Management Accounting Research 30 (2016) 47–61

Contents lists available at ScienceDirect

Management Accounting Research

journal homepage: www.elsevier.com/locate/mar

Exploring differences in budget characteristics, roles and satisfaction:

A configurational approach
Samuel Sponem ∗ , Caroline Lambert
HEC Montreal, 3000, Chemin de la Côte-Sainte-Catherine, Montréal, Québec H3T 2A7, Canada

a r t i c l e i n f o a b s t r a c t

Article history: This study proposes a taxonomy of budget configurations. Combining a qualitative analysis and a cluster
Received 29 March 2012 analysis, we identify five patterns of budget design and budget use: the yardstick budget, the coercive
Received in revised form 26 October 2015 budget, the interactive budget, the loose budget and the indicative budget. Our taxonomy of budget con-
Accepted 26 November 2015
figurations allows us to describe complex arrangements that arise in practice. We observe that the budget
Available online 10 December 2015
is less criticized when the level of participation, the level of involvement of managers and the impor-
tance assigned to action plans during budget negotiations are high. This study refines the representation
Keywords:
of budgetary practices, opening the way to a better understanding of the practice of budgeting.
Budget
Configuration © 2015 Elsevier Ltd. All rights reserved.
Interactive control
Budget role
Budget satisfaction

1. Introduction
While budgets are frequently criticized for being time-
consuming, fostering individualistic logic and short-termism, and
providing poor value to users (Hope and Fraser, 2003; Østergren
and Stensaker, 2011; Henttu-Aho and Järvinen, 2013), they are still
being widely used in most organizations (Libby and Lindsay, 2010).
Extensive research in diverse theoretical perspectives (Covaleski
et al., 2003) focuses on budgets, but often obtains contradictory
results (Derfuss, 2009; Hartmann, 2000; Marginson and Ogden,
2005). These contradictory results might be explained by the fact
that these studies tend to focus on one budget dimension and
neglect the diversity of budget roles in organizations (Hansen and
Van Der Stede, 2004). The inconclusiveness concerning the posi-
tive and negative effects of budgets may also be linked to the fact
that, when investigating budgets, researchers tend to study their
characteristics separately, rather than understanding them as a set
of dimensions that are linked together (Shields and Shields, 1998).
Brownell and Hirst (1986) also point to the importance of interac-
tions between the various dimensions of budgets, by highlighting
that participation appears to exert a moderating effect on the neg-
ative consequences of using budgets to evaluate performance.
Few studies have tried to capture these multiple facets of bud-
geting (Abernethy and Brownell, 1999; Chapman and Kihn, 2009;
∗ Corresponding author.
E-mail addresses: samuel.sponem@hec.ca (S. Sponem), caroline.lambert@hec.ca
(C. Lambert).
Hansen and Van Der Stede, 2004; Hopwood, 1972; Merchant, 1981;
Van der Stede, 2001). These studies enlighten various relevant bud-
get styles, combining a selection of dimensions of design and/or
use. Most of them can be classified as typologies, meaning that
conceptual work guides the selection of relevant dimensions of
budgets to be studied (Bedford and Malmi, 2015). Nevertheless,
tied by this very search for conceptual coherence, these studies
have to make the implicit assumption that the dimensions they
study evolve simultaneously in the same direction. By focusing on
a combination of theoretically selected dimensions, they cannot
capture a complete view of the scope of budget design and use.
While taxonomies propose “simple” descriptions rather than
explanations, they are important for valid theory construction
(Bedford and Malmi, 2015). They provide a rich picture of practices
by capturing configurations, i.e., “multidimensional constella-
tions of conceptually distinct characteristics that commonly occur
together” (Meyer et al., 1993, p. 1175).
Several studies adopt configurational approaches to manage-
ment control systems (Gerdin and Greve, 2004; Gerdin, 2005). For
instance, Chenhall and Langfield-Smith (1998) examine how com-
binations of management techniques and management accounting
practices enhance the performance of organizations depending
on strategic priorities. The formality of attributes of management
accounting systems across life-cycle stages of firms has also been
studied and represented in a five-group taxonomy (Moores and
Yuen, 2001). Recently, Bedford and Malmi (2015) refine and extend
existing control frameworks and theory by identifying five major
control configurations used by managers.

http://dx.doi.org/10.1016/j.mar.2015.11.003
1044-5005/© 2015 Elsevier Ltd. All rights reserved.
48 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
In line with these studies, we aim to develop a taxonomy of bud-
get configurations and to describe complex arrangements that arise
in practice. More specifically, we empirically examine how the dif-
ferent styles of budget design and use combine in configurations
and how the latter are associated with budget satisfaction. Using
a configuration approach to budget practices, we address three
shortcomings of existing typologies on budgets. First, identifying
unexplored budget patterns constitutes a milestone for construct-
ing valid empirical tests of the impacts of budgets on individuals
and organizations. Second, empirically derived budget configura-
tions could extend existing frameworks by enabling us to examine
more than one variable at a time (Miller, 1996, p. 505) with-
out assuming that the different budget characteristics develop in
concert, as suggested in existing typologies. Third, by developing
a taxonomy, we can identify complementarity or substitutabil-
ity between the different characteristics of budgets (Grabner and
Moers, 2013) in order to clarify the impacts of budgets.
The study was carried out in two steps.1 First, we conducted
a series of meetings with CFOs and management accountants to
discuss their budget practices in a focus group. Analyzing the
focus group helped us to highlight major characteristics of bud-
get package design and use (Malmi and Brown, 2008). Second,
taking a configurational approach (Miller and Mintzberg, 1983;
Gerdin and Greve, 2004, 2008; Gerdin, 2005), we performed a clus-
ter analysis incorporating budget design and use characteristics.
This enabled us to propose an inductive typology of the budgetary
practices of 269 French firms and analyze how they relate to budget
roles and budget satisfaction. Our findings are complemented by
respondents’ informative answers to open-ended questions. This
taxonomy reveals five patterns of budget design and use: the yard-
stick budget, the coercive budget, the interactive budget, the loose
budget and the indicative budget. These patterns can be differen-
tiated according to three main characteristics: (1) the extent to
which senior management is involved in the budgetary process,
(2) whether or not the budget is used as a performance evalua-
tion tool, and (3) whether it is possible to revise the budget over
the year. In our subsequent analysis, we explored the differences
between these various budget patterns.
Several contributions of this study can be highlighted. By incor-
porating qualitative data into both the construction of our design
and use characteristics and by integrating a broader array of budget
design and use characteristics, we propose a more complex image
of how the different designs and uses of budgets are combined than
is currently recognized in the literature.
By comparing the roles, criticisms and satisfactions in the var-
ious budget configurations, we also contribute to the literature
on budgets by providing explanations for previous unexpected
or apparently contradictory results. Contrary to previous findings,
our results suggest that budgets with several concurrent roles do
not systematically attract criticisms. In particular, we contribute
to the debates in the Reliance on Accounting Performance Mea-
sures (RAPM) literature and suggest that interactions between the
various components of the budgeting process could explain the
less-than-conclusive findings of research into the effect of using
budgets to evaluate performance (Hartmann, 2000). By analyzing
the budget’s evaluation dimension in its context, as suggested by
Derfuss (2009), we show that the level of criticism relating to using
the budget to evaluate and reward managers is lower when evalu-
ation is coupled with other budget practices. We observe that the
1
In order to increase the internal validity of the research, we give a detailed
account of the construction and use of the variables included in our study. We
are grateful to the editor and the reviewers for giving us the unusual opportunity
(Grafton et al., 2011) to account for the various steps of this exploratory research,
resulting in a long paper.
budget tends to be criticized less when the level of operational
manager participation, senior management involvement and the
importance assigned to action plans during budget negotiations
are high.
The remainder of this paper is structured as follows. In the next
section, we discuss the theoretical and empirical research on bud-
get configurations and the major design and use characteristics of
the budget package. The third section describes the data collection
and the construction of clusters. The fourth section presents five
clusters of budget design and use, and their association with the
roles and criticisms of budgets. In the final section, we discuss our
main findings and conclude.
2. Literature review
2.1. Budget configurations
Recent research emphasizes the necessity to view management
control systems as packages in order to take into account the inter-
dependencies between these systems (Grabner and Moers, 2013;
Bedford and Malmi, 2015). The budget process is a mix of design
and use characteristics that varies across organizations; it is a set
of controls that include planning, measurement and evaluation
(Flamholtz, 1983), which could be considered a control package,
as drawn upon by Malmi and Brown (2008).
Previous studies have contributed to our understanding of bud-
get practices and identified various budget configurations. For
instance, Hopwood (1972) examines the use of accounting data for
performance evaluation and defines three evaluation styles making
distinctly different uses of the data (budget constrained style, profit
conscious style and non-accounting style). Merchant (1981) distin-
guishes administrative budgets (which emphasize explanations of
variances, formal involvement in budgeting and strong reactions
to expected budget over-runs) from non-administrative budgets.
Abernethy and Brownell (1999), based on Simons (1987b, 1990),
contrast the diagnostic use of budgets with the interactive use of
budgets (as a mean of questioning managers’ ongoing decisions
and interacting between top management and department/unit
managers). Van der Stede (2001) distinguishes tight from loose
budgetary control (Anthony, 1988), based on the emphasis on five
different dimensions, i.e., meeting the budget, budget revisions
during the year, the level of detail in the budget, tolerance for
deviations from interim budget targets, and the intensity of budget-
related communications. Chapman and Kihn (2009), based on Adler
and Borys (1996) identify an “enabling style” of budget based on
four design principles: repair; internal transparency; global trans-
parency; and flexibility (as opposed to a “coercive style” of budget).
Although they examine several budget dimensions, these stud-
ies still provide a limited view of the scope of budget design and
use. Hopwood’s (1972) typology mainly focuses on the use of bud-
gets for evaluation purposes. Merchant (1981) neglects this aspect
but emphasizes managers’ involvement in budgeting. Similarly,
Abernethy and Brownell (1999) mainly explore the use of bud-
geting systems and assume that budget design is constant across
organizations. While these studies help to better understand the
budgeting process, they remain limited to certain facets of the
budget package. Moreover, these typologies were developed on dif-
ferent types of samples (some focusing on small and others on large
firms), across an extended period of time, and with incompatible
focuses. Combining their results would be methodologically fragile.
Studies by Van der Stede (2001) and Chapman and Kihn (2009)
provide a more complete description of the budgeting process,
taking into account both the design characteristics and use char-
acteristics of budgets. Van der Stede (2001) draws upon Anthony’s
(1965) classical typology and distinguishes tight budgetary con-
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
trol from loose budgetary control. Chapman and Kihn (2009) build
on work by Adler and Borys (1996) and distinguish enabling
budgets from coercive budgets. However, in both of these cate-
gorizations, Van der Stede (2001) and Chapman and Kihn (2009)
assume that all the dimensions vary simultaneously, in a given
way. For instance, Van der Stede (2001) does not consider the pos-
sibility of a low emphasis on meeting budgets combined with a
high intensity of budget-related communication, and assumes that
the various design characteristics develop concurrently. Chong and
Mahama (2014) suggest that “a budget that is used diagnostically
can simultaneously be used more or less interactively” (p. 213).
They investigate the impact of interactive and diagnostic budgets
on team effectiveness and perceived collective efficacy. They show
that the extent of the interactive use of budgets has direct positive
effects on team effectiveness and that these effects are partially
mediated by perceived collective efficacy. However, they take into
account these various budget dimensions independently, without
looking at the interdependencies between budget characteristics.
Overall, existing typologies only provide a partial understand-
ing of the scope of budget practices. Even when several budget
characteristics are studied, the methods do not account for the mul-
tidimensional nature of budgets or the interdependencies between
the various budget characteristics. Thus, these studies do not exam-
ine the extent to which the various budget characteristics are
substitutable or complementary (Grabner and Moers, 2013). Over-
all, these studies fail to fully capture the multidimensionality of
budget design and use.
One way to expand our knowledge of budgeting practices is to
explore the actual budget design and use configurations formed in
practice. In this paper, we consider the budget process a control
package (Malmi and Brown, 2008) and study the budget process
as a set of dimensions, rather than separately studying its individ-
ual characteristics – motivation, participation, etc. – as suggested
by Shields and Shields (1998), and connect it to satisfaction and
roles (Libby and Lindsay, 2010). An empirical taxonomy of budget
practices can provide more complete descriptions of how different
characteristics of budgeting practice tend to combine and reveal
alternative budgeting patterns not captured or explained in exist-
ing typologies (Bedford and Malmi, 2015).
2.2. Budget design and use characteristics
The budget characteristics retained encapsulate a relatively
broad conceptualization of the budget process. Most of the char-
acteristics have already been established in the literature for a
long time (Chapman and Kihn, 2009; Flamholtz, 1983; Van der
Stede, 2001). In order to design a survey tool that includes the
constructs and measures valid in our context and time (Modell,
2009), we completed and refined our conceptualization by includ-
ing clarifications that emerged from a focus group. We conducted a
yearlong focus group with management accountants on budgetary
practices.2 These qualitative data were used in the initial phase of
the research, with the aim of “facilitating” the study (Hurmerinta-
2
Eleven management accountants attended the focus group once a month from
firms of various sizes. They presented their budgeting processes and the limits of
those processes with the aim of sharing knowledge about one another’s problems
and finding the best practices. These exchanges were both formal, as all manage-
ment accountants prepared their presentation describing their budgeting process
from the angle of their own observations but also the viewpoints of their opera-
tional superior and other management accountants, and informal, in the form of
very informative discussions and subsequent question-and-answer sessions. We
complemented this observation work by carrying out six interviews with manage-
ment accountants and CFOs working in the firms involved in the focus group. Our
objective was twofold: to learn the exact vocabulary used by management accoun-
tants and CFOs and to ensure that the design and use characteristics we selected to
describe the budgeting process were relevant.
49
Peltomäki and Nummela, 2006) and complementing the statistical
analysis. Using quotes from the focus group helped us to reveal
different “slices” of the budgeting practices we wanted to better
capture (De Loo and Lowe, 2011).
We have identified 11 major budget design and use charac-
teristics and grouped them into the three sequential stages that
constitute the budget process: the drafting stage (before action),
the steering stage (during action) and the review and evaluation
stage (after action) (Bouquin, 2010). Three other, transversal design
characteristics are independent of these three stages.
The budget constructs included in the empirical analysis are
detailed in Table 1.
Analyzing the focus group data led us to refine six budget design
and use characteristics.
In the drafting phase, in addition to the commonly used dimen-
sions of “Participation of managers” and “Difficulty to meet budget
targets,” our focus group revealed that firms combine budgeting
with strategy to varying degrees in budget negotiations. Literature
tends to take for granted a strong link between strategy and budget-
ing, defining the budget as a “quantitative expression of a proposed
plan of action by management for a future time period” (Bhimani
et al., 2008, p. 467). Conversely, in the focus group, budgets ranged
from straightforward “copy-paste” of the purely financial com-
mitments made by the CEO to shareholders, to an opportunity to
discuss and shape strategy. We define “Negotiation type” as the
importance granted to action plans and strategic orientations dur-
ing budget negotiations.
In the literature, budget rigidity is considered a central char-
acteristic of the steering phase, and budget revision divides
scholars (Merchant, 1981; Van der Stede, 2001; Chapman and Kihn,
2009) and practitioners (Anthony, 1988). Results obtained through
our focus group suggest that budget rigidity generates two dis-
tinct design characteristics. We distinguish the “Budget revision”
(defined as the possibility to change the initially set budget) and
“Budget reforecast” (defined as the existence of a reforecast during
the year), which could be combined in various ways, as some firms
in our focus groups opted to combine the advantages of a rigidly
set budget with frequent reforecasts.
In the review and evaluation stage, differences can be expected
between the importance attached to the budget for performance
evaluation and the importance attached to the budget for deter-
mining rewards. The distinction between evaluation and reward
appears rarely in the budgeting literature (with the noticeable
exception of van Veen-Dirks, 2010), but the focus group insisted
on the difference between the traditional and well-known use of
budgets to evaluate performance and their use to determine the
level of pay and rewards.
The involvement of senior managers is frequently studied in
budget practices. Two additional transversal design characteris-
tics emerged from the focus group: the level of budget detail and
the level of budget formalization. While the level of the informa-
tion detail in budgets has been previously discussed (Merchant,
1981; Van der Stede, 2001), it has been mostly to inform its report-
ing dimension (in the steering phase). In the focus group, it also
seems clearly associated with the level of autonomy granted to
operational managers to set and then to meet their budget targets,
all through the budget process. A high level of detail reduces the
control managers’ activity to mere execution of their predefined
engagements.
The level of formalization of the structure has been largely
brought to light by the Aston school (see, for instance, Pugh et al.,
1968) but is rarely taken into account in budget research (King et al.,
2010). The focus group has confirmed the importance of formaliza-
tion and associated it with the possibility of adapting the budget to
local contexts.
50 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61

Table 1
Budget constructs: definitions and examples.

Stages Design and use Definition Examples of focus group quotes that led us to refine and add
characteristics design and use characteristics

Drafting stage Participation of Level to which managers are involved with, and have
(before managers influence on, the determination of their budget Shields and
action) Shields, 1998, p. 49; Derfuss, 2009)
Negotiation type Importance granted to action plans and strategic We argue too much about figures and not about action plans.
orientations during budget negotiations It’s difficult to match strategic objectives with budgetary
objectives [. . .] There’s a gulf between the two. At the end of the
day, we focus on variance analysis rather than on the action
plans. (Director of consolidation, Company C)
Budgets should translate the firm’s strategic orientation. In
reality, this is only partially the case, and it’s becoming less and
less so since recently, the budget has been tending to become a
purely financial frame for expected performance targets.
(Director of Management Control Division, Company A)
The budget is used to generate ideas. Commercial action plans
and tactics emerge during the budget drafting. This is when
ideas emerge. (Director of Management Control, Company F)
Difficulty of meeting Level of budget target achievability (Merchant and
budget targets Manzoni, 1989).

Steering stage Variance analysis Level of emphasis put on identifying variances between
(during what is forecasted and what is actually achieved, and
action) assigning responsibility for those variances to the
appropriate managers (Brownell, 1983; Emsley, 2001;
Hirst and Lowy, 1990 Webb, 2002)
Budget revisions Possibility to change the initially set budget At group level, there are three reforecasts a year. At branch
level, there’s a monthly reforecast, which is run by a small team
(management control and consolidation departments). But
operational managers are evaluated solely on their
achievement of targets as set in the initial budget. (Firm B).
The budget itself is never changed once set. However, there are
reforecasts when the strategic plan is elaborated or at the
request of the CEO. (Firm G)
Budget reforecasts Existence of reforecasts throughout the year

Review and Budget-based Extent to which superiors rely on and emphasize the
evaluation evaluation performance criteria, which are quantified in accounting
stage (after and financial terms and specified as budget targets (Briers
action) and Hirst, 1990; Hartmann, 2000; Otley and Fakiolas,
2000; Harrison 1993)
Budget-based rewards Importance attached to the budget for determining To give the budget some power, we link it to a system of
rewards (van Veen-Dirks, 2010) penalties and rewards. (Director of Management Control
branch, Firm I)

Transversal Involvement of senior Level of involvement of senior management in the budget

characteris- management in budget process (Abernethy and Brownell, 1999; Simons, 1990)
tics process
Amount of information Amount of information in the budget If we impose too much specific detail, we soon get bogged down
in the budget in endless discussions about figures, and we lose sight of the
important matters. (Director of Management Control, Firm J)
The proliferation of budget lines should help to keep the budget
meaningful. In our firm, we have a very fine level of detail, but
important variances appear since people aren’t able to “put
figures in the boxes.” However in theory, this fine level of detail
should help the operational managers to build their
budget. . .(Director of Management Control, Firm I)
Extent to which budget Extent to which formalized rules and procedures are The budget is a negotiation. When the players get to understand
is formalized considered important for formulating budgets (Ueno and the rules of the game, they use them. So then we have to change
Sekaran, 1992) the rules of the game. The main problem lies in maintaining the
comparison with n − 1. (Director of branch Management
Control, Firm I)

To summarize, our focus group and literature analysis have 3. Methodology

drawn out 11 budget design and use characteristics: manager par-
ticipation, negotiation type, difficulty of meeting targets, variance
analysis, revisions, reforecasts, budget-based evaluation, budget-
based rewards, senior management involvement, level of detail,
and degree of formalization.
3.1. Data collection
The results presented in this study are mainly survey-based. Our
survey design was pre-tested with four academics and four pro-
fessionals to verify the clarity, understandability and ambiguity of
the questions (Dillman et al., 2008), and adjusted in response to
feedback from our testers.
Then, we sent the questionnaires to CFOs, vice presidents
for finance and management accountants who were members of
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
Table 2
Industry affiliation and organizational size.a
Industry affiliation Organizational size
Manufacturing 40% No. of employees
Wholesale and retail trade 17% 30–49
Business services 15% 50–250
Financial and insurance activities 13% 250–500
Other service activities 10% 500–2,000
Other industries 5% >2,000
Max.
Mean
a
The size reported here is the size of the entity, division or subsidiary studied,
not the overall size of a firm operating in several sectors.
the French professional association of management accountants
(DFCG). We chose CFOs, vice presidents for finance and man-
agement accountants as informants since our exploratory study
showed that they are knowledgeable about their firms’ budget
practices. They have direct, frequent access to senior management
to discuss issues relating to budget practices, roles and criticisms,
and they hold knowledge about strategic issues and other items
covered in the survey. Furthermore, financial managers are usually
much more realistic about the role of a budget tool than operational
managers (Shields and Young, 1993). The target firms were either
independent companies or SBUs with more than 30 employees. We
chose these criteria to ensure that the target group of organizations
was large enough to have formal budgeting systems and that the
managers we contacted would have experience in using budgets
(Libby and Lindsay, 2010).
The survey questionnaire was sent by email to 1451 potential
respondents of different firms and SBUs. However, 190 of the email
addresses used were no longer valid. After two reminders, 314 per-
sons returned their questionnaires within three months (response
rate: 24.9%). Questionnaires with missing data on more than 10%
of the questions necessary for this research were excluded from
our sample, and ultimately 269 questionnaires could be used in
the analysis (useable response rate: 21.3%). This response rate is in
line with other behavioral management accounting research (e.g.,
Chenhall et al., 2011; De Baerdemaeker and Bruggeman, in press).
Of the replies we received, 67% came from CFOs, 28% from
management accountants or their superiors, and 5% from vice pres-
idents for finance, and 75% of respondents work in firms belonging
to corporate groups. Apart from an overrepresentation of firms
operating in the financial sector, the representation of different sec-
tors in our final sample is close to the proportions of French firms
hiring more than 10 workers, as defined by the French national
statistics institute INSEE3 (Table 2).
Early respondents (those who returned the completed question-
naire before the first reminder) and late respondents (those who
returned it after the second reminder) were compared in terms of
sample characteristics (sector, size and sales), and in order to detect
any difference in the mean score for each variable. No significant
differences (p > 0.05) were found concerning firm characteristics. A
comparison of the means for the variables found little difference
between early and late respondents. The t-value was significant
for two variables only (evaluation: t = 2.38, p < 0.05; and formaliza-
tion: t = 1.834, p < 0.1), but we do not believe this poses a serious
problem considering its isolated effect. While this suggests non-
response bias is not a serious concern, it should be acknowledged
that accountants in firms with relatively sophisticated accounting
systems may have been more inclined to respond than those in
firms with under-developed accounting systems.
3
Source: Insee, Répertoire des Entreprises et des Etablissement —Sirene.
51
3.2. Variable measurements
All measures are based on existing instruments where possi-
ble. Following Bisbe et al. (2007), each construct is reflected or
8% manifested by a series of indicators. Forty-one items were initially
32% designed to measure budget practices, and most were measured
20% on a five-point Likert scale (1 = disagree, 5 = agree). To improve the
28%
quality of the constructs, we first performed an exploratory fac-
12%
45,000 tor analysis on each construct, to reduce the number of items and
1,567 retain only the most relevant (Hinkin, 1998). Items with a loading
of 0.50 or greater were considered “practically significant” (Hair
et al., 2010, p. 117), and items with a loading of <0.50 on the first
factor were removed from the analysis. After purification, the 11
constructs were operationalized through 31 items. A confirmatory
factor analysis was then carried out to confirm the convergent and
discriminant validity of the constructs. Appendix A shows the scales
used to measure the budget practice constructs. Table 3 presents
descriptive statistics and the reliability and validity analysis of the
constructs.
The reliability of the constructs was assessed by measuring the
Cronbach alpha (Hair et al., 2010). The Cronbach alpha of our mea-
sures is acceptable for an exploratory study. The convergent validity
was assessed by measuring the average variance extracted (AVE)
and should be greater than 0.50 (Hair et al., 2010). This is true for
all but two constructs (detail: AVE = 0.38; formalization = 0.406), for
which no established measure existed in the literature. Therefore,
most measures indicate good convergent validity. The discriminant
validity was assessed by considering the degree to which a vari-
able measures a concept that is uniquely defined and is not highly
correlated with other variables included in the model (Fornell and
Larcker, 1981). Overall, the reliability and validity of the constructs
was judged acceptable for an exploratory study.
The Pearson correlation coefficients presented in Table 4
indicate several associations between the use and design charac-
teristics of budgets.
Participation, negotiation and involvement are significantly
associated with most of the other design characteristics. Revisions
and reforecasts show close to zero, statistically insignificant corre-
lations with most of the other use and design characteristics.
To measure participation in budgeting, the survey contained
four questions that used a five-point Likert scale ranging from dis-
agree (1) to agree (5). These items draw on questions proposed
by Milani (1975) and Simons (1987a) and are aimed at measur-
ing operational managers’ influence on budgets. Examination of
the Cronbach alpha (0.8), the AVE (0.528) and its correlations with
other constructs validates the reliability and validity of this con-
struct.
The measurement of the importance assigned to action plans
during budget negotiation is made up of four questions adapted
from Ittner and Larcker (1997). This measurement translates the
importance of action plans in drafting the budget. The reliability
(Cronbach alpha = 0.83) and validity (AVE = 0.603) of this construct
are satisfactory.
The measurement of the difficulty of meeting targets derives
from the findings of Merchant and Manzoni (1989) and the scale
used by Van der Stede (2000) to measure slack. The first item is a
fully anchored question asking whether the budget target is (1) very
easy to attain; (2) attainable with reasonable effort; (3) attainable
with considerable effort; (4) practically unattainable; or (5) impos-
sible to attain. The second item deals with the likelihood of a budget
manager meeting the budget targets set. The Cronbach alpha (0.62)
and the AVE (0.534) indicate that this construct is acceptable.
The measurement of variance analysis relies on the scales used
by Brownell (1983) and Van der Stede (2001) to evaluate the
importance senior management assigns to budget variance analy-
sis. Examination of the Cronbach alpha (0.75), the AVE (0.522) and
52 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61

Table 3
Descriptive statistics of the constructs (n = 269).

No. of items Minimum Maximum Mean Standard deviation Cronbach alpha AVE

1-Participation 4 1 5 3.2 0.873 0.8 0.528

2-Negotiation 4 1 5 3.208 1.059 0.83 0.603
3-Difficulty 2 1 4.5 2.185 0.711 0.62 0.534
4-Variance 3 1 5 3.887 0.861 0.75 0.522
5-Reforecasts 2 1 5 3.323 1.04 0.62 0.498
6-Revisions 1 1 5 2.959 1.556 N/A N/A
7-Evaluation 3 1 5 3.203 0.983 0.81 0.599
8-Reward 3 1 5 2.713 1.195 0.88 0.736
9-Involvement 4 1 5 3.565 0.888 0.81 0.542
10-Detail 3 1 5 3.506 0.938 0.63 0.38
11-Formalization 2 1 5 3.695 0.975 0.54 0.406

its correlations with other constructs confirms the reliability and
validity of this construct.
The “reforecasts” and “revisions” constructs derive from analy-
sis of a construct initially designed to analyze budget flexibility. An
exploratory factor analysis performed on this construct reveals two
underlying factor axes. The measurement of reforecasts is made
up of two items drawn from Van der Stede (2001) and Merchant
(1981). Its purpose is to gauge the frequency of reforecasting. The
Cronbach alpha (0.62) and the AVE (0.498) indicate that this mea-
surement is acceptable. The concept of revision is measured by
a single item taken from Van der Stede (2001): “budget targets
cannot be changed over the year.”
An exploratory factor analysis led us to dissociate use of the
budgets to evaluate performance from use to reward. Budget-based
evaluation is measured by three items adapted from Van der Stede
(2001). The Cronbach alpha (0.81) and the AVE (0.599) validate this
measurement. The reliability (Cronbach alpha: 0.88) and validity
(AVE: 0.736) indices for the “budget-based reward” construct are
also acceptable.
The measurement of senior management involvement is made
up of four items taken from Abernethy and Brownell (1999). The
Cronbach alpha (0.81) and the AVE (0.542) validate this construct.
Two additional constructs emerged from our qualitative survey
as highly significant but are barely mentioned in the literature: the
level of detail in the budget (Cronbach alpha: 0.63) and its degree
of formalization (Cronbach alpha: 0.54). We operationalized these
constructs using three items adapted from Van der Stede (2001)
and two items adapted from Simons (1987a).
For a better understanding of the various budgeting styles, we
also measured the roles that are attributed to the budget. Budgets
can be used to motivate managers to make plans, inform managers
about what they are expected to do and accomplish, elicit manager
commitment, coordinate the separate activities of the organiza-
tion, and provide a standard by which to judge actual performance
(Anthony, 1988; Barrett and Fraser, 1977). Budgets can also be
used for strategy formation and implementation (Hansen and Van
Der Stede, 2004), to communicate information to shareholders and
external interested parties by reference to the previous and current
budget in the annual report (Ekholm and Wallin, 2000), to allocate
resources (Bhimani et al., 2008), to authorize spending, and to man-
age risk (Collier and Berry, 2002). Budget roles were operationalized
through 12 items drawn from Lyne (1988), Bunce et al. (1995),
Ekholm and Wallin (2000), and Hansen and Van Der Stede (2004).
Factor analyses performed on budget roles identify four major roles
for budgets (see Table 5): a managerial role (evaluating managers,
incentivizing them, defining responsibilities and contractualizing
commitment); a strategic role (implementing strategy, forecasting
financial needs, and coordinating business activities); an adminis-
trative role (authorizing spending and allocating resources); and a
role concerning communication with external stakeholders.
We also evaluated the budget satisfaction of respondents. Ever
since Argyris’ seminal study (1952), budgets have come under
heavy criticism. Budgets are said to be too rigid and often unre-
alistic and are criticized for being time-consuming and expensive,
providing poor value to users, failing to focus on shareholder
value, concentrating on cost reduction instead of value creation,
encouraging “gaming” and uncooperative behaviors, reinforcing
departmental barriers rather than encouraging knowledge shar-
ing, constraining responsiveness and flexibility, reinforcing the
command and control culture, and adding little value, especially
given the time required to prepare them (Hope and Fraser, 2003).
Budget satisfaction was operationalized through 11 items adapted
from Ekholm and Wallin (2000) and Hope and Fraser (2003).
Rather surprisingly, the descriptive statistics show that the over-
all level of criticism is fairly low. On the scale applied, ranging
from 1 (disagree) to 5 (agree), most answers range between 2.1
and 3.16. Factor analyses highlight four major sources of criti-
cism (see Table 6): a criticism of the adverse effects of budgeting
(opportunistic or conservative behaviors and a negative impact on
cooperation); the constraint that arises from the budget when it
is inappropriate to the environment (the budget is no longer suit-
able in an environment that has become uncertain, because it is

Table 4
Pearson correlation coefficients (n = 269).

1 2 3 4 5 6 7 8 9 10 11

1-Participation 1
2-Negotiation .411** 1
3-Difficulty −.221** −.223** 1
4-Variance .285** .298** −.259** 1
5-Reforecasts .082 .081 −.001 .133* 1
6-Revisions −.023 .019 −.060 −.003 .130* 1
7-Evaluation .163** .175** −.089 .420** .064 −.135* 1
8-Reward .040 .073 .033 .239** .106 −.171** .536** 1
9-Involvement .517** .489** −.241** .570** .099 −.017 .410** .263** 1
10-Detail .267** .298** −.106 .334** .115 .014 .169** .129* .388** 1
11-Formalization .259** .208** −.243** .349** .075 −.072 .262** .199** .420** .309** 1
*
p-value significant at 0.05.
**
p-value significant at <0.01.
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61 53

Table 5
Factor analysis of roles. a

Managerial budget Strategic budget Administrative budget Shareholder budget

Evaluating manager performance .895 −.057 .073 .059

Incentivizing operational managers .848 .084 −.026 .168
Defining responsibilities and contractualizing commitment .693 .279 .294 −.027
Communicating between various levels of the reporting line .568 .346 .041 −.003
Deploying strategy .099 .625 .004 .032
Forecasting financial needs −.022 .611 .097 .048
Managing risks .192 .575 .263 .327
Coordinating various operations .300 .567 .296 .008
Steering the firm’s different business activities .464 .544 −.034 .033
Authorizing spending .065 −.014 .909 .052
Allocating resources .085 .420 .682 −.055
Communicating with external stakeholders (shareholders, creditors). .100 .121 −.008 .970
a
In your organization, to what extent is the budget used for the following purposes (scale: 1 = not used for this purpose to 5 = extensively used for this purpose).

Table 6
Factor analysis of criticisms. a

Perverse effects Constraints Ritual Short term

The budget elicits opportunistic behaviors .751 .020 .287 .142

The budget elicits conservative behaviors .744 .111 .175 .190
The budget deters cooperation .599 .334 .171 .041
The budget is no longer suitable in an environment that has become too uncertain .062 .801 .292 .101
The budget introduces rigidity into the organization .575 .587 -.022 .107
The budget imposes a culture of control rather than a culture of engagement .508 .512 .058 .188
The budget impedes innovation .398 .493 .309 .029
The budget is primarily a ritual .191 .179 .882 .102
The budget takes too much time for little added value .390 .254 .557 .241
The budget translates the predominance of short-term profitability to the detriment of long-term value creation .280 -.042 .136 .793
The budget covers an annual timeframe that is no longer suited to the firm’s business cycle .010 .509 .114 .727
a
Please indicate the degree to which you agree with the following statements as they apply to your organization (scale: 1 = strongly disagree to 5 = strongly agree).

inflexible, imposes a culture of control and impedes innovation);
a criticism of its ritual aspect (the budget is a ritual that takes up
too much time); and a criticism of its short-termist focus (the bud-
get reflects the predominance of short-term thinking and covers an
unsuitable timeframe).
In addition to the above, to fine-tune our understanding of these
criticisms we asked three supplementary questions: “I am satisfied
with the budgeting process (disagree/agree)”; “Managers are satis-
fied with the budgeting process (disagree/agree)”; and “The budget
is a major management instrument (disagree/agree)”.
ance between clusters.4 We examined how the distance between
individuals within clusters increases as the number of clusters
decreases to determine the relevant number of clusters. We also
examined the growth of inertia when the number of clusters
decreases (Husson et al., 2010; Lebart et al., 2006). Both methods
suggested retaining a five-cluster classification.5
Second, an iterative method was used to refine outcomes in
order to minimize the potential impact of outliers (Husson et al.,
2010). This consolidation was carried out with the SPAD software
by performing a clustering on the mobile centers of the clusters
resulting from the agglomerative hierarchical clustering (Lebart

3.3. Data analysis

To capture the multidimensionality of budget design and use,

cluster analysis appeared to be a useful method (Ketchen and
Shook, 1996). This exploratory method takes many variables and
groups them such that the statistical variance between character-
istics in the same group is minimized, while inter-group variance is
maximized (Ketchen and Shook, 1996). The statistical outcome of
this procedure is “the best structuring of the cases that can be done
given the variables that are used and the specifications that the
researcher selected for performing the analysis” (Meyers, Gamst, &
Guarino, 2013, p. 818). This procedure allowed us to categorize the
269 independent companies and SBUs according to their scores on
the 11 budget design characteristics.
We used a two-stage clustering procedure to take advan-
tage of both hierarchical and non-hierarchical clustering methods
(Ketchen and Shook, 1996, p. 452).
First, the cluster analysis was performed using an agglomer-
ative hierarchical technique based on the Ward criterion, which
minimizes the total within-cluster variance and maximizes vari-
4
The clustering was performed on the outcomes of a principle component anal-
ysis to reduce noise and make the clustering more robust (Husson et al., 2010). An
initial principle component analysis was carried out on the factor scores for the dif-
ferent constructs. Factor scores were calculated individually for each construct using
the co-variance matrix and replacing any missing values with the mean score. Two
other principle component analyses were carried out on the scores calculated from
the mean of the items for each dimension and on a multiple factor analysis (Lê et al.,
2008). Preliminary analysis shows that the two types of scores are very highly cor-
related: correlations between the mean-based and factor analysis scores are higher
than 0.99 for all constructs except for the “difficulty in meeting targets” construct,
for which the correlation exceeds 0.97. The outcomes obtained with the mean-based
and multiple factor analysis scores were significantly equivalent to those obtained
using the factor scores.
5
To measure the impact of the choice of method on the number of clusters used,
we carried out ascending hierarchical cluster analyses using various numbers of
factor axes. Regardless of the method chosen, relative stability emerges in cluster
outcomes. The most common optimal split – regardless of the method used – is into
five clusters. The results presented here are generated from weighted data. In line
with recommendations made in the literature, we carried out the same calculations
using non-weighted data (Ketchen and Shook, 1996, p. 444). The results we obtained
were significantly equivalent.
54 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
et al., 2006).6 Four iterations were performed to obtain the optimal
solution.
Two types of analyses were then performed to systematically
explore the differences between the clusters. First, significance
tests were performed on variables used for the cluster construction,
in order to identify clusters for which the means for a dimension
under the t-test (p < 0.05; p < 0.01) were significantly different from
the overall mean. Second, significance tests were performed on
variables not used for the cluster construction: budget roles and
budget satisfaction. For each cluster, a student’s t-test was used to
compare the category mean with the grand mean (Lê et al., 2008)
and identify clusters for which the means for budget roles and bud-
get satisfaction were significantly different from the grand mean
under the t-test (p < 0.05). Roles and satisfaction in each cluster
were thus systematically compared.
Additionally, the cluster analysis and its interpretation were
sent as a “benchmark” to our survey respondents. In this stage
of the study, qualitative data helped us to interpret our results
by providing an illustrative knowledge of our quantitative results
(Hurmerinta-Peltomäki and Nummela, 2006). Feedback and spon-
taneous comments from practitioners corroborated the relevance
of our typology.
4. Results
This section presents and explains the results of the cluster anal-
ysis. We label the clusters by comparing the characteristics of each
cluster with ex ante theory. The interpretation of budget character-
istics, roles and satisfaction are illustrated by the answers to open
questions given by survey respondents in each cluster.
4.1. Cluster 1: yardstick budget
Cluster 1 is characterized by extensive manager involvement
(0.474**)7 . In this cluster, action plans are taken into consideration
during budget negotiations (0.0114).
We mostly run long-range activities. Before beginning the bud-
geting process, we update our strategy according to various
scenarios. During this phase, we arrive at a position that marks
a consensus on the objectives and resources to be implemented.
(Respondent, Cluster 1).
Drawing up a budget is essential in order to plan for potentially
difficult times, and take the right decisions in order to adapt the
structure of the firm. (Respondent, Cluster 1).
The budget is a chance to thoroughly examine the various parts
of the firm and understand how they overlap. It enables us to
ask the right questions. (Respondent, Cluster 1).
The budget is not developed “top-down”, and operational man-
agers actively participate in preparation of the budget (0.375**),
but rewards are not budget-related (−.0126).
Bonuses depend on the overall performance of the firm. They
aren’t individually calculated on the basis of achievement of
budget targets. (Respondent, Cluster 1).
6
To further validate the stability of the clustering results with respect to sampling
variability, we performed a probabilistic clustering (using the procedure described
above). This was done on 500 bootstrap subsets that were resampled from the orig-
inal sample (Hennig, 2007). Each subset consisted of a random sample of 50% of
the initial population. Appendix B presents the mean calculated on the 500 clus-
ter analyses performed on the subsets for each construct. The percentage of correct
classification ranges between 73% and 84% (see Appendix B).
7
Mean of the group. Variables significantly different from the grand mean under
the t-test (p < 0.05/(p < 0.01) are indicated by (*/**).
By involving operational managers in the budget target-setting,
top management hopes to obtain the most accurate forecasts for
the business.
Surprisingly, in our firm, we have to calm down the managers’
excessive optimism during budget drafting. (Respondent, Clus-
ter 1).
The budget targets constitute a yardstick decided on after long
discussions, and this cluster has fewer revisions (−0.756**) and
reforecasts (−1.521**) during the year than other clusters.
Only one-off actions such as large-scale reporting plans, for
instance, can be added to the original budget. But this decision
must be validated by various board committees in the Group.
(Respondent, Cluster 1).
Budget variances are subject to more scrutiny than in other
clusters on average (0.296*). The budget is used more to evaluate
performance (0.288*) than in other clusters on average, but is less
linked to the reward system (−0.126).
Each operational manager receives a detailed report of his indi-
cators. (Respondent, Cluster 1).
In this cluster, the budget is used to define responsibilities, con-
tractualize commitments and communicate between various levels
of the reporting line more than in other clusters (0.275*). It is
still presented as a major instrument (0.190) which both finan-
cial (0.172) and operational managers mostly consider satisfactory
(0.105). The budget is also a central management tool, as illustrated
by open-ended responses made by Cluster 1 respondents.
That’s a way of making all of us work in the same direction.
(Respondent, Cluster 1).
Rather than a very precise and sophisticated exercise, the bud-
get is an opportunity taken by senior management to review the
various sectors of the firm (Respondent, Cluster 1).
Even though it constitutes a yardstick (i.e., there are few revisions
and reforecasts), respondents in this cluster do not feel that the
budget makes the organization inflexible or impedes innovation.
However, they do point out that it can lead to both opportunistic
and conservative behaviors, and sometimes appears to be a ritual
(0.200). In relation to the evaluation dimension, the characteristics
of cluster 1 appear to be in line with the profit-conscious style of
Hopwood (1972). The budget is used to evaluate managers on their
ability to provide accurate forecasts of their activity, and to develop
action plans to increase the general effectiveness of their unit’s
operations in relation to the long-term purposes of the organiza-
tion. It appears as a managerial tool that bridges the gap between
present and future, both in its design process and use. In this cluster,
the budget can be seen as a yardstick .
4.2. Cluster 2: coercive budget
Cluster 2 shows a management-by-numbers style, in which
the budget was used more as an evaluation (0.609**) and reward
(0.806**) tool than in other clusters:
Individual objectives for managers are set according to budget
data. (Respondent, Cluster 2).
Bonuses are calculated on targets set when the budget gets
validated. Incentives are calculated based upon budget-based
indicators such as sales, quality indicators, productivity indica-
tors. (Respondent, Cluster 2).
However, this budget is perceived as less participative than in
the other clusters (−0.292**).
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
Even though the formal process begins by gathering budgets
bottom-up, final decisions, as well as constraints and limits, are
imposed top-down (Respondent, Cluster 2).
The budget is no longer a commitment by the people (i.e., the
operational managers) who make it. It has turned into objec-
tives imposed by top management on the operational managers.
(Respondent, Cluster 2).
Action plans are not taken into account during budget negotia-
tion (−0.210*), the budget has little detail (−0.254*), and is revised
only on rare occasions (−0.230*). Interestingly, in our observations,
on average, firms in this cluster perform more frequent reforecasts
than in other clusters (0.467**).
We do three annual reforecasts and twelve three-month refore-
casts. Budgeting is a constant forecasting exercise, for adapting
initial data to market developments (changes in exchange rates,
cost of materials etc.). But the final objectives, in particular the
profitability ratios, are not revised. (Respondent, Cluster 2).
We have budget reforecasts all year long so that monthly results
forecasts remain as accurate as possible. (Respondent, Cluster
2).
Compared to other clusters, this budget can be seen as coercive.
It is used more to contractualize commitments, and to evaluate and
incentivize managers (0.369**). Overtly contemptuous of the con-
trollability principle, this budgeting practice appears all the more
unfair since it is a major, if not the only, tool of management. It is
highly criticized, and seen as a ritual (0.256**) that deters coopera-
tion, leading to opportunistic and conservative behaviors (0.225*).
The budget is far too rigid when it comes to taking into account
changes in the underlying assumptions. We should change it
to a rolling budget, or at least a three-month budget. Deadlines
are far too short for the operational managers to get involved.
(Respondent, Cluster 2).
In a rapidly changing environment, the budget is energy-
consuming. (Respondent, Cluster 2).
The budget is disconnected from strategy. It’s neither responsive
nor interactive. (Respondent, Cluster 2).
The budget would be meaningful if translated into short-run
action plans. (Respondent, Cluster 2).
The budget is not perceived as a tool to deploy strategy
(−0.426**). It introduces inflexibility into the organization and is
not suited to dealing with changes in the environment (0.192*). It
is seen as a ritual (0.256**) that has adverse effects (0.225*).
Performance indicators and value creation indicators should
be added to the budget. That would help reduce conservative
behaviors and foster innovation. (Respondent, Cluster 2).
This budget can be qualified as coercive (Ahrens and Chapman,
2004; Adler and Borys, 1996). It has similarities with the diag-
nostic budgetary control highlighted by Simons (1990) except for
the reforecast dimension: in Simons’ model, a diagnostic budget is
associated with a low level of reforecasting. There are two possible
explanations for the high frequency of reforecasting in our coer-
cive budget. Since budget targets are so hard to reach, operational
performances tend to fall short. The budget can no longer be used
to steer the firm, and another set of data has to be used to fulfill
this role. An alternative explanation would be that the demand for
reforecasts comes from external stakeholders because they know
that the budget does not provide a reliable benchmark. These reg-
ular reforecasts can generate frustration for operational managers,
as they show the inaccuracy and irrelevance of the initial budget
that is the basis for the managers’ evaluation and rewards.
55
4.3. Cluster 3: interactive budget
For cluster 3, the mean of all budget design and use characteris-
tics is significantly higher than the grand mean for the other clusters
(p < 0.01), except for the difficulty of meeting the budget which is
significantly lower (−0.431**).
In this cluster, operational managers actively participate in
drafting the budget (0.672**) and budget meetings deal with action
plans and operational issues (0.821**).
The budget is a state of mind to be shared with all collaborators
and managers. It’s a planning tool for every actor, who actively
participates in the drafting of his own budget, guided by the sup-
port and the critical eye of the Finance department. The budget
enables us to set objectives, to provide some means to achieve
them and to analyze the variances between expected and actual
results every month. (Respondent, Cluster 3).
The budget is more detailed in cluster 3 than in other clusters
(0.849**), with a large number of reforecasts throughout the year
(0.487**) (more than in any other group). The budget is used as
the basis for discussions of ongoing changes: managers are highly
involved in the budget process (0.928**).
The budget is still a time for communication and sharing with
the various levels of the organization (staff and line). It makes it
possible to reframe the implementation of strategies and their
effects. (Respondent, Cluster 3).
The budget means we can quantify creative ideas, and forces
everyone to develop their arguments. An idea might be good, but
what about its impact on other departments? How far should I
go to make it really worth it? (Respondent, Cluster 3).
The budget is a must-do exercise. It’s the logical step between
the three-year strategic planning and the monthly reforecast
exercises. The budget increasingly has to be completed through
reforecast exercises to update the orientations chosen in the
budget. Our budget is really operational; it’s an interactive,
monthly updated tool. (Respondent, Cluster 3).
The examination of budget variance is more extensive than in
any other group (0.842**) and ultimately, budget targets are used
to evaluate (0.656**) and reward (0.421**) managers.
Budget targets should be ambitious but achievable (which
means that they won’t be systematically achieved). (Respon-
dent, Cluster 3).
In our firm, the budget is steered and analyzed on every detail.
But the targets are set for a bunch of indicators that vary accord-
ing to managers’ accountability. Naturally, managers always try
to get the lowest objectives possible, whereas the finance guys
want to push them as far as they can. The final word on the level
of the objectives depends on the skills of the top management,
their culture. (Respondent, Cluster 3).
In cluster 3, the budget is considered as a major instrument
(0.752**). It is used intensively to fulfill multiple roles: management
(0.500***), strategy deployment (0.559**), resource allocation and
authorizing spending (0.362**), and even discussions with share-
holders (0.149). More than in other clusters, the budget is viewed by
respondents as a crucial instrument, fulfilling all its possible roles.
The budget is designed and used to promote communication
and discussion amongst different levels of the hierarchy.
The budget is an essential tool for management. It facilitates
steering actions and the achievement of objectives. (Respon-
dent, Cluster 3).
The budget is meant to create and diffuse objectives and pro-
vides the means to achieve them within the organization. When
drafting the budget, senior managers take the opportunity to
56 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
diffuse their objectives and piece it all together so that the goals
get achieved. Naturally, as this process requires frequent meet-
ings, it’s also an opportunity to listen to new ideas and take them
up. (Respondent, Cluster 3).
The budget is the backbone of the firm. It directs all energies
towards shared objectives. It can act as a catalyst and unifying
tool if management uses it properly. (Respondent, Cluster 3).
Both financial (0.783**) and operational managers (0.72**) are
satisfied with this budget overall. The level of criticisms is signif-
icantly lower than in the other clusters: it is viewed neither as a
constraint (−0.274**), nor as a ritual (−0.403**) or as having any
adverse effects (−0.327**). This observation leads us to question
the validity of Arwidi and Samuelson’s (1993) assumption that
assigning a large number of roles to the budget can explain the
level of criticisms it attracts. The commitment that this budget
style elicits may be explained by the way it is constructed – it is
highly formalized, very participative and oriented towards opera-
tional issues – and the way it is used—with high levels of senior
management involvement and reforecasts. This type of construc-
tion and use make the budget a central tool in discussing ongoing
changes and keeping the firm in line with shareholders’ expecta-
tions. It seems that these firms apply the budget in the “spirit of
a game” as recommended by Hofstede (1967). The high level of
manager participation in the target-setting process, and the fact
that targets are relatively easy to reach, legitimizes the reward and
evaluation dimension of this budget.
The budget characteristics in this cluster appear to make it an
interactive budgeting tool, as described by Simons (1987b), who
referred to Johnson and Johnson, except on the reward dimension.
It is closely related to the continuous budget highlighted by Frow
et al. (2010) and the enabling budget defined by Chapman and Kihn
(2009).
4.4. Cluster 4: loose budget
In cluster 4, the budget is constructed in a participative (0.317*)
and mostly non-formalized way (−0.331**).
The operational managers draft the budgets, which are pre-
sented and approved (potentially after negotiation) in a meeting
with the CEO of the business unit. (Respondent, Cluster 4).
The drafting phase (of the budget) is a bit too long (two months),
but it does mean managers can be involved in the process.
(Respondent, Cluster 4).
Here the budget is viewed as easier to achieve than in other clus-
ters (−0.189**). Throughout the year, budget variance is analyzed
less than in the other clusters (−0.298**): senior management is
only moderately involved in the budgeting process (−0.123) and
the budget does not serve as a basis for discussion.
Some budget items are monthly reviewed, but, in general, the
review is global and on a three-month basis. Things get more
detailed in case of variances. (Respondent, Cluster 4).
We’re currently moving towards a monthly-reviewed reporting,
with a variance analysis. (Respondent, Cluster 4).
A notable characteristic of firms in cluster 4 is that they hardly
use the budget to evaluate performance (−0.870**) or reward
managers (−0.820**). The budget is used for its strategic role
(0.242*), but is perceived as less important than in the other clus-
ters (−0.347**): it is used less to assign responsibility, incentivize
(−0.440**) or communicate with external stakeholders (−.050).
The budget can create a continuous improvement culture as
long as it’s only used when necessary, to avoid short-term
behavior (such as driving the trucks with the sole aim of using
up the petrol budget by the end of the year. . .). (Respondent,
Cluster 4).
Equally, both operational managers (−0.122) and financial man-
agers (−0.202*) are less satisfied than in other firms. The budget
is not seen as a ritual (−.125) but is criticized, more than in the
other clusters, for its tendency to deter cooperation (despite the
considerable operational manager input it encourages in budget
construction) (0.072) and cause conservative behaviors (0.052), and
for its short-term orientation (0.181*).
The budget is a complex process which lacks reactivity. Bud-
get indicators are too general and take too long to be properly
monitored. (Respondent, Cluster 4).
Used as in cluster 4, the budget displays characteristics close to
Anthony’s loose budgetary control (1988).
4.5. Cluster 5: indicative budget
Cluster 5 is the opposite of cluster 3. The mean for all budget
design and use characteristics is lower than in other clusters, except
for reforecast and revision (not significant). The difficulty of the
budget is considered higher than in other clusters (0.620**).
In this cluster, senior management has very little involvement
in the budget process (−1.390**). This process is barely formalized
(−0.805**) and contains few details (−0.795**).
In my firm, the budgeting process is a mix of theory and good-
will. It can’t be used as a proper management tool correlating
assumptions, objectives and resources. (Respondent, Cluster 5).
Our budget is very (too) general. It’s a vague objective rather
than a management tool. No variances (either against monthly
forecast results, or monthly actual results) are calculated in our
monthly reviews. (Respondent, Cluster 5).
Operational manager participation in budget construction is low
(−1.245**).
The budget is a compulsory exercise in the group. We don’t draft
the budget we need, but the budget we’re required to draft.
The drafting exercise is a total sham! Officially, it’s a bottom-
up process, but the final version is purely top-down. The word
“negotiation” is a misnomer. (Respondent, Cluster 5).
As a result, the budget is poorly linked to action plans (−0.927**).
I’m convinced that budget drafting should rely on action plans,
but in practice this objective is rarely achieved in our firm
(action plans are not very formalized and take second place to
the budgeting process!). (Respondent, Cluster 5).
Action plans are defined after the budget has been set up, as if
they exist to justify the budget rather than as if they were the inter-
pretation of the strategy, and thus the reason behind the budget.
(Respondent, Cluster 5).
Throughout the year, variance analysis is limited and the budget
does not serve as a basis for discussions about ongoing changes
(−1.051**).
The budget is a tool for monitoring charges (or sales). It indicates
the level of consumption of our resources. (Respondent, Cluster
5).
Budget targets are only rarely met and they are not generally
used to evaluate (−0.721**) or reward (−0.428**) managers.
It is a tool for resource allocation which remains strictly finan-
cial. Operational managers did not appropriate the tool for
themselves, and they only use it to get spending authorizations.
(Respondent, Cluster 5).
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
According to respondents from firms in cluster 5, the budget is
used less than in other clusters to assign responsibility, incentivize
or evaluate: it is rarely used as a management tool (−0.773**). Also,
it is not viewed as a strategic tool (−0.274*): it is rarely used to
allocate resources, coordinate, manage risk or steer the firm. More-
over, it is less frequently used than in the other clusters to authorize
spending (−0.556*) or roll out strategy (−0.274*). In short, respon-
dents from firms in cluster 5 do not see the budget as a major
instrument (−0.877**).
Neither operational managers (−0.692**) nor financial man-
agers (−0.720**) appear to be satisfied with the budgeting process:
it is constraining, it imposes a culture of control rather than of
commitment (0.185), and it is seen as a ritual (0.233*).
The budgeting process depends on the top management style.
Depending on the firm, the budget can be just a ritual or a tool
to translate the strategy into action plans, resulting in an instru-
ment for motivation and communication. Unfortunately, in my
firm, the budget is nothing but a ritual. (Respondent, Cluster 5).
The use of the budget in this cluster closely relates to the char-
acteristics of Merchant’s interpersonal budgetary control (1981). In
these organizations, budget control is not considered a predomi-
nant mode of coordination and control. Instead, the budget in these
firms seems to be constructed for indication, and is therefore indica-
tive in nature. Other types of control may consequently have some
level of significance—for instance, direct supervision, the clan or
the hierarchy (Ouchi, 1979).
4.6. Patterns of budget design and use
Table 7 summarizes the budget characteristics, budget roles and
budget satisfactions of the five clusters.
The yardstick budget is intended to mobilize both top and oper-
ational managers on strategy in the long run, and encourages them
to closely follow its deployment in the short run. The coercive bud-
get serves a management by objectives approach. The loose budget
is a tool for deploying the strategy. The interactive budget combines
management by objectives and strategy deployment. The indicative
budget does not have any clearly defined function.
Both interactive and coercive budgets are used to evaluate and
reward. However, high levels of participation, high importance
assigned to action plans during budget negotiations, high levels of
variance analysis and significant senior management involvement
all result in the interactive budget – in clear contrast to the coercive
budget – generating high budget satisfaction and little criticism.
The interactive budget operates as a central tool to structure nego-
tiations and discuss strategic choices: it is therefore legitimate to
use it for performance evaluation and reward. The coercive budget,
on the other hand, does not foster discussion, and is only used for
ex post evaluation: as such, it is perceived as a tool for sanction-
ing rather than a useful management tool for monitoring business
activity. These results suggest that managers do not condemn the
use of budgets to evaluate performance per se; what they object to
is using the budget to this end in a purely mechanical way.
A low level of budget evaluation and weak links between the
budget and incentives characterize loose and indicative budgets.
Specifically, a loose budget is distinguished by strong participa-
tion, numerous revisions during the year, little criticism (except
regarding its short-term vision) and a low level of satisfaction. The
same applies to the indicative budget, which, although unrelated to
performance evaluation, is considered as difficult to achieve. Unre-
alistic goals plus a low impact on evaluations may certainly explain
why the budget is considered as a ritual of little importance that
generates little satisfaction. Analysis of both these groups suggests
57
that budget participation and budget flexibility are not sufficient to
ensure budget satisfaction.
5. Discussion and conclusion
The purpose of our study was to develop a taxonomy of budget
configurations. More precisely, we empirically examined how the
different styles of budget design and use combine in configurations
and the association that can be made with budget satisfaction. Com-
bining a qualitative analysis and an ascending hierarchical cluster
analysis performed on data gathered from 269 firms, we identi-
fied five patterns of budget practices: the yardstick budget, the
coercive budget, the interactive budget, the loose budget and the
indicative budget. Each of these clusters displays different budget
roles and different levels of satisfaction. The taxonomy presented in
this study offers a number of implications for future theory devel-
opment in budgeting research.
First, using a configurational approach combining a variety of
budget characteristics, we identified patterns of budget design and
use that cannot be considered in focused prisms such as those
developed in the Reliance on Accounting Performance Measures
(RAPM) literature. In line with Libby and Lindsay (2010), our study
confirms that using the budget to evaluate performance is not a
universal practice, and that some organizations explicitly link their
budget to strategy. By extending Libby and Lindsay’s (2010) find-
ings to the French context, we answer the call by Derfuss (2009) to
extend our knowledge of budgeting practices in European contexts.
Second, by considering a large number of design and use char-
acteristics in the budgeting process and illustrating them with
qualitative insights, our study reveals a more complex image of
organizational reality. We refined the representation of budgetary
practices, opening the way to a better understanding of budgeting
in practice. The yardstick, coercive and interactive styles of bud-
gets all correspond to different forms of what is generally referred
to as “budgetary control”—i.e., the use of the budget to monitor
results throughout the year and to evaluate performance at the
end of the year (Ansari, 1979; Donner, 1932; Flamholtz, 1983). We
show that each of these styles combines budgetary control with
other design and use characteristics. Certain styles of budgetary
controls are only reactive, ex post controls (the coercive budget)
while others are more proactive (the yardstick and interactive bud-
gets) (Pitkänen and Lukka, 2011; Naranjo-Gil and Hartmann, 2006).
Additionally, we observed that, sometimes, budgets are not used
to control but play a role in defining strategy (the loose budget).
Finally, sometimes, the budget is more akin to a rational myth:
it gives the illusion of rationality (Fernandez-Revuelta Perez and
Robson, 1999), without any apparent use for management (the
indicative budget).
Echoing Bedford and Malmi (2015, p. 17), we consider that “rich
descriptions of the actual choices made in practice may be a prereq-
uisite for empirically valid theoretical developments, or at the very
least, highly instructive.” In analyzing complex budget configura-
tions in their diversity, this study provides empirical observations
to refine and extend existing knowledge about budgets, particularly
the consequences and impacts of budgets.
Third, we have expanded the understanding of budget roles and
budget satisfaction in various ways. By showing that budgets can-
not be described as a universal, fully standardized and stable set of
tools with standardized uses, we cast some doubts on the relevance
of recommending the plain and simple elimination of budgeting
(Hope and Fraser, 2003). In this way, our findings confirm that
“the problems with budgeting are more to do with how [budgets]
are used and some of the roles they are asked to play” (Libby and
Lindsay, 2010, p. 60). More specifically, we have shown that each
of the five budget patterns plays a different role and is character-
58 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61

Table 7
Variables in the different clusters. a,b

Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5

Budget characteristics Participation 0.375(**) −0.292(**) 0.672(**) 0.317(**) −1.245(**)

Negotiation 0.114 −0.210(*) 0.821(**) 0.000 −0.927(**)
Difficulty 0.104 0.116 -0.431(**) −0.189(*) 0.620(**)
Variance 0.296(*) 0.070 0.842(**) −0.298(**) −1.051(**)
Reforecasts −1.521(**) 0.467(**) 0.487(**) 0.014 −0.191
Revisions −0.756(**) −0.230(*) -0.002 0.712(**) −0.120
Evaluation 0.288(*) 0.609(**) 0.656(**) −0.870(**) −0.721(**)
Reward −0.126 0.806(**) 0.421(**) −0.820(**) −0.428(**)
Involvement 0.474(**) −0.045 0.928(**) −0.123 −1.390(**)
Detail 0.146 −0.254(*) 0.849(**) −0.096 −0.795(**)
Formalization 0.122 0.052 0.807(**) −0.331(**) −0.805(**)

Budget roles Shareholder budget 0.089 −0.146 0.149 −0.050 −0.003

Administrative budget 0.009 −0.034 0.362(**) 0.071 −0.556(**)
Managerial budget 0.275(*) 0.369(**) 0.500(**) −0.440(**) −0.773(**)
Strategic budget −0.159 −0.426(**) 0.459(**) 0.242(*) −0.274(*)

Budget satisfaction Constraint −0.191 0.192(*) -0.274(**) 0.052 0.185

Short-term 0.048 0.003 -0.026 0.181(*) −0.256(*)
Adverse effects 0.037 0.225(*) -0.327(**) 0.072 0.016
Ritual 0.200 0.256(**) -0.403(**) −0.125 0.233(*)
The budget is a major management instrument 0.190 0.141 0.752(**) −0.347(**) −0.877(**)
The budget brings satisfaction (to financial managers) 0.172 −0.144 0.783(**) −0.202(*) −0.720(**)
The budget brings satisfaction (to managers) 0.105 −0.183(*) 0.726(**) −0.122 −0.692(**)
a
To facilitate comparison, clusters are described by their factor scores on the different constructs. By design, the overall average for each construct is equal to zero.
b
Clusters in which the mean for a dimension is significantly different from the grand mean under the t-test (p < 0.05/p < 0.01) are identified with (*/**).

ized by a different level of satisfaction. Contrary to the assumption
set out by Arwidi and Samuelson (1993), we observed that mul-
tiplying the roles assigned to a budget does not necessarily lead
to significant criticism. For example, the interactive budget is used
both as a management tool and a strategic tool, and yet it is subject
to very few criticisms. This finding confirms synergies between the
planning and performance control functions of budgetary practices
(Fisher et al., 2002). We have also shown that when budgets are
solely used to assign responsibility, incentivize and contractualize
(the coercive budget), it is highly criticized. Our findings suggest that
the level of criticism concerning the use of budgets to evaluate and
reward managers is lower when evaluation is coupled with other
budget practices: the budget tends to be less criticized when the
level of operational manager participation, the senior management
involvement and the importance assigned to action plans during
budget negotiations are high. In addition, we hope our results gen-
erate a discussion of the criticisms related to budget rigidity (Libby
and Lindsay, 2010; Henttu-Aho and Järvinen, 2013): revisions and
reforecasts are not a guarantee of budget satisfaction.
These results contribute to the debates in the RAPM literature
and suggest that interactions between the various components of
the budget process would explain the less-than-conclusive findings
of research carried out on the effect of using budgets to evalu-
ate performance (Hartmann, 2000). Echoing the results of Derfuss
(2009) and Chong and Mahama (2014), we confirm the impor-
tance of participation and involvement of managers for budget
satisfaction. We also underline the importance of the content of
discussions during budget negotiations: discussing budgets based
on action plans and strategic objectives is more satisfying than
discussing budgets only on a financial level. Associating budget sat-
isfaction with the participation, involvement and consideration of
action plans during budget negotiations suggests that individuals
are more satisfied with the budget when it is used as a tool for dis-
cussion, exchange on business issues, or even socialization within
the organization.
These findings are not free from certain limitations. Method-
ologically speaking, as an explorative data technique, cluster
analysis is a less rigorous set of techniques than more developed
methods. Furthermore, it gives equal weight to different items in
the analysis, and some dimensions might be missing. For instance,
budget-based unit evaluation was not included in the study. We
also acknowledge that some of the reliability alphas and AVEs
are low. It should also be noted that responses were provided
from a sample of management accountants and CFOs and that
the French context in which this study was conducted may have
also influenced our findings. Replicating the study in a different
national context would be useful to increase the generalizability
of the findings. Lastly, this study does not take into account the
fact that individual-level variations occur in the subjective mean-
ings managers give to budgets. Managers’ organizational positions,
functional roles and educational and work histories (Kihn, 2011),
for instance, could be considered in future quantitative studies
to understand budget practices, budget roles and budget satisfac-
tion.
Regardless of these limitations, we think that the configura-
tional approach applied here will be looked at in future research
to better understand budgets as sets of practices. We also hope
that the insights of this study into budget configurations will con-
tribute towards further developing theories of budget practice
variety.
Acknowledgements
We would like to acknowledge the helpful comments of Jonas
Gerdin, Wim Van der Stede and participants at the 7th Conference
on new directions in management accounting (2011) and at the
34th EAA Annual Congress (2011). We would also like to thank
Philippe Caillou for his help on data analysis. We are grateful to the
reviewers and the editor for their comments.
Appendix A. —budget practices: measurement instrument
Please indicate the degree to which you agree with the following
statements as they apply to your organization.
Scale: 1 = strongly disagree to 5 = strongly agree
(*) Item recoded into five categories (1 → 5)
(R) Item reverse coded
Participation (␣ = 0.8; AVE = 0.528)
Operational managers have decisive influence on their budget
targets
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61 59

Budget targets are decided in a bottom-up approach

Operational managers play a significant role in drafting their

Mean on 500
budgets

subsets
What role do operational managers play in drafting their budget

−1.214

0.661

−0.154
−0.125
−0.733
−0.487
−1.413

−0.851
−0.910

−1.120

−0.801
forecasts? (1 = not at all important/5 = very important)
Type of negotiation (␣ = 0.83; AVE = 0.603)
Drafting action plans is a major stage in the budgeting process

Original sample
How important is it to prepare action plans? (1 = not at all impor-
tant/5 = very important)

Cluster 5

−1.245
−0.927

−0.191

−0.721
−0.428

−0.795
0.620
−1.051

−1.390

−0.805
Action plans are constructed before budgets

−0.12

0.73
During negotiations, every significant modification in the bud-
get leads to a modification of the underlying action plans
Difficulty (␣ = 0.52; AVE = 0.534)

Mean on 500
The likelihood of a budget manager meeting his budget target
is low

subsets

0.286

−0.253
−0.238

0.666
−0.881
−0.055

−0.805
−0.108
−0.101
−0.304
0.007
Budget targets are usually met (R)
Budget variance (␣ = 0.75; AVE = 0.522)
Corrective actions must be proposed by managers when a vari-

Original sample
ance emerges between actual performance and the budget over the
course of the year

Cluster 4
Great importance is assigned to budget variances

0.317

−0.189
−0.298

0.712

−0.123

−0.331
0.014

−0.870
−0.820

−0.096
Budget variances give rise to discussions between senior man-

0.74
0
agement and the manager in question
6.6 Reforecasts (␣ = 0.62; AVE = 0.498)
How many reforecasts are there each year? (*)

Mean on 500
Budgets are subject to regular revision to take into account
changes in the environment

subsets

0.637

−0.423
0.799
0.561

0.591
0.447
0.931
0.851
0.758
−0.089
0.900
Revisions (␣ = NA; AVE = NA)
Budget targets cannot be changed over the year (R)
Budget-based evaluation (␣ = 0.81; AVE = 0.599)

Original sample
Not meeting budget targets reflects badly on performance
Not meeting budget targets has a significant impact on the per- Cluster 3
formance evaluations of operational managers

0.672
0.821
−0.431
0.842
0.487

0.656
0.421
0.928
0.849
0.807
−0.002

0.80
The performance of operational managers is primarily judged
on their ability to meet their budget targets
Budget-based reward (␣ = 0.88; AVE = 0.736)
Budget-based rewards make up a high proportion of salary
Mean on 500

What percentage of the budget managers’ compensation is tied

subsets

to meeting budget targets? (*)

−0.311

0.134

0.437
−0.179
0.623
0.793

−0.258
−0.302

0.082

−0.078

0.073
Rewards for operational managers depend largely on their
meeting budget targets
Involvement in budgeting (␣ = 0.81; AVE = 0.542)
Original sample

The budgeting process is subject to frequent scrutiny by man-

agers at all levels
Cluster 2

−0.292

0.116

0.467

−0.254
−0.210

−0.230
0.609
0.806
−0.045

0.052
The budgeting process is subject to frequent and regular scrutiny
0.07

0.84

by senior management
During budget drafting, there are a large number of discussions
between senior management and operational managers
Mean on 500

Senior management uses budget information as a means to

question and debate the decisions and the actions of operational
subsets

0.159
0.318
−1.513
−0.642
0.469

0.494
0.142
0.181
0.430
0.064

−0.044

managers
Budget details (␣ = 0.63; AVE = 0.384)
Budget variance analysis for each operational manager is per-
Original sample

formed line by line

Appendix B. —clusters stability

Budget monitoring reports are not very detailed and only con-
Cluster 1

tain aggregate data (R)

0.375
0.114

0.296
−1.521
−0.756
0.288
−0.126
0.474
0.146
0.122
0.104

Budget negotiations deal with very detailed budgets

% Of correct classification 0.75

Formalization (␣ = 0.54; AVE = 0.406)

The budgeting process is the same for all responsibility centers
Detailed standards frame the budget drafting (the budgeting
process is highly standardized)
Formalization
Participation

Involvement
Negotiation

Reforecasts

Evaluation
Revisions
Difficulty
Variance

Reward

Detail
60 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
References
Abernethy, M.A., Brownell, P., 1999. The role of budgets in organizations facing
strategic change: an exploratory study. Account. Org. Soc. 24, 189–204, http://
dx.doi.org/10.1016/s0361-3682(98) 00059-2.
Adler, P.S., Borys, B., 1996. Two Types of bureaucracy: enabling and coercive. Adm.
Sci. Q. 41, 61–89.
Ahrens, T., Chapman, C.S., 2004. Accounting for flexibility and efficiency: a field
study of management control systems in a restaurant chain*. Contemp.
Account. Res. 21, 271–302.
Ansari, S.L., 1979. Towards an open systems approach to budgeting. Account. Org.
Soc. 4, 149–161, http://dx.doi.org/10.1016/0361-3682(79) 90023-0.
Anthony, R.N., 1965. Planning and Control Systems: A Framework for Analysis, first
edition. Division of Research, Harvard Business School, Boston.
Arwidi, O., Samuelson, L.A., 1993. The development of budgetary control in
Sweden—a research note. Manage. Account. Res. 4, 93–107, http://dx.doi.org/
10.1006/mare.1993.1007.
Barrett, M.E., Fraser III, L.B., 1977. Conflicting roles in budgeting for operations.
Harv. Bus. Rev. 55, 137–146.
Bedford, D.S., Malmi, T., 2015. Configurations of control: an exploratory analysis.
Manage. Account. Res. 27, 2–26, http://dx.doi.org/10.1016/j.mar.2015.04.002.
Bhimani, A., Horngren, C.T., Datar, S.M., Foster, G., 2008. Management and Cost
Accounting. Financial Times Prentice Hall.
Bisbe, J., Batista-Foguet, J.-M., Chenhall, R., 2007. Defining management accounting
constructs: a methodological note on the risks of conceptual misspecification.
Account. Org. Soc. 32, 789–820, http://dx.doi.org/10.1016/j.aos.2006.09.010.
Bouquin, H., 2010. Le contrôle de gestion, 9th ed. Presses Universitaires de
France—PUF.
Briers, M., Hirst, M., 1990. The role of budgetary information in performance
evaluation. Account. Org. Soc. 15, 373–398, http://dx.doi.org/10.1016/0361-
3682(90) 90024-O.
Brownell, P., 1983. The motivational impact of management-by-exception in a
budgetary context. J. Account. Res. 21, 456–472, http://dx.doi.org/10.2307/
2490785.
Brownell, P., Hirst, M., 1986. Reliance on accounting information, budgetary
participation, and task uncertainty: tests of a three-way interaction. J. Account.
Res. 24, 241–249.
Bunce, P., Fraser, R., Woodcock, L., 1995. Advanced budgeting: a journey to
advanced management systems. Manage. Account. Res. 6, 253–265, http://dx.
doi.org/10.1006/mare.1995.1017.
Chapman, C.S., Kihn, L.-A., 2009. Information system integration, enabling control
and performance. Account. Org. Soc. 34, 151–169, http://dx.doi.org/10.1016/j.
aos.2008.07.003.
Chenhall, R.H., Kallunki, J.-P., Silvola, H., 2011. Exploring the relationships between
strategy, innovation, and management control systems: the roles of social
networking, organic innovative culture, and formal controls. J. Manage.
Account. Res. 23, 99–128, http://dx.doi.org/10.2308/jmar-10069.
Chenhall, R.H., Langfield-Smith, K., 1998. The relationship between strategic
priorities, management techniques and management accounting: an empirical
investigation using a systems approach. Account. Org. Soc. 23, 243–264.
Chong, K.M., Mahama, H., 2014. The impact of interactive and diagnostic uses of
budgets on team effectiveness. Manage. Account. Res. 25, 206–222, http://dx.
doi.org/10.1016/j.mar.2013.10.008.
Collier, P.M., Berry, A.J., 2002. Risk in the process of budgeting. Manage. Account.
Res. 13, 273–297, http://dx.doi.org/10.1006/mare.2002.0190.
Covaleski, M.A., Evans, J.H., Luft, J.L., Shields, M.D., 2003. Budgeting research: three
theoretical perspectives and criteria for selective integration. J. Manage.
Account. Res. 15, 3–49, http://dx.doi.org/10.2308/jmar.2003.15.1.3.
De Baerdemaeker, J., Bruggeman, W., 2015. The impact of participation in strategic
planning on managers’ creation of budgetary slack: the mediating role of
autonomous motivation and affective organisational commitment. Manage.
Account. Res. 29, 1–12, http://dx.doi.org/10.1016/j. mar.2015.06.002.
De Loo, I., Lowe, A., 2011. Mixed methods research: don’t-just do it. Qual. Res.
Account. Manage. 8, 22–38, http://dx.doi.org/10.1108/11766091111124685.
Derfuss, K., 2009. The relationship of budgetary participation and reliance on
accounting performance measures with individual-level consequent variables:
a meta-analysis. Eur. Account. Rev. 18, 203–239, http://dx.doi.org/10.1080/
09638180802652371.
Dillman, D.A., Smyth, J.D., Christian, L.M., 2008. Internet, Mail, and Mixed-Mode
Surveys: The Tailored Design Method. John Wiley & Sons.
Donner, F.G., 1932. General motors budgetary control. Account. Rev. 7, 22–30,
http://dx.doi.org/10.2307/239130.
Ekholm, B.-G., Wallin, J., 2000. Is the annual budget really dead? Eur. Account. Rev.
9, 519, http://dx.doi.org/10.1080/09638180020024007.
Emsley, D., 2001. Redesigning variance analysis for problem solving. Manage.
Account. Res. 12, 21–40, http://dx.doi.org/10.1006/mare.2000.0138.
Fernandez-Revuelta Perez, L., Robson, K., 1999. Ritual legitimation, de-coupling
and the budgetary process: managing organizational hypocrisies in a
multinational company. Manage. Account. Res. 10, 383–407, http://dx.doi.org/
10.1006/mare.1999.0114.
Fisher, J., Frederickson, J.R., Peffer, S.A., 2002. The effect of information asymmetry
on negotiated budgets: an empirical investigation. Account. Org. Soc. 27,
27–43, http://dx.doi.org/10.1016/s0361-3682(01) 00046-0.
Flamholtz, E.G., 1983. Accounting, budgeting and control systems in their
organizational context: theoretical and empirical perspectives. Account. Org.
Soc. 8, 153–169, http://dx.doi.org/10.1016/0361-3682(83) 90023-5.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation models with
unobservable variables and measurement error. J. Market. Res.,
39–50.
Frow, N., Marginson, D., Ogden, S., 2010. Continuous budgeting: reconciling budget
flexibility with budgetary control. Account. Org. Soc. 35, 444–461, http://dx.
doi.org/10.1016/j.aos.2009.10.003.
Gerdin, J., 2005. Management accounting system design in manufacturing
departments: an empirical investigation using a multiple contingencies
approach. Account. Org. Soc. 30, 99–126, http://dx.doi.org/10.1016/j.aos.2003.
11.003.
Gerdin, J., Greve, J., 2008. The appropriateness of statistical methods for testing
contingency hypotheses in management accounting research. Account. Org.
Soc. 33, 995–1009, http://dx.doi.org/10.1016/j.aos.2007.07.003.
Gerdin, J., Greve, J., 2004. Forms of contingency fit in management accounting
research—a critical review. Account. Org. Soc. 29, 303–326, http://dx.doi.org/
10.1016/s0361-3682(02) 00096-x.
Grabner, I., Moers, F., 2013. Management control as a system or a package?
Conceptual and empirical issues. Account. Org. Soc. 38, 407–419, http://dx.doi.
org/10.1016/j.aos.2013.09.002.
Grafton, J., Lillis, A.M., Mahama, H., 2011. Mixed methods research in accounting.
Qual. Res. Account. Manage. 8, 5–21, http://dx.doi.org/10.1108/
11766091111124676.
Hair, J.F., Black, W.C., Babin, B.J., Anderson, R.E., 2010. Multivariate Data Analysis,
7th ed. Prentice Hall.
Hansen, S.C., Van Der Stede, W.A., 2004. Multiple facets of budgeting: an
exploratory analysis. Manage. Account. Res. 15, 415–439.
Harrison, G.L., 1993. Reliance on accounting performance measures in superior
evaluative style—the influence of national culture and personality. Account.
Org. Soc. 18, 319–339, http://dx.doi.org/10.1016/0361-3682(93)90019-3.
Hartmann, F.G.H., 2000. The appropriateness of RAPM: toward the further
development of theory. Account. Org. Soc. 25, 451–482, http://dx.doi.org/10.
1016/s0361-3682(98)00036-1.
Hennig, C., 2007. Cluster-wise assessment of cluster stability. Comput. Stat. Data
Anal. 52, 258–271, http://dx.doi.org/10.1016/j.csda.2006.11.025.
Henttu-Aho, T., Järvinen, J., 2013. A field study of the emerging practice of beyond
budgeting in industrial companies: an institutional perspective. Eur. Account.
Rev., 1–21, http://dx.doi.org/10.1080/09638180.2012.758596.
Hinkin, T.R., 1998. A brief tutorial on the development of measures for use in
survey questionnaires. Org. Res. Methods 1, 104–121, http://dx.doi.org/10.
1177/109442819800100106.
Hirst, M.K., Lowy, S.M., 1990. The linear additive and interactive effects of
budgetary goal difficulty and feedback on performance. Account. Org. Soc. 15,
425–436, http://dx.doi.org/10.1016/0361-3682(90) 90026-q.
Hofstede, G.H., 1967. The Game of Budget Control, Technology and Democratic
Society. Van Gorcum & Company, Assen.
Hope, J., Fraser, R., 2003. Beyond Budgeting: How Managers Can Break Free from
the Annual Performance Trap. Harvard Business Press.
Hopwood, A.G., 1972. An empirical study of the role of accounting data in
performance evaluation. J. Account. Res. 10, 156–182.
Hurmerinta-Peltomäki, L., Nummela, N., 2006. Mixed methods in international
business research: a value-added perspective. Manage. Int. Rev. 46, 439–459,
http://dx.doi.org/10.1007/s11575-006-0100-z.
Husson, F., Lê, S., Pagès, J., 2010. Exploratory Multivariate Analysis by Example
Using R. CRC Press.
Ittner, C.D., Larcker, D.F., 1997. Quality strategy, strategic control systems, and
organizational performance. Account. Org. Soc. 22, 293–314, http://dx.doi.org/
10.1016/S0361-3682(96)00035-9.
Ketchen, D.J., Shook, C.L., 1996. The application of cluster analysis in strategic
management research: an analysis and critique. Strategic Manage. J. 17,
441–458.
King, R., Clarkson, P.M., Wallace, S., 2010. Budgeting practices and performance in
small healthcare businesses. Manage. Account. Res. 21, 40–55, http://dx.doi.
org/10.1016/j.mar.2009.11.002.
Kihn, L.A., 2011. How do controllers and managers interpret budget targets? J.
Account. Org. Change 7, 212–236.
Lebart, L., Piron, M., Morineau, A., 2006. Statistique exploratoire
multidimensionnelle: visualisations et inférences en fouille de données.
Dunod.
Lê, S., Josse, J., Husson, F., 2008. FactoMineR: an R Package for multivariate analysis.
J. Stat. Software 25, 1–18.
Libby, T., Lindsay, R.M., 2010. Beyond budgeting or budgeting reconsidered? A
survey of North-American budgeting practice. Manage. Account. Res. 21,
56–75, http://dx.doi.org/10.1016/j.mar.2009.10.003.
Lyne, S.R., 1988. The role of the budget in medium and large uk companies and the
relationship with budget pressure and participation. Account. Bus. Res. 18,
195–212, http://dx.doi.org/10.1080/00014788.1988.9729367.
Malmi, T., Brown, D.A., 2008. Management control systems as a
package—opportunities, challenges and research directions. Manage. Account.
Res. 19, 287–300, http://dx.doi.org/10.1016/j.mar.2008.09.003.
Marginson, D., Ogden, S., 2005. Coping with ambiguity through the budget: the
positive effects of budgetary targets on managers’ budgeting behaviours.
Account. Org. Soc. 30, 435–456, http://dx.doi.org/10.1016/j.aos.2004.05.004.
Merchant, K.A., 1981. The design of the corporate budgeting system: influences on
managerial behavior and performance. Account. Rev. 56, 813–829.
Merchant, K.A., Manzoni, J.-F., 1989. The achievability of budget targets in profit
centers: a field study. Account. Rev. 64, 539–558.
 S. Sponem, C. Lambert / Management Accounting Research 30 (2016) 47–61
Meyer, A.D., Tsui, A.S., Hinings, C.R., 1993. Configurational approaches to
organizational analysis. Acad. Manage. J. 36, 1175–1195.
Meyers, L.S., Gamst, G., Guarino, A.J., 2013. Applied Multivariate Research: Design
and Interpretation, 2nd ed. SAGE.
Milani, K., 1975. The relationship of participation in budget-setting to industrial
supervisor performance and attitudes: a field study. Account. Rev. 50, 274–284.
Miller, D., 1996. Configurations revisited. Strat. Manage. J. 17, 505–512, http://dx.
doi.org/10.1002/(SICI)1097-0266(199607)17:7<505::AID-SMJ852>3.0.CO;2-I.
Miller, D., Mintzberg, H., 1983. The case for configuration. In: Morgan, G. (Ed.),
Beyond Method: Strategies for Social Research. SAGE, pp. 57–73.
Modell, S., 2009. In defence of triangulation: a critical realist approach to mixed
methods research in management accounting. Manage. Account. Res. 20,
208–221, http://dx.doi.org/10.1016/j.mar.2009.04.001.
Moores, K., Yuen, S., 2001. Management accounting systems and organizational
configuration: a life-cycle perspective. Account. Org. Soc. 26, 351–389, http://
dx.doi.org/10.1016/s0361-3682(00) 00040-4.
Naranjo-Gil, D., Hartmann, F., 2006. How top management teams use management
accounting systems to implement strategy. J. Manage. Account. Res. 18, 21–53.
Østergren, K., Stensaker, I., 2011. Management control without budgets: a field
study of beyond budgeting in practice. Eur. Account. Rev. 20, 149–181, http://
dx.doi.org/10.1080/09638180903487842.
Otley, D., Fakiolas, A., 2000. Reliance on accounting performance measures: dead
end or new beginning? Account. Org. Soc. 25, 497–510, http://dx.doi.org/10.
1016/s0361-3682(98) 00007-5.
Ouchi, W.G., 1979. A conceptual framework for the design of organizational
control mechanisms. Manage. Sci. 25, 833–848.
Pitkänen, H., Lukka, K., 2011. Three dimensions of formal and informal feedback in
management accounting. Manage. Account. Res. 22, 125–137, http://dx.doi.
org/10.1016/j.mar.2010.10.004.
Pugh, D.S., Hickson, D.J., Hinings, C.R., Turner, C., 1968. Dimensions of organization
structure. Adm. Sci. Q. 13, 65–105, http://dx.doi.org/10.2307/2391262.
61
Shields, J.F., Shields, M.D., 1998. Antecedents of participative budgeting. Account.
Org. Soc. 23, 49–76, http://dx.doi.org/10.1016/s0361-3682(97) 00014-7.
Shields, M.D., Young, S.M., 1993. Antecedents and consequences of participative
budgeting: evidence on the effects of asymmetrical information. J. Manage.
Account. Res. 5, 265–280.
Simons, R., 1990. The role of management control systems in creating competitive
advantage: new perspectives. Account. Org. Soc. 15, 127–143, http://dx.doi.
org/10.1016/0361-3682(90) 90018-p.
Simons, R., 1987a. Accounting control systems and business strategy: an empirical
analysis. Account. Org. Soc. 12, 357–374, http://dx.doi.org/10.1016/0361-
3682(87) 90024-9.
Simons, R., 1987. Planning, control and uncertainty: a process view. In: Bruns, W.J.,
Kaplan, R.S. (Eds.), Accounting and Management: Field Study Perspectives.
Harvard Business School Press, pp. 339–362.
Ueno, S., Sekaran, U., 1992. The influence of culture on budget control practices in
the USA and Japan: an empirical study. J. Int. Bus. Stud. 23, 659–674.
Van der Stede, W.A., 2000. The relationship between two consequences of
budgetary controls: budgetary slack creation and managerial short-term
orientation. Account. Org. Soc. 25, 609–622, http://dx.doi.org/10.1016/S0361-
3682(99)00058-6.
Van der Stede, W.A., 2001. Measuring tight budgetary control. Manage. Account.
Res. 12, 119–137, http://dx.doi.org/10.1006/mare.2000.0149.
van Veen-Dirks, P., 2010. Different uses of performance measures: the evaluation
versus reward of production managers. Account. Org. Soc. 35, 141–164, http://
dx.doi.org/10.1016/j.aos.2009.02.002.
Webb, R.A., 2002. The impact of reputation and variance investigations on the
creation of budget slack. Account. Org. Soc. 27, 361–378, http://dx.doi.org/10.
1016/s0361-3682(01) 00034-4.