University of Groningen

Co-existing concepts of management control

Tillema, Sandra; van der Steen, Martijn

Published in:
Management Accounting Research

DOI:
10.1016/j.mar.2015.01.002

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.

Document Version
Publisher's PDF, also known as Version of record

Publication date:
2015

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Tillema, S., & van der Steen, M. (2015). Co-existing concepts of management control: The containment of
tensions due to the implementation of lean production. Management Accounting Research, 27, 67-83.
https://doi.org/10.1016/j.mar.2015.01.002

Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.
More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-
amendment.

Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.

Download date: 10-02-2024

 Management Accounting Research 27 (2015) 67–83

Contents lists available at ScienceDirect

Management Accounting Research

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

Co-existing concepts of management control

The containment of tensions due to the implementation
of lean production
Sandra Tillema, Martijn van der Steen ∗
University of Groningen, Department of Accounting, Faculty of Economics and Business, P.O. Box 800, 9700 AV Groningen,
The Netherlands

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

There is substantial evidence that the implementation of lean production and similar
Keywords:
Lean production
innovations is not always successful. One of the explanations provided is that elements
Implementation of traditional control systems may frustrate the transformation process. Although vari-
Management control ous studies have investigated the changes in control systems due to the implementation
Control system of lean production, only a few studies have explored the effects of the remaining tradi-
tional controls on lean implementations. This paper argues that lean production brings
with it a new concept of control, which alters people’s views of being in control. The
new concept of control may co-exist with the traditional concept, but particularly at their
interfaces, tensions may arise. Using case studies in four manufacturing companies in the
Netherlands, this paper explores the various localised ways in which these companies dealt
with such tensions to ensure that lean production continued. The paper concludes that
lean transformations do not require a fundamental resolution of the problems that arise
from inconsistent concepts of control, as long as companies have learned to cope with the
localised tensions that may result from inconsistencies between such concepts.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction production innovations, such as just-in-time (JIT) pro-

duction and total quality management (TQM). Similar
In recent years, fundamental changes have been tak- to its predecessors, it has been developed to improve
ing place in the production domain of many companies. manufacturing and business processes, and to enhance
One of the most influential changes is the implementation competitiveness and performance. Using case studies and
of lean production (Fullerton et al., 2010; Holweg, 2007). surveys, several researchers confirm that lean produc-
Core elements of a lean production system are the con- tion has a positive impact on operating performance (e.g.
tinuous reduction and ultimately elimination of all forms Krafcik, 1988; Shah and Ward, 2003). However, other
of waste, and the continuous improvement of the quality researchers are less optimistic about the successes that can
of products and processes (Cua et al., 2001; see also Shah be attributed to lean production and similar innovations.
and Ward, 2003). Lean production encompasses earlier They describe complications in the implementation of lean
production (Bamber and Dale, 2000; Worley and Doolen,
2006), and they point out that there is a substantial failure
∗ Corresponding author. Tel.: +31 50 363 3993; fax: +31 50 363 7356. rate (Bhasin, 2012; Sohal and Egglestone, 1994). Moreover,
E-mail addresses: [email protected] (S. Tillema), they find that the effects on financial performance are not
[email protected] (M. van der Steen). significant (Sale and Inman, 2003; see also Lewis, 2000).

http://dx.doi.org/10.1016/j.mar.2015.01.002
1044-5005/© 2015 Elsevier Ltd. All rights reserved.
68 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

According to Voss (1995, p. 13), ‘Partial implementation, by control built into operating processes. Although the
failure to achieve desired performance change and aban- evidence of the adoption and success of ‘lean accounting’ is
doned programmes are commonplace’. limited, it is clear that significant changes have taken place
In an attempt to explain lean implementation fail- in the control systems of lean companies. Several studies in
ures, researchers have pointed at the influence of, for the fields of accounting and operations management recog-
instance, management support (Worley and Doolen, 2006), nise that these companies tend to use more organic control
employee education and training (Bamber and Dale, 2000; systems, with a more prominent role given to clan controls
Kassicieh and Yourstone, 1998), and organisational culture and a stronger emphasis on non-financial performance
(Prajogo and McDermott, 2005; Spear and Bowen, 1999). (e.g. Abernethy and Lillis, 1995; Kennedy and Widener,
These are all factors which are widely acknowledged to 2008). Other studies have found that the management
play an important role in any change programme. How- accounting systems of lean companies produce more fre-
ever, some authors also observe that elements of traditional quent and more non-financial information, which is made
control systems may be a hindrance to successfully imple- available to lower levels, and that many of these companies
menting lean production. For example, Maskell et al. (2012, have abandoned the allocation of overheads (e.g. Banker
p. 2) notice that: ‘Traditional [control] systems do not work et al., 1993; Fullerton et al., 2013). Overall, these findings
for companies pursuing Lean thinking; indeed they are show that control systems in lean production companies
actively harmful’. Similarly, Johnson (2006, p. 6) refers to have developed in directions which seek to overcome the
traditional control systems as ‘the number one enemy’ of problems identified by Johnson and Kaplan (1987).
lean production. However, another point made by Johnson and Kaplan
The backbone of traditional control systems is man- (1987) seems to have been ignored by most researchers.
agement accounting (Chenhall, 2003; Kaplan, 1984; Otley, According to Johnson and Kaplan, the newly developed
1994), in the sense that these control systems assign control systems should be used alongside the more tradi-
a central role to planning and budgeting, and empha- tional system. They argued that, as long as a single system
sise financial controls. Key characteristics are that these which can satisfy both internal and external requirements
systems are mainly concerned with providing financial has not been developed, the new systems should be used
information, in an aggregated form, and on a relatively for controlling internal operations, while the more tradi-
infrequent basis (Ittner and Larcker, 1995; Kaplan, 1986, tional system remains relevant for financial reporting. So
1989). Moreover, they generally allocate costs based these authors proposed decoupling as a way of dealing with
on direct labour hours or processing time, emphasise inconsistencies between the traditional and the new con-
variances from budgeted standards, and assess capital trol systems (cf. Meyer and Rowan, 1977). In their view,
investments on financial grounds. Already in the 1980s, the use of separate systems alongside each other should
authors such as Brimson (1987), Brimson and Berliner not be problematic, given the low cost and high power
(1987), Johnson and Kaplan (1987) and Kaplan (1984, 1989) of information-processing technology. Nevertheless, the
concluded that traditional control systems were increas- question of whether this argument holds in practice has
ingly unable to provide support for innovative production received only limited attention in the literature on lean
technologies, including JIT and TQM. They indicated that production.
these systems were incapable of valuing and recognising Researchers have found a reduced emphasis on tra-
the gains from such innovations, and that they encour- ditional accounting controls in lean companies, and a
aged companies to take decisions which would lead to, replacement of these traditional controls with lean controls
for instance, higher levels of inventory and lower qual- (e.g. Kennedy and Widener, 2008; Fullerton et al., 2013).
ity production. According to Johnson and Kaplan (1987), However, they have not investigated how these lean con-
traditional control systems provide information that is too trols interact with the remaining elements of companies’
late, too aggregated, too distorted (due to unnecessary cost traditional accounting-based control systems. As such, they
allocations), and too much focused on short-term finan- seem to assume that the traditional control system sim-
cial performance. They argued that these systems do not ply fades away. This assumption may be unrealistic. For
provide an accurate view of the efficiency and effectiveness example, shareholders or other providers of finance may
of internal operations. As a solution, Johnson and Kaplan take a more traditional, financially oriented perspective
proposed the development of new systems for process con- when evaluating the company. Furthermore, as the way
trol and product costing (see also Kaplan, 1988). in which a company is evaluated externally will have an
Since then, various publications, often aimed at prac- effect on how top management attempts to control the
titioners, have addressed the issue of how accounting operations, this traditional perspective is likely to move
systems should be adjusted to fit lean environments, and down the organisational hierarchy (cf. Cable, 2009). Such
to avoid the above problems (Brosnahan, 2008; Cable, a perspective results in a particular view of being in con-
2009; Johnson, 2006; Maskell and Kennedy, 2007; Maskell trol, which denotes the individual understanding of (1) the
et al., 2012). Under the heading of ‘lean accounting’, these degree to which the organisation is in control; and (2) the
publications call for accounting systems which organise actions that need to be taken to maintain or restore control.
costs by value stream, avoid the use of standard costs, and As lean production and traditional control systems rely on
present information related to changes in inventories and different concepts of control, at some level the traditional
overheads separately. In addition, the authors argue that view of being in control will come into contact with the lean
control based on a detailed tracking of internal transac- view of being in control. At these interfaces, inconsistencies
tions on paper or in computer systems should be replaced between these two concepts may result in disagreements
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 69

on whether the company is in control. The views of being in Lean production is based on a philosophy which empha-
control may come under tension, and potentially give rise sises value, value streams, flow (without interruption),
to conflict. This could become a hindrance to the successful pull production and perfection (Womack and Jones, 1996).
implementation of lean production, as is illustrated in the According to Womack et al. (1990, pp. 13–14), the basic dif-
case study findings of Åhlström and Karlsson (1996) and ference between mass production and lean production lies
Bamber and Dale (2000). in their ultimate objectives. Whereas mass producers aim
More generally, lean companies are likely to combine a at being ‘good enough’ (e.g. having an acceptable number
view of being in control which is derived from lean produc- of defects and a maximum acceptable level of invento-
tion and a more traditional view which gives a central place ries), lean producers aim at perfection. That is, they aim
to accounting. The lean view is more likely to be prevalent at continually declining costs, zero defects, zero invento-
at lower organisational levels, whereas the traditional view ries and endless product variety. To achieve these aims,
may be more prominent at higher levels. In this paper, we lean producers use a multidimensional approach that inte-
will investigate how tensions and potential conflicts due to grates a diverse set of management practices. Shah and
these different views of being in control can be contained. Ward (2003) classify these lean production practices into
We will also explore how companies attempt to resolve four internally consistent and empirically validated ‘bun-
the underlying problems. The paper is based on case stud- dles’, labelled just-in-time (JIT), total quality management
ies in four manufacturing companies in the Netherlands. (TQM), total preventive maintenance (TPM) and human
It shows that, in these companies, the different concepts resource management (HRM). Particularly the first three
of control resulted in localised tensions. The paper investi- ‘bundles’ or elements are clearly defined in the literature,
gates the areas where these tensions arose, and how they as manufacturing programmes that aim at ‘continuously
were contained. It further uncovers how this local contain- reducing and ultimately eliminating all forms of waste’
ment of tensions enabled the companies to continue with (JIT), at ‘continuously improving and sustaining quality
lean production, even though the underlying causes were products and processes [. . .] in order to meet or exceed
not addressed. customer expectations’ (TQM), and at ‘maximiz[ing] equip-
The remainder of this paper is structured as follows. ment effectiveness throughout its entire life’ (TPM) (Cua
Section 2 introduces lean production, and it reviews the et al., 2001, pp. 676–677; see also Shah and Ward, 2003).
literature on its implications for management control sys- The HRM element is not so clearly defined, but various
tems. Section 3 describes the research method employed authors agree that the overarching theme is the empow-
for this paper. Subsequently, Section 4 provides a descrip- erment of (teams of) employees (Cua et al., 2001; Forza,
tion and analysis of our main case companies. Finally, 1996; Shah and Ward, 2003).
Sections 5 and 6 discuss our findings and present the con- It has been observed that control systems in lean pro-
clusions as well as some directions for future research. duction environments have characteristics which differ
from those traditionally used in mass production environ-
2. Lean production and management control ments. One crucial difference concerns the control of the
systems work force (Forza, 1996). Mass producers rely heavily on
breaking down the work into numerous narrowly defined
The spread of lean production across the world is one jobs which require minimal skills, and have a strict sep-
of the most notable developments in operations. The lean aration between those who plan and supervise and those
production system has its origins in Japan, where – in the who do the work. By contrast, lean producers have no strict
late 1940s – the Toyota Motor Company started develop- delineation of job responsibilities and encourage workers
ing a production system that later became known as one to participate in decision making; for example, in decisions
of the purest versions of lean production (Holweg, 2007; about continuously reducing waste and improving quality.
Womack et al., 1990, chapters 2 and 3; see also Forza, 1996; Examples of commonly cited lean control practices are for-
Hines et al., 2004; Katayama and Bennett, 1996). Basically, mal training programmes, job rotation programmes, work
this system was meant to deal with the poor economic teams and problem solving groups (Shah and Ward, 2003).
circumstances in Japan at that time, which required the Therefore, lean production implies that bureaucratic con-
production of a large variety of finished products, in low trols are replaced by clan controls (cf. Ouchi, 1979). Another
volumes, and at competitive costs. In this respect, the sys- key difference between lean production and mass pro-
tem broke with the tradition of mass production, which duction companies involves the control of the production
focused on using standard components, enabling the pro- flow (Benton and Shin, 1998; Forza, 1996). In particular,
duction of large batches, and thus minimising the number in lean companies the production flow is pulled through
of changeovers. However, despite its success at Toyota, the production process by the needs derived from actual
it was not until the mid-1980s – when Japanese imports market demand. This means that removing an ‘end item’
became a growing concern in Western Europe and the US triggers orders to produce or purchase earlier items. The
– that Western motor vehicle manufacturers became inter- demand for these earlier items is visualised by empty
ested in lean production. Since then – and particularly since Kanbans.1 In contrast, in traditional mass production the
the publication of ‘The Machine that Changed the World’
(Womack et al., 1990) – lean production has attracted con-
siderable attention, and is now regarded as one of the most 1
A Kanban is often an essential element of a pull system (Maskell et al.,
influential ‘new’ paradigms in production, both inside and 2012, p. 42). It functions as a buffer between the different production
outside the automotive industry. processes, and moving a Kanban triggers the production of new items. A
70 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

production flow is pushed downstream in accordance with actual cost system. Fullerton et al. (2013) confirm these
pre-determined plans. The overall production plan is trans- results for a broad sample of companies. Abernethy et al.
lated into a series of steps which ensure that materials (2001) also find that advanced manufacturing technolo-
and components are ordered and produced on time to gies can be linked to the use of simple accounting systems.
accomplish the planned production. Hence, lean produc- Based on field study evidence, they argue that the use of
tion replaces production planning systems by physical pull flexible manufacturing technologies can reduce the effec-
production systems which use decentralised information. tiveness of and the need for sophisticated costing systems,
Lean production has various implications for manage- as these technologies reduce the significance of batch or
ment accounting systems. Several researchers find that product-sustaining costs.
the implementation of lean production (or some of its Based on these insights, we differentiate the lean con-
elements) leads to changes in the performance measures cept of control from the traditional concept. The lean
which are used. For example, based on a survey of top concept is characterised by a strong emphasis on clan
US manufacturing executives, Fullerton and McWatters controls and low-level decision making; the use of decen-
(2002) find relationships between the implementation of tralised information; the provision of information to lower
JIT and the use of non-financial performance measures, levels; the distinctly non-financial nature of information;
such as measures of scrap, rework and setups. A simi- and a higher frequency of information dissemination. The
lar conclusion is drawn by Daniel and Reitsperger (1991), aforementioned studies demonstrate that lean production
based on a comparison of Japanese and US managers. In leads to various adjustments to control systems, includ-
addition, Ittner and Larcker (1995) find that, among com- ing a reduced emphasis on traditional accounting controls.
panies from the automobile and computer industries in However, they do not shed light on the appropriateness
four countries, TQM programmes are associated with a of Johnson and Kaplan’s (1987) suggestion that, following
greater use of non-traditional information systems; i.e., the implementation of new production technologies, com-
systems which generate and report strategic planning and panies should use different control systems for different
external benchmarking information, as well as quality and purposes, until one integrated system has been developed.
problem-solving information. Furthermore, in a case study Most of the studies discussed in this section either do not
of a manufacturing plant, Kennedy and Widener (2008) distinguish between the control systems used for different
find that the implementation of lean production leads to purposes, and hence between the systems used at different
the introduction and use of various operating performance levels (e.g. Fullerton and Wempe, 2009; Ittner and Larcker,
measures. 1995), or they concentrate on the systems used to control
Also, the frequency and dissemination of management processes at the operating level (e.g. Banker et al., 1993;
accounting information may change due to the implemen- Sim and Killough, 1998). Moreover, these studies focus on
tation of lean production. For instance, Sim and Killough the differences in control systems between lean produc-
(1998, pp. 327–328) observe that elements of lean pro- tion companies and other companies, while they ignore
duction ‘encourage the continuous flow of information to the similarities between them. As a consequence, we lack
employees who perform the task’, rather than the peri- insight into the role of traditional accounting-based con-
odic dissemination of information to specific managers trol systems within companies that use lean production. In
that is more common in mass production environments. addition, we do not have a clear understanding of the impli-
This observation is supported by empirical evidence found cations of the combined use of different control systems in
by Banker et al. (1993), Daniel and Reitsperger (1991), these companies.
and Kennedy and Widener (2008). The effects on the fre- Contrary to the suggestion by Johnson and Kaplan
quency and dissemination of accounting information can (1987), two studies have revealed that traditional control
be explained by the JIT and HRM elements of lean produc- systems can continue to play a significant role in control-
tion. The JIT element leads to a reduction of slack resources, ling the internal operations of lean companies, and that
which increases the need for coordination (Daniel and this may hinder the implementation of lean production.
Reitsperger, 1991, p. 607), while the HRM element means Åhlström and Karlsson (1996) find that, in the company
putting the control of production into the hands of the they studied, the decision to create flow-lines was post-
workers, who thus need to have appropriate information poned, mainly because productivity had decreased, and
(Banker et al., 1993, p. 37). board members were unsure of the effects of the change.
A final implication of lean production discussed in the This company measured productivity in a traditional way;
literature is that management accounting systems become namely, by comparing the actual number of hours spent
simpler. In their case study, Kennedy and Widener (2008) on operations against the standard number of hours. The
find several examples of ways of streamlining transaction authors conclude that: ‘Taken as a whole, the manage-
processing. These examples include accounting for labour ment accounting system has been a major impediment
as a period cost instead of allocating it to the products, and to the desired changes in the manufacturing strategy’
discontinuing standard costing procedures in favour of an (p. 50). A similar example is provided by Bamber and
Dale (2000), who describe how in their case company a
proposal to change the batching rules in line with lean pro-
duction was rejected. Although the proposal was widely
Kanban can take various forms (e.g. a container or a physical card denoting
the items), but it always has a fixed quantity of items associated with it.
supported within the company, the responsible industrial
This quantity reflects the number of items that are to be manufactured at engineer objected, as the proposed change would increase
any one time. the standard cost of the components involved. The authors
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 71

note that in this company ‘[it] proved difficult to break both companies, interviewees invited us for a site visit and
away from the traditional output-related performance and explained in detail the implications of lean production for
labour utilisation measures’ (p. 294). Operators were eval- everyday practice.
uated mainly on the basis of the number of hours taken Moreover, we enlisted the help of two research assis-
to build a product, despite lean production’s emphasis on tants. In spring 2010 one of these research assistants spent
encouraging workers to make improvements which may a period of 3 months in Pegasus NL to collect data, and
not be directly related to output. in autumn 2010 the other research assistant spent a sim-
Yet, despite the reported difficulties of implementing ilar amount of time in Gryphon NL. During their stay
lean in the context of traditional controls, many compa- in the companies, the research assistants had numerous
nies seem to be able to introduce lean production, including informal talks, and they attended meetings and collected
the lean concept of control, whilst continuing to apply the documents. In addition, they conducted interviews (15 in
traditional concept of control. In this paper, we argue that Pegasus NL and 11 in Gryphon NL) with persons at various
such companies are able to balance these different con- organisational levels. In the period of their stay in the case
cepts of control to the extent that lean production itself is company, we met the research assistants regularly to dis-
not challenged. Given lean production’s ambition to dra- cuss their preliminary findings and to provide them with
matically change the organisation and given its focus on further instructions. The research assistants provided us
perfection, lean implementations are long-term, on-going with recordings of their interviews and written interview
processes. In this time span, the ‘balance’ between the tra- reports, which we then analysed. Thus, although research
ditional and the lean concepts of control may be affected assistants collected part of the data, we analysed all the raw
on various occasions; for example, when new steps in the data (mostly the interview recordings and documentation)
implementation process are taken, or when providers of ourselves. All interviews by us and the research assistants,
finance become unsatisfied with the company’s financial and all the other meetings, are listed in the appendix.
performance. On such occasions, the question of what it The data collection was aimed at mapping the lean
means to be in control comes to the surface, and tensions controls used by the case companies, and exploring the
may arise. This paper aims to uncover how lean companies tensions that were experienced in the use of these lean con-
are able to contain such tensions to the extent that they trols in combination with the companies’ more traditional
avoid failures in the implementation of lean production. control systems. The interviews were also aimed at inves-
tigating the ways in which the companies tried to contain
3. Research method these tensions. We analysed the recordings by marking and
summarising sections which related to the different con-
In order to investigate the ways in which companies trols that were identified by the interviewees. In addition,
are able to contain the tensions that result from having we identified potential areas of tension between controls
multiple concepts of control, we conducted case study and we listed the various ways in which the interviewees
research in four Dutch manufacturing companies. These described how they and the case companies (had) dealt
companies all combined the use of lean production con- with them. To do so, we repeatedly revisited the relevant
trols with the use of a traditional control system. In this parts of the recordings. We used the outcomes of our anal-
paper, we focus primarily on the findings from two of these ysis to reconstruct the implementation of lean production
cases; i.e., Gryphon NL and Pegasus NL.2 Both companies in the two companies, the implications for management
are quite experienced with lean production. At the time of control, the tensions between different controls, and the
our research, Gryphon NL had fully implemented lean pro- containment of these tensions.
duction in one of its assembly lines, while Pegasus NL had To extend our analysis beyond the two main cases,
organised its main production processes according to the we conducted two additional case studies – i.e., in AP
lean philosophy. After a preliminary meeting with a lean Nederland and Swedish Match Assen. These companies
consultant, the empirical research started in March 2010 are interesting because their stage of implementing lean
with a 2-hour meeting with six persons from these two production differs from the main cases. In particular, AP
case companies and the lean consultant. The persons from Nederland is a relatively small privately owned business,
the case companies were all familiar with the control of which – at the time of the study – was in the early stages
lean operations, and they gave us an indication of the ten- of lean production. However, its managing director had
sions between different controls that they had experienced considerable experience in implementing lean production
since the introduction of lean production in their company. in a large multinational company. Swedish Match Assen
Subsequently, we interviewed persons holding key is a production facility of a large multinational corpora-
positions in these companies; i.e., managers and man- tion. Compared to the other cases, this company was at the
agement accountants who were involved in controlling most advanced stage of lean production at the time of the
the lean operations. We used in-depth unstructured inter- study. It had applied lean production to almost all parts of
views, which gave the interviewees the opportunity to its production process, and it was considering the imple-
talk freely about their experiences. The average length of mentation of lean in other parts of the company, including
our interviews was almost 2 h, and with the interviewees’ the finance department. By including this company in our
permission these interviews were digitally recorded. In study, we are able to draw conclusions about whether the
containment of tensions, as it is discussed in this paper, is
a temporary or a more permanent phenomenon. The two
2
Both companies requested us to use fictitious names. additional cases were studied in a similar fashion to the two
72 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

main cases, but based on a more restricted set of data (two industrial cost engineers had become the advocates of lean
interviews in AP Nederland and four in Swedish Match, production in Gryphon NL. One of the problems these lean
with an average length of 1:50 h). The findings from the advocates faced in the early days of the lean transforma-
additional case studies will not be discussed at length in tion, was that they found it difficult to assess the quality of
this paper; instead we will use them to extend and refine lean proposals. For this purpose, the industrial cost engi-
our arguments in the final section. neers tried to use traditional decision-making tools, which
As a final step, we organised a 4-hour meeting in focused on absorption costs and on the implications for
which representatives of the four case companies, and accounting profits. However, this was deemed problem-
also persons from other organisations that had experi- atic. The remarks below from a manager and an industrial
ence with lean production, discussed the relationships cost engineer, respectively, illustrate the frustrations of the
between lean production and management control. This lean advocates in Gryphon NL.
meeting enabled us to obtain feedback from practition-
‘[Considering particular lean production changes] soon
ers on the ideas we were developing based on our case
led us to start discussions, such as: “But wait a minute,
studies.
these machines come with a whole infrastructure; mov-
ing all that will cost millions of Euros, and what exactly
4. Case study findings will be the benefits?” These were quite heated techni-
cal discussions. So, for example, “Should we maintain a
This section presents the findings from the two main stable production volume, or should we follow the mar-
case studies: Gryphon NL and Pegasus NL. ket?” This type of discussions took place. And no matter
what kinds of calculations you made, making changes
4.1. Gryphon NL never paid off, so to speak. So these discussions did not
really help.’
Gryphon NL is a production site of a large multinational
‘Yes sure, we did that several times [i.e. try to calculate
company with headquarters in Western Europe. The pro-
the costs and benefits of implementing lean produc-
duction site is located in the Netherlands, and it operates
tion]. On one occasion, we decided to use one product
as a cost centre. With over 1500 employees, it devel-
family throughout the factory as some kind of pilot. At
ops and manufactures personal appliances. Gryphon NL
that time this seemed to be a good idea. The question
started experimenting with lean initiatives in the mid-
was then: what will it yield? We thought: “First we
1990s. However, the implementation of lean came to full
make a value stream, next a future state value stream,
fruition when corporate headquarters (HQ) decided to
and then we can exactly determine how much it will
introduce lean production companywide, and it selected
yield.” But forget it. Of course this did not work out. It is
Gryphon NL as one of its pilot plants. The lean implementa-
simply not possible. It goes wrong because of the huge
tion in Gryphon NL was characterised by much attention for
quantities of data that you need; actually, it is simply
both the technical tools of lean production as well as a ‘lean
impossible to foresee all these data.’
vision’, which placed considerable emphasis on changing
the organisational culture and people’s behaviours. Cen- These quotes illustrate that in the early days of the
tral to Gryphon NL’s lean production system were twelve implementation process, the lean advocates thought that
‘paradigms’ which had been developed to redirect peo- being in control required traditional decision-making tools
ple’s behaviours; examples were: ‘focus on the longer run’ to give a favourable assessment of the initiatives that were
(rather than ‘focus on short-term financial performance’), to be taken. As attempts to apply these tools to lean initia-
‘produce on customer demand’ (instead of ‘get the product tives were not very successful, the implementation of lean
out’; i.e., out of the factory), and ‘take a look at the shop production was surrounded by much discussion, which
floor’ (rather than ‘go to the online dashboard’; i.e., the ERP hindered the implementation process.
system). At the time of our research, lean production had Nevertheless, the lean advocates in Gryphon NL were
been fully implemented in one of the plant’s assembly lines. convinced of the benefits of lean production. To keep the
All the operators who worked on this line had been trained implementation process going, they were increasingly flex-
in the lean paradigms and were involved in developing ible in their application of the traditional and the lean
lean initiatives. Other parts of the plant were expected to concepts of control. They increasingly dealt with their
implement lean production in the near future. There was inability to apply traditional decision-making tools to lean
considerable support for the lean implementation among initiatives by considering lean criteria (e.g. floor space, flex-
the managers in Gryphon NL. ibility and lead times), in addition to the more traditional
In Gryphon NL, one of the functions involved in the financial criteria (e.g. net present value), when making
company’s lean transformation was the industrial cost capital budgeting decisions. Although the managers were
engineers. Their role had been to set the standards to held accountable for the (financial) performance of their
be included in the company’s standard costing system. unit, both a financial controller and a manager pointed out
However, since the company began implementing lean that the managers’ decisions were increasingly informed
production, the industrial cost engineers had participated – not by traditional financial measures of performance –
in various lean initiatives, such as product development but by the company’s lean paradigms. A manager explained
and process improvements. Together with the managers that whenever a decision had to be taken (whether at the
who were directly involved in the lean transformation, the operational or the strategic level) managers asked each
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 73

other whether the proposed action was consistent with the higher levels in the company and the business unit manage-
behaviour that they wanted to show. ment were still using a traditional budgeting system, with
However, in contrast to the lean advocates in Gryphon an emphasis on standard cost variances. In Gryphon NL, one
NL, HQ did not change its view of what it means to be in con- important set of financial performance measures were the
trol. Despite its support for lean production, it continued to so-called ‘efficiency’ measures, which measured – for both
focus on absorption costs and accounting profits. Their dif- direct labour and direct materials – the difference between
ferent views of control gave rise to tensions between the the standard cost allowed and the actual cost. These ‘effi-
advocates of lean production in Gryphon NL and the finan- ciency’ measures were discussed in the management teams
cial control department at HQ; the financial controllers of all production departments on a weekly basis, and were
at the site level had an intermediate position. One of the reported to the business unit management every month.
examples provided by an industrial cost engineer involved Both the production managers and the shift leaders were
decisions about where particular products should be pack- held accountable for their unit’s scores on these measures.
aged. He pointed out that the financial controllers wanted A financial controller in Gryphon NL explained that the
to take such decisions based on cost price considerations. company’s use of ‘efficiency’ measures was closely related
He argued that such considerations do not take into account to the business unit management’s emphasis on ‘conver-
the risks that are involved in transporting products. Fol- sion costs’.3 He observed that, although the business unit
lowing the lean philosophy, he felt much more in control management required all production sites to use a Bal-
when all the production stages (including packaging) were anced Scorecard for their monthly reports, by far the most
located close to each other, as this would increase the pre- attention was paid to conversion costs.
dictability of the overall process. He also gave the example Also two industrial cost engineers and their manager
of a proposal to adopt a production process which involved understood the need to use financial performance meas-
higher labour intensity but would increase the company’s ures at the higher levels, as the parent company is listed
ability to respond to changes in customer demand. This on the stock exchange and accountable to its shareholders.
proposal was opposed by the financial controllers with However, they explained that it was difficult to reconcile
comments that, in terms of labour costs, ‘the Netherlands the measures of operating performance used at the shop
are not China’. floor level with the measures of financial performance used
Both examples reveal that whereas the lean advocates at the higher levels. They observed that on various occa-
in Gryphon NL stressed predictability and flexibility of sions, the two sets of performance measures suggested
operating processes, the financial controllers at HQ empha- different – sometimes even opposite – actions. As one of
sised traditional accounting concepts. In order to ensure the industrial cost engineers noted:
that future lean initiatives would be approved more eas-
‘Using financial performance measures at the shop floor
ily by the controllers, the lean advocates continued their
level is likely to lead to undesirable behaviour. [. . .] It
earlier attempts to find ways to calculate – in traditional
may induce people to start thinking in terms of effi-
accounting terms – the effects of, for example, increased
ciency again, which will keep them from making that
flexibility or shorter lead times. They wondered how cost
additional changeover altogether.’
prices should be determined and business cases should be
analysed in a context of lean production. With some hesi- And:
tation because of the negative connotation of the words, a
‘What you actually would want, is that you measure,
manager said that they tried ‘to make the numbers come
measure and measure, and then–somewhere [in the
out right’. This way of containing the tensions due to dif-
organisational hierarchy] translate the outcome into
ferent criteria to assess the quality of lean proposals can be
Euros. If this were the case, I would not have so many
seen as attempting to reach local compromises. The lean
problems with using Euros at the lower levels. But what
advocates kept using their broader concept of control, but
you often see, is that at the level where the translation
– in order to convince the financial controllers at HQ – they
into Euros is made, some of the data suddenly comes
also tried to adjust the traditional decision-making tools in
from a different source. If this happens, the coherence
such a way that they would produce similar judgements
is gone.’
about lean proposals.
Tensions also arose due to inconsistencies between per- In the early days of Gryphon’s implementation of
formance measures. As part of the implementation of lean lean, the different suggestions for action led to ‘heated’
production, Gryphon NL had reduced the emphasis on discussions between the proponents and opponents of
financial performance measures at the lower levels of its lean production. However, some of the advocates of lean
organisation, including the shop floor level. Instead, it con- observed that since their company had adopted a broader
centrated on measures of operating performance, such as perspective on performance in recent years, it had become
the output per shift, the ratio of actual and planned produc- easier to follow the signals provided by the lean perfor-
tion orders, defect rates and work-in-progress levels. An mance measures. For example, they reported that there
industrial cost engineer indicated that a great advantage of were no longer any discussions about the need to continue
such operating performance measures is that – in contrast
to financial performance measures – they can be influenced
directly by the operators themselves. The scores on some 3
‘Conversion costs’ refers to the sum of the direct and indirect labour
of these measures were displayed on screens in the pro- cost and the other costs of organisation; it does not include the cost of
duction halls, which were updated every hour. However, materials.
74 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

production when there is no further demand. Although A final tension in Gryphon NL related to the higher
‘efficiency’ (in Gryphon defined in terms of standard cost organisational levels’ reliance on the ERP system to control
variances) was still an important issue in the weekly and the company. As part of the implementation of lean pro-
monthly discussions about the reports prepared by the duction, Gryphon NL had replaced the ERP system as the
financial control department, it no longer had a signifi- basis for planning and recording the production of indi-
cant impact on the operating processes. In the words of a vidual components with a Kanban system. However, as a
financial controller, the production manager had ‘his own result, its ERP system no longer provided real-time infor-
management structure’ to monitor the operating processes mation about the company’s work-in-progress inventories,
and to decide on actions. The financial controllers did not and this meant that the measurement of financial perfor-
provide information to support this structure. As a conse- mance at a particular point in time could be inaccurate. In
quence, the production managers were able to follow their the words of a manager:
own view of control, which was mostly based on meas-
‘In the past, we had plans for finished products, for plas-
ures of operating performance. Hence, potential tensions
tic components, for everything. That is no longer the
due to the reliance on different sets of performance meas-
case: we now only have a plan for finished products. As
ures were contained by decoupling the financial control
a result, the system no longer shows whether a buffer is
department from the operating processes.
already filled. So [based on the system] I can say that I am
Furthermore, the lean advocates we interviewed in
expecting a [Dxxx] loss this month–and it is about large
Gryphon NL expressed a desire for accounting information
numbers–but this might not be accurate, because there
which was consistent with the company’s lean operat-
are still many components stored in the buffer. But the
ing system. They argued that the accounting information
opposite might also be true. So you simply don’t know.
currently available at the higher levels evoked discuss-
There is a range [of possible outcomes for our finan-
ions about topics (including ‘efficiency’) which they no
cial results]. Is that problematic? The financial control
longer considered relevant for controlling the company,
department has the feeling that it is problematic. They
and which – in their view – could lead to ‘undesirable
say: “We can’t give a precise answer, and now you are
behaviour’. Also the financial control department at the site
telling us that there is still a range of 10 to 15%! That
level had questions about the set of performance measures
cannot be true. We can’t explain this to [the business
to be used. A financial controller wondered which perfor-
unit management]. We need to know the exact figures.”
mance measures could be used to control value streams,
So you lose some transparency if at any given moment
and how these measures could be integrated in the reports
you want to take the organisation’s temperature.’
required by HQ.
In a further attempt to contain tensions as a result of People from the financial control department felt that
inconsistent performance measures, Gryphon NL searched they were no longer fully in control. They considered the
for performance measures which could be used at the company’s lack of insight into its financial situation at
higher levels, but which were also consistent with the any point in time to be a risk. By contrast, the advocates
measures used to control the operating processes. During of lean production we interviewed were of the opinion
our visit, we were shown a room with one wall covered by that the ERP system was not needed for controlling pro-
sheets of paper showing the performance measures cur- duction processes, as the actual situation could simply be
rently used at the different organisational levels. It was assessed by looking at the components and products stored
pointed out that there were various problems in these on the shop floor. They regarded much of the resources
measures. For instance, some measures were used on one that were needed to keep a detailed administrative plan-
level, but were not linked to the measures used at the next ning system up-to-date as ‘waste’, and had doubts about
higher level. Furthermore, several measures focused on the accuracy of such a system. At the same time, the lean
‘efficiency’ (and hence stimulated production of standard advocates sympathised with the financial control depart-
volumes in order to cover the fixed costs), whereas other ment’s concerns. They acknowledged that the controllers
measures were linked to the pull production system (which did not know how to assess the situation on the shop
implied that production volumes should not to exceed floor. Moreover, they realised that the controllers were
market demand). By redesigning the sets of performance accountable to higher levels in the Gryphon organisa-
measures, the lean advocates wanted to shape the view tion. The lean advocates suggested that they would like
of being in control held by persons at the higher levels to simplify the accounting system, but also indicated that
in a way that was more consistent with their own view within the company this was a difficult subject. At the
of being in control. This can be seen as a form of coloni- time of the interviews, it seemed that compromise was
sation, where control systems are purposefully adjusted used to contain this tension due to the use of different
to create a misfit with the existing view of control. It is planning systems: a further simplification of the account-
assumed that, over time, an alternative view of control ing system was not planned, and weekly physical counts
will emerge which is more consistent with the new con- of work-in-progress were used for the periodic financial
trol systems (cf. Laughlin, 1987, p. 495). The lean advocates statements.
in Gryphon NL expected that if the traditional performance The above shows that the lean concept of control had
measures available at the higher levels were replaced with gained momentum in Gryphon NL in recent years. More
lean performance measures, people at these levels would and more, people were adopting a view of control which
no longer argue along the lines of the traditional concept of was consistent with this concept. However, some groups
control. – including HQ and financial controllers at various levels
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 75

– continued to rely on tools which reflect the traditional implemented in the plant in 2008. Through a series of infor-
concept of control. To the extent that the two concepts mational sessions, employees were trained in the main
of control were used in distinct domains (e.g. control- ideas of lean production. These sessions included so-called
ling operating processes and financial reporting to higher open space meetings and market meetings, which featured
levels), they could co-exist. However, on some occasions, a variety of stands, including a VSM stand and a 5S stand.4
localised tensions arose at the interface between particular At the time of our research, the implementation of lean
domains; for example, when initiatives in the produc- production was limited mostly to production activities. The
tion domain required the approval of, or support from, main production processes were organised according to the
the financial controllers and/or managers at higher lev- principles of lean production, and this had led to a signifi-
els, or when real-time reliable financial reporting required cant decrease in lead times and a reduction in inventories.
detailed planning and recording of transactions at the oper- Lean production in Pegasus NL is based on the so-
ating level. Several local attempts had been, and were called ‘lean house’. This image reflects a foundation of
being, made to contain such tensions, and to allow lean continuous improvement; pillars based on a variety of pro-
production (and its related concept of control) to expand duction techniques, such as Just-in-Time and VSM; and a
into new areas within the company. roof which reflects improving costs, quality and delivery
times. However, despite the introduction of lean pro-
4.2. Pegasus NL duction, Pegasus NL remained a rather complex matrix
organisation. Although the facility could develop its own
Pegasus is a high-end producer of home fixtures. It strategic direction, departments were also accountable to
employs more than 8000 people worldwide. The com- managers at HQ. For example, the finance department was
pany’s HQ is located in Western Europe, but it has accountable to the manager of Finance located at HQ. The
production and sales facilities located around the world, tensions, which were associated with the introduction of
particularly in Europe, Asia and the United States. In 2010, lean, were mostly related to this organisational structure.
Pegasus reported an EBIT loss of around D50 million, which They emerged in different forms and they were contained
was a significant reduction of the loss in the previous year. in a variety of ways. These tensions were mostly related to
One of the company’s production facilities is Pegasus NL, in the differences between the lean view of being in control
the Netherlands. This production facility employs around espoused by the local organisation and the more traditional
140 people, although the company uses a so-called ‘flexible view held by HQ.
shell’ (i.e. a peripheral workforce) to deal with variations in The decision to embrace lean production was taken at
production volumes. Pegasus NL is a cost centre; its prod- the local level and HQ was initially unaware of this decision.
ucts are sold internally at pre-determined standard cost The decision was not based on financial calculations of costs
prices. and benefits, but rather on the belief that lean would make
The recent history of Pegasus NL has been turbulent. In the company more efficient and responsive. A manager
2006 parts of the production facility were moved to low- explained that in the past the company was segmented,
wage countries. A resulting reorganisation of Pegasus NL and that this impeded people’s ability to cooperate. There-
led to a reduction in employment of more than 20%. In fore, the implementation of lean production was aimed
addition, the facility was threatened with closure by HQ. at improving interdepartmental cooperation. However, the
After several protests, an agreement was put in place which benefits of these improvements could not be expressed in
ensured that the facility would remain open in the com- financial terms. The manager explained:
ing years. The management of Pegasus NL intended to use
‘Only when we are all aware that the organisation can
this timeframe to ensure that the plant became more effi-
proceed only when we cooperate and we offer the struc-
cient and thereby able to compete with Eastern European
tures to make that happen [. . .], and we succeed in
and Asian production facilities. They saw lean production
serving our customers faster than we used to, then
as a promising way to enhance the efficiency of the plant’s
we need to talk to one another. When this intrinsic
production processes. A lean consultant in this company
motivation emerges, there may still be some degree of
explained:
segmentation, but this is no longer an impediment to
‘Lean must become our reason to exist. We have now cooperation, because we have created a collective belief
140 people working here and there is a clearly iden- in lean.’
tifiable reason to change our organisation, as this
This argument was not expected to be acceptable to the
production facility is under pressure. In a sense, it is our
decision-makers at HQ as it represented a large degree of
answer to a changing world.’
‘belief’ in lean production, with little financial underpin-
The introduction of lean production in Pegasus NL began nings. Therefore, Pegasus NL started implementing lean
in 2007 with discussions in the management team and an production without informing HQ. In this way, it tried to
inquiry into the benefits of lean. Although there was quite avoid tensions due to its initial inability to quantify the
a sense of urgency, it took some time for the managers benefits of lean production. To this end, costs which were
to understand the benefits of lean production. However,
when a lean consultant became involved in the company
as an interim manager, the lean efforts gained traction. 4
VSM is an acronym for Value Stream Mapping. 5S refers to the five
Following the initial inquiry into the benefits of lean and workplace organisation principles of lean: sorting, straightening, shining,
the subsequent managerial approval, lean production was standardising and sustaining.
76 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

associated with lean production (for example, the expenses changeover times and takt times5 of processes. These
of a lean coordinator) were reported generically as quality performance measures can be used independently from
costs. This ‘obscuring’ of the specific costs of lean produc- financial reporting. However, traditional performance
tion enabled Pegasus NL to postpone the tensions that were measures and value streams both represented the per-
expected between the facility level and HQ as a result of formance of similar operations, but they often conveyed
their different conceptions of control. The facility bought conflicting messages. As a consequence, the management
additional time to obtain some of the financial benefits of accountants in Pegasus NL could not provide translations
lean production which could eventually convince the man- to HQ as they received ambiguous signals from the differ-
agers at HQ. ent control systems. VSM was therefore a potential source
When HQ became aware of the lean production efforts of tension between Pegasus NL and HQ.
in Pegasus NL, they initially allowed it as a local initia- In attempts to contain such tensions between Pegasus
tive, but they did not change their requirements of the NL and HQ, lean proponents from operating departments
facility, which they continued to treat as a standard cost undertook various lean initiatives (including VSM and 5S)
centre. These requirements included the provision of a very as ‘grassroots’ initiatives. Since there was little budget
detailed budget and an elaborate structure of authorisa- available and the facility was unable to convince HQ about
tions and financial reporting. This meant that the managers the financial benefits, the facility approached these initia-
and accountants in Pegasus NL needed to translate the ben- tives in a cautious and financially conservative manner.
efits of lean improvement programmes into financial terms This resulted in a slow and low-cost localised introduction
to convince upper management at HQ of their value. As and an incremental implementation of lean in the plant.
such, the accountants served as ‘translators’ of lean pro- In Pegasus, the difficulties of bridging the lean and
duction and the traditional control system used elsewhere. the traditional concepts of control were mostly limited to
On occasions, they were able to translate the benefits of the relationship between Pegasus NL and HQ. During the
lean production into benefits recognised by the traditional reorganisation in 2006, HQ had imposed a matrix struc-
control system. For example, the facility was evaluated ture on the facility. This meant that various departments,
by HQ using standard costing. Its main challenge was to including the logistics and the procurement departments,
be able to cover the fixed costs of its expensive capi- were managed from HQ. According to a manager, this
tal goods. As part of the introduction of lean, parts of matrix structure led to isolated ‘vertical pillars’ running
the production were organised into a built-to-order pro- through the organisation, and the emergence of ‘islands’
duction system. This system led to increased production which greatly reduced cohesion in the facility. Although the
capacity which enabled the plant to cover its fixed cost manager acknowledged that this way of organising could
more easily in periods of sufficient demand. These were be beneficial, it was designed to maximise financial per-
advantages that were measurable by the traditional con- formance within the ‘vertical pillars’, rather than to add
trol system and to which the managers at HQ were quite customer value (e.g. by improving quality and reducing
receptive. inventories) which would be consistent with the lean phi-
However, in other instances, tensions arose as the losophy. He illustrated this with the following example:
process improvements could not lead to improved finan-
‘The procurement department here is instructed by the
cial results in themselves, but rather to improvements
managers at HQ to save so-and-so many millions of
in the quality of production processes. Such improve-
Euros on the purchasing of inventories. As a conse-
ments could not easily be measured in traditional financial
quence, our production department may get supplied
terms. As a result, HQ did not approve the associated
with inferior parts. In addition, we may get flooded with
expenses and investments. An example of benefits which
parts, which are shipped from China in such quantities
could not be quantified in traditional terms was that,
that we have no space to store them.’
during periods of low demand, production staff were
put to work on preventative maintenance and tasks In such situations, the cost savings were not necessarily
related to the 5S method of lean, such as cleaning the beneficial from a lean perspective, as the traditional con-
machinery. Although these activities contributed to the trol system emphasised the lower purchasing price of raw
quality of the production processes, the benefits were not materials while ignoring the higher costs of quality and
measurable in financial terms. Traditional performance storage. Although the latter costs were real for Pegasus NL,
measures actually deteriorated in such periods, as standard they were very difficult to quantify. In these circumstances,
costs were not covered due to the lower production the local managers were unable to challenge the decisions
volumes. made by HQ.
Another example was the Value Stream Map. In line In a related fashion, the managers at HQ frequently
with the traditional concept of control, Pegasus NL was made decisions which were based on the traditional
used to reporting performance based on ‘traditional’ accounting-based concept of control, which resulted in
departments. However, at the time of the interviews, tensions between the two views of control. For exam-
it had explored the use of value streams to control its ple, in 2009 and 2010, the demand for Pegasus’s products
processes and to evaluate the performance of products
and processes. These explorations led to the realisation
that hierarchical structures and value streams were not 5
Takt times are the cycle times of production that enable the company
consistent as a basis for control. Value Stream Maps depict to meet customer demand. They are calculated by dividing the available
measures of operating performance, such as cycle times, production time by the required products to meet demand.
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 77

decreased sharply as a result of the recession in Europe, illustrated above, this led to a variety of localised tensions.
and this prompted a local manager to call for a reduction Ever since the introduction of lean production in Pega-
in production volumes. During an interview, he noted: sus NL, the two concepts of control have co-existed. In
many instances, the operating controls of lean production
‘Lean is essentially a pull process. That means that if we
and the financially oriented traditional controls occupied
experience a drop in demand, we produce lower quan-
distinct domains, and tensions occurred mostly at the inter-
tities. For, the waste which is involved with having large
faces between those domains. In case of Pegasus, this was
inventories must be avoided.’
mainly the interface between Pegasus NL and HQ. The
However, a reduction of production volumes would lead source of the tensions seemed to be the stringent empha-
to unfavourable variances, due to the lower contribution sis on traditional concepts of costs at HQ, and the inability
to fixed costs. This was not acceptable for the managers at of Pegasus NL to quantify the benefits of lean production.
HQ, who did not see a need for lower production volumes. However, despite the often-documented high failure rate of
They asked the facility to maintain production levels, and lean implementations, these tensions did not threaten the
to store the surplus products. To the managers at HQ, the existence of the lean operations in the plant. Nevertheless,
unfavourable variances signalled that the facility was out these operations were affected, particularly because the
of control, because they indicated unused capacity. How- lean philosophy could not be applied on all occasions. The
ever, to the lean-oriented managers in Pegasus NL, these tensions between the two concepts of control were con-
variances were largely irrelevant and did not indicate any tained through localised coping. If the continued existence
degree of control. Nevertheless, an employee of the finance of lean production is a measure of success, these coping
department at the facility level explained that attempting mechanisms were successful as lean production is still the
to improve traditional performance measures is sometimes dominant production philosophy in Pegasus NL.
tempting, although it can be at the expense of the prin-
ciples of the lean philosophy, such as the zero-inventory 5. Discussion
principle:
In this section, we will discuss the implications of lean
‘By increasing production we can perform better than
production for the different views of being in control in the
the budget. The result is that you end up with invento-
case companies, and how these companies were able to
ries, but it looks very favourable on the cost calculation.
contain the tensions that arose at their interfaces. We will
So when there is pressure for better results it is very
also discuss the localised and temporary nature of most of
tempting to say: we will increase production.’
the resolutions of the problems that arose at the interfaces
In such situations, the managers in Pegasus NL had between traditional and lean controls.
no choice but to comply with the directives from HQ.
Although this frustrated the degree to which they saw 5.1. Different views of being in control
themselves as ‘lean’, these instances were mostly isolated
events. This meant that the managers occasionally had In Gryphon and Pegasus, the introduction of lean pro-
to make decisions which were not ‘lean’ to contain ten- duction brought with it a new concept of control, which
sions between them and other parties; mostly HQ, but was not consistent with the concept of control that was
sometimes local management accountants. In doing so, the enacted through these companies’ traditional control sys-
managers seemed to temporarily suspend the lean concept tems. This was also observed in the other two companies
of control in favour of the traditional concept of control. we studied: AP Nederland and Swedish Match. Despite
In this way, they shielded the lean transformation process the introduction of the lean concept of control, the role
from outside interference, at the expense of a sub-optimal of the case companies’ traditional control systems was
implementation of lean. Through this cautious incremental not limited to external financial reporting, as suggested
implementation, they kept it a local affair. by Johnson and Kaplan (1987). Instead, the traditional
However, on other occasions, the local managers made accounting-based concept of control continued to play a
attempts to introduce managers from HQ to the bene- role in controlling internal operations. Some interviewees
fits of lean production. For example, a corporate manager explained that the presence of shareholders and other
in charge of procurement was invited to the facility, providers of finance stimulated the use of this concept.
where, through various guided tours, the local managers Others emphasised the powerful position of financial con-
attempted to educate her in the benefits and the tools trollers and auditors, who were trained in a traditional way.
of lean production. Moreover, the facility developed a To these controllers and auditors, the lean concept of con-
Balanced Scorecard combining operating and financial trol was often akin to the loss of control. In many instances
information. As more people became aware of this Bal- the two concepts of control were able to co-exist as deci-
anced Scorecard, it was sent to HQ, where it generated sion makers could select the concept which best suited
considerable interest. This process of ‘colonisation’ was their current tasks. We found that tensions existed pri-
used to pro-actively alter others’ understanding of when marily when the two concepts of control came together;
the company was in control and the consequences thereof. for example, when people needed to communicate the
As such, it aimed at reducing the tensions which could arise benefits of lean production to other parties who used the
from the different conceptions of control. traditional concept of control. In spite of differences in the
In Pegasus, the presence of two distinct concepts of implementation stage of lean production, all four cases
control resulted in different views of being in control. As had experienced such tensions. This suggests that tensions
78 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

between the two concepts of control are not limited to the variances) in the future. Swedish Match’s local response
earlier stages of an implementation, but can persist for a to the opposition of the internal auditors from HQ was
longer period of time. also a form of colonising. In order to convince the audi-
Due to the tensions, the question of what it means to tors that the company had developed a good alternative
be in control could arise at any time. For instance, the to recording inventories in the ERP system, employees at
unfavourable variances at Pegasus NL were an indication the facility level invited the internal auditors to the fac-
to HQ that their processes were not in control. However, in tory, and explained how they had organised the production
the view of the facility’s managers, ‘being in control’ was processes. In addition, a senior financial manager of the
not a matter of avoiding unfavourable variances by min- production facility presented them with historical data
imising unused production capacity, but of being able to which showed that the company was better able to pre-
reduce inventories. The meaning of ‘being in control’ was dict its financial results than when it used the traditional
also discussed in Gryphon. In that company, the financial internal controls favoured by the auditors. In Pegasus NL,
controllers argued that a company can only be in control people attempted to spread the understanding of lean con-
if its ERP system provides accurate information about its trols in a similar fashion. Managers from HQ were given
financial performance on a real-time basis. In contrast, the factory tours in an attempt to convey the benefits of lean
advocates of lean production in Gryphon regarded a com- controls. Additionally, the Balanced Scorecard, which was
pany as ‘in control’ when its operating processes are stable, used in the facility, was shared with other units, including
which – in their view – can be assessed at any time by HQ.
examining the screens in the production halls and look-
ing at the inventories on the shop floor. A similar situation 5.2.2. Decoupling
was observed in Swedish Match when the internal auditors Other tensions were contained by separating the differ-
at HQ objected to plans to minimise inventory counting ent controls for different purposes. Tools consistent with
activities. Interviewees in this company argued that these the lean concept of control were used to control the oper-
activities had become redundant, and thus wasteful, since ating processes, while traditional accounting tools were
the company had replaced control based on the ERP sys- used for financial reporting. For instance, in Gryphon the
tem with control based on an orderly flow of products traditional measures of ‘efficiency’ (which were based on
through the factory combined with visual controls of this variance analysis) did not have a direct impact on actions
flow. However, whereas the interviewees at the facility taken at the operating level. The production manager had
level were convinced that their company was now much ‘his own management structure’, which did not involve
more ‘in control’ than in the past, the auditors did not agree. the financial controllers. A similar situation was found in
In AP Nederland, ‘being in control’ was not a big issue at Swedish Match. A senior financial manager in this com-
the time of study, as the owners (the Van Dool family) had pany explained that he first produced monthly financial
a long-term focus and believed that lean production would statements for the local organisation which did not rely on
be beneficial for the company in the longer term. However, standard costs, and then – as a final step to satisfy the infor-
senior management was aware that operational excellence mation requirements of HQ – he transformed them into
would eventually need to translate into financial results. statements based on standard costs and variance analyses.
A senior manager of Pegasus NL indicated that traditional
5.2. Containment of tensions accounting concepts (e.g. direct labour and direct materi-
als) were important – for instance, to convince HQ that the
In contrast to the substantial failure rates of lean imple- production facility’s performance is satisfactory – but these
mentations documented in the literature, all four case concepts were not applied at operating levels.
companies were able to continue their lean implementa-
tions despite these tensions. Although some interviewees 5.2.3. Compromising
expressed a desire for a single management control sys- Some of the tensions were contained by the mecha-
tem, the case companies coped with the tensions created nism of compromising, which involved the use of a mix
by the potentially conflicting concepts of control in highly of lean and traditional controls. The traditional controls
localised ways. There were no high-level strategies or all- were mostly retained, even though they were no longer
encompassing attempts to integrate the controls into a considered relevant or useful by everyone. This approach
single system, but instead the employees, managers and was used in Gryphon and Swedish Match to contain ten-
accountants coped with their divergent views of being in sions arising from the recording of inventories in the ERP
control in a variety of different ways which we would clas- system. To avoid conflicts between the lean advocates and
sify in the following manner. the financial control department, Gryphon continued to
make weekly counts of work-in-progress inventories to
5.2.1. Colonising ensure that the financial systems remained up-to-date. In
Some of the tensions were contained through attempts contrast, Swedish Match reduced its inventory counting
to influence the others’ view of being in control. We refer activities to a minimum, but to resolve the conflicts with
to this mechanism as colonising. In Gryphon, the lean the internal auditors, it introduced some ‘compensating
advocates were attempting to develop sets of performance controls’, including monthly inventory counts of the most
measures which could be used at the different levels and valuable components together with analyses of the month-
which were consistent with lean production. Their aim was to-month changes in these inventories. Furthermore, the
to avoid discussions about ‘efficiency’ (i.e. standard cost attempts in Gryphon to include the lean principles in cost
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 79

calculations and business cases were also examples of com- replace parts of their traditional (ERP) systems with lean
promising. controls due to concerns about the reliability of financial
information. Table 1 gives an overview of the mechanisms
5.2.4. Implementing incrementally which the case companies used in relation to each of the
An alternative way of containing the tensions created three problems. Below, we will reflect on the conditions
by the different concepts of control was by implementing under which the mechanisms can be applied, and on their
lean production incrementally. This means that a series of suitability to resolve the underlying problems.
small initiatives were taken, through which lean expanded, The investment-related problem occurred because the
but within the boundaries of the lean advocates’ mandate lean proponents in the case companies had not been able
(e.g. budget and/or decision-making authority). In Pega- to develop decision-making tools which could capture
sus, the initiatives to introduce lean production in various the benefits of proposed lean initiatives. As a conse-
domains were very localised. Despite being a production quence, these initiatives could only be ‘justified by faith
facility evaluated on the basis of absorption of costs, the alone’ (cf. Kaplan, 1986). At the operating levels, the faith
local managers were constantly looking for small-scale in lean initiatives was strong, especially as local people
improvements based on the principles of lean production. were impressed by the improvements in operating perfor-
These improvements included better work-place organisa- mance since the introduction of lean. However, large-scale
tion and grass-roots process changes, initiated by engineers investments also required the approval of higher levels.
and operators. Therefore, decoupling was not a viable mechanism for this
type of investments. Furthermore, our cases illustrated that
5.2.5. Obscuring colonising was problematic in settings where the highest
A final way of containing tensions occurred when lean organisational levels lacked a long-term focus (which was
initiatives were taken by obscuring them from persons who obviously present in the family-owned AP Nederland). In
might be expected to object. Both Pegasus and Swedish order to gain faith in future lean initiatives, people at higher
Match started implementing lean production locally, with- levels often demanded ‘evidence’ of the benefits of past ini-
out informing HQ. For example, in Pegasus costs of lean tiatives. Consistent with their traditional view of control,
initiatives were recorded under generic headings, which this evidence had to demonstrate that these initiatives had
obscured their exact purpose. had a positive impact on financial performance. The lean
In the aforementioned list, the ways of containing ten- proponents in both Gryphon and Pegasus were not able
sions are ranked based on the degree of support for lean to provide this type of evidence. They noted that financial
initiatives required from higher levels. Colonising involves performance could only improve if production processes
the highest degree of support. This mechanism relies on were made more flexible, which would ultimately result
a shared view of the appropriateness of lean controls in lower fixed costs. However, as this increased flexibil-
between the operating level, from which the lean initiatives ity required large investments in production technology,
usually emerge, and the higher levels. With decoupling, they observed a vicious circle of discrimination against lean
the lower levels are given the freedom to deal with lean investments. Only Swedish Match had been able to achieve
initiatives in their own way, and they remain relatively substantial improvements in traditional financial terms;
insulated from traditional controls. By contrast, compro- for instance, its direct labour cost per unit had decreased
mising requires operating managers to work with lean by more than one-third. However, as this company was not
controls and the traditional controls that are deemed planning investments at the time of our research, we were
important to the higher levels. Implementing incremen- not able to examine how this impacted on HQ’s support for
tally means that the lower levels are only able to introduce investments in lean.
smaller initiatives in lean. Finally, obscuring does not Because of these problems with colonising and decou-
require higher-level support, as these levels are not even pling, the proponents of lean in Gryphon and Pegasus, and
aware of the lean initiatives. From the point of view of lean in an earlier stage also Swedish Match, had to rely on the
proponents, colonising is most desirable as it expands the three mechanisms on the right-hand side of Table 1 for their
area where people share the lean view of control, and as investments in lean. The events in Gryphon demonstrated
such makes it easier to implement lean initiatives. that compromising is possible, but only if higher levels in
the organisational hierarchy support the implementation
5.3. Resolution of the underlying problems of lean to some degree. It requires higher organisational
levels to be prepared to sacrifice benefits in traditional
Although the five mechanisms enabled the case com- terms for benefits in terms of lean. The other ways to con-
panies to contain tensions between the two concepts of tain tensions are theoretically not suitable to support major
control, they did not necessarily resolve the underlying investments in lean production (as higher hierarchical lev-
problems. More specifically, at the interfaces between els need to authorise large investments), but they can be
the two concepts of control, the companies experienced applied to smaller investments. The benefit of an incremen-
three problems. Firstly, they found it difficult to justify tal implementation of lean – which was applied by Pegasus
their investments in lean production, both before and – is that, through a series of small steps, evidence of the
after the investments had been made. Secondly, they benefits of lean can be obtained and shared. As a result,
were not fully able to apply the lean principles to their more support from senior managers can be obtained, thus
operating decisions due to an emphasis on covering fixed opening the door to other mechanisms in Table 1. The same
costs. Finally, they were hindered in their attempts to could be true for obscuring. Although this mechanism may
80 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

Table 1
Problem areas and ways of containing tensions found in the case companies.

Colonising Decoupling Compromising Implementing incrementally Obscuring

Investments x x x
Operations x x
Financial information x x

not seem legitimate as it ignores the formal lines of author- system ‘more lean’ were frustrated by controllers and
ity, the evidence in Pegasus and Swedish Match showed auditors at higher levels, who relied on financial informa-
that it can shield lean production when it is in its infancy; a tion to control the organisation. Table 1 shows that the
phase in which it is most vulnerable due to the proponents’ case companies contained the resulting tensions through
inability to quantify the benefits of lean. colonising and compromising. The mechanisms of decou-
The operations-related problem arose from a conflict pling, implementing incrementally and obscuring were not
between lean’s emphasis on zero inventories and zero feasible, because the reliability of financial information
defects and the traditional emphasis on covering fixed depends heavily on the tracking of internal transactions
costs. In Pegasus and (in earlier days) Gryphon, this con- and counting of inventories at the lower levels. Any initia-
flict resulted in a debate between the proponents of lean at tive to reduce these activities would be visible at the higher
the lower levels, who stressed the futility of ‘overproduc- levels. Our findings showed that the information-related
tion’, and managers at the higher levels, who emphasised problem can sometimes be resolved through colonising.
the importance of high degrees of utilisation of expensive In particular, by showing evidence of its ability to pre-
machinery. These different points of view were reflected dict financial results, Swedish Match was able to convince
in the use of operating performance measures at the the internal auditors that controls built into operating pro-
lower levels and financial performance measures at the cesses were sufficient to keep the company in control, even
higher levels. The proponents of lean in Gryphon and if its ERP system was not always up-to-date. However, the
Pegasus attempted to contain the tensions between the lean proponents in this company complained that they
two sets of performance measures through colonising and had to repeat this process of creating support with every
decoupling. Their attempts to apply colonising were rather change in the composition of the audit team. Furthermore,
hesitant. They were able to create some enthusiasm for the lean proponents in Swedish Match could not fully rely
the use of non-financial performance measures or for the on the mechanism of colonising. Similar to Gryphon, they
replacement of traditional measures, but the emphasis also used compromising to contain tensions related to the
on traditional financial performance measures remained reliability of financial information. Although the resulting
strong. As a result, the mechanism of compromising was compromises did not provide an efficient resolution of the
unfeasible. The mechanism of decoupling seemed to be information-related problem, they were needed to get the
more effective to alleviate the tensions resulting from a auditors’ approval for an ERP system that was ‘more lean’
strong focus on short-term financial results. This mech- than an ERP system that was uniquely built around a tra-
anism allowed the proponents of lean in Gryphon and ditional view of control.
Swedish Match to operate in a relatively autonomous fash- Although the problems associated with the introduc-
ion. However, they were aware that their level of autonomy tion of lean are well-documented (Brimson, 1987; Kaplan,
could vary over time. The lean proponents in Gryphon had 1984, 1989), this paper shows that the case companies
been able to apply decoupling since HQ had embraced had not found decisive solutions to the problems they
the lean philosophy, but they realised that an increased encountered. From the perspective of the lean proponents,
focus on financial results at the higher levels in the future a comprehensive and sustainable resolution could only
would threaten their ability to apply the lean principles. be achieved through colonising. This mechanism helps
In Swedish Match, decoupling was considered a viable to remove the interfaces between the different views of
mechanism as long as the company was able to achieve control. However, support from higher levels was often
good results in traditional terms. The situation in Pegasus insufficient to apply the mechanism of colonising. There-
demonstrated that, with only limited decoupling, inter- fore, the lean proponents used a mix of mechanisms
ventions from higher organisational levels are likely. Even which was adjusted according to the varying degrees of
though operating decisions are usually the responsibil- higher-level support that they could muster. Over time,
ity of the operating domain and hence seem suitable for this approach enabled them to proceed and accumulate
implementing incrementally and obscuring, such interven- more evidence of the benefits of lean that could appeal
tions limit the autonomy of operating departments and to persons who hold a more traditional view of being in
thus reduce the possibilities of applying these mechanisms. control.
The third problem stemmed from different understand-
ings of the importance of reliable financial information. The 6. Conclusions
lean proponents in the case companies argued that finan-
cial information gives a false sense of security and that For a long time, it has been known that traditional
controls built into operating processes are more reliable control systems may create problems which hinder the
and involve less ‘wasteful’ tracking of internal transac- implementation of lean production. To resolve such prob-
tions. However, their attempts to make the accounting lems, it has been argued that control systems should be
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 81

adjusted to fully realise the benefits of lean transfor- Our study has two limitations, which also provide
mations (Maskell and Kennedy, 2007; see also Goldratt opportunities for future research. First, the study only
et al., 2000). Under the heading of ‘lean accounting’ var- included companies which had been able to contain ten-
ious adjustments have been suggested, and research has sions due to inconsistencies between different controls.
shown that accounting systems have changed in the past As a consequence, we have not been able to investigate
few decades, partly to accommodate the implementa- the conditions under which such inconsistencies can lead
tion of lean production. Nevertheless, our case studies to conflicts which challenge the implementation of lean
revealed that the problems associated with the use of production. These conditions may explain some of the
traditional control systems have not been fundamentally lean implementation failures documented by researchers,
resolved for modern-day practice. Although the lean pro- such as Bashin (2012) and Sohal and Egglestone (1994).
ponents in our case companies had a clear understanding Future research could shed more light on the role of tradi-
of the problems created by their traditional control sys- tional accounting-based control systems in lean failures by
tems, and were well aware of the solutions suggested in investigating companies where lean was abandoned. Such
lean accounting publications, they were only able to resolve research may provide an explanation of why these compa-
the problems to a limited extent. Higher hierarchical lev- nies were not able to contain the tensions that arose at the
els continued to rely on the traditional concept of control, interfaces between the two concepts of control.
and this imposed a constraint on the companies’ lean Second, we were unable to identify companies which
transformations. had been fully able to supplant traditional controls with
However, we found that a lean transformation does not lean controls. Although our results would suggest that it
require a fundamental resolution of the problems that may is unlikely, we cannot exclude the possibility that compa-
arise from inconsistencies between the traditional and the nies can fully rely on lean controls. Future studies could
lean concepts of control, provided that the company has focus on companies where the lean concept of control has
learned to cope with the localised tensions that can occur come to dominate decision-making processes. This paper
at the interfaces between the two concepts. This finding suggests that this would involve an evolutionary process in
is in contrast to Kaplan’s (1986) ideal of a new integrated which an increasing number of stakeholders become satis-
control system. At the interfaces between different con- fied with the quality and the usefulness of lean controls for
cepts of control, the interpretations of what it means to their own purposes. However, it may also be a consequence
be in control can diverge. Therefore, it is here that ten- of more revolutionary ‘jolts’. Such jolts can include changes
sions may arise, giving rise to potential conflicts. In the in regulations on external reporting and audit quality, or
four companies we studied, the two concepts of control changes in governance and financing which reduce the
met at various interfaces, but the resulting tensions were reliance on, for instance, banks. So far, these processes are
contained by the mechanisms of colonising, decoupling, unclear.
compromising, implementing incrementally and obscur- Therefore, we call for additional research on ways
ing. The mechanism of decoupling was already mentioned in which tensions between different concepts of control
by Johnson and Kaplan (1987), but we found that it was can be contained or reduced to contribute to improved
used, not only to separate systems for external reporting outcomes of lean implementations. In the case of lean
from systems to control internal operations, but also to production, the coming together of different concepts of
deal with different views of control at different hierar- control proved to be challenging, but it also created oppor-
chical levels. In addition, we uncovered four alternative tunities to increase the support for lean from higher levels.
mechanisms. The lean proponents in the case companies
preferred the mechanisms which involve a higher degree Acknowledgements
of support from higher organisational levels, but they
were not always able to obtain this support. In practice, The authors would like to thank Bob Scapens and Paula
the five mechanisms were not mutually exclusive and van Veen-Dirks for their comments on and suggestions for
were applied in varying degrees. Together, they enabled earlier versions of this paper.
the companies to continue their implementation of lean
production. Appendix A. Overview of interviews
82 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83

Date Function Company Duration

14-01-10 Lean consultant Ynova 02:00

(consulting
firm)
12-03-10 Lean consultant, Industrial cost engineer Gryphon NL, Industrial cost Ynova, 02:00
engineer #2 Gryphon NL, Lean consultant/former interim Plant Gryphon NL,
manager Pegasus NL, Senior financial manager Pegasus NL, Assistant Pegasus NL
financial manager Pegasus NL
31-03-10 Lean consultant/former interim Plant manager Pegasus NL Pegasus NL 02:30
06-04-10 Industrial cost engineer, Industrial cost engineer #2, Manager Gryphon NL 02:30
13-04-10 Senior financial manager Pegasus NL 02:00
13-04-10 Assistant financial manager Pegasus NL 01:30
03-05-10 Senior manager AP Nederland 01:52
03-05-10 Senior financial manager AP Nederland 01:33
10-05-10 Senior manager Swedish Match 02:39
Assen
10-05-10 Black belt lean manager Swedish Match 01:47
Assen
12-05-10 Senior financial manager Swedish Match 02:13
Assen
12-05-10 Senior manager, Black belt lean manager, Senior financial manager Swedish Match 00:45
Assen
21-05-10 Lean consultant/former interim Plant manager Pegasus NL* Pegasus NL 01:15
28-05-10 Senior financial manager, Assistant financial manager* Pegasus NL 01:15
04-06-10 Senior financial manager, Assistant financial manager* Pegasus NL 01:50
15-06-10 Production manager #1* Pegasus NL 01:30
16-06-10 Production manager #2* Pegasus NL 01:00
24-06-10 Lean manager/Manager quality control* Pegasus NL 00:55
01-07-10 Supply chain manager* Pegasus NL 00:30
01-07-10 Technical project manager* Pegasus NL 00:17
01-07-10 Divisional Procurement manager* Pegasus NL 00:35
08-07-10 Plant manager* Pegasus NL 01:45
09-07-10 Senior financial manager, Assistant financial manager* Pegasus NL 01:47
14-07-10 Team leader production department* Pegasus NL 00:27
21-07-10 Manager product engineering* Pegasus NL 01:18
21-07-10 Lean consultant/former interim Plant manager Pegasus NL* Pegasus NL 01:10
29-07-10 Assistant financial manager* Pegasus NL 00:42
22-09-10 Industrial cost engineer #3* Gryphon NL 00:30
12-10-10 Green belt programme manager* Gryphon NL 01:00
13-10-10 Maintenance employee* Gryphon NL 00:45
14-10-10 Employee technical support group line 2* Gryphon NL 00:45
18-10-10 Production manager hall 4* Gryphon NL 01:30
19-10-10 Lean coach* Gryphon NL 03:00
19-10-10 Quality inspector hall 2* Gryphon NL 00:45
20-10-10 Shift leader assembly line 2* Gryphon NL 00:30
01-11-10 Industrial cost engineer #1 Gryphon NL 01:40
01-11-10 Financial controller Gryphon NL 01:31
03-11-10 Logistics capacity planner* Gryphon NL 01:00
08-11-10 Management team member Supply* Gryphon NL 01:00
08-11-10 Employee logistics department* Gryphon NL 00:45
09-12-10 Meeting about the challenges of accounting and control in a lean Gryphon NL, 04:00
environment. Attended by most interviewees Pegasus NL,
Swedish
Match, AP
Nederland

Interviews indicated by a * were conducted by our research assistants.

References Banker, R.D., Potter, G., Schroeder, R.G., 1993. Reporting manufactur-
ing performance measures to workers: an empirical study. J. Manag.
Abernethy, M.A., Lillis, A.M., 1995. The impact of manufacturing flexibil- Account. Res. 5 (Fall), 33–55.
ity on management control system design. Account. Org. Soc. 20 (4), Benton, W.C., Shin, H., 1998. Manufacturing planning and control: the evo-
241–258. lution of MRP and JIT integration. Eur. J. Oper. Res. 110 (3), 411–440.
Abernethy, M.A., Lillis, A.M., Brownell, P., Carter, P., 2001. Product diver- Bhasin, S., 2012. Prominent obstacles to lean. Int. J. Product. Perform.
sity and costing system design choice: field study evidence. Manag. Manag. 61 (4), 403–425.
Account. Res. 12 (3), 261–279. Brimson, J.A., 1987. CAM-I cost management systems project. In: Capet-
Åhlström, P., Karlsson, C., 1996. Change processes towards lean produc- tini, R., Clancy, D.K. (Eds.), Cost Accounting, Robotics and the New
tion: the role of the management accounting system. Int. J. Oper. Prod. Manufacturing Environment. Edited Presentations of the First Annual
Manag. 16 (11), 42–56. Management Accounting Symposium. American Accounting Associa-
Bamber, L., Dale, B.G., 2000. Lean production: a study of application in tion.
a traditional manufacturing environment. Prod. Plan. Control 11 (3), Brimson, J.A., Berliner, C., 1987. The cost management system project.
291–298. Comput. Aided Eng. J. 4 (5), 198–201.
 S. Tillema, M. van der Steen / Management Accounting Research 27 (2015) 67–83 83

Brosnahan, J.P., 2008. Unleash the power of lean accounting. J. Account. Kassicieh, S.K., Yourstone, S.A., 1998. Training, performance evaluation,
206 (1), 60–66. rewards, and TQM implementation success. J. Qual. Manag. 3 (1),
Cable, J., 2009. Lean accounting’s quest for acceptance. Ind. Week 258 (9), 25–38.
26–29. Katayama, H., Bennett, C., 1996. Lean production in a changing compet-
Chenhall, R.H., 2003. Management control systems design within its orga- itive world: a Japanese perspective. Int. J. Oper. Prod. Manag. 16 (2),
nizational context: findings from contingency-based research and 8–23.
directions for the future. Account. Org. Soc. 28 (2/3), 127–168. Kennedy, F.A., Widener, S.K., 2008. A control framework: insights from
Cua, K.O., McKone, K.E., Schroeder, R.G., 2001. Relationships between evidence on lean accounting. Manag. Account. Res. 19 (4), 301–323.
implementation of TQM, JIT, and TPM and manufacturing perfor- Krafcik, J.F., 1988. Triumph of the lean production system. Sloan Manag.
mance. J. Oper. Manag. 19 (6), 675–694. Rev. 30 (1), 41–52.
Daniel, S.J., Reitsperger, W.D., 1991. Management control systems for J.I.T.: Laughlin, R.C., 1987. Accounting systems in organisational contexts: a case
an empirical comparison of Japan and the U.S. J. Int. Bus. Stud. 22 (4), for critical theory. Account. Org. Soc. 12 (5), 479–502.
603–617. Lewis, M.A., 2000. Lean production and sustainable competitive advan-
Forza, C., 1996. Work organization in lean production and traditional tage. Int. J. Oper. Prod. Manag. 20 (8), 959–978.
plants: what are the differences? Int. J. Oper. Prod. Manag. 16 (2), Maskell, B.H., Baggaley, B., Grasso, L.P., 2012. Practical Lean Accounting:
42–62. A Proven System for Measuring and Managing the Lean Enterprise,
Fullerton, R.R., Kennedy, F.A., Widener, S.K., Available at SSRN: second ed. Boca Raton, CRC Press.
http://ssrn.com/abstract=1659386 2010. Accounting for A Lean Envi- Maskell, B.H., Kennedy, F.A., 2007. Why do we need lean accounting and
ronment. Meeting Paper. AAA 2011 Management Accounting Section., how does it work? J. Corp. Account. Fin. 18 (3), 59–73.
http://dx.doi.org/10.2139/ssrn.1659386. Meyer, J.W., Rowan, B., 1977. Institutionalized organizations: formal
Fullerton, R.R., Kennedy, F.A., Widener, S.K., 2013. Management account- structure as myth and ceremony. Am. J. Sociol. 83 (2), 340–363.
ing and control practices in a lean manufacturing environment. Otley, D., 1994. Management control in contemporary organiza-
Account. Org. Soc. 38 (1), 50–71. tions: towards a wider framework. Manag. Account. Res. 5 (3/4),
Fullerton, R.R., McWatters, C.S., 2002. The role of performance measures 289–299.
and incentive systems in relation to the degree of JIT implementation. Ouchi, W.G., 1979. A conceptual framework for the design of organiza-
Account. Org. Soc. 27 (8), 711–735. tional control mechanisms. Manag. Sci. 25 (9), 833–848.
Fullerton, R.R., Wempe, W.F., 2009. Lean manufacturing, non-financial Prajogo, D.I., McDermott, C.M., 2005. The relationship between total qual-
performance measures, and financial performance. Int. J. Oper. Prod. ity management practices and organizational culture. Int. J. Oper. Prod.
Manag. 29 (3), 214–240. Manag. 25 (11), 1101–1122.
Goldratt, E.M., Schragenheim, E., Ptak, C.A., 2000. Necessary But Not Suffi- Sale, M.L., Inman, R.A., 2003. Survey-based comparison of performance
cient: A Theory of Constraints Business Novel. The North River Press, and change in performance of firms using traditional manufacturing,
Great Barrington, MA. JIT and TOC. Int. J. Prod. Res. 41 (4), 829–844.
Hines, P., Holweg, M., Rich, N., 2004. Learning to evolve: a review of con- Shah, R., Ward, P.T., 2003. Lean manufacturing: context, practice bundles,
temporary lean thinking. Int. J. Oper. Prod. Manag. 24 (10), 994–1011. and performance. J. Oper. Manag. 21 (2), 129–149.
Holweg, M., 2007. The genealogy of lean production. J. Oper. Manag. 25 Sim, K.L., Killough, L.N., 1998. The performance effects of complementar-
(2), 420–437. ities between manufacturing practices and management accounting
Ittner, C.D., Larcker, D.F., 1995. Total quality management and the choice systems. J. Manag. Account. Res. 10, 325–346.
of information and reward systems. J. Account. Res. 33, 1–34. Sohal, A.S., Egglestone, A., 1994. Lean production: experience among Aus-
Johnson, H.T., 2006. Lean accounting: to become lean, shed accounting. tralian organizations. Int. J. Oper. Prod. Manag. 14 (11), 35–51.
Cost Manag. 20 (1), 6–17. Spear, S., Bowen, H.K., 1999. Decoding the DNA of the Toyota production
Johnson, H.T., Kaplan, R.S., 1987. Relevance Lost: The Rise and Fall of Man- system. Harvard Bus. Rev. 77 (5), 96–106.
agement Accounting. Harvard Business School Press, Boston. Voss, C.A., 1995. Alternative paradigms for manufacturing strategy. Int. J.
Kaplan, R.S., 1984. Yesterday’s accounting undermines production. Har- Oper. Prod. Manag. 15 (4), 5–16.
vard Bus. Rev. 62 (4), 95–102. Womack, J.P., Jones, D.T., 1996. Lean Thinking: Banish Waste and Create
Kaplan, R.S., 1986. Must CIM be justified by faith alone. Harvard Bus. Rev. Wealth in Your Corporation. Simon & Schuster, New York, NY.
64 (2), 87–95. Womack, J.P., Jones, D.T., Roos, D., 1990. The Machine that Changed the
Kaplan, R.S., 1988. One cost system isn’t enough. Harvard Bus. Rev. 66 (1), World. Rawson Associates, New York, NY.
61–66. Worley, J.M., Doolen, T.L., 2006. The role of communication and manage-
Kaplan, R.S., 1989. Management accounting for advanced technological ment support in a lean manufacturing implementation. Manag. Decis.
environments. Science 245 (4920), 819–823. 44 (2), 228–245.