Description
Drawing on published work by Williamson Vournal of hw and Economics (October 1979) pp. 233-
261; The Economic Institutions of Capitalism (New York: Free Press, 1985)], Spicer [Accounting,
OrganFzatfons and Society (1988) pp. 302-3221, and Walker [Interfaces (May-June 1988) pp. 62-731
this paper first develops a theory of the sourcing and transfer pricing process from a transaction costs
perspective.
Pergamon Accounting, Organlzatfonr and So&q, Vol. 20, No. 6, pp. 423-456, 1995
copyrighr D 1995 Elsevkr Science Ltd
Printed in Great Brltaln. AU r&M reserved
0361-3682/95 $9.50+0.00
03613682(95)000074
A MULTI-CASE INVESTIGATION OF A THEORY OF THE TRANSFER PRICING
PROCESS*
GARY J. COLBERT
Utahem@ of Col orado at Denver
and
BARRY H. SPICER
The UniversZty of Auckl and
Abstract
Drawing on published work by Williamson Vournal of hw and Economics (October 1979) pp. 233-
261; The Economic I nstitutions of Capitalism (New York: Free Press, 1985)], Spicer [Accounting,
OrganFzatfons and Society (1988) pp. 302-3221, and Walker [Interfaces (May-June 1988) pp. 62-731
this paper first develops a theory of the sourcing and transfer pricing process from a transaction costs
perspective. The paper then reports on an empirical investigation of the theory using a multicase
research design involving internal ttansfers in four large, vertically integrated high-technology compa-
nies in the electronics industry in the U.S.A. The ways in which these firms manage internal transfers and
set transfer prices are found to be positively related to the reported degree of transaction-specific
investment (asset specfficity) associated with each transfer and to the strategic importance of the
transferred component and the nature of the component division’s production capabilities. Suggestions
are made for future research.
The transfer pricing process’ represents an
important and pervasive problem in designing
and implementing management information
and control systems. There has been consider-
able academic research published on these
related topics, yet it remains a troublesome
area for academics and managers alike. Even
though significant strategic and behavioural
consequences including suboptimal decision
making, opportunistic behaviour by subunit
managers, and internal friction and dishar-
mony may ensue from the transfer pricing pro-
cess, we do not understand many of the factors
that condition the way in which managers actu-
ally deal with this process.
There appear to be two reasons for this, the
first being theoretical and the second empiri-
cal. Reviews of the transfer pricing literature
reveal many normative treatments of the trans-
fer pricing issue (see Abdel-khalik & Lusk,
1974, and Grabski, 1985, for extensive
reviews) but fewer attempts such as those
made by Eccles (1985) to induce, or Watson
& Baumler (1975) and Spicer (1988) to
deduce, theories of the transfer pricing
process.
l The authors would like to thank Jean Cooper, gay King, Terry O’Kecfe, Frank Selto and Ralph Viator for their helpful
comments on earlier dnxfts of this paper.
1 We use the tetm transfer pricing process to refer to all related transfer activity, including the decision to source
components internally or externally (make-or-buy), the extent to which senior managers assert control over transfer
relationships between subunits, and how tramfer prices are set.
423
424 G. J. COLBERT and B. H. SPICER
The second and related reason is that much of
the empirical research on actual transfer pricing
processes has been done with firm level surveys
(e.g Tang, 1979, 1980; Vancil, 1979; Borowski,
1990). We believe that the understanding and
insights that have emerged from these types of
studies are limited for two important reasons.
First, they abstract from both industry and orga-
nizational context by choosing samples which
span a number of industries, and second, they
are directed at the wrong level, i.e. at the level
of the firm as a whole rather than at the subunit
level where internal transfers actually take
place. The approach we adopt in this paper
attempts to deal with both of these issues in a
meaningful way.
The objective of this paper is to report on a
theoretically based empirical investigation of
factors relevant to the transfer pricing process
in high-technology companies. Using a multi-
case research design we address two research
questions:
(1) How do the business units of high-tech-
nology companies manage the transfer process
and why do these practices differ?
(2) How do the business units of high-tech-
nology companies set transfer prices for parti-
cular component transfers and why do the
methods used differ?
Drawing on work by Williamson (1979,
1985) Spicer (1988) Walker (1988) and Col-
bet-t (1991) we first set out a theory of the
transfer pricing process from a transaction
costs perspective. We then report on an
empirical investigation of the theory using a
multi-case research design in which the pat-
tern of observations made in each case relative
to those suggested by the theory is evaluated.
Our overriding objective is to assess whether
the evidence from multiple cases is consistent
or inconsistent with the pattern of observations
suggested by theory.
Our paper is structured as follows. The sec-
ond section sets out our theory of the transfer
pricing process. The third section presents the
details of the multi-case research design we
employed to select our cases and collect and
analyse evidence from them. The fourth sec-
tion describes each of our four cases. The fifth
section presents and analyses our case evi-
dence and assesses the support the case evi-
dence provides for the theory of the transfer
pricing process laid out in the second section.
The final section provides a summary and con-
clusions which includes a discussion of the
limitations of our study and some suggestions
for future research.
A THEORY OF THE TRANSFER PRICING
PROCESS
Starti ng poi nts
There are two starting points for the theory of
the transfer pricing process that we develop.
The first is a growing body of work that uses
the existence of differential transaction costs as
an explanation for the existence of variation in
the manner in which organizations conduct
their economic activity. Generally referred to
as transaction costs economics, this framework
can be traced to Coase (1937) and Williamson
(1975, 1979, 1985). Spicer (1988) draws on
transaction costs economics to develop a posi-
tive theory of the transfer pricing process in
which the strategic and transactional character-
istics of specific transfers are related to transfer
pricing issues and the organizational processes
used to manage transfers within firms. Based on
the premise that internal transfers are transac-
tions, he argues that transaction costs econom-
ics can be used to consider contracting issues
involved with the management of internal trans-
fers and the setting of transfer prices.
Although Spicer (1988) argues that firm strat-
egy affects intermediate product design and is
related to the dimensions of intra-Iirm transfers
he provides few particulars. This missing detail
is supplied by Walker (1988) who applies the
concepts of Porter’s (1980, 1985) competitive
strategy framework and transaction costs eco-
nomics to explore the risks associated with a
firm’s sourcing decisions, and the implication
of these risks for competitive strategy. Walker’s
work provides the second starting point for our
development of theory.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 425
Transacti on costs economi cs
Williamson (1985, p. 1) argues that the
appropriate level of economic analysis is the
transaction which he defines as the transfer
of goods or services across a technologically
separable interface. Transaction costs econom-
ics focuses attention on the relative costs and
hazards of conducting transactions within alter-
native governance structures.2 As the costs of
conducting transactions within markets
increase it becomes increasingly likely that
firms will resort to alternative arrangements
such as internalizing the transaction. The level
of transaction costs experienced is determined
by the dimensions of the transaction.
The first of these is asset spec@ci ty3 which is
of special importance because without it trans-
action costs economics would lose much of its
significance (Williamson, 1985, p. 52). Asset
specificity arises when durable investments
are made in relation to a particular transaction
and the value of the investment in its next best
use is considerably lower. The value of such
investments in assets is dependent on the main-
tenance or continuation of a particular transac-
tional relationship. If the asset cannot be
costlessly redeployed from its present use to
some other use, the value of the investment
can be jeopardized by the opportunistic action
of the other party to the transaction. Thus the
parties have an incentive to maintain the rela-
tionship through the implementation of appro-
priate contractual or organizational safeguards.
As the level of asset specificity increases, the
transaction costs associated with firms con-
ducting transactions within markets rise and it
becomes increasingly likely that an alternative
arrangement such as internalization of the
transaction (vertical integration) will have
transaction costs advantages.
Williamson (1985, p. 55) identifies a number
of types of asset specificity which involve
investments in the transactions by one or
both parties to the transaction. These include:
(1) physical asset specificity resulting from
investments in tangible assets, e.g. plant, equip
ment, tooling, dies, etc. that are more or less
unique to this economic relationship; (2)
human asset specificity resulting from invest-
ments in specialized training or specialized
know-how gained through education or experi-
ence which is not easily transferred to other
transactions; (3) site specificity which results
from parties locating their operations in close
physical proximity to one another for flow
economies; and (4) dedicated asset specificity
which arises from investments in generalized
(as contrasted to special purpose) productive
capacity for the explicit purpose of selling a
significant amount of a product to a specific
customer and that capacity cannot be cost-
lessly redeployed or otherwise utilized.
There is growing interest in the notion of
asset specificity as an important explanator of
the boundaries of the firm and the related issues
of strategy; the make-or-buy decision; and verti-
cal, horizontal and diversification alliances
(Alchian, 1984; Alchian & Woodward, 1987;
Riordan & Williamson, 1985; Teece, 1986;
Jones & Hill, 1988; Klein, 1988; Reve, 1990;
Williamson, 1991). There is also a body of
empirical research that provides support for
the relevance of asset specificity. Of particular
pertinence to our study are several papers
which uncover a positive relationship between
the presence of specialized assets and the
a Transaction costs are the costs of engaging in transactions via a particular institutionaI arrangement or governance
stmcturc. They can be separated into exunre and ez-posr transaction costs. Eranfe transaction costs are the costs of
drafting, negotiating and safeguardlng contracts. E&r-post transacti ons costs include maktdaption and haggling costs when
transactions get out of alignment, the mnnlng costs of governance structures to resolve disputes, and the bonding costs to
secure credible commitments. Spicer & Rallew (1983, p. 88) point out that intemai frictions within iirms arising from
contml losses and pIanning difiicuhies can be viewed as the hierarchical counterpart of transaction costs in markets.
3 The terms transaction-specific investment and specialized investments arc also used to refer to asset specificity. We use
these terms interchangeably.
426 G. J. COLBERT and B. H. SPICER
extent of vertical integration (Monterverde &
Teece, 1982; Walker & Weber, 1984; Masten,
1984; Masten et al ., 1989; Levy, 1985).*
Transaction costs economics recognizes two
other dimensions of transactions: the uncer-
tafnnty to which the transaction is subject and
the extent of the transaction. Uncertainty is of a
statecontingent kind. When assets cannot be
costlessly redeployed and uncertainty is pre-
sent, efficient adaptations are needed in order
to preserve the relationship. Extent refers to the
frequency and volume of the transaction and
determines whether the cost of specialized gov-
ernance structures to control the transaction
can be recovered. The cost of specialized gov-
ernance structures (e.g. hierarchy) are easier to
recover for large and recurring transactions.
Appl i cati on of transacti on costs
economi cs to the i nternal transfer process
As Spicer (1988) points out, the extent to
which a Iirm makes internal transfers arises
out of a strategic choice between buying com-
ponents from external sources and making
them itself. This is the issue of the appropriate
extent of vertical integration. Porter (1980, pp.
303-307) cites the potential benefits of vertical
integration as including economies of opera-
tions, tapping into technology, assuring sup
ply, offsetting bargaining power and input
cost distortions in noncompetitive markets,
and enhancing ability to differentiate end pro-
ducts. The transaction costs economics frame-
work offers a particular explanation for vertical
integration. In this framework the make-or-buy
decisions will be made so as to minimize the
sum of production and transaction costs.
Dimensions of the transactions involved (asset
specificity, uncertainty and extent) are thought
to be positively related to the decision to inte-
grate vertically into component production.
With low levels of transaction-specific invest-
ments (asset specificity) it is likely that econo-
mies of scale play an important role in the make-
or-buy decision. Suppliers who can aggregate
demand for multiple customers, thereby obtain-
ing efficient operating scales, will be the low
cost producers. At the same time, with low
levels of transaction-specific investments
involved, the potential transaction costs asso-
ciated with buying externally are likely to be
low. Given these conditions it is less likely that
the firm will source the component internally.
As the investments required to produce a
particular component become more transac-
tion-specific, the production cost advantages
will be less between sellers and buyers but
they will face contracting hazards which will
result in additional contracting costs. These
costs include transaction costs of writing, mon-
itoring and executing contracts. Because pro-
duction cost advantages of buying externally
decrease and the comparative transaction
costs of markets increase as assets involved in
the transaction increase in specificity, it
becomes increasingly likely that the end user
will bring the transaction inside the &rn and
make the component.
It is not only sellers that make transaction-
specific investments. Buyers may also make
investments where the value of their invest-
ment is dependent on the conduct and perfor-
mance of a particular supplier. For example,
investments may be made in equipment, train-
ing of personnel or end product and/or compo-
nent design that are dependent on a
component produced exclusively by one sup
plier. The presence of these transaction-speci-
fic investments by the buyer means that there
are potentially high transaction costs involved
in contracting to acquire the related compo-
nent externally.
* Monterverde & Teece (1982) Walker & Weber (1984) and Masten et al. (1989) provide tests of the relevance of asset
speciticity in the automobile segment. Masten (1984) provides a test in the aerospace industry, and Levy’s (1985) tests
spans a cross-section of industries. The presence of specialized assets is pmxied by design costs and number of compe-
titors in Monterverde & Teece (1982); by supplier competition in Walker & Weber (1984); by engineering effort, the
extent to which components are produced on physical assets specific to an automobile maker, and the closeness of siting
of manufacturing stages in Masten et al. (1989); and by research intensive products in Levy (1985).
TRANSFER PRICING PROCESS
Uncertainty in the presence of transaction-
specific investments creates the need for adap-
tation and co-ordination between the contract-
lng parties. The greater the extent (frequency
and volume) of the transaction the more likely
the firm can obtain production economies by
manufacturing the component itself and lnter-
nalixing the transaction.
However, intemalizatlon of the transaction
within the firm does not fully remove the poten-
tial transaction costs. Where the firm is highly
decentralized, transfers between relatively inde-
pendent subunits are similar in some contract-
ing aspects to external transactions. In the limit
a firm could structure and encourage conduct
between subunits that is essentially equivalent
to that of external parties transacting in the
market-place. Alternatively, firms may arrange
their organization structures and control sys-
tems so as to encourage more cooperative,
interdependent behaviour between subunits.
Where the level of transaction-specific
investments in internal transfers made by sub
units is low, the tirm’s economic interests will
generally be served by allowing external mar-
ket forces to influence the transfer, including
the setting of the transfer price. However, as
investments in assets specialized to internal
transfers increase, the firm faces a trade-off
between the potential efficiencies of trading
in a market and its need to protect the value
of the specialized investment.
As the level of transaction-specitic investment
increases, it becomes increasingly important for
the firm to protect its economic interests in the
transfer relationship by ensuring that conflict
and opportunistic actions by subunits are mod-
THEORY - INVESTIGATION 427
erated or controlled. Numerous researchers,
including Vancil (1979) and Fccles (1985)
have noted that relationships between divisio-
nallxed firms can at times involve considerable
conflict, especially related to transfer pricing.
One possible resolution of this conflict is to
seek alternative sources or customers in the
market. However, from the tirm’s perspective,
market alternatives are less attractive for the
relief or control of disagreements and conflict
between subunits when asset specificity is high.
The most critical issue in managing the tran-
fer process is the degree to which the Iirm acts
to constrain the actions of the subunits in mak-
ing sourcing and selling decisions. Where adap-
tation and coordination of the transfer are
important because of transaction-specific
investments made by the internal selling unit
in an uncertain environment, corporate man-
agers are most likely to protect the firm’s eco-
nomic interests by asserting control over the
transfer, potentially to the point of mandating
the sourcing decision.5 Also, as discussed
above, buyers may also make transaction-speci-
fic investments where the value is dependent
on the conduct and performance of a particular
supplier. Consequently, it follows that when a
buying unit’s level of transaction-specific invest-
ment in the transfer is high, corporate managers
are most likely to protect the firm’s economic
interests by asserting control over the transfer
by constraining the selling decision. However,
when the level of transaction-specific invest-
ment made by a buying unit is low, there will
be few (if any) reasons associated with the need
to protect transaction-specific investments for
constraining the selling decision.6
’ Spicer (1988) suggests that other actions to protect the firm’s interests might Include (1) involving corporate manage-
ment in the arbitration of disputes between the subunits, (2) adjusting performance and Incentive systems to emphasize
performance measutement and incentives that are tied to higher level or joint performance and de-emphasizing perfor-
mance measurements and incentives that focus only on individual subunit performance. However, we conjecture that the
importance of these actions is likely to be much less than the decision to control the transfer process by constraining
sourcing and selling decisions.
6 There may be other factors which are not part of our theory which can influence the extent to which the selling unit is
constrained from selling outside the firm. Some of these other factors observed in our case studies are reported later in the
paper.
428 G. J. COLBERT and B. H. SPICER
Therefore, in our case studies we should
observe that the degree to which Iirms con-
strain subunit managers in sourcing and selling
decisions is, in general, positively related to the
degree of transaction-specific investment (asset
specificity) made by their counterpart(s) to the
internal transfer.
Strategi c sourci ng consi derati ons
Walker (1988) provides a related perspective
by applying the concepts of Porter’s (1980,
1985) competitive strategy framework to
explore the strategic aspects of transaction
costs inherent in the sourcing decision. In Por-
ter’s framework, firms pursue generic strate-
gies such as cost leadership, product
differentiation and focus which are sources of
competitive advantage. Walker assumes that
firms try to order inputs and operations in
terms of their value in contributing to compe-
titive advantage. For example, a supplier may
supply a custom component that provides
important differentiation features in a buyer’s
end product.
Walker’s Insight is that the risk associated
with sourcing relationships can have implica-
tions for the value of the firm’s competitive
strategy. His general argument is that inputs
that are important to the firm’s strategy involve
greatest risk because supply failure can lead to
greater decline in the firm’s performance than
failures with respect to less strategic inputs.
Walker identifies two types of sourcing risk
with strategic implications.
Appropri ati on ri sk arises when a supplier
owns and sells a factor of intrinsic value to
the firm’s strategy or the Iirm has strategically
important assets that are dependent on (i.e.
specialized to) aspects of the supplier’s goods
or services. Firms with suppliers of strategically
important components and high switching
costs are exposed to the risk of the supplier
raising prices or failing to deliver components
to specifications thereby reducing the value of
the firm’s competitive strategy. Appropriation
risk can also arise from dependence on the
supplier’s capabilities necessary for implement-
ing the firm’s strategy, for example the speed
with which a supplier can turn around a pro
totype at the design stage may be crucial to
gaining advantage by being first to market.
Under these conditions, if the buyer has high
switching costs, the supplier can opportunisti-
cally raise prices or fail to perform adequately,
thereby reducing the value of the lirm’s
strategy.
Di ffuston ri sk arises when innovative pro-
prietary products or process technology that
creates competitive advantage can be repli-
cated or Imitated by competitors. In the design
and production of some products suppliers
must have access to information about proprie-
tary processes or designs. A supplier external
to the firm with important knowledge of the
firm’s important processes and designs may
leak this imformation to the firm’s competitors.
To manage or control strategic sourcing risk
the firm may decide to Internalize the transac-
tions by investing in facilities to produce the
component. Walker concludes that vertical
integration will vary directly with the level of
strategic sourcing risk and the qualifications of
the Iii-m’s Internal source relative to those of
the best outside supplier.
However, the strategic implications of sour-
cing risk are also applicable to internal transfers
in a highly decentralized and divisionalized
organization.7 An internal supplier’s actions
can adversely affect the value of the competi-
tive strategy of end product divisions. For
example, by failing to meet contracted deliv-
ery schedules on components an internal sup-
plier may cause an end product division to miss
a market window. Similarly, an end product
division may switch to an external supplier
adversely affecting the value of investment in
component operations. These are examples of
appropriation risk. By selling proprietary com-
’ Walker’s analysis suggests that internal transfers will consist primarily of components where the qualification of the
internal source relative to the best external source and the strategic sourcing risk are high.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
429
ponents externally, internal suppliers may pro-
vide competitors with access to processes,
designs and components that are strategically
important to end product divisions. This is an
illustration of diffusion risk.
Therefore, in our case studies we should
observe that the degree to which firms con-
strain subunit managers in sourcing and selling
decisions is positively related to the strategic
importance of the component and/or the com-
ponent supplier’s capabilities to the firm’s com-
petitive strategy. Constraints on external
sourcing and selling of proprietary compo-
nents will serve to protect proprietary technol-
ogy and large specialized investments in
strategically Important transfers. In this respect
the issues of strategic importance and asset
specificity are closely bound together.
Transfer pri ce method
How transfer prices are set can have impor-
tant consequences for the tirm as a whole and
its subunits through their effects on local deci-
sion making and performance measurement.
When the level of transaction-specific invest-
ment required to support the transfer is low,
intermediate products will have a low degree
of customization and are most likely to be simi-
lar to products available in the market. As Spi-
cer (1988) points out there is no transfer
pricing problem in this case. The firm’s inter-
ests will be best served by using external mar-
ket referents as the primary basis for
determining internal terms of trade and to
inform decision making. Also, subunit man-
agers are likely to accept market prices as a
suitable measure of the value of the internal
transfer and for performance measurement.
Transfers will typically be negotiated by divi-
sional managers at transfer prices close to mar-
ket prices with adjustments confined to such
things as volume discounts, differences in
terms of sale and delivery, and the costs of
customization. For the most nart the firm’s
interests are protected by competition
amongst internal and external suppliers.
As the level of transaction-specific invest-
ment in the transfer by the component divi-
sion increases, intermediate components will
become increasingly customized and idiosyn-
cratic to the end product division. Conse-
quently,
market prices will become
increasingly less relevant for informing the
transfer pricing process and, as noted above,
corporate managers are more likely to impose
constraints on sourcing decisions by end pro-
duct divisions. In some mid-range of transac-
tion-specific investment by the component
division, it is likely that both costs and market
referents in the form of competitive bids and/
or quotations will be used to set internal terms
of trade.
However, at higher levels of asset specificity
in the component division, which results in a
considerable degree of customization (idiosyn-
crasy) in the intermediate product, internal
manufacturing costs will become the primary
basis for setting transfer prices. As noted
above, there will also be strong constraints
placed on the sourcing decision. In these cir-
cumstances the idiosyncratic nature of the
component division’s production process and
the intermediate product being transferred will
make it extremely difficult for managers to find
prices in external markets which are particu-
larly useful for informing the transfer price set-
ting process. With the presence of specialized
assets, any market prices that can be obtained
are likely to be only crude approximations of
the value of the internally transferred compo-
nent. Forcing the use of external prices as the
primary basis for setting internal terms of trade
in the presence of specialized assets may be
disruptive and counterproductive.
This same general point is made strongly by
Ball (1989, pp. 19-20) based on his analysis of
lirms as specialists in minimizing contracting
costs. He writes:
By construction, the more specific the firm’s factors of
production, the less its internal contracts wlII resemble
contracts observed in external markets and the less it
will find market prices to be informative. For highly
specific factors, quasi-prices [which include transfer
prices] will bear little resemblance to any observa-
ble external market prices. Thus, market prices can
provide useful, but not sufficient, information for deter-
430
G. J. COLBERT and B. H. SPICER
mining quasi-prices, and their usefulness varies as an
inverse function of factor specificity.
While asset specificity may not be the only
factor which affects the setting of transfer
prices in divisionalized firms, our theory leads
us to expect that it will exert a systematic and
powerful in8uence.s Therefore, in our case stu-
dies we expect to observe that the weight
given to manufacturing costs in setting transfer
prices is positively associated with the extent
of asset specificity related to the transfer in the
component division.
MULTI-CASE STUDY DESIGN
To investigate the case questions and the
theoretical propositions developed above we
use a multi-case research design with multiple
(embedded) units to collect evidence. This
multi-case design is the most sophisticated of
the four case designs discussed by Yin (1989,
pp. 46-60).9 Yin (1989) was selected because
this book is a standard reference for case
research design and provides the clearest,
most coherent and comprehensive approach
to designing, carrying out and presenting
mutli-case explanatory case research (as
opposed to descriptive or exploratory case stu-
dies) that we were able to find. However, we
also found work by Bisenhardt (1989), McKin-
non (1988) Bruns & Kaplan (1987) Merchant
(1987) Maher (1987), Palepu (1987) Herten-
stem (1987) and Merchant & Manzoni (1989)
helpful to our thinking about how to approach,
design and write up the results of our study.
We chose to use a case research design
because this method is appropriate to the man-
ner in which we have framed the research
questions and our desire to understand which
factors are important in the design and imple-
mentation of transfer processes and the setting
of transfer prices In specific organizational con-
texts. As our theory suggests that the proper
level of analysis is the transfers which occur
within each firm, a number of selected trans-
fers of intermediate products (components)
were studied in each ‘of four large, vertically
integrated high-technology companies in the
electronics industry in the U.S.A.
The firms are treated as the cases, the
embedded units are the divisions between
which the specific transfers take place. Using
a multi-case research method makes it possible
to study particular transfers within each of
these firms.
Case sel ecti on
We targeted firms in the electronics industry
because we believed different degrees of asset
specificity would characterize h-ma-firm trans
fers in this industry. As discussed above, we
expected this factor to be a principal determi-
nant of how these firms manage their internal
transfer pricing processes.
Our analysis of internal transfers in the tar-
geted Iirms focuses primarily upon asset speci-
ficity and related strategic considerations. We
ensure that uncertainty is present by selecting
our cases from an industry which is acknowl-
edged to be volatile and continuously chan-
ging. In addition, within each selected firm,
we focus on significant transfers from large
component divisions. By definition these
s Other Erm-specific factors affecting the setting of transfer prices observed in the course of our cast studies are reported
later in the paper.
9 Types of case research designs arc (1) singlecase design with a single unit of analysis, (2) multiplecase design with a
single unit of analysis, (3) singlecase design with multiple (embedded) units of analysis, and (4) a multiple-case design with
multiple (embedded) units of analysis. Recent case research in management accounting has mostly used either designs of
type 1 or type 2. We are not aware of other contemporary case research in management accounting that uses a type 4
design. The closest is a field study by Merchant & Manzooi (1989) which is conducted at the level of profits centres in
multiple firms (cases). However, their study does not follow case research methods which requires that the data for each
embedded unit (profit centre) be first analyzed for each firm before patterns are analysed and compared across cases.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 431
Involve each firm’s largest component transfers
which are big enough to attract management
attention. However, throughout the cases we
also discuss some transfers of lesser extent.
ponent units making these transfers vary in
terms of their strategic roles and the degree
of asset specificity involved.
General descriptive information about the
cases selected for study and the nature of
embedded units (i.e. the types of transfers
from component divisions studied) are given
in Table 1. The identity of each of the four firms
has been disguised but it can be seen from the
data in the table that they are all large, highly
divisionalized firms in the electronics industry.
Data col l ecti on
The attribute common to these firms is that
each transfers electronic components between
component and end product divisions. The
types of component transfers we focus on are
transfers shown at the bottom of Table 1.
These include the most important and largest
transfers in these companies. It can be seen
that there are overlaps between the types of
transfer studied in each case. This was done
deliberately to provide us with a strong basis
for comparison. Our focus is on the transfers of
similar types of components in each case, i.e.
integrated circuits and circuit boards. The com-
Gai ni ng high level access to the companies
that agreed to participate in this study required
considerable time and effort. We first sent let-
ters of inquiry to corporate level finance or
accounting managers. We followed up our let-
ters with numerous telephone calls to the com-
panies involved.” Our objective was to gain
access to at least four large, vertically inte-
grated, divisionalized, high-technology compa-
nies. The number of cases was determined by
cost, the desire to provide the potential for
replication and, most important of all in case
research, the ability to obtain access.
Not all the firms we approached agreed to
participate. We initially approached nine firms
which we believed would meet our criteria.
One firm was dropped from further considera-
tion when we discovered that it contracted out
most of its manufacturing. Of the eight compa-
nies remaining, one company wished us to
TABLE 1. Descriptive data on cases and embedded units (component divisions)
General
Primary end products
Sales (in billions)
Number of product
divisions
High-Tech Inc.
Instruments
$1-5
> 20
Continental Southern Defense and
World-Wide Inc. Communications Aerospace
Test, measurement and Telecommunications Defence and aerospace
computing systems systems
$10-15 $5-10 $10-15
> 20 > 10 > 20
Embedded uni ts (component di vi si ons)
Component transfer
Integrated circuits or Circuit Technology (CT) Custom Electronic Integrated Circuit (IC) Electronic Circuit (EC)
hybrids division and Packaging Circuit (CEC) division division division
St Interconnect (P&l)
division
circuit boards Circuit Board (CB) Printed Circuit Board Printed Circuit @CD)
division (PCB) division division
lo Many telephone calls to each company were necessary to reach the appropriate corporate level executive. The primary
difficulty we encountered was getting past secretaries and administrative assistants (gate-keepers) who protected their
bosses from what they considered to be unnecessary intrusions. With some companies we logged over 25 telephone calls
before we were able to reach the appropriate executive. In some cases we only started to make progess when we dis-
covered that many top level executives answer their own telephones early in the morning before their secretaries arrive.
432
G. J. COLBERT and B. H. SPICER
defer access until some unspecified time in the
future, and three others decided not to partici-
pate for various reasons.
The numbers of sites and interviews involved
in the course of the research are shown in
Table 2. Our investigation required consider-
able cooperation from participants and was
financially costly. It took us to 21 selling and
buying divisions spread over seven states dis-
persed widely across the U.S.A. Because of
scheduling difficulties this required 15 sepa-
rate interstate trips over an 18 month period.
Interviews were conducted with accounting
and functional managers in these divisions.
Forty-six managers were interviewed in the
course of the study with some individuals
being interviewed on more than one occasion
when we judged there was something more to
be learnt. Interviews ranged between ti and 3
hours in length. A tape recorder was used in
almost all of the interviews and transcripts
were prepared for later analysis.” In some
cases follow-up discussions took place by tele-
phone to clarify gaps in our understanding that
were discovered during the analysis phase. We
also toured manufacturing facilities.
Data collection took place in two phases.
Initial interviews took place with corporate
and/or group level accounting personnel who
were generally familiar with the management
of the transfer pricing process and ,transfer pri-
cing within each of the selected firms. In these
interviews we did two things. First, we gath-
ered general information about the firm’s orga-
nization structure, its performance
measurement system, and its sourcing and
transfer pricing policies. Second, and most
important, we identified particular component
transfers to study and identified the buying and
selling divisions involved. Those selected are
shown in the bottom half of Table 1.
The second phase of interviewing involved
visits to selling and buying divisions on either
side of the selected transfers. The theory of the
transfer pricing process discussed above deter-
mined the data to be collected. To improve the
reliability and comparability of the data col-
lected, a structured questionnaire was used to
guide our interviews and was filled out during
the interview by each interviewee. However,
during the interviews we queried informants
for detail and asked them to comment on mat-
ters they deemed to be important but which
were not adequately captured in our question-
naire. It also allowed us to ask informants to
compare and contrast the particular transfer
under study with others they were knowledge-
able about. At the divisional level we focused
on interviewing divisional controllers and divi-
sional managers.
The information gathered during these inter-
views included the descriptive data about the
subunit, dimensions or characteristics of the
transfer(s) under investigation, the sourcing
Firm
TABLE 2. Number of sites and interviews
Corporate Selling divisions Buying divisions
Number of Number of Number of
interviews Number of sites interviews Number of sites interviews
High-Tech
World-Wide
Continental
Communications
Southern Defense
and Aerospace
2 3 8 3 5
2 2 5 5 9
1 2 2 3 3
1 1 4 2 4
Total 6 8 19 13 21
I1 One lirm would not allow us to tape interviews for security reasons. In this instance we were restricted to note taking.
TRANSFER PRICING PROCESS
and transfer pricing process as it related to the
transfer(s) under study, and the performance
measurement system. The method we used to
obtain information on the key construct of
asset specificity associated with internal trans-
fers is discussed in the Appendix.
Because of the heavy cost of data collection,
the most extensive interviewing took place in
those companies with divisions which were
the least expensive for us to access. These com-
panies were High-Tech and World-Wide. We
started our study at High-Tech. We visited this
company many times over a 6 month period.
This helped us to refine our questions and
approach in order to obtain comparable infor-
mation from the other companies in the study
with divisions which were more widely scat-
tered, further away and more expensive to
visit. Consquently, we conducted fewer inter-
views and spent less time at Continental Com-
munications and Southern Defense and
Aerospace than we did at High-Tech and
World-Wide.
Anal ysi s and i nterpretati on
Analysis and interpretation of our data were
conducted in two stages. First, the data col-
lected for specific transfers within each of the
firms were analysed and interpreted within the
context of that firm. For each individual case
our analysis concentrates on how and why the
pattern of observations and information about
each case is consistent or inconsistent with our
theory. The mode of analysis is referred to by
Yin (1989, p. 109) as pattern matching. The
importance of context in case research
requires that the data be analysed as a pattern
relative to theory, in this instance our theory of
the transfer process. For example, where do
hypothesized relationships line up with the
theory and where do they not? What other
THEORY - INVESTIGATION
433
factors appear to bear on the theoretical rela-
tionship? It is important to note that it was our
theory which influenced not only the data to
be collected but also provided the template
against which the data were analysed. l2
Having analysed the data on individual trans
fers within the context of the firm involved, we
then focused on a cross-case analysis and com-
parison of results. Results from each case were
considered to be findings that were subject to
replication by the other individual cases. If the
findings obtained in subsequent cases also were
consistent with the theory, we concluded that
the viability of the theory was strengthened.
Thus, if the pattern observed in the second
case was similar to the pattern observed in the
first case, we would interpret this as a replica-
tion. If the patterns were not similar we tried to
understand why the patterns observed did not
match those predicted by the theory. Our data
collection efforts were designed to provide suf-
ficient additional information about the tirm,
the divisions and the selected transfers to allow
us to take this additional step.
DESCRIPTION OF CASES
In this section we flesh out the descriptions
of our four cases and embedded units identified
in Table 1. We focus on providing some rele-
vant background about each company and
information about the independent variables;
that is the degree of transaction-specific invest-
ment by each component (selling) and end
product (buying) division, and the strategic
role of each of the component units.
This information is summarized in Table 3
for each firm and component division we stu-
died. We also outline each lirm’s decentraliza-
tion policy as it relates to the divisions involved
” Our research method included Yin’s (1989) recommended methods for dealing with issues involving construct validity,
internal validity, external validity, and reliability in the design of our data collection efforts. Construct validity was
addressed through use of multiple sources of evidence, internal validity through use of a pattern matching approach,
external validity through the use of replication logic (using multiple cases), and reliability through the use of question-
naires and interview transcripts.
C
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
’
s
r
o
l
e
i
n
t
i
r
m
’
s
c
o
m
p
e
t
i
t
i
v
e
s
t
r
a
t
e
g
y
T
A
B
L
E
3
.
S
u
m
m
a
r
y
o
f
i
n
f
o
r
m
a
t
i
o
n
o
f
i
n
t
e
r
e
s
t
b
y
e
m
b
e
d
d
e
d
u
n
i
t
s
(
d
i
v
i
s
i
o
n
s
)
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
t
t
e
d
w
e
i
g
h
t
R
e
p
o
r
t
e
d
l
e
v
e
l
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
r
t
e
d
l
e
v
e
l
o
f
o
n
b
u
y
i
n
g
d
i
v
i
s
i
o
n
’
s
g
i
v
e
n
t
o
o
f
a
s
s
e
t
o
n
s
e
l
l
i
n
g
d
i
v
i
s
i
o
n
’
s
O
r
g
a
n
i
z
a
t
i
o
n
a
s
s
e
t
s
p
e
c
i
f
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
b
u
y
m
a
n
t
l
t
a
c
t
u
r
h
l
g
c
o
s
t
s
p
e
c
i
t
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
s
e
l
l
s
t
m
c
t
u
r
e
c
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
e
x
t
e
r
n
a
l
l
y
i
n
t
r
a
n
s
f
e
r
p
r
i
c
e
b
u
y
i
n
g
d
i
v
i
s
i
o
n
e
n
m
r
m
n
f
y
P
a
n
e
l
A
:
H
i
g
b
T
e
c
b
C
i
r
c
u
i
t
T
e
c
h
n
o
l
o
g
y
S
o
u
r
c
e
o
f
a
d
v
a
n
c
e
d
(
C
T
)
a
n
d
P
a
c
k
a
g
i
n
g
p
r
o
p
r
i
e
t
a
r
y
&
I
n
t
e
r
c
o
n
n
e
c
t
t
e
c
h
n
o
l
o
g
y
c
r
i
t
i
c
a
l
(
P
&
f
)
d
i
v
i
s
i
o
n
s
t
o
t
h
e
e
n
d
p
r
o
d
u
c
t
d
i
v
i
s
i
o
n
s
C
i
r
c
u
i
t
B
o
a
r
d
(
C
B
)
F
i
r
s
t
,
a
s
a
s
o
u
r
c
e
o
f
d
i
v
i
s
i
o
n
e
x
t
e
r
n
a
l
p
r
o
f
i
t
s
;
s
e
c
o
n
d
,
s
u
p
p
o
r
t
s
s
t
r
a
t
e
g
y
b
y
p
r
o
v
i
d
i
n
g
s
u
p
e
r
i
o
r
s
e
r
v
i
c
e
P
a
n
e
l
B
:
W
o
r
l
d
-
w
i
d
e
P
C
w
i
t
h
B
E
o
b
j
e
c
t
i
v
e
o
n
i
n
t
e
r
n
a
l
s
a
f
e
s
.
P
C
o
n
e
x
t
e
r
n
a
l
s
a
l
e
s
’
P
C
o
n
i
n
t
e
r
n
a
l
a
n
d
e
x
t
e
r
n
a
l
s
a
f
e
s
S
i
g
n
i
f
i
c
a
n
t
f
o
r
S
i
g
n
f
k
a
n
t
f
o
r
t
h
e
S
i
g
n
i
k
a
n
t
,
c
o
s
t
-
s
i
g
n
i
f
i
c
a
n
t
S
i
g
n
i
f
i
c
a
n
t
f
o
r
C
T
t
r
a
n
s
f
e
r
s
t
o
l
a
r
g
e
s
t
l
a
r
g
e
s
t
i
n
t
e
r
n
a
l
b
a
s
e
d
w
h
e
n
i
t
i
n
v
o
l
v
e
s
i
n
t
e
r
n
a
l
c
u
s
t
o
m
e
r
,
c
u
s
t
o
m
e
r
,
m
i
n
o
r
n
e
w
t
e
c
h
n
o
l
o
g
y
,
m
i
n
o
r
o
t
h
e
r
w
i
s
e
o
t
h
e
r
w
i
s
e
s
i
g
n
i
f
i
c
a
n
t
f
o
r
P
&
f
t
o
~
a
s
s
u
r
e
s
e
r
v
i
c
e
;
M
i
n
o
r
N
o
n
e
M
i
n
o
r
,
m
a
r
k
e
t
-
b
a
s
e
d
M
i
n
o
r
N
o
n
e
n
9
C
u
s
t
o
m
E
l
e
c
t
r
o
n
i
c
S
e
c
u
r
e
s
o
u
r
c
e
o
f
P
C
w
i
t
h
B
E
S
i
g
n
i
f
i
c
a
n
t
f
o
r
i
t
s
S
i
g
n
i
t
k
a
n
t
w
h
e
n
S
i
g
n
i
k
a
n
t
,
c
o
s
t
-
C
i
m
u
i
t
(
C
E
C
)
i
m
p
o
r
t
a
n
t
o
b
j
e
c
t
i
v
e
o
n
i
n
t
e
r
n
a
l
l
a
r
g
e
s
t
t
r
a
n
s
f
e
r
s
c
a
p
a
c
i
t
y
u
t
i
l
i
z
a
t
i
o
n
b
a
s
e
d
d
i
v
i
s
i
o
n
t
e
c
h
n
o
f
o
g
y
t
h
a
t
s
a
l
e
s
,
P
C
o
n
e
x
t
e
r
n
a
l
i
s
a
n
i
s
s
u
e
p
r
o
v
i
d
e
s
s
a
f
e
s
d
i
f
f
e
r
e
n
t
i
a
t
i
o
n
t
o
e
n
d
p
r
o
d
u
c
t
s
M
i
n
o
r
P
r
i
n
t
e
d
C
i
r
c
u
i
t
B
o
a
r
d
(
P
C
B
)
d
i
v
i
s
i
o
n
S
u
p
p
o
r
t
s
s
t
r
a
t
e
g
y
b
y
p
r
o
v
i
d
i
n
g
s
u
p
e
r
i
o
r
s
e
r
v
i
c
e
P
C
w
i
t
h
B
E
o
b
j
e
c
t
i
v
e
M
o
d
e
r
a
t
e
f
o
r
i
t
s
l
a
r
g
e
s
t
t
r
a
n
s
f
e
r
s
,
m
i
n
o
r
o
t
h
e
r
w
i
s
e
M
i
n
o
r
S
i
g
n
i
l
i
c
a
n
t
,
c
o
s
t
-
b
a
s
e
d
M
i
n
o
r
v
e
r
y
s
i
g
n
i
t
i
c
a
n
t
T
A
B
L
E
3
.
C
o
n
t
.
C
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
’
s
r
o
l
e
i
n
h
r
m
’
s
c
o
m
p
e
t
i
t
i
v
e
s
t
r
a
t
e
g
y
O
r
g
a
n
i
z
a
t
i
o
n
s
t
r
u
c
t
u
r
e
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
r
t
e
d
w
e
i
g
h
t
R
e
p
o
r
t
e
d
l
e
v
e
l
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
t
t
e
d
l
e
v
e
l
o
f
o
n
b
u
y
i
r
q
g
d
i
v
i
s
i
o
n
’
s
g
i
v
e
n
t
o
o
f
a
s
s
e
t
o
n
s
c
l
l
i
a
g
d
i
v
i
s
i
o
n
’
s
4
a
s
s
e
t
s
p
e
c
i
f
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
b
u
y
m
a
n
t
l
k
N
r
i
n
g
c
o
s
t
s
p
e
c
i
f
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
s
e
l
l
c
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
e
x
t
e
m
a
R
y
i
n
t
r
a
n
s
f
e
r
p
r
i
c
e
b
u
y
i
n
g
d
i
v
i
s
i
o
n
e
x
t
e
t
n
a
R
Y
5
5
P
a
n
e
l
C
:
C
O
t
l
t
i
?
W
?
l
k
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
I
n
t
e
g
r
a
t
e
d
C
i
r
c
u
i
t
S
e
c
u
r
e
s
o
u
r
c
e
o
f
P
C
e
x
p
e
c
t
e
d
t
o
B
E
s
i
@
l
i
t
i
c
a
n
t
S
i
g
n
i
t
i
c
a
n
t
o
n
h
i
&
S
&
n
i
f
i
c
a
n
t
,
b
u
t
M
i
n
o
r
V
e
r
y
s
i
g
n
i
f
i
c
a
n
t
g
;
8
M
i
n
o
r
V
e
r
y
s
&
n
i
B
c
a
n
t
!
!
2
8
2
I
(
I
C
)
d
i
v
i
s
i
o
n
P
r
o
P
d
e
t
a
r
y
o
r
m
a
k
e
a
m
o
d
e
s
t
v
o
l
u
m
e
t
r
a
n
s
f
e
r
s
t
e
c
h
n
o
l
o
g
y
p
r
o
f
i
t
p
r
i
n
t
e
d
C
i
r
c
u
i
t
S
u
p
p
o
r
t
s
s
t
r
a
t
e
g
y
b
y
P
C
e
x
p
e
c
t
e
d
t
o
B
E
M
i
n
o
r
M
i
n
o
r
@
C
D
)
d
i
v
i
s
i
o
n
p
r
o
v
i
d
i
n
g
s
u
p
e
r
i
o
r
o
r
m
a
k
e
a
m
o
d
e
s
t
s
e
r
v
i
c
e
p
r
o
f
i
t
P
a
n
e
l
D
:
S
o
u
t
b
e
m
D
e
f
e
n
c
e
a
t
k
i
A
-
p
a
c
e
E
k
c
t
r
o
n
i
c
c
i
r
c
u
i
t
s
S
e
c
u
r
e
s
o
u
r
c
e
o
f
P
C
w
i
t
h
a
B
E
S
i
g
n
i
f
i
c
a
n
t
f
o
r
S
i
g
n
i
f
i
c
a
n
t
f
o
r
s
i
g
r
i
t
i
c
a
n
t
w
e
i
g
h
t
i
s
a
l
s
o
g
i
v
e
n
t
o
m
a
r
k
e
t
M
i
n
o
r
,
m
a
r
k
e
t
-
b
a
s
e
d
s
i
g
n
i
6
c
a
n
t
,
c
o
s
t
-
@
C
)
d
i
v
i
s
i
o
n
2
d
V
Z
U
l
C
C
d
o
b
j
e
c
t
i
v
e
o
n
i
n
t
e
r
n
a
l
h
y
b
r
i
d
s
h
y
b
r
i
d
s
,
m
o
d
e
r
a
t
e
b
a
s
e
d
t
c
c
b
n
o
l
o
g
y
n
o
t
s
a
l
e
s
,
P
C
o
n
e
x
t
e
r
n
a
l
f
o
r
i
n
t
e
g
r
a
t
e
d
r
e
a
d
i
l
y
a
v
a
i
l
a
b
l
e
s
a
l
e
s
c
i
r
c
u
i
t
s
e
-
-
W
l
P
C
m
e
a
n
s
p
r
o
f
i
t
c
e
n
t
r
e
a
n
d
B
E
m
e
a
n
s
b
r
e
a
k
e
v
e
n
.
S
i
@
k
a
n
t
f
o
r
M
o
d
e
r
a
t
e
3
t
h
o
s
e
w
i
t
h
h
i
g
h
9
t
e
c
h
n
o
l
o
g
y
i
j
n
e
e
d
s
5
i
j
2
436 G. J. COLBERT and B. H. SPICER
under the heading organization structure. Our
findings are then discussed in the following
section.
High-Tech I nc.
High-Tech Inc. is a large, divisionalized com-
pany that makes a wide variety of sophisticated
electronic products including display and mea-
surement devices. The company has over 20
divisions that are organized into a small num-
ber of product groupings. Component divisions
are located in the different end product groups.
tical integration strategy. First, commodity type
components (e.g. plastics and metal fabrica-
tion) which provide its end products little or
no differentiation advantage became readily
available from external suppliers at lower
cost. Second, expected growth for end pro
ducts failed to materialize, creating serious
excess capacity problems in its important elec-
tronic component operations.
Since it started up in the 194Os, the compa-
ny’s corporate strategy has been to be a differ-
entiated producer of high-technology, high-
performance, state-of-the-art electronic pro-
ducts; a producer that sets the standards for
the rest of the industry. One divisional control-
ler commented:
Our corporate strategy is to produce products that offer
cost-effective proprietary performance. We have to be
cost-effective, but products have to have a proprietary
performance because we are a technology driven com-
pany-we are not a low cost manufacturer. It is proprie-
tary technology that we need.
These problems forced High-Tech to change.
At the time we conducted our study High-Tech
was attempting to build a stronger market and
customer focus and to improve the company’s
rate of return. As part of this change process top
management divested some facilities producing
commodity type components, reorganized and
made related changes to sourcing and transfer
pricing policies. Some of these changes are dis-
cussed later in the paper. Current corporate
policy is to allow divisions to negotiate over
internal transfers and related transfer prices.
In general, division and group managers are
expected to work things out without calling in
top management to arbitrate disputes.
Historically, because of the technology driven
nature and culture of the company (sometimes
described as a company of engineers produ-
cing products for engineers) High-Tech had
practised backwards vertical integration to
assure a continuous flow of components to its
end product divisions. Large scale component
fabrication and assembly facilities were built to
produce a diverse set of proprietary and non-
proprietary components to end product
divisions.
High-Tech has three primary component
divisions which vary in their strategic roles in
the company and the degree of asset specificity
and proprietary content of internal transfers.
Panel A of Table 3 summarizes information
about these roles and the reported levels of
asset specificity of these component divisions
and their customers. The following sections
describe their operations.
More recently, High-Tech had experienced
financial performance difficulties owing pri-
marily to increasing product competition and
unsuccessful forays into new product areas.
Two other factors disrupted the company’s ver-
Circuit Technology (CD and Packaging C
I nterconnect (PM) divisions. The CT division
produces a wide range of customized inte-
grated circuits while the P&I division produces
many custom circuit modules called hybrids.i3
Because they share many common attributes
these two divisions are discussed together.
There is a consensus among the managers
we interviewed that these two divisions are
._
” Hybrids have been described as “circuits in a can”. They are fully tested circuit modules which provide a way of dealing
with problems of limited space, complex interconnects and heat build-up in electronic circuits. A single hybrid can
functionally replace a fully loaded circuit board in a compact module form. P&I hybrids incorporate custom integrated
circuits supplied by the CT division.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
437
critical to the company’s competitive strategy
because they supply the proprietary technol-
ogy that imparts “cutting edge” performance
and features for the company’s end products.
A significant portion of each division’s sales are
internal although both have acted recently to
build an external customer base as well. How-
ever, their external business is seen primarily
as a way to increase capacity utilization rather
than as a separate source of profits per se. Both
divisions report moderate amounts of unused
capacity. Both divisions are evaluated as profit
centres with a break-even objective on internal
transfers and a positive profit objective for
external sales.
Both divisions report specialized investments
of two types. The first are investments specia-
lized to the production of a type of component,
i.e. integrated circuits or hybrids. These consist
primarily of equipment and training of person-
nel. The fabrication processes employ certain
machines that are very expensive and unique.
Although these machines were purchased with
the primary intent of supporting internal
demand, they can be used to produce a range
of different components. Both CT’s and P&I’s
managers report that there would be moderate
difficulty in redeploying these assets to serve
new external or internal business. Because of
relatively long lead times, curtailment of exist-
ing production owing to lost internal business
would result in lost revenue while capacity is
backfilled with new business.
The second type of specialized investment is
in those assets which are specialized to the
production of particular customized compo-
nents. These investments consist primarily of
engineering design work and software used to
run automated equipment to fabricate and test
each particular component. For example, the
P&I division’s controller comments:
There is no portability with respect to a lot of invest-
ment that we have made here. It is specific to a custom
application. In the custom hybrid business a significant
amount of your investment is your engineering assets
and your process development assets to get a custom
product going. If we were to replace this internal busi-
ness we would have to reincur all that cost over again.
Internal transfers to one end product division
account for over one-half of the total volume of
both the CT and P&I divisions. For each com-
ponent division, a significant level of asset spe-
cificity was reported to be present for these
internal transfers. Both of these component
divisions report that if transfers to this one
end product division were halted there would
be at least a moderate decline in the value of
the investments in tangible assets and person-
nel specialized to the production of compo-
nent types, i.e. integrated circuits and
hybrids, and a significant decline in the value
of investments in design work and training
which are specialized to unique components.
Lower volume transfers involve lower levels of
specialized investments and a minor decline in
investment values if these transfers were
halted.
End product divisions, which are internal
customers of these two buying divisions, also
report a significant level of investment specia-
lized to the components sources from these
two divisions. These consist primarily of invest-
ments in design work and training of person-
nel. These internal customers report that there
would be a significant decline in the value of
these investments if the internal transfers were
halted and they were forced to switch to exter-
nal supply. They attribute this to the significant
difficulty in buying these advanced compo-
nents externally, particularly when the indivi-
dual volumes of specific components are
relatively low. There was also some doubt
expressed about whether they would receive
the same level of service, for example respon-
sive engineering support, from an external ven-
dor.
Ci rcui t Board (CB) di vi si on. Although the
CB division manufactures a diverse range of
custom circuit boards, it specializes in high-
end, multi-layer circuit boards. The division is
noted for delivering these types of boards
quickly and efficiently even in relatively low
volumes. This capability is valuable to external
and internal customers who have low volumes
and fast time-to-market requirements. Because
circuit boards are regarded as something of a
438 G. J. COLBERT and B. H. SPICER
commodity available from external sources,
service in the form of quick turn around, con-
sistent high quallty, reliable delivery times, etc.,
rather than proprietary technology is regarded
as the key to providing competitive advantage
to customers. The division’s controller com-
ments on the service features they offer:
All of those things [quick turn around, electronic design
transfer, close working relationships to the customer,
etc.] make it easy to do business with us as opposed
to someone else. Circuit boards are on the critical path
of time-tomarket. If you screw up and miss getting the
right board in the prototype phase you could comple-
tely miss the market window.
In marked contrast to the CT and P&I com-
ponent divisions, the CB division has a success-
ful and extensive external business and is
viewed as a direct source of profits to the com-
pany somewhat like High-Tech’s end product
divisions. Interestingly, the division views itself
as a “stand alone” business and is evaluated as
a full profit centre on both internal and exter-
nal sales. At the time the study was conducted
the division reported it was operating with
moderately constrained capacity.
The CB division reports a very significant
level of investment specialized to the fabrica-
tion of circuit boards in general, but only a
minor level of investment specialized to the
production of particular circuit boards
whether for internal or external sale. CB
reports that there would only be minor losses
involved if individual internal transfers were
halted. The reason given is that assets used to
fabricate circuit boards can readily be rede-
ployed to produce other boards. Furthermore,
they have a backlog of orders which they could
use to backlill any volume lost. The biggest
issue is the time needed to bring in a new
customer and develop tooling for that custo-
mer’s boards. Tooling in this case is analogous
to a blueprint or map that indicates how to
build a particular board and is unique to that
board.
End product divisions that are internal custo-
mers of CB report moderate to significant levels
of investment specialized to the components
sourced from CB. However, even those with
a need for high-end, high-technology boards
report that there would be only a minor
decline ln the value of investments if the lnter-
nal transfer were halted because even these
boards could be obtained from external
sources, albeit with some difficulty. The com-
ments of a controller of an end product division
that is a major customer of the CB division
reflects this point:
I could find another vendor pretty easily, and I could get
my design tapes and film work from CB. I might have to
put a little bit of effort into reformatting the design
tapes for a different process. My estimate is this would
only involve a minor decline in investment value though
I would have to scramble around a bit.
World- Wtde I nc.
World-Wide is a major manufacturer of pre-
cision electronic equipment used for measure-
ment, analysis and computation. The company
makes thousands of products that cover a
broad range. Operations are concentrated in
the U.S.A., but the company has other research
and manufacturing facilities throughout the
world.
World-Wide’s competitive strategy has a
strong marketing emphasis which is based on
understanding customer needs and then devel-
oping solutions through technological innova-
tion. At the time we conducted our study the
company had been experiencing strong
growth which was attributed to both its inno-
vative capabilities and the success of its end
product divisions in introducing new products
and new generations of existing products to
the market-place. The company’s products
and businesses are very diverse. Some busi-
nesses operate in large, very competitive mar-
kets where they compete against world-class
competitors. Other businesses operate in nar-
row markets where volumes are relatively low
and technology and performance demands are
such that customers are not as price sensitive.
Unsurprisingly, it is the former businesses
which appear to be most sensitive to transfer
pricing issues.
The company is organized into groups on the
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
439
basis of major business segments. The groups
are further decentralized Into divisions. Com-
ponent operations are primarily located in a
separate group. Historically, management’s
strategy has been to utilize its large internal
volume requirements to drive down the cost
of component supply by fully utilizing the capa-
city of its Internal component divisions. The
internal component divisions are expected to
meet the needs of a diverse set of internal cus-
tomers and the company has typically operated
its component operations as captive suppliers.
However, considerable negotiation over trans-
fers and transfer prices takes place between
component and end product divisions but dis-
putes are rarely elevated to the group or corpo-
rate level.
Business units within the groups that sell end
products are operated as profit centres (divi-
sions) that operate very autonomously. The
corporate policy and culture has been to
encourage responsibility and innovation, and
end product divisions, in particular, have
been encouraged to “run their own busi-
nesses” to the greatest extent possible. This
has caused some difficulties when divisions
must cooperate and share resources to pro-
vide Integrated solutions to customer require-
ments. Over time, World-Wide has
experimented with changes in responsibilities
and reporting relationships to Improve co-ordi-
nation.
With World-Wide we focused on two compo-
nent manufacturing units that internally trans-
fer the same or similar types of components to
those studied in the High-Tech case. Like High-
Tech these units vary in terms of their strategic
roles in the company and the characteristics or
dimensions of component transfers to product
divisions. These divisions are located within
the same group and represent the primary
source of components transferred internally
within World-Wide.
Panel B of Table 3 summarizes information
about the roles these two component divisions
play in World-Wide’s competitive strategy,
their reported levels of asset specificity, and
that of their internal customers. The following
sections describe their operations in more
detail.
Custom Electronic Circuit (CEC) division.
World-Wide’s CEC division is similar to High-
Tech’s Circuit Technology (Cl) division. CEC
manufactures custom integrated circuits. The
division is regarded as strategically important
to many of World-Wide’s end product divisions
because it serves as a secure source of custo-
mized advanced circuits that provide differen-
tiation features for the company’s end
products. The division’s origins are rooted in
the belief that at one time the only way to
have access to custom Integrated circuits was
to manufacture them internally. Internal custo-
mers now cite the division’s design capability,
availability of particular manufacturing pro-
cesses, quality, responsiveness, and security
of unique proprietary designs and technology
as advantages of doing business with CEC over
external vendors.
Although it is not intended to be a low cost
producer of components (internal customers
rate CEC as having no price advantage), to
remain economically viable CEC has had to
obtain higher volumes by producing some cir-
cuitry that is less unique and not as advanced
technologically. Like High-Tech’s CT division,
CEC has increasingly pursued external sales to
fill capacity and meet performance objectives.
CEC is evaluated as a profit centre with a break-
even objective on internal sales and a positive
profit objective on external sales. At the time
our study was conducted, CEC reported mod-
erately constrained capacity.
CEC’s managers report a very sign&ant
investment in the facility and its workforce
devoted exclusively to the production of cus-
tom integrated circuits. Although in principle
this Investment could be redeployed, CEC’s
managers report that there would be at least
a moderate decline in investment value if its
largest internal transfers were halted owing to
difficulties in quickly backfilling capacity with
external business. Most importantly, CEC’s
managers report significant investments specia-
lized to particular internal transfers. For exam-
ple, in the case of its largest internal customer,
440 G. J. COLBERT and B. H. SPICER
CEC has invested in an expensive piece of
equipment that is used only to assemble pro
ducts for that customer. In addition there are
significant investments in design work and spe-
cialized training to produce a particular compo-
nent. It is reported that there would be a
significant decline in the value of these invest-
ments if large internal transfers were halted.
Internal customers report moderate levels of
investment consisting of design work and train-
ing and minor levels of tangible assets that are
specialized to components transferred from
CEC. However, they also report that only a
minor decline in the value of these invest-
ments would occur if the internal transfer
were halted. Any decline would be due to the
cost of developing new relationships,
exchange of information about designs and fab-
rication processes, and perhaps some redesign
effort. For example, the controller of the lar-
gest customer division comments:
We had to automate a lot of things to use that [internally
transferred] integrated circuit. The dollar value is high
but the investment is fairly transferable to outside ven-
dors.
Pri nted Ci rcui t Board (PCB) di vi si on. Simi-
lar to the Circuit Board (CB) division at High-
Tech, this division also manufactures circuit
boards which are not regarded as strategically
important. However, in contrast to the High-
Tech CB division, which is operated very
much as a stand-alone business, World-Wide’s
PCB division is operated exclusively as a cap
tive supplier and is several times larger invol-
ving multiple plants. It is evaluated as a profit
centre with a break-even objective. At the time
of the study it was operating at or near capa-
city.
PCB’s strategic role is to support end product
division’s strategy by providing superior ser-
vice including flexible delivery and quick
design turnaround. Internal customers cite
design capability, quality, responsiveness,
delivery and cost as PCB’s advantages over
external suppliers. World-Wide’s product divi-
sions have very demanding needs for design
and redesign of boards since some products
have life cycles as short as a year and a half.
Attempting to stay on the front edge of the
market with each generation of product, end
product divisions make frequent changes to
board layout, etc. Thus the ability to co-ordi-
nate designs and provide quick turnaround
are crucial to obtaining “time-to-market”
advantages.
PCB reports very significant levels of invest-
ment in tangible assets and training specialized
to the production of circuit boards. The decline
in the value of these investments that would
occur if individual internal transfers were
halted is reported to vary from minor to mod-
erate depending on the size of the transfer.
Potentially, specialized assets resulting from
these investments could be switched to produ-
cing boards for external customers. However,
PCB has no external marketing capability and
to establish this would involve substantial
costs. PCB also reports a minor level of invest-
ment in design work related to specific circuit
boards as this is borne primarily by the buying
unit. Hence, there would be only a minor
decline in value if an individual transfer were
halted.
Internal customers report a significant level
of investment in design work that is specialized
to internal transfers of particular circuit boards
from PCB. They report that the decline in
investment value would be minor should the
internal transfer be halted, as the design work
could be shifted easily to an external vendor.
As one manager pointed out “the technology
isn’t that terribly unique that we couldn’t fmd
alternative sources.”
Conti nental Communi cati ons
Continental Communications is a leading
manufacturer of digital telecommunications
systems. The company’s products are focused
on a small number of applications in the tele-
communication’s market. In each of these pro-
duct lines Continental is one of the few major
suppliers in the North American market where
its operations are concentrated.
With the break up of AT&T and the dereg-
TRANSFER PRICING PROCESS
ulation of the telecommunication’s industry,
the company underwent a period of strong
growth but more recently has come under
increasing pressure from competitors vying
for market share and from customers seeking
increasingly sophisticated, higher quality pro-
ducts. At the time we conducted our study
Continental’s competitive strategy centred on
increasing customer satisfaction at the same
time as it reduced costs and increased through-
put in its manufacturing facilities.
The company is divisionalized based on its
major product lines. Component fabrication
and research and development are organized
as separate business units apart from end pro
duct operations. Historically, the company
required its end product divisions to source
components internally but more recently has
given greater latitude to end product divisions
to source some components externally. The
company has streamlined its operations
through closure and consolidation of some of
its manufacturing facilities as it shifted towards
external sourcing of some components such as
low-technology circuit boards. Virtually all
components except custom integrated circuits
and circuit boards and closely related compo-
nents are now sourced from outside vendors.
Owing to competitive pressures, end product
divisions have placed considerable pressure on
component divisions to reduce their transfer
prices.
In our study we looked at transfers from the
company’s two primary component divisions
that supply integrated circuits and circuit
boards. While internal sourcing is in principle
mandated and there is a corporate transfer pri-
cing policy of standard manufacturing cost plus
a set percentage mark-up, the transfers that we
studied are exempted from this policy as a
result of attempts by end product divisions to
reduce costs by forcing internal component
divisions to price against potential external
.‘
THEORY - INVESTIGATION 441
vendors whenever possible. Negotiation and
disagreements over transfer prices occur but
we were told that most disputes are resolved
at the divisional level.
Panel C of Table 3 summarizes information
about the strategic roles these divisions play in
Continental’s competitive strategy and the
reported levels of asset specificity of these
component divisions and their customers.
The following sections describe their opera-
tions.
Integrated Ci rcui t (rC) di vi si on. The I C
division, in contrast to other integrated circuit
plants in this study, specializes in producing
high-volume, mature, custom integrated cir-
cuits.‘* The division is evaluated as a profit
centre with a modest profit objective or at
minimum a breakeven objective.
Internal customers cite proprietary designs,
assured supply and responsiveness as K’s pri-
mary competitive advantages over external
suppliers. Neither IC nor its internal customers
regard price as one of the IC division’s compe-
titive advantages.
The IC division reports a very significant
level of investments in tangible assets, design
work and training specialized to the produc-
tion of integrated circuits. IC produces a lim-
ited number of high-volume components
which are transferred to more than one end
product division. Loss of an individual transfer
would impact on the division through a decline
in capacity utilization. Hence, IC’s managers
report that there would be a significant decline
in the value of these investments if an indivi-
dual transfer were halted. IC’s controller
explains:
This is a highly leveraged business. Because of all the
expensive specialized equipment we have a fairly high
breakeven point. If you pull volume below our break
even point then we are left with a fixed chunk of cost
without volume to cover it. That is the risk you run in
letting buyers redeploy elsewhere.
‘* The company actually has two integrated circuit facilities. The facility we studied focuses on the production of mature
components ln high volumes. The other facility focuses on prototype and advanced technology custom circuits. Transfers
from this division, which is located In the same facility as the company’s research operations, were not studied.
442 G. J. COLBERT and B. H. SPICER
Internal customers report only minor
amounts of investments specialized to internal
transfers from the IC division and that the
potential decline in investment value would
be minor if internal transfers were halted.
They believe that although there would be
some short- to medium-term difficulties their
component requirements could be met by
switching to external sources.
products and services that often involve long-
term programmes or projects.
Southern’s strategy is to be considered as a
technology leader and it is regarded as a pace-
setter in applying advanced electronics in the
defence and aerospace industry. The compa-
ny’s growth and reputation resulted from the
post World War II boom in high-technology
applications in defence and aerospace.
Pri nted Ci rcui t di vi si on (PCD). A large por-
tion of Continental’s board needs are sourced
externally. The boards produced by the PCD
division differ from those sourced externally
as they typically involve more advanced tech-
nology, are more difficult to make and are
lower ln volume.
However, PCD’s managers recognize that
their only real advantage over external suppli-
ers is not cost or quality but rather superior
service and their ability to work closely with
the end product divisions and the company’s
R&D unit to bring new boards or revisions of
old boards into production. Internal customers
concur with this assessment. PCD is evaluated
as a profit centre with a modest profit objective
or at a minimum a break-even objective.
The company is organized into a number of
product groups each of which have several divi-
sions that involve similar technology and appli-
cations. The company is operated on a highly
decentralized basis which allows each group to
respond to its segment of the electronics mar-
ket with a total capability from research to sales.
One exception to this is intracompany activity
that relates to some centralized component
technology and fabrication. Most of Southern’s
internal transfers involve transfers from the
Electronic Circuit division and it is on these
transfers that we focus.
The PCD division reports significant invest-
ments in assets specialized to circuit boards
generally but only minor investments that are
specialized to particular internal transfers. They
further report only a minor potential decline in
investment value if individual internal transfers
were halted.
Internal customers report that although they
have significant investments in design work
related to internally sourced circuit boards,
this investment could be redeployed to outside
vendors; hence, they report that there would
only be a minor decline in investment value if
internal transfers were halted. This is particu-
larly the case for the least complex boards that
are required in large volumes.
Much of the company’s external business is
conducted through prime contracts (as
opposed to subcontracts) with government enti-
ties. These contracts may span several years and
involve hundreds of millions of dollars. Both
cost-plus and fixed-cost contracts are used
with the latter being most common and increas
ing in frequency. Because of its high exposure to
declines in defence budgets, management is
attempting to decrease the company’s depen-
dence on defence-related business. Achieving
this requires changing a rather bureaucratic
organizational infrastructure that has been
long established to service defence contracts.
At the time of our study the company was begln-
ning to explore how to make these changes.
Southern Defense and Aerospace
Panel D of Table 3 summarizes information
about the strategic role the Electronic Circuit
division plays in Southern’s competitive strat-
egy and the reported levels of asset specificity
of this component division and its customers.
The following sections describe its operations.
Southern Defense and Aerospace is a major El ectroni c Ci rcui t (EC) di vfsi on. The major-
manufacturer of high-technology electronic ity of Southern’s internal transfers of compo-
systems for military, commercial and scientific nents comes from the EC division. This
use. The company produces several thousand division produces custom integrated circuits
TRANSFER PRICING PROCESS
and hybrid packaging of these circuits. The
division’s primary mission is to provide end
product divisions with a secure source of
advanced technology components that are
not readily available from external sources.
This component capability, and in particular
their hybrid technology, l5 is crucial to the com-
pany given its markets and strategy to be a
technology leader. In particular it is seen as
giving them control over the interface
between R&D and production in advanced
applications. Given that defence contracts
often include penalties for late delivery, inter-
nal customers also require an assured and reli-
able source of supply and tight security with
respect to designs.
At the time of the study the division had
significant excess integrated circuit capacity
relative to internal volume requirements. In
an attempt to offset the cost of this capacity
the division has aggressively pursued external
sales of Integrated circuits. Hybrid capacity
could be increased with some additional invest-
ment but is presently profitable with only a
minor amount of external sales. EC is evaluated
as a profit centre with a breakeven objective
on internal sales and a positive profit objective
on external sales.
The EC division reports a very sign&ant
level of investment specialized to the produc-
tion of the two component types, integrated
circuits and hybrids. For example, its fabrica-
tion areas require the highest level of “clean”
rooms in the industry and are extremely expen-
sive. The division also reports a significant level
of investment specialized to individual trans
fers. To illustrate, EC recently purchased an
expensive piece of test equipment specifically
to support one product division’s programme.
The division reports that there would be a sig-
nificant decline in both the value of invest-
._
THEORY - INVESTIGATION
443
ments in capacity and investments specialized
to individual transfers of integrated circuits and
hybrids if internal transfers were halted.
Internal customers with advanced technol-
ogy needs report a signiticant level of invest-
ments specialized to internal transfers from
EC. These would involve a significant decline
in investment values if internal transfers were
halted. For example, an end product division
indicates that it is tied by technology and
design decisions to buy certain components
from the EC division. The end product division
has designed these components for fabrication
on uncommon processes owned and operated
by the EC division.
Cfrcuft boar&. Unlike the other companies
in our study, Southern sources all of its circuit
boards externally. However, the company has
recently reconsidered this issue and has
decided to invest in an internal circuit board
facility. In addition to internal sourcing the
company plans to continue sourcing some
boards externally with a few highquality sup
pliers with whom it will develop a long-term
relationship. l6 The anticipated benefits of an
internal board supplier cited by Southern’s
management include lower cost through
economies of scale in production and adminis
nation of the supplier interface, improved qual-
ity and quicker turnaround.
CASE EVIDENCE AND FINDINGS
We initially analysed and wrote up our
results for each case before undertaking a
cross-case analysis. However, this makes for a
long and tedious presentation. Therefore, we
present our evidence and findings here by
directly assessing how well our case evidence
matches the pattern suggested by theory and
” WC were told that because of aerospace requirements EC’s hybrid capability was matched by very few suppliers ln the
world. Packaging layers for current aerospace applications are several times denser than most commercial applications.
Only a small fraction of Southern’s total hybrid requirements are readily available from external vendors.
i6 Southern’s investigation of circuit board procurement activity revealed, to management’s surprise, that product divi-
sions were sourcing boards from several hundred dlfTerent suppliers.
444 G. J. COLBERT and B. H. SPICER
the extent to which the findings are replicated
over the cases. We draw attention to unusual or
interesting issues.
Panels A to D of Table 3 present the indepen-
dent and dependent variables of interest by
embedded unit (division) for High-Tech,
World-Wide, Continental Communications,
and Southern Defense and Aerospace.
The first question we address is:
How do the business units of high-tech-
nology Rrms manage the transfer process
and why do these management practices
differ?
Case evidence pertaining to this question is
set out by case in Table 3 and a cross-case
summary of our findings is set out in Table 4.
Managing sourcing decisions when asset
specijkity is significant in the component
(selling) divisions
When component (selling) divisions have a
sign@cant level of asset specificity related to
TABLE 4. Crosscase data on sourcine and sellina decisions
Panel A: Sourcing decisions
Assets specificity reported by component (selling) divisions
Signifcunt
High-Tech
CT division
P&I division
World-Wide
CEC division
Continental Communications
IC division
Southern Defense and Aerospace
EC division
Mi nor
High-Tech
CB division
Continental Communications
PCD division
Moderate
World-Wide
PCB division
Panel B: Selling decisions
Asset specificity reported by end product @wing) division
Buying unit constrained?
Yes
Yes
Yes
Yes
Yes
No
No
No
SeUing unit constrained?
Sfgni J i cant
High-Tech
CT division
P&I division
Southern Defense and Aerospace
EC division
Mi nor
High-Tech
CB division
World-Wide
CEC division
PCB division
Continental Communications
IC division
PCD division
Yes
Yes
Yes
No
No
Yes
Yes
Yes
TRANSFER PRICING PROCESS
i nternal transfers, we expect to fi nd end pro-
duct @uyi ng) di vi si ons constrai ned i n thei r
abi l i ty to buy the component external l y i n
order to protect the transacti on-speci j k i nvest-
ments made by component di vi si ons.
From Panel A of Table 3 we see that High-
Tech’s Circuit Technology (CT) and Packaging
& Interconnect (P&I) divisions both report a
significant level of asset specificity related to
internal transfers to their largest internal custo-
mer. Consistent with the theory this end pro-
duct division reports that there are significant
constraints on its ability to source components
from an external supplier which are presently
sourced from CT and P&I. This sourcing policy
protects the company’s investments in these
strategically important areas which provide a
secure, assured source of advanced proprie-
tary technology to end product divisions. It is
also consistent with the company-wide strategy
as a differentiated supplier of high technology
state-of-the-art electronic products. ”
From Table 3 and Panel A of Table 4, we see
that the finding in the High-Tech case is
repeated for the integrated circuit facilities in
each of the other three cases, i.e. for World-
Wide’s Custom Electronic Circuit Division
(CEC) division, Continental’s Integrated Circuit
(IC) division and Southern’s Electronic Circuit
(EC) division. In each case, end product divi-
sions report a signiIicant constraint placed on
sourcing particular components externally
albeit with some varying conditions. At World-
Wide the constraint operates primarily when
capacity utilization in the component division
is an issue. Consistent with the company’s man-
agement style this constraint is not formally
imposed but rather the division is expected to
cooperate if asked at the group level.
At Continental the constraint is applied for
high-volume custom circuits in order to main-
tain high levels of utilization of an expensive IC
facility which is dedicated to fabricating
mature, high-volume circuits. Company policy
THEORY - INVESTIGATION 445
is that end product divisions will source intem-
ally unless they have an exemption. At South-
em, where there is a very large specialized
investment in the hybrid unit of the EC divi-
sion, end product (buying) divisions are con-
strained from buying outside without the
approval of the hybrid unit. An important rea-
son for this is to ensure that capacity is utilized.
Another is to centralize supplier certification.
We also observe that there are strategic moti-
vations for each tirm to invest in particular
component technology and capacity. We dis-
cuss this further in a later section.
Managi ng sourci ng deci si ons when asset
speci j ki ty i s mi nor i n the component
(sel l i ng) di vi si ons
Wben component (sel l i n@di vi si ons have a
mi nor l evel of asset speci fi ci ty rel ated to i nter-
nal transfer, we woul d not expect to fi nd end
product (buyi ng) di vi si ons constrai ned i n
thei r abi l i ty to buy the component external l y
i n order to protect mi nor transacti on-speci fi c
i nvestments made by component di vi si ons.
From Panel A of Table 3 we see that for
internal transfers from High-Tech’s CT and
P&I divisions which report minor asset specitl-
city, internal customers report only minor con-
straints on their ability to source components
externally. Similarly, we see that High-Tech’s
Circuit Board (CB) division reports only a
minor level of asset specificity related to inter-
nal transfers. Consistent with the theory, end
product divisions report that they are comple-
tely free to buy their circuit boards from exter-
nal vendors.
As can be seen from Panel C of Table 3 this
result is replicated in Continental Communica-
tion’s Printed Circuit @‘CD) division which
reports a minor level of asset specificity. End
product divisions receiving components from
PCD report only minor constraints on their abil-
ity to source related components externally.
World-Wide’s Printed Circuit Board (PCB)
I7 Also consistent with this strategy, High-Tech has over time divested facilities supplying low-technology, commodity type
components such as plastics and metals.
446 G. J. COLBERT and B. H. SPICER
division reports a moderate level of asset spe-
cificity for its largest transfers and minor other-
wise. Its internal customers, regardless of the
size of the transfer, reported only minor con-
straints on their ability to externally source
boards which were also available from PCB.
With respect to these reported results it is
important to note that our theory predicts dif-
ferent policies on sourcing constraints when
asset specificity in the component division is
either high or low. Where asset specificity is
in the mid-range, as in this instance, other fac-
tors may have more influence. In this case our
interviews lead us to believe that what we
observe is due to World-Wide’s corporate cul-
ture which, as noted in the earlier description
of World-Wide’s operations, placed a strong
emphasis on fostering autonomous divisional
authority and responsibility and minimizing
the impact of imposed corporate policies.
The fact that the PCB division has been able
to operate at or near capacity, without upper
management intervening to require internal
sourcing of circuit boards, has meant that allow-
ing buying divisions autonomy has not resulted
in serious conflict with the wider economic
interest of the firm in capacity utilization.
Managing selling decisions when asset
specificity is signfflcant in the end
product (buying) divisions
When end product (buying) divisions have
a sfgniflcant level of asset specijk&y related to
internal transfers, we expect to find compo-
nent (selling) divisions constrained in their
ability to sell components externally in order
to protect the transaction-specifk investments
made by end product divisions.
From Panel A of Table 3, we see this is the
case with respect to High-Tech’s Circuit Tech-
nology (CT) and Packaging & Interconnect
(P&I) divisions. External sales are pursued, but
only as a way to utilize capacity and not when it
jeopardizes their internal mission or their pro-
prietary technology. Both CT And P&I are con-
strained in their ability to sell advanced and
proprietary technology outside the company
in order to protect the differentiated nature of
the company’s end products. However, on less
advanced processes and technology the divi-
sions have the freedom to pursue external busi-
ness to better utilize capacity. Interestingly,
these component divisions also maintain some
reserve capacity in order to manage their rela-
tionship with the end product divisions. For
example, P&I’s constraint is driven by a policy
that the requirements of end product divisions
must be met first and foremost. Consequently,
the division maintains some reserve capacity for
internal customers even though its manage-
ment believes that it could Iill that capacity
with profitable external business.
The only other instance of buying divisions
reporting significant asset specificity relates to
Southern’s Electronic Circuit (EC) division. As
can be seen from Table 3, there are moderate
constraints on this component division’s ability
to sell externally. EC will not sell advanced tech-
nology without the permission of internal cus-
tomers who would be impacted by a unilateral
decision made by the division. Otherwise, the
EC division is allowed to sell components exter-
nally to help defray the cost of its capacity.
Managing selling decisions when asset
specijkity is minor in the end product
(buying) divisions
When end product (hying) divisions have
a minor level of asset specificity related to
internal transfers, we do not expect to fmd
component (selling) divisions constrained in
their ability to sell components externally in
order to protect the minor transaction-specijk
investments made by end product divisions.
From Panel A of Table 3, we see this is the
case with respect to High-Tech Inc.‘s Circuit
Board (CB) division. Internal customers report
minor asset specificity and the CB division
reports it is completely free to sell externally.
This result is replicated ln World-Wide’s Cus-
tom Electronic Circuit (CEC) division which is
also free to sell to outside customers even
though internal customers have hrst priority
and may, on occasion, object to the external
sale of particular components.
However, from Table 3 and Panel B of Table
TRANSFER PRICING PROCESS
4 we see the opposite result in three instances.
World-Wide’s Printed Circuit Board (PCB) divi-
sion and Continental’s Integrated Clrcult (IC)
and Printed Circuit (PCD) divisions all report
they are constrained from selling externally
(even though end product divisions report
only a minor level of asset specificity related
to internal transfers from these component
divlsions). In these cases there are no reasons
associated with the need to protect the transac-
tion-specific investments of the end product
divisions for constraining these component
divisions from selling externally. Therefore,
other factors must account for these selling
constraints.
Our case-bycase investigation revealed that
each of these component divisions was estab-
lished Initially for the purpose of meeting inter-
nal needs. In this context, the observed
constraints ensure that each component divi-
sion continues to focus on serving internal cus-
tomers. Continental’s PCD division controller
expressed the common dilemma posed by
external sales for component divisions in verti-
cally integrated firms:
who do I serve first, the external customer or the mter-
nal customer? That is the biggest problem that we don’t
have an answer for. We’ve looked at selling externally
but we’re deterred by the external-internal conflict.
Consequently, it is reasonable to conclude
that the reported constraints in these three
cases are due to each company’s vertical inte-
gration strategy. It should also be noted that
each of these divisions has been able to oper-
ate at or near capacity based on internal
volume alone although this has required some
consolidation and downsizing in the two cir-
cuit board operations.
THEORY - INVESTIGATION 447
Strategic sourcing considerations
In each Iirm we observe that there are stra-
tegic motivations for the decisions to invest in
particular component technology and capacity.
Each iirm has invested in integrated circuits
(and in some cases related hybrid facilities) to
assure access to and control over technological
capability that provides differentiated features
in their end products. The comments of the
general manager of a major product division
at High-Tech help make this clear:
IC’s [integrated circuits] are our proprietary technol-
ogy. If we were forced to buy off the shelf ICs it would
compromise part of our competitive advantage: we’d be
giving up our leverage.
The manager in a product division at World-
Wide made a similar comment:
We look to the CEC division for ICs [integrated circuits]
horn a strategic viewpoint and PCB for circuit boards
from an overall production cost and reliability stand-
point. By strategic I am referring to proprietary
designs, processes and technology that translate into
performance and features.
In all four cases it appears that integrated
circuits and hybrid components are regarded
as being of great strategic importance.” Con-
sistent with our expectations, buying units are
generally constrained from externally sourcing
integrated circuit or hybrid components which
are available from internal suppliers. This con-
straint acts to protect the value of the Iirm’s
investment in strategically important compo-
nent facilities.
Circuit boards, on the other hand, are not
generally regarded as having the strategic
importance of integrated circuits. For exam-
ple, in contrast to the integrated circuits pro
duced in High-Tech’s CT division, the circuit
boards produced by the CB division are not
i8 Continental’s IC division is regarded as strategically important because of the critical functionality its components
provide to end products. However, in comparison with High-Tech, World-Wide and Southern’s integrated circuit facilities
the degree of strategic importance appears lower. We believe that this is due to its different supply role. Continental’s IC
division focuses on producing high-volume mature components which could be sourced externally. Another division
produces the low-volume newer technology circuits.
448 G. J. COLBERT and B. H. SPICER
regarded as providing proprietary technology
From our cases it appears that the strategic
to the end product divisions. Although the CB
value of the component is important to the way
division supports end product division strategy
through superior service, for example turn-
in which the internal transfer process and sour-
around time, reliability and quality, the
cing in particular ends up being managed. The
absence of proprietary technology and the
availability of circuit boards from external
following representative comment by a man-
sources means that CB is of lesser strategic
importance to the firm. Consistent with CB’s
ager in an end product division of Continental
role in High-Tech’s competitive strategy CB’s
internal customers are free to buy circuit
Communications captures the importance of
boards externally. Similarly, we observe in
our other cases that buying divisions are lar-
gely free to buy circuit boards from external
strategic considerations in sourcing decisions:
suppliers even when circuit board manufactur-
ing capacity exists internally.
The iirst criterion is always whether the component is
strategic or not. If the component is strategic there will
always be a heavy emphasis on making [or buying
inside]. Where the component is not regarded as strate-
gic and is low in volume then there is no constraint and
we can go to the lowest bidder.
Similar comments were made by the man-
agers we interviewed in our other cases. The
The second research question we address is:
association we found between the strategic
roles of different component facilities and sour-
How do the business units of high-tech-
cing policies is set out in Table 3 case by
nology companies set transfer prices for
case.” As noted earlier, the issues of strategic
importance and asset specificity are closely
bound together and are difficult, if not impos
particular component transfers and why
sible, to disentangle.
do the methods used differ?
Case evidence pertaining to this question is
set out by case in Table 3 and a cross-case
summary of our findings is given in Table 5.
TABLE 5. Cross-case data on transfer pricing
Assets specificity reported by component (se%ng) divisions
SigmjTicant
High-Tech
CT division
P&l division
World-Wide
CEC division
Continental Communications
IC division
Southern Defense and Aerospace
EC division
Minor
High-Tech
CB division
Continental Communications
PCD division
Moderate
World-Wide
PCB division
l Significant weight also given to market reference.
t On occasions may also give weight to market references.
Signiiicant weight given to manufacturing costs?
Yes
Yes
Yes
Yes*
Yes
No
No
Yes t
I9 The dangers of overlooking important strategic issues in making sourcing decisions generally is addressed by Venkatesan
(1992).
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
449
Transfer pri ci ng when asset speci j ki ty i s
si gni fi cant i n the component (sel l i ng)
di vi si on
When component di vi si ons have si gni J ?cant
l evel s of i nvestment speci al i zed to i nternal
transfers, we expect to fi nd greater wei ght
gi ven to manufacturi ng costs i n the setti ng
of transfer pri ces, i .e. we woul d expect to
observe cost-based transfer pri chg.
From Panel A of Table 3 we see that High-
Tech’s Circuit Technology (CT) and Packaging
& Interconnect (P&I) divisions each report a
sign&ant level of asset specificity related to
internal transfers. As expected, both report
they use cost-based pricing. Each of these com-
ponent divisions sets its transfer price at full
manufacturing cost as determined by an activ-
ity-based costing system plus a mark-up to
cover non-manufacturing costs.
As can be seen from a review of the other
panels in Table 3, and as summarized in Table
5, this result is generally replicated for all the
other integrated circuit component divisions,
i.e. World-Wide’s Custom Electronic Circuit
(CEC) division, Continental’s Integrated Cir-
cuit (IC) division, and Southern’s Electronic
Circuit (EC) division. All of these divisions
report significant asset specificity and cost-
based transfer pricing. World-Wide’s CEC divi-
sion transfer prices are usually based on manu-
facturing cost plus a mark-up to cover non-
manufacturing costs. However, it also reports
that infrequently it will also consider external
quotations in transfer pricing negotiations in
order to keep particular component business.
This arises primarily with those components
that are more commonly available and involve
higher volumes, and where market prices are
more informative.
Continental’s Integrated Circuit (IC) division
reports giving significant weight to cost in set-
ting transfer prices. However, it also gives
weight to market referents in neogtiating trans-
fer prices. Internal customers frequently get
quotes from outside vendors which they then
use in negotiations with the IC division. In con-
trast to other IC facilities in this study, this
integrated circuit facility specializes in high-
volume circuit designs that are produced on
mature fabrication processes. This has an
important implication. Market prices are more
likely to be informative and useful in these
circumstances for evaluating the value of a
component and the efficiency of the compo-
nent division. The controller of an end product
division describes how transfer pricing works
with Continental’s IC division:
There arc essentially no constraints on how the transfer
price is set so long as the prime directive of not going
outside is complied with. Consequently, the transfer
price is set by negotiation. We’re always going to out-
side vendors to find out what they could do it for to use
as leverage in negotiations and the IC division tries to
factor their costs into what they try to charge us.
Continental’s IC division’s use of market refer-
ents is in response to the cost pressures placed
on end product divisions and the presence of
competitive external suppliers. Bringing mar-
ket referents into the negotiating process helps
to protect the firm’s transaction-specific invest-
ments since negotiated adjustments to the
transfer price may reduce the motivation to
source the component externally. Reference
to external quotations also places pressures
on the component division to operate effi-
ciently. As suggested above, the use of market
referents appears to be associated with the
mature processes being used to produce com-
ponents. It is important to note that we did not
observe their use where transferred integrated
circuit components were produced on state-of-
the-art processes.
Southern’s EC division also uses cost-based
transfer pricing, typically manufacturing cost
plus a fixed mark-up to recover non-manufac-
turing costs. However, because of its large
amount of government work, the ease of con-
forming to government reporting requirements
is also cited as an important reason for cost-
based pricing. Internal contracts are typically
written to mirror the terms of external prime
contracts (cost-plus or fixed cost). With fixed
cost type internal contracts the component
unit keeps or bears any cost underrun or over-
run. This is done to motivate efficient opera-
450 G. J. COLBERT and B. H. SPICER
tions in the component division and to give the
prime contractor division control over costs.
The internal buying unit’s negotiating concern
stems from the impact of the transfer price on
the total cost of a programme or project. Con-
Bitt over sourcing and transfer pricing is mod-
erately common. On occasions group managers
wilI get involved ln arbitrating a transfer pri-
cing dispute. Also, on occasions, corporate
managment will indicate that winning a speci-
fic contract is most important for the company,
in which case considerable pressure is placed
on the selling unit to reduce its transfer price.
In addition, frequently EC will sell at less than
its full cost to match competitive bids received
by internal customers and to support an end
product division’s efforts to bid successfully
on new programmes and contracts.
nental’s Printed Circuit (PCD) division. PCD’s
transfer pricing is primarily market-based using
a portfolio approach, i.e. circuit boards with
sirnllar characteristics have the same or similar
prices where the base price of portfolios is set
primarily on market quotes. The PCD division
controller explains:
We slice printed circuit boards into technologies or
layers. For example, on a double sided board all con-
tractors who supply Continental with that product
(including us as the inside contractor) arc contracted
at the same price per panel. But there are kickers if it
has Rne lines or a thicker level or ls an odd size, etc. We
charge an extra percentage for these. We don’t know lf
the competition does this, but we think we are ln the
ballpark.
Transfer pri ci ng when asset spec@ci ty i s
mi nor i n the component (sel l i ng) di vi si on
When component di vi si ons have l i ttl e spe-
ci al i zed i nvestment rel ated to i nternal trans-
fers we expect to find market-based pri ci ng I n
use.
This is what we observe for transfers from
High-Tech’s Circuit Board (CB) division which,
as shown in Table 3, reports only a minor level
of asset specificity. The CB division has com-
plete autonomy to sell externally and to set
transfer prices. They use a market-based trans-
fer price which is determined by the same algo-
rithm used for external pricing. There is one
interesting exception to this management pol-
icy. In those infrequent instances where the
price determined by the external pricing algo-
rithm is lower than full manufacturing cost as
determined by CB’s activity-based costing sys
tern, CB will agree to sell Internally, but only at
cost. This represents a concession to internal
customers who, for convenience or other rea-
sons, still desire to purchase the part internally.
In both High-Tech’s CB division and Conti-
nental’s PCD division, the internal division is
most likely to be competitive on manufactur-
ing newer boards which involve more
advanced technology, are most difficult to man-
ufacture, and are lower in volume. As these
products mature these shops are often less
competitive than other external suppliers and
end product divisions have incentives to
reduce costs by exploring options to buy exter-
nally. In both cases end product divisions have
the freedom to do this. Whenever costs exceed
the market price, end product divisions are
encouraged by the upstream internal compo-
nent division to move these boards to external
suppliers who are generally cheaper. These are
typically the more mature, less complex
boards.20
The general finding is repeated with Conti-
Unlike other component divisions studied,
World-Wide’s Printed Circuit Board (PCB) divi-
sion reports a moderate level of asset specifl-
city for its largest Internal transfers and minor
otherwise. Consistent with this mid-way posi-
tion, PCB reports that, as a general policy, it
uses cost-based pricing where cost is manufac-
turing cost (as determined by an activity-based
cost system) plus a mark-up for non-manufac-
^^
‘” In Continental PCD’s case this has resulted ln reducing its capacity. The controller explains: “We are doing a number of
lower technology boards that we price at market. The market price is so low for these that we can’t afford to make them
anymore. It is internally cheaper for us to downsize the plant and just stay with the high technology boards.”
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 451
turlng costs. However, it may infrequently divisions and end product divisions. By
negotiate a different price where an external attempting to meet profit targets by lilling capa-
supplier has provided a lower price. Customer city, component divisions became increasingly
divisions regularly obtain external quotes to focused on external sales opportunities to the
compare with internal transfer prices. In negotia- disadvantage of internal customers and their
tions with end product divisions who have only speciafized investments.
minor constraints on their ability to buy exter-
End product divisions started to pursue
nally, the PCB division argues that its costs are
external sources of key components owing to
close to or consistent with market references.
what they regarded as unreasonable transfer
pricing. Managers quickly realized that this
was to the detriment of the component divi-
SUMMARY, CONCLUSIONS AND
RESEARCH DIRECTIONS
sions and the significant specialized invest-
ments that they had made in the internal
Overall eval uati on of the theory
transfers of components. Divisional and corpo-
High-Tech was the first case we studied and
rate managers began to look for a way out of
the confusion
analysed. Based on our observations which are
, muddle and conflict that
reported in Panel A of Table 3 and summarized
resulted. The consensus that finally emerged
in Tables 4 and 5, we conclude that the empiri-
was that the primary role of the component
cal pattern of findings for High-Tech are consis-
divisions was to service end product divisions
tent with the theoretical predictions. The
by supplying them proprietary h&h-technology
manner in which High-Tech manages the trans
components and service. To support this strat-
fer pricing process and the weight given to
egy, and ostensibly to protect the value of spe-
setting transfer prices are positively related to
cialized investments made by component
the reported degree of asset specificity asso-
(selling) and end product (buying) divisions,
ciated with the transfer. The way in which
the company constrained their selling and buy-
High-Tech manages the transfer process also
ing policies. In addition, a decision was made to
appears to be related to the strategic importance
evaluate component divisions as profit centres
of the component itself and the nature of the
with a break-even objective on internal sales.
component division’s production capabilities.
Transfer prices were to be cost-based, being
Interestingly, High-Tech has made changes
full manufacturing costs as determined by
in its sourcing, selling and transfer pricing poli-
newly installed activity-based costing systems
ties as it has struggled to manage its compo-
plus a mark-up for non-manufacturing costs.
nent operations effectively. These changes and
The CB division was the notable (and speci-
the resulting contrast in management between
fit) exception to the above change in policy.
the CT and P&I divisions on the one hand,
This component division and the end product
versus the CB division on the other, provide
divisions retained the right to sell or buy circuit
additional evidence in support of the theory.
boards externally free of constraint. CB
For a period High-Tech managed all three
remained on market-based pricing with full
component divisions as cost centres. It then
profit centre status on both internal and exter-
switched to managing them as full profit cen-
nal sales. We believe this organizational treat-
tres (on internal and external sales) using quasi-
ment reflects both CB and end product
market-based pricing. Component divisions
divisions’ low asset specificity as well as CB’s
and end product divisions were in principle
less important strategic role in the company.
free to sell and buy externally. Although this The findings for the other cases are similar to
policy helped to change component divisions’ those for High-Tech. From Tables 3,4 and 5 we
orientation and culture it also proved disrup- conclude that there is considerable replication
tive to the relationship between component of specific findings consistent with theory.
452 G. J. COLBERT and B. H. SPICER
Overall, our findings generally match the pat- on the level of asset specificity reported by
tern of relationships suggested by our theory. the selling divisions.
One interesting factor that seems to bear on
the sourcing and transfer pricing process in the
tirms we studied was the circumstances faced
by the end product divisions. In general, end
product divisions were under pressure to
improve their financial performance in the
face of increasingly intense competition in
their respective end product markets. In each
case decision making and accountability had
been decentralized to a considerable extent.
Division management had been left to run
their own businesses but were held accounta-
ble for meeting profit targets. Upper manage-
ment’s role, besides setting profit targets, was
reserved to setting the larger corporate vision
about what (and how) markets and technolo-
gies should be pursued. What can be seen from
the companies studied is that difficult organiza-
tional issues arise where firms are decentralized,
yet the achievement of current and future goals
rests on some interdependent strategy between
units which have made specialized investments
specific to internal transfers. The manner in
which firms we studied manage the transfer pro-
cess and set transfer prices appears to have been
systematically influenced by the presence or
absence of these specialized investments.
For all four cases the strategic importance of
the component or the production capabilities
of the internal component supplier are asso-
ciated with specific investments in component
operations. Each firm has invested in integrated
circuit (and in some cases related hybrid facil-
ities) to ensure access to and control over tech-
nology that provides differentiated features in
their end products. In contrast, circuit board
facilities appear to operate primarily to sup-
port strategy by providing superior service
such as flexible delivery and quick design turn-
around. For component divisions which are
regarded of higher strategic value, buying divi-
sions are constrained from sourcing externally.
Similarly, component divisions are generally
constrained from selling strategically impor-
tant technology externally.
Managi ng the transfer process
From theory we expectprms to put i nsti tu-
ti onal arrangements i nto pl ace to protect
thei r economi c i nterests i n i nvestments spe-
ci al i zed to i nternal transfer acti vi ty and the
val ue of the prm’s competi ti ve strategy.
As can be seen from Panels A and B of Table
4, the firms studied generally constrain subunit
managers in selling and sourcing decisions
when there are significant levels of asset speci-
ficity associated with internal transfers. Conver-
sely, the firms generally do not constrain
buying division’s sourcing decisions when
there are only minor amounts of asset specifi-
city reported by the component (selling) divi-
sions. These fmdings support the theory. Firms
in this study vary the constraints on the sour-
cing decisions of buying divisions depending
However, in three instances, component
divisions were constrained from selling exter-
nally even though the level of asset specificity
in the end product divisions was reported as
minor. As our theory suggests that there is no
need for selling constraints to protect transac-
tion-specific investments in the end product
divisions serviced by World-Wide’s Printed Cir-
cuit Board (PCB) division, and Continental’s
Integrated Circuit (IC) and Printed Circuit
(PCD) divisions, the observed constraints
must exist for other reasons. As noted above,
,in each of these cases it seems that the
intended strategy of the firm in setting up the
component division to serve internal needs and
a continued desire to retain an internal focus
was judged to be more important than the
potential benefits from selling externally. This
factor should be considered for inclusion in a
more expanded theory of constraints in inter-
nal transfer processes.
Transfer pri ci ng
From theory we expect that component
di vi si on transfer pricing pol i cy wi l l be rel ated
to the extent of asset spec@ci ty i nvol ved i n the
transfers. As asset speci j i ci ty of the component
di vi si on i ncreases, market pri ces wi l l become
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 453
i ncreasi ngl y l ess useful and rel evant for
i nformi ng the transfer pri ci ng process and
greater wei gbt wi l l be gi ven to manufactur-
i ng costs i n setti ng transfer pri ces.
As can be seen from Table 5, the findings
generally support the theory. Component divi-
sions with low levels of asset specificity give
market prices most weight. When component
divisions have high levels of asset specificity,
manufacturing costs are given most weight.
We did tind two instances where component
divisions with high levels of asset specificity
made use of market references to inform the set-
ting of transfer prices. The first was High-Tech
which had previously made an attempt to use
quasi-market-based transfer pricing by looking
at prices for very roughly comparable intermedi-
ate products available from external suppliers
and then attempting to make gross adjustments
to these prices for differences in design and func-
tionality. High-Tech’s managers abandoned this
approach because they became convinced that
the so-called “market-based” transfer prices for
components from the Circuit Technology (Cl‘)
and Packaging & Interconnect (P&I) divisions
were not useful and were causing serious deci-
sion-making errors and high levels of conflict
within the firm. By the time we conducted our
study these two component divisions had
abandoned market-based pricing in favour of
cost-based pricing based on activity-based manu-
facturing costs.
We also found one instance where a compo-
nent division (Continental’s Integrated Circuit
(IC) division), which reported a high level of
asset specificity, gave weight to both manufac-
turing costs and external market referents. This
division specializes in high-volume compo-
nents produced on mature fabrication pro-
cesses where market references are more
relevant. Reference to external quotations in
this instance served to protect the component
division’s specialized investment (and the
firm’s interests) by helping to retain large-
volume transfers within the firm.
In the mid-range of asset specificity reported
by World-Wide’s Printed Circuit Board (PCB)
division, weight was also given to market refer-
ents in addition to manufactuimg costs, on
occasions.
Li mi tati ons and suggesti ons for future
research
As we noted in the Introduction, the transfer
pricing literature has both theoretical and
empirical problems. While there is an abun-
dance of normative papers there are far fewer
papers which make a strong theoretical contri-
bution towards explaining how transfer pricing
processes are actually managed. On the empiri-
cal level, most surveys of practice are of limited
value, firstly, because they tend to abstract
from industry and organizational context and,
secondly, because they are wrongly directed at
the level of the firm as a whole rather than at
the subunit level where internal transfers actu-
ally take place.
We attempted to address the theoretical
issue by first developing a theory of the trans-
fer pricing process from a transaction costs
perspective. To avoid the empirical difficulties
that have troubled surveys of transfer pricing
processes and practices and to contribute to an
understanding of transfer pricing processes in
their organizational contexts, we adopted a
multiplecase study approach with embedded
units, i.e. the divisions in each firm and compo-
nent transfers between them. However, like
other research studies, ours has its limitations.
The first has to do with making the asset spe-
cilicity construct operational. We use a subjec-
tive categorical evaluation of the impact of the
potential effect on investment values if internal
transfers were halted. However, we argue that
to the extent that subjective assessments direct
behaviour the subjective categorical evalua-
tions of asset specificity we use are likely to
be consistent with observed management of
the transfer process.*’ Moreover, we were
careful to query each person we interviewed to
” We asked a number of managers whether it was feasible for them to quantify objectively the actual dollar impact on
specific investments of the loss of an internal transfer. The negative responses we received lead us to believe that
dwelopments of such a direct, objective measure of asset specificity is not feasible.
454 G. J. COLBERT and B. H. SPICER
ensure that each question was understood and
was supported by specifics.
A second limitation is that although the rela-
tionship between asset specificity reported by
component (selling) divisions was an important
determinant of resulting transfer prices, in
some instances other firm-specific factors
appear to account for exceptions to policies
based on asset specificity considerations
alone. For example, we noted that transfer pri-
cing policies were influenced, in some cases,
by market conditions for the transferred com-
ponent and competitive pressures in end pro-
duct markets. We also observed circumstances
where managers chose to deviate from their
division’s transfer pricing policy to meet exter-
nal quotes in order to keep internal customers
and/or to accommodate their needs. In addi-
tion, organizational history, management style
and organizational culture appear to play a
moderating role. Ease of conformance to gov-
ernment contracting and reporting require-
ments also influenced a decision to use cost-
based transfer pricing in one case.
A third limitation relates to the inherent
drawbacks of the case study method. Like
other methods of research it has its own
strengths and weaknesses. A major criticism
of case study research relates to the issue of
generalization. Following Yin (1989) our pri-
mary objective was not to generalize from a
sample to a larger population. Bather, our
intent was to provide analytical evidence
about the empirical validity of theory by pat-
tern matching evidence back to theory. How-
ever, by using a multi-case research design we
believe there is some assurance that the empiri-
cal patterns we observed are not idiosyncratic
to a particular firm in the high-technology, elec-
tronics industry.
Clearly, additional case studies would be
helpful. These may involve further studies
amongst high-technology electronics firms
and with vertically integrated firms in other
industries. However, the problems of access,
cost and logistics in undertaking multiple case
research of this type should not be underesti-
mated. It may be more efficient now to use
survey methods to extend the empirical inves-
tigation to a broader cross-section in selected
industries. An extension of the existing survey
literature could focus on particular types of
intra-firm transfers where asset specificity and
the strategic importance of component opera-
tions are expected to differ.
The question of the management of the
transfer process and transfer pricing is an
important one for the managers and academics
alike. Our aim in this paper was to make a
contribution to understanding how and why
managers handle this problem in the complex
and uncertain product environments of high-
technology, electronics firms. Our multicase
investigation has provided empirical evidence
that asset specificity plays an important and
systematic role as predicted by our theory.
However, our empirical evidence revealed
that other firm-specific factors influence man-
agement of the process and the setting of trans
fer prices. Further rounds of theory building
and empirical investigation might incorporate
these additional factors.
BIBLIOGRAPHY
Abdel-khalik, A. & Lusk, E., Transfer Pricing-A Synthesis, Accounffng Revi ew (1974) pp. 8-23.
Alchian, A., Specificity, Specialization, and Coalitions, J ournal of I nstftutfonal and Theoreti cal Econom-
ics (19B4) pp. 34-49.
Alchian, A. & Woodward, S., Reflections on the Theory of the Firm, J ournal of I nsti tuti onal and
Tbeorettcal Economi cs (1987) pp. 110-136.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 455
Ball, R., The Firm as a Specialist Contracting Intermediary: Application to Accounting and Auditing,
Working paper, William E. Simon Graduate School of Business Administration, University of Rochester
(June 1989).
Borowskl, S., Environmental and Organizational Factors Effecting Transfer Pricing, J ournal ofMunuge_
ment Accounttng Research (Fall 1990) pp. 78-99.
Brims, W. Jr & Kaplan, R., Introduction: Field Studies in Management Accounting, in Bruns, W. & Kaplan,
R. (cds), Accounting and Management: Field Study Perspectives (Boston: Harvard Business School
Press, 1987).
Coax, R., The Nature of the Firm, Economica (1937) pp. 386-405.
Colbcrt, G. J., An Empirical Investigation of the Transfer Process and Transfer Pricing: A Multicase
Research Design, Doctoral dissertation, University of Oregon (December 1991).
Ecclcs, R. G., The Transfer Pricing Problem (Lexington: Lexington Books, 1985).
Eiscnhardt, K., Building Theories from Case Study Research, Academy of Management Review (1989)
pp. 532-550.
Grabski, S., Transfer pricing in Complex Organizations: a Review and Integration of Recent Empirical and
Analytical Research, J ournal of Accounffng Ltterature (Spring 1985) pp. 33-71.
Hertcnstein, J., Management Control System Change: the Adoption of Inflation Accounting, in Bruns, W.
% Kaplan, R. (eds), Accountfng and Management: Field Study Perspectives (Boston: Harvard Business
School Press, 1967).
Jones, G. % Hill, C., Transaction Cost Analysis of Strategy-Structure Choice, Strafegfc Management
J ournal (1988) pp. 159-172.
Klein, B., Vertical Integration as Organizational Ownership: the Fisher Body-General Motors Relationship
Revisited, J ournal of Luw, Economtcs and Organization (1988) pp. 199-213.
Levy, D., The Transactions Cost Approach to Vertical Integration: an Empirical Examination, Review of
Economfcs and Statfstfcs (1985) pp. 438-445.
McKinnon, J., Reliability and Validity in Field Research: Some Strategies and Tactics, Accountfng, Au&-
fng and Accountability (1988) pp. 34-54.
Maher, M., The Use of Relative Performance Evaluation in Organizations, in Bruns, W. % Kaplan, R. (eds),
Accountfng and Management: Field Study Perspectfves (Boston: Harvard Business School Press,
1987).
Mastcn, S., The Organization of Production: Evidence from the Aerospace Industry, J ournal of Law and
Economics (1984) pp. 403-418.
Mastcn, S., Mcchan, J. Jr % Snyder, E., Vertical Integration in the U.S. Auto Industry, J ournal of Economic
Behavior and Organization (1989) pp. 265-273.
Merchant, K., How and why Firms Disregard the Controllability Principle, in Bruns, W. & Kaplan, R.
(cds), Accounting and Management: Field Study Perspectives (Boston: Harvard Business School Press,
1987).
Merchant, K. & Manzoni, J. F., Achievability of Budget Targets in Profit Centers: a Field Study, The
Accounting Review (1989) pp. 539-558.
Montcvcrdc, K. & Tccce, D. J., Supplier Switching Costs and Vertical Integration in the Automobile
Industry, BeU J ournal of Economics (1982) pp. 206-13.
Palepu, K., The Anatomy of an Accounting Change, in Bruns, W. & Kaplan, R. (cds), Accounting and
Management: Field Study Perspectives (Boston: Harvard Business School Press, 1987).
Porter, M., Compefftfve Strutegy (New York: Free Press, 1980).
Porter, M., Competitfue Advantage Wcw York: Free Press, 1985).
Rcvc, T., The Finn as a Nexus of Internal and External Contracts, in Aoki, M., Gustafsson, B. % William-
son, 0. (cds), The Firm us u Nexus of Tmulies, pp. 133-161 (Newberry Park: Sage, 1990).
Rio&n, M. L Williamson, O., Asset Specificity and Economic Organization, International J ournal of
I ndustrial Organization (1985) pp. 365-378.
Spiccr, B. H., Towards an Organizational Theory of the Transfer Pricing Process, Accounting, Orgunfzu-
tfons and Society (1988) pp. 302-322.
Spiccr, B. H. % Ballcw, V., Management Accounting Systems and the Economics of Internal Organization,
Accounting, Organfzations and Society (1983) pp. 73-98.
Tang, R., Transfer Pricing Practices I n the United States and J apan (New York: Pragcr, 1979).
Tang, R., Canadian Transfer Pricing Practices, c4 Muguzine (March 1980) pp. 32-38.
Tccce, D., Firm Boundaries, Technological Innovatlon, and Strategic Management, in Thomas, G. (cd.),
The Economics of Strategfc Planning, pp. 187-199. (Lexington: Heath and Co., 1986).
456 G. J. COLBERT and B. H. SPICER
VanciI, R. F., Decentralization: Manageri al Ambi gui ty by Desi gn (Homewood: Dow Jones Irwin, 1979).
Venkatesan, R., Strategic Sourcing: to Make or Not to Make, Harvard Busi ness Reufew (November-
December 1992) pp. 233-261.
WaIker, G., Strategic Sourcing, Vertical Integration, and Transaction Costs, Interfaces (May-June 1988)
pp. 62-73.
WaIker, G. & Weber, D., A Transaction Cost Approach to Make or Buy Decisions, Admi nfstrathe Sci ence
Quarterl y (September 1984) pp. 373-391.
Watson, D. & BaumIer, J., Transfer Pricing: a Behavioral Context, The Accounti ng Ret&w (1975) pp.
466-474.
Wiiamson, 0. E., Markefs and Hi erarchi es (New York Free Press,1975).
Williamson, 0. E., Transaction Cost Economics: the Governance of ContractuaI Relations,JournaC oftaw
and Economi cs (October 1979) pp. 233-261.
Williamson, 0. E., The Economi c I nsti tuti ons of Capi tal i sm (New York: Free Press, 1985).
Williamson, 0. E., Strategizing, Economizing, and Economic Organization, SErategfc ManagementJ our-
naI (l 991) pp. 75-94.
Yin, R. K., Case Study Research: Desi gn and Methods (Newbury Park: Sage, 1989).
APPENDIX
The question used to gather information on the key construct of overall asset specificity associated with an internal
transfer for the internal seiling unit is shown below. Similar questions were fust asked about individuai types of asset
specificity, i.e. physical asset specificity such as plant, equipment, software, moulds, tooling and dies; human asset
specificity such as training and specialized know-how; inteilectual asset specificity such as design work, R&D efforts
and patents; and site specificity.
Each question about asset specificity had two parts. The iirt part was designed to set the stage for the second part
which gauges the degree of asset specificity. The fnst is designed to get the interviewee to thhtk about the concept. The
second question is then more directly focused on the extent to which an attempted redeployment of the transaction-
specific investment would result in loss of value.
To facilitate our written analysis in the text we treated the responses No Decline and Minor Decline as equivalent and
Significant or Very Signiftcant Decline as equivalent.
Overall asset speci fci ty
(a) How would you rate the OVERALL extent to which your unit has made investments SOLELY to manticture the
selected product (i.e. assets that could not be easily transferred to other operations or products)?
Very
No Minor Moderate SigniIicant signiEcant
Investment Investment Investment Investment Investment
1 2 3 4 5
(b) OVERALL, what decline in the investment value (owing to increased costs or reductions in revenue) would be
involved in an attempt to transfer or redirect these investments to other operations or products?
No
Decline in
Investment
Value
1
Minor
Decline in
Investment
Value
2
Moderate
Decline in
Investment
Value
3
significant
Decline in
Investment
Value
4
Very
Signiftcant
Decline in
Investment
Value
5
doc_944646301.pdf
Drawing on published work by Williamson Vournal of hw and Economics (October 1979) pp. 233-
261; The Economic Institutions of Capitalism (New York: Free Press, 1985)], Spicer [Accounting,
OrganFzatfons and Society (1988) pp. 302-3221, and Walker [Interfaces (May-June 1988) pp. 62-731
this paper first develops a theory of the sourcing and transfer pricing process from a transaction costs
perspective.
Pergamon Accounting, Organlzatfonr and So&q, Vol. 20, No. 6, pp. 423-456, 1995
copyrighr D 1995 Elsevkr Science Ltd
Printed in Great Brltaln. AU r&M reserved
0361-3682/95 $9.50+0.00
03613682(95)000074
A MULTI-CASE INVESTIGATION OF A THEORY OF THE TRANSFER PRICING
PROCESS*
GARY J. COLBERT
Utahem@ of Col orado at Denver
and
BARRY H. SPICER
The UniversZty of Auckl and
Abstract
Drawing on published work by Williamson Vournal of hw and Economics (October 1979) pp. 233-
261; The Economic I nstitutions of Capitalism (New York: Free Press, 1985)], Spicer [Accounting,
OrganFzatfons and Society (1988) pp. 302-3221, and Walker [Interfaces (May-June 1988) pp. 62-731
this paper first develops a theory of the sourcing and transfer pricing process from a transaction costs
perspective. The paper then reports on an empirical investigation of the theory using a multicase
research design involving internal ttansfers in four large, vertically integrated high-technology compa-
nies in the electronics industry in the U.S.A. The ways in which these firms manage internal transfers and
set transfer prices are found to be positively related to the reported degree of transaction-specific
investment (asset specfficity) associated with each transfer and to the strategic importance of the
transferred component and the nature of the component division’s production capabilities. Suggestions
are made for future research.
The transfer pricing process’ represents an
important and pervasive problem in designing
and implementing management information
and control systems. There has been consider-
able academic research published on these
related topics, yet it remains a troublesome
area for academics and managers alike. Even
though significant strategic and behavioural
consequences including suboptimal decision
making, opportunistic behaviour by subunit
managers, and internal friction and dishar-
mony may ensue from the transfer pricing pro-
cess, we do not understand many of the factors
that condition the way in which managers actu-
ally deal with this process.
There appear to be two reasons for this, the
first being theoretical and the second empiri-
cal. Reviews of the transfer pricing literature
reveal many normative treatments of the trans-
fer pricing issue (see Abdel-khalik & Lusk,
1974, and Grabski, 1985, for extensive
reviews) but fewer attempts such as those
made by Eccles (1985) to induce, or Watson
& Baumler (1975) and Spicer (1988) to
deduce, theories of the transfer pricing
process.
l The authors would like to thank Jean Cooper, gay King, Terry O’Kecfe, Frank Selto and Ralph Viator for their helpful
comments on earlier dnxfts of this paper.
1 We use the tetm transfer pricing process to refer to all related transfer activity, including the decision to source
components internally or externally (make-or-buy), the extent to which senior managers assert control over transfer
relationships between subunits, and how tramfer prices are set.
423
424 G. J. COLBERT and B. H. SPICER
The second and related reason is that much of
the empirical research on actual transfer pricing
processes has been done with firm level surveys
(e.g Tang, 1979, 1980; Vancil, 1979; Borowski,
1990). We believe that the understanding and
insights that have emerged from these types of
studies are limited for two important reasons.
First, they abstract from both industry and orga-
nizational context by choosing samples which
span a number of industries, and second, they
are directed at the wrong level, i.e. at the level
of the firm as a whole rather than at the subunit
level where internal transfers actually take
place. The approach we adopt in this paper
attempts to deal with both of these issues in a
meaningful way.
The objective of this paper is to report on a
theoretically based empirical investigation of
factors relevant to the transfer pricing process
in high-technology companies. Using a multi-
case research design we address two research
questions:
(1) How do the business units of high-tech-
nology companies manage the transfer process
and why do these practices differ?
(2) How do the business units of high-tech-
nology companies set transfer prices for parti-
cular component transfers and why do the
methods used differ?
Drawing on work by Williamson (1979,
1985) Spicer (1988) Walker (1988) and Col-
bet-t (1991) we first set out a theory of the
transfer pricing process from a transaction
costs perspective. We then report on an
empirical investigation of the theory using a
multi-case research design in which the pat-
tern of observations made in each case relative
to those suggested by the theory is evaluated.
Our overriding objective is to assess whether
the evidence from multiple cases is consistent
or inconsistent with the pattern of observations
suggested by theory.
Our paper is structured as follows. The sec-
ond section sets out our theory of the transfer
pricing process. The third section presents the
details of the multi-case research design we
employed to select our cases and collect and
analyse evidence from them. The fourth sec-
tion describes each of our four cases. The fifth
section presents and analyses our case evi-
dence and assesses the support the case evi-
dence provides for the theory of the transfer
pricing process laid out in the second section.
The final section provides a summary and con-
clusions which includes a discussion of the
limitations of our study and some suggestions
for future research.
A THEORY OF THE TRANSFER PRICING
PROCESS
Starti ng poi nts
There are two starting points for the theory of
the transfer pricing process that we develop.
The first is a growing body of work that uses
the existence of differential transaction costs as
an explanation for the existence of variation in
the manner in which organizations conduct
their economic activity. Generally referred to
as transaction costs economics, this framework
can be traced to Coase (1937) and Williamson
(1975, 1979, 1985). Spicer (1988) draws on
transaction costs economics to develop a posi-
tive theory of the transfer pricing process in
which the strategic and transactional character-
istics of specific transfers are related to transfer
pricing issues and the organizational processes
used to manage transfers within firms. Based on
the premise that internal transfers are transac-
tions, he argues that transaction costs econom-
ics can be used to consider contracting issues
involved with the management of internal trans-
fers and the setting of transfer prices.
Although Spicer (1988) argues that firm strat-
egy affects intermediate product design and is
related to the dimensions of intra-Iirm transfers
he provides few particulars. This missing detail
is supplied by Walker (1988) who applies the
concepts of Porter’s (1980, 1985) competitive
strategy framework and transaction costs eco-
nomics to explore the risks associated with a
firm’s sourcing decisions, and the implication
of these risks for competitive strategy. Walker’s
work provides the second starting point for our
development of theory.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 425
Transacti on costs economi cs
Williamson (1985, p. 1) argues that the
appropriate level of economic analysis is the
transaction which he defines as the transfer
of goods or services across a technologically
separable interface. Transaction costs econom-
ics focuses attention on the relative costs and
hazards of conducting transactions within alter-
native governance structures.2 As the costs of
conducting transactions within markets
increase it becomes increasingly likely that
firms will resort to alternative arrangements
such as internalizing the transaction. The level
of transaction costs experienced is determined
by the dimensions of the transaction.
The first of these is asset spec@ci ty3 which is
of special importance because without it trans-
action costs economics would lose much of its
significance (Williamson, 1985, p. 52). Asset
specificity arises when durable investments
are made in relation to a particular transaction
and the value of the investment in its next best
use is considerably lower. The value of such
investments in assets is dependent on the main-
tenance or continuation of a particular transac-
tional relationship. If the asset cannot be
costlessly redeployed from its present use to
some other use, the value of the investment
can be jeopardized by the opportunistic action
of the other party to the transaction. Thus the
parties have an incentive to maintain the rela-
tionship through the implementation of appro-
priate contractual or organizational safeguards.
As the level of asset specificity increases, the
transaction costs associated with firms con-
ducting transactions within markets rise and it
becomes increasingly likely that an alternative
arrangement such as internalization of the
transaction (vertical integration) will have
transaction costs advantages.
Williamson (1985, p. 55) identifies a number
of types of asset specificity which involve
investments in the transactions by one or
both parties to the transaction. These include:
(1) physical asset specificity resulting from
investments in tangible assets, e.g. plant, equip
ment, tooling, dies, etc. that are more or less
unique to this economic relationship; (2)
human asset specificity resulting from invest-
ments in specialized training or specialized
know-how gained through education or experi-
ence which is not easily transferred to other
transactions; (3) site specificity which results
from parties locating their operations in close
physical proximity to one another for flow
economies; and (4) dedicated asset specificity
which arises from investments in generalized
(as contrasted to special purpose) productive
capacity for the explicit purpose of selling a
significant amount of a product to a specific
customer and that capacity cannot be cost-
lessly redeployed or otherwise utilized.
There is growing interest in the notion of
asset specificity as an important explanator of
the boundaries of the firm and the related issues
of strategy; the make-or-buy decision; and verti-
cal, horizontal and diversification alliances
(Alchian, 1984; Alchian & Woodward, 1987;
Riordan & Williamson, 1985; Teece, 1986;
Jones & Hill, 1988; Klein, 1988; Reve, 1990;
Williamson, 1991). There is also a body of
empirical research that provides support for
the relevance of asset specificity. Of particular
pertinence to our study are several papers
which uncover a positive relationship between
the presence of specialized assets and the
a Transaction costs are the costs of engaging in transactions via a particular institutionaI arrangement or governance
stmcturc. They can be separated into exunre and ez-posr transaction costs. Eranfe transaction costs are the costs of
drafting, negotiating and safeguardlng contracts. E&r-post transacti ons costs include maktdaption and haggling costs when
transactions get out of alignment, the mnnlng costs of governance structures to resolve disputes, and the bonding costs to
secure credible commitments. Spicer & Rallew (1983, p. 88) point out that intemai frictions within iirms arising from
contml losses and pIanning difiicuhies can be viewed as the hierarchical counterpart of transaction costs in markets.
3 The terms transaction-specific investment and specialized investments arc also used to refer to asset specificity. We use
these terms interchangeably.
426 G. J. COLBERT and B. H. SPICER
extent of vertical integration (Monterverde &
Teece, 1982; Walker & Weber, 1984; Masten,
1984; Masten et al ., 1989; Levy, 1985).*
Transaction costs economics recognizes two
other dimensions of transactions: the uncer-
tafnnty to which the transaction is subject and
the extent of the transaction. Uncertainty is of a
statecontingent kind. When assets cannot be
costlessly redeployed and uncertainty is pre-
sent, efficient adaptations are needed in order
to preserve the relationship. Extent refers to the
frequency and volume of the transaction and
determines whether the cost of specialized gov-
ernance structures to control the transaction
can be recovered. The cost of specialized gov-
ernance structures (e.g. hierarchy) are easier to
recover for large and recurring transactions.
Appl i cati on of transacti on costs
economi cs to the i nternal transfer process
As Spicer (1988) points out, the extent to
which a Iirm makes internal transfers arises
out of a strategic choice between buying com-
ponents from external sources and making
them itself. This is the issue of the appropriate
extent of vertical integration. Porter (1980, pp.
303-307) cites the potential benefits of vertical
integration as including economies of opera-
tions, tapping into technology, assuring sup
ply, offsetting bargaining power and input
cost distortions in noncompetitive markets,
and enhancing ability to differentiate end pro-
ducts. The transaction costs economics frame-
work offers a particular explanation for vertical
integration. In this framework the make-or-buy
decisions will be made so as to minimize the
sum of production and transaction costs.
Dimensions of the transactions involved (asset
specificity, uncertainty and extent) are thought
to be positively related to the decision to inte-
grate vertically into component production.
With low levels of transaction-specific invest-
ments (asset specificity) it is likely that econo-
mies of scale play an important role in the make-
or-buy decision. Suppliers who can aggregate
demand for multiple customers, thereby obtain-
ing efficient operating scales, will be the low
cost producers. At the same time, with low
levels of transaction-specific investments
involved, the potential transaction costs asso-
ciated with buying externally are likely to be
low. Given these conditions it is less likely that
the firm will source the component internally.
As the investments required to produce a
particular component become more transac-
tion-specific, the production cost advantages
will be less between sellers and buyers but
they will face contracting hazards which will
result in additional contracting costs. These
costs include transaction costs of writing, mon-
itoring and executing contracts. Because pro-
duction cost advantages of buying externally
decrease and the comparative transaction
costs of markets increase as assets involved in
the transaction increase in specificity, it
becomes increasingly likely that the end user
will bring the transaction inside the &rn and
make the component.
It is not only sellers that make transaction-
specific investments. Buyers may also make
investments where the value of their invest-
ment is dependent on the conduct and perfor-
mance of a particular supplier. For example,
investments may be made in equipment, train-
ing of personnel or end product and/or compo-
nent design that are dependent on a
component produced exclusively by one sup
plier. The presence of these transaction-speci-
fic investments by the buyer means that there
are potentially high transaction costs involved
in contracting to acquire the related compo-
nent externally.
* Monterverde & Teece (1982) Walker & Weber (1984) and Masten et al. (1989) provide tests of the relevance of asset
speciticity in the automobile segment. Masten (1984) provides a test in the aerospace industry, and Levy’s (1985) tests
spans a cross-section of industries. The presence of specialized assets is pmxied by design costs and number of compe-
titors in Monterverde & Teece (1982); by supplier competition in Walker & Weber (1984); by engineering effort, the
extent to which components are produced on physical assets specific to an automobile maker, and the closeness of siting
of manufacturing stages in Masten et al. (1989); and by research intensive products in Levy (1985).
TRANSFER PRICING PROCESS
Uncertainty in the presence of transaction-
specific investments creates the need for adap-
tation and co-ordination between the contract-
lng parties. The greater the extent (frequency
and volume) of the transaction the more likely
the firm can obtain production economies by
manufacturing the component itself and lnter-
nalixing the transaction.
However, intemalizatlon of the transaction
within the firm does not fully remove the poten-
tial transaction costs. Where the firm is highly
decentralized, transfers between relatively inde-
pendent subunits are similar in some contract-
ing aspects to external transactions. In the limit
a firm could structure and encourage conduct
between subunits that is essentially equivalent
to that of external parties transacting in the
market-place. Alternatively, firms may arrange
their organization structures and control sys-
tems so as to encourage more cooperative,
interdependent behaviour between subunits.
Where the level of transaction-specific
investments in internal transfers made by sub
units is low, the tirm’s economic interests will
generally be served by allowing external mar-
ket forces to influence the transfer, including
the setting of the transfer price. However, as
investments in assets specialized to internal
transfers increase, the firm faces a trade-off
between the potential efficiencies of trading
in a market and its need to protect the value
of the specialized investment.
As the level of transaction-specitic investment
increases, it becomes increasingly important for
the firm to protect its economic interests in the
transfer relationship by ensuring that conflict
and opportunistic actions by subunits are mod-
THEORY - INVESTIGATION 427
erated or controlled. Numerous researchers,
including Vancil (1979) and Fccles (1985)
have noted that relationships between divisio-
nallxed firms can at times involve considerable
conflict, especially related to transfer pricing.
One possible resolution of this conflict is to
seek alternative sources or customers in the
market. However, from the tirm’s perspective,
market alternatives are less attractive for the
relief or control of disagreements and conflict
between subunits when asset specificity is high.
The most critical issue in managing the tran-
fer process is the degree to which the Iirm acts
to constrain the actions of the subunits in mak-
ing sourcing and selling decisions. Where adap-
tation and coordination of the transfer are
important because of transaction-specific
investments made by the internal selling unit
in an uncertain environment, corporate man-
agers are most likely to protect the firm’s eco-
nomic interests by asserting control over the
transfer, potentially to the point of mandating
the sourcing decision.5 Also, as discussed
above, buyers may also make transaction-speci-
fic investments where the value is dependent
on the conduct and performance of a particular
supplier. Consequently, it follows that when a
buying unit’s level of transaction-specific invest-
ment in the transfer is high, corporate managers
are most likely to protect the firm’s economic
interests by asserting control over the transfer
by constraining the selling decision. However,
when the level of transaction-specific invest-
ment made by a buying unit is low, there will
be few (if any) reasons associated with the need
to protect transaction-specific investments for
constraining the selling decision.6
’ Spicer (1988) suggests that other actions to protect the firm’s interests might Include (1) involving corporate manage-
ment in the arbitration of disputes between the subunits, (2) adjusting performance and Incentive systems to emphasize
performance measutement and incentives that are tied to higher level or joint performance and de-emphasizing perfor-
mance measurements and incentives that focus only on individual subunit performance. However, we conjecture that the
importance of these actions is likely to be much less than the decision to control the transfer process by constraining
sourcing and selling decisions.
6 There may be other factors which are not part of our theory which can influence the extent to which the selling unit is
constrained from selling outside the firm. Some of these other factors observed in our case studies are reported later in the
paper.
428 G. J. COLBERT and B. H. SPICER
Therefore, in our case studies we should
observe that the degree to which Iirms con-
strain subunit managers in sourcing and selling
decisions is, in general, positively related to the
degree of transaction-specific investment (asset
specificity) made by their counterpart(s) to the
internal transfer.
Strategi c sourci ng consi derati ons
Walker (1988) provides a related perspective
by applying the concepts of Porter’s (1980,
1985) competitive strategy framework to
explore the strategic aspects of transaction
costs inherent in the sourcing decision. In Por-
ter’s framework, firms pursue generic strate-
gies such as cost leadership, product
differentiation and focus which are sources of
competitive advantage. Walker assumes that
firms try to order inputs and operations in
terms of their value in contributing to compe-
titive advantage. For example, a supplier may
supply a custom component that provides
important differentiation features in a buyer’s
end product.
Walker’s Insight is that the risk associated
with sourcing relationships can have implica-
tions for the value of the firm’s competitive
strategy. His general argument is that inputs
that are important to the firm’s strategy involve
greatest risk because supply failure can lead to
greater decline in the firm’s performance than
failures with respect to less strategic inputs.
Walker identifies two types of sourcing risk
with strategic implications.
Appropri ati on ri sk arises when a supplier
owns and sells a factor of intrinsic value to
the firm’s strategy or the Iirm has strategically
important assets that are dependent on (i.e.
specialized to) aspects of the supplier’s goods
or services. Firms with suppliers of strategically
important components and high switching
costs are exposed to the risk of the supplier
raising prices or failing to deliver components
to specifications thereby reducing the value of
the firm’s competitive strategy. Appropriation
risk can also arise from dependence on the
supplier’s capabilities necessary for implement-
ing the firm’s strategy, for example the speed
with which a supplier can turn around a pro
totype at the design stage may be crucial to
gaining advantage by being first to market.
Under these conditions, if the buyer has high
switching costs, the supplier can opportunisti-
cally raise prices or fail to perform adequately,
thereby reducing the value of the lirm’s
strategy.
Di ffuston ri sk arises when innovative pro-
prietary products or process technology that
creates competitive advantage can be repli-
cated or Imitated by competitors. In the design
and production of some products suppliers
must have access to information about proprie-
tary processes or designs. A supplier external
to the firm with important knowledge of the
firm’s important processes and designs may
leak this imformation to the firm’s competitors.
To manage or control strategic sourcing risk
the firm may decide to Internalize the transac-
tions by investing in facilities to produce the
component. Walker concludes that vertical
integration will vary directly with the level of
strategic sourcing risk and the qualifications of
the Iii-m’s Internal source relative to those of
the best outside supplier.
However, the strategic implications of sour-
cing risk are also applicable to internal transfers
in a highly decentralized and divisionalized
organization.7 An internal supplier’s actions
can adversely affect the value of the competi-
tive strategy of end product divisions. For
example, by failing to meet contracted deliv-
ery schedules on components an internal sup-
plier may cause an end product division to miss
a market window. Similarly, an end product
division may switch to an external supplier
adversely affecting the value of investment in
component operations. These are examples of
appropriation risk. By selling proprietary com-
’ Walker’s analysis suggests that internal transfers will consist primarily of components where the qualification of the
internal source relative to the best external source and the strategic sourcing risk are high.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
429
ponents externally, internal suppliers may pro-
vide competitors with access to processes,
designs and components that are strategically
important to end product divisions. This is an
illustration of diffusion risk.
Therefore, in our case studies we should
observe that the degree to which firms con-
strain subunit managers in sourcing and selling
decisions is positively related to the strategic
importance of the component and/or the com-
ponent supplier’s capabilities to the firm’s com-
petitive strategy. Constraints on external
sourcing and selling of proprietary compo-
nents will serve to protect proprietary technol-
ogy and large specialized investments in
strategically Important transfers. In this respect
the issues of strategic importance and asset
specificity are closely bound together.
Transfer pri ce method
How transfer prices are set can have impor-
tant consequences for the tirm as a whole and
its subunits through their effects on local deci-
sion making and performance measurement.
When the level of transaction-specific invest-
ment required to support the transfer is low,
intermediate products will have a low degree
of customization and are most likely to be simi-
lar to products available in the market. As Spi-
cer (1988) points out there is no transfer
pricing problem in this case. The firm’s inter-
ests will be best served by using external mar-
ket referents as the primary basis for
determining internal terms of trade and to
inform decision making. Also, subunit man-
agers are likely to accept market prices as a
suitable measure of the value of the internal
transfer and for performance measurement.
Transfers will typically be negotiated by divi-
sional managers at transfer prices close to mar-
ket prices with adjustments confined to such
things as volume discounts, differences in
terms of sale and delivery, and the costs of
customization. For the most nart the firm’s
interests are protected by competition
amongst internal and external suppliers.
As the level of transaction-specific invest-
ment in the transfer by the component divi-
sion increases, intermediate components will
become increasingly customized and idiosyn-
cratic to the end product division. Conse-
quently,
market prices will become
increasingly less relevant for informing the
transfer pricing process and, as noted above,
corporate managers are more likely to impose
constraints on sourcing decisions by end pro-
duct divisions. In some mid-range of transac-
tion-specific investment by the component
division, it is likely that both costs and market
referents in the form of competitive bids and/
or quotations will be used to set internal terms
of trade.
However, at higher levels of asset specificity
in the component division, which results in a
considerable degree of customization (idiosyn-
crasy) in the intermediate product, internal
manufacturing costs will become the primary
basis for setting transfer prices. As noted
above, there will also be strong constraints
placed on the sourcing decision. In these cir-
cumstances the idiosyncratic nature of the
component division’s production process and
the intermediate product being transferred will
make it extremely difficult for managers to find
prices in external markets which are particu-
larly useful for informing the transfer price set-
ting process. With the presence of specialized
assets, any market prices that can be obtained
are likely to be only crude approximations of
the value of the internally transferred compo-
nent. Forcing the use of external prices as the
primary basis for setting internal terms of trade
in the presence of specialized assets may be
disruptive and counterproductive.
This same general point is made strongly by
Ball (1989, pp. 19-20) based on his analysis of
lirms as specialists in minimizing contracting
costs. He writes:
By construction, the more specific the firm’s factors of
production, the less its internal contracts wlII resemble
contracts observed in external markets and the less it
will find market prices to be informative. For highly
specific factors, quasi-prices [which include transfer
prices] will bear little resemblance to any observa-
ble external market prices. Thus, market prices can
provide useful, but not sufficient, information for deter-
430
G. J. COLBERT and B. H. SPICER
mining quasi-prices, and their usefulness varies as an
inverse function of factor specificity.
While asset specificity may not be the only
factor which affects the setting of transfer
prices in divisionalized firms, our theory leads
us to expect that it will exert a systematic and
powerful in8uence.s Therefore, in our case stu-
dies we expect to observe that the weight
given to manufacturing costs in setting transfer
prices is positively associated with the extent
of asset specificity related to the transfer in the
component division.
MULTI-CASE STUDY DESIGN
To investigate the case questions and the
theoretical propositions developed above we
use a multi-case research design with multiple
(embedded) units to collect evidence. This
multi-case design is the most sophisticated of
the four case designs discussed by Yin (1989,
pp. 46-60).9 Yin (1989) was selected because
this book is a standard reference for case
research design and provides the clearest,
most coherent and comprehensive approach
to designing, carrying out and presenting
mutli-case explanatory case research (as
opposed to descriptive or exploratory case stu-
dies) that we were able to find. However, we
also found work by Bisenhardt (1989), McKin-
non (1988) Bruns & Kaplan (1987) Merchant
(1987) Maher (1987), Palepu (1987) Herten-
stem (1987) and Merchant & Manzoni (1989)
helpful to our thinking about how to approach,
design and write up the results of our study.
We chose to use a case research design
because this method is appropriate to the man-
ner in which we have framed the research
questions and our desire to understand which
factors are important in the design and imple-
mentation of transfer processes and the setting
of transfer prices In specific organizational con-
texts. As our theory suggests that the proper
level of analysis is the transfers which occur
within each firm, a number of selected trans-
fers of intermediate products (components)
were studied in each ‘of four large, vertically
integrated high-technology companies in the
electronics industry in the U.S.A.
The firms are treated as the cases, the
embedded units are the divisions between
which the specific transfers take place. Using
a multi-case research method makes it possible
to study particular transfers within each of
these firms.
Case sel ecti on
We targeted firms in the electronics industry
because we believed different degrees of asset
specificity would characterize h-ma-firm trans
fers in this industry. As discussed above, we
expected this factor to be a principal determi-
nant of how these firms manage their internal
transfer pricing processes.
Our analysis of internal transfers in the tar-
geted Iirms focuses primarily upon asset speci-
ficity and related strategic considerations. We
ensure that uncertainty is present by selecting
our cases from an industry which is acknowl-
edged to be volatile and continuously chan-
ging. In addition, within each selected firm,
we focus on significant transfers from large
component divisions. By definition these
s Other Erm-specific factors affecting the setting of transfer prices observed in the course of our cast studies are reported
later in the paper.
9 Types of case research designs arc (1) singlecase design with a single unit of analysis, (2) multiplecase design with a
single unit of analysis, (3) singlecase design with multiple (embedded) units of analysis, and (4) a multiple-case design with
multiple (embedded) units of analysis. Recent case research in management accounting has mostly used either designs of
type 1 or type 2. We are not aware of other contemporary case research in management accounting that uses a type 4
design. The closest is a field study by Merchant & Manzooi (1989) which is conducted at the level of profits centres in
multiple firms (cases). However, their study does not follow case research methods which requires that the data for each
embedded unit (profit centre) be first analyzed for each firm before patterns are analysed and compared across cases.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 431
Involve each firm’s largest component transfers
which are big enough to attract management
attention. However, throughout the cases we
also discuss some transfers of lesser extent.
ponent units making these transfers vary in
terms of their strategic roles and the degree
of asset specificity involved.
General descriptive information about the
cases selected for study and the nature of
embedded units (i.e. the types of transfers
from component divisions studied) are given
in Table 1. The identity of each of the four firms
has been disguised but it can be seen from the
data in the table that they are all large, highly
divisionalized firms in the electronics industry.
Data col l ecti on
The attribute common to these firms is that
each transfers electronic components between
component and end product divisions. The
types of component transfers we focus on are
transfers shown at the bottom of Table 1.
These include the most important and largest
transfers in these companies. It can be seen
that there are overlaps between the types of
transfer studied in each case. This was done
deliberately to provide us with a strong basis
for comparison. Our focus is on the transfers of
similar types of components in each case, i.e.
integrated circuits and circuit boards. The com-
Gai ni ng high level access to the companies
that agreed to participate in this study required
considerable time and effort. We first sent let-
ters of inquiry to corporate level finance or
accounting managers. We followed up our let-
ters with numerous telephone calls to the com-
panies involved.” Our objective was to gain
access to at least four large, vertically inte-
grated, divisionalized, high-technology compa-
nies. The number of cases was determined by
cost, the desire to provide the potential for
replication and, most important of all in case
research, the ability to obtain access.
Not all the firms we approached agreed to
participate. We initially approached nine firms
which we believed would meet our criteria.
One firm was dropped from further considera-
tion when we discovered that it contracted out
most of its manufacturing. Of the eight compa-
nies remaining, one company wished us to
TABLE 1. Descriptive data on cases and embedded units (component divisions)
General
Primary end products
Sales (in billions)
Number of product
divisions
High-Tech Inc.
Instruments
$1-5
> 20
Continental Southern Defense and
World-Wide Inc. Communications Aerospace
Test, measurement and Telecommunications Defence and aerospace
computing systems systems
$10-15 $5-10 $10-15
> 20 > 10 > 20
Embedded uni ts (component di vi si ons)
Component transfer
Integrated circuits or Circuit Technology (CT) Custom Electronic Integrated Circuit (IC) Electronic Circuit (EC)
hybrids division and Packaging Circuit (CEC) division division division
St Interconnect (P&l)
division
circuit boards Circuit Board (CB) Printed Circuit Board Printed Circuit @CD)
division (PCB) division division
lo Many telephone calls to each company were necessary to reach the appropriate corporate level executive. The primary
difficulty we encountered was getting past secretaries and administrative assistants (gate-keepers) who protected their
bosses from what they considered to be unnecessary intrusions. With some companies we logged over 25 telephone calls
before we were able to reach the appropriate executive. In some cases we only started to make progess when we dis-
covered that many top level executives answer their own telephones early in the morning before their secretaries arrive.
432
G. J. COLBERT and B. H. SPICER
defer access until some unspecified time in the
future, and three others decided not to partici-
pate for various reasons.
The numbers of sites and interviews involved
in the course of the research are shown in
Table 2. Our investigation required consider-
able cooperation from participants and was
financially costly. It took us to 21 selling and
buying divisions spread over seven states dis-
persed widely across the U.S.A. Because of
scheduling difficulties this required 15 sepa-
rate interstate trips over an 18 month period.
Interviews were conducted with accounting
and functional managers in these divisions.
Forty-six managers were interviewed in the
course of the study with some individuals
being interviewed on more than one occasion
when we judged there was something more to
be learnt. Interviews ranged between ti and 3
hours in length. A tape recorder was used in
almost all of the interviews and transcripts
were prepared for later analysis.” In some
cases follow-up discussions took place by tele-
phone to clarify gaps in our understanding that
were discovered during the analysis phase. We
also toured manufacturing facilities.
Data collection took place in two phases.
Initial interviews took place with corporate
and/or group level accounting personnel who
were generally familiar with the management
of the transfer pricing process and ,transfer pri-
cing within each of the selected firms. In these
interviews we did two things. First, we gath-
ered general information about the firm’s orga-
nization structure, its performance
measurement system, and its sourcing and
transfer pricing policies. Second, and most
important, we identified particular component
transfers to study and identified the buying and
selling divisions involved. Those selected are
shown in the bottom half of Table 1.
The second phase of interviewing involved
visits to selling and buying divisions on either
side of the selected transfers. The theory of the
transfer pricing process discussed above deter-
mined the data to be collected. To improve the
reliability and comparability of the data col-
lected, a structured questionnaire was used to
guide our interviews and was filled out during
the interview by each interviewee. However,
during the interviews we queried informants
for detail and asked them to comment on mat-
ters they deemed to be important but which
were not adequately captured in our question-
naire. It also allowed us to ask informants to
compare and contrast the particular transfer
under study with others they were knowledge-
able about. At the divisional level we focused
on interviewing divisional controllers and divi-
sional managers.
The information gathered during these inter-
views included the descriptive data about the
subunit, dimensions or characteristics of the
transfer(s) under investigation, the sourcing
Firm
TABLE 2. Number of sites and interviews
Corporate Selling divisions Buying divisions
Number of Number of Number of
interviews Number of sites interviews Number of sites interviews
High-Tech
World-Wide
Continental
Communications
Southern Defense
and Aerospace
2 3 8 3 5
2 2 5 5 9
1 2 2 3 3
1 1 4 2 4
Total 6 8 19 13 21
I1 One lirm would not allow us to tape interviews for security reasons. In this instance we were restricted to note taking.
TRANSFER PRICING PROCESS
and transfer pricing process as it related to the
transfer(s) under study, and the performance
measurement system. The method we used to
obtain information on the key construct of
asset specificity associated with internal trans-
fers is discussed in the Appendix.
Because of the heavy cost of data collection,
the most extensive interviewing took place in
those companies with divisions which were
the least expensive for us to access. These com-
panies were High-Tech and World-Wide. We
started our study at High-Tech. We visited this
company many times over a 6 month period.
This helped us to refine our questions and
approach in order to obtain comparable infor-
mation from the other companies in the study
with divisions which were more widely scat-
tered, further away and more expensive to
visit. Consquently, we conducted fewer inter-
views and spent less time at Continental Com-
munications and Southern Defense and
Aerospace than we did at High-Tech and
World-Wide.
Anal ysi s and i nterpretati on
Analysis and interpretation of our data were
conducted in two stages. First, the data col-
lected for specific transfers within each of the
firms were analysed and interpreted within the
context of that firm. For each individual case
our analysis concentrates on how and why the
pattern of observations and information about
each case is consistent or inconsistent with our
theory. The mode of analysis is referred to by
Yin (1989, p. 109) as pattern matching. The
importance of context in case research
requires that the data be analysed as a pattern
relative to theory, in this instance our theory of
the transfer process. For example, where do
hypothesized relationships line up with the
theory and where do they not? What other
THEORY - INVESTIGATION
433
factors appear to bear on the theoretical rela-
tionship? It is important to note that it was our
theory which influenced not only the data to
be collected but also provided the template
against which the data were analysed. l2
Having analysed the data on individual trans
fers within the context of the firm involved, we
then focused on a cross-case analysis and com-
parison of results. Results from each case were
considered to be findings that were subject to
replication by the other individual cases. If the
findings obtained in subsequent cases also were
consistent with the theory, we concluded that
the viability of the theory was strengthened.
Thus, if the pattern observed in the second
case was similar to the pattern observed in the
first case, we would interpret this as a replica-
tion. If the patterns were not similar we tried to
understand why the patterns observed did not
match those predicted by the theory. Our data
collection efforts were designed to provide suf-
ficient additional information about the tirm,
the divisions and the selected transfers to allow
us to take this additional step.
DESCRIPTION OF CASES
In this section we flesh out the descriptions
of our four cases and embedded units identified
in Table 1. We focus on providing some rele-
vant background about each company and
information about the independent variables;
that is the degree of transaction-specific invest-
ment by each component (selling) and end
product (buying) division, and the strategic
role of each of the component units.
This information is summarized in Table 3
for each firm and component division we stu-
died. We also outline each lirm’s decentraliza-
tion policy as it relates to the divisions involved
” Our research method included Yin’s (1989) recommended methods for dealing with issues involving construct validity,
internal validity, external validity, and reliability in the design of our data collection efforts. Construct validity was
addressed through use of multiple sources of evidence, internal validity through use of a pattern matching approach,
external validity through the use of replication logic (using multiple cases), and reliability through the use of question-
naires and interview transcripts.
C
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
’
s
r
o
l
e
i
n
t
i
r
m
’
s
c
o
m
p
e
t
i
t
i
v
e
s
t
r
a
t
e
g
y
T
A
B
L
E
3
.
S
u
m
m
a
r
y
o
f
i
n
f
o
r
m
a
t
i
o
n
o
f
i
n
t
e
r
e
s
t
b
y
e
m
b
e
d
d
e
d
u
n
i
t
s
(
d
i
v
i
s
i
o
n
s
)
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
t
t
e
d
w
e
i
g
h
t
R
e
p
o
r
t
e
d
l
e
v
e
l
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
r
t
e
d
l
e
v
e
l
o
f
o
n
b
u
y
i
n
g
d
i
v
i
s
i
o
n
’
s
g
i
v
e
n
t
o
o
f
a
s
s
e
t
o
n
s
e
l
l
i
n
g
d
i
v
i
s
i
o
n
’
s
O
r
g
a
n
i
z
a
t
i
o
n
a
s
s
e
t
s
p
e
c
i
f
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
b
u
y
m
a
n
t
l
t
a
c
t
u
r
h
l
g
c
o
s
t
s
p
e
c
i
t
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
s
e
l
l
s
t
m
c
t
u
r
e
c
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
e
x
t
e
r
n
a
l
l
y
i
n
t
r
a
n
s
f
e
r
p
r
i
c
e
b
u
y
i
n
g
d
i
v
i
s
i
o
n
e
n
m
r
m
n
f
y
P
a
n
e
l
A
:
H
i
g
b
T
e
c
b
C
i
r
c
u
i
t
T
e
c
h
n
o
l
o
g
y
S
o
u
r
c
e
o
f
a
d
v
a
n
c
e
d
(
C
T
)
a
n
d
P
a
c
k
a
g
i
n
g
p
r
o
p
r
i
e
t
a
r
y
&
I
n
t
e
r
c
o
n
n
e
c
t
t
e
c
h
n
o
l
o
g
y
c
r
i
t
i
c
a
l
(
P
&
f
)
d
i
v
i
s
i
o
n
s
t
o
t
h
e
e
n
d
p
r
o
d
u
c
t
d
i
v
i
s
i
o
n
s
C
i
r
c
u
i
t
B
o
a
r
d
(
C
B
)
F
i
r
s
t
,
a
s
a
s
o
u
r
c
e
o
f
d
i
v
i
s
i
o
n
e
x
t
e
r
n
a
l
p
r
o
f
i
t
s
;
s
e
c
o
n
d
,
s
u
p
p
o
r
t
s
s
t
r
a
t
e
g
y
b
y
p
r
o
v
i
d
i
n
g
s
u
p
e
r
i
o
r
s
e
r
v
i
c
e
P
a
n
e
l
B
:
W
o
r
l
d
-
w
i
d
e
P
C
w
i
t
h
B
E
o
b
j
e
c
t
i
v
e
o
n
i
n
t
e
r
n
a
l
s
a
f
e
s
.
P
C
o
n
e
x
t
e
r
n
a
l
s
a
l
e
s
’
P
C
o
n
i
n
t
e
r
n
a
l
a
n
d
e
x
t
e
r
n
a
l
s
a
f
e
s
S
i
g
n
i
f
i
c
a
n
t
f
o
r
S
i
g
n
f
k
a
n
t
f
o
r
t
h
e
S
i
g
n
i
k
a
n
t
,
c
o
s
t
-
s
i
g
n
i
f
i
c
a
n
t
S
i
g
n
i
f
i
c
a
n
t
f
o
r
C
T
t
r
a
n
s
f
e
r
s
t
o
l
a
r
g
e
s
t
l
a
r
g
e
s
t
i
n
t
e
r
n
a
l
b
a
s
e
d
w
h
e
n
i
t
i
n
v
o
l
v
e
s
i
n
t
e
r
n
a
l
c
u
s
t
o
m
e
r
,
c
u
s
t
o
m
e
r
,
m
i
n
o
r
n
e
w
t
e
c
h
n
o
l
o
g
y
,
m
i
n
o
r
o
t
h
e
r
w
i
s
e
o
t
h
e
r
w
i
s
e
s
i
g
n
i
f
i
c
a
n
t
f
o
r
P
&
f
t
o
~
a
s
s
u
r
e
s
e
r
v
i
c
e
;
M
i
n
o
r
N
o
n
e
M
i
n
o
r
,
m
a
r
k
e
t
-
b
a
s
e
d
M
i
n
o
r
N
o
n
e
n
9
C
u
s
t
o
m
E
l
e
c
t
r
o
n
i
c
S
e
c
u
r
e
s
o
u
r
c
e
o
f
P
C
w
i
t
h
B
E
S
i
g
n
i
f
i
c
a
n
t
f
o
r
i
t
s
S
i
g
n
i
t
k
a
n
t
w
h
e
n
S
i
g
n
i
k
a
n
t
,
c
o
s
t
-
C
i
m
u
i
t
(
C
E
C
)
i
m
p
o
r
t
a
n
t
o
b
j
e
c
t
i
v
e
o
n
i
n
t
e
r
n
a
l
l
a
r
g
e
s
t
t
r
a
n
s
f
e
r
s
c
a
p
a
c
i
t
y
u
t
i
l
i
z
a
t
i
o
n
b
a
s
e
d
d
i
v
i
s
i
o
n
t
e
c
h
n
o
f
o
g
y
t
h
a
t
s
a
l
e
s
,
P
C
o
n
e
x
t
e
r
n
a
l
i
s
a
n
i
s
s
u
e
p
r
o
v
i
d
e
s
s
a
f
e
s
d
i
f
f
e
r
e
n
t
i
a
t
i
o
n
t
o
e
n
d
p
r
o
d
u
c
t
s
M
i
n
o
r
P
r
i
n
t
e
d
C
i
r
c
u
i
t
B
o
a
r
d
(
P
C
B
)
d
i
v
i
s
i
o
n
S
u
p
p
o
r
t
s
s
t
r
a
t
e
g
y
b
y
p
r
o
v
i
d
i
n
g
s
u
p
e
r
i
o
r
s
e
r
v
i
c
e
P
C
w
i
t
h
B
E
o
b
j
e
c
t
i
v
e
M
o
d
e
r
a
t
e
f
o
r
i
t
s
l
a
r
g
e
s
t
t
r
a
n
s
f
e
r
s
,
m
i
n
o
r
o
t
h
e
r
w
i
s
e
M
i
n
o
r
S
i
g
n
i
l
i
c
a
n
t
,
c
o
s
t
-
b
a
s
e
d
M
i
n
o
r
v
e
r
y
s
i
g
n
i
t
i
c
a
n
t
T
A
B
L
E
3
.
C
o
n
t
.
C
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
’
s
r
o
l
e
i
n
h
r
m
’
s
c
o
m
p
e
t
i
t
i
v
e
s
t
r
a
t
e
g
y
O
r
g
a
n
i
z
a
t
i
o
n
s
t
r
u
c
t
u
r
e
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
r
t
e
d
w
e
i
g
h
t
R
e
p
o
r
t
e
d
l
e
v
e
l
R
e
p
o
r
t
e
d
c
o
n
s
t
r
a
i
n
t
R
e
p
o
t
t
e
d
l
e
v
e
l
o
f
o
n
b
u
y
i
r
q
g
d
i
v
i
s
i
o
n
’
s
g
i
v
e
n
t
o
o
f
a
s
s
e
t
o
n
s
c
l
l
i
a
g
d
i
v
i
s
i
o
n
’
s
4
a
s
s
e
t
s
p
e
c
i
f
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
b
u
y
m
a
n
t
l
k
N
r
i
n
g
c
o
s
t
s
p
e
c
i
f
i
c
i
t
y
b
y
d
e
c
i
s
i
o
n
t
o
s
e
l
l
c
o
m
p
o
n
e
n
t
d
i
v
i
s
i
o
n
e
x
t
e
m
a
R
y
i
n
t
r
a
n
s
f
e
r
p
r
i
c
e
b
u
y
i
n
g
d
i
v
i
s
i
o
n
e
x
t
e
t
n
a
R
Y
5
5
P
a
n
e
l
C
:
C
O
t
l
t
i
?
W
?
l
k
d
C
o
m
m
u
n
i
c
a
t
i
o
n
s
I
n
t
e
g
r
a
t
e
d
C
i
r
c
u
i
t
S
e
c
u
r
e
s
o
u
r
c
e
o
f
P
C
e
x
p
e
c
t
e
d
t
o
B
E
s
i
@
l
i
t
i
c
a
n
t
S
i
g
n
i
t
i
c
a
n
t
o
n
h
i
&
S
&
n
i
f
i
c
a
n
t
,
b
u
t
M
i
n
o
r
V
e
r
y
s
i
g
n
i
f
i
c
a
n
t
g
;
8
M
i
n
o
r
V
e
r
y
s
&
n
i
B
c
a
n
t
!
!
2
8
2
I
(
I
C
)
d
i
v
i
s
i
o
n
P
r
o
P
d
e
t
a
r
y
o
r
m
a
k
e
a
m
o
d
e
s
t
v
o
l
u
m
e
t
r
a
n
s
f
e
r
s
t
e
c
h
n
o
l
o
g
y
p
r
o
f
i
t
p
r
i
n
t
e
d
C
i
r
c
u
i
t
S
u
p
p
o
r
t
s
s
t
r
a
t
e
g
y
b
y
P
C
e
x
p
e
c
t
e
d
t
o
B
E
M
i
n
o
r
M
i
n
o
r
@
C
D
)
d
i
v
i
s
i
o
n
p
r
o
v
i
d
i
n
g
s
u
p
e
r
i
o
r
o
r
m
a
k
e
a
m
o
d
e
s
t
s
e
r
v
i
c
e
p
r
o
f
i
t
P
a
n
e
l
D
:
S
o
u
t
b
e
m
D
e
f
e
n
c
e
a
t
k
i
A
-
p
a
c
e
E
k
c
t
r
o
n
i
c
c
i
r
c
u
i
t
s
S
e
c
u
r
e
s
o
u
r
c
e
o
f
P
C
w
i
t
h
a
B
E
S
i
g
n
i
f
i
c
a
n
t
f
o
r
S
i
g
n
i
f
i
c
a
n
t
f
o
r
s
i
g
r
i
t
i
c
a
n
t
w
e
i
g
h
t
i
s
a
l
s
o
g
i
v
e
n
t
o
m
a
r
k
e
t
M
i
n
o
r
,
m
a
r
k
e
t
-
b
a
s
e
d
s
i
g
n
i
6
c
a
n
t
,
c
o
s
t
-
@
C
)
d
i
v
i
s
i
o
n
2
d
V
Z
U
l
C
C
d
o
b
j
e
c
t
i
v
e
o
n
i
n
t
e
r
n
a
l
h
y
b
r
i
d
s
h
y
b
r
i
d
s
,
m
o
d
e
r
a
t
e
b
a
s
e
d
t
c
c
b
n
o
l
o
g
y
n
o
t
s
a
l
e
s
,
P
C
o
n
e
x
t
e
r
n
a
l
f
o
r
i
n
t
e
g
r
a
t
e
d
r
e
a
d
i
l
y
a
v
a
i
l
a
b
l
e
s
a
l
e
s
c
i
r
c
u
i
t
s
e
-
-
W
l
P
C
m
e
a
n
s
p
r
o
f
i
t
c
e
n
t
r
e
a
n
d
B
E
m
e
a
n
s
b
r
e
a
k
e
v
e
n
.
S
i
@
k
a
n
t
f
o
r
M
o
d
e
r
a
t
e
3
t
h
o
s
e
w
i
t
h
h
i
g
h
9
t
e
c
h
n
o
l
o
g
y
i
j
n
e
e
d
s
5
i
j
2
436 G. J. COLBERT and B. H. SPICER
under the heading organization structure. Our
findings are then discussed in the following
section.
High-Tech I nc.
High-Tech Inc. is a large, divisionalized com-
pany that makes a wide variety of sophisticated
electronic products including display and mea-
surement devices. The company has over 20
divisions that are organized into a small num-
ber of product groupings. Component divisions
are located in the different end product groups.
tical integration strategy. First, commodity type
components (e.g. plastics and metal fabrica-
tion) which provide its end products little or
no differentiation advantage became readily
available from external suppliers at lower
cost. Second, expected growth for end pro
ducts failed to materialize, creating serious
excess capacity problems in its important elec-
tronic component operations.
Since it started up in the 194Os, the compa-
ny’s corporate strategy has been to be a differ-
entiated producer of high-technology, high-
performance, state-of-the-art electronic pro-
ducts; a producer that sets the standards for
the rest of the industry. One divisional control-
ler commented:
Our corporate strategy is to produce products that offer
cost-effective proprietary performance. We have to be
cost-effective, but products have to have a proprietary
performance because we are a technology driven com-
pany-we are not a low cost manufacturer. It is proprie-
tary technology that we need.
These problems forced High-Tech to change.
At the time we conducted our study High-Tech
was attempting to build a stronger market and
customer focus and to improve the company’s
rate of return. As part of this change process top
management divested some facilities producing
commodity type components, reorganized and
made related changes to sourcing and transfer
pricing policies. Some of these changes are dis-
cussed later in the paper. Current corporate
policy is to allow divisions to negotiate over
internal transfers and related transfer prices.
In general, division and group managers are
expected to work things out without calling in
top management to arbitrate disputes.
Historically, because of the technology driven
nature and culture of the company (sometimes
described as a company of engineers produ-
cing products for engineers) High-Tech had
practised backwards vertical integration to
assure a continuous flow of components to its
end product divisions. Large scale component
fabrication and assembly facilities were built to
produce a diverse set of proprietary and non-
proprietary components to end product
divisions.
High-Tech has three primary component
divisions which vary in their strategic roles in
the company and the degree of asset specificity
and proprietary content of internal transfers.
Panel A of Table 3 summarizes information
about these roles and the reported levels of
asset specificity of these component divisions
and their customers. The following sections
describe their operations.
More recently, High-Tech had experienced
financial performance difficulties owing pri-
marily to increasing product competition and
unsuccessful forays into new product areas.
Two other factors disrupted the company’s ver-
Circuit Technology (CD and Packaging C
I nterconnect (PM) divisions. The CT division
produces a wide range of customized inte-
grated circuits while the P&I division produces
many custom circuit modules called hybrids.i3
Because they share many common attributes
these two divisions are discussed together.
There is a consensus among the managers
we interviewed that these two divisions are
._
” Hybrids have been described as “circuits in a can”. They are fully tested circuit modules which provide a way of dealing
with problems of limited space, complex interconnects and heat build-up in electronic circuits. A single hybrid can
functionally replace a fully loaded circuit board in a compact module form. P&I hybrids incorporate custom integrated
circuits supplied by the CT division.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
437
critical to the company’s competitive strategy
because they supply the proprietary technol-
ogy that imparts “cutting edge” performance
and features for the company’s end products.
A significant portion of each division’s sales are
internal although both have acted recently to
build an external customer base as well. How-
ever, their external business is seen primarily
as a way to increase capacity utilization rather
than as a separate source of profits per se. Both
divisions report moderate amounts of unused
capacity. Both divisions are evaluated as profit
centres with a break-even objective on internal
transfers and a positive profit objective for
external sales.
Both divisions report specialized investments
of two types. The first are investments specia-
lized to the production of a type of component,
i.e. integrated circuits or hybrids. These consist
primarily of equipment and training of person-
nel. The fabrication processes employ certain
machines that are very expensive and unique.
Although these machines were purchased with
the primary intent of supporting internal
demand, they can be used to produce a range
of different components. Both CT’s and P&I’s
managers report that there would be moderate
difficulty in redeploying these assets to serve
new external or internal business. Because of
relatively long lead times, curtailment of exist-
ing production owing to lost internal business
would result in lost revenue while capacity is
backfilled with new business.
The second type of specialized investment is
in those assets which are specialized to the
production of particular customized compo-
nents. These investments consist primarily of
engineering design work and software used to
run automated equipment to fabricate and test
each particular component. For example, the
P&I division’s controller comments:
There is no portability with respect to a lot of invest-
ment that we have made here. It is specific to a custom
application. In the custom hybrid business a significant
amount of your investment is your engineering assets
and your process development assets to get a custom
product going. If we were to replace this internal busi-
ness we would have to reincur all that cost over again.
Internal transfers to one end product division
account for over one-half of the total volume of
both the CT and P&I divisions. For each com-
ponent division, a significant level of asset spe-
cificity was reported to be present for these
internal transfers. Both of these component
divisions report that if transfers to this one
end product division were halted there would
be at least a moderate decline in the value of
the investments in tangible assets and person-
nel specialized to the production of compo-
nent types, i.e. integrated circuits and
hybrids, and a significant decline in the value
of investments in design work and training
which are specialized to unique components.
Lower volume transfers involve lower levels of
specialized investments and a minor decline in
investment values if these transfers were
halted.
End product divisions, which are internal
customers of these two buying divisions, also
report a significant level of investment specia-
lized to the components sources from these
two divisions. These consist primarily of invest-
ments in design work and training of person-
nel. These internal customers report that there
would be a significant decline in the value of
these investments if the internal transfers were
halted and they were forced to switch to exter-
nal supply. They attribute this to the significant
difficulty in buying these advanced compo-
nents externally, particularly when the indivi-
dual volumes of specific components are
relatively low. There was also some doubt
expressed about whether they would receive
the same level of service, for example respon-
sive engineering support, from an external ven-
dor.
Ci rcui t Board (CB) di vi si on. Although the
CB division manufactures a diverse range of
custom circuit boards, it specializes in high-
end, multi-layer circuit boards. The division is
noted for delivering these types of boards
quickly and efficiently even in relatively low
volumes. This capability is valuable to external
and internal customers who have low volumes
and fast time-to-market requirements. Because
circuit boards are regarded as something of a
438 G. J. COLBERT and B. H. SPICER
commodity available from external sources,
service in the form of quick turn around, con-
sistent high quallty, reliable delivery times, etc.,
rather than proprietary technology is regarded
as the key to providing competitive advantage
to customers. The division’s controller com-
ments on the service features they offer:
All of those things [quick turn around, electronic design
transfer, close working relationships to the customer,
etc.] make it easy to do business with us as opposed
to someone else. Circuit boards are on the critical path
of time-tomarket. If you screw up and miss getting the
right board in the prototype phase you could comple-
tely miss the market window.
In marked contrast to the CT and P&I com-
ponent divisions, the CB division has a success-
ful and extensive external business and is
viewed as a direct source of profits to the com-
pany somewhat like High-Tech’s end product
divisions. Interestingly, the division views itself
as a “stand alone” business and is evaluated as
a full profit centre on both internal and exter-
nal sales. At the time the study was conducted
the division reported it was operating with
moderately constrained capacity.
The CB division reports a very significant
level of investment specialized to the fabrica-
tion of circuit boards in general, but only a
minor level of investment specialized to the
production of particular circuit boards
whether for internal or external sale. CB
reports that there would only be minor losses
involved if individual internal transfers were
halted. The reason given is that assets used to
fabricate circuit boards can readily be rede-
ployed to produce other boards. Furthermore,
they have a backlog of orders which they could
use to backlill any volume lost. The biggest
issue is the time needed to bring in a new
customer and develop tooling for that custo-
mer’s boards. Tooling in this case is analogous
to a blueprint or map that indicates how to
build a particular board and is unique to that
board.
End product divisions that are internal custo-
mers of CB report moderate to significant levels
of investment specialized to the components
sourced from CB. However, even those with
a need for high-end, high-technology boards
report that there would be only a minor
decline ln the value of investments if the lnter-
nal transfer were halted because even these
boards could be obtained from external
sources, albeit with some difficulty. The com-
ments of a controller of an end product division
that is a major customer of the CB division
reflects this point:
I could find another vendor pretty easily, and I could get
my design tapes and film work from CB. I might have to
put a little bit of effort into reformatting the design
tapes for a different process. My estimate is this would
only involve a minor decline in investment value though
I would have to scramble around a bit.
World- Wtde I nc.
World-Wide is a major manufacturer of pre-
cision electronic equipment used for measure-
ment, analysis and computation. The company
makes thousands of products that cover a
broad range. Operations are concentrated in
the U.S.A., but the company has other research
and manufacturing facilities throughout the
world.
World-Wide’s competitive strategy has a
strong marketing emphasis which is based on
understanding customer needs and then devel-
oping solutions through technological innova-
tion. At the time we conducted our study the
company had been experiencing strong
growth which was attributed to both its inno-
vative capabilities and the success of its end
product divisions in introducing new products
and new generations of existing products to
the market-place. The company’s products
and businesses are very diverse. Some busi-
nesses operate in large, very competitive mar-
kets where they compete against world-class
competitors. Other businesses operate in nar-
row markets where volumes are relatively low
and technology and performance demands are
such that customers are not as price sensitive.
Unsurprisingly, it is the former businesses
which appear to be most sensitive to transfer
pricing issues.
The company is organized into groups on the
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
439
basis of major business segments. The groups
are further decentralized Into divisions. Com-
ponent operations are primarily located in a
separate group. Historically, management’s
strategy has been to utilize its large internal
volume requirements to drive down the cost
of component supply by fully utilizing the capa-
city of its Internal component divisions. The
internal component divisions are expected to
meet the needs of a diverse set of internal cus-
tomers and the company has typically operated
its component operations as captive suppliers.
However, considerable negotiation over trans-
fers and transfer prices takes place between
component and end product divisions but dis-
putes are rarely elevated to the group or corpo-
rate level.
Business units within the groups that sell end
products are operated as profit centres (divi-
sions) that operate very autonomously. The
corporate policy and culture has been to
encourage responsibility and innovation, and
end product divisions, in particular, have
been encouraged to “run their own busi-
nesses” to the greatest extent possible. This
has caused some difficulties when divisions
must cooperate and share resources to pro-
vide Integrated solutions to customer require-
ments. Over time, World-Wide has
experimented with changes in responsibilities
and reporting relationships to Improve co-ordi-
nation.
With World-Wide we focused on two compo-
nent manufacturing units that internally trans-
fer the same or similar types of components to
those studied in the High-Tech case. Like High-
Tech these units vary in terms of their strategic
roles in the company and the characteristics or
dimensions of component transfers to product
divisions. These divisions are located within
the same group and represent the primary
source of components transferred internally
within World-Wide.
Panel B of Table 3 summarizes information
about the roles these two component divisions
play in World-Wide’s competitive strategy,
their reported levels of asset specificity, and
that of their internal customers. The following
sections describe their operations in more
detail.
Custom Electronic Circuit (CEC) division.
World-Wide’s CEC division is similar to High-
Tech’s Circuit Technology (Cl) division. CEC
manufactures custom integrated circuits. The
division is regarded as strategically important
to many of World-Wide’s end product divisions
because it serves as a secure source of custo-
mized advanced circuits that provide differen-
tiation features for the company’s end
products. The division’s origins are rooted in
the belief that at one time the only way to
have access to custom Integrated circuits was
to manufacture them internally. Internal custo-
mers now cite the division’s design capability,
availability of particular manufacturing pro-
cesses, quality, responsiveness, and security
of unique proprietary designs and technology
as advantages of doing business with CEC over
external vendors.
Although it is not intended to be a low cost
producer of components (internal customers
rate CEC as having no price advantage), to
remain economically viable CEC has had to
obtain higher volumes by producing some cir-
cuitry that is less unique and not as advanced
technologically. Like High-Tech’s CT division,
CEC has increasingly pursued external sales to
fill capacity and meet performance objectives.
CEC is evaluated as a profit centre with a break-
even objective on internal sales and a positive
profit objective on external sales. At the time
our study was conducted, CEC reported mod-
erately constrained capacity.
CEC’s managers report a very sign&ant
investment in the facility and its workforce
devoted exclusively to the production of cus-
tom integrated circuits. Although in principle
this Investment could be redeployed, CEC’s
managers report that there would be at least
a moderate decline in investment value if its
largest internal transfers were halted owing to
difficulties in quickly backfilling capacity with
external business. Most importantly, CEC’s
managers report significant investments specia-
lized to particular internal transfers. For exam-
ple, in the case of its largest internal customer,
440 G. J. COLBERT and B. H. SPICER
CEC has invested in an expensive piece of
equipment that is used only to assemble pro
ducts for that customer. In addition there are
significant investments in design work and spe-
cialized training to produce a particular compo-
nent. It is reported that there would be a
significant decline in the value of these invest-
ments if large internal transfers were halted.
Internal customers report moderate levels of
investment consisting of design work and train-
ing and minor levels of tangible assets that are
specialized to components transferred from
CEC. However, they also report that only a
minor decline in the value of these invest-
ments would occur if the internal transfer
were halted. Any decline would be due to the
cost of developing new relationships,
exchange of information about designs and fab-
rication processes, and perhaps some redesign
effort. For example, the controller of the lar-
gest customer division comments:
We had to automate a lot of things to use that [internally
transferred] integrated circuit. The dollar value is high
but the investment is fairly transferable to outside ven-
dors.
Pri nted Ci rcui t Board (PCB) di vi si on. Simi-
lar to the Circuit Board (CB) division at High-
Tech, this division also manufactures circuit
boards which are not regarded as strategically
important. However, in contrast to the High-
Tech CB division, which is operated very
much as a stand-alone business, World-Wide’s
PCB division is operated exclusively as a cap
tive supplier and is several times larger invol-
ving multiple plants. It is evaluated as a profit
centre with a break-even objective. At the time
of the study it was operating at or near capa-
city.
PCB’s strategic role is to support end product
division’s strategy by providing superior ser-
vice including flexible delivery and quick
design turnaround. Internal customers cite
design capability, quality, responsiveness,
delivery and cost as PCB’s advantages over
external suppliers. World-Wide’s product divi-
sions have very demanding needs for design
and redesign of boards since some products
have life cycles as short as a year and a half.
Attempting to stay on the front edge of the
market with each generation of product, end
product divisions make frequent changes to
board layout, etc. Thus the ability to co-ordi-
nate designs and provide quick turnaround
are crucial to obtaining “time-to-market”
advantages.
PCB reports very significant levels of invest-
ment in tangible assets and training specialized
to the production of circuit boards. The decline
in the value of these investments that would
occur if individual internal transfers were
halted is reported to vary from minor to mod-
erate depending on the size of the transfer.
Potentially, specialized assets resulting from
these investments could be switched to produ-
cing boards for external customers. However,
PCB has no external marketing capability and
to establish this would involve substantial
costs. PCB also reports a minor level of invest-
ment in design work related to specific circuit
boards as this is borne primarily by the buying
unit. Hence, there would be only a minor
decline in value if an individual transfer were
halted.
Internal customers report a significant level
of investment in design work that is specialized
to internal transfers of particular circuit boards
from PCB. They report that the decline in
investment value would be minor should the
internal transfer be halted, as the design work
could be shifted easily to an external vendor.
As one manager pointed out “the technology
isn’t that terribly unique that we couldn’t fmd
alternative sources.”
Conti nental Communi cati ons
Continental Communications is a leading
manufacturer of digital telecommunications
systems. The company’s products are focused
on a small number of applications in the tele-
communication’s market. In each of these pro-
duct lines Continental is one of the few major
suppliers in the North American market where
its operations are concentrated.
With the break up of AT&T and the dereg-
TRANSFER PRICING PROCESS
ulation of the telecommunication’s industry,
the company underwent a period of strong
growth but more recently has come under
increasing pressure from competitors vying
for market share and from customers seeking
increasingly sophisticated, higher quality pro-
ducts. At the time we conducted our study
Continental’s competitive strategy centred on
increasing customer satisfaction at the same
time as it reduced costs and increased through-
put in its manufacturing facilities.
The company is divisionalized based on its
major product lines. Component fabrication
and research and development are organized
as separate business units apart from end pro
duct operations. Historically, the company
required its end product divisions to source
components internally but more recently has
given greater latitude to end product divisions
to source some components externally. The
company has streamlined its operations
through closure and consolidation of some of
its manufacturing facilities as it shifted towards
external sourcing of some components such as
low-technology circuit boards. Virtually all
components except custom integrated circuits
and circuit boards and closely related compo-
nents are now sourced from outside vendors.
Owing to competitive pressures, end product
divisions have placed considerable pressure on
component divisions to reduce their transfer
prices.
In our study we looked at transfers from the
company’s two primary component divisions
that supply integrated circuits and circuit
boards. While internal sourcing is in principle
mandated and there is a corporate transfer pri-
cing policy of standard manufacturing cost plus
a set percentage mark-up, the transfers that we
studied are exempted from this policy as a
result of attempts by end product divisions to
reduce costs by forcing internal component
divisions to price against potential external
.‘
THEORY - INVESTIGATION 441
vendors whenever possible. Negotiation and
disagreements over transfer prices occur but
we were told that most disputes are resolved
at the divisional level.
Panel C of Table 3 summarizes information
about the strategic roles these divisions play in
Continental’s competitive strategy and the
reported levels of asset specificity of these
component divisions and their customers.
The following sections describe their opera-
tions.
Integrated Ci rcui t (rC) di vi si on. The I C
division, in contrast to other integrated circuit
plants in this study, specializes in producing
high-volume, mature, custom integrated cir-
cuits.‘* The division is evaluated as a profit
centre with a modest profit objective or at
minimum a breakeven objective.
Internal customers cite proprietary designs,
assured supply and responsiveness as K’s pri-
mary competitive advantages over external
suppliers. Neither IC nor its internal customers
regard price as one of the IC division’s compe-
titive advantages.
The IC division reports a very significant
level of investments in tangible assets, design
work and training specialized to the produc-
tion of integrated circuits. IC produces a lim-
ited number of high-volume components
which are transferred to more than one end
product division. Loss of an individual transfer
would impact on the division through a decline
in capacity utilization. Hence, IC’s managers
report that there would be a significant decline
in the value of these investments if an indivi-
dual transfer were halted. IC’s controller
explains:
This is a highly leveraged business. Because of all the
expensive specialized equipment we have a fairly high
breakeven point. If you pull volume below our break
even point then we are left with a fixed chunk of cost
without volume to cover it. That is the risk you run in
letting buyers redeploy elsewhere.
‘* The company actually has two integrated circuit facilities. The facility we studied focuses on the production of mature
components ln high volumes. The other facility focuses on prototype and advanced technology custom circuits. Transfers
from this division, which is located In the same facility as the company’s research operations, were not studied.
442 G. J. COLBERT and B. H. SPICER
Internal customers report only minor
amounts of investments specialized to internal
transfers from the IC division and that the
potential decline in investment value would
be minor if internal transfers were halted.
They believe that although there would be
some short- to medium-term difficulties their
component requirements could be met by
switching to external sources.
products and services that often involve long-
term programmes or projects.
Southern’s strategy is to be considered as a
technology leader and it is regarded as a pace-
setter in applying advanced electronics in the
defence and aerospace industry. The compa-
ny’s growth and reputation resulted from the
post World War II boom in high-technology
applications in defence and aerospace.
Pri nted Ci rcui t di vi si on (PCD). A large por-
tion of Continental’s board needs are sourced
externally. The boards produced by the PCD
division differ from those sourced externally
as they typically involve more advanced tech-
nology, are more difficult to make and are
lower ln volume.
However, PCD’s managers recognize that
their only real advantage over external suppli-
ers is not cost or quality but rather superior
service and their ability to work closely with
the end product divisions and the company’s
R&D unit to bring new boards or revisions of
old boards into production. Internal customers
concur with this assessment. PCD is evaluated
as a profit centre with a modest profit objective
or at a minimum a break-even objective.
The company is organized into a number of
product groups each of which have several divi-
sions that involve similar technology and appli-
cations. The company is operated on a highly
decentralized basis which allows each group to
respond to its segment of the electronics mar-
ket with a total capability from research to sales.
One exception to this is intracompany activity
that relates to some centralized component
technology and fabrication. Most of Southern’s
internal transfers involve transfers from the
Electronic Circuit division and it is on these
transfers that we focus.
The PCD division reports significant invest-
ments in assets specialized to circuit boards
generally but only minor investments that are
specialized to particular internal transfers. They
further report only a minor potential decline in
investment value if individual internal transfers
were halted.
Internal customers report that although they
have significant investments in design work
related to internally sourced circuit boards,
this investment could be redeployed to outside
vendors; hence, they report that there would
only be a minor decline in investment value if
internal transfers were halted. This is particu-
larly the case for the least complex boards that
are required in large volumes.
Much of the company’s external business is
conducted through prime contracts (as
opposed to subcontracts) with government enti-
ties. These contracts may span several years and
involve hundreds of millions of dollars. Both
cost-plus and fixed-cost contracts are used
with the latter being most common and increas
ing in frequency. Because of its high exposure to
declines in defence budgets, management is
attempting to decrease the company’s depen-
dence on defence-related business. Achieving
this requires changing a rather bureaucratic
organizational infrastructure that has been
long established to service defence contracts.
At the time of our study the company was begln-
ning to explore how to make these changes.
Southern Defense and Aerospace
Panel D of Table 3 summarizes information
about the strategic role the Electronic Circuit
division plays in Southern’s competitive strat-
egy and the reported levels of asset specificity
of this component division and its customers.
The following sections describe its operations.
Southern Defense and Aerospace is a major El ectroni c Ci rcui t (EC) di vfsi on. The major-
manufacturer of high-technology electronic ity of Southern’s internal transfers of compo-
systems for military, commercial and scientific nents comes from the EC division. This
use. The company produces several thousand division produces custom integrated circuits
TRANSFER PRICING PROCESS
and hybrid packaging of these circuits. The
division’s primary mission is to provide end
product divisions with a secure source of
advanced technology components that are
not readily available from external sources.
This component capability, and in particular
their hybrid technology, l5 is crucial to the com-
pany given its markets and strategy to be a
technology leader. In particular it is seen as
giving them control over the interface
between R&D and production in advanced
applications. Given that defence contracts
often include penalties for late delivery, inter-
nal customers also require an assured and reli-
able source of supply and tight security with
respect to designs.
At the time of the study the division had
significant excess integrated circuit capacity
relative to internal volume requirements. In
an attempt to offset the cost of this capacity
the division has aggressively pursued external
sales of Integrated circuits. Hybrid capacity
could be increased with some additional invest-
ment but is presently profitable with only a
minor amount of external sales. EC is evaluated
as a profit centre with a breakeven objective
on internal sales and a positive profit objective
on external sales.
The EC division reports a very sign&ant
level of investment specialized to the produc-
tion of the two component types, integrated
circuits and hybrids. For example, its fabrica-
tion areas require the highest level of “clean”
rooms in the industry and are extremely expen-
sive. The division also reports a significant level
of investment specialized to individual trans
fers. To illustrate, EC recently purchased an
expensive piece of test equipment specifically
to support one product division’s programme.
The division reports that there would be a sig-
nificant decline in both the value of invest-
._
THEORY - INVESTIGATION
443
ments in capacity and investments specialized
to individual transfers of integrated circuits and
hybrids if internal transfers were halted.
Internal customers with advanced technol-
ogy needs report a signiticant level of invest-
ments specialized to internal transfers from
EC. These would involve a significant decline
in investment values if internal transfers were
halted. For example, an end product division
indicates that it is tied by technology and
design decisions to buy certain components
from the EC division. The end product division
has designed these components for fabrication
on uncommon processes owned and operated
by the EC division.
Cfrcuft boar&. Unlike the other companies
in our study, Southern sources all of its circuit
boards externally. However, the company has
recently reconsidered this issue and has
decided to invest in an internal circuit board
facility. In addition to internal sourcing the
company plans to continue sourcing some
boards externally with a few highquality sup
pliers with whom it will develop a long-term
relationship. l6 The anticipated benefits of an
internal board supplier cited by Southern’s
management include lower cost through
economies of scale in production and adminis
nation of the supplier interface, improved qual-
ity and quicker turnaround.
CASE EVIDENCE AND FINDINGS
We initially analysed and wrote up our
results for each case before undertaking a
cross-case analysis. However, this makes for a
long and tedious presentation. Therefore, we
present our evidence and findings here by
directly assessing how well our case evidence
matches the pattern suggested by theory and
” WC were told that because of aerospace requirements EC’s hybrid capability was matched by very few suppliers ln the
world. Packaging layers for current aerospace applications are several times denser than most commercial applications.
Only a small fraction of Southern’s total hybrid requirements are readily available from external vendors.
i6 Southern’s investigation of circuit board procurement activity revealed, to management’s surprise, that product divi-
sions were sourcing boards from several hundred dlfTerent suppliers.
444 G. J. COLBERT and B. H. SPICER
the extent to which the findings are replicated
over the cases. We draw attention to unusual or
interesting issues.
Panels A to D of Table 3 present the indepen-
dent and dependent variables of interest by
embedded unit (division) for High-Tech,
World-Wide, Continental Communications,
and Southern Defense and Aerospace.
The first question we address is:
How do the business units of high-tech-
nology Rrms manage the transfer process
and why do these management practices
differ?
Case evidence pertaining to this question is
set out by case in Table 3 and a cross-case
summary of our findings is set out in Table 4.
Managing sourcing decisions when asset
specijkity is significant in the component
(selling) divisions
When component (selling) divisions have a
sign@cant level of asset specificity related to
TABLE 4. Crosscase data on sourcine and sellina decisions
Panel A: Sourcing decisions
Assets specificity reported by component (selling) divisions
Signifcunt
High-Tech
CT division
P&I division
World-Wide
CEC division
Continental Communications
IC division
Southern Defense and Aerospace
EC division
Mi nor
High-Tech
CB division
Continental Communications
PCD division
Moderate
World-Wide
PCB division
Panel B: Selling decisions
Asset specificity reported by end product @wing) division
Buying unit constrained?
Yes
Yes
Yes
Yes
Yes
No
No
No
SeUing unit constrained?
Sfgni J i cant
High-Tech
CT division
P&I division
Southern Defense and Aerospace
EC division
Mi nor
High-Tech
CB division
World-Wide
CEC division
PCB division
Continental Communications
IC division
PCD division
Yes
Yes
Yes
No
No
Yes
Yes
Yes
TRANSFER PRICING PROCESS
i nternal transfers, we expect to fi nd end pro-
duct @uyi ng) di vi si ons constrai ned i n thei r
abi l i ty to buy the component external l y i n
order to protect the transacti on-speci j k i nvest-
ments made by component di vi si ons.
From Panel A of Table 3 we see that High-
Tech’s Circuit Technology (CT) and Packaging
& Interconnect (P&I) divisions both report a
significant level of asset specificity related to
internal transfers to their largest internal custo-
mer. Consistent with the theory this end pro-
duct division reports that there are significant
constraints on its ability to source components
from an external supplier which are presently
sourced from CT and P&I. This sourcing policy
protects the company’s investments in these
strategically important areas which provide a
secure, assured source of advanced proprie-
tary technology to end product divisions. It is
also consistent with the company-wide strategy
as a differentiated supplier of high technology
state-of-the-art electronic products. ”
From Table 3 and Panel A of Table 4, we see
that the finding in the High-Tech case is
repeated for the integrated circuit facilities in
each of the other three cases, i.e. for World-
Wide’s Custom Electronic Circuit Division
(CEC) division, Continental’s Integrated Circuit
(IC) division and Southern’s Electronic Circuit
(EC) division. In each case, end product divi-
sions report a signiIicant constraint placed on
sourcing particular components externally
albeit with some varying conditions. At World-
Wide the constraint operates primarily when
capacity utilization in the component division
is an issue. Consistent with the company’s man-
agement style this constraint is not formally
imposed but rather the division is expected to
cooperate if asked at the group level.
At Continental the constraint is applied for
high-volume custom circuits in order to main-
tain high levels of utilization of an expensive IC
facility which is dedicated to fabricating
mature, high-volume circuits. Company policy
THEORY - INVESTIGATION 445
is that end product divisions will source intem-
ally unless they have an exemption. At South-
em, where there is a very large specialized
investment in the hybrid unit of the EC divi-
sion, end product (buying) divisions are con-
strained from buying outside without the
approval of the hybrid unit. An important rea-
son for this is to ensure that capacity is utilized.
Another is to centralize supplier certification.
We also observe that there are strategic moti-
vations for each tirm to invest in particular
component technology and capacity. We dis-
cuss this further in a later section.
Managi ng sourci ng deci si ons when asset
speci j ki ty i s mi nor i n the component
(sel l i ng) di vi si ons
Wben component (sel l i n@di vi si ons have a
mi nor l evel of asset speci fi ci ty rel ated to i nter-
nal transfer, we woul d not expect to fi nd end
product (buyi ng) di vi si ons constrai ned i n
thei r abi l i ty to buy the component external l y
i n order to protect mi nor transacti on-speci fi c
i nvestments made by component di vi si ons.
From Panel A of Table 3 we see that for
internal transfers from High-Tech’s CT and
P&I divisions which report minor asset specitl-
city, internal customers report only minor con-
straints on their ability to source components
externally. Similarly, we see that High-Tech’s
Circuit Board (CB) division reports only a
minor level of asset specificity related to inter-
nal transfers. Consistent with the theory, end
product divisions report that they are comple-
tely free to buy their circuit boards from exter-
nal vendors.
As can be seen from Panel C of Table 3 this
result is replicated in Continental Communica-
tion’s Printed Circuit @‘CD) division which
reports a minor level of asset specificity. End
product divisions receiving components from
PCD report only minor constraints on their abil-
ity to source related components externally.
World-Wide’s Printed Circuit Board (PCB)
I7 Also consistent with this strategy, High-Tech has over time divested facilities supplying low-technology, commodity type
components such as plastics and metals.
446 G. J. COLBERT and B. H. SPICER
division reports a moderate level of asset spe-
cificity for its largest transfers and minor other-
wise. Its internal customers, regardless of the
size of the transfer, reported only minor con-
straints on their ability to externally source
boards which were also available from PCB.
With respect to these reported results it is
important to note that our theory predicts dif-
ferent policies on sourcing constraints when
asset specificity in the component division is
either high or low. Where asset specificity is
in the mid-range, as in this instance, other fac-
tors may have more influence. In this case our
interviews lead us to believe that what we
observe is due to World-Wide’s corporate cul-
ture which, as noted in the earlier description
of World-Wide’s operations, placed a strong
emphasis on fostering autonomous divisional
authority and responsibility and minimizing
the impact of imposed corporate policies.
The fact that the PCB division has been able
to operate at or near capacity, without upper
management intervening to require internal
sourcing of circuit boards, has meant that allow-
ing buying divisions autonomy has not resulted
in serious conflict with the wider economic
interest of the firm in capacity utilization.
Managing selling decisions when asset
specificity is signfflcant in the end
product (buying) divisions
When end product (buying) divisions have
a sfgniflcant level of asset specijk&y related to
internal transfers, we expect to find compo-
nent (selling) divisions constrained in their
ability to sell components externally in order
to protect the transaction-specifk investments
made by end product divisions.
From Panel A of Table 3, we see this is the
case with respect to High-Tech’s Circuit Tech-
nology (CT) and Packaging & Interconnect
(P&I) divisions. External sales are pursued, but
only as a way to utilize capacity and not when it
jeopardizes their internal mission or their pro-
prietary technology. Both CT And P&I are con-
strained in their ability to sell advanced and
proprietary technology outside the company
in order to protect the differentiated nature of
the company’s end products. However, on less
advanced processes and technology the divi-
sions have the freedom to pursue external busi-
ness to better utilize capacity. Interestingly,
these component divisions also maintain some
reserve capacity in order to manage their rela-
tionship with the end product divisions. For
example, P&I’s constraint is driven by a policy
that the requirements of end product divisions
must be met first and foremost. Consequently,
the division maintains some reserve capacity for
internal customers even though its manage-
ment believes that it could Iill that capacity
with profitable external business.
The only other instance of buying divisions
reporting significant asset specificity relates to
Southern’s Electronic Circuit (EC) division. As
can be seen from Table 3, there are moderate
constraints on this component division’s ability
to sell externally. EC will not sell advanced tech-
nology without the permission of internal cus-
tomers who would be impacted by a unilateral
decision made by the division. Otherwise, the
EC division is allowed to sell components exter-
nally to help defray the cost of its capacity.
Managing selling decisions when asset
specijkity is minor in the end product
(buying) divisions
When end product (hying) divisions have
a minor level of asset specificity related to
internal transfers, we do not expect to fmd
component (selling) divisions constrained in
their ability to sell components externally in
order to protect the minor transaction-specijk
investments made by end product divisions.
From Panel A of Table 3, we see this is the
case with respect to High-Tech Inc.‘s Circuit
Board (CB) division. Internal customers report
minor asset specificity and the CB division
reports it is completely free to sell externally.
This result is replicated ln World-Wide’s Cus-
tom Electronic Circuit (CEC) division which is
also free to sell to outside customers even
though internal customers have hrst priority
and may, on occasion, object to the external
sale of particular components.
However, from Table 3 and Panel B of Table
TRANSFER PRICING PROCESS
4 we see the opposite result in three instances.
World-Wide’s Printed Circuit Board (PCB) divi-
sion and Continental’s Integrated Clrcult (IC)
and Printed Circuit (PCD) divisions all report
they are constrained from selling externally
(even though end product divisions report
only a minor level of asset specificity related
to internal transfers from these component
divlsions). In these cases there are no reasons
associated with the need to protect the transac-
tion-specific investments of the end product
divisions for constraining these component
divisions from selling externally. Therefore,
other factors must account for these selling
constraints.
Our case-bycase investigation revealed that
each of these component divisions was estab-
lished Initially for the purpose of meeting inter-
nal needs. In this context, the observed
constraints ensure that each component divi-
sion continues to focus on serving internal cus-
tomers. Continental’s PCD division controller
expressed the common dilemma posed by
external sales for component divisions in verti-
cally integrated firms:
who do I serve first, the external customer or the mter-
nal customer? That is the biggest problem that we don’t
have an answer for. We’ve looked at selling externally
but we’re deterred by the external-internal conflict.
Consequently, it is reasonable to conclude
that the reported constraints in these three
cases are due to each company’s vertical inte-
gration strategy. It should also be noted that
each of these divisions has been able to oper-
ate at or near capacity based on internal
volume alone although this has required some
consolidation and downsizing in the two cir-
cuit board operations.
THEORY - INVESTIGATION 447
Strategic sourcing considerations
In each Iirm we observe that there are stra-
tegic motivations for the decisions to invest in
particular component technology and capacity.
Each iirm has invested in integrated circuits
(and in some cases related hybrid facilities) to
assure access to and control over technological
capability that provides differentiated features
in their end products. The comments of the
general manager of a major product division
at High-Tech help make this clear:
IC’s [integrated circuits] are our proprietary technol-
ogy. If we were forced to buy off the shelf ICs it would
compromise part of our competitive advantage: we’d be
giving up our leverage.
The manager in a product division at World-
Wide made a similar comment:
We look to the CEC division for ICs [integrated circuits]
horn a strategic viewpoint and PCB for circuit boards
from an overall production cost and reliability stand-
point. By strategic I am referring to proprietary
designs, processes and technology that translate into
performance and features.
In all four cases it appears that integrated
circuits and hybrid components are regarded
as being of great strategic importance.” Con-
sistent with our expectations, buying units are
generally constrained from externally sourcing
integrated circuit or hybrid components which
are available from internal suppliers. This con-
straint acts to protect the value of the Iirm’s
investment in strategically important compo-
nent facilities.
Circuit boards, on the other hand, are not
generally regarded as having the strategic
importance of integrated circuits. For exam-
ple, in contrast to the integrated circuits pro
duced in High-Tech’s CT division, the circuit
boards produced by the CB division are not
i8 Continental’s IC division is regarded as strategically important because of the critical functionality its components
provide to end products. However, in comparison with High-Tech, World-Wide and Southern’s integrated circuit facilities
the degree of strategic importance appears lower. We believe that this is due to its different supply role. Continental’s IC
division focuses on producing high-volume mature components which could be sourced externally. Another division
produces the low-volume newer technology circuits.
448 G. J. COLBERT and B. H. SPICER
regarded as providing proprietary technology
From our cases it appears that the strategic
to the end product divisions. Although the CB
value of the component is important to the way
division supports end product division strategy
through superior service, for example turn-
in which the internal transfer process and sour-
around time, reliability and quality, the
cing in particular ends up being managed. The
absence of proprietary technology and the
availability of circuit boards from external
following representative comment by a man-
sources means that CB is of lesser strategic
importance to the firm. Consistent with CB’s
ager in an end product division of Continental
role in High-Tech’s competitive strategy CB’s
internal customers are free to buy circuit
Communications captures the importance of
boards externally. Similarly, we observe in
our other cases that buying divisions are lar-
gely free to buy circuit boards from external
strategic considerations in sourcing decisions:
suppliers even when circuit board manufactur-
ing capacity exists internally.
The iirst criterion is always whether the component is
strategic or not. If the component is strategic there will
always be a heavy emphasis on making [or buying
inside]. Where the component is not regarded as strate-
gic and is low in volume then there is no constraint and
we can go to the lowest bidder.
Similar comments were made by the man-
agers we interviewed in our other cases. The
The second research question we address is:
association we found between the strategic
roles of different component facilities and sour-
How do the business units of high-tech-
cing policies is set out in Table 3 case by
nology companies set transfer prices for
case.” As noted earlier, the issues of strategic
importance and asset specificity are closely
bound together and are difficult, if not impos
particular component transfers and why
sible, to disentangle.
do the methods used differ?
Case evidence pertaining to this question is
set out by case in Table 3 and a cross-case
summary of our findings is given in Table 5.
TABLE 5. Cross-case data on transfer pricing
Assets specificity reported by component (se%ng) divisions
SigmjTicant
High-Tech
CT division
P&l division
World-Wide
CEC division
Continental Communications
IC division
Southern Defense and Aerospace
EC division
Minor
High-Tech
CB division
Continental Communications
PCD division
Moderate
World-Wide
PCB division
l Significant weight also given to market reference.
t On occasions may also give weight to market references.
Signiiicant weight given to manufacturing costs?
Yes
Yes
Yes
Yes*
Yes
No
No
Yes t
I9 The dangers of overlooking important strategic issues in making sourcing decisions generally is addressed by Venkatesan
(1992).
TRANSFER PRICING PROCESS THEORY - INVESTIGATION
449
Transfer pri ci ng when asset speci j ki ty i s
si gni fi cant i n the component (sel l i ng)
di vi si on
When component di vi si ons have si gni J ?cant
l evel s of i nvestment speci al i zed to i nternal
transfers, we expect to fi nd greater wei ght
gi ven to manufacturi ng costs i n the setti ng
of transfer pri ces, i .e. we woul d expect to
observe cost-based transfer pri chg.
From Panel A of Table 3 we see that High-
Tech’s Circuit Technology (CT) and Packaging
& Interconnect (P&I) divisions each report a
sign&ant level of asset specificity related to
internal transfers. As expected, both report
they use cost-based pricing. Each of these com-
ponent divisions sets its transfer price at full
manufacturing cost as determined by an activ-
ity-based costing system plus a mark-up to
cover non-manufacturing costs.
As can be seen from a review of the other
panels in Table 3, and as summarized in Table
5, this result is generally replicated for all the
other integrated circuit component divisions,
i.e. World-Wide’s Custom Electronic Circuit
(CEC) division, Continental’s Integrated Cir-
cuit (IC) division, and Southern’s Electronic
Circuit (EC) division. All of these divisions
report significant asset specificity and cost-
based transfer pricing. World-Wide’s CEC divi-
sion transfer prices are usually based on manu-
facturing cost plus a mark-up to cover non-
manufacturing costs. However, it also reports
that infrequently it will also consider external
quotations in transfer pricing negotiations in
order to keep particular component business.
This arises primarily with those components
that are more commonly available and involve
higher volumes, and where market prices are
more informative.
Continental’s Integrated Circuit (IC) division
reports giving significant weight to cost in set-
ting transfer prices. However, it also gives
weight to market referents in neogtiating trans-
fer prices. Internal customers frequently get
quotes from outside vendors which they then
use in negotiations with the IC division. In con-
trast to other IC facilities in this study, this
integrated circuit facility specializes in high-
volume circuit designs that are produced on
mature fabrication processes. This has an
important implication. Market prices are more
likely to be informative and useful in these
circumstances for evaluating the value of a
component and the efficiency of the compo-
nent division. The controller of an end product
division describes how transfer pricing works
with Continental’s IC division:
There arc essentially no constraints on how the transfer
price is set so long as the prime directive of not going
outside is complied with. Consequently, the transfer
price is set by negotiation. We’re always going to out-
side vendors to find out what they could do it for to use
as leverage in negotiations and the IC division tries to
factor their costs into what they try to charge us.
Continental’s IC division’s use of market refer-
ents is in response to the cost pressures placed
on end product divisions and the presence of
competitive external suppliers. Bringing mar-
ket referents into the negotiating process helps
to protect the firm’s transaction-specific invest-
ments since negotiated adjustments to the
transfer price may reduce the motivation to
source the component externally. Reference
to external quotations also places pressures
on the component division to operate effi-
ciently. As suggested above, the use of market
referents appears to be associated with the
mature processes being used to produce com-
ponents. It is important to note that we did not
observe their use where transferred integrated
circuit components were produced on state-of-
the-art processes.
Southern’s EC division also uses cost-based
transfer pricing, typically manufacturing cost
plus a fixed mark-up to recover non-manufac-
turing costs. However, because of its large
amount of government work, the ease of con-
forming to government reporting requirements
is also cited as an important reason for cost-
based pricing. Internal contracts are typically
written to mirror the terms of external prime
contracts (cost-plus or fixed cost). With fixed
cost type internal contracts the component
unit keeps or bears any cost underrun or over-
run. This is done to motivate efficient opera-
450 G. J. COLBERT and B. H. SPICER
tions in the component division and to give the
prime contractor division control over costs.
The internal buying unit’s negotiating concern
stems from the impact of the transfer price on
the total cost of a programme or project. Con-
Bitt over sourcing and transfer pricing is mod-
erately common. On occasions group managers
wilI get involved ln arbitrating a transfer pri-
cing dispute. Also, on occasions, corporate
managment will indicate that winning a speci-
fic contract is most important for the company,
in which case considerable pressure is placed
on the selling unit to reduce its transfer price.
In addition, frequently EC will sell at less than
its full cost to match competitive bids received
by internal customers and to support an end
product division’s efforts to bid successfully
on new programmes and contracts.
nental’s Printed Circuit (PCD) division. PCD’s
transfer pricing is primarily market-based using
a portfolio approach, i.e. circuit boards with
sirnllar characteristics have the same or similar
prices where the base price of portfolios is set
primarily on market quotes. The PCD division
controller explains:
We slice printed circuit boards into technologies or
layers. For example, on a double sided board all con-
tractors who supply Continental with that product
(including us as the inside contractor) arc contracted
at the same price per panel. But there are kickers if it
has Rne lines or a thicker level or ls an odd size, etc. We
charge an extra percentage for these. We don’t know lf
the competition does this, but we think we are ln the
ballpark.
Transfer pri ci ng when asset spec@ci ty i s
mi nor i n the component (sel l i ng) di vi si on
When component di vi si ons have l i ttl e spe-
ci al i zed i nvestment rel ated to i nternal trans-
fers we expect to find market-based pri ci ng I n
use.
This is what we observe for transfers from
High-Tech’s Circuit Board (CB) division which,
as shown in Table 3, reports only a minor level
of asset specificity. The CB division has com-
plete autonomy to sell externally and to set
transfer prices. They use a market-based trans-
fer price which is determined by the same algo-
rithm used for external pricing. There is one
interesting exception to this management pol-
icy. In those infrequent instances where the
price determined by the external pricing algo-
rithm is lower than full manufacturing cost as
determined by CB’s activity-based costing sys
tern, CB will agree to sell Internally, but only at
cost. This represents a concession to internal
customers who, for convenience or other rea-
sons, still desire to purchase the part internally.
In both High-Tech’s CB division and Conti-
nental’s PCD division, the internal division is
most likely to be competitive on manufactur-
ing newer boards which involve more
advanced technology, are most difficult to man-
ufacture, and are lower in volume. As these
products mature these shops are often less
competitive than other external suppliers and
end product divisions have incentives to
reduce costs by exploring options to buy exter-
nally. In both cases end product divisions have
the freedom to do this. Whenever costs exceed
the market price, end product divisions are
encouraged by the upstream internal compo-
nent division to move these boards to external
suppliers who are generally cheaper. These are
typically the more mature, less complex
boards.20
The general finding is repeated with Conti-
Unlike other component divisions studied,
World-Wide’s Printed Circuit Board (PCB) divi-
sion reports a moderate level of asset specifl-
city for its largest Internal transfers and minor
otherwise. Consistent with this mid-way posi-
tion, PCB reports that, as a general policy, it
uses cost-based pricing where cost is manufac-
turing cost (as determined by an activity-based
cost system) plus a mark-up for non-manufac-
^^
‘” In Continental PCD’s case this has resulted ln reducing its capacity. The controller explains: “We are doing a number of
lower technology boards that we price at market. The market price is so low for these that we can’t afford to make them
anymore. It is internally cheaper for us to downsize the plant and just stay with the high technology boards.”
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 451
turlng costs. However, it may infrequently divisions and end product divisions. By
negotiate a different price where an external attempting to meet profit targets by lilling capa-
supplier has provided a lower price. Customer city, component divisions became increasingly
divisions regularly obtain external quotes to focused on external sales opportunities to the
compare with internal transfer prices. In negotia- disadvantage of internal customers and their
tions with end product divisions who have only speciafized investments.
minor constraints on their ability to buy exter-
End product divisions started to pursue
nally, the PCB division argues that its costs are
external sources of key components owing to
close to or consistent with market references.
what they regarded as unreasonable transfer
pricing. Managers quickly realized that this
was to the detriment of the component divi-
SUMMARY, CONCLUSIONS AND
RESEARCH DIRECTIONS
sions and the significant specialized invest-
ments that they had made in the internal
Overall eval uati on of the theory
transfers of components. Divisional and corpo-
High-Tech was the first case we studied and
rate managers began to look for a way out of
the confusion
analysed. Based on our observations which are
, muddle and conflict that
reported in Panel A of Table 3 and summarized
resulted. The consensus that finally emerged
in Tables 4 and 5, we conclude that the empiri-
was that the primary role of the component
cal pattern of findings for High-Tech are consis-
divisions was to service end product divisions
tent with the theoretical predictions. The
by supplying them proprietary h&h-technology
manner in which High-Tech manages the trans
components and service. To support this strat-
fer pricing process and the weight given to
egy, and ostensibly to protect the value of spe-
setting transfer prices are positively related to
cialized investments made by component
the reported degree of asset specificity asso-
(selling) and end product (buying) divisions,
ciated with the transfer. The way in which
the company constrained their selling and buy-
High-Tech manages the transfer process also
ing policies. In addition, a decision was made to
appears to be related to the strategic importance
evaluate component divisions as profit centres
of the component itself and the nature of the
with a break-even objective on internal sales.
component division’s production capabilities.
Transfer prices were to be cost-based, being
Interestingly, High-Tech has made changes
full manufacturing costs as determined by
in its sourcing, selling and transfer pricing poli-
newly installed activity-based costing systems
ties as it has struggled to manage its compo-
plus a mark-up for non-manufacturing costs.
nent operations effectively. These changes and
The CB division was the notable (and speci-
the resulting contrast in management between
fit) exception to the above change in policy.
the CT and P&I divisions on the one hand,
This component division and the end product
versus the CB division on the other, provide
divisions retained the right to sell or buy circuit
additional evidence in support of the theory.
boards externally free of constraint. CB
For a period High-Tech managed all three
remained on market-based pricing with full
component divisions as cost centres. It then
profit centre status on both internal and exter-
switched to managing them as full profit cen-
nal sales. We believe this organizational treat-
tres (on internal and external sales) using quasi-
ment reflects both CB and end product
market-based pricing. Component divisions
divisions’ low asset specificity as well as CB’s
and end product divisions were in principle
less important strategic role in the company.
free to sell and buy externally. Although this The findings for the other cases are similar to
policy helped to change component divisions’ those for High-Tech. From Tables 3,4 and 5 we
orientation and culture it also proved disrup- conclude that there is considerable replication
tive to the relationship between component of specific findings consistent with theory.
452 G. J. COLBERT and B. H. SPICER
Overall, our findings generally match the pat- on the level of asset specificity reported by
tern of relationships suggested by our theory. the selling divisions.
One interesting factor that seems to bear on
the sourcing and transfer pricing process in the
tirms we studied was the circumstances faced
by the end product divisions. In general, end
product divisions were under pressure to
improve their financial performance in the
face of increasingly intense competition in
their respective end product markets. In each
case decision making and accountability had
been decentralized to a considerable extent.
Division management had been left to run
their own businesses but were held accounta-
ble for meeting profit targets. Upper manage-
ment’s role, besides setting profit targets, was
reserved to setting the larger corporate vision
about what (and how) markets and technolo-
gies should be pursued. What can be seen from
the companies studied is that difficult organiza-
tional issues arise where firms are decentralized,
yet the achievement of current and future goals
rests on some interdependent strategy between
units which have made specialized investments
specific to internal transfers. The manner in
which firms we studied manage the transfer pro-
cess and set transfer prices appears to have been
systematically influenced by the presence or
absence of these specialized investments.
For all four cases the strategic importance of
the component or the production capabilities
of the internal component supplier are asso-
ciated with specific investments in component
operations. Each firm has invested in integrated
circuit (and in some cases related hybrid facil-
ities) to ensure access to and control over tech-
nology that provides differentiated features in
their end products. In contrast, circuit board
facilities appear to operate primarily to sup-
port strategy by providing superior service
such as flexible delivery and quick design turn-
around. For component divisions which are
regarded of higher strategic value, buying divi-
sions are constrained from sourcing externally.
Similarly, component divisions are generally
constrained from selling strategically impor-
tant technology externally.
Managi ng the transfer process
From theory we expectprms to put i nsti tu-
ti onal arrangements i nto pl ace to protect
thei r economi c i nterests i n i nvestments spe-
ci al i zed to i nternal transfer acti vi ty and the
val ue of the prm’s competi ti ve strategy.
As can be seen from Panels A and B of Table
4, the firms studied generally constrain subunit
managers in selling and sourcing decisions
when there are significant levels of asset speci-
ficity associated with internal transfers. Conver-
sely, the firms generally do not constrain
buying division’s sourcing decisions when
there are only minor amounts of asset specifi-
city reported by the component (selling) divi-
sions. These fmdings support the theory. Firms
in this study vary the constraints on the sour-
cing decisions of buying divisions depending
However, in three instances, component
divisions were constrained from selling exter-
nally even though the level of asset specificity
in the end product divisions was reported as
minor. As our theory suggests that there is no
need for selling constraints to protect transac-
tion-specific investments in the end product
divisions serviced by World-Wide’s Printed Cir-
cuit Board (PCB) division, and Continental’s
Integrated Circuit (IC) and Printed Circuit
(PCD) divisions, the observed constraints
must exist for other reasons. As noted above,
,in each of these cases it seems that the
intended strategy of the firm in setting up the
component division to serve internal needs and
a continued desire to retain an internal focus
was judged to be more important than the
potential benefits from selling externally. This
factor should be considered for inclusion in a
more expanded theory of constraints in inter-
nal transfer processes.
Transfer pri ci ng
From theory we expect that component
di vi si on transfer pricing pol i cy wi l l be rel ated
to the extent of asset spec@ci ty i nvol ved i n the
transfers. As asset speci j i ci ty of the component
di vi si on i ncreases, market pri ces wi l l become
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 453
i ncreasi ngl y l ess useful and rel evant for
i nformi ng the transfer pri ci ng process and
greater wei gbt wi l l be gi ven to manufactur-
i ng costs i n setti ng transfer pri ces.
As can be seen from Table 5, the findings
generally support the theory. Component divi-
sions with low levels of asset specificity give
market prices most weight. When component
divisions have high levels of asset specificity,
manufacturing costs are given most weight.
We did tind two instances where component
divisions with high levels of asset specificity
made use of market references to inform the set-
ting of transfer prices. The first was High-Tech
which had previously made an attempt to use
quasi-market-based transfer pricing by looking
at prices for very roughly comparable intermedi-
ate products available from external suppliers
and then attempting to make gross adjustments
to these prices for differences in design and func-
tionality. High-Tech’s managers abandoned this
approach because they became convinced that
the so-called “market-based” transfer prices for
components from the Circuit Technology (Cl‘)
and Packaging & Interconnect (P&I) divisions
were not useful and were causing serious deci-
sion-making errors and high levels of conflict
within the firm. By the time we conducted our
study these two component divisions had
abandoned market-based pricing in favour of
cost-based pricing based on activity-based manu-
facturing costs.
We also found one instance where a compo-
nent division (Continental’s Integrated Circuit
(IC) division), which reported a high level of
asset specificity, gave weight to both manufac-
turing costs and external market referents. This
division specializes in high-volume compo-
nents produced on mature fabrication pro-
cesses where market references are more
relevant. Reference to external quotations in
this instance served to protect the component
division’s specialized investment (and the
firm’s interests) by helping to retain large-
volume transfers within the firm.
In the mid-range of asset specificity reported
by World-Wide’s Printed Circuit Board (PCB)
division, weight was also given to market refer-
ents in addition to manufactuimg costs, on
occasions.
Li mi tati ons and suggesti ons for future
research
As we noted in the Introduction, the transfer
pricing literature has both theoretical and
empirical problems. While there is an abun-
dance of normative papers there are far fewer
papers which make a strong theoretical contri-
bution towards explaining how transfer pricing
processes are actually managed. On the empiri-
cal level, most surveys of practice are of limited
value, firstly, because they tend to abstract
from industry and organizational context and,
secondly, because they are wrongly directed at
the level of the firm as a whole rather than at
the subunit level where internal transfers actu-
ally take place.
We attempted to address the theoretical
issue by first developing a theory of the trans-
fer pricing process from a transaction costs
perspective. To avoid the empirical difficulties
that have troubled surveys of transfer pricing
processes and practices and to contribute to an
understanding of transfer pricing processes in
their organizational contexts, we adopted a
multiplecase study approach with embedded
units, i.e. the divisions in each firm and compo-
nent transfers between them. However, like
other research studies, ours has its limitations.
The first has to do with making the asset spe-
cilicity construct operational. We use a subjec-
tive categorical evaluation of the impact of the
potential effect on investment values if internal
transfers were halted. However, we argue that
to the extent that subjective assessments direct
behaviour the subjective categorical evalua-
tions of asset specificity we use are likely to
be consistent with observed management of
the transfer process.*’ Moreover, we were
careful to query each person we interviewed to
” We asked a number of managers whether it was feasible for them to quantify objectively the actual dollar impact on
specific investments of the loss of an internal transfer. The negative responses we received lead us to believe that
dwelopments of such a direct, objective measure of asset specificity is not feasible.
454 G. J. COLBERT and B. H. SPICER
ensure that each question was understood and
was supported by specifics.
A second limitation is that although the rela-
tionship between asset specificity reported by
component (selling) divisions was an important
determinant of resulting transfer prices, in
some instances other firm-specific factors
appear to account for exceptions to policies
based on asset specificity considerations
alone. For example, we noted that transfer pri-
cing policies were influenced, in some cases,
by market conditions for the transferred com-
ponent and competitive pressures in end pro-
duct markets. We also observed circumstances
where managers chose to deviate from their
division’s transfer pricing policy to meet exter-
nal quotes in order to keep internal customers
and/or to accommodate their needs. In addi-
tion, organizational history, management style
and organizational culture appear to play a
moderating role. Ease of conformance to gov-
ernment contracting and reporting require-
ments also influenced a decision to use cost-
based transfer pricing in one case.
A third limitation relates to the inherent
drawbacks of the case study method. Like
other methods of research it has its own
strengths and weaknesses. A major criticism
of case study research relates to the issue of
generalization. Following Yin (1989) our pri-
mary objective was not to generalize from a
sample to a larger population. Bather, our
intent was to provide analytical evidence
about the empirical validity of theory by pat-
tern matching evidence back to theory. How-
ever, by using a multi-case research design we
believe there is some assurance that the empiri-
cal patterns we observed are not idiosyncratic
to a particular firm in the high-technology, elec-
tronics industry.
Clearly, additional case studies would be
helpful. These may involve further studies
amongst high-technology electronics firms
and with vertically integrated firms in other
industries. However, the problems of access,
cost and logistics in undertaking multiple case
research of this type should not be underesti-
mated. It may be more efficient now to use
survey methods to extend the empirical inves-
tigation to a broader cross-section in selected
industries. An extension of the existing survey
literature could focus on particular types of
intra-firm transfers where asset specificity and
the strategic importance of component opera-
tions are expected to differ.
The question of the management of the
transfer process and transfer pricing is an
important one for the managers and academics
alike. Our aim in this paper was to make a
contribution to understanding how and why
managers handle this problem in the complex
and uncertain product environments of high-
technology, electronics firms. Our multicase
investigation has provided empirical evidence
that asset specificity plays an important and
systematic role as predicted by our theory.
However, our empirical evidence revealed
that other firm-specific factors influence man-
agement of the process and the setting of trans
fer prices. Further rounds of theory building
and empirical investigation might incorporate
these additional factors.
BIBLIOGRAPHY
Abdel-khalik, A. & Lusk, E., Transfer Pricing-A Synthesis, Accounffng Revi ew (1974) pp. 8-23.
Alchian, A., Specificity, Specialization, and Coalitions, J ournal of I nstftutfonal and Theoreti cal Econom-
ics (19B4) pp. 34-49.
Alchian, A. & Woodward, S., Reflections on the Theory of the Firm, J ournal of I nsti tuti onal and
Tbeorettcal Economi cs (1987) pp. 110-136.
TRANSFER PRICING PROCESS THEORY - INVESTIGATION 455
Ball, R., The Firm as a Specialist Contracting Intermediary: Application to Accounting and Auditing,
Working paper, William E. Simon Graduate School of Business Administration, University of Rochester
(June 1989).
Borowskl, S., Environmental and Organizational Factors Effecting Transfer Pricing, J ournal ofMunuge_
ment Accounttng Research (Fall 1990) pp. 78-99.
Brims, W. Jr & Kaplan, R., Introduction: Field Studies in Management Accounting, in Bruns, W. & Kaplan,
R. (cds), Accounting and Management: Field Study Perspectives (Boston: Harvard Business School
Press, 1987).
Coax, R., The Nature of the Firm, Economica (1937) pp. 386-405.
Colbcrt, G. J., An Empirical Investigation of the Transfer Process and Transfer Pricing: A Multicase
Research Design, Doctoral dissertation, University of Oregon (December 1991).
Ecclcs, R. G., The Transfer Pricing Problem (Lexington: Lexington Books, 1985).
Eiscnhardt, K., Building Theories from Case Study Research, Academy of Management Review (1989)
pp. 532-550.
Grabski, S., Transfer pricing in Complex Organizations: a Review and Integration of Recent Empirical and
Analytical Research, J ournal of Accounffng Ltterature (Spring 1985) pp. 33-71.
Hertcnstein, J., Management Control System Change: the Adoption of Inflation Accounting, in Bruns, W.
% Kaplan, R. (eds), Accountfng and Management: Field Study Perspectives (Boston: Harvard Business
School Press, 1967).
Jones, G. % Hill, C., Transaction Cost Analysis of Strategy-Structure Choice, Strafegfc Management
J ournal (1988) pp. 159-172.
Klein, B., Vertical Integration as Organizational Ownership: the Fisher Body-General Motors Relationship
Revisited, J ournal of Luw, Economtcs and Organization (1988) pp. 199-213.
Levy, D., The Transactions Cost Approach to Vertical Integration: an Empirical Examination, Review of
Economfcs and Statfstfcs (1985) pp. 438-445.
McKinnon, J., Reliability and Validity in Field Research: Some Strategies and Tactics, Accountfng, Au&-
fng and Accountability (1988) pp. 34-54.
Maher, M., The Use of Relative Performance Evaluation in Organizations, in Bruns, W. % Kaplan, R. (eds),
Accountfng and Management: Field Study Perspectfves (Boston: Harvard Business School Press,
1987).
Mastcn, S., The Organization of Production: Evidence from the Aerospace Industry, J ournal of Law and
Economics (1984) pp. 403-418.
Mastcn, S., Mcchan, J. Jr % Snyder, E., Vertical Integration in the U.S. Auto Industry, J ournal of Economic
Behavior and Organization (1989) pp. 265-273.
Merchant, K., How and why Firms Disregard the Controllability Principle, in Bruns, W. & Kaplan, R.
(cds), Accounting and Management: Field Study Perspectives (Boston: Harvard Business School Press,
1987).
Merchant, K. & Manzoni, J. F., Achievability of Budget Targets in Profit Centers: a Field Study, The
Accounting Review (1989) pp. 539-558.
Montcvcrdc, K. & Tccce, D. J., Supplier Switching Costs and Vertical Integration in the Automobile
Industry, BeU J ournal of Economics (1982) pp. 206-13.
Palepu, K., The Anatomy of an Accounting Change, in Bruns, W. & Kaplan, R. (cds), Accounting and
Management: Field Study Perspectives (Boston: Harvard Business School Press, 1987).
Porter, M., Compefftfve Strutegy (New York: Free Press, 1980).
Porter, M., Competitfue Advantage Wcw York: Free Press, 1985).
Rcvc, T., The Finn as a Nexus of Internal and External Contracts, in Aoki, M., Gustafsson, B. % William-
son, 0. (cds), The Firm us u Nexus of Tmulies, pp. 133-161 (Newberry Park: Sage, 1990).
Rio&n, M. L Williamson, O., Asset Specificity and Economic Organization, International J ournal of
I ndustrial Organization (1985) pp. 365-378.
Spiccr, B. H., Towards an Organizational Theory of the Transfer Pricing Process, Accounting, Orgunfzu-
tfons and Society (1988) pp. 302-322.
Spiccr, B. H. % Ballcw, V., Management Accounting Systems and the Economics of Internal Organization,
Accounting, Organfzations and Society (1983) pp. 73-98.
Tang, R., Transfer Pricing Practices I n the United States and J apan (New York: Pragcr, 1979).
Tang, R., Canadian Transfer Pricing Practices, c4 Muguzine (March 1980) pp. 32-38.
Tccce, D., Firm Boundaries, Technological Innovatlon, and Strategic Management, in Thomas, G. (cd.),
The Economics of Strategfc Planning, pp. 187-199. (Lexington: Heath and Co., 1986).
456 G. J. COLBERT and B. H. SPICER
VanciI, R. F., Decentralization: Manageri al Ambi gui ty by Desi gn (Homewood: Dow Jones Irwin, 1979).
Venkatesan, R., Strategic Sourcing: to Make or Not to Make, Harvard Busi ness Reufew (November-
December 1992) pp. 233-261.
WaIker, G., Strategic Sourcing, Vertical Integration, and Transaction Costs, Interfaces (May-June 1988)
pp. 62-73.
WaIker, G. & Weber, D., A Transaction Cost Approach to Make or Buy Decisions, Admi nfstrathe Sci ence
Quarterl y (September 1984) pp. 373-391.
Watson, D. & BaumIer, J., Transfer Pricing: a Behavioral Context, The Accounti ng Ret&w (1975) pp.
466-474.
Wiiamson, 0. E., Markefs and Hi erarchi es (New York Free Press,1975).
Williamson, 0. E., Transaction Cost Economics: the Governance of ContractuaI Relations,JournaC oftaw
and Economi cs (October 1979) pp. 233-261.
Williamson, 0. E., The Economi c I nsti tuti ons of Capi tal i sm (New York: Free Press, 1985).
Williamson, 0. E., Strategizing, Economizing, and Economic Organization, SErategfc ManagementJ our-
naI (l 991) pp. 75-94.
Yin, R. K., Case Study Research: Desi gn and Methods (Newbury Park: Sage, 1989).
APPENDIX
The question used to gather information on the key construct of overall asset specificity associated with an internal
transfer for the internal seiling unit is shown below. Similar questions were fust asked about individuai types of asset
specificity, i.e. physical asset specificity such as plant, equipment, software, moulds, tooling and dies; human asset
specificity such as training and specialized know-how; inteilectual asset specificity such as design work, R&D efforts
and patents; and site specificity.
Each question about asset specificity had two parts. The iirt part was designed to set the stage for the second part
which gauges the degree of asset specificity. The fnst is designed to get the interviewee to thhtk about the concept. The
second question is then more directly focused on the extent to which an attempted redeployment of the transaction-
specific investment would result in loss of value.
To facilitate our written analysis in the text we treated the responses No Decline and Minor Decline as equivalent and
Significant or Very Signiftcant Decline as equivalent.
Overall asset speci fci ty
(a) How would you rate the OVERALL extent to which your unit has made investments SOLELY to manticture the
selected product (i.e. assets that could not be easily transferred to other operations or products)?
Very
No Minor Moderate SigniIicant signiEcant
Investment Investment Investment Investment Investment
1 2 3 4 5
(b) OVERALL, what decline in the investment value (owing to increased costs or reductions in revenue) would be
involved in an attempt to transfer or redirect these investments to other operations or products?
No
Decline in
Investment
Value
1
Minor
Decline in
Investment
Value
2
Moderate
Decline in
Investment
Value
3
significant
Decline in
Investment
Value
4
Very
Signiftcant
Decline in
Investment
Value
5
doc_944646301.pdf