the visibility of accounting information
(Roberts, 1991), the manager can be seen to be an
organizationally ‘good citizen’. This is important,
as it means that the manager is able to predict,
with a high degree of con?dence, the consequences
of performance. More importantly still, the
acquisition of such knowledge does not depend,
for example, upon clear communication with the
hierarchical superior about what should be done,
how it should be done, or what kinds of behav-
iours are likely to result in reward. Rather, the
manager autonomously ‘knows’ what needs to be
done to acquire a positive organizational identity,
knowledge which is aided by the ‘mechanistic’
credentials of budgetary systems (e.g. assumptions
of certainty and independence of responsibility––
Ezzamel & Hart, 1987). These make it clear what is
required (achieve the budget), who is responsible
(the manager), and, given the arithmetical nature
of accounting reports (Hopwood, 1976), what
regulatory action must be taken if the manager is
to meet his or her own budgetary target: instruct
the subordinate to do likewise (Hopwood, 1976).
Those who experience role ambiguity may
therefore value the budget for the comfort and
security it o?ers. But, such psychological return is
conditional upon the achievement of targets; there
is little comfort to be gained from exceeding pre-
established targets or missing the budget. Thus, if
budgetary control is to be used as a coping strategy,
the manager must commit to the budget, and must
exercise ‘tight budgetary control’ to achieve bud-
getary targets. Anything less decreases the proba-
bility of meeting the budget and, thereby,
jeopardises the possibility of gaining the sense of
socioemotional security that is desired. There may
be valid ‘strategic reasons’ for exceeding the budget,
but these reasons will be less visible to others and
longer-term in their fruition. Role ambiguity is not
avoided by these actions; rather ambiguity increases
as objects and events recede in time and space, a fact
which most people appear to accept (King & King,
1990). Those who experience role ambiguity are
thus unlikely to adopt a ?exible approach to the
budget or even to forgo budgetary targets in favour
of longer-term payo?s; such behaviour serves to
increase perceptions of uncertainty (e.g. how much
?exibility) and ambiguity (e.g. is it for the ‘right’
strategic reason that the subordinate is being in-
structed to forgo the budget) at a time when cer-
tainty is desired. The coping strategy is committing
to the budget. The reward is a more gratifying role
experience (House & Rizzo, 1970).
An over-riding in?uence?
The corollary of the above argument is that
those who experience little or no role ambiguity
may adopt a more ?exible attitude towards the
issue of budgetary control, because they have little
need to ‘cope’. The absence of role ambiguity re-
moves the need for the manager to commit to
meeting the budget in order to counter or distort
the ambiguity experience. However, in view of the
442 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
?ndings of prior research, we may argue that
budget ?exibility in these circumstances is likely to
be shown only to the extent that variables such as
leadership style (Argyris, 1952; Hopwood, 1974),
the supervisor’s behaviour towards the budget
(Hopwood, 1974), or even the manager’s previous
work experiences and current functional location
(Hirst, 1981) do not themselves result in budgetary
commitment. People for whom the experience of
role ambiguity is minimal or absent may still
exercise tight budgetary control because of their
task-orientated approach to the job (Hopwood,
1974) or because of the e?ects of hierarchical
contagion (Bonini, 1963; Hopwood, 1974). Equal-
ly, of course, these managers may not exercise tight
budgetary control because their style of leadership
is more considerate or employee-oriented, or be-
cause this is what the superior expects, or indeed
because they are merely re?ecting established views
within the particular business unit or function in
regard to the importance of budgets (Hirst, 1981).
In e?ect, therefore, it cannot be suggested cate-
gorically that budget ?exibility is associated with an
absence of role ambiguity. But, what we can argue
with some conviction is that, because role ambigu-
ity triggers a seemingly powerful desire for structure
and certainty (Kahn et al., 1964; Siegall, 2000), and
because budgets o?er a potentially e?ective coping
mechanism, those experiencing role ambiguity are
likely to commit to meeting the budget in these
circumstances. The ambiguity experience crystal-
lises attitudes towards the budget in a way that
non-experience does not. The strength of the rela-
tionship between role ambiguity and budgetary
commitment may be assessed by examining the
extent to which the experience of role ambiguity
overrides or suppresses other explanatory variables
in determining the manager’s behaviour towards
the budget. The most appropriate variables to in-
clude in the analysis are those, such as leadership
style and hierarchical contagion, which have been
shown previously to shape managers’ budgeting
behaviours. We will therefore examine the argu-
ment that the experience of role ambiguity leads the
manager to commit to meeting the budget irre-
spective of (1) his or her natural inclination to be-
have otherwise, as is the case with a considerate or
employee-orientated style of leadership (Hopwood,
1974), (2) what the superior may expect of the
manager (DeCoster & Fertakis, 1968; Hopwood,
1974), and (3) the e?ects of occupational socialisa-
tion, as demonstrated through business unit a?li-
ation.
The formal hypotheses for the above arguments
are as follows:
4
H2 (A crystallisation of attitudes): Those who
experience high levels of role ambiguity are
more likely to commit to meeting the budget
than those whose experience of role ambiguity
is minimal or absent.
H3 (An over-riding in?uence): The experience
of role ambiguity over-rides recognised
explanatory variables (leadership style, hierar-
chical contagion) and other predicted factors
(e.g. occupational socialisation) in shaping
attitudes to the budget in these circumstances.
H4 (A coping strategy for gaining or regain-
ing a gratifying role experience): The higher
the commitment to the budget, the higher
the level of self-reported performance.
5
4
The ?ve hypotheses seek to re?ect the various components
of the ‘theory’ (i.e. source of role ambiguity, asymmetry,
suppression of recognised explanatory variables, and ‘coping’).
Their separate examination is necessary, given the intricacies of
the arguments. In fact, a more complete picture is possible
through this cumulative approach to the empirical analysis.
5
The role theory literature stresses how people invoke
coping mechanisms in a bid to obtain a more gratifying role
experience, but o?ers little guidance on what may constitute a
‘gratifying role experience’ or how this may be measured. The
present study sought to construct this phenomenon in terms of
performance. The reasons are twofold. First, it is reasonable to
argue that one measure of ‘grati?cation’ is performance; those
who are satis?ed with their role are likely to feel that they are
performing well (and vice versa). Second, and more importantly
perhaps, given a lack of formal accountability and reward at the
research site, and given a general emphasis on innovation and
learning, the manager has much less need to evaluate his or her
worth to the organization against accounting numbers (Rob-
erts, 1991). Self-evaluation of performance in this context will
thus represent a feeling of general well-being within the role
rather than a speci?c measure of task accomplishment.
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 443
H5 (A gratifying experience regained): Per-
formance is inversely related to role ambigu-
ity.
Method
The arguments were tested on data collected
from a questionnaire survey involving ‘level three’
managers drawn from ?ve of Infotain’s eight
strategic business units. The business units were
chosen to represent a spectrum of functions,
including engineering, product development and
sales and marketing. All level three managers from
these business units were asked to participate in
the questionnaire survey. Completed question-
naires amounted to 221 out of a possible 270, of
which all were usable. The response rate was 82
percent.
The choice of managerial level was dictated by
several factors. First, level 3 managers represent,
for the ?rm in question, junior executives with
budget responsibilities and regulatory duties (level
4 managers were also allocated resources). Sec-
ond, meetings with accounting and human re-
source personnel and the perusal of documents
relating to formal accountabilities established
several important facts about the way budgetary
control was exercised at this level. One is that
these junior executives were neither assessed nor
rewarded for staying within budget. Another is
that most managerial ranks, including the one in
question, were not, by and large, able to partici-
pate in the setting of their budgetary targets.
Instead, ‘prudent’ budgets tended to be imposed
by the accounting function in a bid to establish
cost consciousness throughout the managerial
hierarchy. When questioned about whether or not
they accepted their budgetary targets, managers
responded by saying they had ‘‘no choice’’! This
is important in the context of the present study as
the lack of participation in setting budgetary
targets is normally expected to make managers
less committed to achieving them. Commitment
to budgetary targets, even for intrinsic reward, is
considered to be inexorably bound with notions
of perceived legitimacy and ‘ego involvement’,
which come from managers’ ability to participate
in the setting of budgetary targets (Collins, 1982).
As regards the achievability of their budgetary
targets, managers when asked in interviews
commented that they were moderately di?cult or
di?cult. None suggested that the budget was
easily achievable.
A ?nal reason for focusing on level 3 managers
is that managers at this level and below were
empowered to behave ‘strategically’; they could, if
necessary, exceed their original budgetary targets
for ‘strategic reasons’. The same applied at the
next level down, but it was felt that lower-ranking
managers may generally feel more insecure, and
therefore might be more predisposed to experi-
encing role ambiguity. The intent of the study was
to focus attention on the highest managerial level
possible where any feelings of anxiety and insecu-
rity may be minimised, thereby establishing a
context which would make signi?cant results less
rather than more likely.
Measurement of the variables
Commitment to the budget: tight budgetary control
Ten questions were initially devised to measure
managers’ commitment to the budget as manifest
in the exercise of ‘tight budgetary control’ (Van
der Stede, 2001). Tightness may be exhibited in
two main ways. One is that which is imposed by
the budgetee on him or herself. The other is the
tightness that is imposed by the budgetee on the
subordinate. Both are necessary if the budgetee is
to obtain security from meeting the budget
(Hopwood, 1976). Hence, both these approaches
were applied in devising the questions. The ten
questions are shown in the Appendix (questions 1–
10).
Con?rmatory factor analysis was employed to
determine validity. This resulted in the original
ten-item instrument being reduced to eight, as
survey items that load above 0.40 were retained
(Dutton et al., 1997; Van der Stede, 2001). A
Cronbach Alpha score of 0.76 was obtained for
this eight-item measure, which is comfortably
above lower limits of acceptability, generally con-
sidered to be around 0.50–0.60 (Nunnally, 1967),
and may be regarded as reasonable for an instru-
ment which is being used for the ?rst time. Cron-
444 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
bach Alpha for all ten original items was 0.66
(Cronbach, 1951). Thus, omitting the two items
with weak factor loadings ‘‘not only makes the
scale more parsimonious, it also enhances its
overall reliability’’ (Van der Stede, 2001, p. 129).
Factor loadings are reported in Table 1. The two
omitted items are statements 3 and 7 (see the
Appendix).
The underlying principle that was applied in
devising the research instrument was that of
‘trade-o?’; budgetary commitment through ‘tight
budgetary control’ (Van der Stede, 2001) was
measured, not by examining, for example, ‘the
amount of emphasis placed on attaining budget
targets’ per se, but by examining the extent to
which such emphasis may be exhibited in the
context of the need for strategic adaptation and
change. Tensions between these important aspects
of organizational endeavour are inherent to ?rms
(Simons, 1995); they were a particular feature at
Infotain. It is likely that most ?rms would like
managers to strike a balance between budgetary
control on the one hand, and innovation and
learning on the other. To that end, getting
respondents to demonstrate their preferences and
priorities in trade-o? situations is a potentially
useful way of establishing strength of budgetary
commitment. Respondents were therefore asked to
what extent they expected budgetary targets to be
met, given the need to consider (1) long-term
?nancial performance, (2) problems of controlla-
bility (Choudhury, 1986), (3) unfolding events and
opportunities, and (4) strategic developments.
Items 6 and 10 are used as overall anchors (Ma-
honey, Jerdee, & Carroll, 1963, 1965). Each item
was measured on a ?ve-point scale, with 1 repre-
senting strong agreement and 5 strong disagree-
ment with the statement presented. Items 1, 7, 8
and 9 were reverse scored.
An expectations approach is used to measure
the degree of tightness as applied to the subordi-
nate. This is in keeping with the ?ndings of the
?eld study and suggestions in the literature (e.g.
Machin, 1979). The communication of expecta-
tions represents an in?uence attempt, which is
aimed at ensuring compliance with what is ex-
pected (Kahn et al., 1964; Machin, 1973, 1979).
Thus, expecting pre-determined budget targets to
be achieved in the context of trade-o? represents
budgetary commitment. It is ‘tight budgetary
control’.
Experienced empowerment
The notion of empowerment has been exten-
sively discussed, but there are surprisingly few re-
search instruments that have been devised to
detect and measure the extent of a manager’s
empowerment in a business context. In this study,
we use a ?ve-item instrument (reduced from six
items on the basis of con?rmatory factor analysis)
to measure respondents’ perceptions as to the de-
gree of strategic empowerment they were a?orded
in the discharge of their role. Consistent with
?ndings from the qualitative study, respondents
were asked to indicate the extent to which they
were expected to indulge in various ‘boundary-
spanning’ activities which took them away from
the line relationship with the superior. Each item
was measured on a ?ve-point scale with 1 repre-
senting strong agreement with the statement pre-
sented and 5 strong disagreement. The Cronbach
Alpha score for the ?ve-item instrument is a highly
respectable 0.72 (Nunnally, 1967). The research
instrument is shown in the Appendix (items 11–15).
Role ambiguity
Role ambiguity is assessed using a slightly
adapted version of the well-tested measure ?rst
developed by Kahn et al. (1964) (see questions 16–
21 in the Appendix). A reliability measure of 0.80
Table 1
Factor loadings for budgetary commitment measures
Items in order shown
in Appendix A
Rotated
factor loading
Squared multiple
correlations
1 0.716 0.525
2 0.651 0.566
3 0.340 0.511
4 0.533 0.483
5 0.814 0.879
6 0.583 0.482
7 0.243 0.548
8 0.708 0.624
9 0.488 0.475
10 0.510 0.897
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 445
was obtained for the resulting six-item instru-
ment.
6
The superior’s budgetary expectations
Hopwood’s (1974) research suggests that tight
budgetary control (in the form of a budget-con-
strained evaluative style) is passed on from one
level to the next through the notion of hierarchical
contagion (Bonini, 1963). However, what is of
greater interest here is the extent to which the
budgetee may be committed to meeting the budget
even though the superior may expect otherwise.
Use of an expectations approach (Machin, 1979)
works both ways, in that the superior may expect
the manager to trade budgetary targets in favour
of strategic adaptation. This was evident at Info-
tain: the few level 2 managers that were inter-
viewed suggested that they would be prepared to
forgo budgetary targets for strategic reasons. The
extent to which level 2 managers expected their
subordinate(s) to trade budgetary targets for
strategic reasons was assessed by asking level 3
managers what they perceived to be expected of
them in relation to the budget. A ?ve-item
instrument was devised and reduced to four items
following factor analysis. The four items are
shown in the Appendix (questions 22–25). A
Cronbach Alpha score of 0.79 was obtained for
the four-item instrument.
Occupational socialisation/business unit a?liation
Occupational socialisation (DiMaggio & Pow-
ell, 1983) at the business unit level was assessed
according to membership of a particular business
unit. This was represented as three separate dum-
my coded (0 or 1) current position variables. For
example, in the ?rst variable members of business
unit one (engineering) were coded 1 and the
remainder 0 (Donabedian, McKinnon, & Bruns,
1998; Sim & Killough, 1998). The process was
repeated for each of the other four business units.
Leadership style
A ?ve-item instrument was devised in order to
detect and measure leadership style in action at the
research site.
7
Broadly speaking, each question
sought to measure leadership style along a high
task/low consideration––low task/high consider-
ation continuum. For example, Question 26 asks
respondents to indicate the extent to which they
expect subordinates to revise their responsibilities/
commitments as circumstances change; the more
revisions are sanctioned or expected, the more
employee-orientated the manager, given the cir-
cumstances encountered. (Expecting original tar-
gets to be met in the context of rapidly changing
conditions demonstrates a higher regard for the
task than the individual, who must confront these
‘di?cult’ circumstances.) Each question in the
present instrument was measured along a ?ve-
6
While the paper focuses on anxiety brought on by the
experience of role ambiguity, all 15 items that Kahn et al. (1964)
?rst composed to measure job-related stress and tension
manifest as role ambiguity, role con?ict and role overload were
included in the questionnaire in order to check the validity of
the role ambiguity measure through factor analysis. The six
items for role ambiguity loaded onto one (of three factors) at
0.50 or above. The analysis yielded three Eigen values >1 which
accounted for 68% of the variance. Only the six items relating to
role ambiguity are shown in Appendix A for reasons of brevity.
7
Due to top management placing restrictions on the size of
the questionnaire, the general circumstances encountered, and
the focus of the study (i.e. the manager rather than the
subordinate), a decision was taken not to use the Stodgill (1963)
two-dimensional instrument of initiating structure and consid-
eration. This measure contains a range of questions which, in
the eyes of Infotain’s senior managers, rendered the question-
naire ‘to big’ in terms of resource implications for the ?rm. In
the event, while this instrument may be well tested, it has several
features which limit its usefulness in the present analysis. First,
it is mainly used to measure subordinates’ perceptions of their
manager’s leadership style; the measure is generally regarded as
being unsuitable for direct assessment because it is vulnerable to
‘favourable self-biasing’ (Lowin & Craig, 1968; Otley & Pierce,
1995). Second, the Stodgill measure has been criticized for not
providing an indication as to how leadership style may manifest
‘on the job’ (Yukl, 1989). This is limiting in the context of the
present analysis, given the focus on examining practical
expressions of both leadership and budgetary style. Finally, it
is possible for the Stodgill measure to identify managers who
score highly on both initiating structure and consideration
dimensions (suggesting that they are both highly task-orien-
tated and highly employee-orientated). The present analysis
seeks a more straightforward instrument which will measure the
relationship between general leadership style (along a broad
continuum from task––employee-oriented) and commitment to
the budget.
446 D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456
point scale. A Cronbach Alpha reading of 0.71 was
obtained. Factor analysis revealed no sub-com-
ponents, while all items loaded above 0.40 (Van
der Stede, 2001).
Managerial performance
In an adaptation of the Mahoney et al. (1963,
1965) overall self-rating measure, a three-item
instrument was used to measure managers’ views
of their performance. This instrument includes
both an absolute (the Mahoney et al measure) and
two relative measures of performance (see ques-
tions 31–33 in the Appendix). These relative
measures were crucial, given the focus on under-
standing how well budgetary performance enables
the manager to achieve a sense of security in the
context of his or her place of employment. A
Cronbach Alpha score of 0.75 was obtained for
this three-item instrument.
While concern has been expressed in the liter-
ature regarding the use of self-report measures to
assess performance, a study by Venkatraman and
Ramamujam (1987) suggests that this concern is
somewhat misplaced. Managers’ self-rating of
performance tends to be less biased than what one
might expect. Moreover, Dunk (1993) points out
that self-rating of performance tends to be less
lenient than supervisory ratings (Heneman, 1974),
while earlier studies by Parker, Taylor, Barret, and
Martens (1959) and Kirchner (1965) provide fur-
ther support ‘‘for the utility of self-report mea-
sures’’ (Dunk, 1993, p. 580). Dunk (1993, p. 580)
goes on to explain that, ‘‘even though the rating of
subordinate has traditionally been the responsi-
bility of supervisors, superior managers may not
necessarily be well placed to rate the performance
of their subordinates.’’ This was precisely the case
at Infotain.
Pre-testing and piloting
Several instruments have been devised for the
?rst time in this study, and as such, particular
attention was paid to the pre-testing and piloting
of the questionnaire. The aim was to ensure that
the questions used to measure each variable were
unambiguous and captured the constructs of
interest. Standard practices were employed,
including the use of a panel of experts to assess the
questionnaire for understanding, style of question
(e.g. trade-o? approach), length of question, lay-
out and so on. The questionnaire was revised on
the basis of comments received. A second phase of
pre-testing and piloting was then undertaken, in
which the 26 managers who formed the interview
set during the earlier ?eld study were given a copy
of the questionnaire and asked to assess the sub-
stance, relevance and clarity of the proposed
questions. Final adjustments to the questionnaire
(mostly to the wording of questions) were made on
the basis of feedback received.
Analysis and results
Descriptive statistics
Table 2 reports means and standard deviations
and, where appropriate, both theoretical and ac-
tual ranges for the variables measured in the study.
Test of hypotheses
Empowerment and role ambiguity
Hypothesis 1 predicts that empowerment will be
related to the experience of role ambiguity. The
Pearson correlation coe?cient is 0.29 ðp < 0:01Þ,
suggesting a signi?cant positive relationship
between perceived empowerment and role ambi-
guity.
8
A crystallisation of attitudes
Hypothesis 2 seeks to re?ect the argument that
attitudes towards the budget are crystallised
as levels of subjective or experienced ambiguity
8
Of course, given the relationships depicted in Fig. 1, it is
important to test for a possible direct link between empower-
ment and budgetary commitment. While the Pearson correla-
tion coe?cient suggests a marginally signi?cant association
between these variables ðp < 0:05Þ, path analysis (Duncan,
1966; Kerlinger, 1986) demonstrates that role ambiguity com-
pletely moderates the relationship between empowerment and
budgetary commitment. From this, we conclude that empow-
erment is linked to budgetary commitment, not directly, but
through the experience role ambiguity.
D. Marginson, S. Ogden / Accounting, Organizations and Society 30 (2005) 435–456 447
increase. To test this, scores for role ambiguity are
divided at roughly the mid-point on the theoretical
range and compared against budgetary commit-
ment (also divided at the mid-point) by way of
Chi-squared analysis.
9
Results are signi?cant at
p < 0:001, indicating support for H2 (see Table 3);
those who experience high levels of role ambiguity
are, it seems, more likely to commit to meeting the
budget than those whose experiences of role
ambiguity are minimal.
10
An over-riding in?uence
Hypothesis 3 predicts that the crystallising of
commitment in situations of high role ambiguity
will suppress the potential for other factors to
in?uence budgeting behaviours in these circum-
stances. This argument is tested using a stepped
process of analysis. In the ?rst instance, it was
necessary to establish the extent to which each of
the variables––role ambiguity, leadership style, the
superior’s expectations of budgetary performance
and occupational socialisation––is signi?cantly
associated with budgetary commitment in its own
right. Table 4 presents Pearson correlation coe?-
cients. All are signi?cant to varying degrees, sug-
gesting that each variable is associated with
budgeting behaviour. Moreover, results for both
leadership style and the expectations of the supe-
rior may be claimed to be in the expected direc-
tion, given the results obtained by Hopwood
(1974).
11
The second stage of the analysis involved
examining the extent to which the experience of
role ambiguity over-rides the ability of each of the
other variables to in?uence budgetary commit-
ments. This was done, in the ?rst instance, through
two multiple regression analyses, which are dif-
ferentiated only by the role ambiguity data used.
For one, the measure of role ambiguity comprises
those scores equal to and in excess of the theo-
Table 2
Descriptive statistics
Variable Items Theoretical range Actual range Mean SD
Min. Max. Min. Max.
Budgetary
commit
8 8 40 11 38 25.87 7.03
Role
ambiguity
6 6 30 11 29 20.35 5.91
Empower-
ment
5 5 25 7 24 14.23 3.76
Leadership
style
5 5 25 8 21 15.67 4.02
Superior’s
expectations
4 4 20 8 18 11.09 3.58
Performance 3 3 15 5 15 10.04 2.56
9
For the purposes of this paper, scores below the theoretical
mean (i.e.