Description
Many firms today form alliances with their suppliers and customers that do not fit into the classical dichotomy of
hierarchies and markets.The emergence of so-called hybrid relational forms makes the make-or-buy decision more
complicated than the neo-classical economic perspective indicates.One outcome of these hybrid relational forms
appears to be the development of cost management techniques that cross the organizational boundary between buyers
and suppliers and whose objective is to reduce costs through collaborative efforts.This paper explores how firms enact
interorganizational cost management during product design and the characteristics of the relational contexts associated
with them.It also discusses the implications of such developments for the make-or-buy decision.
Interorganizational cost management and relational context
Robin Cooper
a,
*, Regine Slagmulder
b
a
Emory University, 1300 Clifton Road, Atlanta, GA 30322-270, USA
b
INSEAD, Boulevard de Constance, F-77305 Fontainebleau Cedex, France
Abstract
Many ?rms today form alliances with their suppliers and customers that do not ?t into the classical dichotomy of
hierarchies and markets. The emergence of so-called hybrid relational forms makes the make-or-buy decision more
complicated than the neo-classical economic perspective indicates. One outcome of these hybrid relational forms
appears to be the development of cost management techniques that cross the organizational boundary between buyers
and suppliers and whose objective is to reduce costs through collaborative e?orts. This paper explores how ?rms enact
interorganizational cost management during product design and the characteristics of the relational contexts associated
with them. It also discusses the implications of such developments for the make-or-buy decision.
#2003 Elsevier Ltd. All rights reserved.
Introduction
The neo-classical economics literature represents
the make-or-buy decision as a dichotomous com-
petition between the ?rm and an external supplier.
Two fundamental assumptions underlying this
model are the absence of information asymmetry
between the buyer and the supplier, and the
incomplete nature of the contract between the two
trading partners (Baiman & Rajan, 2000). Despite
the restrictiveness of these assumptions, such a
pure market approach allows for the acquisition
of a broad range of items, including commodity
products (such as nuts and bolts) and commodity
processes (such as surface-mounting parts onto a
printed circuit board). However, as ?rms increas-
ingly focus their attention on their core compe-
tencies, they outsource both a higher percentage of
the total costs of their products and more sub-
stantial items that do not rely upon their core
competencies (Bryce & Useem, 1998; Gilley &
Raheed, 2000; Kotabe & Murray, 1990; Prahalad
& Hamel, 1990; Quinn, 1992; Venkatraman,
1989). The rationale for using an external supplier
for such items includes the supplier’s superior cost
e?ciency, functionality, and quality, and their
ability to incorporate new technologies in a time-
lier manner (Monczka & Trent, 1991; Nishiguchi,
1994). Consequently, not all items that ?rms out-
source can be described as either product or pro-
cess commodities. Instead, many of them rely
upon knowledge that is proprietary to the buyer
or supplier.
The outsourcing of more signi?cant items intro-
duces the problem of information asymmetry
between the buyer and the supplier into the make-
or-buy decision. This information asymmetry can
cause the buyer to establish speci?cations that
unnecessarily increase the costs incurred by the
supplier. For example, by requiring certain func-
tional speci?cations, the buyer might force the
0361-3682/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0361-3682(03)00020-5
Accounting, Organizations and Society 29 (2004) 1–26
www.elsevier.com/locate/aos
* Corresponding author.
supplier to develop the outsourced part using
unnecessarily expensive raw materials. In order to
reduce the costs associated with this form of
information asymmetry, product engineers at the
buyer and supplier may meet during the product
development process and identify opportunities to
change the buyer’s speci?cations in ways that
lower overall costs. Such formalized buyer-sup-
plier interactions, the objective of which is to
identify opportunities for joint cost reduction, are
the domain of interorganizational cost manage-
ment (IOCM) (Cooper & Slagmulder, 1999;
Cooper & Yoshikawa, 1994).
As ?rms outsource more signi?cant items and
engage in IOCM, they develop relational contexts
that do not fall into the simple dichotomy of
markets and hierarchy (Williamson, 1975, 1979,
1985). Instead, these relationships represent inter-
mediate or hybrid modes of governance that
enable ?rms to access the economies of scale and
scope of their trading partners in more e?cient
ways than are possible through either pure arm’s-
length transactions or through vertical integration
(Powell, 1990; Sheppard & Tuchinsky, 1996; Wil-
liamson, 1991). Many di?erent forms of relational
context between buyers and suppliers have been
observed, ranging from relationships in which the
interactions are close to market driven, to strategic
partnerships in which the ?rms have signaled their
desire to work together closely over the long-term
(Heide & John, 1990). These hybrid relational
contexts are characterized by incomplete con-
tracting, as it is neither possible nor practical to
develop contracts that completely specify all of the
potential outcomes of the interactions between
both parties (Baiman & Rajan, 2000).
One outcome of these hybrid relational contexts
is that the calculus of the make-or-buy decision
becomes more complex (Gietzmann, 1996). This
complexity derives in part from the fact that
transaction costs are di?cult to quantify with any
accuracy or rigor, and therefore are likely to
diminish in importance in the decision-making
process (Walker & Weber, 1984). Management
accounting textbooks typically avoid the
measurement problem and treat these costs as
qualitative factors (Atkinson, Banker, Kaplan, &
Young, 2001; Horngren, Foster, & Datar, 2000).
This treatment is problematic, since the account-
ing justi?cation of the decision to source an item
internally, versus externally, often appears to err
by systematically underestimating the magnitude
of these qualitatively determined costs (Drtina,
1994; Lacity, Wilolocks, & Feeny, 1996). Further-
more, as the complexity of the design project
increases, so too does the scope of management
control systems to include not just the narrow
accounting numbers (cost, pro?tability, and bud-
get), but also a broader information set that cap-
tures customer, product design, and time-sensitive
measures (Davila, 2000).
Thus, the increased complexity of the out-
sourcing decision places new demands upon the
accounting and information systems of ?rms, as
evidenced in part by the emergence of IOCM. Yet,
despite the increased interest in interdependencies
and information ?ows that transcend organiza-
tional boundaries and their potential implications
for the ?eld of management accounting, the topic
has been largely ignored in the accounting
research literature (Hopwood, 1996; Van der
Meer-Kooistra & Vosselman, 2000). In this paper,
we seek to address this limitation by exploring
IOCM practices and the relational contexts asso-
ciated with them at seven Japanese manufacturing
?rms.
The paper is structured as follows. In the next
section, we provide a discussion of the ?eld-based
research method and a brief description of the
?rms involved in the research sample. In Observed
IOCM practices, we give a detailed description of
the three IOCM techniques observed at the sample
?rms. These techniques vary in scope from modest
design changes to a near overhaul of both the end
product and the outsourced item. We also
demonstrate that the ability to undertake IOCM
techniques appears to cluster, such that the ability
to perform a more aggressive technique becomes
diagnostic of the ability to perform less aggressive
ones. In Observed relational contexts, we identify
the characteristics of the relational contexts in
which these IOCM techniques were observed. In
IOCM clusters and relational contexts, we show
how each of the observed clusters of IOCM prac-
tices are associated with a speci?c relational con-
text and we discuss the perceived impact of the
2 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
di?erent levels of IOCMon the joint performance of
the sample ?rms. In the Summary and conclusions,
we discuss the impact of IOCM on the calculus of
the make-or-buy decision. Finally, we evaluate the
implications of the research ?ndings upon the ?eld
of cost management and the make-or-buy decision.
Research method
To gain insights into the IOCM practices of
?rms and the relational contexts associated with
them, an exploratory ?eld-based research project
was undertaken. The choice of exploratory case-
based research was dictated by the nature of the
research problem and the lack of extant literature
about interorganizational cost management (Yin,
1984). We visited three large Japanese manufacturing
enterprises, a ?rst-tier supplier to each of them, and a
second-tier supplier to one of the ?rms and docu-
mented their IOCM practices. We selected product
development as the domain of IOCM to be studied,
as prior research has indicated that joint product
development represents a signi?cant source of eco-
nomic bene?t in buyer-supplier relations (Clark &
Fujimoto, 1991; Cooper & Yoshikawa, 1994).
Research site selection
The research ?ndings reported in this paper are
based on an in-depth examination of the IOCM
practices observed at seven Japanese manufactur-
ing companies, belonging to three di?erent supply
chains: the Komatsu–Toyo Radiator supply chain,
consisting of a large heavy industrial manu-
facturer and an independent manufacturer of heat-
exchange equipment such as radiators, oil coolers,
and condensers; the Isuzu–Jidosha Kiki chain, com-
prised of one of the largest automobile manufactur-
ing companies in Japan and a ?rst-tier supplier of
brakes, clutches, steering systems, and pumps; and
the Tokyo–Yokohama–Kamakura chain, includ-
ing a top 10 automobile manufacturer, a ?rst-tier
supplier of hydraulic systems, and a small second-
tier supplier of automotive parts. We identi?ed
and selected the research sites according to various
criteria. Komatsu was selected because of its
reputation for e?ectively managing costs across its
supply chain. Meanwhile, we chose Isuzu for its
reputation as one of Japan’s best practitioners of
value engineering and cost management during the
product development phase. Finally, we chose to
include Tokyo Motors for its reputation for under-
taking thorough customer requirement analyses tied
to its target costing system.
1
To capture IOCM practices from the supplier’s
perspective, we requested that each of the three
buyer ?rms identify a ?rst-tier supplier that man-
agement considered to be especially adept at under-
taking IOCM. Komatsu identi?ed Toyo Radiator as
an excellent ?rst-tier supplier with which it under-
took particularly e?ective IOCM. Isuzu named
Jidosha Kiki Company (JKC) as one of its most
innovative suppliers with whom it had developed a
highly e?ective cost management relationship.
Finally, Tokyo Motors identi?ed its ?rst-tier sup-
plier, Yokohama Corporation, and its second-tier
supplier, Kamakura Iron Works, as they formed a
three-?rm supply chain that was particularly suc-
cessful at practicing multi-?rm cost management.
The choice of Yokohama and Kamakura was for-
tuitous in that the three ?rms were essentially inde-
pendent, rather than part of a kereitsu.
2
We
considered independence to be particularly impor-
tant because it allowed buyer-supplier interactions to
be observed in their purest form. However, if the
?rms were part of a kereitsu, then ‘‘invisible’’ o?-
setting transactions, such as low-interest loans,
might cause one of the ?rms to agree to ‘‘sub-eco-
nomic’’ selling prices. In addition, while the research
focused primarily on the relationships between the
three buyer ?rms and their identi?ed suppliers, we
also documented the general relationships between
the three buyer ?rms and their entire supplier bases.
Data collection
The ?eld research relied upon open-ended inter-
views with managers, design and manufacturing
engineers, and blue-collar workers at the seven ?rms
in the sample. The interviews focused primarily on
1
For a detailed description of target costing, see Cooper
and Slagmulder (1997).
2
This supply chain was the focus of the Cooper and
Yoshikawa (1994) paper.
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 3
their IOCM practices. Given the objective of the
research to study IOCM practices during product
design, the data collection process was limited to
design-related variables. We chose not to interview
individuals in the sales function, for example, as
they were considered unlikely to interact with the
suppliers’ IOCM systems. This decision was vali-
dated by our discussions with senior management
at all of the sample ?rms, which revealed that the
process of IOCM was largely limited to the design,
engineering, and manufacturing functions of the
?rm.
The individuals interviewed at each site were
actively involved in developing, implementing, and
applying the various IOCM techniques used by the
?rm. The interviews were conducted in English with
translator support as appropriate. Typically, we
interviewed between three to ?ve persons (although
the interview occasionally involved as few as one
and or as many as six subjects). Job titles of those
interviewed included General Manager of Product
Planning, Manager of Corporate Planning, Chief
Engineer, and Senior Manager of Group
Accounting. The total site visits lasted 16 days.
The typical time spent at each ?rm was between 2
and 3 days. Initially, we visited each ?rm for 1–2
days and, from the information collected, pre-
pared a draft of the case. Follow-up visits, typi-
cally lasting half to a whole day, clari?ed any
major outstanding misunderstandings and allowed
the company’s appointed contact manager to sign
and release the ?nal draft of the research case.
Company documents, copious notes, and tape
recordings of the interviews were the basis for
seven research cases of approximately 5000 words
each. The majority of the data collected was qua-
litative in nature. The aim of the research was to
identify and to understand the ?rms’ IOCM prac-
tices and the contexts within which they occurred.
Consequently, the research focused more on
exploring the processes that enabled the ?rms in the
sample to collaborate e?ectively, such as guest
engineer programs,
3
than on collecting quanti?able
metrics, such as the number of engineering hours
dedicated to joint design. With the small number of
?rms inthe sample andthe selectionprocesses usedto
identify them, this research only provides evidence
about the existence of IOCM practices. It should not
be construed as either providing evidence of central
tendency behavior or justifying the investment of
resources made in these techniques.
The cases were sent to the contact manager in
each ?rm for review. The ?rst draft of the cases
contained numerous questions that could not be
answered from the tape recordings and notes. The
cases typically went through two to three revisions
before being cleared by senior management at the
sample ?rms. It took between 3 and 12 months to
clear each case. When necessary, the questions and
appropriate textual portions of the case were
translated into Japanese so that managers with
inadequate English skills could answer the ques-
tions and review the text for accuracy. During a
typical clearance procedure, approximately 30
questions were answered and about one-third of
the case was rewritten or amended in some way.
While the majority of these changes involved
author-initiated questions, others were corrections
to the drafts that the reviewing managers made
before releasing the cases. The purpose of this itera-
tive process was to increase the probability that the
observations captured in the cases were factually
correct and accurately re?ected actual practice.
Data analysis
We undertook data analysis in a modi?ed,
three-stage version of the process suggested by
Eisenhardt (1989). First, we used within-case ana-
lysis to identify the IOCM techniques employed at
each individual ?rm. Second, we undertook a
cross-case analysis of the ?rms within the same
supply chain to identify the unique buyer–supplier
interactions across the ?rms’ boundaries that
enabled them to achieve IOCM. Finally, we used
cross-case analysis to uncover patterns of IOCM
techniques and buyer–supplier interactions com-
mon to multiple ?rms and supply chains in the
sample. We identi?ed patterns of association
between IOCM practices and relational contexts
based on the cross-case analyses, using evidence
3
Guest engineers are employees of the supplier who spend
an extended time at the buyer, or vice versa, in order to resolve
joint design problems. For a discussion of guest engineers, see
Holden and Burgess (1994).
4 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
from each case to support or to extend the emer-
ging theory. We considered the data analysis
complete when no additional general patterns
could be identi?ed from the ?eld observations.
In this paper, we primarily analyze the practices
observed between the dyadic pair of Komatsu and its
supplier, Toyo Radiator, both of whom had devel-
oped sophisticated IOCM skills. Komatsu, Ltd. is
one of Japan’s largest heavy industrial manu-
facturers and the world’s second-largest manu-
facturer of a complete line of construction
equipment. The ?rm’s product line contains over
300 models, including bulldozers, hydraulic exca-
vators, wheel loaders, and dump trucks. Toyo
Radiator Co., Ltd. (Toyo) was founded in 1936 as a
radiator supplier to the ?edgling Japanese auto-
mobile industry. Over the years, it has diversi?ed into
all arenas of heat exchange applications and is now
one of the world’s largest independent heat-exchange
equipment manufacturers for construction equip-
ment. Komatsuhas beena Toyocustomer since 1955.
We chose to focus on the Komatsu–Toyo dya-
dic pair because our observations at those two
?rms captured a contemporaneous change in their
IOCM practices and the relational context
between them. Consequently, the interplay
between buyer and supplier and the role of the
relational context in IOCM were more directly
visible at Komatsu–Toyo than at the other
research sites. We used the observations at the
other ?rms both to ?ll in gaps in our observations
at Komatsu–Toyo (for example, about other IOCM
techniques and their associated relational contexts)
and to provide con?rming evidence of any inter-
pretations of those observations. From our collec-
tive observations, we developed a rich description
and analysis of the observed IOCM techniques and
the relational contexts associated with them.
Limitations
Field research is not without its limitations.
Biased observations and a lack of external validity
remain two common criticisms of ?eld studies
(Yin, 1984). McKinnon (1988) highlights the fol-
lowing three major sources of bias. First, the
researcher’s selective perception and interpretation
of events may in?uence the data collection and/or
analysis stage. Second, the interviewees’ reports
may be biased, due to natural human tendencies
and limitations; for example, interviewees may
forget details or report events in a manner that is
?attering to themselves or their company. Finally,
because the researcher is on site only for a limited
period of time, abnormal instances of the studied
phenomenon might be mistakenly treated as if
they were normal conditions. Our research miti-
gated against such potential sources of bias in
several ways. First, wherever possible, we cross-
checked any statements regarding the operation of
the ?rms’ cost management systems with both
company records and descriptions from other
individuals. Second, by developing long-term
relationships with many of the managers inter-
viewed, we were able to ensure their trust and
cooperation, thus reducing any tendencies on their
part to misreport events in ways that would favor
either the company or themselves. Finally, by visit-
ing the companies over an extended period of time,
we avoided the risk of improperly interpreting aty-
pical events through longitudinal observation.
The second criticism of ?eld research relates to
the risks associated with generalizing case study
?ndings in other settings. A common view is that
one cannot draw generally valid inferences from
observations at a small sample of companies
(Ha? gg & Hedlund, 1979; Lukka & Kasanen,
1995). While accepting the rationale behind such
statements, we believe that the knowledge
obtained through our interviews is interesting in
its own right and potentially o?ers a basis for a
more generally applicable theory. Although the
theory developed in this paper is an attempt to
give meaning to a small number of individual
observations in their particular context, we cau-
tiously suggest that it also could apply in other
settings. To validate this claim will require, at a
minimum, that the theory’s external validity be
tested using large-sample statistical methods.
Observed IOCM practices
IOCM helps to reduce the information asym-
metry between the buyer and supplier regarding
the relationship between the speci?cations for the
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 5
outsourced item established by the buyer, and the
resulting costs at the supplier. An IOCM interven-
tion occurs when this information asymmetry causes
the buyer to set speci?cations that the supplier can-
not meet if the latter is to make an acceptable pro?t.
The primary mechanism for identifying the necessity
of an IOCM intervention is target costing. Target
costing lies at the heart of IOCM in the sample ?rms,
as it links customer demands through product design
to the parts acquisition process (Cooper & Chew,
1996; Cooper & Slagmulder, 1997; Kato, Boer, &
Chow, 1995; Koga, 1998; Monden, 1995). However,
as suggested in the literature, target costing is an
arm’s-length cost management technique; it does not
actively involve the supplier in the buyer’s cost
management program. Instead, the buyer’s target
costing system identi?es the purchase price of the
outsourced item, which signals to the supplier’s
target costing system where cost reduction is
necessary. The key extension of IOCM beyond
other cost management techniques is the active
involvement of both the buyer’s and supplier’s
design teams in the joint management of costs.
We observed three IOCM techniques. The ?rst
technique, functionality–price–quality (FPQ) trade-
o?s, helped to resolve relatively minor cost overrun
problems and involved only modest speci?cation
changes—and hence limited interactions among the
?rms’ design engineers. The second technique,
interorganizational cost investigations, was applied
when FPQ trade-o?s were unable to produce the
desired level of cost reductions. The technique
involved more intense interactions among the design
engineers and more signi?cant changes both to the
design of the outsourced item and occasionally to
the speci?cations of the end product. The ?nal
IOCM technique, concurrent cost management,
addressed cost problems that demanded the most
signi?cant cost reduction levels of all three tech-
niques. It required the most signi?cant interactions
between the buyer’s and supplier’s design engineers
and led to fundamental changes in both the buyer’s
product and the outsourced components.
Functionality–price–quality tradeo?s
An FPQ trade-o? is initiated whenever the sup-
plier determines that the manufacturing cost of the
outsourced item will exceed its target cost and that
the only way to reduce costs to the target level is
to relax the functionality and/or quality speci?ca-
tions of the outsourced item in ways acceptable to
the buyer. Once such relaxations are identi?ed, the
supplier requests a meeting with the buyer and the
two design teams to discuss the proposed changes
and to obtain buyer approval to make them. Suc-
cessfully identi?cation of such opportunities helps
the supplier to ensure that it generates adequate
returns. For example, senior management at
Yokohama described the ?rm’s ability to identify
opportunities for cost reduction through FPQ
trade-o?s as critical to its long-term success.
At the heart of a successful FPQ trade-o? is an
e?ective value-engineering program. Yokohama
applies value engineering to all products and fully
integrates the technique into the ?rm’s new pro-
duct development process. In that process, the
product’s basic functions are ?rst identi?ed and its
target cost established. The next step is to develop
prototypes, analyze their costs, and compare them
to the product’s target cost. If the ?nal prototype’s
costs are considered acceptable, it is subjected to
reliability tests and then submitted to the customer
for approval. Once the product obtains customer
approval, it is subjected to a second design round
and its production costs are re-estimated. If these
costs exceed the target cost, then a ?rst-look
value-engineering project occurs. The aim of this
project is to identify ways to change the design of
the product so that it can be manufactured at its
target cost. Once the second generation of design
is established, another round of cost estimation is
undertaken by the engineers. If the design is con-
sidered acceptable, it is subjected to an analysis to
ensure that the product meets its quality speci?ca-
tions. Upon successful completion of this third
design analysis, experimental mass production
ensues, wherein the manufacturing cost of the
product is again re-estimated and the quality and
functionality of the produced items evaluated.
Sometimes, a fourth design review may be neces-
sary to correct any de?ciencies before the com-
pany can release the product for mass production.
Typically, an FPQ trade-o? is an outcome of a
?rst-look value-engineering project. However, it
can be initiated at any time in the design process,
6 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
up to the release of the item into mass production.
Yokohama’s success at initiating these interven-
tions relies upon the in-depth knowledge that its
engineers have developed regarding its customers’
use of the ?rm’s products. This knowledge allows
the engineers to identify where and to what extent
relaxations in the speci?cations of the outsourced
item will be amenable to the buyer. Examples of
the types of changes in speci?cations that result
from FPQ trade-o?s include a request to limit the
color of a part to black or silver, instead of
matching the color of the end-product; a request
to relax surface tolerances when they are not visi-
ble to the end-user; a reduction in the number of
strengthening bars; a shift to pressing a part as
opposed to machining it; and a reduction in the
material content without reducing the strength of
the part. Under FPQ tradeo?s, only minor chan-
ges can be made to the speci?cations of the out-
sourced item and the speci?cations of the end
product are essentially ?xed.
Interorganizational cost investigations
An interorganizational cost investigation takes
place whenever any ?rm in the supply chain for an
item determines that they cannot manufacture the
part at its target cost, and that an FPQ trade-o?
will not produce su?cient cost reductions to
resolve the problem. In many ways, an inter-
organizational cost investigation is similar to a
just-in-time (JIT) production line with each
worker having the ability to shut down the line
when a defect occurs; in this case, however, the
workers are replaced by the design teams at the
supplier ?rms, and the defect lies in the relation-
ship between the speci?cations of the outsourced
item and its purchase price. Just as in JIT, once
the interorganizational cost investigation is initi-
ated, all of the involved players send representa-
tives from their design teams to resolve the
problem. Thus, the ?rst major di?erence between
an FPQ trade-o? and an interorganizational cost
investigation is the ability to include design engi-
neers from more than two ?rms in the supply
chain. The second major di?erence is the scope of
the design changes contemplated. More speci?-
cally, an interorganizational cost investigation
allows parts to be redesigned so that all of the
steps from raw material to ?nished product are
more cost-e?cient. This increased scope of the
design changes enables the identi?cation and
implementation of greater cost savings across the
entire supply chain. However, just as with FPQ
tradeo?s, the fundamental design of the end pro-
duct still remains essentially ?xed.
There are two ways to reduce costs through
interorganizational cost investigations. First, in
order to take full advantage of the manufacturing
skills located throughout the supply chain, avoid
the need to perform activities by redesigning the
product and its components. Second, change the
location of activities so that they are performed
more e?ciently. For example, Kamakura trig-
gered an interorganizational cost investigation
between the engineers of Tokyo Motors, Yoko-
hama, and Kamakura when Tokyo Motors estab-
lished speci?cations for an internal part that could
be met only if Kamakura forged, rather than cast,
the blank. Tokyo Motors’ target costing system
speci?ed the part’s purchase price and transmitted
it, along with the functionality and quality speci?-
cations, to Yokohama. The engineers at that ?rm
identi?ed the manufacturing processes they would
use and, consequently, the speci?cations of the
item they would source from Kamakura. They
used their target costing system to identify their
purchase price for the outsourced part. When the
speci?cations and associated purchase price were
transmitted to Kamakura, the engineers at that
?rm identi?ed their production processes and
determined the necessity of an expensive forging,
rather than a less expensive casting. Consequently,
they concluded that the purchase price set by
Yokohama was insu?cient for Kamakura to gen-
erate an adequate pro?t.
At this point in the process, Kamakura had two
options. The ?rst was to refuse the business, and
the second was to request an interorganizational
cost investigation. Kamakura’s engineers chose
the latter course and requested a joint meeting of
the engineers from all three ?rms. Since only
Kamakura’s engineers had the expertise in forging
and casting technology necessary to understand
the implications of the shift between the two, they
ran the meeting. The solution they identi?ed was
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 7
for Kamakura to start with a cheaper casting and
to undertake the ?rst stage in the surface prepara-
tion process by removing the ridges inherent to the
casting process. Yokohama would then take
delivery of the part and machine it to the shape
required by Tokyo Motors. In addition, the ?rm
would machine the surface of the part (an extra
step) to ensure that it conformed to the speci?ca-
tions provided by Tokyo Motors. Finally, while
Tokyo Motors was not requested to change the
purchase price of the part, it was required to
change the speci?cations in two ways. First, its
engineers reduced the minimum acceptable tensile
strength of the part and, second, allowed for small
blemishes in the surface of the part. Both changes
were necessary to enable casting technology,
rather than forging, to be utilized for the blank.
The two changes proved acceptable to Tokyo
Motors’ engineers because the part was not sub-
jected to much strain and the imperfections on the
surface were not visible to the owner of the vehi-
cle. In e?ect, the original speci?cations had been
too demanding and thus had caused the excess
manufacturing costs. Once all three ?rms had
agreed to the changes, the engineers at Tokyo
Motors set Kamakura’s selling price and thus
speci?ed the distribution of pro?ts between
Kamakura and Yokohama. It was Tokyo’s
responsibility to distribute the pro?ts, as it was the
most powerful ?rm in the relationship and there-
fore could legislate the split to the other ?rms.
Concurrent cost management
Concurrent cost management aggressively redu-
ces costs by increasing the scope of design changes
that the supplier can undertake. One of the events
triggered the emergence of concurrent cost man-
agement at Komatsu was the requirement that
Toyo Radiator produce an engine cooling system
with 40% more capacity at only 18% higher cost.
Komatsu engineers realized that only heroic
e?orts on both sides would enable them to achieve
that objective. Concurrent cost management helps
achieve greater cost savings through increased
design changes in two ways. First, it increases the
amount of time that a supplier’s engineers may
develop innovative solutions to the customers’
requests; second, it concentrates the sourcing of an
entire major function with a single supplier.
Because of the high cost involved, concurrent cost
management is used only for high-value items,
such as major functions. The aim of involving the
supplier much earlier in the design process is to
provide that ?rm with more time to undertake
fundamental redesigns of the major function. The
increased supplier concentration is important
because it allows the supplier to make more fun-
damental changes to the design of the major
function than would be possible were it sourced
from multiple suppliers.
Under the concurrent cost management
approach, cost negotiations between Toyo and
Komatsu began earlier in the development process
while the product was still in the conceptual
stages. Toyo’s engineers estimated the cost of
manufacture and, if it appeared too high, tried to
?nd ways to alter Komatsu’s speci?cations so that
the part could be manufactured for its target cost.
Changes in the speci?cations for Toyo parts were
allowed only if the functionality of the ?nal
Komatsu product was not excessively compro-
mised. The aim was to make negotiations sur-
rounding the setting of target costs more
substantive and two-sided. Once they established
these target costs, engineers at both ?rms had to
?nd ways to achieve them. The earlier establish-
ment of the target costs meant that Toyo had to be
more aggressive in its negotiations if it felt that
Komatsu’s established target costs were too low.
Komatsu changed the way it enacted target cost-
ing by bundling all of the costs of the major func-
tion together, thus letting Toyo Radiator
determine the appropriate target costs of the indi-
vidual components it designed and produced.
We observed two fundamental approaches to
concurrent cost management: parallel and simul-
taneous. In the parallel approach, the engineering
teams at the buyer and supplier operate indepen-
dently, whereas in the simultaneous approach,
they work together to co-design the end product
and the outsourced major function. The perceived
bene?ts from close interactions of the buyer and
supplier design teams determine the choice
between parallel and simultaneous engineering. If
the value of such interactions is considered high,
8 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
then the ?rms use simultaneous engineering;
otherwise, they use parallel engineering.
The primary advantage of parallel engineering is
the ability of the supplier to uncouple its own
product development program from that of the
buyer. For example, under the old approach,
Komatsu would tell Toyo Radiator that the
engine power of a particular product model was to
be increased by X% and that they would require a
new engine cooling system with the appropriately
increased capacity in 12 months. However,
because Toyo Radiator did not know that
increasing engine power was planned for the fol-
lowing generations, it could only react to the
requests for increased engine cooling capacity as
they were received. Under the new approach,
Komatsu informed Toyo Radiator that, for the
next few generations, they would require increased
engine cooling capacity at essentially the same
cost. Toyo Radiator engineers then could deter-
mine that the best long-term solution would be to
launch a research project to develop a more e?-
cient, low-cost approach to engine cooling. The
?rst new technology engine-cooling system might
not be available until the third generation of
Komatsu designs, but at least it would be avail-
able, whereas, under the old approach, Toyo
Radiator would never have had the con?dence to
launch the project in the ?rst place. As long as
Toyo Radiator’s engineers knew the general level
of increased engine power that Komatsu planned,
they needed only to communicate occasionally
with Komatsu’s engineers, typically via periodic
meetings and telephone calls. Thus, in parallel
engineering, the engineering teams operate essen-
tially independently of each other.
In contrast, the aim of simultaneous engineering
is to allow the design teams to cooperate inten-
sively during the early stages of the design process
so that they can co-design both the end product
and the outsourced item with the objective of
?nding ways to deliver the desired level of func-
tionality and quality of the end product at its tar-
get cost. Successful co-design allows the two
design teams to achieve solutions that they could
not achieve separately. Komatsu and Toyo
Radiator demonstrated the process of simulta-
neous engineering with the development of the
mixed ?ow fan, which allowed the generation of a
more e?ective air ?ow, and thus allowed smaller,
lower-cost radiator to supply the same level of
cooling. However, it also required that Komatsu
engineers work closely with the Toyo Radiator
design team to test the adequacy of the fan’s
cooling e?ciency. This testing was carried out
simultaneously with the development of the fan.
Under the old approach, Toyo Radiator would
have completed the development of the new
engine cooling system and then would have pro-
vided prototypes for Komatsu to test. This
sequential approach would have made the devel-
opment of the new fan design too slow to incor-
porate into the current generation, thus delaying
the product launch and losing sales, or retaining
the existing cooling system and engine designs at
higher cost. Both outcomes were considered
unacceptable, as they led to reduced pro?ts at
both Komatsu and Toyo.
Clusters of IOCM practices
The three observed IOCM techniques were
associated with di?erent magnitudes in the design
changes of the items produced by the interacting
?rms. FPQ tradeo?s are associated with small
design changes that can be accommodated by a
single ?rm in the supply chain with the permission
of at least one other ?rm. Interorganizational cost
investigations are associated with more signi?cant
changes that require modi?cations to the design or
production processes of the items produced by
more than one ?rm in the chain. These design
changes are interrelated, but they can be accom-
plished with relatively low levels of communi-
cation between the design teams. Finally,
concurrent cost management is associated with the
most signi?cant changes. Frequently, these chan-
ges are so substantial that the designs of both the
buyer and supplier’s products must be modi?ed in
a compatible fashion. Although the scope and
magnitude of the changes to the design of the
products constitute a continuum, ranging from
small, nearly imperceptible changes to large chan-
ges obvious to the customer, the observed IOCM
practices consist of three discrete clusters (see
Table 1).
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 9
The ?rst cluster captures the ?rms that were
capable of performing all three IOCM techniques.
For example, one of the product development
projects studied was subjected to both a con-
current cost management intervention and an
interorganizational cost investigation. In the sec-
ond cluster, the ?rms were able to perform inter-
organizational cost investigations and FPQ trade-
o?s, but could not undertake concurrent cost
management. The third cluster captures the ?rms
that had the lowest capacity to perform IOCM.
Here, we observed only FPQ trade-o?s. Managers
at the three top ?rms con?rmed that these patterns
of IOCM capability were replicated throughout
their supplier networks.
Observed relational contexts
Managerial accounting practices cannot be
understood in isolation from the broader organi-
zational settings in which they occur (Hopwood,
1983). In our sample, we observed ?ve di?erent
relational contexts in which IOCM occurs. One of
these contexts ?ts into the pure market perspec-
tive. This context was observed for external sup-
pliers that sold standard products, such as nuts
and bolts, to multiple customers. Another context
maps into a pure hierarchy perspective where the
?rms internally source critical parts; for example,
Komatsu internally sources the engines and
hydraulic systems for its bulldozers and excava-
tors. The other three relational contexts were
hybrid forms that did not ?t into the classical
dichotomy of markets and hierarchies. Since these
hybrid relational contexts were associated with
IOCM, they were the only ones that we studied in
depth. Individuals at both the buyer and supplier
?rms used the common supplier context as the
reference point for their descriptions of the other
three contexts; consequently, we also collected
baseline information about that context.
Hybrid organizations re?ect a wide range of
decisions on the part of the management about the
location of the ?rms’ boundaries and the nature of
the governance structure that controls the rela-
tionship (Dyer & Singh, 1998; Gulati & Singh,
1998; Zaheer & Venkatraman, 1995). However,
for the study of IOCM, we can limit the analysis
to those elements of the organizational setting that
facilitate interactions across the organizational
boundary between buyers and suppliers. Of partic-
ular interest is the way that this relational context
varies in ?rms that undertake di?erent forms of
IOCM. While relational contexts might be expec-
ted to vary continuously from markets to hier-
archies, in practice, a limited number of distinct
relational contexts appear to develope. Six attri-
butes are of particular importance when deter-
mining the appropriate relational context for
IOCM. Four of these attributes—design depen-
dence, resource sharing, supplier participation,
and bilateral commitment—relate to interaction
characteristics of the buyer–supplier relationship.
The ?nal two—incentive and protective mech-
anism—pertain to the choices surrounding the
governance structure of the buyer–supplier
relationship.
Design dependence
According to transaction cost economics, rela-
tionships between buyers and suppliers are char-
acterized by some degree of mutual dependence
contingent upon the speci?c investments made and
the switching costs they entail (Williamson, 1985,
1991). In the context of product development,
mutual dependence arises in the form of design
dependence. Design dependence transpires when
the buyer and supplier split responsibility for the
establishment of the outsourced item’s speci?ca-
tions and/or design. The highest level of design
dependence occurs when the supplier and the
buyer establish joint speci?cations and take joint
responsibility for product design. Under these
conditions, the two ?rms must actively integrate
their product development processes:
Table 1
Clusters of IOCM techniques practiced
Ability to
perform
IOCM
FPQ
trade-o?s
Interorganizational
cost investigations
Concurrent cost
management
High Observed Observed Observed
Medium Observed Observed Not observed
Low Observed Not observed Not observed
10 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
Under the new supplier approach, Toyo
Radiator could negotiate with us to move the
condenser and hence delete a complete fan
and motor assembly from the new design.
Komatsu Design Engineer
The next level of design dependence occurs
when the supplier accepts responsibility for design
and manufacture, but the buyer retains sole
responsibility for establishing high-level speci?ca-
tions. Here, the level of integration is lower, but
still demanding, since the two ?rms must ensure
that the end product and the outsourced item are
compatible. Finally, design dependence is low
when the buyer both establishes the speci?cations
and takes responsibility for design, and the sup-
plier only accepts responsibility for manufacture.
Here, the buyer must ensure that the outsourced
components are designed in a way that enables the
supplier to manufacture them for a reasonable
cost. The supplier has few additional responsi-
bilities other than ensuring that the parts are
delivered on time and on spec.
The improved performance that Komatsu
demanded for its A20 and A21 power shovels
triggered a change in relational context at
Komatsu–Toyo. The new designs required engine-
cooling systems, the radiator size of which was
36% larger than that of the previous generation.
Normally, this increase in size would have raised
Toyo’s costs by approximately the same percen-
tage. However, Komatsu’s customers were only
willing to pay about half of the extra costs for the
proposed increased performance. Komatsu man-
agement realized that the target cost it was forced
to set for Toyo (118% of the prior generation’s
cost) was too aggressive to be achieved without
changing the way the two ?rms interacted. Hence,
it adopted a new relationship with Toyo char-
acterized by higher levels of design dependence.
Under the new relationship, the interdependence
between Komatsu and Toyo was largely recipro-
cal, as both partners exchanged outputs with each
other simultaneously, and the output of each
partner became the input of the other (Gulati &
Singh, 1998; Thompson, 1967). Komatsu used the
term ‘‘family member’’ to di?erentiate its suppliers
with the highest levels of design dependence from
other types of suppliers. Toyo had previously been
a ‘‘major supplier,’’ the term Komatsu used for
suppliers that took responsibility for the design of
the group component they supplied, but not for
the establishment of its speci?cations.
Komatsu used two additional terms to di?er-
entiate between the other types of relational con-
texts it had established with external suppliers,
namely ‘‘subcontractors’’ and ‘‘common suppli-
ers.’’ Subcontractors manufacture outsourced
items that are designed by the buyer. They have
few internal design capabilities, but are typically
highly skilled at manufacturing specialized items.
Consequently, they only accept responsibility for
manufacturing, whereas the buyer is responsible
for establishing speci?cations and designing the
part. Finally, common suppliers typically publish
catalogues that detail their product o?erings.
Common suppliers take responsibility for all
aspects of the items’ design and manufacture.
Consequently, there is no design dependence
between Komatsu and its common suppliers.
Similar relational contexts were observed at the
other ?rms in the sample that occupied the top
position in their supply chains. Table 2 illustrates
the relationship between the level of design
dependence—determined by the responsibility for
establishing speci?cations and product design—
and the relational context.
These ?ndings corroborate earlier research that
focuses on analyzing and categorizing relation-
ships between buyers and suppliers. Asanuma
(1989), for example, identi?ed three types of sup-
pliers based on their responsibilities in product
design, which are equivalent to major suppliers,
subcontractors, and common suppliers; however,
he does not discuss a relational context similar to
family members. Other researchers have identi?ed
types of buyer–supplier relationships in addition
to those discussed in this paper, some of which
appear to be motivated by forms of joint action
other than IOCM (Asanuma, 1989; Bensaou &
Venkatraman, 1995; Kamath & Liker, 1994).
Resource sharing
The observed various relational contexts
enabled the buyer and supplier to share di?erent
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 11
proportions of their design-related resources. The
objective of any increased sharing was to enable
more sophisticated interactions between the two
design teams. The resource sharing took two
major forms. The ?rst was increased asset speci?-
city, while the second was increased sharing of
strategic information. Previous research also has
highlighted the importance of both aspects of
resource sharing to the e?ectiveness of buyer–
supplier relationships (Dyer, 1996, 1997; Hines,
1994; Lamming, 1993; Zaheer & Venkatraman,
1995).
Asset speci?city
Asset speci?city relates to the degree to which
an asset can be redeployed to alternative uses
without sacri?ce of productive value (Williamson,
1979). Most of the extant literature on buyer–
supplier relationships has attributed a central role
to the various types of asset speci?city, in line with
Williamson’s transaction cost economics theory
(Dyer & Ouchi, 1993; Dyer & Singh, 1998;
Mudambi & Helper, 1998; Nishiguchi, 1994). In
our sample, Komatsu and Toyo altered the level
of asset speci?city associated with their relation-
ship in order to increase the level of joint design
changes. The two ?rms had previously invested in
assets that were speci?c to their relationship—for
example, the development of a proprietary stacked
?n oil cooler, the investment in dedicated produc-
tion lines, and the use of guest engineers. How-
ever, we observed several instances of increased
asset speci?city. First, the level of physical asset
speci?city increased both through the develop-
ment of proprietary software to model engine
cooling and through the sharing of prototyping
assets. In addition, the new approach to product
development often led to new solutions that were
proprietary to the Komatsu–Toyo relationship
(such as a new mixed ?ow fan). Manufacture of
these new proprietary parts required additional
dedicated equipment at Toyo.
More important, the level of human asset speci-
?city increased signi?cantly. First, more indivi-
duals were involved in the joint product
development process. For example, engineers at
both ?rms worked jointly to improve the overall
performance of future products. As one manager
at Toyo Radiator put it:
Our new relationship with Komatsu requires
us to dedicate more resources, especially
engineers, to their products. For example, in
the past, only design engineers would visit
Komatsu; now we want cost engineers to visit
as well.
Second, the individuals involved in the joint
development process were more intensely
involved. For example, the guest engineers had
added cost considerations to their design activities.
Third, the design activities were more speci?c to
Komatsu than before, as the new designs were
now the outcome of joint product development.
Finally, the location where individual engineers
spent their time changed. In particular, the engi-
neers spent more time co-located.
4
Much of the
skills and knowledge created through this
increased interaction was viewed as proprietary;
for example, Toyo was not allowed to change the
way that it dealt with the prototypes of the engine
cooling systems of its other customers in the near
future. Thus, for at least the foreseeable future,
these investments had no appreciable value out-
side the relationship.
The observations at the other ?rms in the sam-
ple also indicated that the level of both physical
and human asset speci?city varied across the dif-
ferent relational contexts. Only when the suppliers
Table 2
Design dependence and relational context
Level of
design
dependence
Predominant
speci?cations
responsibility
Predominant
design
responsibility
Relational
context
High Joint Joint Family member
Medium Buyer Supplier Major supplier
Low Buyer Buyer Subcontractor
None Supplier Supplier Common supplier
4
There were other ways that resources were shared among
the network ?rms, such as employee placement. However, since
these were not related to cost management, they were not
documented for the purpose of this paper.
12 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
were considered major suppliers or family mem-
bers did we observe signi?cant levels of design-
related asset speci?city. As an illustration, Isuzu
provided considerable design support to its major
supplier, JKC. This assistance included the reso-
lution of manufacturing problems created by new
Isuzu parts with complicated shapes that were
di?cult to machine. However, none of the ?rms
dedicated any signi?cant design-related assets to
subcontractors or common suppliers.
Strategic information sharing
Achieving cooperative buyer–supplier relation-
ships requires intensive bilateral communication
and information sharing to engender appropriate
levels of learning and trust (Carr & Ng, 1995;
Dyer & Singh, 1998; Lamming, 1993). Increased
sharing of strategic information played an impor-
tant role in enabling Komatsu and Toyo to e?ec-
tively undertake joint product development
projects. E?ective co-design became possible
because Komatsu and Toyo were sharing con-
siderable information about each others’ design
plans early in the product development process.
For example, right at the beginning of the devel-
opment process, Komatsu told Toyo that its plans
for the new A20 and A21 power shovels included a
40% increase in engine power. Early conveyance
of such strategic information was necessary to give
both Komatsu and Toyo adequate time to jointly
make the signi?cant changes in their product
designs necessary for the project to achieve its cost
objectives:
As a family member, we are told about
Komatsu’s long-term development plans a lot
sooner than we used to be. This earlier noti-
?cation allows us to match our research
e?orts with Komatsu’s and ensure that di?-
cult technical and cost challenges created by
Komatsu’s product plans are overcome.
Toyo Radiator Manager
Under family membership, Komatsu and Toyo
shared an extensive range of information with
each other, despite being separate ?rms. Komatsu
had access to all of Toyo’s cost information
regarding Komatsu-related products, even knowing
the price that Toyo paid for a single bolt used in a
Komatsu product. The objective of this cost
information sharing was to allow Komatsu to ?nd
new ways to reduce costs. For example, it might
increase discounts by purchasing the bolt centrally
and have the bolt manufacturer deliver it directly
to all its users in the Komatsu group and its family
members.
At the other ?rms in the sample, the level of
strategic information sharing also varied across
the three relational contexts. While still sharing a
considerable amount of information with their
major suppliers, the buyers did not share as much
strategic information as they did with their family
members. When the suppliers were considered
subcontractors, little strategic information sharing
was necessary, as design responsibility resided
with the buyer, while the supplier needed only to
manufacture the part. Strategic information shar-
ing was markedly absent for common suppliers.
Here, the relationship was essentially arm’s-length
and market-based, and the two ?rms shared
almost no knowledge outside of price and delivery
information.
Supplier participation
The increased resource sharing between
Komatsu and Toyo enabled the suppliers’ product
engineers to play a more substantive role in the
development of the buyer’s products. However, in
order to achieve this objective, the relational con-
text was modi?ed in two major ways. First, more
substantive items and, in particular, the research
and development associated with them, were out-
sourced. Second, family members became involved
at an earlier stage in the product development
process so that they could have more time to
identify innovative, low-cost designs.
Outsourcing more substantive items
In the case of Toyo Radiator, Komatsu man-
agement decided to outsource most of the research
and development and all of the manufacture of the
engine cooling system. Previously, when only sim-
ple components were transferred, Komatsu had
undertaken the majority of the research and
development in-house, had purchased discrete
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 13
components from Toyo and other radiator manu-
facturers, and then had assembled the engine
cooling system in-house. By outsourcing more
substantive items—primarily major functions—
Komatsu was able to rely more heavily upon the
design skills of its new family members. An
important outcome of this outsourcing policy was
a reduction in the number of suppliers. As one
Komatsu manager put it:
Toyo used to produce only the radiator and
other ?rms produced the fan and electric
motor. However, split design made it di?cult
to increase e?ciency, so we decided that
under the new supplier program, Toyo would
produce the motor, fan, and radiator in one
integrated package.
The value and complexity of the outsourced
items varied by relational context. For family
members, the outsourced items were typically
major functions, such as engine cooling systems.
In contrast, for major suppliers, they were group
components (for example, radiators). For sub-
contractors, the items typically outsourced were
simple components, such as radiator fan blades.
Finally, for common suppliers, the outsourced
items were standard components, such as nuts and
bolts. This observation is in keeping with the
results from an empirical study by Heide and John
(1990), who found that closer relationships
between buyer and supplier are associated with
higher values of the outsourced parts as a percen-
tage of the end product.
Timing of supplier involvement
In the case of the A20 and A21 power shovels,
interaction between the design teams of Komatsu
and Toyo Radiator began 12 months earlier than
it would were Toyo still a major supplier. This
additional time was critical, as it allowed Toyo
and Komatsu to consider more fundamental
redesigns of the cooling system and engine.
Previously, we waited until Komatsu pre-
sented us with work to perform before we
began design. Unfortunately, the contracts
were signed too late in the overall design
process to give us the time we really needed to
design low-cost solutions into our products.
Now we work together much earlier in the
process and can propose more substantive
design changes.
Toyo Radiator President
These ?ndings are consistent with previous
research that associates early and extensive sup-
plier involvement with a faster and more e?cient
product development process (Birou & Fawcett,
1994; Clark & Fujimoto, 1991; Ward, Liker, Cris-
tiano, & Sobek, 1995). The buyer–supplier inter-
actions for the other relational contexts occurred
later in the product development process. Typi-
cally, major suppliers were brought in after pro-
duct conceptualization was complete and the
design of the product nearly ?nalized. For sub-
contractors, the involvement occurred even later,
usually after the parts list had been generated. For
common suppliers, there was essentially no inter-
action until mass production was scheduled and
orders were placed. Thus, the timing of the design
interactions between buyer and supplier appeared
relate to the type of relational context and the
magnitude of cost reduction envisioned.
Bilateral commitment
The increased resource sharing and reliance
upon the supplier’s product development skills
associated with some of the observed relational
contexts required that the buyer and supplier
strengthen their bilateral commitment. They
achieved this objective by increasing the stability
of their relationship and the degree of collabora-
tion between the two design teams.
Stability
Stability relates to the bilateral expectation of
continued future interactions between the buyer
and supplier (Heide & John, 1990). The under-
lying stability of the buyer–supplier relationships
observed at the sample ?rms varied with the rela-
tional context. Previously, when Toyo was sup-
plying relatively simple parts designed by
Komatsu, it was easy for Komatsu to ?nd other
?rms to supply those parts, although it typically
14 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
would choose not to do so. Under the new
approach, Toyo supplied Komatsu with highly
specialized parts that relied heavily upon speci?c
knowledge of Komatsu’s products. Therefore,
Komatsu was far more dependent upon Toyo
than they were previously. Because of the high
switching costs, they worked to maintain a stable,
productive relationship. Furthermore, Toyo had
come to rely upon Komatsu for a larger share of
its business, special design skills, and engineering
support. Consequently, it was also less likely to
consider severing the relationship. The two ?rms
were thus highly interdependent and therefore
were more willing to support each other. Toyo
could rely upon Komatsu to continuously give it a
steady stream of business, while Komatsu could
rely upon Toyo to develop customized, innovative
engine cooling systems. The stability of the rela-
tionship was further enhanced by the reduction in
the number of suppliers that accompanied the
adoption of a more integrated buyer–supplier
relationship.
Increased stability is important because it takes
considerable time for the buyer and supplier to
develop the mutual in-depth familiarity required
for the high level of e?ective co-design involved in
family membership. This knowledge includes joint
technical expertise, mature personal relationships,
and extensive prior experience. For example, over
time, Toyo engineers had developed considerable
expertise speci?c to Komatsu. While other engine-
cooling suppliers existed, they lacked this
Komatsu-speci?c know-how and technology. It
would take considerable time and resources on the
part of Komatsu to match the capabilities of Toyo
Radiator.
At the major supplier level, the relationships
were still quite stable. Only if a major supplier
consistently failed to compete would it cease to
remain in that supplier group. In contrast, at the
subcontractor level, the buyer ?rms were willing to
take back in-house certain items that they pre-
viously had outsourced if there was insu?cient
work to keep their own workforce occupied.
Finally, for common suppliers, there was very lit-
tle stability, as the buyers would simply select the
common supplier from a group of certi?ed ?rms
that gave them the best value for each item.
Collaboration
Collaboration between buyers and suppliers is
considered a necessary condition for e?ective
buyer–supplier interactions that go beyond pure
arm’s-length, contractual relationships (Dyer,
1996; Gietzmann, 1996; Heide & John, 1990). The
relationship between Komatsu and Toyo Radiator
had always been collaborative. However, when
Komatsu designed the engine cooling system itself,
its design engineers did not have to interact so
extensively with Toyo’s engineers. Toyo and the
other major suppliers were manufacturing group
components and their primary task was to achieve
high quality and on-time delivery at the right
price. Komatsu’s primary task was to develop
speci?cations, identify capable suppliers, review
their bids, and accept the highest value bid. In
contrast, when Toyo became responsible for the
development and manufacture of entire engine
cooling systems, the product development engi-
neers of the two ?rms had to work closely together
to identify new approaches to engine cooling
technology that might involve concurrent changes
in the design of the engine. Consequently, the new
buyer–supplier relationship created a culture of
intense collaboration.
Toyo Radiator is closely related to Komatsu;
it is part of a cooperative group. Only a few
dozen suppliers have such a rich relationship
with Komatsu.
Toyo Radiator President
The relationship between Komatsu and Toyo
achieved this objective by stimulating regular
meetings between the design teams of the two
?rms. The simultaneous redesign of both the
engine and the engine cooling system required
considerable coordination between the two engi-
neering teams, since the product development
process was highly iterative. The aim of these per-
iodic meetings was to integrate the research and
development e?orts of the two groups, to allow
suppliers to provide greater input earlier in the
development process, and to help ensure that cost
reduction negotiations were more substantive.
The other ?rms in the sample also collaborated
and helped each other overcome cost problems
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 15
during product development. For example, if one
of the ?rms encountered an engineering challenge,
it was not unusual for engineers from other ?rms
in the chain to help solve the problem. Isuzu, in
particular, used value engineers to help its suppli-
ers resolve cost problems. Similarly, the design
teams of all of the ?rms in the Tokyo–Yokohama–
Kamakura chain often got together to identify
ways to jointly reduce costs.
The magnitude and intensity of the collabora-
tion varied with the relational context. For exam-
ple, since subcontractors were not responsible for
any of the design aspects of the parts they manu-
factured, the help they received was limited to the
manufacturing process. In addition, the buyer did
not look to subcontractors for any signi?cant
help, as there was little additional value that they
could provide to the buyer in terms of design sup-
port. Finally, with common suppliers, no assis-
tance was o?ered by either party.
Governance structure
To reap the full bene?ts of enhanced buyer–
supplier relationships, one must consider neces-
sary shifts in governance structure (Gietzmann &
Larsen, 1998). Our observations showed that, as
the relational context between Komatsu and Toyo
changed, the governance structure altered accord-
ingly. Governance structure is de?ned here as the
mechanisms that create both incentives (i.e.
reward and coercion mechanisms) for the buyer
and supplier to interact, and safeguards that pro-
tect each transactor against the risk of opportu-
nistic behavior on the part of the other
(Williamson, 1979, 1985). In particular, our sam-
ple ?rms’ governance structure relied heavily upon
trust rather than the classical disciplining
mechanisms of authority and price (Nooteboom,
Berger, & Noorderhaven, 1997).
Trust played an important role in the relational
contexts associated with both family members and
major suppliers, unlike the other two relational
contexts of subcontractors and common suppliers.
The key role of trust was to stimulate innovation
among the various ?rms in the supply chain. For
example, the relational context and its associated
governance structure allowed Komatsu and Toyo
to ?nd new ways to integrate the engine and its
cooling system for the A20 and A21 power sho-
vels. Our observations regarding the role of trust
and innovation are in line with prior research
?ndings that demonstrated how low trust rela-
tionships fail to stimulate new ideas (Korczynski,
1996). In addition, it has been suggested that nei-
ther the classical hierarchy nor market relational
contexts support innovation:
The hierarchy/authority mode of inter-?rm
relations clearly risks impeding innovation by sti-
?ing the upward ?ow of new ideas from sub-
ordinated suppliers. Their narrow specialization
leaves them without the technological know-how
needed for innovation, and their subordination
leaves them few incentives to contribute innova-
tive ideas to customers. . .The market/price mode
facilitates innovation by creating incentives to
generate new ideas, but his mode, too, impedes
innovation because suppliers and customers of
innovations have di?culty agreeing on a price for
these innovative ideas (Adler, 2001, p. 224).
In contrast, trust-based communities have
indeed been shown to be e?ective when the buyer
needs to encourage the supplier to be both inno-
vative and a source of knowledge (Bensaou &
Venkatraman, 1995; Dyer, 1996; Helper, 1996;
Sako, 1992).
There are three primary sources of trust: famil-
iarity through repeated interaction, calculation,
and behavioral norms (Adler, 2001). In our sam-
ple, two of these sources of trust were essentially
held constant. First, all of the buyer–supplier
relationships between the ?rms had existed for an
extended period of time. The ?rms were familiar
with each other and had built trust through their
repeated interactions. Second, they had developed
norms of behavior that were ‘‘binding.’’ All of the
?rms studied had demonstrated the ability to con-
sistently deliver high quality products in a timely
and accurate manner; therefore, they had already
established a strong reputation as being ‘‘good’’
players (where reputation is de?ned in terms of
repeated consistency over time). Since none of the
?rms in the sample failed to maintain their repu-
tations across the period of observation, we did
16 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
not observe the role of reputation as a di?er-
entiating, self-enforcing safeguard in the buyer–
supplier relationship.
Thus, it was primarily the level of calculative trust
that varied across the observed relational contexts.
Calculative trust develops when each of the parties
undertakes a sober assessment of the costs and ben-
e?ts of exploiting any vulnerability of the other
party, and determines that the calculus favors main-
taining the relationship on both sides (Adler, 2001).
The ?rms in the family member and major supplier
relationships were observed to actively manage
their calculative trust to create adequate self-
enforcing incentive and protection mechanisms.
Incentive mechanisms
The nature and importance of incentive
mechanisms varied across the relational contexts.
For family members, the primary incentive
mechanism was trust-based and took the form of
mutual bene?t. Family members relied largely on
the principle of mutual bene?t by actively working
together to increase the joint economic bene?t
from their cooperative product design e?orts.
Furthermore, the need to maintain an adequate
level of trust played a role in ensuring that some of
the reward and coercion mechanisms remained
e?ective. For example, the incremental value cre-
ated by the relationship had to be shared to some
extent—although not necessarily equitably,
according to both parties—if trust were to evolve
(Contractor & Lorange, 1988). Thus, Komatsu, as
the more powerful ?rm in the chain, had to ensure
that it did not reap all of the additional pro?ts
generated through collaboration.
Sometimes our sharing of cost information,
coupled to our knowledge of Toyo’s pro?ts,
can lead to a con?ict of interest, with pressure
building within Komatsu to reduce target
costs where Toyo’s pro?ts are known to be
high. However, we share a common goal—
getting costs as low as possible—which
ensures that these con?icts rarely become
serious. To reduce the incidence of such con-
?icts, we do not set our targets costs for parts
manufactured by Toyo based upon our
knowledge of Toyo’s costs. Instead, we try to
set our target costs independently of Toyo
and let Toyo make as much pro?t as possible.
Komatsu Purchasing Manager
At the major supplier level, the principle of
mutual bene?t was still operative, but it was less
important. For major suppliers such as Toyo
(before it became a family member), JKC, and
Yokohama, the primary incentive mechanism was
the volume of business that the buyer gave the
supplier. For example, Isuzu used direct competi-
tion between its major suppliers to ensure that the
suppliers were as innovative as possible. When a
supplier failed to remain competitive, Isuzu pun-
ished that ?rm by awarding it slightly less volume
than in previous years. Similarly, it awarded
innovative suppliers with slightly more volume
than in previous years. To help the poorly per-
forming suppliers become competitive, Isuzu pro-
vided them with additional engineering support.
This observation is in line with Helper (1996), who
found that Japanese ?rms typically opt for the
‘‘voice’’ strategy of joint problem solving, rather
than following the ‘‘exit’’ strategy of ending rela-
tions with poorly performing suppliers.
For subcontractors, the primary incentive
appeared to be continued business. While severing
relationships was rare, it was understood that the
supplier would only be retained as a subcontractor
if the ?rm maintained adequate performance
levels. At the subcontractor level, the principle of
mutual bene?t was still in operation, but played a
relatively minor role. It was enacted primarily by
the ?rms sharing engineering expertise where ben-
e?cial. For example, the buyer might provide the
supplier with engineering support to resolve partic-
ularly di?cult manufacturing problems associated
with outsourced items so as to ensure the achieve-
ment of the supplier’s target costs. At the common
supplier level, the only incentive that appeared to be
in e?ect was the economic bene?t both sides derived
from a market price-based transaction. The princi-
ple of mutual bene?t did not appear to operate.
Protection mechanisms
To mitigate against the risk of opportunistic
behavior by their trading partners, the ?rms relied
upon a number of protection mechanisms, which
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 17
varied with the relational context. In trust-based
relationships, the dominant risk is unilateral
defection (Granovetter, 1985). One way that ?rms
can signal a low risk of defection is by structuring
the relationship so that the commitment of both
sides is clearly observable (Parkhe, 1998). In the
case of family members, trust was maintained by
ensuring that both parties were visibly mutually
interdependent. For example, Komatsu openly
relied upon Toyo Radiator for its expertise in
engine cooling systems, while Toyo Radiator
openly relied upon Komatsu for a signi?cant por-
tion of its business and for engineering support to
develop new technologies. Thus, for family mem-
bers, the dominant protection mechanism was
mutual interdependence. This high level of mutual
interdependence led to barriers to unilateral defec-
tion and created an additional safeguard against
opportunistic behavior. More speci?cally, Toyo had
access to Komatsu’s future product plans, which
contained highly valuable information for Komat-
su’s competitors. However, Komatsu in turn had
access to highly proprietary information about
Toyo and, if Toyo were to defect, Komatsu could
retaliate by sharing that information with Toyo’s
competitors. Within these close relationships, per-
sonal connections also play an important role in
reducing the risk of opportunistic behavior (Ring
& Van de Ven, 1989, 1994). For example, both
sides viewed the new relationship as akin to a
strong friendship, an example of a self-enforcing
safeguard that reduces the need for legal and other
formal protection mechanisms (Dyer, 1997).
Komatsu is a very important customer with
whom we have a highly supportive relation-
ship that can be likened to a very strong
friendship.
Toyo Radiator President
For major suppliers, mutual interdependence
was still a major protection mechanism, but less
signi?cant than it was for family members. The
sequential nature of the design process across the
interorganizational boundary for major suppliers,
coupled with the existence of multiple competing
suppliers, reduced the dependence of the buyer
upon the supplier. However, major suppliers still
had extensive and specialized knowledge regarding
the needs of their customers. Therefore, it was not
feasible for the buyer to switch from an existing
major supplier to a new one from outside their
supplier base. Furthermore, the small number of
?rms in the supplier base that could make a given
family of products made it virtually impossible for
the remaining major suppliers to immediately
expand production to o?set the loss of capacity.
Thus, the buyer had visibly rendered itself depen-
dent upon each of its major suppliers for a rea-
sonable duration. The major suppliers had
rendered themselves similarly dependent on their
customers because they only transacted with a
limited number of customers in this type of rela-
tionship.
5
Therefore, they would su?er consider-
able economic hardship by the loss of a customer
to whom they were a major supplier. Since it was
di?cult to create new major supplier relationships
in the short-term, the suppliers were equally com-
mitted for a reasonable period in the future. Thus,
major suppliers and their customers were mutually
interdependent, although not to the same extent as
family members and their customers.
For subcontractors, the level of mutual inter-
dependence was much lower than that of family
members and major suppliers. The buyer typically
dealt with multiple subcontractors and could
easily compensate for the defection of a sub-
contractor. Finally, for common suppliers, there
was little mutual interdependence, since the buyer
could go to any number of equivalent ?rms for the
outsourced items and, typically, the buyer repre-
sented only a small portion of the supplier’s over-
all volume. Here, the dominant protection
mechanism was market price.
Levels and nature of trust
The level and nature of trust di?ered across the
relational contexts at the sample ?rms. Our
observations suggest that the variation in the level
of calculative trust can be viewed as the result of
the interaction between mutual bene?t and mutual
interdependence. For family members, mutual
5
Some of the suppliers transacted with other types of cus-
tomers; for example, it was not usual for some of the major
suppliers to also be subcontractors to other customers.
18 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
bene?t and mutual interdependence were both
high and, consequently, so was the level of calcu-
lative trust. In contrast, the level of mutual bene?t
and interdependence decreased for major suppliers
and even more so for subcontractors; overall trust
was also observed to be low, but still signi?cant.
Finally, for common suppliers, the level of calcu-
lative trust was close to zero, as mutual bene?t
and interdependence were essentially nonexistent.
The nature of the trust engendered by the
observed relational contexts also di?ered. Trust
between Komatsu and its family members took
the form that has been described in the literature
as ‘‘goodwill trust’’ (Sako, 1992; Sako & Helper,
1998). Goodwill trust is characterized by the will-
ingness of both parties to go beyond the contract
and to act in the best interest of the other party,
even at a slight disadvantage to themselves. Such
trust can also be described as taking the strong
form; i.e. while the ?rms were signi?cantly vul-
nerable to each other, they were so committed to
the relationship that they had internalized values,
principles, and standards of behavior that pro-
tected the relationship (Barney & Hansen, 1994).
These ?ndings are in keeping with earlier research
?ndings that concluded that self-enforcing
mechanisms are more e?ective than third-party
enforcement mechanisms (such as contracts) at
both minimizing transaction costs and stimulating
value-creation initiatives (Dyer & Singh, 1998).
Komatsu and its major suppliers were not
mutually interdependent to the same extent as the
?rm was with its family members, nor were the
mutual bene?ts as high. In these relationships,
trust took the form of competence trust (Sako,
1992) or the semi-strong form of trust (Barney &
Hansen, 1994). Competence trust implies that the
buyer believes that the supplier has the compe-
tence to complete the order and need not be mon-
itored during or after the process. The observed
relationships between buyers and subcontractors
were less dependent upon self-enforcing safe-
guards than those observed with major suppliers
and, in particular, family members. Instead, the
predominant form of protection was through the
written contracts designed to maintain those rela-
tionships and through the risk of termination for
poor performance. Thus, the maximum level of
trust between the two ?rms is ‘‘contractual trust’’
(Sako, 1992), or ‘‘weak form trust’’ (Barney &
Hansen, 1994). This form of trust relies upon the
mutual expectation that promises of a written or
verbal nature will be kept. In other words, the two
parties assume that each will adhere to the con-
tract without necessitating any coercion.
The dominant protection mechanism between
the buyers and common suppliers was price. The
buyers would typically choose from a limited
number of common suppliers, depending upon
which one was perceived as o?ering the best value.
If a common supplier failed to deliver upon a
given contract, they were dropped from the list of
acceptable suppliers. Thus, there is no calculative
trust between the two ?rms, and legal action is the
sole remedy for non-performance.
Thus, when few or no bene?ts are to be derived
from joint action, the ?rms in a supply chain avoid
becoming interdependent upon each other and
thus reduce the need for self-enforcing safeguards.
Instead, the ?rms rely upon arm’s-length market
transactions and written contracts. Arm’s-length
contracts work well when the interactions between
the parties are straightforward, but not when they
are highly interdependent. In the former case, only
the quantity, price, quality, and delivery times
need to be speci?ed, whereas in the latter it is
almost impossible to specify a priori all of the
actions that might be required and de?ne payment
schemes for them. According to transaction cost
economics, complex contracts are inevitably
incomplete because it is impossible or too costly to
contract upon all future contingencies as a result
of bounded rationality (Williamson, 1979). There-
fore, an alternative, non-contractual, and more
?exible form of governance based on trust is
required where the two parties act bilaterally to
maximize the joint return.
The relational contexts that we observed are not
simple hybrids of the classical market and hier-
archy contexts. The original markets and hier-
archies model assumed that that there were only
two discrete relational contexts and that trust
played no role in business transactions (William-
son, 1991). This highly constrictive view was later
replaced by a ‘‘swelling middle’’ perspective
whereby the hybrid relational contexts were linear
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 19
mixtures of the two classical ones (Zenger & Hes-
terly, 1997). However, in our sample, the natural
progression is not on a continuum from market to
hierarchy, as a third factor, trust, is getting both
stronger and more encompassing. With the accep-
tance of the role of trust, hybrid relational contexts
can be viewed either as the outcome of a three-way
trade-o? (Ouchi, 1980) or as solutions in a three-
dimensional space consisting of hierarchy/authority,
market/price, and community/trust (Adler, 2001).
IOCM clusters and relational contexts
Each of the three observed IOCM clusters was
associated with a distinct hybrid relational context.
There are two aspects to this association. First, the
motivation behind the link appears to be the level
of interorganizational cooperation and coordina-
tion required by the various IOCM techniques on
the one hand, and the level of cooperation supported
by the relational contexts on the other hand. The
more demanding the technique, the further the rela-
tional context is removed from a pure market. Sec-
ond, the perceived e?ect on joint performance varies
with the IOCM technique and the anticipated
savings apparently play a role in motivating the
adoption of a speci?c type of relational context.
Establishing the linkage between IOCM cluster
and relational context
Our observations suggest that the IOCM tech-
niques practiced by buyers and suppliers are
linked to their relational context. The relational
context of common suppliers was observed to
come closest to the pure market form.
6
If we
de?ne the attributes of the common supplier as the
reference point to compare the three hybrid rela-
tional contexts, as well as their departure from
pure market interactions, then a pattern emerges.
First, each cluster of IOCM techniques was asso-
ciated with a particular relational context, and
vice versa: i.e., we observed a one-to-one relation-
ship between IOCM cluster and relational context.
Second, for all relational context attributes studied
(i.e. those related to design dependence, resource
sharing, supplier participation, bilateral commit-
ment, and governance structure), the sub-
contractor relational context is the closest of the
three to a common supplier, followed by the major
supplier, with the family members being the
furthest removed from the classical market form.
Thus, a systematic, monotonic relationship was
evident in the attributes associated with each rela-
tional context. Third, the IOCM techniques were
associated with di?erent magnitudes in design
changes to the items produced by the interacting
?rms. If we treat the market pricing associated
with common suppliers as the baseline for IOCMon
the basis that no joint design changes are implied,
then the smallest level of design changes are asso-
ciated with the FPQ trade-o?s, the middle level with
interorganizational investigations, and the highest
value design changes with concurrent cost manage-
ment. Thus, a monotonic relationship was observed
again in the magnitude of the design changes and in
the ability to perform IOCM in each cluster.
The observed sets of pairing between IOCM
technique and relational context can now be cau-
tiously interpreted (see Table 3). The higher the
level of design changes envisioned, the more buy-
ers and suppliers are required to interact in a rich
and varied manner. Thus, concurrent cost man-
agement, which involves the largest design chan-
ges, is associated with a family member context.
This context enables the richest interactions
between the design teams of the two ?rms and is
furthest removed from the pure market form. It is
important to observe that family membership and
concurrent cost management evolved simulta-
neously at Komatsu and Toyo Radiator, and that
senior management perceived them as a single
outcome. FPQ trade-o?s, which involve the smal-
lest design changes, are associated with a sub-
contractor relational context, which of all the
hybrid relational contexts observed is the closest
to pure markets. Finally, the middle level of design
changes involved in interorganizational cost
6
There are some minor di?erences in the observed rela-
tional context compared to the theoretical one. For example,
some of the ?rms were loyal to a limited number of common
suppliers and would not switch suppliers simply because of
price considerations. This decision was based in part upon the
perceived bene?ts of long-term relationships in sourcing
decisions, even for common suppliers.
20 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
investigations is associated with the intermediate
relational context, major suppliers. Thus, the three
observed clusters of IOCM practice are associated
with three di?erent relational contexts. The high-
est ability to perform IOCM is associated with
family membership and the lowest ability with a
subcontractor context, whereas the middle IOCM
cluster is associated with the major supplier con-
text. The sample ?rms studied did not undertake
IOCM ?rst and then adapt their relational context to
make it more aligned, nor did they change their
relational context and then subsequently develop the
ability to undertake IOCM. Instead, the relational
context was modi?ed contemporaneously with the
development of the ability to undertake IOCM.
An underlying assumption in this analysis is that
the pressure on the ?rms to undertake IOCM was
su?ciently high to cause them to change their
relational context accordingly; therefore, certain
central associations between IOCM and relational
context can be observed. However, we acknow-
ledge that a large number of factors may well
in?uence the design of relational contexts. Fur-
thermore, we cannot rule out the role of idiosyn-
cratic in?uences on IOCM and relational contexts.
However, since our objective is to highlight the
existence of a pattern of relationships between
IOCM practice and relational context and not to
suggest causal relationships, we can ignore idio-
syncratic in?uences for the purpose of this paper.
Perceived e?ect on joint performance
The managers in the sample ?rms appear to adopt
the IOCM techniques and associated relational
contexts because they believe that these combina-
tions will lead to superior joint performance, as
illustrated by a Toyo Radiator manager:
Under the old approach to supplier relations,
Komatsu would identify the performance
speci?cations for their next model and
develop the basic design concepts. Toyo
would then develop new technologies on its
own and present Komatsu with their ideas of
how to achieve the basic concept. However,
with the A20 and A21 power shovels, that
approach would not have achieved our
objective. In order to get the costs low
enough, we had to undertake a concurrent
cost management program.
It is extremely di?cult to generate a priori
numerical support for the contention that the
application of IOCM in the relational context of
hybrid organizations is more e?ective than con-
ventional cost management in the context of a
market or a hierarchy. Consequently, the decision to
adopt a new cost management technique can be
viewed as a calculated gamble based upon the
expected bene?ts and costs of both developing the
expertise necessary to undertake the IOCM tech-
nique and creating its associated relational context.
There are two ways to measure the e?ectiveness
of a given IOCM technique. The ?rst way com-
pares the savings to the incremental cost of an
individual IOCM intervention. The second way
compares the total savings associated with a given
IOCM technique from all a?ected suppliers dis-
counted over time, to the cost of creating the
ability to undertake the technique and establishing
the appropriate relational context.
E?ectiveness of individual IOCM interventions
The economic justi?cation of an individual
IOCM intervention compares the anticipated cost
savings against the cost of undertaking the inter-
vention. The anticipated joint savings of a single
intervention, when considered as a percentage of
the overall value of the outsourced item, were
considered to vary with the utilized IOCM tech-
nique. The savings from FPQ tradeo?s were
expected to be modest, representing only a few
Table 3
Relational context and IOCM techniques utilized
Relational context Level of
design change
supported
Predominant IOCM
technique utilized
Family Member High Concurrent cost
management
Major Supplier Medium Interorganizational
cost investigation
Subcontractor Low FPQ trade-o?
Common Supplier None None
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 21
percentage points of the cost of the outsourced
item. For interorganizational cost investigations,
the savings were typically expected to be in the 5–
10% range. Finally, for concurrent cost manage-
ment, the savings were expected to be between 10
and 15%. Since the value of the outsourced item
increases with the distance of the relational con-
text from a pure market situation, so do the
expected cost savings (see Table 4).
Some ex post evidence supports the adoption of
concurrent cost management for the A20 and A21
power shovels at Komatsu and Toyo Radiator.
Using simultaneous engineering and an inter-
organizational cost investigation, the two design
teams managed to overcome a particularly severe
cost problem: an a priori predicted 36% cost
overrun. These savings were achieved by jointly
designing a new engine cooling system and engine,
changing the shape of the cooling fan blades,
reducing the number of fans and electric motors
from two to one, and changing the design of the
diesel engine so that it could accommodate the
new engine cooling system. Komatsu and Toyo
engineers involved in the project believed that the
new approach was essential to the project’s success
in that the ability to reduce a cooling system’s costs
by 18% demonstrated the power of a simultaneous
engineering approach. Neither Komatsu nor Toyo
undertook a formal analysis of the costs and bene?ts
of the intervention. However, the engineers at both
?rms were convinced that the savings far out-
weighed the incremental costs associated with the
concurrent cost management and interorganiza-
tional cost investigations interventions.
E?ectiveness of an IOCM technique
The e?ectiveness of an IOCM technique is cap-
tured by the net present value of the savings
derived from all suppliers using the technique
during a given time frame, minus the cost of
establishing and maintaining the technique and its
associated relational context. None of the ?rms in
the sample had even attempted a formal economic
justi?cation of their IOCM programs. Instead,
they relied upon their perceptions of the economic
success of the programs. For example, Komatsu
management adopted a holistic view of the make-
or-buy decision and the development of con-
current cost management. They believed that the
bene?ts of the new technique, quanti?ed by the
reduction in the costs of outsourced items, excee-
ded any additional investment, as well as any
coordination and transaction costs associated with
the new approach. However, they admitted that
there was no easy way for them to evaluate the net
bene?ts inherent to the new buyer–supplier rela-
tionship. The problem was that, while the
increased investment in speci?c assets theoretically
could be measured, along with their costs, the
incremental coordination and transaction costs
were di?cult to observe and to incorporate into
any formal economic evaluation of the overall
process.
Most of the bene?ts to Komatsu from its new
relationship with Toyo emerged from the
improved design capabilities of the dyadic pair.
These bene?ts included the lower costs and higher
functionality of the end products that resulted
from the joint design activities. The bene?ts to
Toyo, however, were less obvious. While the
value-added and the pro?t margin of an engine
cooling system were indeed higher than that of
Komatsu-designed components, Toyo now had to
invest in a more extensive research and develop-
ment program. Unfortunately, there was no way
to determine if Toyo’s pro?ts from its business
with Komatsu were higher under the new
approach, compared to the old one, because of the
Table 4
IOCM technique utilized and typical level of cost savings
IOCM technique utilized Type of item outsourced Relative value
of item outsourced
Typical level
of cost savings (%)
Concurrent cost management Major function High 10–15
Interorganizational cost investigation Group component Medium 5–10
FPQ trade-o? Component Low 0–5
22 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
problem of isolating the incremental revenues and
costs—in particular, the costs associated with the
new relational context. Neither ?rm had main-
tained detailed records of their prior investments
with one another; therefore, there was no way to
determine the incremental costs. Consequently, a
formal incremental pro?t analysis was not con-
sidered feasible. However, both Komatsu and
Toyo management stated that, in their opinion,
the two ?rms were better o? and that they actively
supported the new relationship. Literature sug-
gests that the decision to outsource is not a simple
trade-o? between increased performance and
increased transaction and coordination costs.
Dyer (1997) has suggested that in the type of rela-
tionship we have characterized as family member-
ship, ongoing transaction costs (and presumably
coordination costs) actually decrease, although
initial set-up costs may be higher.
Re-contracting
The success of Komatsu’s ?rst concurrent cost
management project with its ?rst family member
renders moot any discussion about how the ?rms
would have reacted, had the interaction resulted
into failure. Two possible alternatives suggest
themselves. First, the two ?rms could have
returned to their prior relational context and lim-
ited their IOCM practices to interorganizational
cost investigations and FPQ trade-o?s. Alter-
natively, they could have tried to rede?ne the
relational context so that concurrent cost man-
agement was successfully supported. Discussions
with both management teams reveal that the two
?rms had entered into a calculated gamble that,
together, they could reduce overall costs to a
greater extent than they could in isolation, or even
jointly under the relational context of a major
supplier. The process of achieving the desired sav-
ings for the A20 and A21 power shovels was a
risky and complex undertaking that required mul-
tiple joint design e?orts and encompassed multiple
solutions. The fact that neither ?rm had the tech-
nical expertise to solve the problem on its own was
critical. More speci?cally, the degree to which
overall cost was a function of joint design e?orts
appeared to be an important consideration.
Finally, the perceived inability to solve the cost
challenge using conventional approaches was a
signi?cant motivating factor for adopting a family
membership context and undertaking concurrent
cost management.
Summary and conclusion
The make-or-buy decision at the sample ?rms is
far more complex than the neo-classical economic
perspective suggests. Rather than being a simple
dichotomous competition between the ?rm and its
external suppliers, the process often involves a rich
interplay between buyer and supplier to ?nd ways
to take advantage of their disparate capabilities.
In particular, the sample ?rms have developed
IOCM techniques to overcome the information
asymmetry that exists between buyers and suppli-
ers, and to enable their design teams to coordinate
and cooperate e?ectively in order to identify low-
cost solutions by changing the speci?cations of the
outsourced items and sometimes the end product
itself.
The sample identi?ed three distinct clusters of
IOCM, each of which was associated with a dif-
ferent hybrid relational context. The relational
context established between the buyer and supplier
appeared to be linked to the IOCM technique that
supports the highest level of design changes
undertaken by the two ?rms, and thus involves the
highest level of interaction between the two design
teams. The ?rst cluster contained ?rms that could
undertake all three of the observed IOCM tech-
niques (FPQ tradeo?s, interorganizational cost
investigations, and concurrent cost management);
this cluster was associated with the relational con-
text family member, which provided the richest
support for interactions between buyer and sup-
plier. The second cluster contained ?rms that
could undertake only two of the IOCM techniques
(FPQ trade-o?s and interorganizational cost
investigations); it was associated with the rela-
tional context major supplier, which could support
an intermediate level of interaction between the
design teams at the buyer and supplier ?rms.
Finally, the third cluster consisted of ?rms that
could only undertake FPQ tradeo?s and was asso-
ciated with the relational context subcontractor,
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 23
which only supported low levels of buyer–supplier
interactions.
For ?rms that undertake IOCM, at least two
additional layers of complexity must enter into the
make-or-buy calculus. The ?rst layer deals with
the costs associated with an individual IOCM
intervention, including the time spent by the buy-
er’s engineers discussing and negotiating solutions
with the supplier, as well as the cost of formalizing
any design changes to the buyer’s product(s) and
the outsourced item. The second layer of com-
plexity further distances the make-or-buy decision
from the classical neo-economic analysis of inter-
nal versus external costs of outsourcing by intro-
ducing a higher level of analysis into the calculus
(Anderson, Glenn, & Sedatole, 2000). A single
outsourcing decision will rarely result in savings
su?cient to justify the adoption of a new IOCM
technique and its associated relational context.
Instead, the analysis must be applied to the sav-
ings associated with all of the suppliers likely to be
a?ected by the adoption of the new IOCM tech-
nique discounted over the time period that they
occur. These discounted savings then must be
compared to the incremental cost of developing
and maintaining the ability to undertake the tech-
nique and its associated relational context. Thus,
the calculus not only includes which parts will be
outsourced and to whom, but also the nature of
the relationship between the buyer and supplier
that should be adopted for the transaction.
Because of incomplete contracting, the calculus
has to be expanded to incorporate both bargain-
ing costs, where the parties are acting in their own
self-interest but in good faith, and opportunism
costs, where they are acting in their own self-
interest but in poor faith (Vining & Globerman,
1999).
The small number and limited diversity of the
sample ?rms used in this study re?ects the
exploratory nature of our research. To increase
the validity of the results and to help determine
the extent to which they can be generalized, the
?ndings need to be tested on a larger and more
diverse sample that includes ?rms from di?erent
countries and from a wider range of industries. In
addition, there are several extensions of this
research that appear worthy of further study.
First, in this paper, we have explored only a lim-
ited range of relational contexts and IOCM tech-
niques. One possible extension is to study other
types of relational contexts between ?rms, such as
strategic alliances and joint ventures, in which
di?erent forms of IOCM may emerge. Second, in
this paper, we explored IOCM practices that are
design-driven. However, nothing we observed
appears to be idiosyncratic to product design.
Another logical extension of this research would
be to examine the extent to which other forms of
joint action between ?rms trigger IOCM interven-
tions.
Acknowledgements
We thank Shannon Anderson, Jacob Birnberg,
Robert Kazanjian, Richard Makadok, Haya-
greeva Rao, the referees and participants at the
2000 AAA/IMA Conference, Anthony Hopwood,
the editor, and two anonymous reviewers for
helpful comments on the research.
References
Adler, P. (2001). Market, hierarchy, and trust: the knowledge
economy and the future of capitalism. Organization Science,
12(2), 215–234.
Anderson, S. W., Glenn, D., & Sedatole, K. L. (2000). Sourcing
parts of complex products: evidence on transactions costs,
high-powered incentives and ex-post opportunism. Account-
ing, Organizations and Society, 25(2000), 383–409.
Asanuma, B. (1989, March). Manufacturer-supplier relation-
ships in Japan and the concept of relation-speci?c skill. Jour-
nal of the Japanese and International Economies, 3(1), 1–30.
Atkinson, A. A., Banker, R. D., Kaplan, R. S., & Young, S. M.
(2001). Management accounting (10th ed). Upper Saddle
River, NJ: Prentice Hall.
Baiman, S. & Rajan, M. V. (2000). Incentive issues in inter-?rm
relationships [Working paper]. The Wharton School.
Barney, J. B., & Hansen, M. H. (1994). Trustworthiness as a
source of competitive advantage. Strategic Management
Journal, 15, 175–190.
Bensaou, M., & Venkatraman, N. (1995). Con?gurations of inter-
organizational relationships: a comparison between U.S. and
Japanese automakers. Management Science, 41(9), 1471–1492.
Birou, L. M., & Fawcett, S. E. (1994). Supplier involvement in
integrated product development: a comparison of US and
European practices. International Journal of Physical Distri-
bution & Logistics Management, 24(5), 4–14.
24 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
Bryce, D., & Useem, M. (1998). The impact of corporate out-
sourcing on company value. European Management Journal,
16(6), 635–643.
Carr, C., & Ng, J. (1995). Total cost control: Nissan and its
U.K. supplier partnerships. Management Accounting
Research, 6, 347–365.
Clark, K. B., & Fujimoto, T. (1991). Product development per-
formance. Boston: Harvard Business School Press.
Contractor, F. J., & Lorange, P. (1988). Why should ?rms
cooperate? The strategy and economics basis for cooperative
ventures. In F. J. Contractor, & P. Lorange (Eds.), Coop-
erative strategies in international business. Lexington: Lex-
ington Books.
Cooper, R., & Chew, W. B. (1996). Control tomorrow’s costs
through today’s designs. Harvard Business Review, 1, 88–97.
Cooper, R., & Slagmulder, R. (1997). Target costing and value
engineering. Portland, OR: Productivity Press.
Cooper, R., & Slagmulder, R. (1999). Supply chain development
for the lean enterprise. Interorganizational cost management.
Portland, OR: Productivity Press.
Cooper, R., & Yoshikawa, T. (1994). Inter-organizational cost
management systems: the case of the Tokyo-Yokohama-
Kamakura supplier chain. International Journal of Produc-
tion Economics, 37, 51–62.
Davila, T. (2000). An empirical study on the drivers of manage-
ment control systems’ design in new product development.
Accounting, Organizations and Society, 25(4–5), 383–409.
Drtina, R. E. (1994, March). The outsourcing decision. Man-
agement Accounting (USA), 56–62.
Dyer, J. H. (1996). Specialized supplier networks as a source of
competitive advantage: evidence from the auto industry.
Strategic Management Journal, 17(4), 271–291.
Dyer, J. H. (1997). E?ective inter?rm collaboration: how ?rms
minimize transaction costs and maximize transaction value.
Strategic Management Journal, 18(7), 535–556.
Dyer, J. H., & Ouchi, W. (1993). Japanese-style partnerships:
giving companies a competitive edge. Sloan Management
Review, 35(1), 51–63.
Dyer, J. H., & Singh, H. (1998). The relational view: coopera-
tive strategy and sources of interorganizational competitive
advantage. Academy of Management Review, 23(4), 660–679.
Eisenhardt, K. (1989). Building theories from case study
research. Academy of Management Review, 14(4), 532–550.
Gietzmann, M. B. (1996). Incomplete contracts and the make
or buy decision: governance design and attainable ?exibility.
Accounting, Organizations and Society, 21(6), 611–626.
Gietzmann, M. B., & Larsen, J. G. (1998). Motivating sub-
contractors to perform development and design tasks. Man-
agement Accounting Research, 9, 285–309.
Gilley, K. M., & Raheed, A. (2000). Making more by doing
less: an analysis of outsourcing and its e?ects on ?rm per-
formance. Journal of Management, 26(4), 763–790.
Granovetter, M. (1985). Economic action and social structure:
the problem of embeddedness. American Journal of Sociol-
ogy, 91, 481–510.
Gulati, R., & Singh, H. (1998). The architecture of coopera-
tion: managing coordination costs and appropriation con-
cerns in strategic alliances. Administrative Science Quarterly,
43(4), 781–815.
Ha? gg, I., & Hedlund, G. (1979). ‘‘Case studies’’ in accounting
research. Accounting, Organizations and Society, 4(1/2), 135–
143.
Heide, J. B., & John, G.(1990).Alliances in industrial purchas-
ing: the determinants of joint action in buyer-supplier rela-
tionships (1990). Journal of Marketing Research, XXVII, 24–
36.
Helper, S. (1996). Incentives for supplier participation in pro-
duct development: evidence from the U.S. auto industry. In
T. Nishiguchi (Ed.), Managing product development (pp. 165–
189). New York: Oxford University Press.
Hines, P. (1994). Creating world-class suppliers. London: Pit-
man.
Holden, N., & Burgess, M. (1994). Japanese led companies.
London: McGraw Hill.
Hopwood, A. G. (1983). On trying to study accounting in the
contexts in which it operates. Accounting, Organizations and
Society, 8(2–3), 287–305.
Hopwood, A. G. (1996). Looking across rather than up and
down: on the need to explore the lateral processing of infor-
mation. Accounting, Organizations and Society, 21(6), 589–
590.
Horngren, C. T., Foster, G., & Datar, S. M. (2000). Cost
accounting. A managerial emphasis. Upper Saddle River, NJ:
Prentice Hall.
Kamath, R. R., & Liker, J. K. (1994). A second look at Japa-
nese product development. Harvard Business Review, 72(6),
154–170.
Kato, Y., Boer, G., & Chow, C. W. (1995). Target costing: an
integrative management process. Journal of Cost Manage-
ment, 9, 39–51.
Koga, K. (1998). Determinants of e?ective product cost man-
agement during product development: opening the black box of
target costing [Working paper]. Harvard Business School.
Kotabe, M., & Murray, J. (1990). Linking product and process
innovations and modes of international sourcing in global
competition: a case of foreign multinational ?rms. Journal of
International Business Studies, 3, 383–408.
Korczynski, M. (1996). The low trust route to economic devel-
opment: inter?rm relations in the U.K. engineering con-
struction industry in the 1980s and 1990s. Journal of
Management Studies, 33(6), 787–808.
Lacity, M. C., Wilolocks, L. P., & Feeny, D. F. (1996). The
value of selective IT sourcing. Sloan Management Review,
Spring, 13–25.
Lamming, R. (1993). Beyond partnership. Strategies for innova-
tion and lean supply. New York: Prentice Hall.
Lukka, K., & Kasanen, E. (1995). The problem of general-
izability: anecdotes and evidence in accounting research.
Accounting, Auditing, & Accountability Journal, 8(5), 71–90.
McKinnon, J. (1988). Reliability and validity in ?eld research:
some strategies and tactics. Accounting, Auditing & Account-
ability Journal, 1(1), 34–54.
Monden, Y. (1995). Cost reduction systems. Target costing and
Kaizen costing. Portland, OR: Productivity Press.
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 25
Monczka, R. M., & Trent, R. J. (1991). Evolving sourcing
strategies for the 1990s. International Journal of Physical
Distribution and Logistics Management, 23(4), 42–54.
Mudambi, R., & Helper, S. (1998). The ‘close but adversarial’
model of supplier relations in the U.S. auto industry. Strat-
egic Management Journal, 19(8), 775–792.
Nishigushi, T. (1994). Strategic industrial sourcing. New York:
Oxford University Press.
Nooteboom, B., Berger, H., & Noorderhaven, N. (1997).
E?ects of trust and governance on relational risk. Academy
of Management Journal, 40(2), 308–338.
Ouchi, W. (1980). Markets, bureaucracies, and clans. Admini-
strative Science Quarterly, 25, 125–141.
Parkhe, A. (1998, Winter). Building trust in international alli-
ances. Journal of World Business, 33(4), 417–437.
Powell, W. W. (1990). Neither market nor hierarchy: network
forms of organization. In L. L. Cummings & B. Staw (Eds.),
Research in Organizational Behavior, 12, 295–336.
Prahalad, C. K., & Hamel, G. (1990). The core competence of
the corporation. Harvard Business Review, 68(3), 79–93.
Quinn, J. B. (1992). Intelligent enterprise: a knowledge and ser-
vice based paradigm for industry. New York: Free Press.
Ring, P. S., & Van de Ven, A. H. (1989). Formal and informal
dimensions of transactions. In A. H. Van de Ven, H. Angle,
& M. S. Poole (Eds.), Research on the management of inno-
vation: the Minnesota studies (pp. 171–192). New York: Bal-
linger/Harper Row.
Ring, P. S., & Van de Ven, A. H. (1994). Developmental pro-
cesses of cooperative interorganizational relationships.
Academy of Management Review, 19(1), 90–118.
Sako, M. (1992). Prices, quality and trust: inter-?rm relationships
in Britain and Japan. Cambridge: Cambridge University Press.
Sako, M., & Helper, S. (1998). Determinants of trust in sup-
plier relations: evidence from the automotive industry in
Japan and the United States. Journal of Economics, Behavior
& Organization, 34, 387–417.
Sheppard, B. H. & Tuchinsky, M. (1996). Inter?rm relation-
ships: a grammar of Pairs. In L. L. Cummings & B. Staw
(eds.), Research in Organizational Behavior, 18, 331–373.
Thompson, J. D. (1967). Organizations in action. New York:
McGraw Hill.
Van der Meer-Kooistra, J., & Vosselman, E. G. J. (2000).
Management control of inter?rm transactional relationships:
the case of industrial renovation and maintenance. Account-
ing, Organizations and Society, 25(1), 51–77.
Venkatraman, N. (1989). Beyond outsourcing: managing IT
resources as a value center. Sloan Management Review, 38,
51–64.
Vining, A., & Globerman, S. (1999). A conceptual framework
for understanding the outsourcing decision. European Man-
agement Journal, 17(6), 645–654.
Walker, G., & Weber, D. (1984). A transaction cost approach
to make-or-buy decisions. Administrative Science Quarterly,
29, 373–391.
Ward, A., Liker, J. K., Cristiano, J. J., & Sobek, D. K. (1995).
The second Toyota paradox: how delaying decisions can
make better cars faster. Sloan Management Review, 36(3),
43–61.
Williamson, O. E. (1975). Markets and hierarchies. New York:
Free Press.
Williamson, O. E. (1979). Transaction cost economics: the
governance of contractual relations. Journal of Law and
Economics, 22, 233–261.
Williamson, O. E. (1985). The economic institutions of capital-
ism. New York: The Free Press.
Williamson, O. E. (1991). Comparative economic organization:
the analysis of discrete structural alternatives. Administrative
Science Quarterly, 36, 269–296.
Yin, R. K. (1984). Case study research, design and methods.
Sage Publications.
Zaheer, A., & Venkatraman, N. (1995). Relational governance
as an interorganizational strategy: an empirical test of the
role of trust in economic exchange. Strategic Management
Journal, 16(5), 373–392.
Zenger, T., & Hesterly, W. (1997). The disaggregation of
corporations: selective intervention, high-powered incen-
tives, and molecular units. Organization Science, 8(3), 209–
222.
26 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
doc_197179554.pdf
Many firms today form alliances with their suppliers and customers that do not fit into the classical dichotomy of
hierarchies and markets.The emergence of so-called hybrid relational forms makes the make-or-buy decision more
complicated than the neo-classical economic perspective indicates.One outcome of these hybrid relational forms
appears to be the development of cost management techniques that cross the organizational boundary between buyers
and suppliers and whose objective is to reduce costs through collaborative efforts.This paper explores how firms enact
interorganizational cost management during product design and the characteristics of the relational contexts associated
with them.It also discusses the implications of such developments for the make-or-buy decision.
Interorganizational cost management and relational context
Robin Cooper
a,
*, Regine Slagmulder
b
a
Emory University, 1300 Clifton Road, Atlanta, GA 30322-270, USA
b
INSEAD, Boulevard de Constance, F-77305 Fontainebleau Cedex, France
Abstract
Many ?rms today form alliances with their suppliers and customers that do not ?t into the classical dichotomy of
hierarchies and markets. The emergence of so-called hybrid relational forms makes the make-or-buy decision more
complicated than the neo-classical economic perspective indicates. One outcome of these hybrid relational forms
appears to be the development of cost management techniques that cross the organizational boundary between buyers
and suppliers and whose objective is to reduce costs through collaborative e?orts. This paper explores how ?rms enact
interorganizational cost management during product design and the characteristics of the relational contexts associated
with them. It also discusses the implications of such developments for the make-or-buy decision.
#2003 Elsevier Ltd. All rights reserved.
Introduction
The neo-classical economics literature represents
the make-or-buy decision as a dichotomous com-
petition between the ?rm and an external supplier.
Two fundamental assumptions underlying this
model are the absence of information asymmetry
between the buyer and the supplier, and the
incomplete nature of the contract between the two
trading partners (Baiman & Rajan, 2000). Despite
the restrictiveness of these assumptions, such a
pure market approach allows for the acquisition
of a broad range of items, including commodity
products (such as nuts and bolts) and commodity
processes (such as surface-mounting parts onto a
printed circuit board). However, as ?rms increas-
ingly focus their attention on their core compe-
tencies, they outsource both a higher percentage of
the total costs of their products and more sub-
stantial items that do not rely upon their core
competencies (Bryce & Useem, 1998; Gilley &
Raheed, 2000; Kotabe & Murray, 1990; Prahalad
& Hamel, 1990; Quinn, 1992; Venkatraman,
1989). The rationale for using an external supplier
for such items includes the supplier’s superior cost
e?ciency, functionality, and quality, and their
ability to incorporate new technologies in a time-
lier manner (Monczka & Trent, 1991; Nishiguchi,
1994). Consequently, not all items that ?rms out-
source can be described as either product or pro-
cess commodities. Instead, many of them rely
upon knowledge that is proprietary to the buyer
or supplier.
The outsourcing of more signi?cant items intro-
duces the problem of information asymmetry
between the buyer and the supplier into the make-
or-buy decision. This information asymmetry can
cause the buyer to establish speci?cations that
unnecessarily increase the costs incurred by the
supplier. For example, by requiring certain func-
tional speci?cations, the buyer might force the
0361-3682/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0361-3682(03)00020-5
Accounting, Organizations and Society 29 (2004) 1–26
www.elsevier.com/locate/aos
* Corresponding author.
supplier to develop the outsourced part using
unnecessarily expensive raw materials. In order to
reduce the costs associated with this form of
information asymmetry, product engineers at the
buyer and supplier may meet during the product
development process and identify opportunities to
change the buyer’s speci?cations in ways that
lower overall costs. Such formalized buyer-sup-
plier interactions, the objective of which is to
identify opportunities for joint cost reduction, are
the domain of interorganizational cost manage-
ment (IOCM) (Cooper & Slagmulder, 1999;
Cooper & Yoshikawa, 1994).
As ?rms outsource more signi?cant items and
engage in IOCM, they develop relational contexts
that do not fall into the simple dichotomy of
markets and hierarchy (Williamson, 1975, 1979,
1985). Instead, these relationships represent inter-
mediate or hybrid modes of governance that
enable ?rms to access the economies of scale and
scope of their trading partners in more e?cient
ways than are possible through either pure arm’s-
length transactions or through vertical integration
(Powell, 1990; Sheppard & Tuchinsky, 1996; Wil-
liamson, 1991). Many di?erent forms of relational
context between buyers and suppliers have been
observed, ranging from relationships in which the
interactions are close to market driven, to strategic
partnerships in which the ?rms have signaled their
desire to work together closely over the long-term
(Heide & John, 1990). These hybrid relational
contexts are characterized by incomplete con-
tracting, as it is neither possible nor practical to
develop contracts that completely specify all of the
potential outcomes of the interactions between
both parties (Baiman & Rajan, 2000).
One outcome of these hybrid relational contexts
is that the calculus of the make-or-buy decision
becomes more complex (Gietzmann, 1996). This
complexity derives in part from the fact that
transaction costs are di?cult to quantify with any
accuracy or rigor, and therefore are likely to
diminish in importance in the decision-making
process (Walker & Weber, 1984). Management
accounting textbooks typically avoid the
measurement problem and treat these costs as
qualitative factors (Atkinson, Banker, Kaplan, &
Young, 2001; Horngren, Foster, & Datar, 2000).
This treatment is problematic, since the account-
ing justi?cation of the decision to source an item
internally, versus externally, often appears to err
by systematically underestimating the magnitude
of these qualitatively determined costs (Drtina,
1994; Lacity, Wilolocks, & Feeny, 1996). Further-
more, as the complexity of the design project
increases, so too does the scope of management
control systems to include not just the narrow
accounting numbers (cost, pro?tability, and bud-
get), but also a broader information set that cap-
tures customer, product design, and time-sensitive
measures (Davila, 2000).
Thus, the increased complexity of the out-
sourcing decision places new demands upon the
accounting and information systems of ?rms, as
evidenced in part by the emergence of IOCM. Yet,
despite the increased interest in interdependencies
and information ?ows that transcend organiza-
tional boundaries and their potential implications
for the ?eld of management accounting, the topic
has been largely ignored in the accounting
research literature (Hopwood, 1996; Van der
Meer-Kooistra & Vosselman, 2000). In this paper,
we seek to address this limitation by exploring
IOCM practices and the relational contexts asso-
ciated with them at seven Japanese manufacturing
?rms.
The paper is structured as follows. In the next
section, we provide a discussion of the ?eld-based
research method and a brief description of the
?rms involved in the research sample. In Observed
IOCM practices, we give a detailed description of
the three IOCM techniques observed at the sample
?rms. These techniques vary in scope from modest
design changes to a near overhaul of both the end
product and the outsourced item. We also
demonstrate that the ability to undertake IOCM
techniques appears to cluster, such that the ability
to perform a more aggressive technique becomes
diagnostic of the ability to perform less aggressive
ones. In Observed relational contexts, we identify
the characteristics of the relational contexts in
which these IOCM techniques were observed. In
IOCM clusters and relational contexts, we show
how each of the observed clusters of IOCM prac-
tices are associated with a speci?c relational con-
text and we discuss the perceived impact of the
2 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
di?erent levels of IOCMon the joint performance of
the sample ?rms. In the Summary and conclusions,
we discuss the impact of IOCM on the calculus of
the make-or-buy decision. Finally, we evaluate the
implications of the research ?ndings upon the ?eld
of cost management and the make-or-buy decision.
Research method
To gain insights into the IOCM practices of
?rms and the relational contexts associated with
them, an exploratory ?eld-based research project
was undertaken. The choice of exploratory case-
based research was dictated by the nature of the
research problem and the lack of extant literature
about interorganizational cost management (Yin,
1984). We visited three large Japanese manufacturing
enterprises, a ?rst-tier supplier to each of them, and a
second-tier supplier to one of the ?rms and docu-
mented their IOCM practices. We selected product
development as the domain of IOCM to be studied,
as prior research has indicated that joint product
development represents a signi?cant source of eco-
nomic bene?t in buyer-supplier relations (Clark &
Fujimoto, 1991; Cooper & Yoshikawa, 1994).
Research site selection
The research ?ndings reported in this paper are
based on an in-depth examination of the IOCM
practices observed at seven Japanese manufactur-
ing companies, belonging to three di?erent supply
chains: the Komatsu–Toyo Radiator supply chain,
consisting of a large heavy industrial manu-
facturer and an independent manufacturer of heat-
exchange equipment such as radiators, oil coolers,
and condensers; the Isuzu–Jidosha Kiki chain, com-
prised of one of the largest automobile manufactur-
ing companies in Japan and a ?rst-tier supplier of
brakes, clutches, steering systems, and pumps; and
the Tokyo–Yokohama–Kamakura chain, includ-
ing a top 10 automobile manufacturer, a ?rst-tier
supplier of hydraulic systems, and a small second-
tier supplier of automotive parts. We identi?ed
and selected the research sites according to various
criteria. Komatsu was selected because of its
reputation for e?ectively managing costs across its
supply chain. Meanwhile, we chose Isuzu for its
reputation as one of Japan’s best practitioners of
value engineering and cost management during the
product development phase. Finally, we chose to
include Tokyo Motors for its reputation for under-
taking thorough customer requirement analyses tied
to its target costing system.
1
To capture IOCM practices from the supplier’s
perspective, we requested that each of the three
buyer ?rms identify a ?rst-tier supplier that man-
agement considered to be especially adept at under-
taking IOCM. Komatsu identi?ed Toyo Radiator as
an excellent ?rst-tier supplier with which it under-
took particularly e?ective IOCM. Isuzu named
Jidosha Kiki Company (JKC) as one of its most
innovative suppliers with whom it had developed a
highly e?ective cost management relationship.
Finally, Tokyo Motors identi?ed its ?rst-tier sup-
plier, Yokohama Corporation, and its second-tier
supplier, Kamakura Iron Works, as they formed a
three-?rm supply chain that was particularly suc-
cessful at practicing multi-?rm cost management.
The choice of Yokohama and Kamakura was for-
tuitous in that the three ?rms were essentially inde-
pendent, rather than part of a kereitsu.
2
We
considered independence to be particularly impor-
tant because it allowed buyer-supplier interactions to
be observed in their purest form. However, if the
?rms were part of a kereitsu, then ‘‘invisible’’ o?-
setting transactions, such as low-interest loans,
might cause one of the ?rms to agree to ‘‘sub-eco-
nomic’’ selling prices. In addition, while the research
focused primarily on the relationships between the
three buyer ?rms and their identi?ed suppliers, we
also documented the general relationships between
the three buyer ?rms and their entire supplier bases.
Data collection
The ?eld research relied upon open-ended inter-
views with managers, design and manufacturing
engineers, and blue-collar workers at the seven ?rms
in the sample. The interviews focused primarily on
1
For a detailed description of target costing, see Cooper
and Slagmulder (1997).
2
This supply chain was the focus of the Cooper and
Yoshikawa (1994) paper.
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 3
their IOCM practices. Given the objective of the
research to study IOCM practices during product
design, the data collection process was limited to
design-related variables. We chose not to interview
individuals in the sales function, for example, as
they were considered unlikely to interact with the
suppliers’ IOCM systems. This decision was vali-
dated by our discussions with senior management
at all of the sample ?rms, which revealed that the
process of IOCM was largely limited to the design,
engineering, and manufacturing functions of the
?rm.
The individuals interviewed at each site were
actively involved in developing, implementing, and
applying the various IOCM techniques used by the
?rm. The interviews were conducted in English with
translator support as appropriate. Typically, we
interviewed between three to ?ve persons (although
the interview occasionally involved as few as one
and or as many as six subjects). Job titles of those
interviewed included General Manager of Product
Planning, Manager of Corporate Planning, Chief
Engineer, and Senior Manager of Group
Accounting. The total site visits lasted 16 days.
The typical time spent at each ?rm was between 2
and 3 days. Initially, we visited each ?rm for 1–2
days and, from the information collected, pre-
pared a draft of the case. Follow-up visits, typi-
cally lasting half to a whole day, clari?ed any
major outstanding misunderstandings and allowed
the company’s appointed contact manager to sign
and release the ?nal draft of the research case.
Company documents, copious notes, and tape
recordings of the interviews were the basis for
seven research cases of approximately 5000 words
each. The majority of the data collected was qua-
litative in nature. The aim of the research was to
identify and to understand the ?rms’ IOCM prac-
tices and the contexts within which they occurred.
Consequently, the research focused more on
exploring the processes that enabled the ?rms in the
sample to collaborate e?ectively, such as guest
engineer programs,
3
than on collecting quanti?able
metrics, such as the number of engineering hours
dedicated to joint design. With the small number of
?rms inthe sample andthe selectionprocesses usedto
identify them, this research only provides evidence
about the existence of IOCM practices. It should not
be construed as either providing evidence of central
tendency behavior or justifying the investment of
resources made in these techniques.
The cases were sent to the contact manager in
each ?rm for review. The ?rst draft of the cases
contained numerous questions that could not be
answered from the tape recordings and notes. The
cases typically went through two to three revisions
before being cleared by senior management at the
sample ?rms. It took between 3 and 12 months to
clear each case. When necessary, the questions and
appropriate textual portions of the case were
translated into Japanese so that managers with
inadequate English skills could answer the ques-
tions and review the text for accuracy. During a
typical clearance procedure, approximately 30
questions were answered and about one-third of
the case was rewritten or amended in some way.
While the majority of these changes involved
author-initiated questions, others were corrections
to the drafts that the reviewing managers made
before releasing the cases. The purpose of this itera-
tive process was to increase the probability that the
observations captured in the cases were factually
correct and accurately re?ected actual practice.
Data analysis
We undertook data analysis in a modi?ed,
three-stage version of the process suggested by
Eisenhardt (1989). First, we used within-case ana-
lysis to identify the IOCM techniques employed at
each individual ?rm. Second, we undertook a
cross-case analysis of the ?rms within the same
supply chain to identify the unique buyer–supplier
interactions across the ?rms’ boundaries that
enabled them to achieve IOCM. Finally, we used
cross-case analysis to uncover patterns of IOCM
techniques and buyer–supplier interactions com-
mon to multiple ?rms and supply chains in the
sample. We identi?ed patterns of association
between IOCM practices and relational contexts
based on the cross-case analyses, using evidence
3
Guest engineers are employees of the supplier who spend
an extended time at the buyer, or vice versa, in order to resolve
joint design problems. For a discussion of guest engineers, see
Holden and Burgess (1994).
4 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
from each case to support or to extend the emer-
ging theory. We considered the data analysis
complete when no additional general patterns
could be identi?ed from the ?eld observations.
In this paper, we primarily analyze the practices
observed between the dyadic pair of Komatsu and its
supplier, Toyo Radiator, both of whom had devel-
oped sophisticated IOCM skills. Komatsu, Ltd. is
one of Japan’s largest heavy industrial manu-
facturers and the world’s second-largest manu-
facturer of a complete line of construction
equipment. The ?rm’s product line contains over
300 models, including bulldozers, hydraulic exca-
vators, wheel loaders, and dump trucks. Toyo
Radiator Co., Ltd. (Toyo) was founded in 1936 as a
radiator supplier to the ?edgling Japanese auto-
mobile industry. Over the years, it has diversi?ed into
all arenas of heat exchange applications and is now
one of the world’s largest independent heat-exchange
equipment manufacturers for construction equip-
ment. Komatsuhas beena Toyocustomer since 1955.
We chose to focus on the Komatsu–Toyo dya-
dic pair because our observations at those two
?rms captured a contemporaneous change in their
IOCM practices and the relational context
between them. Consequently, the interplay
between buyer and supplier and the role of the
relational context in IOCM were more directly
visible at Komatsu–Toyo than at the other
research sites. We used the observations at the
other ?rms both to ?ll in gaps in our observations
at Komatsu–Toyo (for example, about other IOCM
techniques and their associated relational contexts)
and to provide con?rming evidence of any inter-
pretations of those observations. From our collec-
tive observations, we developed a rich description
and analysis of the observed IOCM techniques and
the relational contexts associated with them.
Limitations
Field research is not without its limitations.
Biased observations and a lack of external validity
remain two common criticisms of ?eld studies
(Yin, 1984). McKinnon (1988) highlights the fol-
lowing three major sources of bias. First, the
researcher’s selective perception and interpretation
of events may in?uence the data collection and/or
analysis stage. Second, the interviewees’ reports
may be biased, due to natural human tendencies
and limitations; for example, interviewees may
forget details or report events in a manner that is
?attering to themselves or their company. Finally,
because the researcher is on site only for a limited
period of time, abnormal instances of the studied
phenomenon might be mistakenly treated as if
they were normal conditions. Our research miti-
gated against such potential sources of bias in
several ways. First, wherever possible, we cross-
checked any statements regarding the operation of
the ?rms’ cost management systems with both
company records and descriptions from other
individuals. Second, by developing long-term
relationships with many of the managers inter-
viewed, we were able to ensure their trust and
cooperation, thus reducing any tendencies on their
part to misreport events in ways that would favor
either the company or themselves. Finally, by visit-
ing the companies over an extended period of time,
we avoided the risk of improperly interpreting aty-
pical events through longitudinal observation.
The second criticism of ?eld research relates to
the risks associated with generalizing case study
?ndings in other settings. A common view is that
one cannot draw generally valid inferences from
observations at a small sample of companies
(Ha? gg & Hedlund, 1979; Lukka & Kasanen,
1995). While accepting the rationale behind such
statements, we believe that the knowledge
obtained through our interviews is interesting in
its own right and potentially o?ers a basis for a
more generally applicable theory. Although the
theory developed in this paper is an attempt to
give meaning to a small number of individual
observations in their particular context, we cau-
tiously suggest that it also could apply in other
settings. To validate this claim will require, at a
minimum, that the theory’s external validity be
tested using large-sample statistical methods.
Observed IOCM practices
IOCM helps to reduce the information asym-
metry between the buyer and supplier regarding
the relationship between the speci?cations for the
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 5
outsourced item established by the buyer, and the
resulting costs at the supplier. An IOCM interven-
tion occurs when this information asymmetry causes
the buyer to set speci?cations that the supplier can-
not meet if the latter is to make an acceptable pro?t.
The primary mechanism for identifying the necessity
of an IOCM intervention is target costing. Target
costing lies at the heart of IOCM in the sample ?rms,
as it links customer demands through product design
to the parts acquisition process (Cooper & Chew,
1996; Cooper & Slagmulder, 1997; Kato, Boer, &
Chow, 1995; Koga, 1998; Monden, 1995). However,
as suggested in the literature, target costing is an
arm’s-length cost management technique; it does not
actively involve the supplier in the buyer’s cost
management program. Instead, the buyer’s target
costing system identi?es the purchase price of the
outsourced item, which signals to the supplier’s
target costing system where cost reduction is
necessary. The key extension of IOCM beyond
other cost management techniques is the active
involvement of both the buyer’s and supplier’s
design teams in the joint management of costs.
We observed three IOCM techniques. The ?rst
technique, functionality–price–quality (FPQ) trade-
o?s, helped to resolve relatively minor cost overrun
problems and involved only modest speci?cation
changes—and hence limited interactions among the
?rms’ design engineers. The second technique,
interorganizational cost investigations, was applied
when FPQ trade-o?s were unable to produce the
desired level of cost reductions. The technique
involved more intense interactions among the design
engineers and more signi?cant changes both to the
design of the outsourced item and occasionally to
the speci?cations of the end product. The ?nal
IOCM technique, concurrent cost management,
addressed cost problems that demanded the most
signi?cant cost reduction levels of all three tech-
niques. It required the most signi?cant interactions
between the buyer’s and supplier’s design engineers
and led to fundamental changes in both the buyer’s
product and the outsourced components.
Functionality–price–quality tradeo?s
An FPQ trade-o? is initiated whenever the sup-
plier determines that the manufacturing cost of the
outsourced item will exceed its target cost and that
the only way to reduce costs to the target level is
to relax the functionality and/or quality speci?ca-
tions of the outsourced item in ways acceptable to
the buyer. Once such relaxations are identi?ed, the
supplier requests a meeting with the buyer and the
two design teams to discuss the proposed changes
and to obtain buyer approval to make them. Suc-
cessfully identi?cation of such opportunities helps
the supplier to ensure that it generates adequate
returns. For example, senior management at
Yokohama described the ?rm’s ability to identify
opportunities for cost reduction through FPQ
trade-o?s as critical to its long-term success.
At the heart of a successful FPQ trade-o? is an
e?ective value-engineering program. Yokohama
applies value engineering to all products and fully
integrates the technique into the ?rm’s new pro-
duct development process. In that process, the
product’s basic functions are ?rst identi?ed and its
target cost established. The next step is to develop
prototypes, analyze their costs, and compare them
to the product’s target cost. If the ?nal prototype’s
costs are considered acceptable, it is subjected to
reliability tests and then submitted to the customer
for approval. Once the product obtains customer
approval, it is subjected to a second design round
and its production costs are re-estimated. If these
costs exceed the target cost, then a ?rst-look
value-engineering project occurs. The aim of this
project is to identify ways to change the design of
the product so that it can be manufactured at its
target cost. Once the second generation of design
is established, another round of cost estimation is
undertaken by the engineers. If the design is con-
sidered acceptable, it is subjected to an analysis to
ensure that the product meets its quality speci?ca-
tions. Upon successful completion of this third
design analysis, experimental mass production
ensues, wherein the manufacturing cost of the
product is again re-estimated and the quality and
functionality of the produced items evaluated.
Sometimes, a fourth design review may be neces-
sary to correct any de?ciencies before the com-
pany can release the product for mass production.
Typically, an FPQ trade-o? is an outcome of a
?rst-look value-engineering project. However, it
can be initiated at any time in the design process,
6 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
up to the release of the item into mass production.
Yokohama’s success at initiating these interven-
tions relies upon the in-depth knowledge that its
engineers have developed regarding its customers’
use of the ?rm’s products. This knowledge allows
the engineers to identify where and to what extent
relaxations in the speci?cations of the outsourced
item will be amenable to the buyer. Examples of
the types of changes in speci?cations that result
from FPQ trade-o?s include a request to limit the
color of a part to black or silver, instead of
matching the color of the end-product; a request
to relax surface tolerances when they are not visi-
ble to the end-user; a reduction in the number of
strengthening bars; a shift to pressing a part as
opposed to machining it; and a reduction in the
material content without reducing the strength of
the part. Under FPQ tradeo?s, only minor chan-
ges can be made to the speci?cations of the out-
sourced item and the speci?cations of the end
product are essentially ?xed.
Interorganizational cost investigations
An interorganizational cost investigation takes
place whenever any ?rm in the supply chain for an
item determines that they cannot manufacture the
part at its target cost, and that an FPQ trade-o?
will not produce su?cient cost reductions to
resolve the problem. In many ways, an inter-
organizational cost investigation is similar to a
just-in-time (JIT) production line with each
worker having the ability to shut down the line
when a defect occurs; in this case, however, the
workers are replaced by the design teams at the
supplier ?rms, and the defect lies in the relation-
ship between the speci?cations of the outsourced
item and its purchase price. Just as in JIT, once
the interorganizational cost investigation is initi-
ated, all of the involved players send representa-
tives from their design teams to resolve the
problem. Thus, the ?rst major di?erence between
an FPQ trade-o? and an interorganizational cost
investigation is the ability to include design engi-
neers from more than two ?rms in the supply
chain. The second major di?erence is the scope of
the design changes contemplated. More speci?-
cally, an interorganizational cost investigation
allows parts to be redesigned so that all of the
steps from raw material to ?nished product are
more cost-e?cient. This increased scope of the
design changes enables the identi?cation and
implementation of greater cost savings across the
entire supply chain. However, just as with FPQ
tradeo?s, the fundamental design of the end pro-
duct still remains essentially ?xed.
There are two ways to reduce costs through
interorganizational cost investigations. First, in
order to take full advantage of the manufacturing
skills located throughout the supply chain, avoid
the need to perform activities by redesigning the
product and its components. Second, change the
location of activities so that they are performed
more e?ciently. For example, Kamakura trig-
gered an interorganizational cost investigation
between the engineers of Tokyo Motors, Yoko-
hama, and Kamakura when Tokyo Motors estab-
lished speci?cations for an internal part that could
be met only if Kamakura forged, rather than cast,
the blank. Tokyo Motors’ target costing system
speci?ed the part’s purchase price and transmitted
it, along with the functionality and quality speci?-
cations, to Yokohama. The engineers at that ?rm
identi?ed the manufacturing processes they would
use and, consequently, the speci?cations of the
item they would source from Kamakura. They
used their target costing system to identify their
purchase price for the outsourced part. When the
speci?cations and associated purchase price were
transmitted to Kamakura, the engineers at that
?rm identi?ed their production processes and
determined the necessity of an expensive forging,
rather than a less expensive casting. Consequently,
they concluded that the purchase price set by
Yokohama was insu?cient for Kamakura to gen-
erate an adequate pro?t.
At this point in the process, Kamakura had two
options. The ?rst was to refuse the business, and
the second was to request an interorganizational
cost investigation. Kamakura’s engineers chose
the latter course and requested a joint meeting of
the engineers from all three ?rms. Since only
Kamakura’s engineers had the expertise in forging
and casting technology necessary to understand
the implications of the shift between the two, they
ran the meeting. The solution they identi?ed was
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 7
for Kamakura to start with a cheaper casting and
to undertake the ?rst stage in the surface prepara-
tion process by removing the ridges inherent to the
casting process. Yokohama would then take
delivery of the part and machine it to the shape
required by Tokyo Motors. In addition, the ?rm
would machine the surface of the part (an extra
step) to ensure that it conformed to the speci?ca-
tions provided by Tokyo Motors. Finally, while
Tokyo Motors was not requested to change the
purchase price of the part, it was required to
change the speci?cations in two ways. First, its
engineers reduced the minimum acceptable tensile
strength of the part and, second, allowed for small
blemishes in the surface of the part. Both changes
were necessary to enable casting technology,
rather than forging, to be utilized for the blank.
The two changes proved acceptable to Tokyo
Motors’ engineers because the part was not sub-
jected to much strain and the imperfections on the
surface were not visible to the owner of the vehi-
cle. In e?ect, the original speci?cations had been
too demanding and thus had caused the excess
manufacturing costs. Once all three ?rms had
agreed to the changes, the engineers at Tokyo
Motors set Kamakura’s selling price and thus
speci?ed the distribution of pro?ts between
Kamakura and Yokohama. It was Tokyo’s
responsibility to distribute the pro?ts, as it was the
most powerful ?rm in the relationship and there-
fore could legislate the split to the other ?rms.
Concurrent cost management
Concurrent cost management aggressively redu-
ces costs by increasing the scope of design changes
that the supplier can undertake. One of the events
triggered the emergence of concurrent cost man-
agement at Komatsu was the requirement that
Toyo Radiator produce an engine cooling system
with 40% more capacity at only 18% higher cost.
Komatsu engineers realized that only heroic
e?orts on both sides would enable them to achieve
that objective. Concurrent cost management helps
achieve greater cost savings through increased
design changes in two ways. First, it increases the
amount of time that a supplier’s engineers may
develop innovative solutions to the customers’
requests; second, it concentrates the sourcing of an
entire major function with a single supplier.
Because of the high cost involved, concurrent cost
management is used only for high-value items,
such as major functions. The aim of involving the
supplier much earlier in the design process is to
provide that ?rm with more time to undertake
fundamental redesigns of the major function. The
increased supplier concentration is important
because it allows the supplier to make more fun-
damental changes to the design of the major
function than would be possible were it sourced
from multiple suppliers.
Under the concurrent cost management
approach, cost negotiations between Toyo and
Komatsu began earlier in the development process
while the product was still in the conceptual
stages. Toyo’s engineers estimated the cost of
manufacture and, if it appeared too high, tried to
?nd ways to alter Komatsu’s speci?cations so that
the part could be manufactured for its target cost.
Changes in the speci?cations for Toyo parts were
allowed only if the functionality of the ?nal
Komatsu product was not excessively compro-
mised. The aim was to make negotiations sur-
rounding the setting of target costs more
substantive and two-sided. Once they established
these target costs, engineers at both ?rms had to
?nd ways to achieve them. The earlier establish-
ment of the target costs meant that Toyo had to be
more aggressive in its negotiations if it felt that
Komatsu’s established target costs were too low.
Komatsu changed the way it enacted target cost-
ing by bundling all of the costs of the major func-
tion together, thus letting Toyo Radiator
determine the appropriate target costs of the indi-
vidual components it designed and produced.
We observed two fundamental approaches to
concurrent cost management: parallel and simul-
taneous. In the parallel approach, the engineering
teams at the buyer and supplier operate indepen-
dently, whereas in the simultaneous approach,
they work together to co-design the end product
and the outsourced major function. The perceived
bene?ts from close interactions of the buyer and
supplier design teams determine the choice
between parallel and simultaneous engineering. If
the value of such interactions is considered high,
8 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
then the ?rms use simultaneous engineering;
otherwise, they use parallel engineering.
The primary advantage of parallel engineering is
the ability of the supplier to uncouple its own
product development program from that of the
buyer. For example, under the old approach,
Komatsu would tell Toyo Radiator that the
engine power of a particular product model was to
be increased by X% and that they would require a
new engine cooling system with the appropriately
increased capacity in 12 months. However,
because Toyo Radiator did not know that
increasing engine power was planned for the fol-
lowing generations, it could only react to the
requests for increased engine cooling capacity as
they were received. Under the new approach,
Komatsu informed Toyo Radiator that, for the
next few generations, they would require increased
engine cooling capacity at essentially the same
cost. Toyo Radiator engineers then could deter-
mine that the best long-term solution would be to
launch a research project to develop a more e?-
cient, low-cost approach to engine cooling. The
?rst new technology engine-cooling system might
not be available until the third generation of
Komatsu designs, but at least it would be avail-
able, whereas, under the old approach, Toyo
Radiator would never have had the con?dence to
launch the project in the ?rst place. As long as
Toyo Radiator’s engineers knew the general level
of increased engine power that Komatsu planned,
they needed only to communicate occasionally
with Komatsu’s engineers, typically via periodic
meetings and telephone calls. Thus, in parallel
engineering, the engineering teams operate essen-
tially independently of each other.
In contrast, the aim of simultaneous engineering
is to allow the design teams to cooperate inten-
sively during the early stages of the design process
so that they can co-design both the end product
and the outsourced item with the objective of
?nding ways to deliver the desired level of func-
tionality and quality of the end product at its tar-
get cost. Successful co-design allows the two
design teams to achieve solutions that they could
not achieve separately. Komatsu and Toyo
Radiator demonstrated the process of simulta-
neous engineering with the development of the
mixed ?ow fan, which allowed the generation of a
more e?ective air ?ow, and thus allowed smaller,
lower-cost radiator to supply the same level of
cooling. However, it also required that Komatsu
engineers work closely with the Toyo Radiator
design team to test the adequacy of the fan’s
cooling e?ciency. This testing was carried out
simultaneously with the development of the fan.
Under the old approach, Toyo Radiator would
have completed the development of the new
engine cooling system and then would have pro-
vided prototypes for Komatsu to test. This
sequential approach would have made the devel-
opment of the new fan design too slow to incor-
porate into the current generation, thus delaying
the product launch and losing sales, or retaining
the existing cooling system and engine designs at
higher cost. Both outcomes were considered
unacceptable, as they led to reduced pro?ts at
both Komatsu and Toyo.
Clusters of IOCM practices
The three observed IOCM techniques were
associated with di?erent magnitudes in the design
changes of the items produced by the interacting
?rms. FPQ tradeo?s are associated with small
design changes that can be accommodated by a
single ?rm in the supply chain with the permission
of at least one other ?rm. Interorganizational cost
investigations are associated with more signi?cant
changes that require modi?cations to the design or
production processes of the items produced by
more than one ?rm in the chain. These design
changes are interrelated, but they can be accom-
plished with relatively low levels of communi-
cation between the design teams. Finally,
concurrent cost management is associated with the
most signi?cant changes. Frequently, these chan-
ges are so substantial that the designs of both the
buyer and supplier’s products must be modi?ed in
a compatible fashion. Although the scope and
magnitude of the changes to the design of the
products constitute a continuum, ranging from
small, nearly imperceptible changes to large chan-
ges obvious to the customer, the observed IOCM
practices consist of three discrete clusters (see
Table 1).
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 9
The ?rst cluster captures the ?rms that were
capable of performing all three IOCM techniques.
For example, one of the product development
projects studied was subjected to both a con-
current cost management intervention and an
interorganizational cost investigation. In the sec-
ond cluster, the ?rms were able to perform inter-
organizational cost investigations and FPQ trade-
o?s, but could not undertake concurrent cost
management. The third cluster captures the ?rms
that had the lowest capacity to perform IOCM.
Here, we observed only FPQ trade-o?s. Managers
at the three top ?rms con?rmed that these patterns
of IOCM capability were replicated throughout
their supplier networks.
Observed relational contexts
Managerial accounting practices cannot be
understood in isolation from the broader organi-
zational settings in which they occur (Hopwood,
1983). In our sample, we observed ?ve di?erent
relational contexts in which IOCM occurs. One of
these contexts ?ts into the pure market perspec-
tive. This context was observed for external sup-
pliers that sold standard products, such as nuts
and bolts, to multiple customers. Another context
maps into a pure hierarchy perspective where the
?rms internally source critical parts; for example,
Komatsu internally sources the engines and
hydraulic systems for its bulldozers and excava-
tors. The other three relational contexts were
hybrid forms that did not ?t into the classical
dichotomy of markets and hierarchies. Since these
hybrid relational contexts were associated with
IOCM, they were the only ones that we studied in
depth. Individuals at both the buyer and supplier
?rms used the common supplier context as the
reference point for their descriptions of the other
three contexts; consequently, we also collected
baseline information about that context.
Hybrid organizations re?ect a wide range of
decisions on the part of the management about the
location of the ?rms’ boundaries and the nature of
the governance structure that controls the rela-
tionship (Dyer & Singh, 1998; Gulati & Singh,
1998; Zaheer & Venkatraman, 1995). However,
for the study of IOCM, we can limit the analysis
to those elements of the organizational setting that
facilitate interactions across the organizational
boundary between buyers and suppliers. Of partic-
ular interest is the way that this relational context
varies in ?rms that undertake di?erent forms of
IOCM. While relational contexts might be expec-
ted to vary continuously from markets to hier-
archies, in practice, a limited number of distinct
relational contexts appear to develope. Six attri-
butes are of particular importance when deter-
mining the appropriate relational context for
IOCM. Four of these attributes—design depen-
dence, resource sharing, supplier participation,
and bilateral commitment—relate to interaction
characteristics of the buyer–supplier relationship.
The ?nal two—incentive and protective mech-
anism—pertain to the choices surrounding the
governance structure of the buyer–supplier
relationship.
Design dependence
According to transaction cost economics, rela-
tionships between buyers and suppliers are char-
acterized by some degree of mutual dependence
contingent upon the speci?c investments made and
the switching costs they entail (Williamson, 1985,
1991). In the context of product development,
mutual dependence arises in the form of design
dependence. Design dependence transpires when
the buyer and supplier split responsibility for the
establishment of the outsourced item’s speci?ca-
tions and/or design. The highest level of design
dependence occurs when the supplier and the
buyer establish joint speci?cations and take joint
responsibility for product design. Under these
conditions, the two ?rms must actively integrate
their product development processes:
Table 1
Clusters of IOCM techniques practiced
Ability to
perform
IOCM
FPQ
trade-o?s
Interorganizational
cost investigations
Concurrent cost
management
High Observed Observed Observed
Medium Observed Observed Not observed
Low Observed Not observed Not observed
10 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
Under the new supplier approach, Toyo
Radiator could negotiate with us to move the
condenser and hence delete a complete fan
and motor assembly from the new design.
Komatsu Design Engineer
The next level of design dependence occurs
when the supplier accepts responsibility for design
and manufacture, but the buyer retains sole
responsibility for establishing high-level speci?ca-
tions. Here, the level of integration is lower, but
still demanding, since the two ?rms must ensure
that the end product and the outsourced item are
compatible. Finally, design dependence is low
when the buyer both establishes the speci?cations
and takes responsibility for design, and the sup-
plier only accepts responsibility for manufacture.
Here, the buyer must ensure that the outsourced
components are designed in a way that enables the
supplier to manufacture them for a reasonable
cost. The supplier has few additional responsi-
bilities other than ensuring that the parts are
delivered on time and on spec.
The improved performance that Komatsu
demanded for its A20 and A21 power shovels
triggered a change in relational context at
Komatsu–Toyo. The new designs required engine-
cooling systems, the radiator size of which was
36% larger than that of the previous generation.
Normally, this increase in size would have raised
Toyo’s costs by approximately the same percen-
tage. However, Komatsu’s customers were only
willing to pay about half of the extra costs for the
proposed increased performance. Komatsu man-
agement realized that the target cost it was forced
to set for Toyo (118% of the prior generation’s
cost) was too aggressive to be achieved without
changing the way the two ?rms interacted. Hence,
it adopted a new relationship with Toyo char-
acterized by higher levels of design dependence.
Under the new relationship, the interdependence
between Komatsu and Toyo was largely recipro-
cal, as both partners exchanged outputs with each
other simultaneously, and the output of each
partner became the input of the other (Gulati &
Singh, 1998; Thompson, 1967). Komatsu used the
term ‘‘family member’’ to di?erentiate its suppliers
with the highest levels of design dependence from
other types of suppliers. Toyo had previously been
a ‘‘major supplier,’’ the term Komatsu used for
suppliers that took responsibility for the design of
the group component they supplied, but not for
the establishment of its speci?cations.
Komatsu used two additional terms to di?er-
entiate between the other types of relational con-
texts it had established with external suppliers,
namely ‘‘subcontractors’’ and ‘‘common suppli-
ers.’’ Subcontractors manufacture outsourced
items that are designed by the buyer. They have
few internal design capabilities, but are typically
highly skilled at manufacturing specialized items.
Consequently, they only accept responsibility for
manufacturing, whereas the buyer is responsible
for establishing speci?cations and designing the
part. Finally, common suppliers typically publish
catalogues that detail their product o?erings.
Common suppliers take responsibility for all
aspects of the items’ design and manufacture.
Consequently, there is no design dependence
between Komatsu and its common suppliers.
Similar relational contexts were observed at the
other ?rms in the sample that occupied the top
position in their supply chains. Table 2 illustrates
the relationship between the level of design
dependence—determined by the responsibility for
establishing speci?cations and product design—
and the relational context.
These ?ndings corroborate earlier research that
focuses on analyzing and categorizing relation-
ships between buyers and suppliers. Asanuma
(1989), for example, identi?ed three types of sup-
pliers based on their responsibilities in product
design, which are equivalent to major suppliers,
subcontractors, and common suppliers; however,
he does not discuss a relational context similar to
family members. Other researchers have identi?ed
types of buyer–supplier relationships in addition
to those discussed in this paper, some of which
appear to be motivated by forms of joint action
other than IOCM (Asanuma, 1989; Bensaou &
Venkatraman, 1995; Kamath & Liker, 1994).
Resource sharing
The observed various relational contexts
enabled the buyer and supplier to share di?erent
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 11
proportions of their design-related resources. The
objective of any increased sharing was to enable
more sophisticated interactions between the two
design teams. The resource sharing took two
major forms. The ?rst was increased asset speci?-
city, while the second was increased sharing of
strategic information. Previous research also has
highlighted the importance of both aspects of
resource sharing to the e?ectiveness of buyer–
supplier relationships (Dyer, 1996, 1997; Hines,
1994; Lamming, 1993; Zaheer & Venkatraman,
1995).
Asset speci?city
Asset speci?city relates to the degree to which
an asset can be redeployed to alternative uses
without sacri?ce of productive value (Williamson,
1979). Most of the extant literature on buyer–
supplier relationships has attributed a central role
to the various types of asset speci?city, in line with
Williamson’s transaction cost economics theory
(Dyer & Ouchi, 1993; Dyer & Singh, 1998;
Mudambi & Helper, 1998; Nishiguchi, 1994). In
our sample, Komatsu and Toyo altered the level
of asset speci?city associated with their relation-
ship in order to increase the level of joint design
changes. The two ?rms had previously invested in
assets that were speci?c to their relationship—for
example, the development of a proprietary stacked
?n oil cooler, the investment in dedicated produc-
tion lines, and the use of guest engineers. How-
ever, we observed several instances of increased
asset speci?city. First, the level of physical asset
speci?city increased both through the develop-
ment of proprietary software to model engine
cooling and through the sharing of prototyping
assets. In addition, the new approach to product
development often led to new solutions that were
proprietary to the Komatsu–Toyo relationship
(such as a new mixed ?ow fan). Manufacture of
these new proprietary parts required additional
dedicated equipment at Toyo.
More important, the level of human asset speci-
?city increased signi?cantly. First, more indivi-
duals were involved in the joint product
development process. For example, engineers at
both ?rms worked jointly to improve the overall
performance of future products. As one manager
at Toyo Radiator put it:
Our new relationship with Komatsu requires
us to dedicate more resources, especially
engineers, to their products. For example, in
the past, only design engineers would visit
Komatsu; now we want cost engineers to visit
as well.
Second, the individuals involved in the joint
development process were more intensely
involved. For example, the guest engineers had
added cost considerations to their design activities.
Third, the design activities were more speci?c to
Komatsu than before, as the new designs were
now the outcome of joint product development.
Finally, the location where individual engineers
spent their time changed. In particular, the engi-
neers spent more time co-located.
4
Much of the
skills and knowledge created through this
increased interaction was viewed as proprietary;
for example, Toyo was not allowed to change the
way that it dealt with the prototypes of the engine
cooling systems of its other customers in the near
future. Thus, for at least the foreseeable future,
these investments had no appreciable value out-
side the relationship.
The observations at the other ?rms in the sam-
ple also indicated that the level of both physical
and human asset speci?city varied across the dif-
ferent relational contexts. Only when the suppliers
Table 2
Design dependence and relational context
Level of
design
dependence
Predominant
speci?cations
responsibility
Predominant
design
responsibility
Relational
context
High Joint Joint Family member
Medium Buyer Supplier Major supplier
Low Buyer Buyer Subcontractor
None Supplier Supplier Common supplier
4
There were other ways that resources were shared among
the network ?rms, such as employee placement. However, since
these were not related to cost management, they were not
documented for the purpose of this paper.
12 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
were considered major suppliers or family mem-
bers did we observe signi?cant levels of design-
related asset speci?city. As an illustration, Isuzu
provided considerable design support to its major
supplier, JKC. This assistance included the reso-
lution of manufacturing problems created by new
Isuzu parts with complicated shapes that were
di?cult to machine. However, none of the ?rms
dedicated any signi?cant design-related assets to
subcontractors or common suppliers.
Strategic information sharing
Achieving cooperative buyer–supplier relation-
ships requires intensive bilateral communication
and information sharing to engender appropriate
levels of learning and trust (Carr & Ng, 1995;
Dyer & Singh, 1998; Lamming, 1993). Increased
sharing of strategic information played an impor-
tant role in enabling Komatsu and Toyo to e?ec-
tively undertake joint product development
projects. E?ective co-design became possible
because Komatsu and Toyo were sharing con-
siderable information about each others’ design
plans early in the product development process.
For example, right at the beginning of the devel-
opment process, Komatsu told Toyo that its plans
for the new A20 and A21 power shovels included a
40% increase in engine power. Early conveyance
of such strategic information was necessary to give
both Komatsu and Toyo adequate time to jointly
make the signi?cant changes in their product
designs necessary for the project to achieve its cost
objectives:
As a family member, we are told about
Komatsu’s long-term development plans a lot
sooner than we used to be. This earlier noti-
?cation allows us to match our research
e?orts with Komatsu’s and ensure that di?-
cult technical and cost challenges created by
Komatsu’s product plans are overcome.
Toyo Radiator Manager
Under family membership, Komatsu and Toyo
shared an extensive range of information with
each other, despite being separate ?rms. Komatsu
had access to all of Toyo’s cost information
regarding Komatsu-related products, even knowing
the price that Toyo paid for a single bolt used in a
Komatsu product. The objective of this cost
information sharing was to allow Komatsu to ?nd
new ways to reduce costs. For example, it might
increase discounts by purchasing the bolt centrally
and have the bolt manufacturer deliver it directly
to all its users in the Komatsu group and its family
members.
At the other ?rms in the sample, the level of
strategic information sharing also varied across
the three relational contexts. While still sharing a
considerable amount of information with their
major suppliers, the buyers did not share as much
strategic information as they did with their family
members. When the suppliers were considered
subcontractors, little strategic information sharing
was necessary, as design responsibility resided
with the buyer, while the supplier needed only to
manufacture the part. Strategic information shar-
ing was markedly absent for common suppliers.
Here, the relationship was essentially arm’s-length
and market-based, and the two ?rms shared
almost no knowledge outside of price and delivery
information.
Supplier participation
The increased resource sharing between
Komatsu and Toyo enabled the suppliers’ product
engineers to play a more substantive role in the
development of the buyer’s products. However, in
order to achieve this objective, the relational con-
text was modi?ed in two major ways. First, more
substantive items and, in particular, the research
and development associated with them, were out-
sourced. Second, family members became involved
at an earlier stage in the product development
process so that they could have more time to
identify innovative, low-cost designs.
Outsourcing more substantive items
In the case of Toyo Radiator, Komatsu man-
agement decided to outsource most of the research
and development and all of the manufacture of the
engine cooling system. Previously, when only sim-
ple components were transferred, Komatsu had
undertaken the majority of the research and
development in-house, had purchased discrete
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 13
components from Toyo and other radiator manu-
facturers, and then had assembled the engine
cooling system in-house. By outsourcing more
substantive items—primarily major functions—
Komatsu was able to rely more heavily upon the
design skills of its new family members. An
important outcome of this outsourcing policy was
a reduction in the number of suppliers. As one
Komatsu manager put it:
Toyo used to produce only the radiator and
other ?rms produced the fan and electric
motor. However, split design made it di?cult
to increase e?ciency, so we decided that
under the new supplier program, Toyo would
produce the motor, fan, and radiator in one
integrated package.
The value and complexity of the outsourced
items varied by relational context. For family
members, the outsourced items were typically
major functions, such as engine cooling systems.
In contrast, for major suppliers, they were group
components (for example, radiators). For sub-
contractors, the items typically outsourced were
simple components, such as radiator fan blades.
Finally, for common suppliers, the outsourced
items were standard components, such as nuts and
bolts. This observation is in keeping with the
results from an empirical study by Heide and John
(1990), who found that closer relationships
between buyer and supplier are associated with
higher values of the outsourced parts as a percen-
tage of the end product.
Timing of supplier involvement
In the case of the A20 and A21 power shovels,
interaction between the design teams of Komatsu
and Toyo Radiator began 12 months earlier than
it would were Toyo still a major supplier. This
additional time was critical, as it allowed Toyo
and Komatsu to consider more fundamental
redesigns of the cooling system and engine.
Previously, we waited until Komatsu pre-
sented us with work to perform before we
began design. Unfortunately, the contracts
were signed too late in the overall design
process to give us the time we really needed to
design low-cost solutions into our products.
Now we work together much earlier in the
process and can propose more substantive
design changes.
Toyo Radiator President
These ?ndings are consistent with previous
research that associates early and extensive sup-
plier involvement with a faster and more e?cient
product development process (Birou & Fawcett,
1994; Clark & Fujimoto, 1991; Ward, Liker, Cris-
tiano, & Sobek, 1995). The buyer–supplier inter-
actions for the other relational contexts occurred
later in the product development process. Typi-
cally, major suppliers were brought in after pro-
duct conceptualization was complete and the
design of the product nearly ?nalized. For sub-
contractors, the involvement occurred even later,
usually after the parts list had been generated. For
common suppliers, there was essentially no inter-
action until mass production was scheduled and
orders were placed. Thus, the timing of the design
interactions between buyer and supplier appeared
relate to the type of relational context and the
magnitude of cost reduction envisioned.
Bilateral commitment
The increased resource sharing and reliance
upon the supplier’s product development skills
associated with some of the observed relational
contexts required that the buyer and supplier
strengthen their bilateral commitment. They
achieved this objective by increasing the stability
of their relationship and the degree of collabora-
tion between the two design teams.
Stability
Stability relates to the bilateral expectation of
continued future interactions between the buyer
and supplier (Heide & John, 1990). The under-
lying stability of the buyer–supplier relationships
observed at the sample ?rms varied with the rela-
tional context. Previously, when Toyo was sup-
plying relatively simple parts designed by
Komatsu, it was easy for Komatsu to ?nd other
?rms to supply those parts, although it typically
14 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
would choose not to do so. Under the new
approach, Toyo supplied Komatsu with highly
specialized parts that relied heavily upon speci?c
knowledge of Komatsu’s products. Therefore,
Komatsu was far more dependent upon Toyo
than they were previously. Because of the high
switching costs, they worked to maintain a stable,
productive relationship. Furthermore, Toyo had
come to rely upon Komatsu for a larger share of
its business, special design skills, and engineering
support. Consequently, it was also less likely to
consider severing the relationship. The two ?rms
were thus highly interdependent and therefore
were more willing to support each other. Toyo
could rely upon Komatsu to continuously give it a
steady stream of business, while Komatsu could
rely upon Toyo to develop customized, innovative
engine cooling systems. The stability of the rela-
tionship was further enhanced by the reduction in
the number of suppliers that accompanied the
adoption of a more integrated buyer–supplier
relationship.
Increased stability is important because it takes
considerable time for the buyer and supplier to
develop the mutual in-depth familiarity required
for the high level of e?ective co-design involved in
family membership. This knowledge includes joint
technical expertise, mature personal relationships,
and extensive prior experience. For example, over
time, Toyo engineers had developed considerable
expertise speci?c to Komatsu. While other engine-
cooling suppliers existed, they lacked this
Komatsu-speci?c know-how and technology. It
would take considerable time and resources on the
part of Komatsu to match the capabilities of Toyo
Radiator.
At the major supplier level, the relationships
were still quite stable. Only if a major supplier
consistently failed to compete would it cease to
remain in that supplier group. In contrast, at the
subcontractor level, the buyer ?rms were willing to
take back in-house certain items that they pre-
viously had outsourced if there was insu?cient
work to keep their own workforce occupied.
Finally, for common suppliers, there was very lit-
tle stability, as the buyers would simply select the
common supplier from a group of certi?ed ?rms
that gave them the best value for each item.
Collaboration
Collaboration between buyers and suppliers is
considered a necessary condition for e?ective
buyer–supplier interactions that go beyond pure
arm’s-length, contractual relationships (Dyer,
1996; Gietzmann, 1996; Heide & John, 1990). The
relationship between Komatsu and Toyo Radiator
had always been collaborative. However, when
Komatsu designed the engine cooling system itself,
its design engineers did not have to interact so
extensively with Toyo’s engineers. Toyo and the
other major suppliers were manufacturing group
components and their primary task was to achieve
high quality and on-time delivery at the right
price. Komatsu’s primary task was to develop
speci?cations, identify capable suppliers, review
their bids, and accept the highest value bid. In
contrast, when Toyo became responsible for the
development and manufacture of entire engine
cooling systems, the product development engi-
neers of the two ?rms had to work closely together
to identify new approaches to engine cooling
technology that might involve concurrent changes
in the design of the engine. Consequently, the new
buyer–supplier relationship created a culture of
intense collaboration.
Toyo Radiator is closely related to Komatsu;
it is part of a cooperative group. Only a few
dozen suppliers have such a rich relationship
with Komatsu.
Toyo Radiator President
The relationship between Komatsu and Toyo
achieved this objective by stimulating regular
meetings between the design teams of the two
?rms. The simultaneous redesign of both the
engine and the engine cooling system required
considerable coordination between the two engi-
neering teams, since the product development
process was highly iterative. The aim of these per-
iodic meetings was to integrate the research and
development e?orts of the two groups, to allow
suppliers to provide greater input earlier in the
development process, and to help ensure that cost
reduction negotiations were more substantive.
The other ?rms in the sample also collaborated
and helped each other overcome cost problems
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 15
during product development. For example, if one
of the ?rms encountered an engineering challenge,
it was not unusual for engineers from other ?rms
in the chain to help solve the problem. Isuzu, in
particular, used value engineers to help its suppli-
ers resolve cost problems. Similarly, the design
teams of all of the ?rms in the Tokyo–Yokohama–
Kamakura chain often got together to identify
ways to jointly reduce costs.
The magnitude and intensity of the collabora-
tion varied with the relational context. For exam-
ple, since subcontractors were not responsible for
any of the design aspects of the parts they manu-
factured, the help they received was limited to the
manufacturing process. In addition, the buyer did
not look to subcontractors for any signi?cant
help, as there was little additional value that they
could provide to the buyer in terms of design sup-
port. Finally, with common suppliers, no assis-
tance was o?ered by either party.
Governance structure
To reap the full bene?ts of enhanced buyer–
supplier relationships, one must consider neces-
sary shifts in governance structure (Gietzmann &
Larsen, 1998). Our observations showed that, as
the relational context between Komatsu and Toyo
changed, the governance structure altered accord-
ingly. Governance structure is de?ned here as the
mechanisms that create both incentives (i.e.
reward and coercion mechanisms) for the buyer
and supplier to interact, and safeguards that pro-
tect each transactor against the risk of opportu-
nistic behavior on the part of the other
(Williamson, 1979, 1985). In particular, our sam-
ple ?rms’ governance structure relied heavily upon
trust rather than the classical disciplining
mechanisms of authority and price (Nooteboom,
Berger, & Noorderhaven, 1997).
Trust played an important role in the relational
contexts associated with both family members and
major suppliers, unlike the other two relational
contexts of subcontractors and common suppliers.
The key role of trust was to stimulate innovation
among the various ?rms in the supply chain. For
example, the relational context and its associated
governance structure allowed Komatsu and Toyo
to ?nd new ways to integrate the engine and its
cooling system for the A20 and A21 power sho-
vels. Our observations regarding the role of trust
and innovation are in line with prior research
?ndings that demonstrated how low trust rela-
tionships fail to stimulate new ideas (Korczynski,
1996). In addition, it has been suggested that nei-
ther the classical hierarchy nor market relational
contexts support innovation:
The hierarchy/authority mode of inter-?rm
relations clearly risks impeding innovation by sti-
?ing the upward ?ow of new ideas from sub-
ordinated suppliers. Their narrow specialization
leaves them without the technological know-how
needed for innovation, and their subordination
leaves them few incentives to contribute innova-
tive ideas to customers. . .The market/price mode
facilitates innovation by creating incentives to
generate new ideas, but his mode, too, impedes
innovation because suppliers and customers of
innovations have di?culty agreeing on a price for
these innovative ideas (Adler, 2001, p. 224).
In contrast, trust-based communities have
indeed been shown to be e?ective when the buyer
needs to encourage the supplier to be both inno-
vative and a source of knowledge (Bensaou &
Venkatraman, 1995; Dyer, 1996; Helper, 1996;
Sako, 1992).
There are three primary sources of trust: famil-
iarity through repeated interaction, calculation,
and behavioral norms (Adler, 2001). In our sam-
ple, two of these sources of trust were essentially
held constant. First, all of the buyer–supplier
relationships between the ?rms had existed for an
extended period of time. The ?rms were familiar
with each other and had built trust through their
repeated interactions. Second, they had developed
norms of behavior that were ‘‘binding.’’ All of the
?rms studied had demonstrated the ability to con-
sistently deliver high quality products in a timely
and accurate manner; therefore, they had already
established a strong reputation as being ‘‘good’’
players (where reputation is de?ned in terms of
repeated consistency over time). Since none of the
?rms in the sample failed to maintain their repu-
tations across the period of observation, we did
16 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
not observe the role of reputation as a di?er-
entiating, self-enforcing safeguard in the buyer–
supplier relationship.
Thus, it was primarily the level of calculative trust
that varied across the observed relational contexts.
Calculative trust develops when each of the parties
undertakes a sober assessment of the costs and ben-
e?ts of exploiting any vulnerability of the other
party, and determines that the calculus favors main-
taining the relationship on both sides (Adler, 2001).
The ?rms in the family member and major supplier
relationships were observed to actively manage
their calculative trust to create adequate self-
enforcing incentive and protection mechanisms.
Incentive mechanisms
The nature and importance of incentive
mechanisms varied across the relational contexts.
For family members, the primary incentive
mechanism was trust-based and took the form of
mutual bene?t. Family members relied largely on
the principle of mutual bene?t by actively working
together to increase the joint economic bene?t
from their cooperative product design e?orts.
Furthermore, the need to maintain an adequate
level of trust played a role in ensuring that some of
the reward and coercion mechanisms remained
e?ective. For example, the incremental value cre-
ated by the relationship had to be shared to some
extent—although not necessarily equitably,
according to both parties—if trust were to evolve
(Contractor & Lorange, 1988). Thus, Komatsu, as
the more powerful ?rm in the chain, had to ensure
that it did not reap all of the additional pro?ts
generated through collaboration.
Sometimes our sharing of cost information,
coupled to our knowledge of Toyo’s pro?ts,
can lead to a con?ict of interest, with pressure
building within Komatsu to reduce target
costs where Toyo’s pro?ts are known to be
high. However, we share a common goal—
getting costs as low as possible—which
ensures that these con?icts rarely become
serious. To reduce the incidence of such con-
?icts, we do not set our targets costs for parts
manufactured by Toyo based upon our
knowledge of Toyo’s costs. Instead, we try to
set our target costs independently of Toyo
and let Toyo make as much pro?t as possible.
Komatsu Purchasing Manager
At the major supplier level, the principle of
mutual bene?t was still operative, but it was less
important. For major suppliers such as Toyo
(before it became a family member), JKC, and
Yokohama, the primary incentive mechanism was
the volume of business that the buyer gave the
supplier. For example, Isuzu used direct competi-
tion between its major suppliers to ensure that the
suppliers were as innovative as possible. When a
supplier failed to remain competitive, Isuzu pun-
ished that ?rm by awarding it slightly less volume
than in previous years. Similarly, it awarded
innovative suppliers with slightly more volume
than in previous years. To help the poorly per-
forming suppliers become competitive, Isuzu pro-
vided them with additional engineering support.
This observation is in line with Helper (1996), who
found that Japanese ?rms typically opt for the
‘‘voice’’ strategy of joint problem solving, rather
than following the ‘‘exit’’ strategy of ending rela-
tions with poorly performing suppliers.
For subcontractors, the primary incentive
appeared to be continued business. While severing
relationships was rare, it was understood that the
supplier would only be retained as a subcontractor
if the ?rm maintained adequate performance
levels. At the subcontractor level, the principle of
mutual bene?t was still in operation, but played a
relatively minor role. It was enacted primarily by
the ?rms sharing engineering expertise where ben-
e?cial. For example, the buyer might provide the
supplier with engineering support to resolve partic-
ularly di?cult manufacturing problems associated
with outsourced items so as to ensure the achieve-
ment of the supplier’s target costs. At the common
supplier level, the only incentive that appeared to be
in e?ect was the economic bene?t both sides derived
from a market price-based transaction. The princi-
ple of mutual bene?t did not appear to operate.
Protection mechanisms
To mitigate against the risk of opportunistic
behavior by their trading partners, the ?rms relied
upon a number of protection mechanisms, which
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 17
varied with the relational context. In trust-based
relationships, the dominant risk is unilateral
defection (Granovetter, 1985). One way that ?rms
can signal a low risk of defection is by structuring
the relationship so that the commitment of both
sides is clearly observable (Parkhe, 1998). In the
case of family members, trust was maintained by
ensuring that both parties were visibly mutually
interdependent. For example, Komatsu openly
relied upon Toyo Radiator for its expertise in
engine cooling systems, while Toyo Radiator
openly relied upon Komatsu for a signi?cant por-
tion of its business and for engineering support to
develop new technologies. Thus, for family mem-
bers, the dominant protection mechanism was
mutual interdependence. This high level of mutual
interdependence led to barriers to unilateral defec-
tion and created an additional safeguard against
opportunistic behavior. More speci?cally, Toyo had
access to Komatsu’s future product plans, which
contained highly valuable information for Komat-
su’s competitors. However, Komatsu in turn had
access to highly proprietary information about
Toyo and, if Toyo were to defect, Komatsu could
retaliate by sharing that information with Toyo’s
competitors. Within these close relationships, per-
sonal connections also play an important role in
reducing the risk of opportunistic behavior (Ring
& Van de Ven, 1989, 1994). For example, both
sides viewed the new relationship as akin to a
strong friendship, an example of a self-enforcing
safeguard that reduces the need for legal and other
formal protection mechanisms (Dyer, 1997).
Komatsu is a very important customer with
whom we have a highly supportive relation-
ship that can be likened to a very strong
friendship.
Toyo Radiator President
For major suppliers, mutual interdependence
was still a major protection mechanism, but less
signi?cant than it was for family members. The
sequential nature of the design process across the
interorganizational boundary for major suppliers,
coupled with the existence of multiple competing
suppliers, reduced the dependence of the buyer
upon the supplier. However, major suppliers still
had extensive and specialized knowledge regarding
the needs of their customers. Therefore, it was not
feasible for the buyer to switch from an existing
major supplier to a new one from outside their
supplier base. Furthermore, the small number of
?rms in the supplier base that could make a given
family of products made it virtually impossible for
the remaining major suppliers to immediately
expand production to o?set the loss of capacity.
Thus, the buyer had visibly rendered itself depen-
dent upon each of its major suppliers for a rea-
sonable duration. The major suppliers had
rendered themselves similarly dependent on their
customers because they only transacted with a
limited number of customers in this type of rela-
tionship.
5
Therefore, they would su?er consider-
able economic hardship by the loss of a customer
to whom they were a major supplier. Since it was
di?cult to create new major supplier relationships
in the short-term, the suppliers were equally com-
mitted for a reasonable period in the future. Thus,
major suppliers and their customers were mutually
interdependent, although not to the same extent as
family members and their customers.
For subcontractors, the level of mutual inter-
dependence was much lower than that of family
members and major suppliers. The buyer typically
dealt with multiple subcontractors and could
easily compensate for the defection of a sub-
contractor. Finally, for common suppliers, there
was little mutual interdependence, since the buyer
could go to any number of equivalent ?rms for the
outsourced items and, typically, the buyer repre-
sented only a small portion of the supplier’s over-
all volume. Here, the dominant protection
mechanism was market price.
Levels and nature of trust
The level and nature of trust di?ered across the
relational contexts at the sample ?rms. Our
observations suggest that the variation in the level
of calculative trust can be viewed as the result of
the interaction between mutual bene?t and mutual
interdependence. For family members, mutual
5
Some of the suppliers transacted with other types of cus-
tomers; for example, it was not usual for some of the major
suppliers to also be subcontractors to other customers.
18 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
bene?t and mutual interdependence were both
high and, consequently, so was the level of calcu-
lative trust. In contrast, the level of mutual bene?t
and interdependence decreased for major suppliers
and even more so for subcontractors; overall trust
was also observed to be low, but still signi?cant.
Finally, for common suppliers, the level of calcu-
lative trust was close to zero, as mutual bene?t
and interdependence were essentially nonexistent.
The nature of the trust engendered by the
observed relational contexts also di?ered. Trust
between Komatsu and its family members took
the form that has been described in the literature
as ‘‘goodwill trust’’ (Sako, 1992; Sako & Helper,
1998). Goodwill trust is characterized by the will-
ingness of both parties to go beyond the contract
and to act in the best interest of the other party,
even at a slight disadvantage to themselves. Such
trust can also be described as taking the strong
form; i.e. while the ?rms were signi?cantly vul-
nerable to each other, they were so committed to
the relationship that they had internalized values,
principles, and standards of behavior that pro-
tected the relationship (Barney & Hansen, 1994).
These ?ndings are in keeping with earlier research
?ndings that concluded that self-enforcing
mechanisms are more e?ective than third-party
enforcement mechanisms (such as contracts) at
both minimizing transaction costs and stimulating
value-creation initiatives (Dyer & Singh, 1998).
Komatsu and its major suppliers were not
mutually interdependent to the same extent as the
?rm was with its family members, nor were the
mutual bene?ts as high. In these relationships,
trust took the form of competence trust (Sako,
1992) or the semi-strong form of trust (Barney &
Hansen, 1994). Competence trust implies that the
buyer believes that the supplier has the compe-
tence to complete the order and need not be mon-
itored during or after the process. The observed
relationships between buyers and subcontractors
were less dependent upon self-enforcing safe-
guards than those observed with major suppliers
and, in particular, family members. Instead, the
predominant form of protection was through the
written contracts designed to maintain those rela-
tionships and through the risk of termination for
poor performance. Thus, the maximum level of
trust between the two ?rms is ‘‘contractual trust’’
(Sako, 1992), or ‘‘weak form trust’’ (Barney &
Hansen, 1994). This form of trust relies upon the
mutual expectation that promises of a written or
verbal nature will be kept. In other words, the two
parties assume that each will adhere to the con-
tract without necessitating any coercion.
The dominant protection mechanism between
the buyers and common suppliers was price. The
buyers would typically choose from a limited
number of common suppliers, depending upon
which one was perceived as o?ering the best value.
If a common supplier failed to deliver upon a
given contract, they were dropped from the list of
acceptable suppliers. Thus, there is no calculative
trust between the two ?rms, and legal action is the
sole remedy for non-performance.
Thus, when few or no bene?ts are to be derived
from joint action, the ?rms in a supply chain avoid
becoming interdependent upon each other and
thus reduce the need for self-enforcing safeguards.
Instead, the ?rms rely upon arm’s-length market
transactions and written contracts. Arm’s-length
contracts work well when the interactions between
the parties are straightforward, but not when they
are highly interdependent. In the former case, only
the quantity, price, quality, and delivery times
need to be speci?ed, whereas in the latter it is
almost impossible to specify a priori all of the
actions that might be required and de?ne payment
schemes for them. According to transaction cost
economics, complex contracts are inevitably
incomplete because it is impossible or too costly to
contract upon all future contingencies as a result
of bounded rationality (Williamson, 1979). There-
fore, an alternative, non-contractual, and more
?exible form of governance based on trust is
required where the two parties act bilaterally to
maximize the joint return.
The relational contexts that we observed are not
simple hybrids of the classical market and hier-
archy contexts. The original markets and hier-
archies model assumed that that there were only
two discrete relational contexts and that trust
played no role in business transactions (William-
son, 1991). This highly constrictive view was later
replaced by a ‘‘swelling middle’’ perspective
whereby the hybrid relational contexts were linear
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 19
mixtures of the two classical ones (Zenger & Hes-
terly, 1997). However, in our sample, the natural
progression is not on a continuum from market to
hierarchy, as a third factor, trust, is getting both
stronger and more encompassing. With the accep-
tance of the role of trust, hybrid relational contexts
can be viewed either as the outcome of a three-way
trade-o? (Ouchi, 1980) or as solutions in a three-
dimensional space consisting of hierarchy/authority,
market/price, and community/trust (Adler, 2001).
IOCM clusters and relational contexts
Each of the three observed IOCM clusters was
associated with a distinct hybrid relational context.
There are two aspects to this association. First, the
motivation behind the link appears to be the level
of interorganizational cooperation and coordina-
tion required by the various IOCM techniques on
the one hand, and the level of cooperation supported
by the relational contexts on the other hand. The
more demanding the technique, the further the rela-
tional context is removed from a pure market. Sec-
ond, the perceived e?ect on joint performance varies
with the IOCM technique and the anticipated
savings apparently play a role in motivating the
adoption of a speci?c type of relational context.
Establishing the linkage between IOCM cluster
and relational context
Our observations suggest that the IOCM tech-
niques practiced by buyers and suppliers are
linked to their relational context. The relational
context of common suppliers was observed to
come closest to the pure market form.
6
If we
de?ne the attributes of the common supplier as the
reference point to compare the three hybrid rela-
tional contexts, as well as their departure from
pure market interactions, then a pattern emerges.
First, each cluster of IOCM techniques was asso-
ciated with a particular relational context, and
vice versa: i.e., we observed a one-to-one relation-
ship between IOCM cluster and relational context.
Second, for all relational context attributes studied
(i.e. those related to design dependence, resource
sharing, supplier participation, bilateral commit-
ment, and governance structure), the sub-
contractor relational context is the closest of the
three to a common supplier, followed by the major
supplier, with the family members being the
furthest removed from the classical market form.
Thus, a systematic, monotonic relationship was
evident in the attributes associated with each rela-
tional context. Third, the IOCM techniques were
associated with di?erent magnitudes in design
changes to the items produced by the interacting
?rms. If we treat the market pricing associated
with common suppliers as the baseline for IOCMon
the basis that no joint design changes are implied,
then the smallest level of design changes are asso-
ciated with the FPQ trade-o?s, the middle level with
interorganizational investigations, and the highest
value design changes with concurrent cost manage-
ment. Thus, a monotonic relationship was observed
again in the magnitude of the design changes and in
the ability to perform IOCM in each cluster.
The observed sets of pairing between IOCM
technique and relational context can now be cau-
tiously interpreted (see Table 3). The higher the
level of design changes envisioned, the more buy-
ers and suppliers are required to interact in a rich
and varied manner. Thus, concurrent cost man-
agement, which involves the largest design chan-
ges, is associated with a family member context.
This context enables the richest interactions
between the design teams of the two ?rms and is
furthest removed from the pure market form. It is
important to observe that family membership and
concurrent cost management evolved simulta-
neously at Komatsu and Toyo Radiator, and that
senior management perceived them as a single
outcome. FPQ trade-o?s, which involve the smal-
lest design changes, are associated with a sub-
contractor relational context, which of all the
hybrid relational contexts observed is the closest
to pure markets. Finally, the middle level of design
changes involved in interorganizational cost
6
There are some minor di?erences in the observed rela-
tional context compared to the theoretical one. For example,
some of the ?rms were loyal to a limited number of common
suppliers and would not switch suppliers simply because of
price considerations. This decision was based in part upon the
perceived bene?ts of long-term relationships in sourcing
decisions, even for common suppliers.
20 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
investigations is associated with the intermediate
relational context, major suppliers. Thus, the three
observed clusters of IOCM practice are associated
with three di?erent relational contexts. The high-
est ability to perform IOCM is associated with
family membership and the lowest ability with a
subcontractor context, whereas the middle IOCM
cluster is associated with the major supplier con-
text. The sample ?rms studied did not undertake
IOCM ?rst and then adapt their relational context to
make it more aligned, nor did they change their
relational context and then subsequently develop the
ability to undertake IOCM. Instead, the relational
context was modi?ed contemporaneously with the
development of the ability to undertake IOCM.
An underlying assumption in this analysis is that
the pressure on the ?rms to undertake IOCM was
su?ciently high to cause them to change their
relational context accordingly; therefore, certain
central associations between IOCM and relational
context can be observed. However, we acknow-
ledge that a large number of factors may well
in?uence the design of relational contexts. Fur-
thermore, we cannot rule out the role of idiosyn-
cratic in?uences on IOCM and relational contexts.
However, since our objective is to highlight the
existence of a pattern of relationships between
IOCM practice and relational context and not to
suggest causal relationships, we can ignore idio-
syncratic in?uences for the purpose of this paper.
Perceived e?ect on joint performance
The managers in the sample ?rms appear to adopt
the IOCM techniques and associated relational
contexts because they believe that these combina-
tions will lead to superior joint performance, as
illustrated by a Toyo Radiator manager:
Under the old approach to supplier relations,
Komatsu would identify the performance
speci?cations for their next model and
develop the basic design concepts. Toyo
would then develop new technologies on its
own and present Komatsu with their ideas of
how to achieve the basic concept. However,
with the A20 and A21 power shovels, that
approach would not have achieved our
objective. In order to get the costs low
enough, we had to undertake a concurrent
cost management program.
It is extremely di?cult to generate a priori
numerical support for the contention that the
application of IOCM in the relational context of
hybrid organizations is more e?ective than con-
ventional cost management in the context of a
market or a hierarchy. Consequently, the decision to
adopt a new cost management technique can be
viewed as a calculated gamble based upon the
expected bene?ts and costs of both developing the
expertise necessary to undertake the IOCM tech-
nique and creating its associated relational context.
There are two ways to measure the e?ectiveness
of a given IOCM technique. The ?rst way com-
pares the savings to the incremental cost of an
individual IOCM intervention. The second way
compares the total savings associated with a given
IOCM technique from all a?ected suppliers dis-
counted over time, to the cost of creating the
ability to undertake the technique and establishing
the appropriate relational context.
E?ectiveness of individual IOCM interventions
The economic justi?cation of an individual
IOCM intervention compares the anticipated cost
savings against the cost of undertaking the inter-
vention. The anticipated joint savings of a single
intervention, when considered as a percentage of
the overall value of the outsourced item, were
considered to vary with the utilized IOCM tech-
nique. The savings from FPQ tradeo?s were
expected to be modest, representing only a few
Table 3
Relational context and IOCM techniques utilized
Relational context Level of
design change
supported
Predominant IOCM
technique utilized
Family Member High Concurrent cost
management
Major Supplier Medium Interorganizational
cost investigation
Subcontractor Low FPQ trade-o?
Common Supplier None None
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 21
percentage points of the cost of the outsourced
item. For interorganizational cost investigations,
the savings were typically expected to be in the 5–
10% range. Finally, for concurrent cost manage-
ment, the savings were expected to be between 10
and 15%. Since the value of the outsourced item
increases with the distance of the relational con-
text from a pure market situation, so do the
expected cost savings (see Table 4).
Some ex post evidence supports the adoption of
concurrent cost management for the A20 and A21
power shovels at Komatsu and Toyo Radiator.
Using simultaneous engineering and an inter-
organizational cost investigation, the two design
teams managed to overcome a particularly severe
cost problem: an a priori predicted 36% cost
overrun. These savings were achieved by jointly
designing a new engine cooling system and engine,
changing the shape of the cooling fan blades,
reducing the number of fans and electric motors
from two to one, and changing the design of the
diesel engine so that it could accommodate the
new engine cooling system. Komatsu and Toyo
engineers involved in the project believed that the
new approach was essential to the project’s success
in that the ability to reduce a cooling system’s costs
by 18% demonstrated the power of a simultaneous
engineering approach. Neither Komatsu nor Toyo
undertook a formal analysis of the costs and bene?ts
of the intervention. However, the engineers at both
?rms were convinced that the savings far out-
weighed the incremental costs associated with the
concurrent cost management and interorganiza-
tional cost investigations interventions.
E?ectiveness of an IOCM technique
The e?ectiveness of an IOCM technique is cap-
tured by the net present value of the savings
derived from all suppliers using the technique
during a given time frame, minus the cost of
establishing and maintaining the technique and its
associated relational context. None of the ?rms in
the sample had even attempted a formal economic
justi?cation of their IOCM programs. Instead,
they relied upon their perceptions of the economic
success of the programs. For example, Komatsu
management adopted a holistic view of the make-
or-buy decision and the development of con-
current cost management. They believed that the
bene?ts of the new technique, quanti?ed by the
reduction in the costs of outsourced items, excee-
ded any additional investment, as well as any
coordination and transaction costs associated with
the new approach. However, they admitted that
there was no easy way for them to evaluate the net
bene?ts inherent to the new buyer–supplier rela-
tionship. The problem was that, while the
increased investment in speci?c assets theoretically
could be measured, along with their costs, the
incremental coordination and transaction costs
were di?cult to observe and to incorporate into
any formal economic evaluation of the overall
process.
Most of the bene?ts to Komatsu from its new
relationship with Toyo emerged from the
improved design capabilities of the dyadic pair.
These bene?ts included the lower costs and higher
functionality of the end products that resulted
from the joint design activities. The bene?ts to
Toyo, however, were less obvious. While the
value-added and the pro?t margin of an engine
cooling system were indeed higher than that of
Komatsu-designed components, Toyo now had to
invest in a more extensive research and develop-
ment program. Unfortunately, there was no way
to determine if Toyo’s pro?ts from its business
with Komatsu were higher under the new
approach, compared to the old one, because of the
Table 4
IOCM technique utilized and typical level of cost savings
IOCM technique utilized Type of item outsourced Relative value
of item outsourced
Typical level
of cost savings (%)
Concurrent cost management Major function High 10–15
Interorganizational cost investigation Group component Medium 5–10
FPQ trade-o? Component Low 0–5
22 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
problem of isolating the incremental revenues and
costs—in particular, the costs associated with the
new relational context. Neither ?rm had main-
tained detailed records of their prior investments
with one another; therefore, there was no way to
determine the incremental costs. Consequently, a
formal incremental pro?t analysis was not con-
sidered feasible. However, both Komatsu and
Toyo management stated that, in their opinion,
the two ?rms were better o? and that they actively
supported the new relationship. Literature sug-
gests that the decision to outsource is not a simple
trade-o? between increased performance and
increased transaction and coordination costs.
Dyer (1997) has suggested that in the type of rela-
tionship we have characterized as family member-
ship, ongoing transaction costs (and presumably
coordination costs) actually decrease, although
initial set-up costs may be higher.
Re-contracting
The success of Komatsu’s ?rst concurrent cost
management project with its ?rst family member
renders moot any discussion about how the ?rms
would have reacted, had the interaction resulted
into failure. Two possible alternatives suggest
themselves. First, the two ?rms could have
returned to their prior relational context and lim-
ited their IOCM practices to interorganizational
cost investigations and FPQ trade-o?s. Alter-
natively, they could have tried to rede?ne the
relational context so that concurrent cost man-
agement was successfully supported. Discussions
with both management teams reveal that the two
?rms had entered into a calculated gamble that,
together, they could reduce overall costs to a
greater extent than they could in isolation, or even
jointly under the relational context of a major
supplier. The process of achieving the desired sav-
ings for the A20 and A21 power shovels was a
risky and complex undertaking that required mul-
tiple joint design e?orts and encompassed multiple
solutions. The fact that neither ?rm had the tech-
nical expertise to solve the problem on its own was
critical. More speci?cally, the degree to which
overall cost was a function of joint design e?orts
appeared to be an important consideration.
Finally, the perceived inability to solve the cost
challenge using conventional approaches was a
signi?cant motivating factor for adopting a family
membership context and undertaking concurrent
cost management.
Summary and conclusion
The make-or-buy decision at the sample ?rms is
far more complex than the neo-classical economic
perspective suggests. Rather than being a simple
dichotomous competition between the ?rm and its
external suppliers, the process often involves a rich
interplay between buyer and supplier to ?nd ways
to take advantage of their disparate capabilities.
In particular, the sample ?rms have developed
IOCM techniques to overcome the information
asymmetry that exists between buyers and suppli-
ers, and to enable their design teams to coordinate
and cooperate e?ectively in order to identify low-
cost solutions by changing the speci?cations of the
outsourced items and sometimes the end product
itself.
The sample identi?ed three distinct clusters of
IOCM, each of which was associated with a dif-
ferent hybrid relational context. The relational
context established between the buyer and supplier
appeared to be linked to the IOCM technique that
supports the highest level of design changes
undertaken by the two ?rms, and thus involves the
highest level of interaction between the two design
teams. The ?rst cluster contained ?rms that could
undertake all three of the observed IOCM tech-
niques (FPQ tradeo?s, interorganizational cost
investigations, and concurrent cost management);
this cluster was associated with the relational con-
text family member, which provided the richest
support for interactions between buyer and sup-
plier. The second cluster contained ?rms that
could undertake only two of the IOCM techniques
(FPQ trade-o?s and interorganizational cost
investigations); it was associated with the rela-
tional context major supplier, which could support
an intermediate level of interaction between the
design teams at the buyer and supplier ?rms.
Finally, the third cluster consisted of ?rms that
could only undertake FPQ tradeo?s and was asso-
ciated with the relational context subcontractor,
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 23
which only supported low levels of buyer–supplier
interactions.
For ?rms that undertake IOCM, at least two
additional layers of complexity must enter into the
make-or-buy calculus. The ?rst layer deals with
the costs associated with an individual IOCM
intervention, including the time spent by the buy-
er’s engineers discussing and negotiating solutions
with the supplier, as well as the cost of formalizing
any design changes to the buyer’s product(s) and
the outsourced item. The second layer of com-
plexity further distances the make-or-buy decision
from the classical neo-economic analysis of inter-
nal versus external costs of outsourcing by intro-
ducing a higher level of analysis into the calculus
(Anderson, Glenn, & Sedatole, 2000). A single
outsourcing decision will rarely result in savings
su?cient to justify the adoption of a new IOCM
technique and its associated relational context.
Instead, the analysis must be applied to the sav-
ings associated with all of the suppliers likely to be
a?ected by the adoption of the new IOCM tech-
nique discounted over the time period that they
occur. These discounted savings then must be
compared to the incremental cost of developing
and maintaining the ability to undertake the tech-
nique and its associated relational context. Thus,
the calculus not only includes which parts will be
outsourced and to whom, but also the nature of
the relationship between the buyer and supplier
that should be adopted for the transaction.
Because of incomplete contracting, the calculus
has to be expanded to incorporate both bargain-
ing costs, where the parties are acting in their own
self-interest but in good faith, and opportunism
costs, where they are acting in their own self-
interest but in poor faith (Vining & Globerman,
1999).
The small number and limited diversity of the
sample ?rms used in this study re?ects the
exploratory nature of our research. To increase
the validity of the results and to help determine
the extent to which they can be generalized, the
?ndings need to be tested on a larger and more
diverse sample that includes ?rms from di?erent
countries and from a wider range of industries. In
addition, there are several extensions of this
research that appear worthy of further study.
First, in this paper, we have explored only a lim-
ited range of relational contexts and IOCM tech-
niques. One possible extension is to study other
types of relational contexts between ?rms, such as
strategic alliances and joint ventures, in which
di?erent forms of IOCM may emerge. Second, in
this paper, we explored IOCM practices that are
design-driven. However, nothing we observed
appears to be idiosyncratic to product design.
Another logical extension of this research would
be to examine the extent to which other forms of
joint action between ?rms trigger IOCM interven-
tions.
Acknowledgements
We thank Shannon Anderson, Jacob Birnberg,
Robert Kazanjian, Richard Makadok, Haya-
greeva Rao, the referees and participants at the
2000 AAA/IMA Conference, Anthony Hopwood,
the editor, and two anonymous reviewers for
helpful comments on the research.
References
Adler, P. (2001). Market, hierarchy, and trust: the knowledge
economy and the future of capitalism. Organization Science,
12(2), 215–234.
Anderson, S. W., Glenn, D., & Sedatole, K. L. (2000). Sourcing
parts of complex products: evidence on transactions costs,
high-powered incentives and ex-post opportunism. Account-
ing, Organizations and Society, 25(2000), 383–409.
Asanuma, B. (1989, March). Manufacturer-supplier relation-
ships in Japan and the concept of relation-speci?c skill. Jour-
nal of the Japanese and International Economies, 3(1), 1–30.
Atkinson, A. A., Banker, R. D., Kaplan, R. S., & Young, S. M.
(2001). Management accounting (10th ed). Upper Saddle
River, NJ: Prentice Hall.
Baiman, S. & Rajan, M. V. (2000). Incentive issues in inter-?rm
relationships [Working paper]. The Wharton School.
Barney, J. B., & Hansen, M. H. (1994). Trustworthiness as a
source of competitive advantage. Strategic Management
Journal, 15, 175–190.
Bensaou, M., & Venkatraman, N. (1995). Con?gurations of inter-
organizational relationships: a comparison between U.S. and
Japanese automakers. Management Science, 41(9), 1471–1492.
Birou, L. M., & Fawcett, S. E. (1994). Supplier involvement in
integrated product development: a comparison of US and
European practices. International Journal of Physical Distri-
bution & Logistics Management, 24(5), 4–14.
24 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
Bryce, D., & Useem, M. (1998). The impact of corporate out-
sourcing on company value. European Management Journal,
16(6), 635–643.
Carr, C., & Ng, J. (1995). Total cost control: Nissan and its
U.K. supplier partnerships. Management Accounting
Research, 6, 347–365.
Clark, K. B., & Fujimoto, T. (1991). Product development per-
formance. Boston: Harvard Business School Press.
Contractor, F. J., & Lorange, P. (1988). Why should ?rms
cooperate? The strategy and economics basis for cooperative
ventures. In F. J. Contractor, & P. Lorange (Eds.), Coop-
erative strategies in international business. Lexington: Lex-
ington Books.
Cooper, R., & Chew, W. B. (1996). Control tomorrow’s costs
through today’s designs. Harvard Business Review, 1, 88–97.
Cooper, R., & Slagmulder, R. (1997). Target costing and value
engineering. Portland, OR: Productivity Press.
Cooper, R., & Slagmulder, R. (1999). Supply chain development
for the lean enterprise. Interorganizational cost management.
Portland, OR: Productivity Press.
Cooper, R., & Yoshikawa, T. (1994). Inter-organizational cost
management systems: the case of the Tokyo-Yokohama-
Kamakura supplier chain. International Journal of Produc-
tion Economics, 37, 51–62.
Davila, T. (2000). An empirical study on the drivers of manage-
ment control systems’ design in new product development.
Accounting, Organizations and Society, 25(4–5), 383–409.
Drtina, R. E. (1994, March). The outsourcing decision. Man-
agement Accounting (USA), 56–62.
Dyer, J. H. (1996). Specialized supplier networks as a source of
competitive advantage: evidence from the auto industry.
Strategic Management Journal, 17(4), 271–291.
Dyer, J. H. (1997). E?ective inter?rm collaboration: how ?rms
minimize transaction costs and maximize transaction value.
Strategic Management Journal, 18(7), 535–556.
Dyer, J. H., & Ouchi, W. (1993). Japanese-style partnerships:
giving companies a competitive edge. Sloan Management
Review, 35(1), 51–63.
Dyer, J. H., & Singh, H. (1998). The relational view: coopera-
tive strategy and sources of interorganizational competitive
advantage. Academy of Management Review, 23(4), 660–679.
Eisenhardt, K. (1989). Building theories from case study
research. Academy of Management Review, 14(4), 532–550.
Gietzmann, M. B. (1996). Incomplete contracts and the make
or buy decision: governance design and attainable ?exibility.
Accounting, Organizations and Society, 21(6), 611–626.
Gietzmann, M. B., & Larsen, J. G. (1998). Motivating sub-
contractors to perform development and design tasks. Man-
agement Accounting Research, 9, 285–309.
Gilley, K. M., & Raheed, A. (2000). Making more by doing
less: an analysis of outsourcing and its e?ects on ?rm per-
formance. Journal of Management, 26(4), 763–790.
Granovetter, M. (1985). Economic action and social structure:
the problem of embeddedness. American Journal of Sociol-
ogy, 91, 481–510.
Gulati, R., & Singh, H. (1998). The architecture of coopera-
tion: managing coordination costs and appropriation con-
cerns in strategic alliances. Administrative Science Quarterly,
43(4), 781–815.
Ha? gg, I., & Hedlund, G. (1979). ‘‘Case studies’’ in accounting
research. Accounting, Organizations and Society, 4(1/2), 135–
143.
Heide, J. B., & John, G.(1990).Alliances in industrial purchas-
ing: the determinants of joint action in buyer-supplier rela-
tionships (1990). Journal of Marketing Research, XXVII, 24–
36.
Helper, S. (1996). Incentives for supplier participation in pro-
duct development: evidence from the U.S. auto industry. In
T. Nishiguchi (Ed.), Managing product development (pp. 165–
189). New York: Oxford University Press.
Hines, P. (1994). Creating world-class suppliers. London: Pit-
man.
Holden, N., & Burgess, M. (1994). Japanese led companies.
London: McGraw Hill.
Hopwood, A. G. (1983). On trying to study accounting in the
contexts in which it operates. Accounting, Organizations and
Society, 8(2–3), 287–305.
Hopwood, A. G. (1996). Looking across rather than up and
down: on the need to explore the lateral processing of infor-
mation. Accounting, Organizations and Society, 21(6), 589–
590.
Horngren, C. T., Foster, G., & Datar, S. M. (2000). Cost
accounting. A managerial emphasis. Upper Saddle River, NJ:
Prentice Hall.
Kamath, R. R., & Liker, J. K. (1994). A second look at Japa-
nese product development. Harvard Business Review, 72(6),
154–170.
Kato, Y., Boer, G., & Chow, C. W. (1995). Target costing: an
integrative management process. Journal of Cost Manage-
ment, 9, 39–51.
Koga, K. (1998). Determinants of e?ective product cost man-
agement during product development: opening the black box of
target costing [Working paper]. Harvard Business School.
Kotabe, M., & Murray, J. (1990). Linking product and process
innovations and modes of international sourcing in global
competition: a case of foreign multinational ?rms. Journal of
International Business Studies, 3, 383–408.
Korczynski, M. (1996). The low trust route to economic devel-
opment: inter?rm relations in the U.K. engineering con-
struction industry in the 1980s and 1990s. Journal of
Management Studies, 33(6), 787–808.
Lacity, M. C., Wilolocks, L. P., & Feeny, D. F. (1996). The
value of selective IT sourcing. Sloan Management Review,
Spring, 13–25.
Lamming, R. (1993). Beyond partnership. Strategies for innova-
tion and lean supply. New York: Prentice Hall.
Lukka, K., & Kasanen, E. (1995). The problem of general-
izability: anecdotes and evidence in accounting research.
Accounting, Auditing, & Accountability Journal, 8(5), 71–90.
McKinnon, J. (1988). Reliability and validity in ?eld research:
some strategies and tactics. Accounting, Auditing & Account-
ability Journal, 1(1), 34–54.
Monden, Y. (1995). Cost reduction systems. Target costing and
Kaizen costing. Portland, OR: Productivity Press.
R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26 25
Monczka, R. M., & Trent, R. J. (1991). Evolving sourcing
strategies for the 1990s. International Journal of Physical
Distribution and Logistics Management, 23(4), 42–54.
Mudambi, R., & Helper, S. (1998). The ‘close but adversarial’
model of supplier relations in the U.S. auto industry. Strat-
egic Management Journal, 19(8), 775–792.
Nishigushi, T. (1994). Strategic industrial sourcing. New York:
Oxford University Press.
Nooteboom, B., Berger, H., & Noorderhaven, N. (1997).
E?ects of trust and governance on relational risk. Academy
of Management Journal, 40(2), 308–338.
Ouchi, W. (1980). Markets, bureaucracies, and clans. Admini-
strative Science Quarterly, 25, 125–141.
Parkhe, A. (1998, Winter). Building trust in international alli-
ances. Journal of World Business, 33(4), 417–437.
Powell, W. W. (1990). Neither market nor hierarchy: network
forms of organization. In L. L. Cummings & B. Staw (Eds.),
Research in Organizational Behavior, 12, 295–336.
Prahalad, C. K., & Hamel, G. (1990). The core competence of
the corporation. Harvard Business Review, 68(3), 79–93.
Quinn, J. B. (1992). Intelligent enterprise: a knowledge and ser-
vice based paradigm for industry. New York: Free Press.
Ring, P. S., & Van de Ven, A. H. (1989). Formal and informal
dimensions of transactions. In A. H. Van de Ven, H. Angle,
& M. S. Poole (Eds.), Research on the management of inno-
vation: the Minnesota studies (pp. 171–192). New York: Bal-
linger/Harper Row.
Ring, P. S., & Van de Ven, A. H. (1994). Developmental pro-
cesses of cooperative interorganizational relationships.
Academy of Management Review, 19(1), 90–118.
Sako, M. (1992). Prices, quality and trust: inter-?rm relationships
in Britain and Japan. Cambridge: Cambridge University Press.
Sako, M., & Helper, S. (1998). Determinants of trust in sup-
plier relations: evidence from the automotive industry in
Japan and the United States. Journal of Economics, Behavior
& Organization, 34, 387–417.
Sheppard, B. H. & Tuchinsky, M. (1996). Inter?rm relation-
ships: a grammar of Pairs. In L. L. Cummings & B. Staw
(eds.), Research in Organizational Behavior, 18, 331–373.
Thompson, J. D. (1967). Organizations in action. New York:
McGraw Hill.
Van der Meer-Kooistra, J., & Vosselman, E. G. J. (2000).
Management control of inter?rm transactional relationships:
the case of industrial renovation and maintenance. Account-
ing, Organizations and Society, 25(1), 51–77.
Venkatraman, N. (1989). Beyond outsourcing: managing IT
resources as a value center. Sloan Management Review, 38,
51–64.
Vining, A., & Globerman, S. (1999). A conceptual framework
for understanding the outsourcing decision. European Man-
agement Journal, 17(6), 645–654.
Walker, G., & Weber, D. (1984). A transaction cost approach
to make-or-buy decisions. Administrative Science Quarterly,
29, 373–391.
Ward, A., Liker, J. K., Cristiano, J. J., & Sobek, D. K. (1995).
The second Toyota paradox: how delaying decisions can
make better cars faster. Sloan Management Review, 36(3),
43–61.
Williamson, O. E. (1975). Markets and hierarchies. New York:
Free Press.
Williamson, O. E. (1979). Transaction cost economics: the
governance of contractual relations. Journal of Law and
Economics, 22, 233–261.
Williamson, O. E. (1985). The economic institutions of capital-
ism. New York: The Free Press.
Williamson, O. E. (1991). Comparative economic organization:
the analysis of discrete structural alternatives. Administrative
Science Quarterly, 36, 269–296.
Yin, R. K. (1984). Case study research, design and methods.
Sage Publications.
Zaheer, A., & Venkatraman, N. (1995). Relational governance
as an interorganizational strategy: an empirical test of the
role of trust in economic exchange. Strategic Management
Journal, 16(5), 373–392.
Zenger, T., & Hesterly, W. (1997). The disaggregation of
corporations: selective intervention, high-powered incen-
tives, and molecular units. Organization Science, 8(3), 209–
222.
26 R. Cooper, R. Slagmulder / Accounting, Organizations and Society 29 (2004) 1–26
doc_197179554.pdf